You are on page 1of 9

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 31, NO.

5, OCTOBER 2016 1971

A Method to Estimate the Impact of Harmonic


and Unbalanced Currents on the Ampacity of
Concentric Neutral Cables
Jing Yong, Member, IEEE, and Wilsun Xu, Fellow, IEEE

AbstractConcentric neutral cables are commonly used in and concentric wires serving as the neutral conductor. Three
medium-voltage power distribution systems. Presently, the cable concentric neutral cables form a three-phase, multigrounded
ampacity calculation methods can take into account only the underground distribution feeder. Unfortunately, the published
fundamental frequency current and the harmonic currents in
the phase conductors. In reality, the cable neutral can carry cable derating methods shown in [1][7] cannot be applied to
fundamental frequency and harmonic currents (induced and such cases, as the concentric neutral cable cannot be treated as a
unbalanced), leading to a further increase of cable temperature. single-core cable with a metallic sheath due to its different op-
In view of the increased harmonics and load imbalance in power erating arrangements. In a multigrounded underground feeder,
distribution systems and the need to determine their impact on the loads are connected between the (concentric) neutral and the
concentric neutral cables, this paper presents a method to estimate
the ampacity of concentric neutral cables, especially for multi- phase conductors. This makes the neutral carry (unbalanced or
grounded neutral systems. The proposed cable derating factor zero sequence) load current, in addition to the induced current.
can be used to determine the increase in cable loading caused by On the other hand, the sheath of a cable is not connected to the
unbalanced load and harmonics. A chart is also presented as a sim- load. As a result, the metallic sheath carries only the induced
plied method for cable ampacity and cable loading estimation, current. In addition, the unbalanced loads may induce more
for which only the total harmonic distortion and zero-sequence
current ratio of the phase current are needed. current to the neutral in comparison with the balanced load
case. Therefore, the unbalanced load currents and the induced
Index TermsCable ampacity and concentric neutral cable, har-
currents (at the fundamental and harmonic frequencies) must
monics, power quality (PQ).
be considered together to derive a derating method for the
concentric neutral cables. To the authors' best knowledge,
I. INTRODUCTION little research has been published in the subject area. In view
of the widespread adoption of harmonic-producing loads in
power systems, it has become timely and important to research

I N THE past 20+ years, nonlinear devices have become a


signicant part of the electrical loads in industrial, commer-
cial, and residential power systems. Nonlinear loads inject har-
the derating problem of concentric neutral cables. The work
presented in this paper is an attempt to solve this problem,
and it can be considered as an extension of the concepts and
monic currents into the phase conductor, resulting in additional methods reported in [2].
conductor heating and temperature increase. Research has been
conducted on cable derating under harmonic conditions [1][3]. II. PROPOSED DERATING FACTOR FOR THE CONCENTRIC
The published methods are able to estimate the impact of phase NEUTRAL CABLE SYSTEM
harmonics on the ampacity of single conductor cables. Refer- A lot of research has been done in the eld on calculating
ences [3] and [4] further presented methods to estimate the in- the ampacity of cables [1][15]. Among them, the NeherMc-
duced current on the cable sheath. Grath model has become widely accepted and has formed the
In North America, concentric neutral cables are commonly theoretical basis of the cable ampacity methods in IEC, IEEE,
used in power distribution systems. A concentric neutral cable and NEC standards [5][7]. The NeherMcGrath model con-
consists of a single-core conductor, an XLPE insulation layer, sidered the steady-state load conditions and involved the appli-
cation of the thermal equivalent of Ohm's and Kirchoff's Laws.
The model assumed that all cables in the systems carry the same
Manuscript received December 23, 2013; revised April 29, 2014; accepted
June 16, 2014. Date of publication July 18, 2016; date of current version
fundamental frequency current and operate under the same load
September 21, 2016. This work was supported in part by NFC (50177138) and cycle.
in part by the 111 project under Grant B08036, Chongqing University. Paper Based on the aforementioned theory, IEC standard (60287-
no. TPWRD-01443-2013.
J. Yong is with the State Key Laboratory of Power Transmission Equipment
1-1) established a set of comprehensive cable ampacity calcu-
& System Security and New Technology, Chongqing University, Chongqing lation equations, including considering the heating effect of the
400044, China (e-mail: yongjing@cqu.edu.cn). fundamental frequency current in the cable sheath. One of the
W. Xu is with the Department of Electrical and Computer Engineering, Uni-
versity of Alberta, Edmonton, AB T6G 2V4 Canada (e-mail: wxu@ualberta.ca).
key equations of the IEC method is shown as
Color versions of one or more of the gures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identier 10.1109/TPWRD.2014.2331852 (1)

