6
Infrastructure and Security

INTRODUCTION

The energy system depends on a massive infrastructure to produce and distribute energy to households and businesses. Table 6-1 gives a rough idea of the amount of fixed assets related to energy production. The infrastructure related to energy production amounted to nearly $2.9 trillion in 2007, or 12% of the value of the net stock of nonresidential fixed assets in that year.1 The bulk of energy-related assets are structures—electricity-generation facilities and mining exploration, shafts, and wells.

In this chapter, the committee considers a variety of externalities that are associated with the energy infrastructure. In particular, we consider disruption externalities in the electricity-transmission grid, the vulnerability of energy facilities to accidents and possible attack, the external costs of oil consumption, supply security considerations, and national security externalities. Where possible, we quantify the externalities that we identify.

DISRUPTION EXTERNALITIES IN THE ELECTRICITY-TRANSMISSION GRID

In the interconnected electric-power system, “reliability” is the degree to which the system delivers power to consumers within accepted standards

1

This estimate does not include energy capital in the U.S. military, nor does it include the value of transportation assets or computers and other equipment used in the production and distribution of energy. Adding transportation-related equipment and structures alone would add $1.3 trillion to the value of energy-related fixed assets.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 309
6 Infrastructure and Security INTRODUCTION The energy system depends on a massive infrastructure to produce and distribute energy to households and businesses. Table 6-1 gives a rough idea of the amount of fixed assets related to energy production. The infra- structure related to energy production amounted to nearly $2.9 trillion in 2007, or 12% of the value of the net stock of nonresidential fixed assets in that year.1 The bulk of energy-related assets are structures—electricity- generation facilities and mining exploration, shafts, and wells. In this chapter, the committee considers a variety of externalities that are associated with the energy infrastructure. In particular, we consider disruption externalities in the electricity-transmission grid, the vulnerability of energy facilities to accidents and possible attack, the external costs of oil consumption, supply security considerations, and national security exter- nalities. Where possible, we quantify the externalities that we identify. DISRUPTION EXTERNALITIES IN THE ELECTRICITY-TRANSMISSION GRID In the interconnected electric-power system, “reliability” is the degree to which the system delivers power to consumers within accepted standards 1 This estimate does not include energy capital in the U.S. military, nor does it include the value of transportation assets or computers and other equipment used in the production and distribution of energy. Adding transportation-related equipment and structures alone would add $1.3 trillion to the value of energy-related fixed assets. 30

OCR for page 309
310 HIDDEN COSTS OF ENERGY TABLE 6-1 Net Stock of Energy-Related Fixed Assets in 2007 ($Billions) Priate Fixed Assets Equipment and Software 523.9 Engines and turbines 83.5 Electrical transmission, distribution, and industrial apparatus 358.4 Mining and oilfield machinery 49.5 Electrical equipment, not elsewhere classified 32.5 Structures 2,120.4 Power 1,230.6 Mining exploration, shafts, and wells 889.8 Goernment Fixed Assets 241. Power 241.5 TOTAL 2,885.8 SOURCE: BEA 2009. and in the amount desired (Abel 2006). Typically, reliability is good—the nation’s electric-power grid delivers power when needed and within an ac- ceptable quality range. Occasionally, however, electric outages occur when the demand for electricity exceeds the supply. There are various causes of outages, includ- ing equipment failure, extreme weather events, such as ice storms and hur- ricanes; trees or animals physically damaging parts of the electric system; accidents that damage parts of the system; equipment failure; and operator error. Outages solely from overloads are rare. However, other things being equal, a greater load increases the likelihood of transmission congestion and of decreased reliability. Consequently, there are externalities associated with the consumption of electricity in the sense that when an electricity consumer draws from the grid, this increases the probability that demand will exceed supply and that an outage will occur. In addition to outages or interruptions in electricity service, voltage sags, harmonic distortions, and other power-quality events occur. Although each event generally causes little damage, except for customers whose com- mercial activities depend on very-high-quality power, these events occur much more frequently and result in significant annual damage. The possible externality that an individual consumer imposes when using additional electricity is the expected damage to all other users of the grid from an outage or power-quality event (the damages of an event weighted by the increased risk of an outage from the marginal consumer’s use). The optimal price to internalize this externality would include this marginal damage (unless the costs of implementing such a pricing scheme exceed the benefits). This externality has long been recognized, and as we discuss later, various means of internalizing this externality are in place to varying degrees.

OCR for page 309
311 INFRASTRUCTURE AND SECURITY To the extent transmission externalities exist, they apply to all of the electricity-generation options, such as coal, natural gas, oil, wind, and hy- dropower. However, intermittency in generation has the potential to affect the frequency of outage events, as well as power quality. The Magnitude of the Electricity-Disruption Externality To calculate the magnitude of this externality, one needs an estimate of the damages from outages and power-quality events and an estimate of the increased probability of these events occurring because of the consumption of additional electricity. The nature and severity of the impacts of an out- age or power-quality disturbance vary and depend on the affected sector (manufacturing, commercial, or residential) and on the specific functions affected, as well as the availability of backup power, the duration of the outage, the time of the year, the time of day, the geographic region, and the extent to which customers are notified prior to the outage. A number of empirical studies estimating the damages from outages and power-quality disturbances have been undertaken. Some are based on estimates of lost output and damage from actual outages; others value the prevention of an outage. Previous estimates are typically of the total an- nual cost, cost per kilowatt hour, value of lost load, or damage divided by total kilowatt hour (Primen 2001; Overdomain 2002; Lawton et al. 2003; LaCommare and Eto 2004; Layton and Moeltner 2005; van der Welle and van der Zwaan 2007; Mount et al. 2008). None of these estimates measures marginal damages per se. Table 6-2 provides estimates of average damages for different sectors. In viewing the estimates in Table 6-2, it is important to note that SAIFI and SAIDI2 estimates were unavailable for the different sectors, so the same overall averages were used for the residential, commercial, and industrial sectors. Anecdotal evidence strongly indicates, however, that commercial and industrial establishments are likely to have backup power and thus less frequent loss of power (lower SAIFI values). Thus, the estimates in Table 6-2 of the average damage per kWh for the commercial and industrial sectors have probably been significantly over-estimated. The most striking observation from Table 6-2 is that smaller commer- cial businesses are most vulnerable to outages and power-quality distur- bances, especially the latter. Assuming for the purpose of discussion that the estimates of average damage per kWh consumed are over-estimates of these damages in the commercial and industrial sectors by, say, an order of 2 SAIFI is the System Average Interruption Frequency Index—the average number of inter- ruptions a customer experiences in a year. SAIDI is the System Average Interruption Duration Index—the average cumulative duration of outages a customer experiences in a year.

