APPENDIX F ELECTRIC AND MAGNETIC FIELDS GLENBROOK-NORWALK TRANSMISSION LINE REINFORCEMENT

Transcription

1 APPENDIX F ELECTRIC AND MAGNETIC FIELDS GLENBROOK-NORWALK TRANSMISSION LINE REINFORCEMENT

2 Electric and Magnetic Fields: Glenbrook Cables Project Prepared for Northeast Utilities P.O. Box 270 Hartford, CT Prepared by Exponent 420 Lexington Ave Suite 408 New York, NY January 20, 2004

3 Contents Page List of Figures List of Tables Executive Summary iv v vi 1 Introduction 1 2 Sources of Power-Frequency Fields Electric and Magnetic Fields from Power Lines and Other Sources Magnetic Fields Encountered in Everyday Environments Project Sources of Electric and Magnetic Fields Magnetic Fields Measured Along Selected Routes Method for Measuring Magnetic Fields Primary Route Alternative Route Calculated Magnetic Fields from Underground XLPE and HPFF Cable Systems Methods for Calculating Fields from Underground Cable Systems Calculated Magnetic Field Values for XLPE and HPFF Cable Systems 12 3 EMF and Health Assessment of Potential EMF Health Risk Methods for Evaluating Scientific Research Epidemiology and Experimental Studies are Complementary Evaluating Epidemiology Weight-of-the-Evidence Evaluation Epidemiology Studies Experimental Studies Weight-of-the-Evidence Conclusions by Multidisciplinary Groups The IARC Working Group Conclusions of Other Multidisciplinary Review Panels Other EMF Perspectives 27 ii

4 3.4.1 California EMF Program World Health Organization National Institute of Environmental Health Sciences (NIEHS) 28 4 Overall Project EMF Assessment 30 5 References 31 iii

5 List of Figures Page Figure 1. Primary route and northern Alternative route between Norwalk substation and Glenbrook substation under consideration for the 115-kV cable systems 1 Figure 2. Magnetic field levels in the environment 4 Figure 3. Figure 4. Figure 5. Figure 6. Typical magnetic field personal exposures (the same data is plotted in the two graphs; the lower graph expands the vertical axis) 5 Magnetic field recorded at one-second intervals while driving along Primary southern route 7 Magnetic field recorded at one-second intervals while driving along Alternative route 8 Trench cross section of XLPE cable system with two cables per phase 115-kV XLPE cable Installed 10 Figure 7. Typical trench cross-section, 115-kV cables in 2x3 ductbank 10 Figure 8. Trench cross section of HPFF cable system with two cables per phase 11 Figure 9. Magnetic flux density one meter above ground level, 115-kV XLPE Norwalk- Glenbrook underground cable system at normal loading (204 MVA) and peak loading (342 MVA) 12 Figure 10. Magnetic flux density one meter above ground level, Alternative 115-kV XLPE Norwalk-Glenbrook underground cable system at normal loading (204 MVA) and peak loading (342 MVA) 13 Figure 11. Magnetic flux density one meter above ground level, 115-kV HPFF Norwalk- Glenbrook underground cable system at normal loading (204 MVA) and peak loading (342 MVA) 14 Figure 12. Trench cross section of XPLE cables in single conduit 15 iv

6 List of Tables Page Table 1. Summary of magnetic fields measured in a Connecticut town (Bethel) 6 Table 2. Summary of magnetic field levels (mg) measured along Primary route 8 Table 3. Summary of magnetic field levels measured along Alternative route 9 Table 4. Table 5. Table 6. Table 7. Magnetic field values (mg) at one meter above ground 115-kV XLPE cables in duct bank 13 Magnetic field values (mg) at one meter above ground 115-kV XLPE cables in duct bank 13 Magnetic field values (mg) at one meter above ground 115 kv HPFF cable circuits 14 Magnetic field values (mg) at one meter above ground XLPE cables in single conduit 16 Table 8. Hill s criteria for causation 20 Table 9. Conclusions of large multidisciplinary review groups assembled by health agencies and scientific organizations 27 v

7 Executive Summary This report summarizes the assessment by Exponent and Power Delivery Consultants, Inc. (PDC) of the potential effect of the Project on existing levels of electric and magnetic fields (EMF), evaluates health research on EMF, including reviews of the literature published by scientific advisory organizations, and assesses compliance with the Connecticut Siting Council s EMF Best Management Practices. Over the last 30 years, research has been conducted in the United States and around the world to examine whether exposures to EMF have health or environmental effects. These fields are produced by both natural and man-made sources that surround us in our daily lives. They are found throughout nature and in our own bodies, and the earth itself. The earth produces a static direct current magnetic field it is this field that is used for compass navigation. EMF that oscillates 60 times per second a frequency of 60 Hertz (Hz) is found wherever electricity is generated, delivered, or used. Power lines, wiring in homes, workplace equipment, electrical appliances, and motors all produce alternating current (AC) EMF at this frequency in North America. Fields at this frequency differ significantly from fields at the higher frequencies characteristic of radio and television signals, microwaves from ovens, cellular phones, and radar (which can have frequencies up to billions of Hz). The proposed Project will only affect ambient levels of magnetic fields, with the greatest effect closest to the underground line. The proposed line will not affect ambient levels of electric fields above ground because the conductors will be surrounded by metal cladding and the earth. The Project will have a limited effect on ambient magnetic field outside the town roads and substation properties. For this report, the magnetic fields from two types of underground cable systems were calculated. At a distance of 50 feet, the magnetic field from cross-linked polyethylene (XPLE) cables in concrete conduits would be less than four milligauss (mg) under normal loading conditions; for the same conditions the conductors contained within high-pressure fluid-filled steel pipe (HPFF) would produce a magnetic field too small to measure. The differences in field levels between these two designs, or between different current levels on these cable systems, become smaller as distance from the lines increases. The consensus of scientists who have reviewed the literature for scientific and regulatory organizations including the International Agency for Research on Cancer, the National Institute of Environmental Health Sciences, the Health Council of the Netherlands, and the National Radiological Protection Board of Great Britain is that no cause and effect relationship between EMF from any source and ill health has been established at the levels generally found in residential environments. Moreover, the information provided in this report demonstrates that the proposed Project complies with the Council s Electric and Magnetic Field Best Management Practices. vi

