FMS Publications

Optimum Reduction of Power Frequency Magnetic Fields from Transformer Vaults

INTRODUCTION:

Substantial research exists regarding the characterization and reduction of extremely low frequency (ELF) magnetic fields associated with power systems. However, increasing sensitivity of computer monitors and other electronic equipment to ELF magnetic fields at low levels (3-5 mG) coupled with potential health concerns, has created interest in reducing magnetic fields from sources such as building transformer vaults to significantly lower values than is typical with conventional mitigation schemes. Traditional shielding methods typically lower field strengths into the 10 mG range.

Power frequency magnetic fields from commercial transformer vaults are very difficult to mitigate to values of less than 10 mG due to the extremely complex character of magnetic fields produced by transformer vault components. This study evaluates methods to substantially reduce magnetic field levels from transformer vaults by installation configuration and passive shielding combinations.

OBJECTIVE:

Commercial buildings with internal large electrical transformer vaults, commonly produce magnetic field strengths in the order of 10 to several 100 mG in spaces directly over or adjacent to the transformer room installation. Substantial magnetic field strengths in adjacent areas are usually complex in nature and are the result of multiple sources represented by the electrical components, wiring and connections present in typical transformer vault installations.

The objective of this study was (1) to closely characterize, in a full scale laboratory setting, magnetic fields produced by single and three-phase transformers, their phase conductors and peripheral equipment as typically found in commercial building transformer vault installations, (2) to develop, test and document the effectiveness of configuration and installation techniques intended to reduce magnetic field strengths into the 3 to 5 mG range, and (3) to implement field reduction schemes at several commercial building sites to validate reduction techniques developed in the laboratory setting.

METHOD:

At a Southern California Edison Laboratory, a 20 ft x 16 ft measurement grid was marked on the floor to simulate the approximate size of a typical transformer vault. Three single-phase 500 kVA 12,000-480 volt transformers were placed in the center of the measurement grid in a linear array to form a three-phase 1500 kVA 12,000-277/480 volt transformer bank. The linear arrangement of such transformers is a widely utilized installation configuration by utilities throughout the United States. The test transformers were connected utilizing common industry installation practices to a 12,000 volt primary source. The secondary phase conductors were connected to bus and cable commonly used to connect secondary phase connectors to building electrical distribution equipment. All secondary circuit paths in the test set up were connected together beyond the measurement grid area to create three-phase current circulation. This arrangement permitted 1,000 amps of current to be circulated through the secondary conductors to generate a wide variety of magnetic field conditions in the area of the measurement grid.

Magnetic field spot measurements including x, y, z and resultant values were then taken at all measurement grid locations at 3, 6, 9 and 12 feet above the floor. These measurements provided a comprehensive view of magnetic fields present in areas immediately adjacent to and above the transformer arrangement. Close analysis of the test magnetic field measurement data provided the basis for design of an alternate arrangement wherein the three transformers were placed in a triangular or "delta" configuration to minimize secondary connection lengths and to promote natural cancellation effects. The triangular configuration of transformers was centered in the laboratory measurement grid and connected to the same test power system used for the linear arrangement. Magnetic field measurements were taken at the same measurement grid location at 3, 6, 9 and 12 feet to evaluate the anticipated field reductions.

After completion of the laboratory testing and evaluation of magnetic field strength reduction techniques, two "real world" sites with transformer vaults similar to the linear array laboratory transformer configuration were selected for field characterization and implementation of field reduction measures. A commercial building and a High School were selected for evaluation sites. Each site had linear array transformer vaults located underneath rooms with elevated magnetic field strengths present. Magnetic field spot measurements taken at these sites confirmed magnetic fields strengths in rooms above the transformer vaults were comparable to those noted in the laboratory.

It was not possible to reconfigure the linear array transformers at the two sites with the laboratory developed triangular arrangement. Wiring and installation modifications developed for use in the triangular configuration testing were implemented to improve natural cancellation of magnetic fields by improved adjacency of conductors. Using techniques previously developed in the laboratory, area shielding was also installed in the affected room above each vault, to insure that the average magnetic field levels were lowered to the 3 to 5 mG range.

LINEAR ARRAY TEST

In commercial building transformer vaults, it is common practice to arrange three single phase transformers in a straight line to provide three phase power. A neutral bus of copper strap is customarily attached to an adjacent wall to which the low voltage neutral conductor from each transformer and the building load are connected to form the neutral (star) point. Low voltage phase connections from each transformer are often oriented either perpendicular to or parallel with, the neutral bus often with significant separations between the phase conductors and the neutral bus bar. Also the building (load) connection to the neutral bus is frequently separated from the phase conductors.

TRIANGULAR ARRAY TEST

In the triangular array test, the three single phase transformers were repositioned in a triangular arrangement and the low voltage connections from each transformer were oriented toward the center of the triangle. The neutral point (star) was created at a common location near the center of the triangle at which point the neutral load was also connected. The triangular arrangement of the transformers provided a very short neutral bus to form the common point with transformer phase leads immediately adjacent. The low voltage phase conductors from each transformer were placed adjacent to the transformer and load neutral conductors. The triangular configuration assured that all conductors associated with the transformers were in close proximity to one another thereby optimizing natural cancellation of magnetic fields.

