- Email: [email protected]
Implementation of shielding principles for magnetic field management of power cables
Electric Power Systems Research 48 (1999) 193 – 209
Implementation of shielding principles for magnetic field management of power cables A.S. Farag *, M.M. Dawoud, I.O. Habiballah King Fahd Uni6ersity of Petroleum and Minerals, KFUPM Box 371, Dhahran 31261, Saudi Arabia Received 13 November 1997; received in revised form 11 May 1998; accepted 5 June 1998
Abstract Adverse health effects due to magnetic fields is a matter of great concern and has been widely debated in recent years. Managing these fields is a challenge to researchers. One of the important sources of magnetic fields is power cables. Different management techniques have been studied. In this paper, passive shielding schemes are implemented to manage the magnetic fields of power cables. The various shielding schemes can be separated into two broad categories: shielding subject and shielding source. Both schemes are implemented in this paper. Passive shielding schemes are found to be the most powerful technique as the reductions obtained are sometimes as high as 97–98%. This scheme is a costly one, should be used only in assigned locations, and as such, there has to be a trade-off between the cost and the level for reduction desired and the health desired and the health hazards. © 1999 Elsevier Science S.A. All rights reserved. Keywords: Magnetic field management; Power cables; Passive shielding
1. Introduction Questions about effects from exposure to power systems AC electric and magnetic fields are of continuing interest to utilities, regulators, consumers and power equipment suppliers. The possible effect of low frequency electromagnetic field (EMF) exposures in occupational and residential environments raises the questions of how electromagnetic fields are created and what effects they may have. There is substantial scientific uncertainty and no widespread agreement among scientists as to how the presently available information regarding the possible health effects of magnetic fields should be interpreted. It is also not clear as to what aspect, if any, of the average magnetic field might be of significance. Those aspects currently being examined include the average field strength, peak field strength, switching of fields, transients, time spent in field, and frequency of field [1 – 6]. There is no doubt that many people are concerned about the magnetic field effects of power frequency electric currents associated with AC transmission and * Corresponding author. Tel.: + 966 3 8602288/3535; fax: + 966 3 8603535; e-mail: [email protected]
distribution underground cables. While health studies are in progress, it appears desirable to conduct parallel technical studies related to the magnetic field management that serve as a guideline to the utilities in practical implementation in occupational areas. Magnetic field management techniques for cables being a recent area of research has not received much of an attention from researchers and utility engineers. The application of some of these techniques to multi-conductor underground lines have not been carried out so far and they are very much a source of concern because of the high currents they carry. Attempts have been made to see the reduction over a distance by the application of some of the management techniques for the case of multi-conductor underground lines [7–14]. In this paper, the implementation of shielding schemes and a comparison between the shielded and unshielded transmission power cables magnetic field values is carried out. A very high reduction is obtained through passive shielding, however, the scheme is quite costly. Different magnetic field profiler showing trapping and field profiles showing trapping and field diversions for different plate geometries are investigated through shielding materials of good conductivity and high relative permeability.
0378-7796/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved. PII S0378-7796(98)00108-4
194
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
2. Factors affecting magnetic fields of power cables There are numerous factors which affect the values of magnetic fields produced by underground transmission cables. These factors may be grouped into the following general areas: 1. System parameters such as current magnitude and phase balance, and system grounding, 2. Cable installation parameters such as depth of burial, installation configuration and relative placement of the cable phases when there is more than one circuit, 3. Cable construction parameters: primarily shield/ sheath resistance and type of material for non-pipetype cables, and 4. External factors such as the presence of nearby underground conductors or other current sources which may flow on the cable sheath/shield or ground continuity conductor.
land availability in the vicinity of the sources. This approach consists of finding ways to avoid the presence of living beings near the location of high magnetic fields by modifying work rules, e.g. limiting access to high magnetic field zones and installations of equipment far away from the areas frequently used by the people.
3.2. Manipulation of source geometry and current This approach is, at present, widely accepted by utilities. Techniques pertaining to his approach are compacting source geometry, modifying circuit currents or characteristics of magnetic materials of facilities and equipment [9–12]. Some of these techniques include: (a) In case of double circuits of cables (two cables per phase), there is a mutual cancellation of magnetic fields. Cables are placed so that they are point symmetric to avoid the induction of zero sequence currents on the two cable circuits. A vertical duct bank installation is probably the most practical installation method to take advantage of magnetic field reduction through mutual
3. Magnetic field management techniques in cables structures Magnetic field management is basically concerned with the minimization of the effect of a magnetic field on the public health front without sacrificing the effectiveness and reliability of the power system. Since magnetic field sources in power systems have been identified, the interest of utilities have now shifted to magnetic field management of these sources. The application of magnetic field exposure reduction techniques to existing facilities and equipment is much more difficult than for new constructions as they have severe constraints which limit the choice of suitable magnetic field reductions and offer a new set of engineering challenges. For each magnetic field source, there may be a number of design options that would reduce its exposure without altering the function job for which the system was intended. Consideration of the factors which significantly affects the magnetic field produced by transmission cables, gives a general approach to the reduction of the magnetic field produced by single-conductor cables. The drawback is that the current carrying capacity of the cable becomes reduced in most of the methods due to the increase in mutual heating between the cables or by an increase in losses. A general classification of the different approaches for magnetic field reduction is given in the following sections [9–12].
3.1. Increased distance between the sources and point of interest This approach is simple and straightforward provided there are no physical constraints, such as space or
Fig. 1. Shielding a subject.
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
195
Table 1 Maximum value of magnetic field for two cables per phase (single phase cables) Cable diameter (in.) Plate thickness (in.)
