1

Comparative Analysis of Transmission Line Falling

Conductor Protection Methods
Charles Adewole and Carlos Aguilar
GE Vernova, USA
Alfredo Marquez, Arturo Torres, Ed Wong, and Nagy Abed
Southern California Edison, USA

Abstract— An energized overhead conductor may break and a fire. An undetected energized downed conductor poses an
fall to the ground or on surrounding objects due to several reasons electrical hazard to people in the vicinity. Traditional protective
such as conductor aging, hardware failures, pole knock-over, relaying methods cannot detect Hi-Z faults and may reclose on
severe weather conditions, and natural disasters. When an
a line that may have a broken conductor. Thus, it is expected
energized conductor falls to the ground or other grounded objects,
this may cause a ground fault or arcing. This poses a risk to utility that protective relays detect the broken conductor condition in
personnel, public safety, equipment, and may ignite wildfires. mid-air and trip the corresponding breaker(s) before the
Traditional broken conductor detection or Falling Conductor conductor hits the ground.
Protection (FCP) is typically based on the current unbalance Electric power utilities around the world are beginning to
calculated as the ratio of negative sequence current (I2) to positive implement various strategies for grid resiliency and wildfires
sequence (I1), with an I2/I1 detection slope set to 20%-30%. mitigation. Such strategies may include changes in operational
However, detecting broken conductor events during lightly-loaded
operating conditions may be challenging. Also, it may be
practices, system hardening, asset inspections, situational
challenging to distinguish between broken conductor events and awareness, weather monitoring, etc [1]. Transmission line
asymmetrical faults within and outside of the zone of protection. falling conductor protection is one of the wildfires mitigation
The mechanics of a conductor separation depends on the type of strategies utilized by utilities to detect a broken or falling
hardware failures and may be sudden or gradually evolving over conductor in midair, trip the appropriate circuit breakers, and
a period of time. This affects the electrical parameters that are block line reclosing.
continuously measured by Intelligent Electronic Devices (IEDs)
and may result in the misoperation of traditional FCP methods.
While some Hi-Z fault detection functions/elements have
This paper evaluates the effectiveness of existing FCP methods for been developed and are commercially available in some digital
broken-conductor detection in transmission lines. The FCP protective relays, their effectiveness in transmission, sub-
methods evaluated include current-based (using the conventional transmission, and distribution systems have not been widely
I2/I1 and the modified I2/I1 methods) and impedance-based documented [2]. Existing broken conductor detection methods
methods. The performance of these FCP methods is validated are generally categorized into current-based methods and
using field events playback and hardware-in-the-loop simulations
voltage-based methods [1]-[4].
using the Real-Time Digital Simulator (RTDS). Extensive testing
using in-zone and out-of-zone broken conductor events, in-zone This paper evaluates the effectiveness of two current-based
and out-of-zone fault events, and transient events were tested for and one impedance-based transmission line Falling Conductor
an actual transmission system. Lessons learned from the testing Protection (FCP) methods for both instantaneous and
and evaluation of the FCP methods are also discussed. mechanically-evolving breaks. Current-based methods utilize
the currents received from both line ends. The impedance-based
Index Terms: Broken conductor, charging currents, falling method utilizes the voltages and currents received from the
conductor protection, Impedance Change Ratio (ICR), power lines ends. Transmission line differential relays exchange the
factor, RTDS, synchrophasors, and transmission line.
voltages and currents of the local and remote line terminals with
each other, either via an existing line differential
I. INTRODUCTION
communication channel or a separate communication medium.
Utilities are facing an increase in wildfire risks resulting from Each relay will process this data to detect falling conductors by
faults and failures in overhead lines and equipment, aging identifying a specific pattern in the line impedances.
electric power apparatus equipment, severe weather conditions, Alternatively, a central real-time controller may be used to
etc [1]. collect and process the data.
An energized overhead conductor may break and fall to the The rest of this paper is organized as follows: Section II
ground or on surrounding objects due to several reasons such as describes the current-based broken conductor detection
such as severe weather conditions, conductor aging, hardware methods, while section III presents the impedance-based high-
failures, pole knock-over, etc. When the falling conductor speed falling conductor method. Section IV describes the
touches the ground or other grounded objects, the resulting hardware-in-the-loop test platform and the results obtained. The
high-impedance (Hi-Z) fault may be difficult to detect by contribution of this paper is summarized in Section V.
conventional protective relaying schemes. Also, as the
energized conductor hits the ground, it can produce electrical
arcing that may ignite dry flammable vegetation and may start
 2