OCR for page 309
312 TABLE 6-2 Estimates of the Average Cost of Outagesa,b Source Estimate of Cost of Outages per kWh Method Used for Estimate Lawton et Aerages over all U.S. regions (costs vary by region) for Meta-analysis—Reviewed previous studies over the past 15 years al. 2003 studies done in 2001 and 2002 ending in 2002. Studies were of outage costs conducted by seven electric utilities in 20 studies. Authors also provided ranges, regional Large commercial and industrial—Outages breakdown, and duration. All durations: $0.0038 1-h duration: $0.0029 Small-medium commercial and industrial—Outages All durations: $0.015 1-h duration: $0.0083

OCR for page 309
This Average Calculated based on estimates of costs per event and on ranges of the committee damage per Average System Average Interruption Frequency Index, the System Average Type of year per damage in $ Interruption Duration Index, and the Momentary Average Interruption Sector occurrence customer per kWh Frequency Index in LaCommare and Eto (2004), combined with the Energy Information Administration data on annual sales to each sector Residential Power $7.90-$40.00 $0.00072- (in MWh). quality $0.0037 Estimates do not account for anecdotal evidence that larger commercial Outages $2.80-$6.40 $0.00025- and industrial establishments have backup power and thus experience $0.00072 fewer outages than smaller establishments and residential consumers. Thus, the estimates of average cost per kWh are significantly over- Commercial Power $2,800-$9,100 $0.038- estimated for the larger establishments. quality $0.13 Outages $1,200-$2,300 $0.016- $0.03 Industrial Power $9,400-$25,000 $0.0069- quality $0.018 Outages $4,800-$8,500 $0.0036- $0.0062 NOTE: Estimates for the residential sector have probably been under-estimated, and for the commercial and industrial sectors, they have probably been oer-estimated. a2007 U.S. dollars per kilowatt hour (USD/kWh). bThe top part of this table provides a summary of a meta-analysis on outage costs based on studies done between 1987 and 2002. Outage costs are reported by utilities taking part in the studies. Costs are reported in USD/kWh consumed. The bottom part of the table reports estimates for power-quality events and outages for different customer classes. The right-hand column explains how the numbers in the table are obtained (top half) or constructed (bottom half). 313

OCR for page 309
314 HIDDEN COSTS OF ENERGY magnitude, they are nevertheless sizeable compared with the average U.S. retail commercial electricity price of $0.094/kWh in January 2008 (EIA 2009k, Table 5.6a). Although an outage, if it were to occur, would cause greater damage to industrial firms, there are far fewer of them compared with commercial establishments, and industrial firms use much more electric power than commercial establishments. The upshot of these differences is that aver- age damage per kWh consumed in the industrial sector is almost an order of magnitude less than that in the commercial sector. Average damage per kWh consumed in the residential sector is, in turn, about an order of magnitude less than that in the industrial sector. As previously noted, these numbers measure average damages per kWh consumed rather than mar- ginal damages. Not all damages are congestion-related network-disruption damages. It is not possible to disentangle the marginal effect of network congestion’s contribution to outages, apart from other factors. Thus, the numbers in Table 6-1 are upper bounds on (and are probably significantly higher than) the marginal damages from outages and power-quality events due to congestion. Differences in the Effects of Alternative Electricity-Generation Technology and Fuel Options on Grid Reliability The effects of renewable energy sources—especially wind and solar—on the reliability of the electric-power grid might be different from those of conventional sources. The timing and duration of wind and sunshine can- not be controlled. Because wind velocities and sun intensity determine power output, it is variable and not entirely predictable. Electricity gener- ally cannot be stored, and transmission is costly. Although reserves and transmission must be provided with any type of electricity-generation tech- nology, the issues are more prominent with some of the renewable energy resources. There are concerns that, if interconnected to the grid, wind and so- lar facilities might reduce grid reliability or power quality. More backup sources and power-quality control devices, as well as additional wind or solar capacity to account for their lower capacity factors, might be needed. Transmission lines would be needed to carry power from more remote areas, where some of these facilities would be located, to where power is needed. These measures are costly but could be internalized in market transactions.3 3 There has been an interesting debate about how great these costs would be (for example, Jacobson and Masters 2001; DeCarolis and Keith 2001).