8 1 Introduction Connecticut Light and Power (CL&P) has proposed to construct two underground 115,000-volt (115-kV) transmission circuits (cable systems) between the Glenbrook and Norwalk Substations (the Glenbrook Cables Project ). The double circuit 115-kV cable system would extend approximately eight to ten miles, from CL&P s existing Glenbrook Substation in Stamford to the existing Norwalk Substation in Norwalk. The line would be located within or adjacent to existing public roadway rights-of-way (ROW). For planning purposes, multiple routes were considered, as well as several variations on these routes. The southern Primary route is identified on an aerial photograph of the study area (Figure 1; as shown in the Application). Also shown is an Alternative northern route that follows routes 106 and 123 to the North. Figure 1. Primary route and northern Alternative route between Norwalk substation and Glenbrook substation under consideration for the 115-kV cable systems 1

9 This report summarizes measurements made by Exponent of ambient magnetic field levels along roads that are considered for the Primary and Alternative routes of the underground line. To characterize magnetic fields produced by the proposed underground cable system, PDC calculated magnetic field levels associated with two cable system designs that have been most commonly used for transmission at 115 kv (Section 2). An up-to-date assessment of current health research on magnetic fields is provided in Section 3. The report concludes with an overall assessment based on relevant guidelines and standards (Section 4). 2

10 2 Sources of Power-Frequency Fields 2.1 Electric and Magnetic Fields from Power Lines and Other Sources Electricity in our homes and workplaces is transmitted over considerable distances from generation sources by transmission and distribution systems. Electricity is transmitted as alternating current (AC) through electric lines that deliver power to our neighborhoods, factories and commercial establishments. The power provided by electric utilities in North America oscillates 60 times per second, i.e., at a frequency of 60 hertz (Hz). Electric fields are the result of voltages applied to electrical conductors and equipment and are quite stable over time. The electric field is expressed in measurement units of volts per meter (V/m) or kilovolts per meter (kv/m); a kilovolt per meter is equal to 1,000 V/m. Most objects including fences, shrubbery, and buildings easily block electric fields. Therefore, certain appliances and the wiring within homes and the workplace are the major sources of electric fields indoors, while power lines are the major sources of electric fields outdoors. Magnetic fields are produced by the flow of electric currents and therefore vary over time. However, unlike electric fields, most materials do not readily block magnetic fields. The level of the magnetic field is commonly expressed as magnetic flux density in units called gauss, or in milligauss (mg), where 1 G = 1,000 mg 1. The level of the magnetic field at any point depends on characteristics of the source, including the arrangement of conductors, the amount of current flow through the source, and its distance from the point of measurement. The intensity of both electric and magnetic fields diminishes with increasing distance from the source. In most of our homes, background AC magnetic field levels average about 1 mg, even when not near a particular source such as an appliance. Higher magnetic field levels are measured in the vicinity of distribution lines, subtransmission lines, transmission lines and appliances (Figure 2). 1 Scientists more commonly refer to magnetic flux density at these levels in units of microtesla (µt). Magnetic flux density in milligauss units can be converted to µt by dividing by 10, i.e., 1 milligauss = 0.1 µt. 3

11 Figure 2. Magnetic field levels in the environment Source: Savitz et al, 1989 The strongest sources of AC magnetic fields that we encounter indoors in residential settings are electrical appliances (fields near appliances vary over a wide range, from a fraction of a milligauss to a thousand milligauss or more). For example, Gauger (1985) reports the maximum AC magnetic field at three centimeters from a sampling of appliances as 3,000 mg (can opener), 2,000 mg (hair dryer), 5 mg (oven), and 0.7 mg (refrigerator). Similar measurements have shown that there is a tremendous variability in the fields generated among appliances made by different manufacturers. 2.2 Magnetic Fields Encountered in Everyday Environments Considering magnetic fields from a perspective of specific sources or environments, as in Figure 2, does not fully describe the variations in a person s personal exposure encountered in everyday life. To illustrate this, magnetic field measurements that were recorded by a meter worn at the waist while going about daily activities in Bethel, Connecticut for two hours are shown in Figure 3. Activities included a visit to the post-office, the library, walking along the street, getting ice cream, browsing in the bicycle shop, stopping in the chocolate shop, going to the bank/atm, driving along streets, shopping in a supermarket, stopping for gas, and getting something to eat at a fast food restaurant. 4

