Energy and Power Engineering, 2013, 5, 258-263
doi:10.4236/epe.2013.54B050 Published Online July 2013 (http://www.scirp.org/journal/epe)
Mitigation of Distributed Generation Impact on Protective
Devices in a Distribution Network by Superconducting
Fault Current Limiter*
Yu Zhao1, Yong Li2, Tapan Kumar Saha1, Olav Krause1
1School of Information Technology and Electrical Engineering, University of Queensland, Australia
2College of Electrical and Information Engineering, Hunan University, Changsha, China
Received April, 2013
ABSTRACT
Protection of radial distribution networks is widely based on coordinated inverse time overcurrent relays (OCRs) en-
suring both effectiveness and selectivity. However, the integration of distributed generation (DG) into an existing dis-
tribution network not only inevitably increases fault current levels to levels that may exceed the OCR ratings, but it may
also disturb the original overcurrent relay coordination adversely effecting protection selectivity. To analyze the poten -
tially adverse impact of DG on distribution system protective devices with respect to circuit breaker ratings and OCR
coordination fault current studies are carried out for common reference test system under the influence of additional DG.
The possible advantages of Superconducting Fault Current Limiter (SFCL) as a means to limit the adverse effect of DG
on distribution system protection and their effectiveness will be demonstrated. Furthermore, minimum SFCL imped-
ances required to avoid miss-operation of the primary and back-up OCRs are determined. The theoretical analysis will
be validated using the IEEE 13-bus distribution test system is used. Both theoretical and simulation results indicate that
the proposed application of SFCL is a viable option to effectively mitigate the DG impact on protective devices, thus
enhancing the reliability of distribution network interfaced with DG.
Keywords: Overcurrent Relay (OCR); Distribution Protection Coordination; Distributed Generation (DG); Fault
Current; Supercondu cting Fault Current Limiter (SFCL)
1. Introduction
In recent years, mainly due to environmental concerns
and in preparation for an expected shortage of traditional
fossil fuel based energy, distributed generation (DG)
based on renewable energy sources is attracting more and
more attention. The advantages of introducing DG into a
distribution system are generally called “system support
benefits” such as voltage support, improved power qual-
ity, loss reduction as well as transmission and distribu-
tion capacity release [1]. However, there are several dis-
advantages introduced by DG. For instance, the increas-
ing fault current levels may exceed the current ratings of
circuit breakers (CBs), which lead to the need for expen-
sive upgrading of CBs [2]. Another impact of DG is the
disturbance on existing protection coordination. Since
distribution systems are for the predominant part of ra-
dial structure, the inverse time Overcurrent Relays
(OCRs) [3] are the most applied protection device in dis-
tribution systems. However, when DGs are installed in a
distribution system, the typical one-direction nature of
power flow can be lost. In such case, there is a risk of
existing relay coordination to be disturbed or even be-
coming ineffective [4].
Several possible solutions have been proposed to
overcome the above problems, such as upgrading circuit
breakers, installing microprocessor based recluses [5],
employing adaptive protection [6], decreasing the gen-
eration capacity of DGs o r even cu t off the DGs from th e
main grid during fault conditions [7]. These methods are
complex and expensive, and in many cases put constraint
in using DG capacity and limiting the benefits from DG
units.
As will be shown in this paper, Superconducting Fau lt
Current Limiter (SFCL) can be used to minimize the ad-
verse impact of DGs on distribution system protection.
SFCL represents a near-zero impedance during normal
operating conditions, thus causing a negligible voltage
drop and power loss. However, during fault condition, it
introduces high serial impedance limiting the short- cir-
*This paper was partly supported by the national Nature Science
Foundation of China (NSFC) under Grant 51007020.