OCR for page 309
31 INFRASTRUCTURE AND SECURITY Extent to Which Grid Externalities Are Internalized The degree to which externalities are internalized is difficult to de- termine, as there are several ways in which internalization could occur. First, distinguish between the local distribution network and the regional transmission grids. For the former, some of the externalities are probably internalized through pricing by the local distribution companies. Electric- ity pricing can take many forms (see Borenstein [2005] for a description of different forms of peak-load pricing). For example, it can be implemented through differing block prices for peak and off-peak periods. If these peri- ods and tariffs are pre-established, then this is a form of time-of-use pricing and is common for commercial and industrial customers. Dynamic pricing, which itself has several forms, varies the price dynamically, depending on the load at different times of the day each day. As of 2004, over 70 utili- ties have experimented with dynamic pricing systems (see Barbose et al. 2004). For the bulk-level transmission grid, the rule making, standards, and regulations set forth by the Federal Energy Regulatory Commission (FERC) and the North American Electric Reliability Corporation (NERC) help increase reliability and act to internalize some of the externalities. Utilities have long been required to have operating reserve margins. More recently, the Energy Policy Act of 2005 empowered FERC to enforce mandatory reli- ability standards, set by NERC, which apply to all participants in the bulk power system. In 2007, monetary penalties of up to $1 million per day were established for noncompliance with these standards. Performance-based incentives that allow increased rates of returns on transmission projects intended to reduce transmission congestion costs or increase reliability are another recent development. Regulators may also use performance-based rates that reward utilities for good reliability and penalize them for poor reliability. Electricity markets are still in their infancy in providing incentives for socially efficient investment in grid infrastructure, especially transmission (Hogan 2008). Transmission-system operators responsible for providing transmission services have an incentive to reduce their private (as opposed to social) costs.4 The transition toward more competitive power markets has resulted in declining reserve capacity because producers are striving to minimize their costs; this situation reflects the lack of adequate incentives to improve power reliability.5 4 Social costs exceed private costs to the extent that disruption affects customers in other regions, and out-of-region impacts are not taken into account by system operators. 5 Reserve capacity is the amount of generating capacity in excess of capacity required for electricity needs at any time. At periods of peak usage, reserve capacity is low. Lowering re- serve capacity reduces costs for utilities but raises the risks of outages.

OCR for page 309
316 HIDDEN COSTS OF ENERGY Reserve capacity also has public-good attributes. For technical and eco- nomic reasons, it is not possible to prevent customers from benefiting from it, even when they are not paying for the power delivered by that reserve capacity (van der Welle and van der Zwaan 2007). Reserve requirements, the possibility of fines for noncompliance with NERC reliability standards, and performance-based incentives internalize some of the externalities. However, the extent to which these standards, associated fines, and incentives ensure a certain level of reliability—as measured by frequency and duration of outages—is uncertain. Compli- ance with many of the standards relies on self-certification, and penalties levied by FERC may be negotiated, possibly with economically inefficient agreements. Many of the problems with the transmission grid are due to the age of the infrastructure.6 Investing in a modern, “smart grid” will alleviate many of the problems that were noted above. New capital investment also will probably reduce risks associated with intermittent renewable electricity sources. This is an example where technological innovation can reduce or eliminate externalities. FACILITY VULNERABILITY TO ACCIDENTS AND ATTACKS The United States has over 1.5 million miles of oil and gas pipelines, 104 operating nuclear plants, 9 liquefied natural gas (LNG) import facili- ties, 100 LNG peaking facilities, and over 17,000 non-nuclear electricity generators in the United States. The domestic infrastructure to deliver en- ergy is complex and critical to the smooth functioning of the economy. The committee focuses in this section on the following questions: 1. To what extent is the U.S. energy infrastructure vulnerable to ac- cidents and terrorist attacks? 2. To what extent should infrastructure vulnerability be regarded as an externality? The following subsections explore these questions in the context of four infrastructure areas. Liquefied natural gas is a growing source of natural gas in the United States. LNG facilities are large and complex and have been the source of some community concern over safety. Oil transportation and stor- age are vulnerable to accidents and spills. The oil and gas pipeline network is extensive and critical to the flow of gas and oil around the United States. Finally nuclear power accidents continue to be of concern to the public. 6 See the discussion in America’s Energy Future, Chapter 9.

OCR for page 309
31 INFRASTRUCTURE AND SECURITY Analyzing these forms of energy infrastructure sheds light on the nature of externalities associated with our production and consumption of energy. LNG Infrastructure and Hazards The infrastructure associated with LNG distribution in the United States includes tankers used to transport the gas from foreign ports, import terminals which are dedicated to LNG, and inland storage facilities. There are nine operating terminals within the United States (Parfomak 2008). The most significant hazards associated with LNG infrastructure include a pool fire, which occurs if there is a spill that is ignited, or flammable vapor clouds, which occur if there is a spill that is not immediately ignited. In this case, the evaporating natural gas can travel some distance and during that time is at risk of ignition. Other safety hazards include LNG spilled on water that may be able to regasify and cold LNG that can injure people and damage physical structures. Because LNG dissipates with no residue, the only environmental damages that might occur would be linked to fire or cold damage. Finally, terrorism hazards that are directed at ships carrying LNG and at land facilities are also a risk. These damages can be associated with damage to a tanker, terminal, or inland storage facility. The most dangerous possibilities include a spill on water because such a spill could spread the farthest and, if ignited, seriously harm people and property at some distance. The safety record of tankers carrying LNG is quite good; since international shipping started in 1959, no spills have occurred, although a few groundings and collisions have oc- curred (Parfomak 2008). Tankers are double-hulled and engage GPS, radar, and other safety systems to reduce the risk of accidents and grounding. In contrast to tanker safety, a number of incidents at terminals and onshore storage facilities have occurred. Worldwide, there are over 40 ter- minals and 150 onshore storage facilities, and there are 13 known accidents at those sites reported since 1944. These data are summarized in Table 6-3. No LNG tankers fly under the U.S. flag, so only a worldwide number is reported for number of tankers. The number of serious accidents has been quite small, only 13 since 1959, with 29 fatalities and 74 injuries. With such a small number of incidents, it would be difficult to extrapolate from the historical data to estimate expected damages in the future. On average, over 260,000 mil- lion cubic feet of LNG has been imported or exported by the United States annually since 1985 (EIA 2009l). In this context, the apparent risk of ac- cidental injury, death, or property damage is small. A number of studies have quantitatively evaluated the hazards and risk of accidents or terrorist attacks, largely on a facility by facility basis. Parfomak and Vann (2008) summarized recent and most cited studies in