12 Post Office Library Ice Cream Parlor Bicycle Shop Chocolate Shop Bank - ATM << Supermarket >> < Gas > < Fast Food > Post Office Library Greenwood Ave Town Clock Ice Cream Parlor Greenwood Ave Window Shopping Bicycle Shop Chocolate Shop Greenwood Ave Town Clock Greenwood Ave Window Shopping Driving - Greenwood Ave < Gas > Bank - ATM Driving - Greenwood Ave << Supermarket >> < Fast Food > Figure 3. Typical magnetic field personal exposures (the same data is plotted in the two graphs; the lower graph expands the vertical axis) This figure shows that we encounter magnetic fields whose intensity varies over a wide range from moment to moment in everyday life. The maximum average and median magnetic field levels encountered are listed in Table 1 below. 5

13 Table 1. Summary of magnetic fields measured in a Connecticut town (Bethel) Magnetic Field Levels (milligauss, mg) Median Average Maximum * *Maximum occurred in the supermarket. Still higher levels can occur in other community and residential settings. 2.3 Project Sources of Electric and Magnetic Fields Existing sources of magnetic fields along any potential route between the substations will include transmission and distribution lines and currents flowing on other conductors of electricity, such as telephone and cable television cables and water pipes. The major sources of power-frequency magnetic fields associated with the project will be the cable systems beneath town streets and the transformers and other equipment within the associated substations. 2.4 Magnetic Fields Measured Along Selected Routes Existing ambient levels of magnetic fields were surveyed along the routes described below on November 25, The magnetic field was measured primarily along the right hand lane of the roads along the Primary route between Glenbrook and Norwalk substations, and also along roads from Glenbrook Substation back to Norwalk Substation along the Northern Route Alternative Method for Measuring Magnetic Fields Measurements were taken at a height of one meter (3.28 feet) above ground in accordance with the industry standard protocol for taking measurements near power lines (IEEE Std b). Magnetic fields were expressed as the total field computed as the resultant of field vectors measured in the x, y, and z-axes (rms 2 ). The magnetic field was measured in units of milligauss in x, y and z-axes by orthogonally mounted sensing coils whose output was logged by a digital recording meter (Dexsil Corp) at one-second time intervals while driving through town streets with the meter suspended three feet behind the vehicle. The Dexsil meter meets the IEEE instrumentation standard for obtaining valid and accurate field measurements at power line frequencies (IEEE Std a). The meter was calibrated by the manufacturer by methods like those described in IEEE Std b. It is important to remember that measurements of the magnetic field present a snapshot of the conditions at a point in time. Within a day, or over the course of months, and even seasons, the 2 Root-mean-square (rms) refers to the common mathematical method of defining the effective voltage, current, or field of an AC system. 6

14 magnetic field can change depending upon the amount and the patterns of power demand within the state and surrounding region Primary Route The magnetic field was measured along a route that covers almost the entire Primary route (Figure 4). The route of measurements begins at Route 123 and Riverside Avenue under some AC overhead power lines out of Norwalk Substation. The route then proceeds approximately south along Riverside Avenue changing to Van Buren Avenue and then to Connecticut Avenue traveling in a southwesterly direction, where it becomes Boston Post Road. The route continues along Boston Post Road to West Avenue and then along West Avenue. West Avenue becomes Maple Tree Avenue and then the route crosses Courtland Avenue and drops south on Taylor Reed Place to Crescent Street. The route heads west on Crescent Street to Glenbrook Road, then south on Glenbrook Road to Hamilton Avenue, where it continues east to the north end of the substation. Electric distribution lines are present along almost the entire route and are the dominant source of the measured magnetic fields. The route also crosses and parallels the Amtrak rail line, which is also a source of magnetic fields Total RMS Magnetic Field Magnetic Field - mg Norwalk Substation Riverside Van Buren < > <---> < Connecticut > <- Boston Post -> <----- West Avenue -----> Crescent Glen Brook Hamilton <------> <----> <----> Glen Brook Substation 10 0 Figure Time - seconds Magnetic field recorded at one-second intervals while driving along Primary southern route 7

15 A statistical summary of the magnetic field levels recorded along the Primary route is provided below in Table 2. Table 2. Summary of magnetic field levels (mg) measured along Primary route Alternative Route Magnetic Field Level (milligauss, mg) 5 th Percentile Average Median 95 th Percentile Maximum The route along which measurements were taken closely follows the Northern Alternative Route described in the Application, extending west from the Glenbrook Substation along Hamilton Avenue, north on Glenbrook Road to Church Street, east on Church Street continuing to Route 106, north on Route 106 to Farm Road, east on Farm Road, north on White Oak Shade Road, then east on Old Norwalk Road to Route 123, and finally southeast on Route 123 to the Norwalk Substation (Figure 5). As for the primary route, electric distribution lines along streets are the dominant source of magnetic fields. The route also crosses the MetroNorth rail line several times, which is also a source of magnetic fields Total RMS Magnetic Field Magnetic Field - mg Glen Brook Substation Hamilton Old Stamford Old Norwalk <-Glenbrook-> < > < > <--> < Hoyt > <- Farm Road -> New Norwalk / New Canaan < > Norwalk Substation 0 Figure Time - seconds Magnetic field recorded at one-second intervals while driving along Alternative route 8