0885-8977 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
1972 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 31, NO. 5, OCTOBER 2016

where is the permissible temperature rise above the am- A. Proposed Derating Factor for Concentric Neutral Cable
bient temperature; is the cable ampacity (fundamental
current in the phase conductor); is the ac resistance The structure of a single conductor cable with metallic sheath
of the phase conductor at the fundamental frequency; is the is similar to that of the concentric neutral cable except the con-
ratio of loss in the metallic sheath to the loss in the phase con- centric neutral is connected to the load and carries load current.
ductor; are the thermal resistances between the Therefore, in (1), replacing power losses on the sheath with the
phase conductor (sheath) and the surrounding medium; power losses on concentric neutral, (3) must hold true.
is the temperature increase caused by the dielectric loss, solar
radiation, and so on. The rst and second items in (1) are ac- (3)
tually the temperature increases caused by power losses on the
phase conductor and the sheath, respectively. In fact, in this where and are the power losses on the phase con-
equation considers the derating impact of the metallic sheath on ductor and concentric neutral conductor, respectively.
the cable ampacity. Since is the temperature rise caused by the dielectric
Although (1) can include the cable sheath, it cannot be di- loss which is determined by voltage, solar radiation, and so on,
rectly used for concentric neutral cables. A concentric neutral it is reasonable to assume is independent of load. In addi-
cable has loads connected between the phase conductor and the tion, since the phase conductor is located in the center of cable,
concentric neutral whereas the sheath does not have loads con- the thermal resistance between the phase conductor and the sur-
nected to it. In addition, the equation has not included the effect rounding medium is bigger than that between the sheath
of harmonic currents. and the surrounding medium . It is also acceptable to as-
Separately, [1] and [2] proposed a method to consider the ef- sume for conservative estimation, which is pre-
fect of harmonic currents on single-core, nonconcentric neutral ferred when dealing with loading issues.
cables. An index called harmonic derating factor was As a result, the maximum permissible power losses of
established as follows: concentric neutral cable can be obtained by

(2)
(4)

For the generic unbalanced three-phase loads containing har-


where is the equivalent fundamental current, which pro- monics, the power losses on the phase (or concentric neutral)
duces the same power loss as the distorted current does, conductor consist of two parts: one is caused by the fundamental
is the individual harmonic distortion (IHD) of the actual cur- frequency current (or ) and the other one is
rent, is the fundamental frequency component of caused by harmonics (or ). Therefore, the total
the distorted current, and are the ac resis- power loss can be expressed by
tances of the conductor at the fundamental frequency and the th
harmonic in ohms per unit length, and is the maximum har-
monic order of interest. can be calculated according (5)
to equations presented in [1], which include the skin effect and
proximity effect of other conductors. In (5)
Equation (2) assumed that the power loss caused by distorted
current is equal to the power loss caused by an equivalent funda-
(6)
mental current. If the cable ampacity is adopted as ,
the calculated actual fundamental current by (2) is (7)
the derated cable ampacity considering harmonic impact. How-
(8)
ever, the aforementioned harmonic derating factor was devel-
oped only for single-core, nonconcentric neutral cables.
It can be seen that there is a gap in cable ampacity calcu- (9)
lation, that is, how to estimate the ampacity or derating factor
for concentric neutrals under unbalanced and harmonic current
conditions. In view of the fact that concentric neutral cables are where and are the th harmonic currents in the
widely used in distribution systems where unbalanced and har- phase and neutral conductors, respectively, and are
monic currents are a norm rather than an exception, it is impor- the ac resistances of the neutral conductor at the th harmonic.
tant to address this knowledge gap. To this end, the following Subscript can be A, B, or C to indicate phases A, B, or C.
research works are conducted and presented in this paper: Since the concentric neutral behaves as the metallic sheath
proposition of a derating factor for concentric neutral cable under balanced load and harmonic-free conditions, employing
based on the IEC method and those in [1] and [2]; as the equivalent fundamental current, which produces the
development of a method to estimate the neutral current of total power loss as the actual current does, we have
the cable based on its three-phase currents and grounding
conditions. (10)
YONG AND XU: METHOD TO ESTIMATE THE IMPACT OF HARMONIC AND UNBALANCED CURRENTS 1973