OCR for page 309
318 HIDDEN COSTS OF ENERGY TABLE 6-3 LNG Infrastructure and Safety Recorda Serious Infrastructure Worldwide U.S. Expected Incidents Type Numbers Numbers Growth Since 1959 Fatalities Injuries Tanker Ships 200 Not 200 more 0 0 0 relevant worldwide by 2013 Terminalsd 21b 13c 40 10 29 74 Storage >150 103 Not Facilities available aThe data in this table are taken from Parfomak and Vann (2008). bSix terminals have been approved and are under construction in the United States; another 15 have been approved, but construction has not yet begun. cData on accidents, fatalities, and injuries are not separately reported for terminals and storage facilities. These numbers represent the worldwide sum since records are available. dThese numbers reflect import and export terminals. their Table 6-2. A report prepared for the U.S. Department of Energy by Sandia National Laboratories summarizes and assesses four of those stud- ies to develop estimates of the damages of a large pool spill from a tanker over water. The report does not monetize damages but provides estimates of numerous end points of damage (for example, asphyxiation, cryogenic burns, and structural damage). Based on its review, the Sandia report provides estimates of the effects of various small- and large-scale accidents and intentional damage. For example, the report provides estimates of the size of a pool fire from dif- ferent sizes of holes of a tanker breach, the size of the pool, the distance of thermal hazards, and the burn time. This information could be combined with data on population density and on monetary damages of death, injury, and property loss to quantify this externality at least for specific locations. However, the committee did not attempt such quantification. Further, the small risk of these incidents suggested by the historical record implies that the magnitude of this externality relative to other energy-based externalities is likely to be small. The LNG industry faces unlimited liability for damages from accidents. In fact, fines have been large in the case of a pipeline fire in Bellingham, Washington, in 1999 and in training violations at the Everett, Massachu- setts, LNG terminal. These facts suggest that the risk of accidents and spills is internalized in the LNG industry. In addition, private insurance held by facility owners contributes to internalization of the externality. Finally, LNG tankers and facilities face considerable regulatory over- sight. The Coast Guard has responsibility (and bears the cost) for shipping

OCR for page 309
31 INFRASTRUCTURE AND SECURITY and terminal security, the Office of Pipeline Safety and the Transportation Security Administration have security authority for LNG storage plants and terminals, and the Federal Energy Regulatory Commission approves facility siting in conjunction with the other three agencies. The Coast Guard estimated that it incurs about $62,000 to accom- pany a tanker through Boston harbor to the Everett facility. This figure, combined with estimates of policy, fire, and security costs incurred by the cities of Boston and Chelsea and the state of Massachusetts, suggests that these “shepherding” costs run to about $100,000 per tanker. Although public costs of providing safe passage to tankers or of assessing health and safety standards are not externalities (they represent the costs of mitigating the externalities), these costs will not be represented in the market price of energy unless there are requirements that industry pay some of the costs. This provides another example of a situation in which private costs do not reflect social costs in ways that are unrelated to externalities. Here, the divergence arises from the government provision of services that are not priced in the fuel. Oil Spills from Ships and Facilities Oil is transported by tankers, barges, and other vessels where oil can be accidentally spilled during its transfer between vessels or where an acci- dent can occur on a vessel itself. Oil is stored in a variety of facilities where spills are possible. A number of highly publicized oil spills have occurred in the previous two decades that have increased the public’s awareness of the ecological harm and other damages that such spills can cause. Seabirds, marine mammals, a variety of reptiles and amphibians, fish, and inverte- brates are all at risk of death or injury when they come into contact with oil. In addition, many coastal and wetland habitats can be significantly altered by the presence of large quantities of spilled oil, and oil spills del- eteriously affect human use of the environment when recreational sites are damaged and when commercial activities are harmed (such as fishing or shrimping). Important cultural sites can also be damaged by these spills (Ramseur 2008). Data on the prevalence and size of oil spills that occur on U.S. soil are collected and maintained by the U.S. Coast Guard. Table 6-4 reports the average number of spills in the United States between 1990 and 1998 as well as the average amount of oil spilled (in gallons).7 A number of studies of the damages caused by oil spills have been com- 7 More detailed information on the source and type of spills can be obtained in Table 2-2 from Oil in the Sea III: Inputs, Fates, and Effects (NRC 2003c). The report also contains detailed estimates of spills worldwide.

