Life Cycle Assessment of Overhead and Underground Primary Power

Jun 16, 2010 - Collected primary data include material and energy requirements of aluminum and copper rod production, wire drawing, stranding and test...
0 downloads 0 Views 1MB Size
Environ. Sci. Technol. 2010, 44, 5587–5593

Life Cycle Assessment of Overhead and Underground Primary Power Distribution SARAH BUMBY,† EKATERINA DRUZHININA,† REBE FERALDI,† DANAE WERTHMANN,† R O L A N D G E Y E R , * ,† A N D J A C K S A H L ‡ Bren School of Environmental Science and Management, University of California, Santa Barbara, California 93106-5131, and Southern California Edison, 2244 Walnut Grove, Rosemead, California 91770

Received December 14, 2009. Revised manuscript received June 4, 2010. Accepted June 8, 2010.

Electrical power can be distributed in overhead or underground systems, both of which generate a variety of environmental impacts at all stages of their life cycles. While there is considerable literature discussing the trade-offs between both systems in terms of aesthetics, safety, cost, and reliability, environmental assessments are relatively rare and limited to power cable production and end-of-life management. This paper assesses environmental impacts from overhead and underground medium voltage power distribution systems as they are currently built and managed by Southern California Edison (SCE). It uses processbased life cycle assessment (LCA) according to ISO 14044 (2006) and SCE-specific primary data to the extent possible. Potential environmental impacts have been calculated using a wide range of midpoint indicators, and robustness of the results has been investigated through sensitivity analysis of the most uncertain and potentially significant parameters. The studied underground system has higher environmental impacts in all indicators and for all parameter values, mostly due to its higher material intensity. For both systems and all indicators the majority of impact occurs during cable production. Promising strategies for impact reduction are thus cable failure rate reduction for overhead and cable lifetime extension for underground systems.

1. Introduction This article is concerned with assessing the environmental impacts of overhead and underground primary power, or medium voltage (MV), distribution systems in Southern California. In California and elsewhere electricity demand is projected to grow, which will lead to growth in power transmission and distribution (1–4). Overhead distribution systems create vehicle collision hazards, visual obstruction, and increased damage in fires (5–7). The California Public Utilities Commission’s Rule 20 provides undergrounding conversion funds from ratepayer fees, giving priority to congested, civic, and scenic areas (8). Likewise, other areas within the United States, the European Union, and Australia are recommending mandates for installing new systems * Corresponding author e-mail: [email protected]; tel: 1 805 893 7234; fax: 1 805 893 6113. † University of California. ‡ Southern California Edison. 10.1021/es9037879

 2010 American Chemical Society

Published on Web 06/16/2010

underground and converting existing overhead infrastructure for aesthetic and safety purposes (5–7). However, there are trade-offs between overhead and underground power distribution. The most widely discussed trade-offs in literature are the following: aesthetics, safety, cost, and reliability. Underground systems are concealed, thus increasing nearby property values and preserving local aesthetics. Also, underground systems reduce the possibility for live-wire contacts and vehicle accidents from collisions with utility poles (9). Although installing underground distribution presents a substantial initial investment, costing four to twenty times more than overhead systems, it may improve reliability and decrease maintenance costs (6, 7, 10). While underground systems may improve reliability due to fewer outages, the time required to repair an outage event is considerably longer than that for overhead systems (5–7, 10). Thus, reliability is a contentious topic and depends significantly on the location of the power distribution system. All of these factors are discussed extensively in literature, yet studies of the environmental impacts of electrical distribution systems are rare. A few life cycle assessments (LCAs) exist that examine individual components of the power grid infrastructure, but to date no LCAs of entire power distribution systems exist (11–14). Southern California Edison (SCE) delivers power to 13 million people in a 50,000 square-mile service area, and is in what is considered one of the most rapidly developing areas in the U.S. (15). SCE’s load-growth for 2008-2017 is estimated at 2.22% per year (615 megawatts per year) systemwide. This growth will require 564 new distribution circuits. Focusing on SCE’s service area, this study evaluates the life cycles of cables and infrastructure associated with medium voltage distribution. Power distribution, which is typically located in densely populated areas, comprises 87% of SCE’s electrical line length (16). It is this urban and suburban area of power delivery where the majority of stakeholders assess the choice between overhead and underground systems. This study provides a basis for more informed decision-making in electricity grid planning and management by adding a new dimension to the discussion, namely environmental impacts of each MV power distribution system drawn from comprehensive process-based LCA.