1.07 1.07 1.55 1.55 1.89 1.89
0.005 0.030 0.002 0.030 0.005 0.030
Derated current (Amp) Without shielding
87.5 87.5 190 190 272.5 272.5
With shielding
Stack
Triangular (mG)
Flat
Stack
Triangular (mG)
Flat
8.74 8.74 27.14 27.14 47.04 47.04
4.40 4.40 13.71 13.71 23.72 23.72
0.901 0.901 4.07 4.07 8.62 8.62
0.48 0.50 0.67 0.68 0.97 0.98
0.59 0.66 0.98 1.00 1.45 1.47
0.58 0.66 0.48 0.49 0.55 0.56
Table 2 Maximum value of magnetic field for three cables per phase (single phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.030 0.005 0.030 0.005 0.030
Derated current (Amp) Without shielding
87.5 87.5 190 190 272.5 272.5
With shielding
Stack
Triangular (mG)
Flat
Stack
Triangular (mG)
Flat
4.69 4.69 15.32 15.32 27.54 27.54
6.56 6.56 20.33 20.33 34.92 34.92
13.21 13.21 32.19 32.19 55.96 55.96
0.50 0.54 0.54 0.546 0.68 0.69
0.83 0.95 1.37 1.40 2.03 2.07
0.63 0.74 0.87 0.88 1.35 1.37
Table 3 Maximum value of magnetic field for four cables per phase (single phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.030 0.005 0.030 0.005 0.030
Derated current (Amp) Without shielding
87.5 87.5 190 190 272.5 272.5
cancellation produced by the phase conductors. Horizontal configuration magnetic field reduction is mot as great as vertical configuration. Stacked horizontal configuration would also permit effective magnetic field cancellation between the two circuits. (b) Close triangular configuration results in relatively low magnetic fields but increases mutual heating and therefore reduces the current carrying capacity of the cable. (c) The use of multi-point grounding and a cable with low shield/sheath resistance. This option results in a significant reduction in capacity for a given conductor size. This option is used only when the above two
With shielding
Stack
Triangular (mG)
Flat
Stack
Triangular (mG)
Flat
1.73 1.73 7.69 7.69 16.08 16.08
6.48 6.48 20.08 20.08 34.70 34.70
1.38 1.38 6.23 6.23 13.19 13.19
0.39 0.41 0.52 0.53 0.64 0.65
0.90 0.87 1.13 1.16 1.61 1.65
0.57 0.65 0.51 0.53 0.62 0.64
options do not reduce the magnetic field to acceptable levels. Multi-point grounding makes the most engineering sense for cables installed in a close triangular configuration since the reduction in capacity is much less for horizontal or vertical configurations. If multipoint grounding is used to reduce the magnetic field levels, a shield/sheath resistance lower than that required by fault currents considerations will probably be necessary. (d) Placing any ground continuity conductors also reduces the induced current. The reduction in the magnetic field by optimum placement of the ground conductor is usually small compared to the other options,
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
196
Table 4 Maximum value of magnetic field for five cables per phase (single phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.030 0.005 0.030 0.005 0.030
Derated current (Amp) Without shielding
87.5 87.5 190 190 272.5 272.5
With shielding
Stack
Triangular (mG)
Flat
Stack
Triangular (mG)
Flat
4.69 4.69 15.54 15.54 28.35 28.35
8.93 8.93 26.96 26.96 46.66 46.66
17.56 17.56 14.98 14.98 26.95 26.95
0.82 0.95 0.61 0.63 0.78 0.77
0.52 0.57 1.34 1.38 1.31 1.34
0.82 0.98 0.62 0.64 0.76 0.77
Table 5 Maximum value of magnetic field for six cables per phase (single phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.030 0.005 0.030 0.005 0.030
Derated current (Amp) Without shielding
87.5 87.5 190.0 190.0 272.5 272.5
and hence is to be avoided for significant field reductions. (e) An increase in the depth of burial also reduces the magnetic field directly above the cables. This is effective only in reducing the maximum magnitude of the field. However, increasing the burial depth leads to a decrease in capacity and an increase in the cost of installation.
3.3. Shielding with conducting materials Magnetic field shielding with conducting (including high permeability) materials is a common practice in industry and has received extensive attention recently. Shielding methods may include the use of induced currents, modification of magnetic flux patterns using high permeability and/or high conductivity materials, addition of a second field that tends to reduce the original field, and even a change in technology to eliminate 60 Hz magnetic fields for specific applications. Current shielding designs or implementations are just a practice of experience. The higher the permeability of the material, the better the shielding [11 – 14]. Gaps in shielding can radically affect the flux within the shield. In practice, there is a trade off between the effectiveness of shielding and its type and parameters such as the radius, thickness and size as the radius, thickness and
With shielding
Stack
Triangular (mG)
Flat
Stack
Triangular (mG)
Flat
1.73 1.73 7.69 7.69 16.08 16.08
6.48 6.48 20.08 20.08 34.70 34.70
1.38 1.38 6.23 6.23 13.19 13.19
0.39 0.41 0.52 0.53 0.64 0.65
0.90 0.87 1.13 1.16 1.61 1.65
0.57 0.65 0.51 0.53 0.62 0.64
size. It is very important that the shields remain continuous for maximum effectiveness[11 –14]. The various existing shielding schemes can be separated into two broad categories: (a) shielding a subject; and (b) shielding a source. There are very distinct differences between these two categories. Shielding a subject means implementing a shield of some sort in order to reduce the field in some relatively small, well defined volume, due to sources of field outside the volume. Shielding the source involves placing a shield to reduce the magnetic field in the ‘outside’ world due to some localized source. An example of shielding a subject would be to place a highly permeable material around a subject (Fig. 1 (a)), as the magnetic flux coming from the top takes the path of least reluctance (highest m) rather than going through the high reluctance of the air. It is shunted around the subject (Fig. 1 (b)) where the unperturbed field B0 induces a current to flow in the loop which will produce a field in the opposite direction of B0 and the field at the subject is reduced. It should be noted that there are regions of space in which the magnetic field is increased due to the shield. The shield would need to be designed in such a way as to designed in such a way that those regions are located where they have no effect. Another way of shielding can be carried out with materials of high permeability and/or conductivity.
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
197
3.4. Shielding with current carrying wires A magnetic field can be shielded (i.e. reduced) by establishing currents in wires such that the fields produced by those current oppose the fields to be reduced. The shielding effectiveness is generally a function of the magnitude and the phase of the current. Currents in the wires can be established in two ways. They can be induced by the external source of the magnetic field (passive shielding), or they can imposed by external devices (active shielding).
3.4.1. Passi6e shielding Passive shielding (or shielding with ‘passive’ conductors) refers to the use of currents induced in conductors by existing (or ambient) magnetic fields to reduce (shields) these fields in a certain region. A time varying magnetic field passing through a closed loop will induce a voltage into the loop (Faraday’s law) as: emf= −(F/ (t. If the loop is a conductive wire, the voltage will cause a current to flow. This current will set up its own magnetic field. the net magnetic field in the region will be a superposition of the original field and the field due to the induced current which is produced to oppose the original field. Since the induced voltage onto the loop is proportional to the derivative of the flux through the loop, the voltage will be 90° shift of the current with respect to the voltage, and therefore, a 180° difference will exist between the original field and the induced current fields to oppose each other. It should be noted that in some cases, it is possible to over-compensate the original field and actually cause an increase in the field level over some regions, i.e. regions very close to the wire of the loop. However, these regions are generally very localized, and their locations can be controlled.
Fig. 2. Comparison of magnetic field levels for cables with and without shielding for different configurations.
3.4.2. Acti6e shielding Active shielding (or shielding with ‘active’ conductors) refers to any scheme which reduces the magnetic field in certain regions of space by the use of conductors with an imposed current whose magnitude, direction, and phase angle create fields in opposition to the ambient fields and thereby reduce the overall magnetic field in a region. Currents (magnitude and phase) can be imposed in conductors by electric devices to reduce the magnetic field levels in certain regions. These devices generally operate in two ways. The first involves placing a magnetic field sensor (transducer) in the region to be shielded. The sensor provides feedback to the electric device that adjusts the magnitude and the phase of a current being fed to a wire loop around the region to be shielded. The electric device continually adjusts the current as it attempts to null out the field at the sensor [11,14]. A second method is to manually adjust the magnitude and phase of the current being fed to the shielding loop until the field id nullified. This is a very effective means of
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
198
Table 6 Maximum value of magnetic field for two cables per phase (three phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.03 0.005 0.03 0.005 0.03
Derated current (Amp)
87.5 87.5 190.0 190.0 272.5 272.5
Without shielding
With shielding
Flat
Triangular (mG)
Flat
Triangular (mG)
0.321 4.379 1.456 1.456 2.698 2.698
4.379 0.297 13.604 13.604 23.586 23.586
0.297 0.307 0.529 0.523 0.788 0.759
0.525 0.540 1.111 1.115 2.941 2.647
Table 7 Maximum value of magnetic field for three cables per phase (three phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.03 0.005 0.03 0.005 0.03
Derated current (Amp)
87.5 87.5 190.0 190.0 272.5 272.5
Without shielding
With shielding
Flat
Triangular (mG)
Flat
Triangular (mG)
6.675 6.675 20.762 20.762 35.937 35.937
6.543 6.543 20.242 20.242 34.96 34.96
0.965 1.082 1.233 1.276 1.719 1.754
0.653 0.672 1.452 1.442 3.571 3.621
Table 8 Maximum value of magnetic field for four cables per phase (three phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.03 0.005 0.03 0.005 0.03
Derated current (Amp)
87.5 87.5 190.0 190.0 272.5 272.5
reducing fields, however, an obvious drawback is that the device must be adjusted every time the field changes [11,14].