Extending the concept of (2) to concentric neutral cable, the


of (5) and (11) must be equal. It brings a new derating
factor

(11)

where
Fig. 1. Typical distribution feeder.
.
In the above equation
approximately equal to the power loss on concentric neu-
tral when the loads are three-phase balanced and harmonic
free. Therefore

(12) (15)

where is the individual harmonic distortion (IHD) of Thus, in (14).


the phase current, is the concentric neutral current The loading increase caused by harmonics can only be es-
IHD normalized to the corresponding fundamental phase timated by
current. and can be considered as the normalized
th harmonic ac conductor resistances for phase and neutral, (16)
respectively.
To compute the index, one needs to know the neutral where is the equivalent harmonics in phase and con-
current . However, this current cannot be measured, so it has centric neutral conductors, and
to be estimated. The method for neutral current estimation will
be presented in Section III.

B. Applications of
From (13), we have
Before explaining the methods to determine the additional
power loss, it is useful to discuss the potential applications of
. has three applications: (17)
1) Select cable size. The new derating factor can be used at the Since , and in practice, (18) can be
system design stage to consider the effects of harmonics easily proved mathematically.
and load imbalance on cable-size selection. 3) Estimate the equivalent cable loading level. The equivalent
2) Quantify the increase in cable loading caused by load im- loading level can be calculated by (19) to esti-
balance and harmonics. The inverse of represents mate the actual loading level
the total cable loading normalized to the fundamental fre-
quency current loading. Therefore, the increase of normal- (18)
ized cable loading , called loading increase here-
inafter, can be obtained as follows: where is the cable ampacity provided by the manu-
facturer. is the fundamental loading level. The
(13) equivalent loading level is very helpful for utilities to per-
form estimates of the online load condition since
The impacts of load imbalance and harmonics on the (1) can be measured directly. In order to prevent the cable
loading increase can be estimated separately by (14) and from overheating, the maximum permissible
(16). is 1.0 p.u. and the higher values imply an overloading
The loading increase caused by fundamental load imbal- situation.
ance only can be estimated by (15)
III. DETERMINATION OF THE NEUTRAL CURRENT
(14) Fig. 1 illustrates a typical distribution feeder using multi-
grounded concentric neutral cable. is the equivalent system
where is the equivalent unbalanced fundamental impedance. and in this gure represent the grounding
current in concentric neutral. The power loss on neutral resistances of the substation and neutral, respectively. In most
caused by the fundamental current is con- practices, is much smaller than . Since some of the
tributed by induced current and partially triplex harmonics and unbalanced load current will return
unbalanced current return . is through the ground resistance and the amount of return current
1974 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 31, NO. 5, OCTOBER 2016

Fig. 3. Equivalent circuit of the MGN system for neutral current calculation.

where is the zero-sequence current

Fig. 2. Equivalent circuit of three neutral conductors. (24)