2. Methods and Data Electric power transmission and distribution systems generate a wide variety of environmental impacts at all stages of their life cycles. Life cycle assessment (LCA) methodology is therefore ideally suited to quantify and compare the overall environmental impacts of different power distribution systems. This study quantifies the cradle-to-grave environmental impacts of overhead and underground primary power distribution systems in Southern California using LCA methodology and terminology as described in ISO 14044 (17). In contrast to other studies, which only assess cable production and end-of-life management (11, 12, 14), the reported research quantifies the environmental impacts from entire power distribution systems throughout their life cycles. 2.1. Scope and System Boundary. The life cycle inventory models of overhead and underground MV power distribution are specific to SCE’s situation to the extent possible. Inventory data was gathered in close collaboration with SCE and their primary suppliers and contractors. SCE-specific data was collected using site visits, on-site measurements, and personal communications. Sites visited include the cable supplier’s manufacturing facilities, SCE service centers, SCE warehouses, and the waste management facilities of SCE’s VOL. 44, NO. 14, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5587

FIGURE 1. Process flow diagram for the overhead power distribution system.

FIGURE 2. Process flow diagram for the underground power distribution system. contractors. These methods facilitated measurement, calculation, or robust estimation of SCE-specific values for production, installation, maintenance, decommissioning, and waste management processes. The processes for which no sufficient SCE-specific inventory data could be collected were modeled using process inventories from LCA databases, including GaBi 4.3 and Ecoinvent 2.0, and literature sources (18, 19). Figures 1 and 2 show material and process flow diagrams of the two product systems. The color coding of processes indicates their data sources. The inventory models for cable manufacturing are based on primary data collected from SCE’s primary cable supplier (20). Collected primary data include material and energy requirements of aluminum and copper rod production, wire drawing, stranding and testing, and cable extrusion. Cradleto-gate inventories for the material inputs to cable manufacturing are based on industry averages (18, 19). 5588

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 14, 2010

The use phase of each product system, which includes installation, maintenance, repair, and decommissioning, was modeled almost entirely based on typical requirements as reported by SCE specialists. Details on all use phase processes can be found in the Supporting Information. The use of dieselfueled utility vehicles makes up the bulk of the use-phase inventory. Inventory models for these processes represent either distances driven by utility vehicles or stationary engine use for activities such as hydraulic and auxiliary work, digging, and pumping. For the overhead system, installation processes include digging holes for poles, setting poles, and stringing the cable. Installation of an underground system typically requires digging the trench, placing vaults and conduits, mixing and pouring concrete, filling the remaining space with backfill, and pulling the cable. Modeling of maintenance accounts for impacts from transportation, tree trimming for the overhead system, and pumping vault water out of the

TABLE 1. Calculation of Reference Flow Masses from Material Requirements of Circuit Installation components

materials

installation flows (kg)/(circuit mile)

overhead 4 cables 1.85 S-77 reels 25 utility poles and 30 crossarms 30 steel brackets 100 insulators

3.96 S-300 reels 1 cable duct and 5.3 vaults 5.3 vaults 6 cable conduits

) reference flows (kg)/(circuit mile · year)

aluminum alloy 1350 galvanized steel galvanized steel wood (douglas fir) pentachlorophenol (PCP) galvanized steel polyethylene (PE)

3,020.25 474.34 336.43 9,070.59 266.38 419.57 90.72

(1/63 + 0.038) (1/63 + 0.038) (1/63 + 0.038) 1/50 1/50 1/50 1/50

163.75 25.72 18.24 181.41 5.33 8.39 1.81

aluminum alloy 1350 copper polyethylene (PE) galvanized steel concrete steel rebar polyvinyl chloride (PVC)

6,734.06 2,277.96 7,407.73 2,277.61 1,096,593.49 15,909.87 66,985.78

(1/30 + (1/30 + (1/30 + (1/30 + 1/125 1/125 1/125

352.01 119.08 387.22 119.06 8,772.75 127.28 535.89

underground 3 cables

× conversion factor (1)/(year)