Y. ZHAO ET AL. 259
cuit current flowing though the SFCL. Therefore SFCL
in series with DG units are able to decrease the adverse
impacts of DG to the distribution system protection dur-
ing network faults.
This ability as well as the minimum SFCL impedance
requirements during fault conditions is analyzed in this
paper, which is organized as follows. Section 2 analyses
the effect of DG on the protection coordination of distri-
bution systems. In Section 3, the functioning prin ciple of
SFCL is analyzed and its mathematical model is estab-
lished. In Section 4, a case study is used to investigate
the impact of SFCL on protection of distribution systems
with DG with respect to CB fault current levels and co-
ordination of OCRs. Finally, the conclusions are given in
Section 5.
2. Impact of DG on Overcurrent Relay
(OCR) Coordination
2.1. Principle of OCR
The protective devices in a power system are used to
operate CBs correctly to detect and clear faults with
minimum customer interruption and as quickly as possi-
ble. Inverse time OCRs are most commonly used in a
radial distribution system. The operating time of OCRs is
inversely proportional to the current flow
I
through the
relay exceeding the pick-up current threshold Ipick-up. The
tripping time ttrip(I) is given by the following equation,
where TDS is the time-dial setting, which adjusts the
time-delay curve between minimum and maximum
curves for the particular relay. Here A, B, p and K are
constants that represent different types of OCR:
() ()
1
trip p
pick up
A
tITDS BK
M
MII
 




(1)
Results presented in this paper were obtained using
,, and based on
the “very inverse type OCR” defined by IEEE Std. C37.
112-1996 [8]. For this study TDS values are taken from
an interval between 0.5 and 11s, which are used to tune
response times at same current levels.
3.922A0.0982B2p0K
For systems with multiple installed OCRs, relays in-
stalled in series should be coordinated to ensure relay
response in a specified operation sequence, that is to say,
primary relay near the fault location is supposed to trip
first, and the back-up relay is supposed to trip only in
case of a primary relay fails. This is to ensure maximum
selectivity and to limit the number of customers affected
by the required de-energization of sections of the net-
work. Therefore, the Coordinated Time Interval CTI,
specifying the time between the primary relay’s tripping
time ttrip,primary and the back-up relay’s tripping time
ttrip,back-up is defined as follows:
,,trip backtripprimary
CTI tt

 (2)
Typical CTI values range between 0.2 s and 0.5 s. For
the results presented in this paper CTI values are set to
around 0.25 s. Value for Ipick-up and TDS are chosen ac-
cording to the magnitude of load and fault current flow-
ing through each OCR and the required operating times
to clear the corresponding fault. The selection of these
two values should satisfy the following conditions:
 The primary relay must trip over the level of 1/3 of
minimum fault current of the back-up relay;
 The CTI between primary and back-up relays are set
around 0.25 s and must be over 0.2 s to avoid miss-trip-
ping.
2.2. Analysis of DG Impact
When DGs are integrated into a distribution system, the
Thévenin impedance seen from a possible fault location
will decrease and thus the corresponding fault current
level will increase, which may exceed the interrupting
capacity of the installed CBs. For example, when a fault
F1 occurs in Figure 1, the fault current flowing through
CB2 (2
CB
I
) is calculated as:
I
CB
2
I
s
I
DG
(3)
where,
s
I
is the fault current flowing through CB 2 from
the source feeder before the presence of DG, then the
resulting 2
CB
I
will be greater than
s
I
with help of
D
G
I
supplied by DG. Therefore, in some cases the fault
current 2
CB
I
in the system with DG may exceed the
rated current of the specific CB, which is selected in ac-
cordance with
s
I
.
Additionally, the application of DG in a distribution
network may cause wrong relay coordination. For in-
stance, the OCRs R1, R2 and R3 in Figure 1 have been
coordinated properly for a fault at F1 and F2.The operat-
ing time of R2 is larger than that of R3 by a certain CTI
value while the operatin g sequence for relay R1 and R2 is
similar. However, when DG is connected, the coordina-
tion between these two pairs of relays (R1-R2 and R2-R3)
is likely to be disturbed by the decreasing operation time
of R2 and R3, which is determined by th e increasing fault
current flowing through them. Therefore, the CTI be-
tween R2 and R3 may decrease and CTI between R1 and
R2 may increase.
Figure 1. DG impact analysis.