OCR for page 309
326 HIDDEN COSTS OF ENERGY TABLE 6-7 U.S. Oil Dependence 1990 2000 2007 2030 Net oil imports as percentage of total U.S. Supply 42.2 52.9 58.2 55.5 World oil price (2007 $/BBL) 38 35 72 60 World crude production (million BBD) 65.5 74.9 81.5 102.9 OPEC share (percentage) 38.3 42.9 43.2 46.4 U.S. petroleum consumption (million BBD) 17 19.7 20.7 22.8 U.S. share of world production (percentage) 26.0 26.3 25.4 22.2 Oil intensity (1,000 Btu/GDP) $2000 4.7 3.9 3.4 2.2 Oil intensity (value of oil as a percentage of GDP) 2.6 2.0 3.6 1.9 ABBREVIATIONS: BBL = billion barrels, BBD = barrels per day, Btu = British thermal units; GDP = gross domestic product. SOURCES: BP 2008; EIA 2008a,b,g. economy continues to decline. Oil intensity (measured as 1,000 British thermal units of oil consumption per dollar of gross domestic product [GDP]) has fallen by over one-quarter since 1990 and is projected to fall an additional one-third by 2030. Rising oil prices, however, offset the declining physical intensity; thus, the value of oil consumption in GDP is projected to remain at about 2% (although down from its anomalous level of 3.6% in 2007 and the sharp run-up in prices that year). The importance of oil in the U.S. economy has given rise to a large amount of literature measuring the external costs of oil consumption.13 Parry and Darmstadter (2003), for example, attempt to quantify the mar- ginal external cost of petroleum consumption, defined as “the difference between the costs to the U.S. economy as a whole and that to individuals or firms from additional oil consumption. Marginal external costs, expressed in $/BBL [barrels], are referred to as the oil premium” (p. 11). In this section, we consider three questions: • What is the oil premium? • Is the oil premium an externality? • How does the oil premium relate to the optimal tax on oil con- sumption or imports? What Is the Oil Premium? As the quotation from Parry and Darmstadter (2003) above indicates, the oil premium is a measure of the difference between the private and so- 13 See, for example, Bohi and Toman (1993, 1995); Greene and Leiby (2006); Leiby (2007); Greene (2009).

OCR for page 309
32 INFRASTRUCTURE AND SECURITY cial costs of petroleum consumption measured in dollars per barrel. The lit- erature identifies two major quantifiable sources of the discrepancy between private and social costs—U.S. monopsony power and economic disruptions arising from unanticipated price shocks. We discuss these in turn. The literature on monopsony power takes as its point of departure the observation that the United States is a large consumer of oil. As such, any policy to reduce domestic oil demand reduces the world oil price and benefits the United States through lower prices on the remaining oil it im- ports. The oil premium arising from monopsony power reflects the lack of recognition by individual consumers to the buying power that the nation has if it acts in a coordinated fashion. Figure 6-1 illustrates the idea of monopsony power. On the basis of individual demands for oil, aggregate demand is given by the downward sloping curve marked D. With an upward sloping supply curve for oil, the market equilibrium occurs at point e where Q0 barrels of oil are consumed at price per barrel P0. If the U.S. government takes some action to reduce oil demand from D to D′, the world oil price falls from P0 to P1. The gain to consumers from the fall in oil price is the rectangle P0P1fg. This is offset P h S g P0 e f P1 D D' Q1 Q0 Q FIGURE 6-1 Illustration of monopsony. D = aggregate demand; e = market equi- librium Qo consumption and P0 price; f = market equilibrium for Q1 consumption and P1 price; P= price per barrel of oil; Q = barrels of oil consumed. Figure 6-1 R01631 editable vectors

OCR for page 309
328 HIDDEN COSTS OF ENERGY by losses to producers in oil revenue (equal to the same rectangle).14 If all supply comes from domestic production, there is no gain to the United States. The gain to producers is exactly offset by the loss to U.S. producers. If all supply comes from non-U.S. producers, the gain to U.S. consumers is financed by a transfer from other oil-producing countries. This gain is the monopsony benefit identified in the literature. The marginal oil premium is the incremental income transfer to U.S. consumers from foreign producers from a small reduction in demand for oil arising from a U.S. policy. One policy that would give rise to this income transfer is a tax on oil consump- tion. In Figure 6-1, an excise tax of fh per barrel would shift demand from D to D’ and lower demand from Q0 to Q1. Because the tax (ignoring the efficiency costs of taxation for the moment) is simply a transfer within the United States, the result follows. The literature on oil consumption correctly notes that private markets in the United States do not account for the potential market power that U.S. consumers could wield in world oil markets. The literature also generally recognizes that any policy to take advantage of consumer purchasing power affects a transfer from foreign oil producing nations to the United States. Such a “beggar thy neighbor” policy has been justified on the grounds that OPEC is artificially inflating world oil prices at the expense of consum- ing nations and that the exercise of monopsony power is a countervailing policy. Disruption costs have also been identified as a cost that is not incor- porated into the price of oil. Leiby (2007) identifies two components of a disruption premium. First, a cost increase increases transfers of U.S. wealth from domestic consumers to foreign producers. The magnitude of this transfer depends on the price increase and the amount of oil imported into the United States. Second, cost increases induce shocks to the economy, resulting in losses in economic output, income, and jobs. Is the Oil Premium an Externality? The committee considers the two components of the oil premium in turn. With respect to the issue of monopsony power, it is unquestionable that domestic policy can reduce aggregate demand and lead to a reduc- tion in the world price of oil. Such a policy would generate a transfer in wealth from foreign oil producing nations to the United States.15 However, the ability to exercise monopsony power is not the same as an externality. 14 Thisignores efficiency losses for the moment. We return to this issue in a moment. 15 Thistransfer occurs even if foreign oil producing countries curtail production to stem the price reduction. Reductions in oil supply and demand from the countervailing policies will lead to a reduction in expenditures on imported oil in the United States.