underground infrastructure. Repair consists of replacement of the cable sections due to failure events. Cables decommissioned from the SCE service area are sorted and baled by SCE’s waste management contractor. The bales are then shipped to China for recycling. Inventory data required to model the end-of-life management processes of cable sorting, packaging, and chopping were gathered in close collaboration with SCE’s cable waste management facility. Recycling rates were parameterized in the model, so a range of plausible values could be analyzed. Transport of materials was modeled using data specific to each trip including payload capacity and utilization ratio of the vehicle, trip distance and profile, fuel specifications, and vehicle emission standards. The travel paths between specific SCE supply chain facilities and the receiving SCE warehouses and service centers were specified using Google Maps (21). For an accurate representation of the electricity used within Southern California, power inputs required for the use phase were specific to the SCE utility profile. Power inputs required for the production and end-of-life phases used the average U.S. electricity mix, except for cable recycling, which occurs in China and thus, uses the average Chinese electricity mix (22–25). The three electricity profiles, together with the California mix, can be found in the Supporting Information. 2.2. Functional Unit and Reference Flows for the Baseline Case. This study focuses on primary, or medium voltage (MV), power distribution, which comprises 87% of SCE’s total electrical line length (16). Much of the debate on alternative power delivery methods is about primary distribution, which involves many stakeholders in its planning and management. Inventory model and data are specific to SCE, their upstream suppliers, and downstream waste management providers. The study focuses only on those materials and processes that are used in new SCE installations. For instance, many utility poles were previously treated with coal tar creosote and are still in use as part of existing infrastructure. However, SCE has shifted to using pentachlorophenol (PCP) as wood preservative and thus, only impacts from PCP treated poles were analyzed in this study. In summary, the goal of the inventory models is to represent the life cycles of new SCE overhead and underground MV distribution lines. The functional unit is defined as the distribution of MV power in one circuit over one mile (1.609 km) and for one year. The reference flows include cables, infrastructural components, and processes, but exclude capital equipment and human labor, which are estimated to be similar between both systems. The specific overhead and underground cables

0.019) 0.019) 0.019) 0.019)

used in the study were selected based on their high purchase volume and comparable capacity for power delivery. The overhead electrical circuit requires four cables to accommodate three alternating currents in different phases and a neutral wire which is connected to the ground. The underground circuit requires only three cables, because each includes a copper concentric neutral wire. Table 1 lists the flows required to install one mile of circuit (installation flows) and translates these values into reference flows according to component durability and frequency of failure events. Comparing overhead and underground power distribution systems is complicated by the facts that different system components have different theoretical lifetimes and failure events affect the actual lifetimes of the components. The average failure frequency for overhead cables is 0.9 events per year and mile (26). Usually only one section of overhead cable is replaced in the event of a failure; a length equal to the distance between two utility poles (225 feet or 68.6 m on average). A circuit mile of overhead cable might thus be replaced in pieces in as little as 5280/(0.9 × 225) ≈ 26 years and is most unlikely to ever reach its theoretical cable lifetime of 40 years (26). The environmental impacts of initial installation are therefore distributed over the planning horizon, which has been set as 63 years in the baseline scenario (26, 27). For the overhead cable and its associated components and processes, the factor that converts the installation flows into reference flows is thus the sum of one over the planning horizon, 1/63, plus the average fraction of circuit mile replaced every year due to failure events, i.e., (0.9 × 225)/5280 ) 0.038. The average failure frequency for underground cables is 0.1 events per year and mile, and the average length of replaced cable is 1000 feet or 304.8 m, which is the average distance between two vaults (26). The underground cable will therefore reach its theoretical lifetime and the environmental impacts of initial installation are distributed over the underground cable lifetime. For the underground cable and its associated components and processes, the factor converting the installation flows into reference flows is thus the sum of one over the theoretical lifetime, 1/30, plus the average fraction of circuit mile replaced every year due to failure events, i.e., (0.1 × 1000)/5280 ) 0.019. Insulators, steel castings, and PCP are part of the overhead utility poles and replaced together with the poles. The conversion factor for those three components is thus one over the pole lifetime, i.e., 1/50 (27). In the underground system, the conversion factor for vaults, ducts, and conduits is one over their estimated lifetime, i.e., 1/125 (27). Data for VOL. 44, NO. 14, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5589

TABLE 2. Baseline Values of Inventory Model Parameters and Value Ranges Used in Sensitivity Analysis parameter

baseline value

range

unit

source

overhead cable lifetime PCP leaching to soil recycling rate failure frequency planning horizon

40 0 0.94 0.9 63

30 - 50 0-1 0.91 - 0.95 0.63 - 1.17 26 - 100

years fraction of PCP leaching into soil fraction of recovered cable mass events per mile and year years

26 28–30 12, 31 26 27

underground cable lifetime recycling rate infrastructure lifetime failure frequency