4. Quantity shield effectiveness In order to evaluate the effectiveness of various shielding schemes, it is necessary to define terms which quantify the degree of shielding and there are several ways to do so. The terms shielding effectiveness and shielding factor (SF) are most often used. Shielding effectiveness is a generic term without a rigorous mathematical definition and is the standard for quantifying shielding effectiveness [10 – 14]. The shielding factor is defined as a ratio of resultant field after shielding to that before shielding (open air), and can be expressed as:
Without shielding
With shielding
Flat
Triangular (mG)
Flat
Triangular (mG)
6.586 6.586 20.525 20.525 35.590 35.590
8.677 8.677 26.689 26.689 45.853 45.853
0.879 0.964 1.266 1.304 1.785 1.816
0.776 0.797 1.712 1.718 4.660 4.215
(1)
SF= B/B0
where B is the rms value of the magnetic field with the shield in place, and B0 is the rms value of the unperturbed magnetic field, i.e. the magnetic field without the shield. Clearly, SF is a function of position and always lies between 0 and 1. The smaller SF is, the more significant the magnetic field reduction. A unitary value of SF means no shielding, and zero SF means perfect shielding sometimes, the shielding factor is called the attenuation factor and is expressed in dB as: SF= 20 log (B/B0)
dB
(2)
Another quantity, which is sometimes used to quantify shielding effectiveness is the shielding efficiency (SE), which is defined as: SE= (1−SF)*100
(3)
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
199
Table 9 Maximum value of magnetic field for five cables per phase (three phrase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.03 0.005 0.03 0.005 0.03
Derated current (Amp) Without shielding With shielding
87.5 87.5 190.0 190.0 272.5 272.5
Flat
Triangular (mG)
Flat
Triangular (mG)
8.831 8.831 27.580 27.580 47.915 47.915
10.772 10.772 24.238 24.238 41.734 41.734
1.203 1.326 1.625 1.677 2.251 2.297
0.895 0.918 1.850 1.750 7.207 6.375
Table 10 Maximum value of magnetic field for six cables per phase (three phrase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.03 0.005 0.03 0.005 0.03
Derated current (Amp) Without shielding With shielding
87.5 87.5 190.0 190.0 272.5 272.5
Another quantity is the field reduction factor which is simply the reciprocal of the shielding factor.
5. Implementation of shielding principles
5.1. Passi6e shielding of the source The principle of passive shielding is applied to the sources which are single phase and three phase cables. Different cases are simulated, using the software package PCFIELD by EPRI, without shielding and after shielding with the same material as the conductor itself. Five different plate thickness of 0.005, 0.003, 0.001, 0.003 and 0.05 in. are used and the field reductions are calculated. However, the results of only two plate thickness (0.005 and 0.03 in.) are reported here as the difference between other thickness is insignificant. Three different cable sizes namely, c 2/0 AWG, 500 MCM, and 1000 MCM, with conductor diameters of 1.07, 1.55 and 1.89 in., respectively were also selected. A 2-in. thickness for insulation is also incorporated. The results obtained show significant reductions in the magnetic field levels.
5.1.1. Single phase cables The results for single phase cables having different configurations and different number of circuits are shown in Tables 1–5. For two cables per phase (Table 1), if considering a cable of 1.89 in. diameter, the field reduces from a level of 47.04 – 0.97 mG and 0.98 (stack
Flat
Triangular (mG)
Flat
Triangular (mG)
11.039 11.039 34.391 34.391 59.611 59.611
12.818 12.818 31.061 31.061 53.581 53.581
1.466 1.640 1.822 1.884 2.487 2.540
1.008 1.003 1.755 1.714 5.430 4.900
configuration) with 0.005 and 0.03 in. shielding plate, respectively. In terms of the shielding factor, it is only 0.0206, which is a very high reduction. For other cable diameters, the reduction is generally of the comparable level. The simulated values with shielding do not necessarily correspond to optimal conditions of simulated valves without shielding. For three cables per phase (Table 2), the shielding factor varies between 0.0025 and 0.144 for the cases considered. Consequently, the field level comes sown to as low as only 2–3% of the simulated value without shielding. The lower the value of the shielding factor, the more effective the shielding is. The reductions are more pronounced for larger sizes of cables. It should be noted that for a thicker plate, the reduction is generally less due to the fact that these plates are reducing the level of the magnetic field over a particular region. Thicker plates generally reflect the field closer to the source compared to thinner plates. The region taken in simulation is − 50–50 ft on either side of the reference axis which is generally the center of the cable configuration. A sample of the plot showing the comparison between the shielded and unshielded case for the stack, triangular and flat layout is illustrated in Fig. 2 for five cables per phase with a cable diameter of 1.55 in.. For four to six cables per phase, the stack and flat configurations have approximately the same levels for magnetic fields for large cable diameters as illustrated in Tables 3–5), while for small diameters, the stack configuration is better. For five cables per phase, the triangular configuration is better than both the stack and flat configurations for small cable diameters.
200
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 3. Comparison of magnetic field levels for cables with and without shielding for different configurations.
5.1.2. Three phase cables For a three phase cable, two types of configurations —the flat and the triangular — are simulated. The chosen configurations do not always correspond to the
phase locations that gives the minimal field. To be precise, with the exception of two cables per phase, all of the configurations of flat arrangements are not the optimal conditions. For triangular, the values are the
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
201
Table 11 Maximum value of magnetic field for four cables per phase (single phrase cables) Cable diameter (in.)
1.89 1.89 1.89 1.89 1.89 1.89
Conductor current (Amp)
272.5 272.5 272.5 272.5 272.5 272.5
Shield current (Amp)
27.25 −27.25 27.25 −27.25 27.25 −27.25
Shield angle (°)
Magnetic field values (mG)
a
b
c
0 0 0 0 0 0
0 0 −120 −120 120 120
0 0 −240 −240 240 240
165.683 167.014 2.524 2.065 2.431 2.211
Fig. 4. Plot of the best and worst case for active shielding for stack configuration (cable diameter 1.89 in., two cables per phase, single phase cable).
minimum possible one could get out of the phase relations. Tables 6– 10 show the values of the magnetic fields before and after shielding for 1.07, 1.55, and 1.89 in. diameter cables. In the case of a double circuit line, Table 6 gives the values before and after shielding and the reduction in the field level varies between 8 and 92%. For three cables per phase, the maximum reduction is 95%. The reduction is more generally in the case of a thinner plate but the difference is not significant. In some of the cases of triangular configuration, the thicker plates give less fields. This is due to a different configuration in the conductor arrangements. For five and six cables per phase, the level of reductions obtained is very high and this leads to the conclusion that the passive shielding scheme of the magnetic field reduction is a very successful technique but at an extra cost. Samples of the plots showing the comparison between the shielded and the unshielded cases for flat and triangular layouts are illustrated in Fig. 3 for five cables per phase of cable diameter of 1.55 in.