OCR for page 309
32 INFRASTRUCTURE AND SECURITY Externalities create a market failure. Exercising monopsony power creates a market failure where one did not exist before. In Figure 6-1, the exercise of market power creates deadweight loss equal to the triangle efh.16 In fact, this market failure is designed purely to transfer income from another country to the United States. Bohi and Toman (1995) noted that the exercise of monopsony power is an example of a pecuniary externality designed to shift wealth from one nation to another.17 They pointed out that the existence of market power on the part of energy producers complicates the analysis slightly. For one thing, market power leads to the creation of rents that transfer wealth from energy-consuming to energy-producing countries. The exercise of monopsony power can then perhaps be justified as a countervailing policy to prevent the excessive transfer of wealth from consuming to producing countries. Although this justification might provide a legitimate political reason to undertake such an action, we should stress that no externality in the sense considered in this report exists for this policy.18 That the ability of the United States to exercise monopsony power is not an externality is recognized by Leiby (2007) and Greene (2009) among others. Green noted that the “costs of oil dependence are not external costs and neither a tax on oil nor a tax on imported oil is an adequate solution to the problem,” although Green argued that either of these taxes can ameliorate the problem (see pp. 10-11 in Green 2009). Turning to disruption costs, we consider the following questions. First, is macroeconomic disruption an externality? Second, if it is an externality, is the cost substantive and quantifiable? On the first question, most econo- mists that have studied this issue would agree that abrupt increases in oil prices adversely impact the economy. Differences arise over the magnitude 16 If OPEC is exercising cartel output restricting power, then the exercise of monopsony power adds to a pre-existing distortion, and the deadweight loss is slightly more complicated than suggested by the triangle in Figure 6-1. 17 A pecuniary externality is not an externality in the sense defined in Chapter 1. Rather it is a transfer of income or wealth arising from some action or policy that is transmitted through the marketplace. Unlike standard externalities, pecuniary externalities do not involve any loss of efficiency. 18 The use of monopsony power to extract rents from an energy cartel raises the question of the response by the cartel to the use of this power. An optimizing cartel will wish to raise the price to offset the exercise of monopsony power. However, it will be unable to recover all of the rents extracted by the use of monopsony power. A clumsy cartel (viz. Adelman [1980]) may be able to retaliate in a way that raises their profits. Such retaliation would simply reflect their previously nonoptimizing behavior. The analysis in Figure 6-1 assumes a competitive market supplying the product. How a cartel responds affects the welfare transfer to the United States. Alhajji and Huettner (2000) statistically rejected the hypothesis that OPEC acts as a cartel. They could not reject the hypothesis that Saudi Arabia acts as a dominant firm. In this view, Saudi Arabia can influence world oil prices but not OPEC member production decisions.

OCR for page 309
330 HIDDEN COSTS OF ENERGY of the impact and the extent to which other events and actions play a role in magnifying the impacts.19 The most recent run-up in oil prices in 2007 and 2008 was in large part a demand shock coupled with stagnation in supply, according to Hamilton (2009). Hamilton argued that the onset of the current recession would have been delayed from the fourth quarter of 2007 to the third quarter of 2008 in the absence of the price run-up (Hamilton 2009). That there are links between oil shocks and economic performance is uncontroversial. Leiby (2007) estimated the macroeconomic disruption and adjustment costs for 2006 market conditions. Leiby reported a mean estimate of $5.14 per barrel (2007 dollars) and a range from $2.39 to $8.57 per barrel.20 Does this imply an externality associated with oil consump- tion? The literature on the oil premium and the oil disruption component focuses on measuring the relationship between incremental oil consump- tion and its effect on disruptions to economic activity. We believe that oil disruption costs are not an externality. That said, it is certainly the case that policies that result in a reduction in oil consumption in the United States will most assuredly reduce vulnerability to future oil shocks. In summary, quantifying this possible externality is a challenge. The cost depends importantly on the type of shock and policy response. Changes over time in economic institutions also pose a challenge to measuring the size of this externality. Given the conceptual difficulties in identifying the basis for and size of the externality, we do not think that it makes sense to include a disruption cost as a component in the list of externalities associ- ated with the production or consumption of energy. We do recommend that further research be carried out to better understand this issue. SECURITY OF ENERGY SUPPLY Concerns about the security of the energy supply (as distinct from na- tional security as discussed in the next subsection) arise from the possibility that resources may become unavailable. Security concerns may pertain to energy sources (for example, oil, natural gas, and uranium) or materials that are critical for energy production, distribution, or consumption (for example, lithium for lithium-ion batteries). Risk of disruption exists when supply is dominated by one or a few countries (or facilities) that are unreli- able (for example, unstable in ways that may disrupt operations, as from 19 Bernanke et al. (1997), for example, argued that tightening of monetary policy exacerbated the output effects of the 1973 oil shock. Hamilton and Herrera (2004) presented results sug- gesting that monetary policy did not play a role. Blanchard and Gali (2008) argued that real wage rigidities in the 1970s exacerbated oil shocks. 20 Leiby reported values in 2004 dollars.

OCR for page 309
331 INFRASTRUCTURE AND SECURITY civil strife within the country) or that may choose to restrict supply for political or other objectives (for example, the OPEC oil price shocks). We argue that these sources of insecurity are not an externality. They are supply conditions that are presumably incorporated in market out- comes. For example, buyers of a resource that is subject to risk of supply interruption will seek ways to reduce the risk of disruption, or the harm if disruption occurs, through reducing reliance on the resource, seeking alternative suppliers, maintaining a stockpile, having financial insurance, and other measures. The expectation that demand for the resource would increase if its supply were more secure provides an incentive for suppliers to develop methods for enhancing security. NATIONAL SECURITY EXTERNALITIES Energy is inextricably linked with national security. The U.S. demand for oil contributes to high oil prices that provide support for hostile foreign regimes with large reserves of oil. Second, dependence on foreign energy sources creates dependencies that may constrain foreign policy. Third, some have argued that the oil price paid by U.S. consumers does not reflect the true cost of oil, in particular the cost of U.S. military presence in the Middle East or of maintaining military readiness to protect oil supply lines. The committee discusses these issues in this section. Energy and Foreign Policy Considerations High oil prices provide a source of revenue for countries with foreign policies at odds with the United States (for example, Iran and Venezuela). One could make the argument that U.S. consumers do not take into account that their oil consumption contributes to actions by foreign countries that negatively impact the United States. A simple analogy illustrates the problem with viewing that situation as an externality. Let us assume that my neighbor burns trash in his backyard that causes pollution that adversely affects my household. This is a clear externality. Further assume that I purchase commodities in a store owned by my neighbor. My consumption thus provides income for my neighbor that leads him to purchase more commodities and produce more trash to be burned. My purchase of goods from my neighbor’s store is not an exter- nality. Rather, the neighbor’s burning of trash is the externality. Restricting (or taxing) my purchases indirectly reduces the externality, but it does so in a highly inefficient manner. It would be more efficient to address the externality directly. In a similar vein, U.S. oil consumption that enriches countries with which the United States has differences is not an externality. Rather, U.S.