30 0.84 125 0.1

20 - 40 0.73 - 0.95 100 - 150 0.07 - 0.13

years fraction of recovered cable mass years events per mile and year

26 12, 31 27 26

maintenance of both systems are already in annual terms and can thus be used directly as reference flows. Infrastructural components such as poles, vaults, ducts, and conduits can support more than one power circuit. Underground MV distribution systems in the SCE service area are designed to accommodate one, four, or six circuits, while overhead poles can carry one, two, or four circuits. Meaningful comparisons require the number of circuits per infrastructure to be the same, i.e., either one or four for both systems. Both cases were studied, yet because the results were very similar, only the first one is reported here. In other words, overhead and underground MV power distribution systems with one circuit per supporting infrastructure are modeled and assessed in this paper. 2.3. Sensitivity Analysis. Several important aspects in each power distribution system are subject to significant uncertainty. For each of these uncertainties, parameters were introduced into the inventory model to conduct sensitivity analysis and examine the ensuing range of results (see Table 2). The cable lifetime ranges are from literature and reflect various designed lifetimes for MV power cables (26). The designed lifetime of the underground cable is shorter than that of the overhead cable. This is mainly due to the plastic insulation of the underground cable being susceptible to degradation, such as water treeing which generates waterfilled microcavities in the polymer. There is large uncertainty in literature regarding rates of pentachlorophenol (PCP) leaching from utility poles (28–30). Thus, leaching is varied from 0 to 100% to capture all possibilities of contaminant release during pole use. For both overhead and underground systems, cable recycling rates are uncertain because the sorted cables are shipped to China for material recovery and reprocessing. The upper bound of the recycling rate is an international average from literature, whereas the lower bound is estimated from material market values to account for the uncertainty of material recovery rates and reprocessing yields overseas (12, 31). Whereas the lifetime of the overhead infrastructure is fairly well-known, that of the underground system infrastructure is not. The oldest underground power delivery infrastructures have been in place for about 100 years and have yet to be decommissioned (27). Underground infrastructure materials (concrete, steel, and PVC) are very durable. For these reasons, the underground infrastructure lifetime is assumed to be 125 years with a 20% error margin (27). Another uncertain factor is cable failure frequencies due to external events. The overhead lines are more exposed to forces of wind, fire, ice, and human interaction than underground lines and thus, their failure frequency is higher. The baseline failure frequency for each system is based on literature values, and a 30% error margin was selected for each to account for variation in external influences (26). The planning horizon was introduced into the inventory model of the overhead system because the high failure frequency 5590

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 14, 2010

makes it very unlikely that the overhead cable will reach its designed lifetime. A wide range of values (26-100 years) was chosen in order to assess the sensitivity of the results with respect to this parameter. 2.4. Assumptions and Limitations. Allocation, i.e., the partitioning of inputs and outputs among product systems, is avoided in the inventory models by using consequential system expansion, also called avoided burden method (32). In the case of cable recycling, end-of-life cable collection and recycling are added to the product system and the production processes assumed to be avoided by the recycled materials are subtracted. Collection and recycling includes transportation, disassembly, sorting, and secondary processing of constituent materials. The same approach is used to account for the energy recovery from incinerating wood chips during pole and crossarm production, the energy recovery from incinerating the tree trimmings from overhead system maintenance, and the reuse of the cable reels. Cable reels have a reuse rate of approximately 80%, while the remaining 20% are assumed to be recycled. Because the details of end-of-life overhead and underground cable recycling are unknown, a range of plausible recycling rates was used (see Table 2). The upper bound of the recycling rate is 95%, which is a literature value for wire and cable products in general, and is very high and only achievable if cable collection, material separation, and material recycling are highly efficient (12). Material recycling also does not automatically translate into displacement of primary production; particularly if the collection, separation, and recycling processes change the inherent properties of the materials. The underground cable has higher material complexity than the overhead cable and thus increased likelihood of material loss and downcycling. The lower bounds for the recycling rates of each cable product (see Table 2 and Supporting Information) account for possible reprocessing losses and limitations in the displacement potential of the resulting secondary materials (31, 33). Transformers were excluded in both product systems. Material compositions of typical and equivalent overhead and pad mount transformers are very similar and yield reference flows of nearly identical mass. A first-order assessment concluded that environmental impacts from transformer production and end-of-life management are very similar and, depending on the impact category, would add 0.3-0.6% to the UG system and 0.4-3.2% to the OH system (see Supporting Information). All process types and activity levels were estimated for one mile of straight circuit with no topographical barriers and no unusual obstacles for installation, such as roads, hard rocks, hills, or corners. In many cases, geological, terrain, and land use conditions will affect the quantity of infrastructural components needed and energy required for installation, which may significantly change the relative impacts of the two systems. For example, assuming no