5.2. Acti6e shielding of the source The theory of active shielding is still in its early stage and very little work has been done in this area. However, one case of simulation result is shown. The cable size selected is 1000 MCM having a diameter of 1.89 in. The rated current for this size of cable is 550 amps and a 50% derating is applied. The shield current is taken as 10% of the derated conductor current. the simulation is performed for two cables per phase (single phase cable) having stack configuration. The phase location does not correspond here for minimal field. The results are tabulated in Table 11. As evident from the results, the minimum field is obtained when the shield current direction is opposite to the conductor current and have the same phase sequence as the conductor. The value obtained is 2.065 mG, while for the same case when it is not shielded, the field value was found to be 2.295 mG. Hence, there is a reduction of 10%. The plot of the best and worst case is shown in Fig. 4.
202
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 5. Contour map and profile graph for case 1.
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 6. Contour map and profile graph for case 2.
203
204
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 7. Contour map and profile graph for case 3.
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
205
Fig. 8. Contour map and profile graph for case 4.
5.3. Shielding the subject 5.3.1. Simulation methodology The software MAGNETO uses the boundary element method (BEM) for calculation of field distribu-
tions. Basically, the software is used to solve the effects of shielding using different magnetic materials having different geometry of plates. The method adopted is as follows: (a) the limits for different quantities are set; (b) the geometry of the problem is defined (Xmin, Ymin, and
206
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 9. Contour map and profile graph for case 5.
Ymax) beside the grid size and conductors are defined by their coordinates; (c) currents are assigned to the conductors; (d) materials are chosen from the tables including their permeability’s and the materials are
placed in their appropriate regions; and (e) the boundary elements are defined which is crucial to achieve valid results by proper distribution of boundaries.
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
207
Fig. 10. Contour map and profile graph for case 6.
6. Results and discussions Different cases of shielding with different plate ge-
ometry and different materials are simulated. the results of different cases are documented. In case 1, two rectangular plates having a thickness of 0.3 in. and
208
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 11. Contour map and profile graph for case 7.
separated by 1 in. are used for shielding purposes. The plate above the cable is ferrite possessing a permeability of 2000 and is at a distance of 6 in. from the conductors. The other plate is made of aluminum (relative
permeability= 1). The contour map and profile graph are shown in Fig. 5. The ferrite plate practically traps all of the fields and the ends of the plates being stressed more. In case 2, the position of the two plates are
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
interchanged and the field easily passes through the aluminum plate, which carries out practically no shielding and only the ferrite plate helps in trapping the fields (Fig. 6) for the contour map and profile graph. In case 3, two high permeable materials (ferrite) and cast iron (relative permeability =165.96) plates are placed over each other. Ferrite is farther away from the conductor. A Contour map is shown in Fig. 7, where some of the fields are trapped by the lower plate and the rest by the ferrite plate. The contour of x-component of the field are diverted to the surroundings. In case 4, the positions of the two plates are interchanged. Fig. 8 shows that the lower high permeable plate does most of the trapping while the upper plate of cast iron takes care of the rest so that the region behind it is perfectly clear of magnetic fields. Moreover, the field near the conductor surface is not diverted to the surroundings as in case 3. Therefore, it is better to keep the high permeable plate near the conductors in the case of multilayer shielding. Case 5 shoes three levels of plates with an air gap on the middle plate level. The middle plate is made of aluminum, while the top and bottom plates are made up of highly permeable ferrite. The lower plate traps some of the field while the aluminum does little to reduce the field in a particular region. The upper ferrite plate diverts some of these into other regions (Fig. 9). If the positions of the plates are interchanged leading to case 6, the situation is much better compared to case 5, as predicted from Fig. 10. Gaps in shielding greatly affect the field. The ferrite plate, which is at the middle, provides a good shielding to the region of interest. The ends of the plates get further stressed. The lower aluminum plate does nothing to the field and it simply passes through it. In case 7, a single plate of ferrite is used for shielding but with an inverted U-shape. This will shield the conductors from three sides. The contour plot and profile graph are shown in Fig. 11, which demonstrates that the bend points are more stressed, it gives a better shielding performance, and there is not much change in the profiles over the region.
7. Conclusions In this paper, one of the major sources of magnetic fields, which is underground transmission and distribution cables, has been identified. The state-of-the-art magnetic field simulation packages (PCFIELD and MAGNETO) has been used to quantify and manage the field values due to underground cables. The implementation of passive shielding techniques for field management has been successfully carried out to shield source of fields. Two plate thickness were taken and the results show a very high value of shielding techniques for field
209
management has been successfully done to shield source of fields. Two plate thickness were taken and the results show a very high value of shielding effectiveness. In some cases, the reduction is as high as 98%. Successful attempts have also been made to implement the subject shielding using MAGNETO simulation package. The trapping and field diversion have been shown in the contour maps and the magnetic field profiles. Effect of different materials (different permeability) and different plate geometry has been investigated. Materials having good conductivity and high relative permeability have been shown to be very effective for shielding purposes.
Acknowledgements The authors acknowledge KFUPM support.
References [1] D.A. Savitz, et al., Case-control study of childhood cancer and exposure to 60 Hz magnetic fields, Am. J. Epidemiol. 128 (1988) 21 – 38. [2] S.J London, et al., Exposure to residential electric and magnetic fields and the risk of childhood leukemia, Am. J. Epidemiol. 134 (1991) 923 – 937. [3] T.S. Perry, Today’s view of magnetic fields, IEEE Spectrum (1994) 14 – 23. [4] A. Ahlkom, M. Feychting, Magnetic fields and cancer in children residing near Swedish high voltage power lines, Am. J. Epidemiol. 138 (1993) 467 – 481. [5] B. Floderus, et al., Occupational exposure to electromagnetic fields in relation to leukemia and brain tumors: a case-control study in Sweden, Cancer causes control 4 (1993) 465 – 467. [6] Ad Hoc Working Group, Extremely low-frequency electric and magnetic fields and risk of human cancer, Bioelectromagn., 11 (1990) 91 – 99. [7] EPRI Report EL-6759-D, Transmission Cable Magnetic Field Management PR 7898-21, April, 1990. [8] EPRI Report EL-7340-D, Single Conductor Transmission Cable Magnetic Fields PR 7898-29, August 1991. [9] EPRI Report TR-102003., Transmission Cable Magnetic Field Management RP 7898-37, September, 1993. [10] A.S. Farang, A.M. Al-Sheri, J. Bakhashwain, T.C. Cheng, G. Zheng, V. Curci, M. Ghaemi, M. Khajavi, Magnetic Field Modeling and Management for Transmission Underground Cables, JI Cable’95 Proceedings, Paris, June 1995, pp. 596–601. [11] Y. Du, T.C Cheng, A.S. Farag, Principles of power frequency magnetic field shielding with flat sheets in a source of long conductors, IEEE Trans. Electromagn. Compat. 38 (3) (1996) 450 – 459. [12] I.O. Habiballah, A.S. Farag, M.M. Dawoud, A. Firoz., Underground cable magnetic field simulation and management using new design configurations, J. Elec. Power Syst. Res., 45 (2) (1998) 141 – 148. [13] H. Nakanishi, H. Inoguchi, I. Hashimoto, Y. Ito, H. Takehana, N. Nagaoka, A. Ametani, A study of zero-sequence current induced in a cable system, IEEE Trans. Power Appar. Syst., 6 (4) (1991) 1352 – 1358. [14] EPRI., Handbook of Shielding Principles For Power System Magnetic Fields RP 3335-02, EPRI Report Tr-103630, April, 1994.