II. CURRENT-BASED FALLING CONDUCTOR DETECTION Three conditions must be satisfied in order to detect a broken
METHODS conductor condition. These are: (i) the phase current magnitude
The detection of broken falling conductors in transmission in all the three phases should be below a maximum (fault)
systems may pose a challenge since it depends on the system threshold, (ii) the |I2/I1| criterion must assert, and (iii) the third
topology, dynamic line loading, location of the conductor break, criterion requires the assertion of either a Power Factor (‘PF
mechanics of the conductor separation, availability of Detector’) AND a phase current magnitude detector (‘Current
information along the power line, etc. Magnitude Detector’) when below minimum threshold values.
A. Traditional I2/I1 Current-Based Method Criterion 1
|I1| < I1max
The implementation of the traditional I2/I1 FCP method Criterion3
tested requires three criteria to be satisfied as shown in Fig. 1. |IA| < Ich Criterion 2 TIMER
FCP PKP PU FCP TRIP
The first criterion requires that the phase current magnitudes |I2|/|I1| > RATIO
DO
PF DetectorPhA
in all the three phases be less than a threshold for a faulted event,
to ensure that there is no fault event at the time of detection. IA OPEN
ROCOCPhA
The second criterion is the I2/I1 requirement for detecting
TIMER
current imbalance on the transmission line. Broken or falling |IB| < Ich IB OPEN
PU
DO
conductors can be detected using current imbalance (|I2/I1|),
IC OPEN
which is an estimated representation of Z0/(Z1+Z0), where Z0 |IC| < Ich

and Z1 denote the total zero-sequence impedance and the total

positive-sequence impedance of the circuit, respectively Fig. 2. Modified |I2/I1| current-based FCP method using Power Factor
Detectors.
(including local and remote sources and the line). If the
calculated |I2/I1| ratio is greater than a setpoint, there may be a
For a broken conductor condition, the phase current angle
broken conductor along the line. The |I2/I1| ratio of the
remaining two phases when one phase is broken is theoretically would lead the phase voltage angle by ∠90̊ due to charging
50%. A typically |I2/I1| ratio threshold setting is between 20% current on the line. A broken conductor may be detected using
- 30% [3]. a power factor setpoint of ∠90° ± ∠x°.
The positive sequence line charging current Ich for a
Criterion 1
|I1| < I1max
transmission line under balanced condition can be calculated
Criterion 2 TIMER using (1):
FCP PKP PU FCP TRIP
|I2|/ |I1| > RATIO
DO 𝑉𝐿−𝐿
𝐼𝑐ℎ = 𝑗𝑤𝐶1 (1)
Criterion 3 √3
|IA| < I1 min
IA OPEN TIMER
PU
DO 𝐶1 𝑖s the positive-sequence line capacitance, ω is the angular
ROCOCPhA
IB OP EN
frequency, and 𝑉𝐿−𝐿 is the line voltage.
IC OP EN
ROCOCPhB
ROCOCPhC For the Current Magnitude Detector to assert, the phase
current magnitude (|IPhase|) in one of the phases (phasex) must be
Fig. 1. Conventional |I2/I1| current-based FCP method. less than the total line charging current (Ich). That is, |IPhase,x| <
Ich. The phase current magnitude in the other two phases
The third criterion requires the current on the broken phase
(phasey and phasez) must be healthy and greater than Ich, i.e
to be near zero and should have experienced a significant drop
|IPhase,y,z| > Ich. Also, the ROCOC operand of the phase with the
in current magnitude within a time period (Rate-of-Change of
broken conductor must assert.
Current, ROCOC), while the remaining two healthy phases
The ‘PF Detector’ and ‘Current Magnitude Detector’ provide
have a current above a minimum threshold. Since asymmetrical
overlapping zones of detection across the entire Transmission
shunt faults within or beyond the protection zone may also
line. For example, the Current Magnitude Detector is more
cause high |I2/I1| values that are similar to broken conductor
effective for close-in breaks, while the PF Detector is more
events, this method poses some challenges for traditional
suitable for conductor breaks away from the line terminals since
protection schemes to detect. Intentional time delays may be
the relay might be unable to calculate the power factor at that
added using timers for coordination between the FCP method
line terminal from small charging current magnitudes.
and the operating or reclosing times of conventional protective
relaying schemes. III. IMPEDANCE-BASED FALLING CONDUCTOR DETECTION
B. Modified I2/I1 Current-Based Method METHOD
This section describes a modified |I2/I1| current-based This section presents an impedance-based High-speed
method implemented using charging current and power factor Falling Conductor Protection (HFCP) method which utilizes the
detectors as shown in Fig. 2. local and remote voltages and currents already available in line
differential protection through existing communication
channels. Fig. 3(a) shows an implementation using line currents
 3

and extended Clarke’s voltage communicated from the remote positive-sequence voltage, and synchrophasor data. Normally,
end relay to the relay at the local end which then communicates one set of this voltage data is available in each system. The logic
these quantities to a Real-Time Controller running the FCP diagram is presented for a three-terminal line. However, the
algorithm. Another implementation requires relays, IEDs, or configuration for a two-terminal line is the same, except the
Phasor Measurement Units (PMUs) to stream synchrophasor third terminal is removed from the configuration.
measurements of three phase currents and phase voltages (from In the block diagram of Fig. 4, Va, Vb, and Vc are voltage
each of the line terminals) to the Real-Time Controller running synchrophasors, V1 denotes positive-sequence voltage, and Vcl
the FCP algorithm. The main requirement is that the exchanged is extended Clarke voltage. The main building blocks of the
data is synchronized. If the phase voltages of a two- or three- proposed HFCP algorithm are impedance calculation, Delta-Z
terminal line are available, utilizing synchrophasor data, this calculation and detection algorithm as shown in Fig. 6. They are
data may be streamed to a real-time controller (RTC) or phasor discussed in detail below.
data concentrator (PDC) to detect broken falling conductor
conditions as shown in Fig.4. The logic diagram in Fig. 4 shows
the proposed detection logic using extended Clarke voltage,