16 A statistical summary of the magnetic field levels recorded along the Alternative route is provided below in Table 3. Table 3. Summary of magnetic field levels measured along Alternative route Magnetic Field Level (milligauss, mg) 5 th Percentile Median Average 95 th Percentile Maximum Calculated Magnetic Fields from Underground XLPE and HPFF Cable Systems Methods for Calculating Fields from Underground Cable Systems PDC performed calculations of the magnetic fields that will be produced above ground by two different types of underground cable systems under consideration. The magnetic field values produced by the cable systems are a linear function of the currents flowing in the cables XLPE Cable System (345-kV compatible) The first type of underground cable system considered will use XLPE cables rated at 115 kv. The cables will be installed in concrete-encased eight-inch duct banks as shown in Figure 6. A spare set of ducts is included for potential future use. The following assumptions were made concerning installation and operating conditions for the 115-kV XLPE underground cable system lines shown in Figure The 115-kV XLPE cables have a copper conductor with a conductor size of 3,500 kcmil. 2. The relative phase placements of the cables are shown in Figure The currents flowing in the 115-kV underground cable system will be balanced three-phase currents (i.e., the zero sequence current would be neglibible) and the total cable system load divided equally on the two cable sets. PDC s computer program, CableZ, was used to model the magnetic field produced by the underground transmission cables. This program is based on the Biot-Savart law of fundamental magnetic field theory implemented as described in the Electric Power Research Institute Underground Transmission Systems Reference Book (EPRI, 2002). The accuracy of magnetic field calculations made by CableZ has been verified by field measurements performed in the vicinity of 138-kV XLPE cable systems in Hawaii and Maryland. 9

17 20-60 IN IN. COMMUNICATIONS DUCTS AND GROUND CONTINUITY CONDUCTORS 11.9 IN. A C IN IN. B B 39 IN. 8-IN. CABLE DUCTS C A Figure IN. Trench cross section of XLPE cable system with two cables per phase 115-kV XLPE cable installed Alternative XLPE Cable System in Duct Bank (not 345-kV compatible) An alternative XLPE design that would be more compact but could not be upgraded to 345 kv is shown in Figure 7. PAVEMENT, IF PRESENT IN. LOW STRENGTH CONCRETE, OR FILL HIGH-STRENGTH CONCRETE B A C IN. 23 IN. GCC C A B COMMUNICATIONS DUCTS Figure IN. Typical trench cross-section, 115-kV cables in 2x3 ductbank 10

18 High-Pressure Fluid-Filled (HPFF) Cable System The 115-kV HPFF underground cable system could also be constructed with two, three-phase HPFF transmission circuits in parallel and the total cable system load divided equally between them. The 115-kV HPFF cables will be manufactured with copper conductors and Laminated Paper Polypropylene (LPP) high voltage insulation. Three of the HPFF cables will be installed in each of the two eight-inch carbon steel pipes. The horizontal separation between the centerlines of the two pipes will be 24 inches. The depth from the surface of the ground to the centerlines of the eight-inch pipes will vary between 38 and 78 inches along the alignment of the cable trench depending on local construction conditions (Figure 8). PAVEMENT, IF PRESENT IN IN. THERMAL SAND OR FLUIDIZED THERMAL BACKFILL IN. 24 IN. 28 IN. CABLE PIPES COMMUNICATIONS DUCTS Figure IN. Trench cross section of HPFF cable system with two cables CableZ was used for the HPFF cable calculations. This program is based on the Biot-Savart law of fundamental magnetic field theory implemented as described in the Electric Power Research Institute Underground Transmission Systems Reference Book (EPRI, 2002). The accuracy of magnetic field calculations made by CableZ has been verified by field measurements performed in the vicinity of 138-kV XLPE cable systems in Hawaii and Maryland. The following assumptions were made concerning installation and operating conditions for the 115-kV HPFF underground cable system lines shown in Figure The HPFF cables will have a copper conductor with a conductor size of 2,500 kcmil and inches of PPP insulation (345-kV construction). 2. The copper conductor 345-kV rated HPFF cables will be installed in an eightinch carbon steel pipe with a wall thickness of one-quarter inch (Schedule 20 pipe). 11

19 3. The magnetic properties of the carbon steel pipes vary depending on carbon content as well as the manufacturing process. The magnetic field calculations assume typical magnetic properties for the carbon steel pipes. 4. The currents flowing in the 115-kV underground cable system will be balanced three-phase currents (i.e. the zero sequence current would be negligible) and the total cable system load divided equally on the two cables Calculated Magnetic Field Values for XLPE and HPFF Cable Systems XLPE Cable System in Duct Bank (345-kV compatible) The calculated magnetic field, as a function of distance from the centerline of the cable trench, is shown in Figure 9. Magnetic field values at 0, 25, and 50 feet from the center of the trench are also shown in Table 4. The magnetic field values shown in Figure 9 are the resultant magnetic field that would be measured with a conventional three-axis gauss meter at one meter above ground level. The total cable system load is assumed to divide equally between two cable sets. The phasing of the two sets of cables has been optimized to reduce the magnetic field. This method can minimize the magnetic field with minimal additional cost Magnetic Flux Density - mg MVA MVA Distance from Centerline - feet Figure 9. Magnetic flux density one meter above ground level, 115-kV XLPE Norwalk-Glenbrook underground cable system at normal loading (204 MVA) and peak loading (342 MVA) 12