is location dependent, the neutral current at each feeder span is Since all neutrals are bonded at two ends of a span, the fol-
different as well. lowing equation holds true:
The goal of the neutral current estimation is to determine the
highest neutral current among all locations, based on known (25)
three-phase currents (magnitudes and phase angles), the feeder
Solving (19)(25) for yields
impedance parameters, and the cable conguration data.
It is worth noting that [3] and [4] presented methods to es-
timate the cable sheath current. The method assumes that the
sheath is solidly grounded to remote earth (i.e., 0). This
assumption signicantly simplies the problem, but it can be
too conservative, that is, it can overestimate the sheath current. (26)
Such a conservative approach cannot be applied to the case of Equation (26) reveals that the concentric neutral current de-
the concentric neutral cable. This is because the unbalanced cur-
termined not only the inductive coupling, but also the unbal-
rent will ow into the ground instead of into the cable neutral if
anced load current. So far, the remaining problem is how to de-
0, which will result in an underestimate of the current in
termine the total neutral current .
cable neutral.
B. Determination of Total Neutral Current
A. Derivation of Neutral Current of Concentric Neutral Cable
Assume each neutral span has the same length and each neu-
The cable system can be installed either in trefoil or in par- tral grounding resistance is equal. The equivalent circuit
allel, and the cables can be placed either in a touching formation for a multigrounded neutral (MGN) system can be illustrated in
or separately. The trefoil and one-cable-diameter separation in- Fig. 3. In the gure, indicates the unbalanced load cur-
stallation has been considered in the following analysis.
rent, that is, three times that of zero-sequence current, injected
A detailed circuit for a span of cable neutral is shown in Fig. 2.
In the gure, , , and are induced into node .
voltages from the other conductors (both phase and neutral) in For the cascade circuit, one can obtain the total equivalent
the cable system. is the self-impedance of each neutral con- impedance of the three concentric neutral based on
ductor while is the total neutral current, which is the phasor Fig. 2 for the th span
summation of three neutral currents in three neutral conductors

(19)
(27)
The voltage drops , , and on the neutral where is one-third of the summation of the three
span for each neutral conductor can be expressed as phase load currents corresponding to span

(20)
(21)
(22) By establishing the node and loop equations, the equation in
matrix form shown in (22) can be obtained
Since the three cables are identical and trefoil installed, the
mutual impedances between the phase and neutral in the same (28)
cable (denoted as ), the mutual impedances between
the phase and neutral in different cable (denoted as ), where as shown in the equation at the bottom of the next page.
and the mutual impedances between the neutrals of different Let , and we have
cable (denoted as ) are equal, respectively. Therefore,
the induced current on phase can be expressed by (29)

Obviously, the unbalanced current is the essential driving


(23) force of the total neutral current. In order to estimate the ratio of
YONG AND XU: METHOD TO ESTIMATE THE IMPACT OF HARMONIC AND UNBALANCED CURRENTS 1975

the total neutral current to the unbalanced three-phase current, TABLE I


, called unbalanced current return ratio, is dened as SYSTEM PARAMETERS

(30)

When loads are evenly distributed from node to ,


can be obtained by (29)

(31)
It can be seen that is determined only by the param- In summary, the steps in determining the neutral current of
eters of power line and ground resistances, that is, , each phase are as follows.
, , and and the rst load location . Equation 1) Measure the three-phase currents experienced by a cable
(31) suggests that for a given concentric neutral cable system where the cable is already in operation. For the planned
with evenly distributed loads, the total neutral current of span cable installation, the three-phase currents are calculated
can be obtained by multiplying the summation of three-phase using distribution power-ow programs.
currents by , and the unavailable neutral current can 2) Calculate zero-sequence harmonic currents from the
be obtained by (26). aforementioned results.
Sensitivity studies have been performed to investigate the 3) Estimate the neutral current of each cable by using (31).
variation of the with the above factors. The results With the calculated neutral current, the proposed new der-
are presented in the Appendix. The main conclusions are drawn ating factor can be obtained by (11). More indices can
as follows. also be calculated to estimate the impact of harmonics and
The unbalanced current return ratio is among 0.9 unbalanced loads on the ampacity of concentric neutral cable
0.98 for the typical system of which the parameters are shown systems.
in Table I. The range of does not have a signi- More complicated calculations for the neutral current, as-
cant change with the substation and neutral grounding resis- suming the three-phase cables are in parallel installation, using
tances and the separation distance of cables. It will be slightly the aforementioned process but considering the actual mutual
higher in harmonics. The highest value of occurs impedances, are also done. The results show that the different
when the loads are intensively located at the end of the feeder. installations do not impact the neutral current signicantly.
For a conservative estimation, it is reasonable to assume that
and and are in phase. Thus, the IV. EXAMPLE CABLE DERATING RESULTS
neutral current can be obtained through
In order to demonstrate the impact of harmonics and unbal-
anced load on the concentric neutral cable ampacity, case studies
using actual eld data have been conducted. Those case studies
are based on measurements taken from 15 actual feeders, which
(32) supply residential and industrial loads. In the following case
1976 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 31, NO. 5, OCTOBER 2016

