FMS Publications

Power Frequency Magnetic Field Management Using a Combination Of Active And Passive Shielding Technology

Abstract

Power frequency magnetic fields are difficult and expensive to shield, particularly when the fields originate from sources with complicated field pattern. A method is discussed and test data presented which suggests that carefully combining active and passive shielding techniques can produce results which are importantly superior to either technique when used alone. The results are generally consistent with an analysis which considers sources as a superposition of elementary source fields with different order terms.

Key Words - Magnetic Fields, Active Field Cancellation, Passive Shielding

Introduction

Questions about possible adverse health effects of power frequency magnetic fields, coupled with the capacity of such fields to interfere with sensitive instruments and computer monitors, have spurred efforts to find effective and economical means to reduce fields in a variety of settings. Attention has focused on power lines but other sources are both as prevalent and as difficult to manage. Examples are transformers, cables, bus bars and service entrance panels. Field management techniques proposed for power lines have typically employed either compaction, reconfiguration or distance. When these "utility-specific" methods are impractical, active cancellation, either by induction in a passive conductor loop or by active cancellation from directly energizing a conductor current loop, are options. These techniques are well documented, though not universally employed, and essentially involve cancellation of one field by another of opposite phase.

Such techniques are most effective on those fields which are most uniform: in direction, intensity and polarity. Accordingly, fields from transformers, cable and bus bars and service panels are not effectively managed by such active cancellation devices due to their geometry and the complexity of the fields produced. These sources and the fields produced by them are typically managed, more or less effectively, by shielding with either permeable or conductive materials, or a combination of both.

This paper presents a design for a field management solution applied to electric panels which combines, selectively, elements of both of the two general methods in current use (passive and active shielding). This is done on such a way as to take advantage of their desirable characteristics but uses them in combination to minimize their respective undesirable characteristics.

This paper discusses both single-phase and three-phase cases and presents three-phase test results of this method. However, the method applies to single-phase service panel applications and can be of significant benefit in other applications as well, including transformers, cables and bus bars. Future papers are envisioned to discuss those areas and their differences from panel treatment.

PASSIVE AND ACTIVE MAGNETIC FIELD REDUCTION PRINCIPLES

Overview of Magnetic Field Reduction Method

The geometry of a typical switch panel is shown in Fig. 1. Note that the depth (d) of the panel is small compared to its height (h) and width (w). Inside the box is a set of wires or busses which distributes currents to different circuits within the building.

One or more sheets of magnetic and/or conductive material (passive shield) are placed in front of the panel. In addition, a slightly smaller loop of current (active shield) is placed between the passive shield and the test panel to produce a magnetic field which partially cancels that of the panel. The amplitude and phase of the loop current are determined by a remote coil sensor. The shielding scheme is illustrated in Fig. 2.

Multipole Expansion of Source Magnetic Fields

To understand the shielding mechanism, it is useful to describe the source of magnetic fields as a set of quasi-static magnetic multipole sources [4]. Using this description, the magnetic fields can be written as a superposition of elementary source fields. The least complicated of these elementary sources are dipoles and are illustrated in Fig. 3. The dipoles are equivalent to small loop currents in the xy, xz and yz planes as shown. The arrows on the loops indicate the reference directions for the currents. One characteristic of dipole terms is that their associated magnetic fields decay as 1/r³, where r is the distance from the center of the source.

The next set of terms, the "quadrupole" terms, can be described as pairs of dipoles. Some of these terms are illustrated in Fig. 4.

In all cases, the currents on the two dipoles which constitute the quadrupole are equal and opposite. One consequence of this is that the magnetic fields of quadrupoles decay as 1/r⁴, more rapidly than the dipole fields. Thus, these fields are more confined tot he region near the source. Higher order mulipole terms have the characteristic that their fields decay even more rapidly than either dipole or quadrupole terms. Specific formulas for the spatial variation of all these fields are available in the literature [4].

Characterization of the Experimental Source Field Management Services

Since the depth of the service panel is much smaller than either of its other dimensions, the wires within it are roughly parallel to the xy plane shown in Fig. 1. Given this, the dominant dipole term is the one shown in Fig. 3(a). This is important because the active shield will be designed to cancel this dipole term. A loop oriented as shown in Fig. 2. Cannot effectively cancel the magnetic fields of dipoles in planes other than the xy plane.