20 Table 4. Magnetic field values (mg) at one meter above ground 115-kV XLPE cables in duct bank Norwalk-Glenbrook Line XLPE Cable Circuit MVA Horizontal Distance = 0 ft Horizontal Distance = 25 ft Horizontal Distance = 50 ft Normal Load Case Peak Normal Load Case Alternative XLPE Cable System in Duct Bank (not 345-kV compatible) An alternative XLPE cable system is also under consideration that would be more compact and could not be upgraded to 345-kV specifications in the future. The calculated magnetic field from this system design is shown in Figure 10 and Table Magnetic Field (mg) MVA MVA Distance from Centerline (ft) Figure 10. Magnetic flux density one meter above ground level, Alternative 115-kV XLPE Norwalk-Glenbrook underground cable system at normal loading (204 MVA) and peak loading (342 MVA) Table 5. Magnetic field values (mg) at one meter above ground 115-kV XLPE cables in duct bank Norwalk-Glenbrook Line Alternative XLPE Cable Circuit MVA Horizontal Distance = 0 ft Horizontal Distance = 25 ft Horizontal Distance = 50 ft Normal Load Case Peak Normal Load Case

21 HPFF Cable System The calculated magnetic field, as a function of distance from the centerline of the cable trench, is shown in Figure 11. Magnetic field values at 0, 25, and 50 feet from the center of the trench are also shown in Table 6. The magnetic flux density values shown in Figure 11 are the resultant magnetic field that would be measured with a conventional three-axis gauss meter at one meter above ground level. The carbon steel pipe significantly attenuates the magnetic field produced by the current flowing in the 2,500-kcmil cables inside of the pipe. The magnetic field attenuation increases with the magnitude of the current in the HPFF cables because of the nonlinear magnetic properties of the steel pipe Magnetic Field (mg) MVA 342 MVA 0 Figure Distance from Centerline (ft) Magnetic flux density one meter above ground level, 115-kV HPFF Norwalk-Glenbrook underground cable system at normal loading (204 MVA) and peak loading (342 MVA) Table 6. Magnetic field values (mg) at one meter above ground 115 kv HPFF cable circuits Norwalk-Glenbrook Line 115-kV HPFF Cable Circuit MVA Horizontal Distance = 0 ft Horizontal Distance = 25 ft Horizontal Distance = 50 ft Normal Load Case Peak Normal Load Case

22 XPLE Cables in Single Conduit The Glenbrook Cables Project will cross nine rivers and streams, ranging from approximately feet wide (Table J-1, Application). For these crossings, special methods will be required for the construction of the underground XPLE cable system. PDC and Burns and McDonnell have developed a design for trenchless installation. The trenchless construction consists of installing the two 115-kV XLPE cable circuits in six-inch PE ducts within a common steel casing and bore. The casing is installed into the bore using jack and bore techniques. The bore arrangement for a single duct is shown in Figure 12. In some locations it may be necessary have the circuits carried in two bores located some distance apart. The need for site-specific designs for trenchless installation is still being studied 3. A table of the magnetic field values at 0, 25, and 50 feet from the trench centerline is provided in Table 7. ~ 10 FEET FOR JACK & BORE B GCC C A A SPARE B C 6-INCH DUCTS 30 INCH CASING ~ 44-INCH BORE Figure 12. Trench cross section of XPLE cables in single bore 3 It should be noted that the ampacity ratings per circuit may be appreciably lower for two cables in a common bore than compared to the typical duct bank arrangements. The reduced spacing between cables, deeper burial depth, poor thermal properties of the drilling fluid, and loss of a thermal backfill envelope will tend to reduce the ratings of the cables. The power-carrying capacity of this design in duct banks should be calculated for the trenchless installation to be sure that all ampacity requirements are met. It may be feasible to install a larger conductor in cable sections that have trenchless installations to bring the ampacities close to those for trenched sections 15

23 Table 7. Magnetic field values (mg) at one meter above ground XLPE cables in single conduit Loading Scenario MVA Horizontal Distance = 0 ft Horizontal Distance = 25 ft Horizontal Distance = 50 ft Normal Load Peak Normal Load The magnetic field level for the bore arrangement is lowered by the mutual cancellation of the fields of each conductor and by the steel casing. The steel casing alone reduces the magnetic field in the order of 8:1 to 10:1 compared to a similar installation without a steel casing. This reduction is similar to the effect of the steel pipe in a HPFF cable systems. The same calculation assumptions used for the XPLE installations were used for the trenchless installation. The calculation procedure was modified to account for the steel casing. The values calculated by CableZ were then reduced by a factor of 8:1 to model the effect of the steel casing. This reduction is in line with field measurements, which demonstrate a reduction of the magnetic field levels in the order of 8:1 to 10:1 compared to similar installations without a steel casing. The calculated magnetic flux density values are the RMS values that would be measured by most three-axis Gauss meters. The calculations assume that balanced three-phase currents are present in the three high voltage cables and typical magnetic properties for the carbon steel casing Anticipated Magnetic Fields at Public Facilities Public facilities categorically described in the Connecticut Siting Council Application guidelines are being identified. Measurements of existing magnetic fields in the street in front of those identified at this time along the Primary route ranged from 0.2 to 13.7 mg. Based on the expected distance of the underground cable circuits to these facilities, the calculated fields for the preferred cable configuration (Figure 6) would be less than 2 mg. Additional information will be provided in the final Application. 16