TABLE II
NORMALIZED AC RESISTANCES OF PHASE AND NEUTRAL CONDUCTORS

studies, the basic parameters of the cable feeder are shown in


Table I. To consider the skin effect and proximity effect of
other conductors, the ac resistances of phase and neutral con-
ductors at each harmonic frequency were calculated by using the
method proposed in [1]. Since the neutral conductor of the con- Fig. 4. Averaged spectrum for Phase A current of each case.
centric neutral cable is always made of many small conductors
of which skin effect and proximity effect of other conductors can
be neglected, the ac resistance of the neutral conductor at each
frequency can be treated as equal and considered as .
can be considered as a constant value. The normalized ac
resistances of phase and neutral conductors are listed in Table II.
The one-day averaged spectra for phase A and zero se-
quence currents of 15 cases are shown in Fig. 4 and Fig. 5. The
averaged s of each case are in the range of 2.5 12% and
listed in the legend of Fig. 4. Employing the ratio of fundamental
zero sequence current to positive sequence current to
represent the load imbalance rate, the load imbalance rates of
these 15 cases are less than 4.5%.

A. Impact of Load Imbalance and THD on the Loading Fig. 5. Averaged spectrum for zero sequence harmonic current of each
case.
Increase
In order to know the impacts of various unbalanced and THD
loads, the relationships between the load imbalance rate and the
corresponding loading increase and between the THDs of loads
and the corresponding loading increase of the 15 cases have
been investigated.
The results for various feeders are sorted by load imbalance
rate and THD and presented in Fig. 6. The left gure
in Fig. 6 presents the impact of load imbalance on the loading
increase, and the right one presents the impact of THD on the
loading increase. The total loading increase sorted by and
THD is also shown in the gure for comparison.
In this measured load imbalance and phase-current Fig. 6. Comparison of loading increase for different THD and .
range, the following conclusions can be drawn.
The load imbalance has a slightly signicant impact on
the loading increase in comparison with the harmonics,
especially in the low THD range.
Higher THD leads to closer contributions caused by har-
monics and load imbalance.
The above conclusions imply that if it's necessary to consider
the impact of harmonics, the load imbalance has to be consid-
ered in estimating the level of loading.

B. Variation of Loading Level and the Loading


Increase During One Day
Fig. 7 (left) presents the comparison of equivalent loading
level for an actual load to that for the corresponding funda-
mental current. It provides a general scenario of the increase in
loading level during one day because of the effects of harmonics
and load imbalance. Fig. 7 shows that the increase in equivalent
loading level is not very signicant for an actual feeder. Fig. 7. Variation of the loading level and the equivalent loading increase.
YONG AND XU: METHOD TO ESTIMATE THE IMPACT OF HARMONIC AND UNBALANCED CURRENTS 1977

Fig. 8. Comparison of loading increase for different THDs and load


imbalances.

The amounts of loading increase, , ,


, are shown in Fig. 7 (right). Fig. 9. Contours of increase in cable loading .
The actual cases show that the concentric neutral cable
doesn't cause a signicant increase in loading since the phase
current and the rate of load imbalance of those measured sequence current, which will partially return through the neutral
feeders are relatively low. If the THD or the rate of load and can be calculated by the known phase current. In this case,
imbalance gets higher, the concentric neutral cables could be
overloaded even more.

