Documentos de Académico
Documentos de Profesional
Documentos de Cultura
6, NOVEMBER/DECEMBER 1973
can be resolved into a Fourier series which contains sinusoidal Fig. 1. Fundamental and third harmonic frequency voltage wave-
terms of the fundamental and all odd-order (3, 5, 7, 9, etc.) forms on three-phase system. Note that phase one voltage third
harmonic frequencies. Each of these sinusoidal flux compo- harmonic is in time-phase with both the phase two and phase three
harmonic wave.
nents gives rise to a voltage, at its frequency, in the stator of
the machine. Some of the harmonic frequency voltages can
cause nuisance circulating currents. The effect of the fifth harmonic voltage in contrast, appears
The magnitude of harmonic voltage is a function of generator in the negative sequence circuit, and of the seventh harmonic
loading as well as machine design. For example, a round rotor voltage, in the positive sequence circuit; both consist of three
machine may produce two to five percent third harmonic phase-to-neutral components equal in magnitude but displaced
voltage (in percent of rated line-to-neutral voltage) at no load, 120 electrical degrees apart on the fundamental scale. Thus
and four to ten percent at full load. The harmonic content of they form a balanced set of voltage phasors; any currents these
salient pole machines will, in general, be greater than that of voltages would drive in a balanced system would add vector-
round rotor generators. For any specific machine, an estimate ially to zero at the generator neutral. Also, the magnitude of
of the harmonic voltage can be obtained from the manufac- the higher order (9, 15, etc.) harmonic voltages in the zero
turer of the generator. sequence circuit is usually small. Moreover, since inductive
In general, only the third harmonic voltage can cause appre- reactance is proportional to frequency, the magnitude of the
ciable circulating currents in the neutral of the power system. zero sequence impedances in the system and in the neutral
There are several reasons for this. Equal third harmonic circuit at the higher harmonic frequencies is relatively great.
voltages exist in each of the three phases. While the funda- These effects combined mean that the circulating currents
mental frequency voltages in the three phases are displaced driven by higher order harmonic voltages within the machine
120 electrical degrees in time-phase, the third harmonic volt- itself can usually be neglected.
ages are displaced 3 X 120 = 360 electrical degrees (see Fig. 1). This is not to say that the existence of the fifth, seventh, or
Thus the three third harmonic phase voltages are in phase higher order harmonic currents should be ignored. The
with each other and their effect is felt only in the zero presence of higher order harmonic currents may reduce the
sequence circuit where it appears in the form of a circulating fundamental frequency thermal load capability of a generator
current at the third harmonic frequency. The magnitude of and produce difficulties with internal heating, including rotor
this current is determined by the third harmonic driving voltage heating problems which are not significantly present with
and the third harmonic impedence of the zero sequence circuit. balanced third harmonic currents alone. Concern for the
higher order harmonics usually comes about when the mill
Paper TOD-73-63, approved by the Pulp and Paper Industry Com- power system voltage waveform is distorted by the presence of
mittee of the IEEE Industry Applications Society for presentation at
the Pulp and Paper Industry Technical Conference, Jacksonville, Fla., harmonic-generating loads such as phase-controlled rectifiers.
May 1-4, 1973. Manuscript released for publication August 6, 1973. Whenever applications ofthis sort are contemplated, the manu-
The author is with The General Electric Company, Schenectady, N.Y.
12345. facturer of the machine should be consulted.
POWELL: THIRD HARMONIC CIRCULATING CURRENTS 673
TABLE I
COMPARISON OF NEUTRAL GROUNDING METHODS
System Grounding Figure Fault Current Characteristics Magnitude of Expected Comments
Method of Grounding Method Third Haromics Currents
Individual Resistance 3 (a) Relatively low magnitude of Negligible SmalL A Sndll COnItillUOUS
Grounding of each ground fault current is ample current will exist in
Source for selective relaying yet the neutral resistors.
reduces fault point damage.