24 3 EMF and Health Research on electric and magnetic fields (EMF) in residential settings and health was prompted by an epidemiology study of children exposed to EMF, mostly from neighborhood distribution lines in the U.S. (Wertheimer and Leeper, 1979). The results of subsequent studies are mixed: some studies have reported associations between estimated exposures to magnetic fields and childhood cancers but others have not. The purpose of this report is to describe methods scientists use to evaluate the scientific research (Section 3.1) and provide an up-to-date assessment of the current epidemiologic and experimental research on EMF relevant to power lines (Section 3.2). The report also presents conclusions of large multidisciplinary groups who have reviewed the body of EMF research (Section 3.3) and other recent perspectives (Section 3.4). 3.1 Assessment of Potential EMF Health Risk Science is more than a collection of facts, it a method of obtaining information and of reasoning to ensure that the information is accurate and correctly describes physical and biological phenomena. Many limitations and misconceptions in human reasoning occur when people casually observe and interpret their observations and experience. Therefore, scientists use systematic methods to evaluate scientific research and assess the potential impact of a specific agent on human health. This process is designed to ensure that more weight is given to those studies of better quality, and to ensure that studies with a given result are not selected out from all of the studies available to advocate or suppress a preconceived idea of an adverse effect. These methods include an assessment of the kind of effect that can be caused by exposure (qualitative assessment), as well as a quantitative assessment of the levels of exposure that can produce these effects. These methods have been used to assess potential health effects of electric and magnetic fields (EMF). The following review has been prepared to update the Connecticut Siting Council (CSC) on the status of recent scientific research regarding the potential for health effects of exposure to EMF Methods for Evaluating Scientific Research Scientific methods for assessing potential risk to human health from environmental exposures have been defined by several agencies. These methods are described as a weight-of-evidence approach and are used by scientists, scientific organizations, and regulatory agencies worldwide. These agencies include: the International Agency for Research on Cancer (IARC), which routinely evaluates substances such as drugs, chemicals, physical agents and occupational exposures for their ability to cause cancer; the World Health Organization s (WHO) International Programme for Chemical Safety; and the U.S. Environmental Protection Agency (USEPA), which determines whether exposure to an agent can increase cancer or other adverse health effects and regulates public exposure (WHO, 1994; USEPA, 1993; USEPA, 1996). The weight-of-evidence approach is based on a comprehensive assessment of all of the relevant scientific research, which includes epidemiologic and experimental studies of humans, 17

25 experimental studies in animals (in vivo) and experimental studies in isolated cells and tissues (in vitro) Epidemiology and Experimental Studies are Complementary To assess the potential health effects from any exposure, data from several types of studies must be critically evaluated. These include non-experimental, epidemiologic observations of people, and experimental studies on animals, humans, and tissues in laboratory settings. Epidemiology is the science that studies the patterns of health and disease in human populations. The objective of environmental epidemiology is to quantify and evaluate the associations between exposures to environmental factors (e.g., vegetables in the diet) and health outcomes (e.g., coronary artery disease). Epidemiology studies are non-experimental; they observe people in their normal daily life. In contrast to epidemiology investigations, the experimental studies in controlled laboratory conditions are designed to isolate the subjects (human volunteers, animals, tissues, cells, or molecules) from environmental fluctuations, so that the effects of one variable can be studied in relative isolation. Epidemiologic studies can help suggest risk factors that may contribute to a disease risk, but they usually cannot be used as the sole basis for inferences about cause-and-effect relationships, and they usually only provide information on a limited range of exposures. Experimental studies designed to test specific hypotheses under controlled conditions are generally required to establish cause-and-effect relationships. Conversely, the results of experimental studies by themselves may not always be directly extrapolated to human populations. It is therefore both necessary and desirable that biological responses to agents that could present a potential health threat be explored by epidemiologic methods in human populations, as well as by experimental studies in the research laboratory. Toxicology is an important part of laboratory research designed to evaluate the potential beneficial or harmful effects of an agent (e.g., a chemical or a magnetic field). The goal of toxicology studies is to identify the nature of effects that result from exposure and the dose of the agent in the target tissue that elicits that effect. A most critical distinction, therefore, must be made between harmless biological responses or effects, and those that are truly adverse or deleterious. Many agents produce biological responses in organisms like the response of the eye to light or the influence of food and water on growth and cellular metabolism at quite low concentrations or intensities. Hence, the mere demonstration of a biological response or effect does not indicate that an exposure to an agent is hazardous per se. Rather, it is imperative to ascertain whether biological responses are deleterious or merely normal, and to establish what, if any, exposure concentrations may be toxic and under what conditions Evaluating Epidemiology In studies of leukemia and other cancers in children, the epidemiologists tested for an association between magnetic fields and disease by comparing the exposures of individuals with the disease (cases) to other individuals without the disease (controls). If the exposures of the cases are not higher or lower than those of the controls, then there is no association. While the 18