Y. ZHAO ET AL.
260
3. Application of SFCL
A Fault Current Limiter (FCL) is a device for detecting,
triggering and limiting fault currents in power systems.
An ideal FCL works in low impedance at standby state
thus causes little contribution to power loss of a healthy
system. However, it rapidly converts to a high impedance
when a fault occurs, decreasing the fault current. Among
all types of FCL, the usefulness and usability of Super-
conducting Fault Current Limiters (SFCLs) are widely
investigated due to the advantage of inherent
self-triggering, fast response and self-recovery. The
quenching and recovery characteristics of a resistive type
SFCL can be described as follows:


1
2
0
1exp1
f
f
SFCL n
F
r
tt
tt
RtR ttt
T
at tbtt







 




 


f r
r
(4)
In Equation 4 Rn refers to the maximum resistance of
the specific SFCL, TF refers to the time constant of tran-
sition from the superconducting state to the non-super-
conducting state, while tf and tr are the time intervals for
SFCL starting quenching and starting recovery respec-
tively. Variables a and b are constants related to recovery
characteristic.
The impact of the SFCL on a connected DG unit dur-
ing fault conditions is determined by its current limiting
performance on the DG current. A model of DG-SFCL
unit has been developed in the environment of
PSCAD/EMTDC [9], based on the mathematical model
defined by Equation 4. To illustrate the SFCL perform-
ance Figure 2 depicts the DG fault current contribution
Ia resulting from a simulated fault in the network, with Ia0
depicting the non-limited DG fault current contribution
as a reference. Here, the quenched impedance of SFCL is
selected equals to the line impedance of the small sys-
tem.
Figure 2. DG fault current limitation by a SFCL.
As can be seen in Figure 2, the peak value of fault
current Ia0 before the installation of a SFCL (approxi-
mately 1.8 kA) is more than 10 times of the normal op-
eration current (the magnitude is around 150 A). By in-
stalling a SFCL in series to the DG, this fault current
peak can be limited by around 50% (being reduced to
about 1 kA) of its non-limited value. Note that the cur-
rent limiting performance greatly depends on the
non-superconducting impedance of SFCL, which will be
discussed later on.
4. Case study
4.1. IEEE 13-bus Distribution Network
The test network used in this study is 4.16kV IEEE
13-bus distribution network, which is a radial unbalanced
power system with three-phase, two-phase and sin-
gle-phase lines as well as unbalanced wye load and delta
load [10]. Figure 3 shows the single-line diagram of
13-bus system protected by 10 protection units (OCRs
and CBs).
The system configuration, line impedance and load
data with no DG are given by IEEE PES Distribution
Systems Analysis Subcommittee [10]. The first step of
the study carried out for this paper was to calculate the
load and fault currents, in purpose of determining the
operating times and required CTI between each pair of
OCRs based on the two conditions described in Section 2.
The second step consisted of introducing an additional
three-phase load S= (600+j30) kVA at bus 680 and an
additional 660k VA wind turbine at bus 675 to supp ly the
increasing power. This changes the power and current
flows, leading to fault current increasing and disturbance
of the protective coordination between some pairs of
relays during fault conditions. As the final step, a model
of SFCL is developed and added to the DG connection.
Figure 3. IEEE 13-bus distribution network [10].