V. SENSITIVITY STUDIES Fig. 8 (right) shows the comparison of for different


THDs (phase current). The phase current THD and the loading
For the medium voltage distribution systems, especially those increase don't follow a linear relationship. When the THD gets
supplying residential loads, the harmonic current distortion be- higher, the loading increase will increase rapidly. Once the THD
comes severe, and the balanced load condition is rare to achieve. of load reaches 28%, the maximum possible loading increase
This section investigates the impacts of highly unbalanced load caused by the concentric neutral cable can be over 10%. This
and severe harmonic distortion on the cable loading increase, amount of loading increase cannot be neglected as well.
separately.
C. Contour of the Increase in Cable Loading
A. Highly Unbalanced Load Cases Without Harmonics
In actual distribution systems, both load imbalance and har-
This study assumes that there are only unbalanced loads monics contribute to the loading increase.
without harmonics in the current. Obviously in this case, there Loading increase in Fig. 8 (left) can be considered
is no contribution of harmonics to the cable loading. In this case as in actual systems, since only the unbalanced fun-
damental frequency current is taken into account. Meanwhile,
loading increase in Fig. 8 (right) can be considered as
A comparison of for variation of the rate of load im- in actual systems, ignoring unbalanced non-triplex
balance from 0 30% is shown in Fig. 8 (left). It can harmonics which are generally very small in distribution sys-
be seen that when the load imbalance gets bigger, its impact on tems in comparison with the triplex harmonics and harmonics
loading increase becomes noticeable. The rate of load imbal- owing in phase. Therefore, as illustrated in (16), to obtain the
ance and the loading increase almost follow a linear relation- loading increase caused by both harmonic and funda-
ship. When the rate of load imbalance reaches 20%, the loading mental load imbalance, one can simply add up the loading in-
increase will be over 10%. In transmission systems, such high crease values from Fig. 8 (left) considering only and those
load imbalance is not the case. However, in distribution systems, from Fig. 8 (right) considering only.
a load imbalance with a rate of 20% is not impossible, especially For more convenient use, assuming a conservative current
during some abnormal operation conditions. return ratio of , the contours were obtained
by combining Fig. 8 (left) and Fig. 8 (right) and is shown
B. High THD Cases With Balanced Loads in Fig. 9. In this gure, the x-axis stands for the rate of load
To investigate the impact of higher THD loads on the in- imbalance in percentage, and the y-axis stands for the
crease in equivalent loading, the THDs of loads are extended THD of the phase current in percentage. The number on the
to a higher range by simply magnifying the harmonic magni- contour curve stands for the amount of loading increase (i.e.,
tudes to 2.5 times of the previous values for the above 15 cases. in percentage) caused by the two factors. This chart
In this way, the proportion among harmonics is still practical. can be used as a simplied tool for utility engineers to estimate
The THDs of those hypothetical loads are distributed in a wide cable overloading. For example, a load with 15% THD and
range of 4.39 29.3%. Loads with those THDs are also reason- 15% may lead to 11% increase in loading. It means that
able and may occur in some situations. In the study, the cable if the loading level with regard to fundamental current is 80%,
feeder is the same as the previous one, and loads are considered the actual loading level is 88.8%. It also suggests that if one
as balanced. Therefore, only triplex harmonics exist in the zero intends to select a cable to supply for this type of load, the cable
1978 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 31, NO. 5, OCTOBER 2016

Fig. 10. Ratios for different load distributions. Fig. 11. Ratios for different load distributions with different cable separation
distances.