Individual Reactance 4 (a) Required when phase-to-neutral Appreciable May .equire derdtiniq of
Grounding of each connected single-phase loading generators and traris-
Generator is needed. Reactor must be formers, special reactor
sized to limit ground fault ratings, and desensitized
current from the generator ground overcurrent
to no greater than the three relays.
phase fault current available
from the generator.
Solid Grounding of 5(a) Ground fault magnitude may be Appreciable y require deratinig Of
Generators significantly greater than generators and trans-
three phase fault magnitude. formers and desensitized
ground overcurrent
relays. Should be
avoided.
Group Resistance 6(a) Fault current to single Appreciable May require desensitizinig
Grounding of phase to ground faults the otf ground relays and
Generators through same or less than in derating of generators.
Common Neutral individual resistance Requires means to dis-
Resistors. grounded system. conniiect out-of-service
generators from neutral
bus. Should be avoided.
Group Resistance Fault current to single Negligibly Snmall A smnall continuous currerit
Grounding of phase to ground faults the will exist in the neutraI.
Generators through same or less than in resistors.
Switched Neutral individual resistance
Resistor grounded system.
=
0.11
0.065
on generator thermal capabilities and the continuous current E,3 = 0.05 at rno load
Eo3 = 0.10 at full load
ratings of neutral grounding device, and since their presence
can lead to false relay operations or necessitate desensitizing
essential protective equipment, it is important to recognize 20 MVA self-cooled rating
X1 = 6.5%
ways in which the magnitude of such currents can be con- X0 = 5.5%
trolled. The remainder of this paper will deal with that subject.
20 ampere, 10 second, for use in
LIMITING THIRD HARMONIC NEUTRAL CURRENTS a system rated 13.8 kV line-to-line
harmonic impedance of the power system zero sequence net- Typically, resistors will be selected to limit the maximum
work. The only system components that can be conveniently single-phase-to-ground fault to 400 to 2000 A. Low-resistance
manipulated to obtain a desired magnitude of zero sequence grounding offers ample fault current for selective relaying with
impedance are the neutral grounding devices. Thus, the a minimum risk of damage to motors or generators with inter-
method employed to ground the power system neutral is the nal faults.
key to controlling the magnitude of third harmonic current. The third harmonic zero sequjence circuit for the simple
Several grounding techniques are in common use in power system of Fig. 3(a) is shown in Fig. 3(b). Note that the zero
systems; these will be discussed separately. Table I summarizes sequence transformer and generator fundamental frequency
the conclusions reached for each grounding method. reactances must be multiplied by three to account for their
frequency dependence. Note also that the neutral resistances
Low-Resistance Grounding must be muiltiplied by three to convert to zero sequence
The most common approach to medium-voltage system quantities because they are connected in the system neutral.
grounding is to connect a resistor between each wye-connected In Fig. 3(b) two interesting facts can be observed.
POWELL: THIRD HARMONIC CIRCULATING CURRENTS 675
TABLE III
CALCULATED THIRD HARMONIC CURRENTS IN TYPICAL INDUSTRIAL
POWER SYSTEM ARRANGEMENTS
System Grounding Figure Both Generators at Full Load One Generator Fully Loaded - One Unloaded (Note 3)
Arrangement Generators Transformer Loaded Generator Unloaded Generator Transformer
(Note 1) Stator Neutral Phase Neutral Stator Neutral Stator Neutral Phase Neutral
Current Current Current Current Current Current Current Current Current Current
Individual resistance 3(a) 4.42A 13.26A 8.84A 26.52A 11.2A 33.6A 8.81A 26.43A 2.21A 6.63A
grounding of each
source
Individual reactance 4(a) 129A 387A 258A 774A 229A 687A 165A 495A 64.5A 193.5A
grounding of each
generator (Note 2)
Solid grounding of 5(a) 172.5A 517.5A 345A 1035A 362A 1086A 282A 846A 79.6A 238.8A
generators
Group resistance 6(a) 3.315A 19.95A 6.65A 19.95A 323A 969A 322A 965A 1.65A 4.95A
grounding of
generators through
common resistor
Group resistance 7 6.64A 19.2A 6.64A 19.2A 6.64A 19.2A 0 0 6.64A 19.2A
grounding of
generators through
switched resistor
WA
J1~A
r
(a)
(a)
j3Xo {J3XO2 j 3XOT
(b) b)
Fig. 4. (a) Reactance grounded power system. (b) Third harmonic Fig. 5. (a) Solidly grounded power system. (b) Third harmonic zero
zero sequence equivalent circuit. sequence equivalent circuit.