TABLE 3. Life Cycle Impact Assessment (LCIA) Results for Baseline, Best Case, and Worst Case overhead

underground

impact category (indicator unit)

baseline

best

worst

baseline

best

worst

ADP (kg SBeq) AP (kg SO2 eq) EP (kg PO4 eq) FAETP (kg DCB eq) GWP (kg CO2 eq) HTP (kg DCB eq) POCP (kg C2H4 eq) TETP (kg DCB eq)

7.09 4.09 2.42 91.67 1419.14 256.51 0.45 5.92

3.46 2.67 2.25 53.39 1120.50 181.73 0.32 4.12

14.31 6.95 2.75 174.73 2019.30 410.74 0.73 9.72

63.66 32.68 3.73 527.08 7682.8 1376.29 3.65 29.16

48.61 25.01 2.91 243.85 5939.30 907.81 2.78 19.16

88.13 45.20 5.08 1024.80 10527.00 2160.80 5.07 45.98

obstacles implies that the underground system does not require landscaping or surface repavement after installation. For data quality reasons, the truck transportation process inventories are based on EU, rather than U.S., diesel fuel emission standards. Although the current U.S. regulation for new diesel vehicles is comparable to EU Euro 5 emission standards, diesel vehicles currently in use in the U.S. are older and the majority of vehicles are below this standard. Therefore, it was estimated that the EU Euro 4 diesel fuel emission standards are the most suitable, conservative assumption for our analysis (34, 35). This study does not assess land use impacts. Our research indicates that in many cases land use is similar for both distribution systems. California’s General Orders 95 and 128, which contain the rules for construction of overhead and underground electric supply, do not contain standard mandatory horizontal land clearances (36, 37). Also, distribution lines are typically colocated on public land, like city streets, with other utilities, like water, gas, and sewer, and other land uses, like sidewalks and driveways. The effects of electromagnetic fields (EMFs) were not considered in this study because they are quite controversial and there are many competing claims about the possible harm caused (9, 38–40). While the potential harm from EMFs is still an open question, EMF concentrations around primary power distribution lines are mainly defined by the distance from the power line, design of the line, and the amount of current the line is carrying (41). Underground primary power delivery systems would thus result in higher EMF exposure for power consumers and the general public, since underground circuits are usually closer to the ground’s surface than overhead lines. Loss of power in an alternating current circuit is related to its resistance and reactance, i.e., its opposition to steady current and change in current. The studied UG cable has a lower rated resistance than the studied OH cable, and the studied UG distribution circuit is likely to have a lower reactance than the OH circuit (27). However, no scientifically robust and generally accepted way has been identified to convert these properties into differences in power loss and resulting differences in environmental impact. For this reason the relative environmental impact of power losses has been left to further study.

restrial ecotoxicity potential (TETP) (32). For each system they show indicator results for the baseline model, and the best and worst case, which are defined as those combinations of parameter values that yield the largest and smallest environmental impacts. These parameter combinations are the same for all indicators, since each parameter impacts all indicators in the same direction. The joint impact of the model parameter ranges shown in Table 2 results in a wide range of results. Remarkably, the worst case of the overhead system has lower impacts than the best case of the underground system for all indicators. This provides significant evidence that the results in Table 3 and Figure 3 are robust. With the exception of eutrophication (EP), the baseline impacts of the underground system are between five and nine times larger than those of the overhead system. The baseline eutrophication impact of the underground system is only 50% larger than the baseline result of the overhead system, which is due to the large EP from disposal of the wooden poles and crossbars. The global warming impact related to one mile of the overhead power distribution system for one year is equivalent to the direct CO2 emissions of generating 1494 KWh of coal-based electricity (42) or combusting 161 gallons of gasoline (43). For the underground power distribution system, the numbers are 8087 KWh of coal-based electricity or 873 gallons of gasoline. A typical U.S. household consumes about 11 000 KWh per year (44). The contribution analysis shown in Table 4 reveals that, for both systems and all impact categories, the majority of impact occurs during cable production. Within cable production, aluminum production is the single largest contributor. Contribution analysis is not entirely straightforward because the consequential system expansion used to account for the impacts of cable recycling and energy recovery through byproduct combustion reduces impacts, i.e., creates negative impacts. In the overhead system, the negative GWP of “other production” comes from the CO2 sequestration of timber growth, while the negative impacts of the use phase are from the avoided burdens caused by incinerating the tree trim-