ampacity should have a derating of 0.90, that is, a larger cable


may need to be selected. 3) The length of span does not impact signicantly.
4) When the separation distance between cables gets bigger,
the mutual impedance will be reduced. Fig. 11 provides the
VI. CONCLUSION
results when the separation distance of cables is 10 times
A new derating factor has been proposed to evaluate the im- of the cable's diameter. It shows that the current return ratio
pact of harmonics and unbalanced currents on the ampacity of decreased to 0.8 for most cases; however, when the loads
concentric neutral cables. This index can be used to estimate the are concentrated at the end of feeder, the current return
level of increase in cable loading, to determine if a cable is over- ratio is still very close to 1.
loaded, and to nd the cable derating value for design purposes. Overall, the assumption of is reasonable.
Based on the analytical and sensitivity studies of the index, the
following overall conclusions are drawn:
1) The increase in loading of a concentric neutral cable is
REFERENCES
caused by two factors: the phase current harmonics and the
phase current imbalance. A set of contour curves has been [1] A. P. S. Meliopoulos and M. A. Martin, Jr., Calculation of secondary
cable losses and ampacity in the presence of harmonics, IEEE Trans.
developed to facilitate a quick estimate of increase in cable Power Del., vol. 7, no. 2, pp. 451459, Apr. 1992.
loading caused by these two factors. [2] A. Hiranandani, Calculation of cable ampacities including the effects
2) The results show that moderate phase current harmonics of harmonics, IEEE Ind. Appl. Mag., vol. 4, no. 2, pp. 4251, Mar./
Apr. 1998.
and load imbalance will not increase a cable's loading level [3] G. J. Anders, Rating of Electric Power Cables. New York, USA: Mc-
signicantly. However, the loading level will increase Graw-Hill, 1998, pp. 7-217-24.
rapidly when either harmonics or unbalance increases. [4] Application of Sheath-Bonding Methods for Single Conductor Cables
and the Calculation of Induced Voltages and Currents in Cable
They do not follow a linear relationship. Sheaths, ANSI/IEEE Standard 575-1988.
3) The load imbalance is the dominant factor for the loading [5] Calculation of the Continuous Current Rating of Cables (100% Load
increase in comparison with the phase harmonic currents Factor) (Second Edition), IEC Standard 60287, 1993.
[6] IEEE Recommended Practice for Electric Power Distribution for In-
in a low THD and load imbalance range. dustrial Plants, IEEE Standard 141-1993.
[7] National Electric Code Handbook, NFPA, 2011.
APPENDIX [8] J. H. Neher and M. H. McGrath, The calculation of the temperature
CHARACTERISTICS OF CURRENT RETURN RATIO rise and load capability of cable systems, AIEE Trans., pp. 752773,
Oct. 1957.
Using typical system parameters as shown in Table I, the un- [9] A. Hiranandani, Calculation of conductor temperatures and ampaci-
balanced current return ratio for each span can be ob- ties of cable systems using a generalized nite difference model, IEEE
Trans. Power Del., vol. 6, no. 1, pp. 1524, Jan. 1991.
tained. The results are shown in Fig. 10. in the gure legend [10] G. Engmann, Ampacity of cable in covered tray, IEEE Trans. Power
indicates the start span of the evenly distributed loads. This App. Syst., vol. PAS-103, no. 2, pp. 1524, Feb. 1990.
gure suggests that for a specied load amount, no matter the [11] C. Demoulias, D. P. Labridis, P. S. Dokopoulos, and K. Gouramanis,
Ampacity of low-voltage power cables under non-sinusoidal cur-
starting point of load connecting, the ratio does not have rents, IEEE Trans. Power Del., vol. 22, no. 1, pp. 584594, Jan. 2007.
much variation and can be considered as 1 for conservative [12] J. J. desmet, D. J. Putman, G. M. Vanalme, and R. J. Belmans, Mod-
estimation. eling and sensitivity analysis of the thermal behavior of cable for dif-
ferent current conditions, in Proc. 11th Int. Conf. Harmonics Quality
is determined by ground resistances, the cable imped- Power, 2004, pp. 471476.
ances, the harmonic order and the length of span. The impacts [13] Y. G. Sahin and F. Aras, Investigation of harmonic effects on under-
of those factors are summarized as follows. ground power cables, in Proc. POWERENG, 2007, pp. 589594.
[14] J. A Palmer, R. C. Degeneff, T. M. Mckeman, and T. M. Halleran,
1) The grounding resistances at substation and feeder and Pipe-type cable ampacities in the presence of harmonics, IEEE
do not impact signicantly. Trans. Power Del., vol. 8, no. 4, pp. 16891695, Oct. 1993.
2) Values of at harmonic frequencies slightly increase [15] G. Carpinelli, A. Andreotti, P. Caramia, and P. Verde, On four-con-
ductors low voltage cables ampacity in balanced and unbalanced
with the harmonic order. When the harmonic order gets non-sinusoidal conditions, in Proc. 11th Int. Conf. Harmonics Quality
higher, for this harmonic almost reaches 1. Power, 2004, pp. 442449.
YONG AND XU: METHOD TO ESTIMATE THE IMPACT OF HARMONIC AND UNBALANCED CURRENTS 1979

Jing Yong (M08) received the B.Sc., M.Sc., and Ph.D. degrees in electrical Wilsun Xu (M90SM95F05) received the Ph.D. degree from the University
engineering from Chongqing University, Chongqing, China, in 1985, 1988, and of British Columbia, Vancouver, BC, Canada in 1989.
2007, respectively. He was an Engineer with BC Hydro, Burnaby, BC, Canada, from 1990 to
Currently, she is a Professor at Chongqing University and a Visiting Professor 1996. Currently, he is a Professor and an NSERC/iCORE Industrial Research
at the University of Alberta, Edmonton, AB, Canada. Her research interests are Chair at the University of Alberta, Edmonton, AB, Canada. His research inter-
the analysis of power distribution systems and power quality. ests are power quality and harmonics.

You might also like