Single-Phase vs. Three-Phase

Before approaching the question of shielding, it will be useful to discuss the difference between the single-phase and the three-phase case. Fig. 5 presents a simplified mapping of the conductors within the test service entrance panel.

The Single-Phase Case

In the single phase experiment, wires A and C are in parallel and the wires B and N are in parallel and the lower ends of all of the wires are connected together. Thus, in the single-phase experiment, the current flow through wires A and C is in the counterclockwise direction while the current flow through wires B and N is in the clockwise direction (i.e. 180 degrees out of phase). If each wire inside the panel is considered as a loop of current, a rough model of the source can be drawn as shown in Fig. 6. Clearly, the radius of the loop for wire A (r_A) will be greater than those of the others (r_B, r_C and r_N, respectively). The arrows on Fig. 6 indicate relative current direction for the different loops.

In front of the service panel, the dominant magnetic fields of each loop are z directed. Further, the magnetic field of each loop is proportional to its current ad its area. Thus the dominant magnetic field in front of the panel (in phasor form) can be written as:

Where k is an unspecified real constant. Clearly, the phase of this field is either 0 or 180 degrees (depending on the relative sizes of the loop radii) from that of the current.

Given this, it is relatively simple to cancel the dipole portion of the magnetic field in front of the panel using an active cancellation loop. A current on the cancellation loop (either in phase or 180 degrees out of phase with the source current) is adjusted for the amplitude which will roughly cancel the source dipole field. Cancellation will not be perfect everywhere since the effective size of the cancellation loop may well not be the same as that of the source dipole. Improved cancellation will be obtained by adjusting the size of the cancellation loop.

The Three-Phase Case

In the three-phase case, loops A, B and C carry currents with phase 0, -120 and 120 degrees (0, -2pi/3 and 2pi/3 radians) respectively. If hte currents in A, B and C are balanced, then wire N carries no current. The dominant magnetic fields in front on the panel is then:

Clearly, the phase of this magnetic field can vary depending on the relative sizes of the individual loops. Effective cancellation of this field requires adjusting both the amplitude and the phase of the current in the cancellation loop.

The unbalanced three-phase case can be treated as a superposition of balanced three-phase and single-phase current loops.

Basic Theory of Shielding With Materials

Consider a finite sized flat passive shield and a source consisting of a number of multipoles. The amount of field reduction in front of the shield is a function of two things: 1) the thickness and composition of the shield and 2) the height and width of the shield. The effect of the first can be determined by assuming the shield to be of infinite height and width and analyzing the magnetic field with and without the shield [5]. The effect of the second can be described qualitatively in the following way. Consider a magnetic dipole and a magnetic quadrupole source behind the shield as shown in Fig. 79a) and Fig. 7(b) respectively. The fields of each will extend beyond the height and width of the shield as shown by the dashed lines in Fig. 7. Since the quadrupole fields decay at 1/r⁴ compared to the 1/r³ decay of the dipole, the quarupole fields are smaller outside the shield than the dipole fields. The smaller the source fields outside of the shield's dimensions, the smaller the leakage around the shield and hence the more effective the shield. It follows that a shield of fixed size is more effective at shielding a quadrupole than at shielding a dipole. This principle can be extended to higher order mulitpoles. The higher the order, the more effective the shield.

Basic Cancellation Theory

Consider now the use of an active shield (loop) to cancel the dipole term of a set of multipole fields. Perfect shielding of the dipole term can be achieved if a loop of the same size is placed at the same location and driven with an equal and opposite current. This, of course, is not possible. However, the closer the size and placements of the cancellation loop to those of the dipole source, the more complete the cancellation.

APPLICATION OF ACTIVE AND PASSIVE SHIELDING THEORY TO THE PROPOSED METHOD

The Single-Phase Case

The field of the magnetic field source (as mentioned above) can be thought of as that of a dipole in the xy plane plus a number of other higher order multipole terms. In principle, the dipole term can be approximately canceled if the amplitude of the cancellation current is chosen correctly. The phase of the cancellation current need not be varied since it will be the same (or opposite) as that of the source field. Once the dipole sources are eliminated, the remaining source of fields is a set of higher order multipoles. As stated in the last section, the size of shield needed to efficiently shield these terms is smaller than that needed is a dipole term were present.