Implementation of shielding principles for magnetic field management of power cables A.S. Farag *, M.M. Dawoud, I.O. Habiballah King Fahd Uni6ersity of Petroleum and Minerals, KFUPM Box 371, Dhahran 31261, Saudi Arabia Received 13 November 1997; received in revised form 11 May 1998; accepted 5 June 1998
Abstract Adverse health effects due to magnetic fields is a matter of great concern and has been widely debated in recent years. Managing these fields is a challenge to researchers. One of the important sources of magnetic fields is power cables. Different management techniques have been studied. In this paper, passive shielding schemes are implemented to manage the magnetic fields of power cables. The various shielding schemes can be separated into two broad categories: shielding subject and shielding source. Both schemes are implemented in this paper. Passive shielding schemes are found to be the most powerful technique as the reductions obtained are sometimes as high as 97–98%. This scheme is a costly one, should be used only in assigned locations, and as such, there has to be a trade-off between the cost and the level for reduction desired and the health desired and the health hazards. © 1999 Elsevier Science S.A. All rights reserved. Keywords: Magnetic field management; Power cables; Passive shielding
1. Introduction Questions about effects from exposure to power systems AC electric and magnetic fields are of continuing interest to utilities, regulators, consumers and power equipment suppliers. The possible effect of low frequency electromagnetic field (EMF) exposures in occupational and residential environments raises the questions of how electromagnetic fields are created and what effects they may have. There is substantial scientific uncertainty and no widespread agreement among scientists as to how the presently available information regarding the possible health effects of magnetic fields should be interpreted. It is also not clear as to what aspect, if any, of the average magnetic field might be of significance. Those aspects currently being examined include the average field strength, peak field strength, switching of fields, transients, time spent in field, and frequency of field [1 – 6]. There is no doubt that many people are concerned about the magnetic field effects of power frequency electric currents associated with AC transmission and * Corresponding author. Tel.: + 966 3 8602288/3535; fax: + 966 3 8603535; e-mail: [email protected]
distribution underground cables. While health studies are in progress, it appears desirable to conduct parallel technical studies related to the magnetic field management that serve as a guideline to the utilities in practical implementation in occupational areas. Magnetic field management techniques for cables being a recent area of research has not received much of an attention from researchers and utility engineers. The application of some of these techniques to multi-conductor underground lines have not been carried out so far and they are very much a source of concern because of the high currents they carry. Attempts have been made to see the reduction over a distance by the application of some of the management techniques for the case of multi-conductor underground lines [7–14]. In this paper, the implementation of shielding schemes and a comparison between the shielded and unshielded transmission power cables magnetic field values is carried out. A very high reduction is obtained through passive shielding, however, the scheme is quite costly. Different magnetic field profiler showing trapping and field profiles showing trapping and field diversions for different plate geometries are investigated through shielding materials of good conductivity and high relative permeability.
0378-7796/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved. PII S0378-7796(98)00108-4
194
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
2. Factors affecting magnetic fields of power cables There are numerous factors which affect the values of magnetic fields produced by underground transmission cables. These factors may be grouped into the following general areas: 1. System parameters such as current magnitude and phase balance, and system grounding, 2. Cable installation parameters such as depth of burial, installation configuration and relative placement of the cable phases when there is more than one circuit, 3. Cable construction parameters: primarily shield/ sheath resistance and type of material for non-pipetype cables, and 4. External factors such as the presence of nearby underground conductors or other current sources which may flow on the cable sheath/shield or ground continuity conductor.
land availability in the vicinity of the sources. This approach consists of finding ways to avoid the presence of living beings near the location of high magnetic fields by modifying work rules, e.g. limiting access to high magnetic field zones and installations of equipment far away from the areas frequently used by the people.
3.2. Manipulation of source geometry and current This approach is, at present, widely accepted by utilities. Techniques pertaining to his approach are compacting source geometry, modifying circuit currents or characteristics of magnetic materials of facilities and equipment [9–12]. Some of these techniques include: (a) In case of double circuits of cables (two cables per phase), there is a mutual cancellation of magnetic fields. Cables are placed so that they are point symmetric to avoid the induction of zero sequence currents on the two cable circuits. A vertical duct bank installation is probably the most practical installation method to take advantage of magnetic field reduction through mutual
3. Magnetic field management techniques in cables structures Magnetic field management is basically concerned with the minimization of the effect of a magnetic field on the public health front without sacrificing the effectiveness and reliability of the power system. Since magnetic field sources in power systems have been identified, the interest of utilities have now shifted to magnetic field management of these sources. The application of magnetic field exposure reduction techniques to existing facilities and equipment is much more difficult than for new constructions as they have severe constraints which limit the choice of suitable magnetic field reductions and offer a new set of engineering challenges. For each magnetic field source, there may be a number of design options that would reduce its exposure without altering the function job for which the system was intended. Consideration of the factors which significantly affects the magnetic field produced by transmission cables, gives a general approach to the reduction of the magnetic field produced by single-conductor cables. The drawback is that the current carrying capacity of the cable becomes reduced in most of the methods due to the increase in mutual heating between the cables or by an increase in losses. A general classification of the different approaches for magnetic field reduction is given in the following sections [9–12].
3.1. Increased distance between the sources and point of interest This approach is simple and straightforward provided there are no physical constraints, such as space or
Fig. 1. Shielding a subject.
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
195
Table 1 Maximum value of magnetic field for two cables per phase (single phase cables) Cable diameter (in.) Plate thickness (in.)
1.07 1.07 1.55 1.55 1.89 1.89
0.005 0.030 0.002 0.030 0.005 0.030
Derated current (Amp) Without shielding
87.5 87.5 190 190 272.5 272.5
With shielding
Stack
Triangular (mG)
Flat
Stack
Triangular (mG)
Flat
8.74 8.74 27.14 27.14 47.04 47.04
4.40 4.40 13.71 13.71 23.72 23.72
0.901 0.901 4.07 4.07 8.62 8.62
0.48 0.50 0.67 0.68 0.97 0.98
0.59 0.66 0.98 1.00 1.45 1.47
0.58 0.66 0.48 0.49 0.55 0.56
Table 2 Maximum value of magnetic field for three cables per phase (single phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.030 0.005 0.030 0.005 0.030
Derated current (Amp) Without shielding
87.5 87.5 190 190 272.5 272.5
With shielding
Stack
Triangular (mG)
Flat
Stack
Triangular (mG)
Flat
4.69 4.69 15.32 15.32 27.54 27.54
6.56 6.56 20.33 20.33 34.92 34.92
13.21 13.21 32.19 32.19 55.96 55.96
0.50 0.54 0.54 0.546 0.68 0.69
0.83 0.95 1.37 1.40 2.03 2.07
0.63 0.74 0.87 0.88 1.35 1.37
Table 3 Maximum value of magnetic field for four cables per phase (single phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.030 0.005 0.030 0.005 0.030
Derated current (Amp) Without shielding
87.5 87.5 190 190 272.5 272.5
cancellation produced by the phase conductors. Horizontal configuration magnetic field reduction is mot as great as vertical configuration. Stacked horizontal configuration would also permit effective magnetic field cancellation between the two circuits. (b) Close triangular configuration results in relatively low magnetic fields but increases mutual heating and therefore reduces the current carrying capacity of the cable. (c) The use of multi-point grounding and a cable with low shield/sheath resistance. This option results in a significant reduction in capacity for a given conductor size. This option is used only when the above two
With shielding
Stack
Triangular (mG)
Flat
Stack
Triangular (mG)
Flat
1.73 1.73 7.69 7.69 16.08 16.08
6.48 6.48 20.08 20.08 34.70 34.70
1.38 1.38 6.23 6.23 13.19 13.19
0.39 0.41 0.52 0.53 0.64 0.65
0.90 0.87 1.13 1.16 1.61 1.65
0.57 0.65 0.51 0.53 0.62 0.64
options do not reduce the magnetic field to acceptable levels. Multi-point grounding makes the most engineering sense for cables installed in a close triangular configuration since the reduction in capacity is much less for horizontal or vertical configurations. If multipoint grounding is used to reduce the magnetic field levels, a shield/sheath resistance lower than that required by fault currents considerations will probably be necessary. (d) Placing any ground continuity conductors also reduces the induced current. The reduction in the magnetic field by optimum placement of the ground conductor is usually small compared to the other options,
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
196
Table 4 Maximum value of magnetic field for five cables per phase (single phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.030 0.005 0.030 0.005 0.030
Derated current (Amp) Without shielding
87.5 87.5 190 190 272.5 272.5
With shielding
Stack
Triangular (mG)
Flat
Stack
Triangular (mG)
Flat
4.69 4.69 15.54 15.54 28.35 28.35
8.93 8.93 26.96 26.96 46.66 46.66
17.56 17.56 14.98 14.98 26.95 26.95
0.82 0.95 0.61 0.63 0.78 0.77
0.52 0.57 1.34 1.38 1.31 1.34
0.82 0.98 0.62 0.64 0.76 0.77
Table 5 Maximum value of magnetic field for six cables per phase (single phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.030 0.005 0.030 0.005 0.030
Derated current (Amp) Without shielding
87.5 87.5 190.0 190.0 272.5 272.5
and hence is to be avoided for significant field reductions. (e) An increase in the depth of burial also reduces the magnetic field directly above the cables. This is effective only in reducing the maximum magnitude of the field. However, increasing the burial depth leads to a decrease in capacity and an increase in the cost of installation.