Y. ZHAO ET AL. 261
The purpose is to investigate its performance during
faults by minimizing the DG’s adverse impact on protec-
tion coordination.
4.2. DG and SFCL Impact on CB Fault Currents
The required rated current level of CB is determined by
the highest fault current that might have to be cleared by
it. In this section, some simulations are carried out to
analyze the impact of DG and SFCL to the fault current,
when faults occur at the terminal of each CB. Table 1
shows the highest RMS value of the fault currents flow-
ing through CB1, CB2, CB3, CB7 and CB8, for the pro-
posed distribution network without DG, with DG and
with DG-SFCL unit.
As shown in Table 1, SFCL can decrease the fault
current effectively for all CBs while comparing without a
use of SFCL. As the SFCL is installed in series with DG
unit, it is used to limit the DG current contribu tion to the
main grid during a fault. Its cu rrent limiting performance
turns out to be better for tho se CBs that located closer to
it. This can also be observed in Table 1 that the current
limiting performance is more significant for nearby CBs
(CB1, CB2 and CB3). Moreover, the limiting performance
is highly affected by the parameter of SFCL (RSFCL).
Figure 4 shows the relationship between fault current
(RMS) and SFCL resistivity (RSFCL). It can be observed
that the fault current limiting performance becomes bet-
ter with increasing resistivity of the SFCL.
4.3. DG and SFCL Impact on OCR Coordination
As the load and the fault current of this 13-bus network
can be calculated, the OCRs are modified in accordance
with the pre-defined Ipick-up and TDS, aiming at setting
CTIs of each pair of OCRs around 0.25s and in the range
between 0.2s and 0.5s. However, when a DG is con-
nected into the network, the protection coordination will
be disturbed. In purpose of investigating the changes, a
number of simulations in PSCAD/EMTDC environment
have been carried out. The results of three-phase and
single line to ground faults at the terminal of different
buses before and after the introduction of DG are shown
in Tables 2 and 3 respectively. For a fault occurrence at
Table 1. Fault current of each CB (kA).
with DG with DG and SFCL(RSFCL=2pu)
CB No DG
Current increase rate Current increase rate
CB1 2.34 2.71 15.8% 2.48 6.0 %
CB2 2.32 2.65 14.2% 2.48 6.9 %
CB3 2.68 3.17 18.3% 2.90 8.2 %
CB7 3.61 3.80 5.3% 3.71 2.8 %
CB8 4.30 4.58 6.5% 4.44 3.2 %
Figure 4. Fault current limiting effect of SFCL.
Table 2. Setting value of each OCR (three-phase faults).
Without DG With DG
Fault
Location Relay No. Trip time(s)CTI(s) Trip time(s)CTI(s)
R0 0.160 0.155
680 R5 0.385 0.225 0.385 0.230
R4 0.109 0.109
675 R5 0.337 0.228 0.337 0.228
R5 0.285 0.285
692 R9 0.568 0.283 0.568 0.283
R6 0.129 0.129
633 R9 0.342 0.213 0.342 0.213
Table 3. Setting value of each OCR (single-phase faults).
Without DG With DG
Fault
LocationRelay No. Trip time(s)CTI(s) Trip time(s)CTI(s)
R1 0.555 0.400
652
(A-G) R3 0.793 0.238 0.581
0.181
R3 0.606 0.430
684
(A-G) R5 0.831 0.225 0.735 0.305
R7 0.620 0.543
646
(B-G) R8 0.838 0.218 0.743 0.200
R8 0.601 0.527
645
(B-G) R9 0.873 0.272 0.832 0.305
R2 0.575 0.421
611
(C-G) R3 0.818 0.243 0.608
0.187
R3 0.645 0.470
684
(C-G) R5 0.881 0.236 0.800 0.330
R7 0.634 0.568
646
(C-G) R8 0.866 0.232 0.784 0.216
R8 0.612 0.550
645
(C-G) R9 0.845 0.233 0.828 0.278