26 presence of a statistical association does not mean that that the exposure caused (or prevented) the disease (i.e., a cause and effect relationship), it can assist in identifying associations that should be investigated in further research by a variety of methods. To evaluate whether epidemiologic studies suggest an association, it is first necessary to evaluate whether the observed association is likely to be real or whether it is possible that a spurious association was produced due to bias or confounding. Bias refers to any systematic error in the design, implementation or analysis of a study that results in a mistaken estimate of an exposure s effect on the risk of disease. A confounder is something that is related to both the disease (or condition) under study and the exposure of interest 4. If care is not taken to minimize bias and confounding by the design and analysis of the study, these factors can distort the study s results. For studies of magnetic fields and leukemia additional studies were designed to improve exposure assessment and minimize bias and confounders. Still, the effects of bias and confounding have not been fully eliminated from EMF studies (Greenland et al., 2000; IARC, 2002). Conclusions about cause and effect cannot be reached until the totality of the evidence is taken into account. In order to support a cause and effect relationship, the data must present a logically coherent and consistent picture. The process of evaluating statistical associations in epidemiological data is generally guided by a set of criteria outlined by the Surgeon General of the U.S. in an early health report (USDHEW, 1964). They are often called Hill s criteria after the British physician who originally outlined them. These criteria are not met with a simple yes or no ; rather, they serve as guidance for weighing the evidence to reach a decision about cause and effect. The more firmly these criteria are met by the data, the more convincing the evidence. The more important criteria in this guidance are listed in Table 8. These criteria are discussed in epidemiology textbooks as factors to consider in making a judgment about causation. These criteria indicate associations that: are strong, exhibit an increase in risk with increasing exposure, are consistently reported among studies, show coherence with existing information, are biologically plausible, and are more likely to be causal. 4 For example, a link between coffee drinking in mothers and low birth weight babies has been reported in the past. To evaluate this question properly, we must not only know the amount of coffee the mothers drank, but also their exposures to confounding factors like cigarette smoking. Some women who drink coffee also smoke cigarettes. In this example, it was found that when the smoking habits of the mothers are taken into account, coffee drinking was not associated with low birth weight babies (Kelsey et al., 1996). 19

27 Table 8. Hill s criteria for causation Strength Consistency Dose-response Biological plausibility Temporality The stronger the associations between the disease and the exposure or characteristic in question, the more persuasive the evidence. Consistent results across different study populations are more convincing than isolated observations. An increase in exposure is related to an increase in disease if the exposure is associated to the disease. Epidemiologic results are much more convincing if they are coherent with what is known about biology, that is, there must be a biological mechanism that can explain the effect. The data must provide evidence of correct temporality, that is, the exposure must be documented to have occurred before the observed effect, with sufficient time for any induction period related to the disease. 3.2 Weight-of-the-Evidence Evaluation Over the last 30 years, research has been conducted in the U.S. and around the world to examine whether exposures to alternating electric and magnetic fields (EMF) have health or environmental effects. EMF are produced by both natural and man-made sources that surround us in our daily lives. They are found throughout nature and in our own bodies. The earth itself produces a static magnetic field it is this field that is used for compass navigation. Man-made EMF are found wherever electricity is generated, delivered, or used. Power lines, wiring in homes, workplace equipment, electrical appliances, and motors produce EMF. EMF from such sources changes direction and intensity 60 times, or cycles, per second a frequency of 60 Hertz (Hz). These extremely low frequency (ELF) fields differ significantly from fields at the higher frequencies characteristic of radio and television signals, microwaves from ovens, cellular phones, and radar (which can have frequencies up to billions of Hz). Using the weight-of-the-evidence approach described in the previous section, both epidemiology and laboratory studies as well as scientific reviews of large multidisciplinary groups through 2003 were reviewed Epidemiology Studies In some epidemiology studies over the past 30 years, electric and magnetic fields from a variety of electrical sources, including power lines, have been linked to cancer and other illnesses. The most comprehensive review of epidemiology studies and other data was performed by scientists in a Working Group drawn from 10 countries on behalf of the International Agency for Research in Cancer (IARC), a division of the World Health Organization (2002). The conclusion of the 395-page report was that the strength of the scientific evidence was inadequate to conclude that there is a statistical association between EMF and risk of any cancer, with the exception of childhood leukemia, and then only at higher levels of exposure (greater than 4 mg). No other diseases were identified as linked to EMF. As explained above, a statistical association is not the same as a causal relationship. 20

28 Childhood Cancer Epidemiology studies with small numbers of children reported a statistical association between magnetic fields and leukemia. In those studies, more of the children who had cancer lived closer to certain types of power lines, or were exposed to higher estimated magnetic fields (e.g., Savitz et al., 1988; Wertheimer and Leeper, 1979; Feychting and Ahlbom, 1993) than other children without cancer. However, because of limitations of these specific studies, such as the small numbers of children, the meaning of these results was not clear. So larger and better studies were undertaken. The investigators who conducted the newer studies did not find convincing, consistent links between power lines and leukemia, even when children had been exposed to higher levels of magnetic fields (e.g., Linet et al., 1997; McBride et al., 1999; Kleinerman et al., 2000; UK Childhood Study Investigators, 1999, 2000). Summaries of the two largest studies of childhood leukemia are provided below. Both groups of investigators concluded that their data provided little evidence for an association of magnetic fields with leukemia in children. National Cancer Institute The National Cancer Institute (NCI) completed a large and comprehensive study of childhood leukemia in the U.S. in This study compared exposure to magnetic fields in children who did not have cancer to the exposure of those who had acute lymphocytic leukemia (ALL), the most common form of leukemia in children (Linet et al., 1997). The major advantage of this study was the short time between exposure assessment and diagnosis compared to previous studies, and the assessment of exposure by a variety of methods. In addition, the investigators obtained magnetic field measurements from multiple rooms in each child s home, which included magnetic field exposures from household appliances. No association was found between ALL and the wiring configuration code at the residences occupied by the children before they had cancer. The researchers observed a statistical association between leukemia and magnetic field levels in the category mg, but not for exposures less than 4 mg or for exposures greater than or equal to 5 mg, the highest exposure category. There was no overall trend for a stronger association with increased exposure. Further analyses indicated that distance from high-voltage lines and an improved exposure index was not related to risk for ALL (Kleinerman et al., 2000). United Kingdom Childhood Cancer Study The largest childhood cancer study to date was completed in the United Kingdom (U.K.) in The United Kingdom Childhood Cancer Study (UKCCS) investigators reported on magnetic field measurements on a portion of the cases and controls evaluated in a previous study (UKCCS, 1999). To obtain additional information, they used a method to assess exposure to magnetic fields without entering homes (UKCCS, 2000) and were able to analyze 50% more subjects (a total of 3,380 cases and 3,390 controls). For all these children, they measured distances to power lines and substations. This information, combined with data 21