3.3. Shielding with conducting materials Magnetic field shielding with conducting (including high permeability) materials is a common practice in industry and has received extensive attention recently. Shielding methods may include the use of induced currents, modification of magnetic flux patterns using high permeability and/or high conductivity materials, addition of a second field that tends to reduce the original field, and even a change in technology to eliminate 60 Hz magnetic fields for specific applications. Current shielding designs or implementations are just a practice of experience. The higher the permeability of the material, the better the shielding [11 – 14]. Gaps in shielding can radically affect the flux within the shield. In practice, there is a trade off between the effectiveness of shielding and its type and parameters such as the radius, thickness and size as the radius, thickness and
With shielding
Stack
Triangular (mG)
Flat
Stack
Triangular (mG)
Flat
1.73 1.73 7.69 7.69 16.08 16.08
6.48 6.48 20.08 20.08 34.70 34.70
1.38 1.38 6.23 6.23 13.19 13.19
0.39 0.41 0.52 0.53 0.64 0.65
0.90 0.87 1.13 1.16 1.61 1.65
0.57 0.65 0.51 0.53 0.62 0.64
size. It is very important that the shields remain continuous for maximum effectiveness[11 –14]. The various existing shielding schemes can be separated into two broad categories: (a) shielding a subject; and (b) shielding a source. There are very distinct differences between these two categories. Shielding a subject means implementing a shield of some sort in order to reduce the field in some relatively small, well defined volume, due to sources of field outside the volume. Shielding the source involves placing a shield to reduce the magnetic field in the ‘outside’ world due to some localized source. An example of shielding a subject would be to place a highly permeable material around a subject (Fig. 1 (a)), as the magnetic flux coming from the top takes the path of least reluctance (highest m) rather than going through the high reluctance of the air. It is shunted around the subject (Fig. 1 (b)) where the unperturbed field B0 induces a current to flow in the loop which will produce a field in the opposite direction of B0 and the field at the subject is reduced. It should be noted that there are regions of space in which the magnetic field is increased due to the shield. The shield would need to be designed in such a way as to designed in such a way that those regions are located where they have no effect. Another way of shielding can be carried out with materials of high permeability and/or conductivity.
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
197
3.4. Shielding with current carrying wires A magnetic field can be shielded (i.e. reduced) by establishing currents in wires such that the fields produced by those current oppose the fields to be reduced. The shielding effectiveness is generally a function of the magnitude and the phase of the current. Currents in the wires can be established in two ways. They can be induced by the external source of the magnetic field (passive shielding), or they can imposed by external devices (active shielding).
3.4.1. Passi6e shielding Passive shielding (or shielding with ‘passive’ conductors) refers to the use of currents induced in conductors by existing (or ambient) magnetic fields to reduce (shields) these fields in a certain region. A time varying magnetic field passing through a closed loop will induce a voltage into the loop (Faraday’s law) as: emf= −(F/ (t. If the loop is a conductive wire, the voltage will cause a current to flow. This current will set up its own magnetic field. the net magnetic field in the region will be a superposition of the original field and the field due to the induced current which is produced to oppose the original field. Since the induced voltage onto the loop is proportional to the derivative of the flux through the loop, the voltage will be 90° shift of the current with respect to the voltage, and therefore, a 180° difference will exist between the original field and the induced current fields to oppose each other. It should be noted that in some cases, it is possible to over-compensate the original field and actually cause an increase in the field level over some regions, i.e. regions very close to the wire of the loop. However, these regions are generally very localized, and their locations can be controlled.
Fig. 2. Comparison of magnetic field levels for cables with and without shielding for different configurations.
3.4.2. Acti6e shielding Active shielding (or shielding with ‘active’ conductors) refers to any scheme which reduces the magnetic field in certain regions of space by the use of conductors with an imposed current whose magnitude, direction, and phase angle create fields in opposition to the ambient fields and thereby reduce the overall magnetic field in a region. Currents (magnitude and phase) can be imposed in conductors by electric devices to reduce the magnetic field levels in certain regions. These devices generally operate in two ways. The first involves placing a magnetic field sensor (transducer) in the region to be shielded. The sensor provides feedback to the electric device that adjusts the magnitude and the phase of a current being fed to a wire loop around the region to be shielded. The electric device continually adjusts the current as it attempts to null out the field at the sensor [11,14]. A second method is to manually adjust the magnitude and phase of the current being fed to the shielding loop until the field id nullified. This is a very effective means of
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
198
Table 6 Maximum value of magnetic field for two cables per phase (three phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.03 0.005 0.03 0.005 0.03
Derated current (Amp)
87.5 87.5 190.0 190.0 272.5 272.5
Without shielding
With shielding
Flat
Triangular (mG)
Flat
Triangular (mG)
0.321 4.379 1.456 1.456 2.698 2.698
4.379 0.297 13.604 13.604 23.586 23.586
0.297 0.307 0.529 0.523 0.788 0.759
0.525 0.540 1.111 1.115 2.941 2.647
Table 7 Maximum value of magnetic field for three cables per phase (three phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.03 0.005 0.03 0.005 0.03
Derated current (Amp)
87.5 87.5 190.0 190.0 272.5 272.5
Without shielding
With shielding
Flat
Triangular (mG)
Flat
Triangular (mG)
6.675 6.675 20.762 20.762 35.937 35.937
6.543 6.543 20.242 20.242 34.96 34.96
0.965 1.082 1.233 1.276 1.719 1.754
0.653 0.672 1.452 1.442 3.571 3.621
Table 8 Maximum value of magnetic field for four cables per phase (three phase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.03 0.005 0.03 0.005 0.03
Derated current (Amp)
87.5 87.5 190.0 190.0 272.5 272.5
reducing fields, however, an obvious drawback is that the device must be adjusted every time the field changes [11,14].
