Jain et al.

Protection and Control of Modern Power Systems (2019) 4:26

https://doi.org/10.1186/s41601-019-0140-6
Protection and Control of
Modern Power Systems

ORIGINAL RESEARCH Open Access

Augmentation of situational awareness by

fault passage indicators in distribution
network incorporating network
reconfiguration
Tanmay Jain*, Debomita Ghosh and Dusmanta Kumar Mohanta

Abstract
Power distribution systems are profoundly inclined to disturbances like untimely switching of breakers & relays,
sympathetic tripping, and uncertainties regarding fault location. Thus, system stability and reliability are greatly
affected. In this way, situational awareness and system integrity are the crucial factors in developing power system
security, as it empowers successful decision making & timely reaction by the operators to any disturbance and also
maintaining continuity of power supply. This paper focuses on the enhancement of situational awareness by fault
location through fault passage indicators (FPI) to improve nominal impedance-based methods in distribution
networks. Also, the proposed method is validated by comparing it with Intelligent Electronic Device (IED) based
fault location method. Further, simultaneous reconfiguration of the system is incorporated to maintain the
continuity of supply. The analysis has been tested on IEEE 33 bus distribution system.
Keywords: Situational awareness, Fault passage Indicator, Tie switching, Reconfiguration

1 Introduction taking an intelligent scheme based on graph-theoretic

Nowadays blackouts in power systems is a major con- fuzzy cognitive maps (FCM) i.e., external or internal fail-
cern for utilities, especially in distribution systems. As ures and degradation of an event has been proposed by
the complexity in present networks is increasing, an ab- heuristic rules and unsupervised learning methods.
normality at any side may result in cascading outages Kundu and Pradhan in [5], presented that uncertainties
[1]. Due to this, the concept of situational awareness like faults, loss of load or generator outage leading to
came into the picture. Situational Awareness (SA) is de- unwanted relay operations and makes system unhealthy
fined as “the view of the components in a situation in- and can be prevented by system integrated protection
side a volume of existence, the cognizance of their scheme (SIPS) by locating fault locations, and other dis-
significance, and the projection of their status sooner ra- turbance points in the system. Fault location in power
ther than later”. The sufficient information required for systems, especially in the distribution network, is essen-
SA is divided into three levels i.e., perception, compre- tial as most of the disturbances occur at the load side.
hension, and projection [2]. Many approaches have been carried out considering
Quite a few researches have been carried out on situ- impedance-based fault location methods in distribution
ational awareness and system integrity in power systems. networks. In [6], Gong & Guzman proposed fault
In [3], panteli & kirschen discussed major challenges in location technique that uses current and voltage mea-
implementing SA in power systems which further com- surements obtained from digital relays during fault con-
prises its applications in the power industry. Mohaghe- ditions and adjusts the nonhomogeneous characteristics
ghi in [4], explained about SA in automated systems by of the distribution. The model proposed consist of relay,
recloser and FCIs. In [7], Abo-shady et.al proposed an
* Correspondence: [email protected] analytical approach for determining the fault location in
Department of Electrical and Electronics Engineering, Birla Institute of
Technology, Mesra, Ranchi, India
distribution networks in presence of distributed

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made.
Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 2 of 14