Vx Vy
Ix Zx T Zy Iy

Zz Vcl,y
Iy,abc
IEDx IEDy
Vcl,z Iz
Iz,abc
Vcl,x-Vcl,y-Vcl,z Existing Differential Channel
Ix,abc-Iy,abc-Iz,abc Analog Goose or PMU Data
IEDz
Real-Time
Vz
Controller

Located Inside Substation

(a) (b)
Fig. 3. Data flow for (a) HFCP using extended Clarke voltage or positive-sequence voltage, and (b) HFCP using synchrophasor measurements.

Va
Vb V1 Vcl
ΔZ Phase A BCD
Vc ICharge (GOOSE Item x)
Relay Phase A 0
Ia ITerm1 0
@ 0.5s
Terminal 1
Ib ITerm2
Ic Phase B BCD
q Phase A
0 (GOOSE Item x)
t Phase B 0
ΔZ 0.5s
Broken Conductor Detection
Impedance Calculation

Va ICharge Phase C BCD

Vb V1 Vcl ITerm1 0 (GOOSE Item x)
Vc ITerm2 Phase C 0
Relay 0.5s
Ia
@ Phase B
Ib
Terminal 2
Ic
q ΔZ
t ICharge
ITerm1
Va ITerm2
Vb V1 Vcl Phase C
Vc
Relay
Ia voltage_low_a
@
Ib
Terminal 3 voltage_low_b
Ic
Specific to three-terminal q voltage_low_c
line t Line_energize
Current_high

Setpoint BCD Enabled

Impedance Change Setpoint (Default=15)
Security Count Setpoint (Default=2)

Fig. 4. Block diagram of the impedance-based HFCP method.

 4

′
𝑍𝑥𝑧,𝑎 −𝑍𝑥𝑧,𝑎
A. Impedance Calculation ᵟ𝑍𝑥𝑧,𝑎 = | ′ | (4)
𝑍𝑥𝑧,𝑎
The impedance calculation block calculates the impedance of
where:
the line using the available local and remote data from a two-
Zxy,a is the calculated A-phase line impedance from
terminal line (Fig. 5a) or a three-terminal line (Fig. 5b).
terminal X to terminal Y.
Traditional protection relays, typically poses multi-ended fault
Zxz,a is the calculated A-phase line impedance from
location features which requires the voltages of the line
terminal X to terminal Z.
terminals to be exchanged over the line current differential
Z’xy,a’ is the calculated A-phase line impedance from
protection communication channel. Consequently, the local
terminal X to terminal Y from a few cycles ago.
voltages are communicated to the remote line terminal through
Z’xz,a’ is the calculated A-phase line impedance from
direct channel (e.g., using IEEE C37.94 standard). The
terminal X to terminal Z from a few cycles ago.
proposed method uses the existing communicated data between
line current differential relays to detect a broken/falling
The impedance-based HFCP function will identify a falling
conductor. For a three-terminal line, two impedances are
conductor condition when the rate of change of impedance for
calculated for each phase of the line at each terminal (i.e.,
the transmission line exceeds a threshold (defaulted at 15 times
impedances between the
the normal value). Only single-phase broken/open conductors
local terminal and two remote terminals). Thus, there will be a
can be detected with this algorithm. To prevent incorrect
total of six and eighteen impedance calculations for two-
operation of the impedance-based HFCP, for a fault occurring
terminal and three-terminal lines, respectively. Fig. 5 and 6
on the transmission line, a high current threshold is utilized to
show the PI model of the two- and three- terminal lines.
block the algorithm if the line current exceeds a predefined
value (defaulted at 1.2 pu). The algorithm will also be blocked
when the phase voltage is outside a pre-defined range at all line
terminals, indicating abnormal conditions other than a broken
(a) falling conductor. The impedance-based method is immune to
existing system imbalance and transient events since it is using
the ICR over time.