3. Results and Discussion The life cycle inventories for the overhead and underground systems have been modeled in the GaBi LCA software, which offers characterization factors for a wide range of impact assessment methods (18). Table 3 and Figure 3 display the results for eight pertinent indicators from the CML2001 suite of impact categories, which are abiotic depletion potential (ADP), acidification potential (AP), eutrophication potential (EP), freshwater aquatic ecotoxicity potential (FAETP), global warming potential (GWP), human toxicity potential (HTP), photochemical ozone creation potential (POCP), and ter-

FIGURE 3. LCIA results for baseline, best case, and worst case, normalized to estimated U.S. total in 2007. VOL. 44, NO. 14, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5591

TABLE 4. Contribution Analysis for the Baseline Scenario life cycle stage f impact category V

cable production

overhead ADP (kg SBeq) AP (kg SO2 eq) EP (kg PO4 eq) FAETP (kg DCB eq) GWP (kg CO2 eq) HTP (kg DCB eq) POCP (kg C2H4 eq) TETP (kg DCB eq)

18.08 10.34 1.24 920.15 2276.89 977.33 1.17 23.87

0.79 0.63 0.09 7.95 -217.58 61.63 0.12 1.48

-3.30 0.18 0.24 -0.58 996.37 0.97 0.02 -0.26

-10.23 -8.29 -1.04 -856.47 -1877.67 -899.06 -0.97 -22.08

1.75 1.23 1.89 20.61 271.13 115.64 0.11 2.90

underground ADP (kg SBeq) AP (kg SO2 eq) EP (kg PO4 eq) FAETP (kg DCB eq) GWP (kg CO2 eq) HTP (kg DCB eq) POCP (kg C2H4 eq) TETP (kg DCB eq)

55.17 28.35 3.11 1995.41 6203.44 2457.46 3.21 52.19

23.87 11.19 1.08 102.04 3584.61 196.65 1.58 8.18

10.53 6.54 1.10 4.10 1542.96 47.29 0.63 1.16

-32.96 -20.39 -2.29 -1670.87 -4668.57 -2036.06 -2.37 -44.23

7.05 6.99 0.73 96.40 1020.39 710.94 0.60 10.86

other production

mings from overhead system maintenance for energy recovery. The avoided burdens from cable recycling are of the same order of, and sometimes even fairly close to, the impacts from cable production. This is due to the fact that recycling of aluminum, copper, and steel is assumed to displace their primary production, which has much higher impacts than recycling. The importance of cable recycling and how it is accounted for in the LCA is further highlighted by the fact that for many categories the impact of cable production alone is higher than the totals shown in Table 3, sometimes several times as high. Another observation from Table 4 is that cable production is considerably less dominant in the underground system. This means that the production, use, and end-of-life management of the additional infrastructure for underground power distribution is much more significant than in overhead systems. This is due to the substantial amounts of concrete, steel, and PVC required for vaults, ducts, and conduits (see Table 1). Overall, the large difference in impact between the overhead and underground system is to a large extent due to the material intensity of underground power distribution infrastructure. Compared to overhead distribution infrastructure, it requires over twice the amount of aluminum to provide the functional unit, over four times as much copper as overhead requires steel, and large additional amounts of PE sheath, PVC conduits, and reinforced concrete vaults and ducts (see Table 1). The additional underground infrastructure is required to protect the conductor from mechanical damage. Mass reduction measures, such as not using conduits or using less concrete, would expose the cables and thus likely shorten their lifetime. One important reason for the high mass of underground conductors themselves is their larger crosssectional area. This is driven by the need to decrease the electrical current density to ensure a safe and efficient temperature range of the cable. In the ground and enclosed by PVC conduits and concrete ducts, the heat from cables does not dissipate as easily as it does in open air. Temperature increases in the underground cable would not only pose the risk of melting the plastic insulation layers, but would also decrease the conductive properties of the cable. On the other hand, increasing the cross-sectional area of conductors reduces their electrical resistance which, other things being equal, would reduce distribution losses. The environmental trade-off between material intensity of distribution cables 5592

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 14, 2010

use phase

gross benefits from cable recycling

other end-of-life management

and associated power loss, which applies to both OH and UG cables, is highly nontrivial and should be studied further. The most important environmental impact reduction strategies that avoid this trade-off are cable failure rate reduction for overhead cables (45) and cable lifetime expansion for underground systems (46, 47).