generation (DG). The proposed technique presented a 2. Reliable detection and reporting of ground & phase
method to overcome the error caused due to distribution faults in radial and open-ring networks.
of load along the feeder, non-homogenous sections, high 3. They do not require auxiliary voltage i.e., they are
DG penetration level, and different fault resistance values. self-sustaining because of the presence of lithium
Salim et al. in [8], presented improvements related to the battery.
capacitive effect consideration on impedance-based fault 4. Noise immune plastic fiber-optic connections
location techniques, by considering an exact line segment between sensors & FPI.
model for the distribution line and also new algorithm is
developed which considers line shunt admittance matrix. Locating the exact fault location in a distribution sys-
Mora-Florez et al. in [9], proposed a comparative analysis tem which ultimately sends a signal to relay to operate is
of the performance of 10 impedance-based fault location one of the key steps in promoting SA. For this the fault
methods for power distribution networks. The limitation passage indicator (FPI) is used which is defined as “the
of the above techniques is that in most cases unique iden- devices that give the signal regarding the exact location
tification of location of the fault is not obtained. This of fault i.e., the passage of fault current in the system to
brings the concept of fault locators or indicators (FL) in the crew or operator which ultimately helps in sending
distribution network. the signal to the protection system”. A few researches on
FLs or FIs are located in a substation of the distribu- FPIs have been carried out. In [17], Almeida et al.
tion network, are used to find the location of fault. FLs proposed the algorithm for the optimal placement of FPI
calculate the distance to the fault by calculating faulty in the distribution system by binary approach and
line impedance. In [10], Ho, Lee & Lin proposed the al- optimization through Chu-Beasley genetic algorithm
location of fault indicators (FI) by an immune algorithm (CBGA) which ultimately improves impedance-based ap-
in order to minimize the total cost of customer service proach. In [18], Chollet et al., proposed an overview of
outage and investment cost of fault indicators. Farajol- comparison between FPIs and protection relay over
lahi, Firuzabad and Safdarian in [11, 12] proposed phase and earth fault detection and developed an algo-
mixed-integer programming (MIP) approach for alloca- rithm for the earth fault detection in MV networks. In
tion of fault indicators and also worked on simultaneous [19], Altonen et al., proposed a novel approach of FPI
placement of FI and sectionalizing switches in distribu- indication by using multi-frequency neutral admittance
tion network by considering reliability perspective. In measurement together with a cumulative phasor sum-
[13] Shahsavari, Mazhari, Fereidunian & Lesani consid- ming technique in MV networks. In [20], Mršić et al.,
ered a multi-objective formulation approach by modified outlined the importance of FLs with FPIs and developed
particle swarm optimization (MPSO) for the placement an algorithm for determining the location of FPIs in dis-
of FI in distribution system. Usida, Coury, Flauzino and tribution network with installed FLs. Zeljković et al., in
da Silva in [14] proposed an evolutionary computation [21], assessed the performance of FLs and FPIs separ-
approach for solving the problem of placement of FI’s in ately and by combining them together in distribution
distribution networks. The drawback of FLs is that faulty networks using non-sequential Monte-Carlo simulation
line impedance depends on four parameters i.e., distance and further highlighted the advantages of FPIs over FLs.
to fault (line length), fault type (fault impedance), cross- From the above literature survey, it can be concluded
section of the lines and the number of branches from that many researches are done towards fault identifica-
the substation bus. To overcome this drawback, fault tion and optimal placement of FPI in different power
passage indicators (FPIs) are installed in the network. system networks but in most cases it does not outlines
FPIs determine the location of the fault by detecting the the unique identification of the fault and operator’s deci-
passage of fault current through the lateral coming from sion making & action plan (comprehension) and future
bus on which they are located. By placing the FPIs at state of the system (projection). This paper focuses on
suitable locations in a distribution network, the fault the enhancement of SA by perceiving data of the fault
location time increases, which is in between the last in- and determining its exact (unique) or suspected loca-
dicator that detected fault current and the next indicator tions by FPI placement using binary approach so to
that did not indicate. improve the impedance-based approach. Also, the pro-
The most common makers of FPIs in distribution posed method is compared with impedance-based fault
systems are Siemens (SICAM FPI), Schneider Electric location method using IEDs & FCI (FPI). Further, on the
(Flair 21D) [15, 16], etc. basis of FPI placed, faults are created at different loca-
Features of FPI: tions and then simultaneously tie switching or reconfig-
uration is incorporated in order to maintain continuity
1. Simple & economical for fast fault location in of supply at all buses which accounts on the timely and
modern distribution networks. corrective action plan in regard to the perceived data
Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 3 of 14

i.e., comprehension. This tie switching or reconfigur- 3. Projection – In this level, future system state and
ation is carried out on the basis of power loss time for actions implementation contributes
minimization by backward forward sweep load flow towards developing strategies to avoid undesirable
technique. The efficacy of this analysis is tested on IEEE events.
33 bus distribution system.
The paper has been categorized as follows. Section 2 3 Fault susceptibility by FPI placement
illustrates a brief review of SA in the distribution net- As it is explained already, the first level of SA i.e.,
work. Section 3 outlines fault susceptibility by FPI place- perception which deals with the retrieval of informa-
ment using binary approach. Further, in Sect. 4, tie tion using real-time measurements, electronic dis-
switching incorporation is outlined on the basis of power plays, indicators, etc. Here triggering of FPI in
loss minimization by backward forward sweep load flow distribution network indicates the occurrence of fault
method. In Sect. 5 the experimental results of IEEE 33 in the network and this information is sent to the
bus distribution network are discussed. The conclusion operator. Suppose the network detects the fault
of the paper is comprised in Sect. 6. without installation of FPI, the faulty points which
will be obtained will create uncertainty regarding the
2 Brief review of situational awareness in exact location of fault (shown in Table 1). This
distribution network wrong information percepted by the operator will re-
SA in distribution networks is very important as most of sult in an increase in the fault location time and ul-
the disturbances occur at the customer side. Deployment timately hamper the power supply to the affected
of FPIs in distribution networks enhances SA as retrieval areas for a longer time. So, by installing FPIs in the
of information from FPIs (perception) makes the oper- network, the appropriate information will be per-
ator in control room aware regarding the fault and en- cepted to the operator personnel, as only the con-
hances his decision making by analyzing it & plan the cerned FPI(s) will trigger and the suspected location
necessary actions (comprehension) and which further of fault will be obtained. Hence, the appropriate in-
decreases restoration time & cascading outages can be formation regarding any abnormal event (fault) re-
prevented (projection). The three levels of SA in power sults in enhancement of SA.
systems is depicted in Fig. 1. Distribution systems are vulnerable to faults by dif-
SA is divided into three levels: ferent sources, for example, extreme climatic condi-
tions and equipment failure. So as to improve the
1. Perception – In this level, to carry out SA in power reliability of the system and keep away penalties, it
system data or information is retrieved by real-time is alluring for utilities to rapidly discover faulted
measurements, electronic displays, etc. Along with areas and therefore diminish restoration time to en-
this communication between crew to share the in- hance SA.
formation is equally important. For ex, if any fault The utilization of FPIs to improve impedance-based
occurs in a system, subsequent devices and its dis- approach is an intriguing methodology for electric util-
plays give the information to the crew in the con- ities. It can accelerate the restoration time, staying away
trol room. from penalties and enhancing quality indices. At the
2. Comprehension – After the first level, the perceived point when a fault happens downstream of the indicator,
information is analyzed and according to crew it naturally sends a signal to the supervisory control
objectives its action plan is decided. crew. Appropriately apportioning a specific number of