IV. SIMULATION STUDIES AND RESULT ANALYSIS

The performance of the three FCP methods was validated
using both hardware-in-the-loop Real-Time Digital Simulator
(RTDS) simulations and field events playback, respectively.
Fig. 6 shows a section of the 220 kV transmission system
used in testing the FCP methods presented in this paper. The
transmission system was built in the RSCAD (RTDS) software.
The protected line is a 220 kV, 38-mi long transmission line
(b) between Bus-5 and Bus-6. The CT and PT ratios used were 400
and 2000, respectively. The calculated line charging current
Fig. 5. PI model representation of a (a) two-terminal transmission line, and (b) was 28 A.
three-terminal transmission line. Per phase broken conductor (BC) events were applied at 5%
interval (5% to 95%) on Line 5-6 and at 20% of lines adjacent
B. Impedance-Based HFCP Algorithm to Bus-5 and Bus-6 (Line 5-7, Line 5-8, Line 3-6). The security
of the FCP methods was tested for ground and phase faults on
Once the line impedances are calculated, the ICR of the
Line 5-6, at lines adjacent to Line 5-6, and at buses two stations
transmission line (δZ) is derived by subtracting the previous
away from Bus-5 and Bus-6 as indicated in Fig. 6.
impedance Z’ (Z’ = Zt0-n) from the current impedance Z (Z t0)
The HIL testbed consists of line protection IEDs, real-time
and then divided by previous impedance Z’, as follows:
automation controller, ethernet switch, GPS clock, low voltage
𝑍−𝑍 ′
amplifiers, and the RTDS. The broken conductor protection
ᵟ𝑍 = | | (2) scheme logic using the conventional |I2/I1| method and the
𝑍′
Taking Phase ‘a’ at Terminal X of a three-terminal line as an Power Factor-based |I2/I1| method were implemented and
example, the ICR (ᵟ𝑍𝑥𝑦,𝑎 ᵟ𝑍𝑥𝑧,𝑎 ) can be calculated using the deployed on two pairs of line protective IEDs located at both
ends of the transmission line under test.
following formulas (other phase ICRs are calculated similarly):
′
𝑍𝑥𝑦,𝑎 −𝑍𝑥𝑦,𝑎
ᵟ𝑍𝑥𝑦,𝑎 = | ′ | (3)
𝑍𝑥𝑦,𝑎
 5

Bus-3 Bus-4 Bus-5

Bus-1 Bus-6 Bus-7

Bus-2
FL7
Bus-9
G
G
BC1/FL1
FL9 FL13
BC2/FL2 FL11
FL10 BC4/FL4 Bus-10
Bus-8

FL8 FL6
G

BC3/FL3
FL5 FL12 FL14

Fig. 6. A section of the 220-kV transmission system.

These IEDs are interfaced with the RTDS through low- coordination with conventional protective relaying schemes
voltage amplifiers. The HFCP function was implemented on a during faulted events.
real-time automation controller that is interfaced with the Fig. 7 shows the three phase currents recorded by the relay at
RTDS. One of the IED pairs at both ends of the transmission Bus-5. The measured current on the Phase-A conductor before
line was configured to stream synchrophasor measurements at the break were 126 A on all the phases. After the break, the
60 frames per second (fps) onto the network for the impedance- relays at Bus-5 and Bus-6 both measured a continuous current
based HFCP method. The automation controller has PDC, PLC, drop as the conductor separated. The conductor physically
and gateway capability. Thus, it has the capability to receive broke after about 410 ms. From Fig. 3, three criteria must be
and time-align the synchrophasors streamed from the line satisfied for the conventional |I2/I1| FCP method to operate.
terminals. Also, the HFCP algorithm is implemented on the From Fig. 7, the |I1|<I1max (Phase TOC1 DPO) was satisfied,
automation controller. The communication of the while Criterion 2 (|I2/I1| ratio) asserted at 23 ms after the
synchrophasor measurements from the IEDs at both line conductor completely separated, and Criteria 3 (IA OPEN)
terminals may utilize wideband ethernet, point-to-point radio, asserted 13 ms after the conductor completely separated.
LTE cellular radios, fiber optics. Any robust communication TABLE I
with a latency or time delay under 50 ms is sufficient. SETTINGS FOR CONVENTIONAL |I2/I1| FCP METHOD
A comprehensive set of test cases including internal/external Settings Parameters
broken conductor events, internal/external faults, dynamic line
I1 min (pu) 0.05
loading, and transient incidents were executed for the I1max (pu) 1.20
transmission system using the RTDS-HIL testbed. The Rate-of-change of current (pu) 0.02
proceeding sub-sections present and discuss the use cases I2/I1 Slope (%) 15.00
considered. Pickup time delay (ms) 100.00
A. Case Study-1: Lightly-Loaded Operating Condition
In this case study, a slowly-evolving close-in broken
conductor event was simulated on Phase-A conductor of Line
5-6 at 5% from the local end (Bus-5) for a lightly-loaded
operating condition.
The load flow on the protected transmission line was 52 MW
(6.5% of the maximum line carrying capability). Tests were
performed for the three FCP methods presented in Sections II
and III.
1) Conventional I2/I1 FCP Method
The settings used in testing the conventional I2/I1 FCP
settings are presented in Table I. An I1min setting of 0.05 pu
applies a cut-off value to the magnitudes and angles of the
measured currents and substitutes the measurements with zero Fig. 7. Results for conventional |I2/I1| FCP method for a Phase-A broken
conductor event at 5% of Line 5-6.
if below this cut-off. An I1max threshold setting of 1.2 pu
indicates the presence of a fault condition. An |I2/I1| setting of The pick-up (BRKNCND PKP) operand asserted after 45
15% was used in these tests to increase the sensitivity of the ms, while the conventional FCP method (BROKEN COND)
method and test the dependability and security of the signal operated 145 ms after the conductor completely broke
conventional |I2/I1| FCP for a sensitive |I2/I1| setting. An due to the intentional time delay setting of 100 ms. The
intentional time delay pickup of 100 ms was added to provide operating time is calculated as the elapsed time from when
the conductor completely broke to when the broken
 6