Acknowledgments We thank Carol Godfrey and Denise Quarles (Southwire Company), Ray Harlow (Young & Company), A. Greg Tellier (Alpert & Alpert Iron & Metal Inc.), Shannon Terrell (Brooks Manufacturing), and Michael Hughes, Tom Beamish, and Glenn Sias (SCE) profusely for their invaluable support.

Supporting Information Available Detailed descriptions of the overhead and underground system components, as well as process inventory data, parameters, and model diagrams. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) EIA. International Energy Outlook 2008; DOE/EIA-0484(2008); U. S. Department of Energy, Energy Information Administration (EIA): Washington, DC, September 2008. (2) EIA. Annual Energy Outlook 2009; DOE/EIA-0383(2009); U. S. Department of Energy, Energy Information Administration (EIA): Washington, DC, April 2009. (3) IEA. World Energy Outlook 2008; International Energy Agency (IEA): Paris, France, 2008. (4) CEC. California Energy Demand 2008-2018: Staff revised forecast; CEC-200-2007-015-SF2; California Energy Commission (CEC): Sacramento, CA, 2007. (5) SCD. Undergrounding electricity cables in NSW; Sydney Cables Downunder (SCD): Pennant Hills, NSW, Australia, 2002. (6) EEI. Out of sight, out of mind: A study on the costs and benefits of undergrounding overhead power lines; Edison Electric Institute (EEI): Washington, DC, 2006. (7) NCPSUC. The feasibility of placing electric distribution facilities underground; North Carolina Public Staff Utilities Commission (NCPSUC): Raleigh, NC, 2003. (8) CPUC. Energy division resolution E-3757; California Public Utilities Commission (CPUC): San Francisco, CA, March 2002. (9) Quanta Technology. Undergrounding assessment phase 1 final report: Literature review and analysis of electric distribution overhead to underground conversion; Florida Public Service Commission, Florida Electric Utilities: Tallahassee, FL, February, 2007. (10) Maney, C. T. Benefits of urban underground power delivery. IEEE Technol. Soc. Mag. 1996, 15 (1), 12–22.

(11) Barras, R. E.; Bow, K. E.; Snow, J. H.; Voltz, D. A. A new era in cable designs and materials to resolve environmental issues. IEEE Trans. Ind. Appl. 1997, 33 (5), 1321–1330. (12) EPA. Wire and cable insulation and jacketing: Life-cycle assessments for selected applications; EPA 744-R-08-001; U.S. Environmental Protection Agency (EPA), Office of Pollution Prevention and Toxics: Washington, DC, 2008. (13) Sedjo, R. A. Wood materials used as a means to reduce greenhouse gases (GHGs): An examination of wooden utility poles. Mitigation Adapt. Strat. Global Change 2002, 7 (2), 191–200. (14) Simonson, M.; Andersson, P.; Emanuelsson, V.; Stripple, H. A life-cycle assessment (LCA) model for cables based on the fireLCA model. Fire Mater. 2003, 27, 71–89. (15) SCE. Company overview; Southern California Edison (SCE): Rosemead, CA, 2009; available at http://www.sce.com. (16) SCE. Piloting a Centralized Remedial Action Scheme (C-RAS) with emerging telecomm/protection technologies; Southern California Edison (SCE): Rosemead, CA, 2007. (17) ISO 14044: Environmental management - Life cycle assessment - Principles and framework; International Organization for Standardization (ISO): Geneva, Switzerland, 2006. (18) PE International. GaBi 4.3 Software and Databases; PE International: Leinfelden-Echterdingen, Germany, 2009. (19) Ecoinvent. Life Cycle Inventory Data v.2.0; Ecoinvent Centre: Du ¨ bendorf, Switzerland, 2007. (20) Godfrey, C.; Quarles, D. Personal communications; Southwire: Carrollton, GA, 2008-2009. (21) Google maps. Available at http://maps.google.com. (22) SCE. Power Content Label 2006; Southern California Edison (SCE): Rosemead, CA, 2007; available at http://www.sce.com. (23) CEC. California Electricity Sector Overview; California Energy Commission (CEC): Sacramento, CA, 2008; available at http:// www.energy.ca.gov. (24) EIA. Electric Power Annual with Data for 2007; U. S. Department of Energy, Energy Information Administration (EIA): Washington, DC, 2009; available at http://www.eia.doe.gov. (25) IEE. Energy in China: Transportation, Electric Power and Fuel Markets; Asia Pacific Energy Research Centre, Institute of Energy Economics (IEE): Japan, 2004. (26) Short, T. A. Electric Power Distribution Handbook; CRC Press: Boca Raton, FL, 2004. (27) Hughes, M. Personal Communications; Southern California Edison (SCE): Rosemead, CA, 2008-2009. (28) Bulle, C.; Bertrand, R.; Samson, R.; Deschenes, L. Sensitivity study of an OCDD environmental fate screening model in soils in the presence of PCP wood-preserving oil. Chemosphere 2008, 73, S149–S157. (29) Lorber, M. N.; Barton, R. G.; Winters, D. L.; Bauer, K. M.; Davis, M.; Palausky, J. Investigation of the potential release of polychlorinated dioxins and furans from PCP-treated utility poles. Sci. Total Environ. 2002, 290, 15–39. (30) Goodrich-Mahoney, J. W.; Muraka, I. P.; Holcombe, L. J.; Horn, M. E. Pentachlorophenol-treated wood poles and crossarms: Toxicity characteristic leaching procedure (TCLP) results. Environ. Int. 1993, 19 (6), 535–542.