Fig. 1 Schematic model of SA in the power system

Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 4 of 14

Table 1 Comparison of discrete impedance points of fault in substation i.e., bus 1. For example, the impedance at bus
the network (Case 1) without & with FPIs 2 (25 Ω) is the sum of impedance at bus 1(0 Ω) and
Actual fault Discrete impedance point of Discrete impedance point branch 1(25 Ω). Similarly, for bus 3, the sum of imped-
impedance fault detected by the system of fault detected by the ance up to bus 2(25 Ω) and branch 3(20 Ω) and so on.
(Ω) without FPI system with FPI
Here the impedance step size is 10 Ω and Z○ = 25 Ω.
10 1 1
Now FPIs are placed in a radial network on any of the
20 2 2 laterals whose source bus contains more than one lat-
30 3, 5 5 (if sensed by F1) eral. Also, the fault location is determined through sub-
3 (if sensed by F2) station, so it is mandatory to place FPI at branch coming
40 4, 6, 8, 10 4 (if sensed by F2) out from substation bus. Here in this network, 3 FPIs
6, 10 (if sensed by F1)
8 (if sensed by F3)
are placed at branch 1(F1), 3(F2) and 4(F3).
Suppose in Fig. 2a, the fault occurs at branch 1 at
50 7, 9, 11 7, 11 (if sensed by F1)
9 (if sensed by F3) 10 Ω, so both nominal impedance-based technique and
FPI placed at branch 1 will give one location of fault.
60 12 12
But if the fault occurs at branch 2 at 30 Ω, the nominal
impedance-based technique will give two suspected loca-
FPIs may diminish or even wipe out the vulnerabilities tions of fault i.e., at branch 2 and 3. In this fault case the
related to the exact fault area. But, because of economic FPI(s) will reduce or even eliminate this uncertainty by
restrictions, it is difficult to locate FPIs at all the areas of giving the exact location of fault i.e., at branch 2. This
the system to make it uncertainty free upon the happen- comparison of fault locations in a network consisting of
ing of any fault [17, 18]. FPIs and without FPIs is depicted in Table 1.
For determining the suspected locations of fault by From Table 1, it is concluded that by placing FPIs in
FPI placement, two cases are possible in distribution the network the uncertainty regarding the fault is re-
networks i.e., when only one branch is coming out duced or even eliminated which enhances SA
from substation bus and other is more than one is considerably.
coming out from substation bus. Also, this approach In accordance to FPIs placed, suspected locations for
is considered when only one fault has occurred in the the fault can be drawn by binary approach. In this ap-
network. proach, the total number of combinations drawn for
FPIs flagged status is given by Eq. (2)
3.1 Case 1: when only one branch from substation bus
In a radial network, before placing FPIs the key step is N ¼ 2m ð2Þ
to decide the impedance step size as it was done in nom- Here, m equal to 3 so, the number of combinations N
inal impedance-based methods [17]. is equal to 8. In accordance with it, following Table 2
Impedance step size is decided on the basis of Eq. (1) i.e., using Eq. (3). is created showing the FPIs flag status and
its corresponding discrete impedance points i.e., the sus-
0:1 ≤ Δz ≤ Z ∘ ð1Þ
pected locations of fault in the network shown in
Fig. 2b.
The impedance step size cannot be less than 0.1 be-
cause the results obtained by taking Δz less than 0.1 is K ¼ ½A jBpq ð3Þ
approximately similar to results obtained when Δz = 0.1.
By taking a smaller step size the computational time is Where,
increased means number of discrete impedance points
are increased. Also, the upper limit of Δz is the imped- A has order Rp × (q − q′).
ance of line between substation bus and adjacent bus be- B has order Rp × q′
cause the placing of discrete impedance points starts
from the substation bus. Δz value is appropriate where Unique location of the fault is governed by Eq. (4) for
at least one impedance point is placed in each line i.e., every combination of the FPI status.
all the sections of the network are covered by the


discrete impedance points. This concept has been ex- 1 Unique location of fault
Sj ¼ ð4Þ
plained in [17]. > 1 Multiple suspected locations
Consider Fig. 2a. which depicts a seven-bus radial
feeder which contains seven buses and six branches. The It is important to note that when a fault occurs, all
values in braces are the branch numbers. The cumula- FPIs placed between substation & fault location is
tive impedance is shown at every bus calculated from flagged 1.
Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 5 of 14

Fig. 2 a Seven bus radial feeder. b FPIs placed in seven bus feeder with an impedance step size of 10 Ω