conductor method operated. The remote end relay at Bus-6 The Phase-A current recorded by the relay before the
sees the conductor break located 95% from Bus-6, recorded conductor break was 126 A. After the break, the relays at Bus-
the continuous current drop from the inception of the 5 and Bus-6 both measured a continuous current drop as the
conductor break, and had a similar response to the relay at conductor separated. The charging current measured by Bus-5
Bus-5. The FCP relay at Bus-6 picked-up and operated at 53 and Bus-6 relays were 1.6 A and 27 A, respectively.
ms and 153 ms, respectively. The conventional I2/I1 FCP From Fig. 4, three criteria must be satisfied for the modified
method was dependable, detected all in-zone broken |I2/I1| FCP method to operate. Criterion 1 |I1|<I1max (Phase
conductor conditions, and was secure for internal phase and TOC1 DPO) was satisfied, while Criterion 2 (|I2/I1|) asserted
ground faults. However, it wrongly operated for external 20 ms after the conductor break. For a close-in conductor break
faults at some remote locations such as Line 5-7, Line 5-8, at 5% of the line, the ‘PF detector’ will not assert or will
and at Bus-1, Bus-2, Bus-5, Bus-7, Bus-8, and Bus-10. experience instability (‘PFA OP’ signal shown in Fig. 8)
because of the measurement inaccuracies for close-in conductor
2) Modified I2/I1 FCP Method
breaks. This is because the charging current measured by the
The settings used in testing the modified I2/I1 FCP method
protective relay is negligible and the protective relay may not
are presented in Table II. An I1min setting of 0.05 pu is the
calculate the phase current angle from such measurements.
measured current cut-off value, while I1max threshold setting
However, the overlapping zone ‘|IA|<Ich’ Detector’ (PHASE
of 1.2 pu indicates the presence of a fault condition. An |I2/I1|
TOC2 DPO A signal) asserted at 23 ms and Criterion 3 (IA
setting of 15% was also used to increase the sensitivity of the
OPEN) was satisfied 107 ms after the conductor break.
method and test the dependability and security of the modified
The pick-up (BRKNCND PKP) operand of the modified FCP
|I2/I1| FCP for a sensitive setting.
method asserted 140 ms after the complete separation, while the
For the transmission system tested, the transmission line
modified FCP method (BROKEN COND) signal operated 240
angles for a single conductor per phase were approximately 79° ms after the conductor break due to the configured intentional
or less during short circuit faults. Therefore, a power factor time delay setting of 100 ms. The remote end relay (at Bus-6)
setting of 90° ± 9° was selected since this provided consistent also recorded the continuous current drop following the
results for close-in breaks and did not encroach on the inception of the evolving conductor break. In this case, the
transmission line angles during fault events. A charging current break was 95% from Bus-6 relay. The PF Detector operated
detector setting of 0.007 pu was determined through testing to instead of the ‘|IA|<Ich’ Detector since the latter will not detect
provide a 30% reach from both line terminals. The lower- and for remote conductor breaks. The combined use of the ‘PF
upper- current setpoints used in calculating the rate-of-change Detector’ and the ‘Phase Detector’ in Criteria 3 provided an
of current are determined by applying multiplication- and overlapping (100%) zone of detection for a falling conductor
summation- factors to the line charging current. Fig. 8 shows event. The modified |I2/I1| FCP method was dependable,
the three phase currents and the Phase-A power factor recorded detected all the in-zone broken conductor conditions, and did
by the relay at Bus-5 configured with the modified |I2/I1| FCP not misoperate for phase or ground short circuit faults.
method. 3) Impedance-Based FCP Method
TABLE II The settings used in testing the impedance-based FCP
SETTINGS FOR MODIFIED |I2/I1| FCP METHOD method are presented in Table III. An I1min setting of 0.002 pu
Settings Parameters is the minimum current cut-off, while an I1max setting of 1.2
I1 min (pu) 0.05 pu indicates the presence of a fault condition and blocks the
I1max (pu) 1.20 impedance-based algorithm from operating. The current delta
Upper current setpoint (pu) 0.034 setpoint is used for fault detection and interlocking of the output
Lower current setpoint (pu) 0.017 of the FCP method for 2 seconds. The charging current factor
Rate-of-change of current calculation time (ms) 100.00 ensures that the current phasors are less than the line charging
Charging current detector (pu) 0.007
Power factor detector pickup 90°±9°
current. The pickup time delay is an intentional time delay
I2/I1 Slope (%) 15.00 before tripping once the impedance rate of change conditions
Pickup time delay (ms) 100.00 are met. All other settings are self-explanatory.
Fig. 9 shows the results for the impedance-based HFCP
method for the same broken conductor event at 5% from the
local end (Bus-5). The maximum calculated Impedance Change
Ratio (ICR) before the break was 32 on all the phases. While
the calculated ICRs after the break were approximately 350, 40,
32, on Phases A, B, C, respectively.
The impedance-based HFCP method operated on Phase A
since its ICR exceeded the pickup threshold. The pickup
threshold used for these tests was 65. The results obtained show
that the impedance-based method was very sensitive and was
able to detect conductor breaks for lightly-loaded conditions as
low as 32 MW (3% of the maximum line loading).
Fig. 8. Results for the modified |I2/I1| FCP method for a broken conductor
event at 5% of Line 5-6.
 7

