[James_H._Harlow]_Electric_Power_Transformer_Engin(BookSee.org)

They did have the advantage of little harmonic production, but they are less efficient than semiconductor
devices. Nevertheless, they are still used in many applications today with diodes.
The new problem presented by semiconductor technology was harmonic current. The operation of
the semiconductor rectifier produces harmonic voltages and currents. The harmonic currents are at
higher frequencies than the fundamental frequency of the transformer. These higher-frequency currents
cause high levels of eddy-current losses and other stray losses in other parts of the transformer. This can
create potentially high temperatures, which degrade the insulation of the transformer and can cause early
failure of the transformer.
Problems of failures were reported to the IEEE Power Engineering Society Transformer Committee.
These were typically hottest-spot failures in the windings. In 1981 a new standard began development.
Rather than just updating the old mercury-arc rectifier standard, a new standard was created — ANSI/
IEEE C57.18.10-1998, Practices and Requirements for Semiconductor Power Rectifier Transformers. In
a similar time frame, IEC developed its own standard, IEC 61378-1-1997, Convertor Transformers —
Part 1: Transformers for Industrial Applications. The IEEE standard took an inordinate amount of time
to develop due to the development of new products, new terminology, the definition of harmonics, the
estimated harmonic effects on losses and heating, individual company practices, conflicting standards,
harmonization with the IEC’s efforts to develop a standard, and development of appropriate test methods,
to name a few of the obstacles.
Important developments occurred during the preparation of these standards. Some involve the specification
of transformers, some involve performance characteristics and calculations, and some involve
the testing of transformers.
2.4.2 New Terminology and Definitions
At least two new terms were defined in ANSI/IEEE C57.18.10. Both are important in the specification
of rectifier transformers.
2.4.2.1 Fundamental kVA
The traditional IEEE method for rating a rectifier transformer has always been the root-mean-square
(rms) kVA drawn from the primary line. This is still the method used to develop all of the tables and
figures given in ANSI/IEEE C57.18.10, Clause 10. However, the IEC converter transformer standards
define the kVA by the fundamental kVA drawn from the primary line. The rms-rated kVA method is
based on the rms equivalent of a rectangular current waveshape based on the dc rated load commutated
with zero commutating angle. The fundamental kVA method is based on the rms equivalent of the
fundamental component of the line current. There are pros and cons to both methods. IEC allows only
the fundamental-kVA-method rating with an “in some countries” clause to accommodate North American
practice. The logic behind this rating is that the transformer manufacturer will only be able to
accurately test losses at the fundamental frequency. The manufacturer can not accurately test losses with
the complex family of harmonics present on the system. Therefore, according to IEC, it is only proper
to rate the transformer at the fundamental frequency. Transformer rating and test data will then correspond
accurately.
The traditional IEEE rms-kVA rating method will not be exactly accurate at test. However, it does
represent more accurately what a user sees as meter readings on the primary side of the transformer.
Users feel strongly that this is a better method, and this is what their loading is based on.
ANSI/IEEE C57.18.10 allows for both kVA methods. It is important for a user to understand the
difference between these two methods so that the user can specify which rating is wanted.
2.4.2.2 Harmonic Loss Factor
The term harmonic-loss factor, F HL , was developed by IEEE and IEC as a method to define the summation
of harmonic terms that can be used as a multiplier on winding eddy-current losses and other stray losses.
These items are separated into two factors, winding eddy-current harmonic-loss factor, F HL-WE , and the
other-stray-loss harmonic-loss factor, F HL-OSL . These are used as multipliers of their respective losses as
measured at test at the fundamental frequency. Both factors can be normalized to either the fundamental
current or the rms current.
These terms are similar to the values used by the Underwriters Laboratories (UL) K-factor multiplier,
except that stray losses are amplified by a lesser factor than the winding eddy-current losses. The term
F HL-WE comes mathematically closest to being like the term K-factor. It must be noted that the term K-
factor was never an IEEE term but only a UL definition. The new IEEE Recommended Practice for
Establishing Transformer Capability when Supplying Non-Sinusoidal Load Currents, ANSI/IEEE
C57.100, gives a very good explanation of these terms and comparisons to the UL definition of K-factor.
The Transformers Committee of the IEEE Power Engineering Society has accepted the term harmonicloss
factor as more mathematically and physically correct than the term K-factor. K-factor is used in UL
standards, which are safety standards. IEEE standards are engineering standards.
In their most simple form, these terms for harmonic-loss factor can be defined as follows:
and
n
F HL-WE = I h (pu) 2 h 2 (2.4.1)
1
n
F HL-OSL = I h (pu) 2 h 0.8 (2.4.2)
1
where
F HL-WE = winding eddy-current harmonic-loss factor
F HL-OSL = other-stray-loss harmonic-loss factor
I h = harmonic component of current of the order indicated by the subscript h
h = harmonic order
As is evident, the primary difference is that the other stray losses are only increased by a harmonic
exponent factor of 0.8. Bus-bar, eddy-current losses are also increased by a harmonic exponent factor of
0.8. Winding eddy-current losses are increased by a harmonic exponent factor of 2. The factor of 0.8 or
less has been verified by studies by manufacturers in the IEC development and has been accepted in
ANSI/IEEE C57.18.10. Other stray losses occur in core clamping structures, tank walls, or enclosure walls.
On the other hand, current-carrying conductors are more susceptible to heating effects due to the skin
effect of the materials. Either the harmonic spectrum or the harmonic-loss factor must be supplied by
the specifying engineer to the transformer manufacturer.
2.4.3 Rectifier Circuits
Rectifier circuits often utilize multiple-circuit windings in transformers. This is done to minimize harmonics
on the system or to subdivide the rectifier circuit to reduce current or voltage to the rectifier.
Since different windings experience different harmonics on multiple-circuit transformers, the kVA ratings
of the windings do not add arithmetically. Rather, they are rated on the basis of the rms current carried
by the winding. This is an area where the rms-kVA rating of the windings is important. The fundamentalkVA
rating would add arithmetically, since harmonics would not be a factor.
Rectifier circuits can be either single way or double way. Single-way circuits fire only on one side of
the waveform and therefore deliver dc. Double-way circuits fire on both side of the waveform.
A variety of common rectifier circuits are shown in Figure 2.4.1 through Figure 2.4.11. Table 9 in
ANSI/IEEE C57.18.10 shows the properties of the more common circuits, including the currents and
voltages of the windings. The dc winding — the winding connected to the rectifier — is usually the
secondary winding, unless it is an inverter transformer. The ac winding — the winding connected to the
system — is usually the primary winding, unless it is an inverter transformer.
© 2004 by CRC Press LLC
© 2004 by CRC Press LLC