It is important to note that when the impedance step

size decreases the number of discrete impedance points
Table 2 FPIs flag status and corresponding discrete impedance increases. Also, from [17] it is proved that taking imped-
points (suspected locations of fault) for Case 1 ance step size less 0.1 Ω will increase number of sus-
FPIs status Discrete impedance points pected locations and increases the computational time.
F3 F2 F1 10 Ω 20 Ω 30 Ω 40 Ω 50 Ω 60 Ω So, the results obtained by taking step size less than or
0 0 0 NO FAULT
equal to 0.1 Ω appeared to be same.
From Table 2 it is important to note that, when F1 is
0 0 1 1 1 1 2 2 1
(1) (2) (5) (6,10) (7,11) (12) flagged zero regardless of flag status of F2 and F3, it will
be a no-fault condition i.e., substation FPI flag status is
0 1 0 NO FAULT
vital in knowing that fault location in the network. Also,
0 1 1 0 0 1 1 0 0
(3) (4)
when F1 is flagged 1 and other FPIs are flagged 0, then
at 40 Ω and 50 Ω two suspected fault locations are seen
1 0 0 NO FAULT
at branch 5 and 6, whereas, when FPI is not there in net-
1 0 1 0 0 0 1 1 0 work the nominal impedance-based method will give
(8) (9)
four suspected locations of fault for 40 Ω. So, when FPIs
1 1 0 NO FAULT
are installed in a network the uncertainty regarding the
1 1 1 0 0 1 2 1 0 fault location is reduced or even eliminated. Therefore,
(3) (4,8) (9)
by placing FPIs, the fault identification time is reduced
Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 6 of 14

for the crew or operator as compared to nominal imped- Here no-fault condition will occur when both F1 and F2
ance techniques [17]. will be flagged zero regardless of the status of F3 and F4.

3.2 Case 2: when multiple branches from substation bus 3.3 Case 3 (special case): when substation FPI has 80%
Now consider Fig. 3a which shows an eight bus feeder reach
which has eight buses and seven branches. Here also im- Both above cases are designed when the substation bus
pedance step size is 10 Ω. As discussed in case 1, it is FPI has 100% reach i.e., F1 will operate on 100% of the
mandatory to place FPI(s) at the branches coming out impedances present in its downstream i.e., 0 Ω to 65 Ω.
from substation bus. So as seen in Fig. 3b, 4 FPIs are So, if F1 is designed up to 80% of 65 Ω i.e., 0 Ω to
placed at branches 1(F1), 2(F2), 3(F3) and 7(F4). 52 Ω of the impedance present in its downstream then
In Fig. 3b, four FPIs are placed. Here the number of the FPIs location and suspected location of fault based
combinations of FPI flag status using Eq. (2) is equal to on FPI flag status will be different from above cases.
16. Similarly, its FPIs flag status vs discrete impedance In Fig. 2b, the number of FPI will increase when sub-
points table is created using Eq. (4). station FPI has 80% reach i.e., instead of 3 FPIs, 5 FPIs

Fig. 3 a Eight bus radial feeder. b FPIs placed in eight bus feeder with an impedance step size of 10 Ω
Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 7 of 14

will be placed. The two more FPIs will be placed at 52 Ω  If both master and lateral FPI malfunctions (extreme
at branch 5(F4) and 6(F5) to uniquely locate faults after case) – If both master and lateral FPI malfunctions,
52 Ω. then the PMU data will give the information
When substation FPI has 80% reach and other FPIs regarding the fault in the network.
have not 100% reach, then the impedance step size of
the network will change according to the impedance of Thus, in all the cases the determination of the exact
the lines. In both cases, the number of FPIs will increase fault location to take corrective measures i.e., perception
i.e., a new FPI will be placed where the reach of previous in the power system network is enhanced ultimately en-
FPI in the lateral ends. For ex, in case 1 the reach of F2 hancing the SA of the network.
is 80%, so, it will cover up to 80% of 25-45 Ω i.e., 41 Ω, Based on this binary approach, the analysis is tested
hence a new FPI will be placed after 41 Ω to cover rest on IEEE 33 bus distribution system which is discussed in
of the section. Sect. 5 and further tie switching or reconfiguration is in-
A generalized 4 bus radial feeder showing the FPIs and corporated by backward forward load flow analysis to
discrete impedance points are shown in Fig. 4. On the maintain continuity of supply which is the second stage
basis of network shown in Fig. 4, the flowchart for the of SA i.e., comprehension is explained in Sect. 4.
FPI placement and identification of the suspected loca-
tion of fault by binary approach is depicted in Fig. 5 to 4 Tie switching by backward forward load flow
enhance perception of the network which is the first technique
level of SA. After retrieval of information regarding an event, the
Time delay, false negatives, and false positives problem data related to it is analyzed and corrective & timely ac-
occur when FPI malfunctions i.e., either FPI fails to give tions are implemented. This is called the second level of
the status when the fault has occurred or it triggers even SA i.e., comprehension. Here, when the perceived data
if there is no fault. To deal with this condition when FPI of the fault is analyzed means by checking the FPI flag
malfunctions, the status of master FPI (substation FPI) status the fault section is located (explained in Sect. 3),
and any voltage & current magnitude measuring device tie switching in the network is implemented means the
like phasor measurement units (PMUs) will give the in- tie associated with the faulty section is closed on the
dication regarding the fault. The measuring device with basis of minimum power loss to maintain the continuity
master FPI in the network is just to reconfirm the detec- of power supply. This timely implementation of neces-
tion of the fault if the adjoining FPI malfunctions i.e., sary actions improves level 2 of SA (comprehension) and
they both operate parallel to each other. further enhances the SA.
The expansion in the power demand and high load
 If there is no fault in the network and concerned FPI density in the urban areas make the chore of distribution
flagged 1 – In this case the status of master FPI will system complex. Because of low voltage level of distribu-
be flagged 0 which means there is no fault in the tion systems, active power losses are significantly more.
network. To limit these losses, the status of normally closed
 If there is fault in the network and concerned FPI flag sectionalizing switches and normally opened tie switches
status is 0 – In this case the status of master FPI will is modified by keeping up the radial topology of the
be flagged 1 and the PMU installed at the substation system. This is called the reconfiguration of the system
bus & at buses where more than one branches are [22]. The merits of system reconfiguration are load
coming out will let the crew know the distance to balancing, minimizing power losses, improvement of
fault using Thevenin equivalent model by voltage profile, enhancing power system security &
monitoring the voltage & current magnitude. reliability and upgrade of power quality [23]. Many