Y. ZHAO ET AL.
262
bus 680, R0 works as the primary relay while R5 is the
back-up relay. The coordinated conditions of primary
and back-up relays are similar for faults located at other
buses in Tables 2 and 3.
It can be seen from Tables 2 and 3 that most of CTIs,
especially for faults close to the DG are affected by the
presence of DG. The values of CTIs may increase or de-
crease with respects to their locations and distance to the
DG unit as analyzed in Section 2. In this case, the in-
creasing CTIs are still in the range and need no adjust-
ment. However, among those decreased CTIs, the
CTI_1,3 (phase A) and CTI_2,3 (phase C) drop below
0.2s, which is out of the acceptable range. Therefore, the
coordination of these two pairs of OCRs needs to be re-
stored, e.g. by means of a SFCL.
Figure 5 shows the improvement of these two CTIs
when a SFCL is installed, where . Under
the presence of a SFCL, both of these two CTIs have
been improved to over 0.2s, which satisfy the range re-
quirement mentioned in Section 2. In addition, the con-
tribution of SFCL to the improvement of the CTIs is
more significant when the OCR pairs are located closer
to the DG-SFCL unit. For instance, compared with
CTI_1,3 (increasing by 0.034s), CTI_7,8 (phase C) just
increases from 0.200s to 0.209s under the same situation.
2pu
SFCL
R
To further investigate the relationship between differ-
ent values of SFCL parameter SFCL and CTIs, SFCL is
set to 1pu, 1.5pu, 2pu, 2.5pu, and 3pu. The simulation
results are shown in Table 4. It is found that the larger
the SFCL resistivity, the closer is the CTIs to their pre-
vious determined setting values (see Table 3).
R R
Figure 5. Comparison of CTIs.
Table 4. Comparison of CTIs with different value of
RSFCL.
RSFCL 0pu 1pu 1.5pu2pu 2.5pu3pu
CTI_1,3 0.181 0.198 0.2080.215 0.2210.226
CTI_2,3 0.187 0.199 0.2060.214 0.2190.224
With the last part of this study, the minimum value of
RSFCL, which improves all CTIs to the range between 0.2s
and 0.5 s should be determined. As can be observed in
Table 4, when RSFCL set as 1pu, CTI_1, 3 and CTI_2, 3
are slightly under 0.2s, while when , both
of them are over 0.2s. Therefore, some specific tests
were carried out to find the minimum value of RSFCL in
the range between 1pu and 1.5pu. The results turn out
that when RSFCL is set to 1.1pu, CTI_1, 3 (phase A) and
CTI_2, 3 (phase C) are equal to 0.200s and 0.201s re-
spectively, both of them are in the required range. At th e
same time, all of the increasing CTIs are under 0.5s. In
other words, for this case study, a minimum value 1.1pu
is needed for RSFCL to avo id any alteration of the original
OCR settings.
1.5pu
SFCL
R
5. Conclusions
The application of DG in a distribution network increases
the fault current level and disturbs the protection coordi-
nation. To overcome these problems, this paper proposed
a resistive type of SFCL to mitigate the adverse impact
of DG to the protective devices in a radial distribution
network. Simulations on the IEEE 13-bus distribution
test network are carried out by using PSCAD/ EMTDC
software. For this study, the issues of CB rating current
levels and OCR coordination are considered. Particularly,
the fault current flows through CB at the tripping mo-
ment is used to evalu ate the current limiting performance
while the CTIs between the primary and back-up OCRs
operating times are used to investigate the SFCL behav-
ior on OCR restoration. Besides, the minimum parameter
of the proposed SFCL is also d etermined to avoid wrong
coordination of all the OCR pairs. Results show that the
proposed SFCL installation in series with a DG unit is
able to effectively limit the fault current and at the same
time improve the CTIs to its required value.
REFERENCES
[1] P. P. Barker and R. W. de Mello, “Determine the Impact
of Distribution Generation on Power Systems: Part 1- ra-
dial Distribution Systems,” Power Technologies, Inc.,
2000.
[2] N. Hadjsaid, J. F. Canard and F. Dumas, “Dispersed Gen-
eration Impact on Distribution Networks,” IEEE Com-
puter Application in Power, Vol. 12, 1999, pp.
22-28.doi:10.1109/67.755642
[3] C. R. Mason, “The Art and Science of Protection Relay-
ing,” General Electric Company, 1965.
[4] A. A. Girgis and S. M. Brahma, “Effect of Distributed
Generation on Protective Device Coordination in Industry
System”, in Proceedings of Large Engineering Systems
Conference on Power Engineering, 2001, pp. 115-119.
[5] S. M. Brahma and A. A. Girgis, “Microprocessor Based
Reclosing to Coordinate Fuse and Recloser in a System

Y. ZHAO ET AL.
263
with High Penetration of Distributed Generation,” in
Proceedings of IEEE Power Engineering and Society,
Winter Meeting, Vol. 1, 2002, pp. 453-458.
[6] S. M. Brahma and A. A. Girgis, “Development of Adap-
tive Protection Scheme for Distribution Systems with
high Penetration of Distributed Generation,” IEEE trans-
action Power Delivery, Vol. 19, No. 1, 2004, pp. 56-63.
doi:10.1109/TPWRD.2003.820204
[7] S. Chaitusaney and A. Yokoyama, “Impact of Protection
Coordination on Sizes of Several Distributed Generation
Sources,” 7th International Power Engineering Confer-
ence, Vol. 2, 2005, pp. 669-674.
[8] IEEE Standard Inverse –Time Characteristic Equations
for Overcurrent Relays, 1997, IEEE Std. C37, pp. 112-
1996.
[9] Applications of PSCAD/EMTDC, https://hvdc.ca/pscad/
[10] IEEE PES Distribution Systems Analysis Subcommittee
Radial Test Feeders,
http://ewh.ieee.org/soc/pes/dsacom/testfeeders/index.html.