OCR for page 309
332 HIDDEN COSTS OF ENERGY consumption makes inimical actions possible. In the absence of any ability to address the foreign policy problem directly, it may be desirable to reduce oil consumption to lower world prices. However, such an effort would be an imperfect proxy for better targeted instruments and would hurt oil- producing friends and foe alike.21 In addition to funding activities that are inimical to U.S. interests, rising oil prices may weaken the instruments of economic statecraft. One could argue, for example, that high oil prices through the latter half of 2008 rendered economic sanctions on Iran for its nuclear activities ineffective.22 Dependence on foreign energy sources may constrain U.S. foreign pol- icy. For example, the Bush administration’s goal of furthering the spread of democracy in the world was constrained by U.S. ties to major oil-producing states with autocratic regimes in control. The 2006 report by the Council on Foreign Relations on U.S. oil dependency noted that oil dependence can cause “political realignments that constrain the ability of the United States to form partnerships to achieve common objectives. Perhaps the most per- vasive effect arises as countries dependent on imports subtly modify their policies to be more congenial to suppliers. For example, China is aligning its relationships in the Middle East (e.g., Iran and Saudi Arabia) and Af- rica (e.g., Nigeria and Sudan) because of its desire to secure oil supplies” (Deutch and Schlesinger 2006, pp. 26-27). Deutch and Schlesinger also noted that oil revenues can undermine ef- forts to support good governance. This is another example of the way in which oil revenues can undermine the tools of economic statecraft. Russia, for example, is less responsive to efforts to promote democracy when it has ample oil and gas revenues that reduce its reliance on Western economic assistance. Having constraints placed on foreign policy goals because of oil depen- dency is arguably an externality that is not recognized in the price of oil, but consider two points. First, it is not clear what the incremental reduction in these costs would be were the United States to reduce its oil consumption by a modest amount (say, 10%). Second, it is not clear that this cost could be monetized even if the marginal cost were positive. Therefore, the commit- tee notes the possibility of dependence on imported oil being an externality and recommends further research on this topic to better understand these important issues. 21 See Fullerton et al. (2001) for a discussion of the efficiency of imperfectly targeted instruments. 22 See Deutch and Schlesinger (2006) for further discussion of the role oil funds play in pro- viding flexibility to countries to pursue policies at odds with those of the United States.

OCR for page 309
333 INFRASTRUCTURE AND SECURITY Energy and Military Considerations The argument has been made that the true cost of oil does not reflect the cost of maintaining a military presence in the Middle East or of main- taining a military preparedness. Parry and Darmstadter (2003) reported that analysts generally do not include this cost in any exercise to measure an oil premium for two reasons. First, it is difficult to disentangle military spending for such political goals as reducing terrorism or providing support for Israel from spending to protect oil supply routes. It is also unlikely that whatever spending is specific to securing the supply routes would change appreciably for a moderate reduction in oil flowing from that region to the United States. In other words, the marginal cost is essentially zero. This view is held by a number of other researchers in the area, including Bohi and Toman (1995). The committee adopts this position. We note, however, that military expenditures could be affected by a large drop in oil consump- tion—for example, a reduction in oil consumption to zero. Measuring the impacts on military spending (or for that matter on a host of economic and political responses) from a large change in oil consumption would require extrapolating existing statistical evidence well out of a sample. To do so would give rise to—at best—speculative estimates. We would go further and argue that military spending—to the extent it occurs to safeguard oil-production sites and transportation lanes—is a government subsidy to production. It replaces the need for private security expenditures that would otherwise have to be incurred to provide equivalent protection for oil production and transport. Nuclear Waste and Security In addition to the potential health and environmental damages associ- ated with generation of electricity through nuclear technologies (see Chap- ter 2), several potential external costs are associated with nuclear security. Specifically, the following considerations have been raised: 1. To what extent does the transportation and deposition of fission- able material post-use represent an increased opportunity for terrorists or other parties interested in unlawful use of the material? 2. To what extent does the long-term deposition of fissionable mate- rial create risk of catastrophic accidents above and beyond the theoretical risk of nuclear accidents at the sites? 3. To the extent that either of the above costs can be quantified, to what extent have they been internalized by existing regulations, insurance requirements (including liability costs required by regulations), or other market mechanisms?