Fig. 4 Generalized 4 bus radial feeder

Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 8 of 14

Fig. 5 Flow chart to enhance network perception

approaches have been proposed on network reconfig- approach. In [23–27], reconfiguration of distribution
uration of distribution networks over the years. Baran networks is carried out by considering different ob-
and Wu in [24] highlighted the concept of distribu- jective functions and by using methods such as gen-
tion network reconfiguration by considering loss etic algorithm (GA), fuzzy logic approach, particle
minimization and load balancing by branch exchange swarm optimization (PSO), etc.
Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 9 of 14

The backward and forward load flow technique is uti- voltage and if the error is inside tolerance limits, at that
lized to comprehend the power flow examination of dis- point the procedure is halted, also the line losses are fig-
tribution networks with recursive equations. The ured out otherwise, the process is rehashed until criteria
backward and forward technique depends on the conditions are met [29, 30].
Kirchhoff’s voltage and current law and in every em- The bus voltages are calculated in the forward path
phasis, two calculation stages happen, the forward way from the upstream bus to the downstream bus. They are
and the backward way [28]. represented by branch currents and line parameters
shown in Eq. (10).
S x ¼ P x þ jQx ð5Þ
 
S x ¼ V x I x ð6Þ V y ¼ V x −I xy Rxy þ jX xy ð10Þ

The apparent power at each bus in terms of active & The active and reactive power losses are given by Eqs.
reactive power, voltage and current are given by Eqs. (5) (11) and (12).
and (6).
X
PL ¼ I 2xy  Rxy ð11Þ
4.1 Backward sweep
The motive is to calculate current in each branch by
considering the steady value of voltages found in the
previous iteration while a voltage value is assumed ini- X
tially for analysis. The backward way begins from the last QL ¼ I 2xy  X xy ð12Þ
node to the source node [29].
  The backward forward load flow algorithm is depicted
Px þ jQx  from the flowchart in Fig. 7.
Ix ¼ ð7Þ
Vx Tie switching or reconfiguration can be incorporated
  using backward forward sweep load flow technique.
Py þ jQy 
Iy ¼ ð8Þ Whenever a fault occurs in a distribution network, loss
Vy
of load takes place i.e., continuity of supply to the cus-
X tomers have been hampered. On this basis tie switching
I xy ¼ I y þ I yr ð9Þ
rϵαy
is carried out simultaneously i.e., normally opened tie
switches are closed on the basis of minimum active
For Fig. 6. the bus currents at each bus are calculated power loss in order to feed the loads which are suffered
using Eqs. (7) and (8). The bus currents are summed up by the faults.
in backward path to give branch currents towards cen- The algorithm for tie switching on the basis of mini-
tral bus which is at upstream by Eq. (9). mum active power loss by backward forward sweep load
flow is as follows:
4.2 Forward sweep
This begins from the source node to the far end node i. Run load flow when network is not subjected to any
means to compute the voltages at each bus while keep- faults.
ing the current acquired from previous iteration con- ii. Calculate the initial active power losses in a healthy
stant implying that the current acquired in the backward condition.
path will be constant amid the forward sweep. The de- iii. Now, if the fault occurs, identify the faulted line by
termined voltages are compared with the predetermined FPI flag status.
iv. Correspondingly, check the impedance point of
fault from FPI flag status vs discrete impedance
points table.
v. Identify the possible associated ties for the faulted
line by seeing the network topology.
vi. Close the associated ties one by one.
vii. Run load flow again on every tie closing and
calculated the active power losses in each case
using Eq. (11).
viii.From the data obtained of active power losses by
running the load flow on every tie closing, choose
Fig. 6 Two Bus radial system the case where minimum power loss has occurred.
Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 10 of 14

5 Case study & results

5.1 FPI placement and determination of suspected
locations
Single line diagram of IEEE 33 bus distribution network
is represented in Fig. 8. The network consists of 37
branches and 33 buses which further includes (normally
closed) 32 sectionalizing switches (normally closed) & 5
tie switches (normally opened). The tie switches are A33,
A34, A35, A36, A37. The distribution system load is 3715
kW & 2300 kVAr and voltage is 12.6 kV. Branch 1 to 9
has the current capacity of 400A & other lines (including
tie lines) has 200A. The Vmin and Vmax are 0.95 p.u &
1.05 p.u. 1000 kW DG sources are present on buses 6,
11, and 29 taken from [31]. The impedance of each line
is shown in Table 3.
As it can be seen from the network, line 3 to 5, line 6
to 17, line 22 to 24, line 25 to 32 and line 18 to 31 are
connected in series. Figure 9 depicts the equivalent net-
work of IEEE 33 bus distribution system by combining
the buses which are connected in series and also sum-
ming their impedances. Also, the placement of FPIs and
the discrete impedance points in the equivalent network
of IEEE 33 bus distribution network are shown in Fig. 9.
As the FPIs are usually placed at the branches whose ini-
tial bus contains laterals. So according to it, four FPIs
are placed at branch 1, 4, 5 and 7 on the equivalent net-
work of IEEE 33 bus distribution system. Here Z∘ =
0.1034 Ω and impedance step size Δz = 0.1 Ω. Impedance
step size should be in constraints using Eq. (1).
From Fig. 9, it has been observed that branch 1 has
the FPI but its initial bus does not have any laterals. This
is because if any fault happens in branch 1, then it can-
not be uniquely located with the other FPIs placed on
other branches. So, it is mandatory to place FPI at
branch 1. Also, as branch 1 is connected to the sub-
station, so, FPI placed at branch 1 plays a vital role in
determining fault location in the network.
Using Eq. (2), the number of combinations of FPI sta-
tus is equal to 16. Similar to the table created using Eq.
(3) in Sect. 2, the same will be created for IEEE 33 bus
distribution system. Also, the presence of priority areas
in the network are the buses where the FPI is connected
i.e., in IEEE 33 bus distribution network the priority
buses are bus 1, 2, 3 & 6.