Resistor
f- rA
Resistor
(a)
(b)
Fig. 6. (a) Group resistance grounded power system. (b) Third har-
monic zero sequence equivalent circuit. Fig. 7. Preferred method of group resistance grounding.
POWELL: THIRD HARMONIC CIRCULATING CURRENTS 677
RE FERENCES third harmonics will flow; or, referring to Fig. 2 of the paper,
[11 Standard Requirements, Terminology and Test Procedures for break the Io circuit without otherwise jeopardizing fault detec-
Neutral Grounding Devices, IEEE Standard 32, 1972 tion equipment.
[21 "Grounding of power system neutrals," in Electrical Transmissionz If generator is provided with a unit transformer, a delta/
a
and Distribution Reference Book, Westinghouse Electric Corp., E.
Pittsburgh, Pa., 1950, ch. 13. wye connection (delta being on the generator side) means no
[31 G. Angst and J. L. Oldenkamp, "Third-harmonic voltage genera- third harmonics will flow.
tion in salient-pole, synchronous machines," AIEE Trans. (Power
App. Syst.), vol. 75, pp. 434-441, June 1956. Another alternative has been used on several occasions where
[41 D. Gingsberg and A. L. Jokl, "Voltage harmonics of salient-pole generators are to be paralleled through a common bus. The
generators under balanced 3-phase loads - II," AIEE Trans. (Power
App. Syst.), vol. 79, pp. 560-566, Aug. 1960. generator neutral is ungrounded providing for ground fault
[5] T. W. Schroeder, "Power system harmonics," Gen. Elec. Rev., detection through insertion of a synthetic ground point, gener-
vol. 45, pp. 399402, July 1942.
[61 C. E. Webb, "Determining the rating of a generator neutral ground- ally consisting of a zig-zag grounding transformer giving the
ing reactor," Industrial Power Systems Magazine, Schenectady, same performance to the network as a low resistance neutral
N.Y., vol. 13, pp. 16-17, Dec. 1970.
point ground.
Although the use of a synthetic ground poinlt is not
mentioned in the paper, we have found it will generally meet
the requirements with relatively modest cost, allowing 5/6-
Discussion winding pitch generators to be applhed in parallel with 2/3-
pitch generators without any need to modify existing ground
John E. McConnell (ASEA Inc., White Plains, N.Y.): The fault detection and relaying practices commonly used in North
author has put forth a very comprehensive paper on influence America.
of third harmonic currents on selecting neutral grounding
devices. I believe it appropriate to share ASEA experiences on
this topic.
ASEA builds generators using a 5/6-armature winding pitch.