TABLE III B. Case Study-2: Heavy Loading Operating Condition

SETTINGS FOR THE IMPEDANCE-BASED FCP METHOD
A slowly evolving close-in broken conductor event was
Settings Parameters simulated on Phase-A conductor at 5% from the local end (Bus-
Imin (pu) 0.002
5) for a heavy-loading operating condition. The load flow on
Imax (pu) 1.20
the transmission line was 430 MW. Tests were performed for
Current delta setpoint 5.00
Charging current factor 30.00
the three FCP methods discussed in Sections II and III,
Number of samples used for delta 10.00 respectively.
comparison 1) Conventional I2/I1 FCP Method
Rate of change of impedance 3.00 Fig. 10 shows the three phase currents and binary signals
Per phase impedance delta threshold 65.00 recorded by the line protective relay at the local end (Bus-5) for
Per phase impedance calculation threshold 150.00
a Phase-A broken conductor event at 5% of Line 5-6. The three
Vmin (pu) 0.001
phase currents before the conductor break were 1080 A on all
27V pickup (pu) 0.85
Pickup time delay (ms) 50.00
phases. All the three criteria (|I1|<I1max (Phase TOC1 DPO),
|I2/I1| ratio, and IA OPEN) required for the conventional |I2/I1|
The above case studies were repeated for fast conductor FCP method were satisfied. The conventional |I2/I1| FCP
breaks and similar characteristics were obtained for the three method at Bus-5 operated at 155 ms and the relay at Bus-6 had
FCP methods. Table IV summarizes the findings for the lightly- an operating time of 156 ms.
loaded operating condition. The conventional |I2/I1| FCP Criteria 1-3 all operated, the FCP pick-up (BRKNCND PKP)
method was able to detect broken conductor events as low as 52 operand asserted 54 ms after the conductor completely broke,
MW (6.5% of the maximum line loading), was dependable, while the FCP trip (BROKEN COND) operand operated after
detected all the in-zone broken conductor conditions, and was about 155 ms. The relay at Bus-6 also detected the break, the
secure for internal phase and ground faults. However, it BRKNCND PKP operand asserted after 54 ms, while the FCP
wrongly operated for external faults at some locations such as trip (BROKEN COND) operand operated after the configured
Line 5-7, Line 5-8, and at Bus-1, Bus-2, Bus-5, Bus-7, Bus-8, time delay of 100 ms.
Bus-10. The modified |I2/I1| method was dependable, secure,
and detected broken conductor events for lightly-loaded
operating conditions of 41 MW (5% of the maximum line
loading). The impedance-based method was dependable, secure,
sensitive, and was able to detect conductor breaks for lightly-
loaded conditions as low as 32 MW (4% of the maximum line
loading).
Phase A

Phase A

Za_delta
Za_delta

Phase B
Phase B
Zb_delta
Zb_delta

Fig. 10. Results for conventional |I2/I1| FCP method for a broken conductor
Phase C
event at 5% of Line 5-6.
Zc_delta Phase C
Zc_delta

2) Modified I2/I1 FCP Method

For the modified I2/I1 FCP method, the three phase currents
recorded by the relays before the conductor break was 1080 A.
Fig. 9. Results for impedance-based FCP method for a broken conductor event The relays at Bus-5 and Bus-6 both measured a continuous
at 5% of Line 5-6
current drop as the conductor separated mechanically. The
TABLE IV charging current on the broken phase (Phase-A) after the
SUMMARY OF TEST RESULTS FOR LIGHTLY-LOADED conductor break was about 1.7 A at End-1, and 28 A at End-2.
OPERATING CONDITION As shown in Fig. 11, the three criteria required by the
Characteristics Conventional Modified Impedance- modified I2/I1 FCP method to operate were satisfied. The PF
|I2/I1| FCP |I2/I1| FCP based FCP Detector will not assert for a close-in conductor break at 5% of
Security during ✓ ✓ ✓ the line. However, the back-up ‘|IA|<Ich’ Phase Detector
internal faults asserted in 22 ms after the conductor break and resulted in the
Security during – ✓ ✓ assertion of Criterion 3 (IA OPEN) at about 120 ms. The
external faults
Coverage for line ✓ ✓ ✓
modified |I2/I1| FCP method at Bus-5 operated at 248 ms, while
loading greater than the relay at Bus-6 operated 250 ms after the conductor
0.13 pu completely separated.
Minimum arming 52 41 32
threshold (MW)
Operating time (ms) 150 250 185
 8