Fig. 7 Flow chart of Backward Forward Load 5.1.1 Comparison of the proposed method with IED based
Flow Technique method
The proposed method is compared with the impedance-
Finally, the level 3 of SA i.e., projection is implemented based fault location method using IEDs & FCI (FPI) [6]
which accounts on the future behavior of the system in has been summarized in Table 4 and explained as
relation to perceived information Here, after timely ac- follows:
tion implementation, the future system state is analyzed
and after the fault is cleared, the system is reverted back 1. The IED based method account on possible fault
to its original state. locations in the network i.e., the accuracy of fault
Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 11 of 14

Table 3 Line impedances of IEEE 33 bus distribution network

Line number Impedance (Ω)
1 0.1034
2 0.5532
3 0.4107
4 0.4276
5 1.0819
6 0.6464
7 0.7492
8 1.2682
9 1.2796
10 0.2070
11 0.3943
12 1.8678
13 0.8952
14 0.7911
15 0.9241
16 2.1502
17 0.9302
18 0.2266
19 2.0247
20 0.6297
21 1.1751
22 0.5464
23 1.1442
24 1.1376
25 0.2278
26 0.3189
27 1.4118
28 1.0665
29 0.5695
30 1.3699
31 0.4768
32 0.6303

location is around 85–90% whereas the proposed

method accuracy is near to 95% because in most of
the cases the fault is uniquely located.
2. In IED based method there is no such information
regarding “no-fault” condition whereas in the
proposed method the master FPI status (flagged 0)
gives the information to the control center
personnel regarding the no-fault condition.
3. Due to presence of recloser action in IED based
method for fault location, the method is only meant
for overhead distribution network whereas, in the
case of the proposed method, it is possible for both
Fig. 8 IEEE 33 bus distribution network
overhead & underground network.
Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 12 of 14

Table 4 Comparison of the proposed method with IED based

method
Analyzing aspect IED based method Proposed method
Accuracy 85–90% 95%
Detection of No Yes
no-fault Scenario
Type of network Only possible for Possible for both overhead
overhead distribution and underground
network distribution network

Based on the fault location by binary approach, tie

switching is incorporated in IEEE 33 bus distribution
network on the basis of power loss minimization.

5.2 Tie switching or reconfiguration

Distribution system reliability and power security are de-
graded heavily when any abnormality or fault occurs.
Here in IEEE 33 bus distribution network faults are cre-
ated at different locations and then in order to maintain
of supply to each and every load, tie switching or recon-
figuration is incorporated on the basis of power loss
minimization.
Seven faults are created at different locations and simul-
taneous tie switching or reconfiguration is incorporated to
maintain continuity of supply. These observations and
analysis are considered by running backward forward
sweep load flow analysis and are shown in Table 5.
From Table 5, it has been observed that when

i. The fault occurs at A23, only one tie switch closing

is possible i.e., A37.
ii. The fault occurs at A27, A14, and A19 two tie switch
closing is seen but by closing tie A37, A34 and A35
receptively, the minimum loss is observed.
iii. The fault occurs at A10, A7 and A4 three tie switch
closing are seen but by closing tie A35, A35, and A37
respectively, minimum loss is observed.

6 Discussion
Fault location by FPIs using binary approach to improve
nominal impedance-based methods is more advanta-
geous over IED based fault location method as accuracy
is quite higher and it can be used for any type of radial
network (discussed in Sect. 5.1.1). The limitation of the
study is that this fault location method by FPIs does not
account for multiple faults in the network so, this aspect
can be considered as a future research area for this
Fig. 9 Four FPI placed with discrete impedance points for study. Also, only power loss parameter has been taken
IEEE 33 bus distribution network under scrutiny in analyzing the optimal tie at the fault
condition when tie switching is incorporated i.e., param-
eters like reliability indices, bus voltages, etc. can also be
considered.
Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 13 of 14