(31) Villalba, G.; Segarra, M.; Fernandez, A. I.; Chimenos, J. M.; Espiell, F. A proposal for quantifying the recyclability of materials. Resour., Conserv. Recycl. 2002, 37, 39–53. (32) Guine´e, J. B., Ed. Handbook on Life Cycle Assessment, Operational Guide to the ISO Standards; Kluwer Academic Publishers: Dordrecht, The Netherlands, 2002. (33) Werner, F.; Richter, K. Economic allocation in LCA: A case study about aluminium window frames. Int. J. LCA 2000, 5 (2), 79–83. (34) Dieselnet. Emission standards - European Union: Heavy-duty diesel truck and bus engines; Ecopoint, Inc.: Mississauga, Canada, September, 2009. (35) Dieselnet. Emission standards - United States: Heavy-duty diesel truck and bus engines; Ecopoint, Inc.: Mississauga, Canada, September, 2009. (36) CPUC. Rules for Overhead Electric Line Construction; General Order No. 95; California Public Utilities Commission (CPUC): San Francisco, CA, 2009. (37) CPUC. Rules for Construction of Underground Electric Supply and Communication Systems; General Order No. 128; California Public Utilities Commission (CPUC): San Francisco, CA, 2006. (38) Ahlbom, A.; Cardis, E.; Green, A.; Linet, M.; Savitz, D.; Swerdlow, A. Review of the epidemiologic literature on EMF and health. Environ. Health Perspect. 2001, 109 (6), 911–933. (39) Greenland, S.; Sheppard, A. R.; Kaune, W. T.; Poole, C.; Kelsh, M. A. A pooled analysis of magnetic fields, wire codes, and childhood leukemia. Epidemiology 2000, 11 (6), 624–634. (40) Feychting, M.; Jonsson, F.; Pedersen, N. L.; Ahlbom, A. Occupational magnetic field exposure and neurodegenerative disease. Epidemiology 2003, 14 (4), 413–419. (41) SCE. Electric magnetic fields (EMF); Southern California Edison (SCE): Rosemead, CA, 2009; available at http://www.sce.com. (42) U.S. DOE and EPA. Carbon Dioxide Emissions from the Generation of Electric Power in the United States; US Department of Energy (DOE) & U.S. Environmental Protection Agency (EPA): Washington, DC, 2000. (43) EPA. Average Carbon Dioxide Emissions from Gasoline and Diesel Fuel; EPA420-F-05-001; U.S. Environmental Protection Agency (EPA), Office of Transportation and Air Quality: Washington, DC, 2005. (44) EIA. Residential Energy Consumption Survey; U.S. Department of Energy, Energy Information Administration (EIA): Washington, DC, 2009. (45) Brown, R. E. Electric Power Distribution Reliability; Marcel Dekker, Inc.: New York, 2002. (46) Prysmian. Prysmian Cables & Systems: Superseal; Product Catalog No i-3; Prysmian: Milan, Italy, November, 2007; available at www.prysmianusa.com. (47) UtilX. Underground Power Solutions; UtilX: Kent, WA, 2009; available at www.utilx.com.

ES9037879

VOL. 44, NO. 14, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5593