based method may not be tested due to the unavailability of

event records synchronized to a common time reference and
with the same record length.
A. Field Event-1: Broken Conductor Event at 25% of the
Transmission Line
This field event was a broken conductor event on Phase-A of
the protected line (Line 5-6 shown in Fig. 6) at about 25% from
the local bus (Bus-5).
1) Conventional I2/I1 FCP Method
Fig. 13 shows the results when the DFR field event record
was played back to the protective relays at Bus-5 and Bus-6
configured with the conventional |I2/I1| BCD method.
The line loading before the conductor break was 108 MW
Fig. 11. Results for the modified |I2/I1| FCP method for a broken conductor
(270 A). As shown in Fig. 13, the transmission line experienced
event at 5% of Line 5-6. a continuous drop in Phase-A current magnitude as the
conductor slowly separated mechanically.
3) Impedance-Based FCP Method All the three criteria (|I1|<I1max (Phase TOC1 DPO), |I2/I1|
Fig. 12 shows the results for the impedance-based HFCP ratio, and IA OPEN) required for the conventional |I2/I1| FCP
method for the same broken conductor event. The maximum method were satisfied. The conventional |I2/I1| FCP method
ICR before the break was 32 on all the phases. The ICRs after operated in 110 ms at Bus-5, and in 120 ms at Bus-6.
the break were 3630, 37, and 33 on Phases A, B, C, respectively.
The impedance-based HFCP method operated on Phase A in
about 185 ms since the calculated ICR exceeded the pickup
threshold. The pickup threshold used for these tests was 65.
The above tests were repeated for fast conductor breaks and
similar characteristics were obtained for the three FCP methods.
Table V summarizes the finding for this case study.
Phase A

Phase A
Phase A

Za_delta
Za_delta
Za_delta

Phase B
Phase B Phase B
Zb_delta
Zb_delta Zb_delta

Phase C
Fig. 13. Results for the conventional |I2/I1| FCP method for a broken conductor
Zc_delta Zc_delta Phase C Phase C
Zc_delta event at 25% of Line 5-6.

2) Modified I2/I1 FCP Method

Fig. 14 shows the results when the DFR field event record
Fig. 12. Results for impedance-based FCP method for a broken conductor event was played back onto the protective relays at Bus-5 and Bus-6
at 5% of Line 5-6. configured with the modified |I2/I1| BCD method.
All the three criteria required for this FCP method to operate
TABLE V
SUMMARY OF TEST RESULTS FOR HEAVY LOADING CONDITION
were satisfied. The Power Factor logic detected and operated
Characteristics Conventional Modified Impedance- for this broken conductor event. The power factor operands for
|I2/I1| FCP |I2/I1| FCP based FCP Phases A and B (PFA OP and PFB OP) on the Bus-5 Protective
Security during ✓ ✓ ✓ IED showed some chattering as shown in (Fig. 14).
internal faults Similar results were seen by the protective relay at Bus-6.
Security during ✓ ✓ ✓ Instability was seen on Phases A, B, and C (PFA OP, PFB OP,
external faults
Operating time (ms) 150 250 185
PFC OP) operands. However, the Phase-A Power Factor (PFA)
operand correctly asserted first and resulted in the operation of
the FCP method.
V. PLAYBACK TESTING WITH FIELD EVENTS The three criteria (|I1|<I1max (Phase TOC1 DPO), |I2/I1|,
and IA OPEN) required for the conventional |I2/I1| FCP method
This section presents the playback of field event records
were satisfied. The conventional |I2/I1| FCP method at Bus-5
captured by Digital Fault Recorders (DFRs) located at the local
operated in 220 ms and in 237 ms at Bus-6.
(Bus-5) and remote (Bus-6) substations, respectively.
Protection relay test-set were used in playing back the DFR
records to protection relays configured with the conventional
|I2/I1| and the modified |I2/I1| FCP methods. The impedance-
 9

asserted. Both PFA and PFB power factor detectors asserted.

However, the IA OPEN and IB OPEN operands correctly
restrained and did not assert for this external fault. This FCP
method was secure and did not detect or operate for this remote
ground fault.

Fig. 14. Results for the modified |I2/I1| FCP method for a broken conductor
event at 25% of Line 5-6.