Table 5 Fault locations and tie switching in IEEE 33 bus distribution network
Faulted FPI flag status Fault Power loss Tie Power Tie % Loss
line impedance after the closed losses after chosen reduction
F4 F3 F2 F1
range (Ω) fault (no the fault on the on
tie (with tie basis of comparing
switching) switching) minimum power loss
in kW in kW power after fault
loss (kW) without tie
switching
& with tie
switching
A23 0 1 0 1 1.2030–2.34720 910.03600 A37 172.1304 A37 81.0353
(172.1304)
A27 1 0 0 1 3.1236–4.53540 834.10000 A37 132.8096 A37 84.0774
(132.8096)
A36 141.0619
A14 0 0 0 1 9.8846–10.6757 384.50340 A36 114.0669 A34 75.9695
(92.39790)
A34 92.39790
A10 0 0 0 1 6.5203–6.72730 617.26090 A36 139.3396 A35 88.0307
(73.88150)
A34 76.35050
A35 73.88150
A7 0 0 0 1 3.2233–3.97250 933.05120 A36 166.6502 A35 92.1796
(73.00310)
A35 73.00310
A33 75.46760
A4 0 0 0 1 1.0673–1.49490 1243.9911 A37 143.5136 A37 88.4634
(143.5136)
A35 157.8183
A33 160.3537
A19 0 0 1 1 0.3300–2.35470 353.12300 A33 95.59170 A35 73.9674
(91.92680)
A35 91.92680

7 Conclusion at every load. This accounts in the timely and corrective

Nowadays situational awareness (SA) is very essential for action plan of the operator and enhances the level 2 of
any power system network. As maximum uncertainties SA i.e., comprehension and further after fault is cleared,
and abnormalities occur in distribution networks, so SA network is reverted back to its original state which in-
plays an important role in these networks to quickly re- corporate to level 3 of SA i.e., projection.
spond to the abnormities by the maintenance crew to
avoid the loss of power to customers and enhancing sys- 8 Nomenclature
tem reliability. The information of fault occurrence is
perceived by FPIs and determination of suspected loca- FPI Fault Passage Indicator
tions of fault by FPI placement using binary approach to Δz Impedance step size
improve nominal impedance-based method in IEEE 33 Z∘ Impedance between substation bus and its adjacent
bus distribution network to enhance the level 1 of SA bus in a network
i.e., perception has been proposed in this paper. This N Number of combinations of FPIs flag status
method reduced or even eliminated the uncertainty of m Number of FPIs in a system
fault locations in the system as it was seen in the nom- Sj Suspected location of fault for jth combination of FPI
inal impedance-based method of the same impedance status
point which further enhances the situational awareness p FPIs flag status combinations
of the distribution network. The proposed method has q Total columns of flag status of each FPI and number
been validated by comparing it with the IED based fault of discrete impedance points
location method. Also, in regard to the faults created at q’ Number of discrete impedance points
different locations in the network tie switching is incor- (q-q’) Flag status of each FPI
porated on the basis of power loss minimization at that Px, Qx Active and Reactive power at bus x
particular fault to maintain continuity of power supply Vx, Vy Bus Voltage at bus x and y
Jain et al. Protection and Control of Modern Power Systems (2019) 4:26 Page 14 of 14