OCR for page 309
334 HIDDEN COSTS OF ENERGY Concerns about the environmental hazard of nuclear waste, which may continue to emit radioactive particles for thousands to millions of years, has been the subject of national debate as an environmental question for decades and came to a head after the incident at Three Mile Island. At that juncture, congressional legislation created an affirmative obligation of the federal government to provide long-term storage in the future, although this was not accomplished by 1998, and the U.S. government has been pay- ing liability payments of approximately 0.5 billion dollars per year to the operators of the existing 104 nuclear power facilities for this failure. The facilities remain under private control but are regulated tightly by the U.S. Nuclear Regulatory Commission. During the cold war, the Nuclear Regulatory Commission developed plans to protect sites from “enemies of the United States” based on “design- based threat” scenarios contained in a series of classified documents, but not until after 9/11 did the reference point change to consider nuclear waste as potential materials for harm—probably one of the reasons for legisla- tive progress since that time, most notably the Energy Policy Act of 2005 (EPACT), which by all accounts has led to substantial upgrades in security procedures and oversight by the Nuclear Regulatory Commission (NRC 2005, GAO 2006). However, the NRC report in 2005 concluded that, de- spite progress, there remained at least some finite risk of security breaches at existing or planned sites and made a series of recommendations to fur- ther enhance security. However, neither the probability of such a breach nor the damages that might ensue, either locally or in the aggregate, have been estimated in quantitative terms. Specific to the waste storage issue, DOE finally submitted to the Nu- clear Regulatory Commission in 2008 the formal license application to operate a national waste-material repository at Yucca Mountain, Nevada, and provided extensive estimates of the cost in the range of $100 billion through 2030. Almost as quickly, the new administration announced its intent to abandon this project, but the future strategy for waste disposal remains open. The case for a deep central repository such as the one pro- posed, in addition to the potential economic efficiency and compliance with earlier legislation, is that security would be more readily achieved at a single site than at many, an argument that has not won favor among some near the site. As with the broader security issues at the nuclear facilities, probabilities of an adverse event involving waste storage, under the current (disseminated) or envisioned (single-site) schemes have not been quantified, nor have the potential damages under any scenario. A further level of complexity relates to emerging technologies to modify the life cycle of nuclear fuels at the back end to reduce the long-term storage need. Simplifying the principle, reprocessing of spent fuel could, based on current knowledge, result in reuse of the material to extract almost all of

OCR for page 309
33 INFRASTRUCTURE AND SECURITY its radioactivity; this approach is also being taken to achieve sustainability in all materials cycles. Although this may in the future displace in part or whole the security risks associated with storage, the technology itself, including the new facilities created and the likelihood that the reprocessed materials would need at various phases to be transported, creates new risks, for which neither the costs of appropriate controls nor any estimate of risk of breach has been calculated.23 Finally, it must be mentioned that there is also some potential in the security arena for external benefit from expanding the American nuclear energy capability, namely, the likelihood that the United States and its government could be proportionately more influential in global nuclear negotiations. Based on current developments, it is a certainty that many countries will turn to nuclear energy as the best solution to their energy needs, including many that are politically unstable or hostile to the United States; the potential that the United States could be a leader, both techno- logically but also politically, hinges, in the views of some (NEAC 2008), on the degree to which the United States also follows this energy pathway. Taking the available information, the committee concludes as follows: 1. The direct cost of nuclear storage under present and envisioned scenarios is high, but the potential for damages from security breaches not incorporated in these costs cannot be quantified. Even if the probability of such an event or its damages could be quantified, it would still be impos- sible to calculate the marginal cost—that is, the risk of an additional facility to a world still populated with nuclear warheads and with many foreign countries already committing to a nuclear energy future. 2. As with other damage possibilities associated with the generation of electricity, the distribution of potential damages is certain to be unevenly shared. The move to Yucca Mountain or another centralized storage site, if approved, would probably reduce aggregate risk but obviously increase local and regional risk; conversely, a centralized site would reduce local and regional risks at the 100-plus sites where waste is currently disposed at U.S. government cost and at all future locations. 3. It is also difficult to assess the extent to which the potential damage from security risk has already been internalized. Certainly, the net upgrade of security requirements brought about by EPACT and other post-9/11 Nuclear Regulatory Commission changes has internalized some of the costs. However, because taxpayers presumably bear some of the costs in the event of a high-cost security incident (through an implicit commitment to compensate victims of the event through government relief), the degree 23 Note,however, that the MIT 2003 study concluded that once-through technology with permanent storage of waste material was preferable to closed fuel-cycle technologies.

OCR for page 309
336 HIDDEN COSTS OF ENERGY to which the market has internalized these risks is difficult if not impossible to measure. CONCLUSION In conclusion, the committee finds the following: 1. The nation’s electricity-transmission grid is vulnerable to failure at times because of transmission congestion and the lack of adequate reserve capacity. Electricity consumption generates an externality because indi- vidual consumers do not take into account the impact their consumption has on aggregate load. Damages from consumption could be significant, and it underscores the importance of investing in a modernized grid that takes advantage of new smart technology and that is better able to handle intermittent renewable power sources. 2. Externalities from accidents at facilities are largely internalized and—in the case of the oil and gas transmission network—of negligible magnitude per barrel of oil or thousand cubic feet of gas trans-shipped. We find that the monopsony component of the oil consumption premium is not an externality. 3. Although government policy may be desirable to serve as a coun- tervailing force to monopoly or cartel producer power, it is a separate issue from the focus of this report. 4. We find that macroeconomic disruptions from oil supply shocks are not an externality. We also find that sharp and unexpected increases in oil prices adversely affect the U.S. economy. Estimates in the literature of the macroeconomic costs of disruption and adjustment range from $2 to $8 per barrel in 2007 dollars. 5. Dependence on imported oil has implications for foreign policy, and we find that some of the effects can be viewed as an externality. We find, however, that it is impossible to quantify these externalities. The role of the military in safeguarding foreign supplies of oil is often identified as a potential externality. We find it difficult if not impossible to disentangle nonenergy-related reasons for a military presence in certain regions of the world from energy-related reasons. Moreover much of the military cost is likely to be fixed in nature. A 20% reduction in oil consumption, for example, would probably have little impact on the strategic positioning of military forces in the world. 6. Nuclear waste and proliferation raise important issues and pose difficult policy challenges. The extent to which uninternalized externalities exist is difficult to measure. Moreover, it is very difficult to quantify them. Thus, we do not report numerical values in this report but recognize the importance of studying this issue further.