B. Field Event-2: Remote A-g Fault

This field event was a remote single-line-to-ground (SLG)
fault on Phase-C of Line 2-3 (FL9) of the transmission system
shown in Fig. 6.
1) Conventional I2/I1 FCP Method Fig. 16. Results for the modified |I2/I1| FCP method at Bus-5 for a remote C-g
Fig. 15 shows the shows the results when the DFR field event fault event at Line 2-3.
record for Line 5-6 was played back onto the protective relays
at Bus-5 and Bus-6 configured with the conventional |I2/I1| VI. CONCLUSION
FCP method. The line loading was 83 MW (209 A) prior to the This paper presented a comparative analysis of three
external C-g fault event. The conventional |I2/I1| FCP method transmission line Falling Conductor Protection (FCP) methods
incorrectly identified this ground fault as a broken conductor – the conventional |I2/I1|, modified |I2/I1|, and the impedance-
event due to the drop in magnitude on Phase-A current. This based HFCP methods, respectively. The three FCP methods
resulted in the assertion of the IA OPEN criterion. Also, the were compared with respect to operating speed, dependability,
|I1|<I1max (Phase IOC1 DPO) criterion and |I2/I1| criterion and security for various conductor breaks, fault events, and
asserted as well. The conventional |I2/I1| FCP method picked- switching events on the protected line and adjacent buses and
up (BRKNCND PKP) 50 ms after fault inception. However, the lines. Also, slowly evolving conductor breaks which are often
conventional |I2/I1| FCP method correctly restrained and did difficult to detect by FCP methods based on rate-of-change
not operate because of the configured time delay. were tested and validated.
The tests were performed using the Real-Time Digital
Simulator (RTDS) and playback of past field events. The
conventional |I2/I1| FCP and modified |I2/I1| FCP methods
correctly operated for all the broken conductor events within
150 ms and 250 ms, respectively.
However, the conventional |I2/I1| FCP misoperated for
ground faults at some remote locations. These scenarios are
highly unlikely because line protection, bus protection, and
breaker failure relays are expected to operate for these ground
faults before the conventional |I2/I1| FCP method operates.
The modified |I2/I1| FCP method correctly restrained for all
in-zone phase- and ground-faults, and out-of-zone broken
Fig. 15. Results for conventional |I2/I1| FCP method at Bus-5 for a remote C-g
conductor and fault conditions. This FCP method is suitable for
fault event at Line 2-3.
transmission lines that have measurable charging current.
2) Modified I2/I1 FCP Method Before deploying the modified FCP method, the appropriate
Fig. 16 shows the three phase currents, per phase power settings must be derived from the line charging current. The
factors, and binary signals recorded by the relay at Bus-5 action of shunt reactors may lead to the absorption of reactive
configured with the modified |I2/I1| FCP method. The power/capacitive currents in the system. Consequently, this
|I1|<I1max (Phase TOC1 DPO) and |I2/I1| ratio operands may lead to varying line charging current in the system. Also,
 10

since different lines may have different line charging currents The evaluation presented in this paper provided a great
depending on the prevailing operating condition of the system, opportunity to thoroughly validate the performance of various
it is recommended that RTDS testing be performed for each line transmission line falling conductor protection methods before
before deploying the modified |I2/I1| FCP method. field deployment in utilities. Also, the RTDS testing of the
Furthermore, CT accuracy may affect the sensitivity of the protective algorithms or hardware using realistic power system
modified |I2/I1| FCP method. The analog current inputs to models and scenarios not only allows protection engineers to
protective relays are from protection CTs which has a Class C verify the settings philosophy of the fundamental components
accuracy with a 1% – 3% ratio error at rated nominal current. within a protection scheme, but also greatly assists utilities in
The relay introduces an additional error of ±0.25% of uncovering, and fixing errors or limitations before field
measurement or ±0.1% of rated current, whichever is greater. deployment.
Therefore, the overall accuracy of the input current at the rated
input is less than 4%. In protection applications, currents below VII. REFERENCES
0.05 pu are generally not used as a result. Therefore, further [1] T. Rahman, C. Bolton, and E. A. Udren, “Transmission Line Falling
research and validation might be required when measuring very Conductor Protection System development at SDG&E”, 49 th Annual
low current inputs. This may be a concern for close-in Western Protective Relay Conference (WPRC), October 10-13, 2022, pp.
1-22.
conductor breaks. However, the remote terminal would reliably
[2] Yujie Yin, Hasan Bayat, Nathan Dunn, Matthew Webster, Alfredo
detect these breaks and initiate a transfer trip (e.g., a Direct Marquez, Kiet Tran, and Arturo Torres, “High-Speed Falling Conductor
Transfer Trip (DTT)) to de-energize the transmission line. Protection for Electric Power Transmission Systems”, 49 th Annual
The impedance-based HFCP method performed well for all Western Protective Relay Conference (WPRC), Spokane, Washington,
the in-zone broken conductor events and operated within 200 October 10-13, 2022, pp. 1-5.
ms. The impedance-based HFCP method correctly restrained [3] GE L90 Line Current Differential Protection Instruction Manual, “Broken
Conductor Detection”, 2021.
for all in-zone phase- and ground-faults, and out-of-zone [4] Kanchanrao Dase, Sajal Harmukh, and Arunabha Chatterjee, “Detecting
broken conductor and fault conditions. Also, the impedance- and Locating Broken Conductor Faults on High-Voltage Lines to Prevent
based HFCP solution had the best sensitivity and was able to Autoreclosing onto Permanent Faults”, 46th Annual Western Protective
operate for broken conductor events during lightly-loaded Relay Conference (WPRC), Spokane, Washington, October 22-24, 2019,
conditions as low as 32 MW (3% of line loadability). The pp. 1-20.
impedance-based HFCP method uses synchrophasor data
available on most Line protective relays, IEDs, or PMUs and
can take advantage of the existing IEEE C37.94 communication
channel between line differential relays. Therefore, it does not
require a separate communication channel. Furthermore, the
impedance-based HFCP method is vendor-agnostic, and the
application can be deployed on any synchrophasor-compliant
real-time controller.