The Three-Phase Case

The only difference between the three-phase case and the single-phase case is that the amplitude and phase of the cancellation must be set. Without this, the cancellation field cannot match and hence cancel that of the source dipole. The passive shielding theory is the same as that for the single-phase case.

TEST PROCEDURE

Configuration of Equipment

The test configuration was designed to model an office located adjacent to the electric equipment room of a building, inside of which a three-phase panel was acting as a source of elevated magnetic fields. A three-phase switch panel was placed at floor level parallel to and 178 mm (7 in) removed (the thickness of a standard building wall) from one side of a 4.9m x 6.1m (16ft x 20ft) grid. Fig. 8 shows a plan view of the test area with the measurement grid.

The panel is rated at 1000 amps and was fed by a power network which is shown in Fig. 9.

Three single-phase variable transformers were fed by a 4-wire, three-phase circuit. The transformers are capable of supplying independently variable voltage from 0 to 140 volts to a set of three loading transformers which were configured for 120 volt input to 5 volt output. The output from the loading transformers was fed tot he three-phase 4-wire switch panel. This setup allowed for continuous, dynamic control of phase-to-phase balance on each of the three phases from 0 to 1000 amps.

The four conductors of the panel were connected together beyond the panel breaker to create a three-phase current circulation.

Shielding effectiveness tests were conducted using both conductive (aluminum) and ferromagnetic (steel) passive shielding materials and an active cancellation coil, independently and in combination. Fig. 10 presents a section view of the shielding configurations relative to the source switch panel.

Active Cancellation Sensor and Feedback System

A multi-turn coil was placed in a plane parallel to the xy plane of the test panel (perpendicular to the direction of the dominant dipole magnetic field component). The output of the sensor coil fed an active feedback network which drove the cancellation coil so as to minimize the field strength at the sensor location. Although several locations of the sensor coil produced satisfactory shielding, optimal performance was achieved with the sensor located behind the test panel (at z = -d) and at approximately the center of the panel. Although this location was optimal for this test, other configurations will likely require other preferred sensor locations.

Measurement Protocol

All measurements were taken with an EMDEX-C at an elevation of 1 meter (39in) and at 61 cm (2ft) intervals for all intersections of the measurement grid in front of the electric panel. Comparative data presented in this paper is of the longitudinal profile line parallel with the panel at 61 cm (2ft).

Background fields were principally generated by a distribution line near the laboratory and parallel to the panel, which produced regular patterns across the grid, from a high of approximately 1 milligauss (mG) at 0 to 0.4 mG at the 6.1m (20 ft) line.

TEST RESULTS

Fig. 11 presents graphical comparisons between a base, unshielded case in which the panel had 400 amps of circulating current and the same base case but with selected individual passive shields, an active cancellation coil, and a combination of a passive shield and the active cancellation coil.

The results demonstrate that any of a number of combinations of passive shields are capable of reducing the fields from this test panel to values in the range of 10 mG. Other tests performed as part of the underlying data of this report achieved significant further reductions of these fields with passive shielding alone. However, the effort (cost) required for field reductions below 10 mG is disproportionately high and increases with each lower value achieved, and may not be economically feasible for target field strength values in the 2 to 3 mG range.

However, the combination of active cancellation and passive shields offers field reduction advantages over any of the individual passive shields or the singular active coil such that these lower values are achievable without the consequential increases in difficulty.

Although shielding is affected by the materials selected for the passive shield, the major improvement in shielding effectiveness in front of the test panel results from the combination of active shielding (dipole cancellation) with passive shielding (higher order multipole shielding). Further experiments confirm that the dipole field created by the active cancellation coil partially cancels the dipole field from the source at other locations as well, including areas above and along the sides of the test panel. Moreover, field levels at all testpoints within the test grid were reduced by the combination of active and passive shielding.

CONCLUSIONS

Many sources of magnetic fields exist inside buildings. One example is an electric service panel. Such a source can be considered as a set of multipole sources (e.g. dipole, quadrupole, octupole etc.).

Shielding the magnetic fields of an electric service panel has been studied theoretically and experimentally. It has been demonstrated that the combination of a loop with appropriately set current (active shield) and a passive shield of relatively small size is an effective shield of these fields.

The active shield is used to approximately cancel the dipole component of the source fields. With this term significantly reduced, only a smaller passive shield is required to reduce the remaining higher order multipole terms. This is possible because the fields of higher order multipole sources are more confined to the source region.