29 on historical current flow, was used to calculate the magnetic field from these external field sources, based on power line characteristics related to production of magnetic fields. The results of the second UKCCS study showed no evidence for an association with leukemia for magnetic fields calculated to be between 1 mg 2 mg, 2 mg 4 mg, or 4 mg or greater at the residence, which is consistent with the results of the earlier report in which magnetic field exposure was estimated by measurement (UKCCS, 1999). Children with leukemia were also reported not to be more likely to live near distribution or high-voltage power lines than control children. Nevertheless, when the data from the more recent studies cited above and a larger number of earlier studies are pooled, a weak association of leukemia with estimates of long-term average magnetic exposures greater than 3 4 mg is reported (Ahlbom et al., 2000; Greenland et al., 2000). These magnetic field values are interpreted as descriptors of long-term average exposure Studies of Electric Field Exposures Assessing electric field exposures is more difficult than for magnetic field exposures because electric fields are easily blocked by objects. A few epidemiology studies of children, however, have focused on exposures to electric fields from cable systems and electrical appliances. Childhood cancer was not found to be associated with electric fields regardless of how electric fields were measured (Savitz, 1988; London, 1991; Coghill et al., 1996; Dockerty et al., 1998; McBride et al., 1999; Green et al., 1999a). The UKCCS recently re-evaluated a subset of subjects from their 2000 study in which magnetic and electric field exposures were measured simultaneously at the residence. Measurements, that were verified, were recorded for 549 subjects (273 cases, 276 controls). No reliable associations between measured electric field exposure and all cancers, leukemia, or central nervous system cancers were reported (UKCCS, 2002) Research Related to Adult Cancer and Reproduction Research Related to Adult Cancer Occupational studies have varied greatly in the methods used to estimate exposure (e.g., type of industry, exposure based only on job titles, direct electric and magnetic field measurements), study design (e.g., retrospective cohort studies based on death records, case-control studies with direct magnetic field measurements) and source of exposure to EMF (e.g., specific occupations [i.e., railway workers, electricity generation and transmission industry] or multiple industries). Recent studies have greatly improved estimates of EMF exposures. Occupational studies published through 2002 are described in IARC (2002). No consistent relationship between occupational exposures to magnetic or electric fields has been found for any type of cancer in adults, including leukemia and types of cancer affecting the brain and breast. 22

30 Research Related to Reproduction Several epidemiology studies have examined effects of exposures to magnetic fields on pregnancy, including miscarriages (spontaneous abortion). They reported no association with birth weight or fetal growth retardation after exposure to sources of relatively strong magnetic fields such as electric blankets, or sources of typically weaker magnetic fields such as power lines (Bracken et al., 1995; Belanger et al., 1998; Lee et al., 2000). Two recent studies of EMF and miscarriage reported a positive association between miscarriage and exposure to high maximum, or instantaneous, peak magnetic fields (Li et al., 2002; Lee et al., 2002). However, no reliable associations were found with higher average magnetic field levels during the day, the typical way of assessing exposure. Neither study found that miscarriage was associated with residential wiring codes, another method presumed to identify higher magnetic fields from power lines. There are several possible issues to be considered in assessing whether these statistical associations with the maximum magnetic field exposure during the day are possibly causal in nature. First, the studies include possible biases. For example, each of the studies had a low response rate, which means that the study groups may not be comparable because those who participated in the studies may have differed from those who declined (selection bias). Second, these studies found no reliable association with higher daily average exposure, that is, the average of the measurements recorded throughout the day. Third, despite years of research, there is no biological basis to indicate that EMF increases the risk of miscarriage Experimental Studies A wide variety of methods are available for assessing possible harm or toxicity associated with exposures to EMF in experimental studies. There are two types of experimental studies: studies of human volunteers or whole animals called in vivo studies, and studies of isolated cells and tissues (obtained from human or animal sources), called in vitro studies. Studies in which laboratory animals (in vivo) receive high exposures provide an important basis for evaluating the safety of environmental exposures, chemicals and medicines. From a public health perspective, chronic (long-term) bioassay studies in which animals are exposed over most of their entire lifetime are of particular value in assessing potential risks of cancer and chronic disease and are widely used by health agencies to assess health risks to humans from medicines, chemicals and physical agents. Clinical studies of humans exposed to EMF have evaluated short-term effects of exposures. Adverse health effects have not been reported at exposures many times higher than those from cable systems (IARC, 2002). The value of in vitro tests to human health risk assessment is more problematic. Responses of cells and tissues outside the body may not reflect the response of those same cells in a living system and so their relevance cannot be assumed (IARC, 1992). The mechanism underlying effects observed in vitro may not correspond to mechanisms underlying complex processes like carcinogenesis (able to produce cancer). It may be difficult to extrapolate from simple cellular systems to complex, higher organisms to predict risks to health. In addition, the results of in vitro studies cannot be interpreted in terms of potential human health risks unless they are performed in a well-studied and validated test system. Under such conditions, in vitro studies 23