Ix, Iy Bus currents at bus x and y 13. Shahsavari, A., Mazhari, S. M., Fereidunian, A., & Lesani, H. (2014). Fault
Ixy Branch current between bus x and y Indicator deployment in distribution systems considering available control
and protection devices: A multi-objective formulation approach. IEEE
Rxy, Xxy Resistance and Reactance of xy line Transactions on Power Systems, 29(5), 2359–2369.
αy Set of all buses adjacent to y bus towards its 14. Usida, W. F., Coury, D. V., Flauzino, R. A., & da Silva, I. N. (2012). Efficient
downstream placement of fault indicators in an actual distribution system using
evolutionary computing. IEEE Transactions on Power Systems, 27(4), 1841–
PL, QL Active & reactive power losses 1849.
15. w3.siemens.com. (2016). SICAM FPI – Digital Grid – Siemens. [online]
Acknowledgements Available at: https://w3.siemens.com/smartgrid/global/en/products-systems-
Not Applicable solutions/substation-automation/short-circuit-indicator/pages/sicam-fpi.aspx
[Accessed 8 Jul. 2019].
Authors’ contributions 16. download.schneider-electric.com. (2018). [online] Available at: https://
TJ has carried out the research on fault susceptibility by FPI placement in download.schneider electric.com/files?p_enDocType=User+guide&p_File_
distribution networks by considering different network cases and further tie Name=NT00240-FR-EN-08+-+Flair+21D+User+manual.pdf&p_Doc_Ref=
switching has been incorporated so has to maintain continuity of supply. DG NT00240 [Accessed 8 Jul. 2019].
and DKM has been the technical adviser of the complete research work. 17. De Almeida, M. C., Costa, F. F., Xavier-de-Souza, S., & Santana, F. (2011).
They have supported to overcome various challenges that appeared in the Optimal placement of faulted circuit indicators in power distribution
work and helped in drafting the manuscript. All authors read and approved systems. Electric Power Systems Research, 81(2), 699–706.
the final manuscript. 18. Chollot, Y., Mecreant, J., Leblond, D., & Cumunel, P. (2017). New solution of
fault directional detection for MV fault passage indicators. CIRED – Open
Funding Access Proceedings Journal, 2017(1), 1326–1329.
Not Applicable 19. Altonen, J., Wahlroos, A., & Vähäkuopus, S. (2017). Application of multi-
frequency admittance-based fault passage indication in practical
Availability of data and materials compensated medium-voltage network. CIRED – Open Access Proceedings
Data sharing not applicable to this article as no datasets were generated or Journal, 2017(1), 947–951.
analyzed during the current study. 20. Mršić, P. et al. (2018) Minimization of power interruption time in MV
distribution networks with fault locators based on optimal placement of
fault passage indicators, 2018 International symposium on industrial
Competing interests
electronics (INDEL), Banja Luka, Bosnia and Herzegovina, pp. 1–7.
The authors declare that they have no competing interests.
21. Zeljković, Č. et al. (2018) Performance assessment of fault locators and fault
passage indicators in distribution networks by the non-sequential Monte
Received: 18 April 2019 Accepted: 5 November 2019
Carlo simulation, 2018 International symposium on industrial electronics
(INDEL), Banja Luka, Bosnia and Herzegovina, pp. 1–7.
22. Atteya, I. I., Ashour, H., Fahmi, N., & Strickland, D. (2017). Radial distribution
References network reconfiguration for power losses reduction using a modified particle
1. Begovic, M., Madani, V., Novosel, D. (2007, December 26) System Integrity swarm optimization. CIRED-Open Access Proceedings Journal, 1, 2505–2508.
Protection Schemes (SIPS), 2007 iREP Symposium – Bulk Power System Dynamics 23. Savier, J. S., & Das, D. (2007). Impact of network reconfiguration on loss
and Control – VII. Revitalizing Operational Reliability, Charleston, SC, pp. 1–6. allocation of radial distribution systems. IEEE Transactions on Power Delivery,
2. Panteli, M., Crossley, P. A., Kirschen, D. S., & Sobajic, D. J. (2013). Assessing 22(4), 2473–2480.
the impact of insufficient situation awareness on power system operation. 24. Baran, M. E., & Wu, F. F. (1989). Network reconfiguration in distribution
IEEE Transactions on Power Systems, 28(3), 2967–2977. systems for loss reduction and load balancing. IEEE Transactions on Power
3. Panteli, M., & Kirschen, D. S. (2015). Situation awareness in power systems: Delivery, 4(2), 1401–1407.
Theory, challenges and applications. Electric Power Systems Research, 122, 25. Ching-Tzong Su, & Chu-Sheng Lee. (2003). Network reconfiguration of
140–151. distribution systems using improved mixed-integer hybrid differential
4. Mohagheghi, S. (2014). Integrity assessment scheme for situational evolution. IEEE Transactions on Power Delivery, 18(3), 1022–1027.
awareness in utility automation systems. IEEE Transactions on Smart Grid, 26. Das, D. (Jan. 2006). A fuzzy multi-objective approach for network reconfiguration
5(2), 592–601. of distribution systems. IEEE Transactions on Power Delivery, 21(1), 202–209.
5. Kundu, P., & Pradhan, A. K. (2016). Enhanced protection security using the 27. Rao, R. S., Ravindra, K., Satish, K., & Narasimham, S. V. L. (2013). Power loss
system integrity protection scheme (SIPS). IEEE Transactions on Power minimization in distribution system using network reconfiguration in the
Delivery, 31(1), 228–235. presence of distributed generation. IEEE Transactions on Power Systems,
6. Gong, Y., Guzmán, A. (2011) Distribution feeder fault location using IED and 28(1), 317–325.
FCI information, 64th Annual Conference for Protective Relay Engineers, 28. Chang, G. W., Chu, S. Y., & Wang, H. L. (2007). An improved backward/
College Station, TX, pp. 168–177. forward sweep load flow algorithm for radial distribution systems. IEEE
7. Abo-Shady, F.M., Alaam, M.A., Azmy, A.M. (2013) Impedance-based fault Transactions on Power Systems, 22(2), 882–884.
location technique for distribution systems in presence of distributed 29. Sunisith, S. and Meena, K., 2014. Backward/Forward sweep based
generation, IEEE International Conference on Smart Energy Grid Engineering distribution load flow method. International electrical engineering journal
(SEGE), Oshawa, ON, pp. 1–6. (IEEJ), 5(9), pp. 1539–1544.
8. Salim, R. H., Salim, K. C. O., & Bretas, A. S. (2011). Further improvements on 30. Hameed, F., Al Hosani, M., & Zeineldin, H. H. (2019). A modified backward/
impedance-based fault location for power distribution systems. IET forward sweep load flow method for islanded radial microgrids. IEEE
Generation, Transmission & Distribution, 5(4), 467–478. Transactions on Smart Grid, 10(1), 910–918.
9. Mora-Florez, J., Melendez, J., & Carrillo-Caicedo, G. (2008). Comparison of 31. Ameli, A., Bahrami, S., Khazaeli, F., & Haghifam, M. (2014). A multiobjective
impedance based fault location methods for power distribution systems. particle swarm optimization for sizing and placement of DGs from DG
Electric Power Systems Research, 78(4), 657–666. Owner’s and distribution Company’s viewpoints. IEEE Transactions on Power
10. Ho, C., Lee, T., & Lin, C. (2011). Optimal placement of fault indicators using Delivery, 29(4), 1831–1840.
the immune algorithm. IEEE Transactions on Power Systems, 26(1), 38–45.
11. Farajollahi, M., Fotuhi-Firuzabad, M., & Safdarian, A. (2018). Deployment of
fault Indicator in distribution networks: A MIP-based approach. IEEE
Transactions on Smart Grid, 9(3), 2259–2267.
12. Farajollahi, M., Fotuhi-Firuzabad, M., & Safdarian, A. (2019). Simultaneous
placement of fault Indicator and sectionalizing switch in distribution
networks. IEEE Transactions on Smart Grid, 10(2), 2278–2287.