4. Quantity shield effectiveness In order to evaluate the effectiveness of various shielding schemes, it is necessary to define terms which quantify the degree of shielding and there are several ways to do so. The terms shielding effectiveness and shielding factor (SF) are most often used. Shielding effectiveness is a generic term without a rigorous mathematical definition and is the standard for quantifying shielding effectiveness [10 – 14]. The shielding factor is defined as a ratio of resultant field after shielding to that before shielding (open air), and can be expressed as:
Without shielding
With shielding
Flat
Triangular (mG)
Flat
Triangular (mG)
6.586 6.586 20.525 20.525 35.590 35.590
8.677 8.677 26.689 26.689 45.853 45.853
0.879 0.964 1.266 1.304 1.785 1.816
0.776 0.797 1.712 1.718 4.660 4.215
(1)
SF= B/B0
where B is the rms value of the magnetic field with the shield in place, and B0 is the rms value of the unperturbed magnetic field, i.e. the magnetic field without the shield. Clearly, SF is a function of position and always lies between 0 and 1. The smaller SF is, the more significant the magnetic field reduction. A unitary value of SF means no shielding, and zero SF means perfect shielding sometimes, the shielding factor is called the attenuation factor and is expressed in dB as: SF= 20 log (B/B0)
dB
(2)
Another quantity, which is sometimes used to quantify shielding effectiveness is the shielding efficiency (SE), which is defined as: SE= (1−SF)*100
(3)
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
199
Table 9 Maximum value of magnetic field for five cables per phase (three phrase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.03 0.005 0.03 0.005 0.03
Derated current (Amp) Without shielding With shielding
87.5 87.5 190.0 190.0 272.5 272.5
Flat
Triangular (mG)
Flat
Triangular (mG)
8.831 8.831 27.580 27.580 47.915 47.915
10.772 10.772 24.238 24.238 41.734 41.734
1.203 1.326 1.625 1.677 2.251 2.297
0.895 0.918 1.850 1.750 7.207 6.375
Table 10 Maximum value of magnetic field for six cables per phase (three phrase cables) Cable diameter (in.)
1.07 1.07 1.55 1.55 1.89 1.89
Plate thickness (in.)
0.005 0.03 0.005 0.03 0.005 0.03
Derated current (Amp) Without shielding With shielding
87.5 87.5 190.0 190.0 272.5 272.5
Another quantity is the field reduction factor which is simply the reciprocal of the shielding factor.
5. Implementation of shielding principles
5.1. Passi6e shielding of the source The principle of passive shielding is applied to the sources which are single phase and three phase cables. Different cases are simulated, using the software package PCFIELD by EPRI, without shielding and after shielding with the same material as the conductor itself. Five different plate thickness of 0.005, 0.003, 0.001, 0.003 and 0.05 in. are used and the field reductions are calculated. However, the results of only two plate thickness (0.005 and 0.03 in.) are reported here as the difference between other thickness is insignificant. Three different cable sizes namely, c 2/0 AWG, 500 MCM, and 1000 MCM, with conductor diameters of 1.07, 1.55 and 1.89 in., respectively were also selected. A 2-in. thickness for insulation is also incorporated. The results obtained show significant reductions in the magnetic field levels.
5.1.1. Single phase cables The results for single phase cables having different configurations and different number of circuits are shown in Tables 1–5. For two cables per phase (Table 1), if considering a cable of 1.89 in. diameter, the field reduces from a level of 47.04 – 0.97 mG and 0.98 (stack
Flat
Triangular (mG)
Flat
Triangular (mG)
11.039 11.039 34.391 34.391 59.611 59.611
12.818 12.818 31.061 31.061 53.581 53.581
1.466 1.640 1.822 1.884 2.487 2.540
1.008 1.003 1.755 1.714 5.430 4.900
configuration) with 0.005 and 0.03 in. shielding plate, respectively. In terms of the shielding factor, it is only 0.0206, which is a very high reduction. For other cable diameters, the reduction is generally of the comparable level. The simulated values with shielding do not necessarily correspond to optimal conditions of simulated valves without shielding. For three cables per phase (Table 2), the shielding factor varies between 0.0025 and 0.144 for the cases considered. Consequently, the field level comes sown to as low as only 2–3% of the simulated value without shielding. The lower the value of the shielding factor, the more effective the shielding is. The reductions are more pronounced for larger sizes of cables. It should be noted that for a thicker plate, the reduction is generally less due to the fact that these plates are reducing the level of the magnetic field over a particular region. Thicker plates generally reflect the field closer to the source compared to thinner plates. The region taken in simulation is − 50–50 ft on either side of the reference axis which is generally the center of the cable configuration. A sample of the plot showing the comparison between the shielded and unshielded case for the stack, triangular and flat layout is illustrated in Fig. 2 for five cables per phase with a cable diameter of 1.55 in.. For four to six cables per phase, the stack and flat configurations have approximately the same levels for magnetic fields for large cable diameters as illustrated in Tables 3–5), while for small diameters, the stack configuration is better. For five cables per phase, the triangular configuration is better than both the stack and flat configurations for small cable diameters.
200
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 3. Comparison of magnetic field levels for cables with and without shielding for different configurations.
5.1.2. Three phase cables For a three phase cable, two types of configurations —the flat and the triangular — are simulated. The chosen configurations do not always correspond to the
phase locations that gives the minimal field. To be precise, with the exception of two cables per phase, all of the configurations of flat arrangements are not the optimal conditions. For triangular, the values are the
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
201
Table 11 Maximum value of magnetic field for four cables per phase (single phrase cables) Cable diameter (in.)
1.89 1.89 1.89 1.89 1.89 1.89
Conductor current (Amp)
272.5 272.5 272.5 272.5 272.5 272.5
Shield current (Amp)
27.25 −27.25 27.25 −27.25 27.25 −27.25
Shield angle (°)
Magnetic field values (mG)
a
b
c
0 0 0 0 0 0
0 0 −120 −120 120 120
0 0 −240 −240 240 240
165.683 167.014 2.524 2.065 2.431 2.211
Fig. 4. Plot of the best and worst case for active shielding for stack configuration (cable diameter 1.89 in., two cables per phase, single phase cable).
minimum possible one could get out of the phase relations. Tables 6– 10 show the values of the magnetic fields before and after shielding for 1.07, 1.55, and 1.89 in. diameter cables. In the case of a double circuit line, Table 6 gives the values before and after shielding and the reduction in the field level varies between 8 and 92%. For three cables per phase, the maximum reduction is 95%. The reduction is more generally in the case of a thinner plate but the difference is not significant. In some of the cases of triangular configuration, the thicker plates give less fields. This is due to a different configuration in the conductor arrangements. For five and six cables per phase, the level of reductions obtained is very high and this leads to the conclusion that the passive shielding scheme of the magnetic field reduction is a very successful technique but at an extra cost. Samples of the plots showing the comparison between the shielded and the unshielded cases for flat and triangular layouts are illustrated in Fig. 3 for five cables per phase of cable diameter of 1.55 in.