TEST SITE # 1

Excessive computer monitor distortion in several offices of this commercial building, had caused the tenant to relinquish occupancy and use of the affected spaces. The affected offices were located directly above the building’s basement transformer vault and main electrical switch room. Three 333 kVA single phase transformers were positioned in a linear array with a neutral bus positioned on top of the transformer cabinets at some distance from the low voltage phase conductors. Magnetic fields from net currents on a number of ground bonding cables and cable ducts were also documented. A magnetic field survey of the affected office areas confirmed that field characteristics present were generally approximate to the laboratory transformer linear array.

Due to physical constraints, it was not feasible to reconfigure the three single phase transformers in a triangular arrangement. It was possible however, to relocate the neutral bus closer and immediately adjacent to the transformer secondary phase conductors. Additional grounding wiring reconfigurations and corrections were also implemented to minimize the presence of net current conditions. As projected, the neutral bus relocation and wiring changes dramatically reduced peak values and lowered the average B-field level in the office area to the 10 - 20 mG range. A layer of magnetic field shielding was installed on the office floor thereby further reducing the average magnetic field level to the 3 to 5 mG range.

TEST SITE #2

Elevated magnetic fields had caused the test site High School to abandon use of a large classroom located directly over the school’s basement transformer vault and main electrical switch room. Three 250 kVA single phase transformers were positioned in a linear array with low voltage connections made to a series of exposed copper bus bars suspended from the basement vault ceiling. A comprehensive magnetic field survey of the classroom revealed that the bus bars were the primary source of elevated magnetic fields. Fields from net currents present on a number of ground bonding cables and cable ducts were also documented. The classroom magnetic field survey data confirmed that field characteristics present in the classroom were generally approximate to the laboratory transformer linear array.

Due to constraints, it was not possible to implement a triangular reconfiguration of the three single phase transformers. Alternately, a custom robust multi-layer magnetic field shield was designed and installed around the suspended bus bars. Wiring reconfigurations and corrections were also implemented to minimize the presence of net current conditions. As anticipated, the wiring reconfigurations and bus bar shield dramatically reduced peak values and lowered the average B-field level in the classroom to the 10 - 20 mG range. A layer of magnetic field shielding was then installed on the classroom floor to further reduce the average magnetic field level to the target level of 3 to 5 mG.

RESULTS:

Analysis of laboratory measurement data from both transformer test configurations, confirmed that secondary low voltage phase conductors and neutral connections were the largest source of magnetic fields when measured adjacent to and at various distances from, the transformer(s) location. Significantly elevated magnetic field values from the commonly used linear array of single phase transformers were noted at a distance (15 to 20 feet) equal to an office over the vault. These elevated values appear to be primarily the consequence of large separations between the phase and neutral transformer conductors which result in poor self-cancellation.

The triangular configuration of single phase transformers substantially improved the proximity of phase and neutral secondary conductors and minimized separations. Magnetic field values measured at 15 to 20 feet from the transformer were considerably diminished when compared to the linear transformer array measurements. The data suggests that magnetic field reductions of 50% or more can be achieved by compacting phase conductors and optimum arrangement of transformers. However, even after such compaction schemes, magnetic field strengths in the 10's of mG remain at a distance equal to an office over the vault. It appears from the laboratory experience, that further reduction of maximum fields to the range of 3-5 mG or less requires shielding in addition to reconfiguration options.

In both real world test sites, implementation of equipment reconfiguration measures and wiring changes improved natural cancellation between the low voltage phase and neutral conductors, and yielded considerable reductions of magnetic field strengths in affected areas directly over each transformer vault. Magnetic field measurements at the two study sites however, confirmed that magnetic field levels in the 10 mG range remained as predicted from the laboratory testing.

The installation of large planes of conductive shielding on the floor of the affected area above the transformer vault at each test site, successfully lowered the remaining magnetic field strengths to the 3 to 5 mG range as anticipated.

CONCLUSION:

The laboratory test data suggests that ELF magnetic field levels in spaces adjacent to commercial building transformer vaults are due primarily from the low voltage load conductors attached to the transformers, rather than fields produced by the transformer(s). Magnetic fields at distances from transformer vault installations, can be substantially reduced by electrical system design and installation modifications which promote natural cancellation among the transformer phase and neutral conductors. Even modest reconfigurations to bring the neutral bus in close proximity to phase conductors will cause substantial reductions in magnetic field values.

It appears from laboratory tests, that it is impractical to achieve low (3-5 mG) magnetic field values exclusively by electrical system design and installation configurations which promote self-cancellation. The two test sites confirmed that lower values of 3 to 5 mG may be achieved by the careful use of sufficiently large planes of ferromagnetic and conductive shielding material when used in combination with electrical system design modifications which promote self-cancellation. Although elevated magnetic fields in commercial buildings are very difficult to reduce to low values, the test site installations demonstrate that it is possible. Equipment configuration and installation practices at the time of construction are the best options for managing magnetic fields from commercial transformer vaults.