The practice in the United States and the United Kingdom has Louie J. Powell, Jr.: The author wishes to thank Mr. Mc-
been to use a 2/3 pitch. The 5/6-pitch generators normally Connell for his compliments and his comments on grounding
produce a higher third harmonic voltage than the 2/3-pitch practices in Europe.
winding. Mr. McConnell is wrong in stating that the practice among
The common grounding practice in Europe for industrial generator manufacturers in this country has been to use a 2/3-
power systems is to use high resistance grounding with sensi- winding pitch. To the contrary, except in special instances all
tive ground relays. Thus the third harmonic will be suppressed. domestic generator builders use a pitch close to 5/6. It is
In addition, transient overvoltages on arcing grounds will be recognized that the use of 2/3 pitch will eliminate third har-
effectively eliminated. monic waveform distortion almost completely, but such designs
However, the United States and Canada have not readily are also known to invariably be larger, more costly, less effi-
accepted high resistance grounding practices for reasons that cient, and operate at higher field temperatures than machines
probably can be explained by not wishing to change relaying with higher pitch factors. The point of view is taken that
practices, even though ASEA can readily supply the relays controlling currents circulating in the neutral due to third
required for use with high resistance neutral grounding equip- harmonics in the stator voltage is a system consideration
ment. In other occasions, generators having a 5/6-pitch wind- separate from machine design problems and therefore is not
ing may often be applied for parallel operation with generators of primary concern to the generator designer. This is because
that have 2/3-pitch windings, resulting in the possibility of the neutral circulating current problem can be easily handled
ruling out high resistance grounding. Thus it becomes necessary by appropriate system neutral grounding.
to make a more thorough review. Mr. McConnell has noted that the voltage waveform of ASEA
The first step toward investigation was to review the merits generators has an abnormally high third harmonic content and
and reasons for the 5/6-pitch winding. As with most anything, that, as a result, ASEA insists on high resistance grounding of
tradeoffs exist. The result is that ASEA found no compelling the machine neutral or the use of neutral deriving transformers
cause to change to a 2/3 pitch at this time. and resistors on the bus. However, I do not know of any
On the plus side, 5/6 pitch produces much lower fifth and instances in which a domestically produced round-rotor ma-
seventh harmonic voltages that cause eddy currents and heating chine generated sufficient harmonics to necessitate the use of
in the surface of the rotor. In addition, the 2/3-pitch winding special grounding techniques, even though these machines also
utilizes a significant portion of the magnetic flux in the rotor have 5/6 pitch. Thus, I suspect there may be influences other
for suppressing the third harmonic, lowering the efficient use than pitch factor involved in ASEA's choice of a grounding
of the field and detracting by about 12 percent from the method.
utilization of the field to produce the fundamental frequency. All of the other means of grounding Mr. McConnell men-
It is also recognized that other additional alternatives exist tioned are well known in this country, and their perforinance
to eliminating the third harmonic circulating current which in controlling circulating neutral currents can be determined
have not been presented in the author's paper. The principle by extension of the cases considered in the paper. The reason-
involved is to restrain or eliminate the path through which the ing for their use, or lack of use in this country has been wefl
POWELL: THIRD HARMONIC CIRCULATING CURRENTS 679
documented in the technical literature so there is no need to characteristics, safety, economics, and neutral circulating cur-
discuss it further here. rents. Generators with relatively low third harmonic voltage
It will always be the case that selection of neutral grounding output, such as those produced in this country, are more
devices is a system consideration involving an examination of flexible on this basis and can usually be applied with any of
influences such as system load requirements, relaying, generator the grounding techniques available today.
Louie J. Powell, Jr. (S'67-M'69) was born in Jacksonville, Fla. He received the B.E.E. degree,
cooperative plan, from the Georgia Institute of Technology, Atlanta, and has done graduate
work in the Power Systems Engineering Course sponsored jointly by General Electric Company
and Union College.
While in college he was employed by Reynolds, Smith and Hills, Architects and Engineers,
Jacksonville, where he worked on electric utility generating plant and transmission substation
designs. After graduation he joined the General Electric Company and was initially assigned to
the Chicago district office, where his principle work was on electrical systems for large steel-
making facilities. In 1969 he transferred to the Energy Systems Operation in Schenectady,
N.Y., as an Application Engineer, working primarily in the field of power systems analysis. He
has made electrical system studies for plants in the steel, shipbuilding, petroleum, paper, and
electrical manufacturing industries, both in the United States and abroad.