5.2. Acti6e shielding of the source The theory of active shielding is still in its early stage and very little work has been done in this area. However, one case of simulation result is shown. The cable size selected is 1000 MCM having a diameter of 1.89 in. The rated current for this size of cable is 550 amps and a 50% derating is applied. The shield current is taken as 10% of the derated conductor current. the simulation is performed for two cables per phase (single phase cable) having stack configuration. The phase location does not correspond here for minimal field. The results are tabulated in Table 11. As evident from the results, the minimum field is obtained when the shield current direction is opposite to the conductor current and have the same phase sequence as the conductor. The value obtained is 2.065 mG, while for the same case when it is not shielded, the field value was found to be 2.295 mG. Hence, there is a reduction of 10%. The plot of the best and worst case is shown in Fig. 4.
202
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 5. Contour map and profile graph for case 1.
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 6. Contour map and profile graph for case 2.
203
204
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 7. Contour map and profile graph for case 3.
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
205
Fig. 8. Contour map and profile graph for case 4.
5.3. Shielding the subject 5.3.1. Simulation methodology The software MAGNETO uses the boundary element method (BEM) for calculation of field distribu-
tions. Basically, the software is used to solve the effects of shielding using different magnetic materials having different geometry of plates. The method adopted is as follows: (a) the limits for different quantities are set; (b) the geometry of the problem is defined (Xmin, Ymin, and
206
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 9. Contour map and profile graph for case 5.
Ymax) beside the grid size and conductors are defined by their coordinates; (c) currents are assigned to the conductors; (d) materials are chosen from the tables including their permeability’s and the materials are
placed in their appropriate regions; and (e) the boundary elements are defined which is crucial to achieve valid results by proper distribution of boundaries.
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
207
Fig. 10. Contour map and profile graph for case 6.
6. Results and discussions Different cases of shielding with different plate ge-
ometry and different materials are simulated. the results of different cases are documented. In case 1, two rectangular plates having a thickness of 0.3 in. and
208
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
Fig. 11. Contour map and profile graph for case 7.
separated by 1 in. are used for shielding purposes. The plate above the cable is ferrite possessing a permeability of 2000 and is at a distance of 6 in. from the conductors. The other plate is made of aluminum (relative
permeability= 1). The contour map and profile graph are shown in Fig. 5. The ferrite plate practically traps all of the fields and the ends of the plates being stressed more. In case 2, the position of the two plates are
A.S. Farag et al. / Electric Power Systems Research 48 (1999) 193–209
interchanged and the field easily passes through the aluminum plate, which carries out practically no shielding and only the ferrite plate helps in trapping the fields (Fig. 6) for the contour map and profile graph. In case 3, two high permeable materials (ferrite) and cast iron (relative permeability =165.96) plates are placed over each other. Ferrite is farther away from the conductor. A Contour map is shown in Fig. 7, where some of the fields are trapped by the lower plate and the rest by the ferrite plate. The contour of x-component of the field are diverted to the surroundings. In case 4, the positions of the two plates are interchanged. Fig. 8 shows that the lower high permeable plate does most of the trapping while the upper plate of cast iron takes care of the rest so that the region behind it is perfectly clear of magnetic fields. Moreover, the field near the conductor surface is not diverted to the surroundings as in case 3. Therefore, it is better to keep the high permeable plate near the conductors in the case of multilayer shielding. Case 5 shoes three levels of plates with an air gap on the middle plate level. The middle plate is made of aluminum, while the top and bottom plates are made up of highly permeable ferrite. The lower plate traps some of the field while the aluminum does little to reduce the field in a particular region. The upper ferrite plate diverts some of these into other regions (Fig. 9). If the positions of the plates are interchanged leading to case 6, the situation is much better compared to case 5, as predicted from Fig. 10. Gaps in shielding greatly affect the field. The ferrite plate, which is at the middle, provides a good shielding to the region of interest. The ends of the plates get further stressed. The lower aluminum plate does nothing to the field and it simply passes through it. In case 7, a single plate of ferrite is used for shielding but with an inverted U-shape. This will shield the conductors from three sides. The contour plot and profile graph are shown in Fig. 11, which demonstrates that the bend points are more stressed, it gives a better shielding performance, and there is not much change in the profiles over the region.
7. Conclusions In this paper, one of the major sources of magnetic fields, which is underground transmission and distribution cables, has been identified. The state-of-the-art magnetic field simulation packages (PCFIELD and MAGNETO) has been used to quantify and manage the field values due to underground cables. The implementation of passive shielding techniques for field management has been successfully carried out to shield source of fields. Two plate thickness were taken and the results show a very high value of shielding techniques for field
209
management has been successfully done to shield source of fields. Two plate thickness were taken and the results show a very high value of shielding effectiveness. In some cases, the reduction is as high as 98%. Successful attempts have also been made to implement the subject shielding using MAGNETO simulation package. The trapping and field diversion have been shown in the contour maps and the magnetic field profiles. Effect of different materials (different permeability) and different plate geometry has been investigated. Materials having good conductivity and high relative permeability have been shown to be very effective for shielding purposes.
Acknowledgements The authors acknowledge KFUPM support.
References [1] D.A. Savitz, et al., Case-control study of childhood cancer and exposure to 60 Hz magnetic fields, Am. J. Epidemiol. 128 (1988) 21 – 38. [2] S.J London, et al., Exposure to residential electric and magnetic fields and the risk of childhood leukemia, Am. J. Epidemiol. 134 (1991) 923 – 937. [3] T.S. Perry, Today’s view of magnetic fields, IEEE Spectrum (1994) 14 – 23. [4] A. Ahlkom, M. Feychting, Magnetic fields and cancer in children residing near Swedish high voltage power lines, Am. J. Epidemiol. 138 (1993) 467 – 481. [5] B. Floderus, et al., Occupational exposure to electromagnetic fields in relation to leukemia and brain tumors: a case-control study in Sweden, Cancer causes control 4 (1993) 465 – 467. [6] Ad Hoc Working Group, Extremely low-frequency electric and magnetic fields and risk of human cancer, Bioelectromagn., 11 (1990) 91 – 99. [7] EPRI Report EL-6759-D, Transmission Cable Magnetic Field Management PR 7898-21, April, 1990. [8] EPRI Report EL-7340-D, Single Conductor Transmission Cable Magnetic Fields PR 7898-29, August 1991. [9] EPRI Report TR-102003., Transmission Cable Magnetic Field Management RP 7898-37, September, 1993. [10] A.S. Farang, A.M. Al-Sheri, J. Bakhashwain, T.C. Cheng, G. Zheng, V. Curci, M. Ghaemi, M. Khajavi, Magnetic Field Modeling and Management for Transmission Underground Cables, JI Cable’95 Proceedings, Paris, June 1995, pp. 596–601. [11] Y. Du, T.C Cheng, A.S. Farag, Principles of power frequency magnetic field shielding with flat sheets in a source of long conductors, IEEE Trans. Electromagn. Compat. 38 (3) (1996) 450 – 459. [12] I.O. Habiballah, A.S. Farag, M.M. Dawoud, A. Firoz., Underground cable magnetic field simulation and management using new design configurations, J. Elec. Power Syst. Res., 45 (2) (1998) 141 – 148. [13] H. Nakanishi, H. Inoguchi, I. Hashimoto, Y. Ito, H. Takehana, N. Nagaoka, A. Ametani, A study of zero-sequence current induced in a cable system, IEEE Trans. Power Appar. Syst., 6 (4) (1991) 1352 – 1358. [14] EPRI., Handbook of Shielding Principles For Power System Magnetic Fields RP 3335-02, EPRI Report Tr-103630, April, 1994.