16th NATIONAL POWER SYSTEMS CONFERENCE, 15th-17th DECEMBER, 2010 653

Estimation and Elimination of DC Component in

Digital Relaying
K Venkatesh and K S Swarup, Senior Member, IEEE

Abstract—This paper propose a new decaying DC component effect of an exponential decaying DC component based on the
estimation algorithm for digital relaying. Fault currents tend to weighting least error square(LES) technique is proposed [8]
include a DC decaying component. This component decreases In this paper,describe a new and unique algorithm to estimate
the accuracy and speed of the protection relay operation. The
proposed algorithm can estimate and eliminate the DC decaying and eliminate the decaying DC component in a fault curre
component from fault current signals after one cycle from the nt signal.and also describe the mathematical derivation of the
fault instant. Also, it can be applied to a conventional discrete proposed algorithm.The magnitude and time constant of the
Fourier transform to calculate phasor quantities of fault currents DC component were estimated by integrating the fault current.
in a digital protection relay. In the proposed algorithm, the DC
decaying magnitude and time constant are estimated exactly by
integrating fault currents during one cycle. The DC decaying II. P ROBLEM F ORMULATION
component is eliminated by subtracting the DC value at each
sampling instant. To verify the performance of the proposed Generally, the fault current not only has a fundamental
algorithm, we performed a DC component estimation test and component but also harmonics and a decaying DC component.
distance protection test using PSCAD/EMTDC. The results of the Fundamental and harmonic components can be represented as
PSCAD/EMTDC simulation showed that the proposed algorithm a sinusoidal function. The decaying DC component can be
can estimate DC components exactly from fault currents and can
represented as a decaying exponential function. So the fault
be applied to digital protection relays for phasor extraction
current can be mathematically expressed
Index Terms—DC estimation, DC magnitude, DC time con-
p
stant, digital protection, DFT, Fault current −t X
i(t) = I0 e τ + IK sin(kω1 t + θk ) (1)
k=1
I. I NTRODUCTION
Where I0 is the magnitude of the decaying dc offest,τ is the
T HE Power system protection is a branch of electrical
power engineering that deals with the protection of
electrical power systems from faults through the isolation of
time constant of the decaying dc offset,K is the harmonic
order, IK is the magnitude of the Kth harmonic component,θk
is the phase angle of the kth harmonic component, and P is
faulted parts from the rest of the electrical network. DFT
the maximum harmonic order.
is the most preferable method to extract the fundamental
If equation (1) integrates during one period (T) the integral
phasor quantities from waveforms [1]. DFT has immunity
of the second term in is zero, and only the integral of the first
from harmonic components and has a relatively fast response
term, which is related to the decaying DC component, remains
time for the fundamental component calculation However, the
Z t Z t p
DFT is not immune from the DC component,and the decaying −t X
DC component in the fault current can cause undesirable i(t) d = [I0 e τ + IK sin(kω1 t + θk )]dt
t−T t−T k=1
oscillations in the DFT results[2,3] Up till now, many research Z t
−t
studies were conducted to remove the DC component from = I0 e τ dt
fault current waveforms for protection relaying[2,4]. t−T
−t −T
One is the DC component filtering method,which extracts = −I0 τ.e τ (1 − e τ )
the fundamental component only from the original signal
= Z(t) (2)
without a DC component calculation. In [2], a mimic filter
was proposed to remove the decaying DC component over a In equation 2, let the integral of the DC component during
wide range of time constant. Recently an adaptive compen- one period at time t be Z(t) Then, Z(t + ∆t), which is
sation method to remove a decaying DC offset component represented as the integral of the DC component after a small
from the fault signals has been described in [5] A modified time step, is expressed
DFT algorithm to efficiently compute and eliminate the DC
component using full cycle or half cycle data windows has −t+∆t T
been proposed in [6] The technique for removal of a decaying Z(t + ∆t) = −I0 τ.e τ (1 − e τ ) (3)
DC offset on phasor estimates using the DFT is described in = Z(t).e
−∆t
τ (4)
[7] An adaptive phasor estimation algorithm to suppress the
From 4, if we know the integrals of fault current during one
K.Venkatesh and K. S. Swarup are with the Department of Electrical
Engineering, Indian Institute of Technology, Madras, Chennai, India. e-mail: period, the equation can be used to calculate the time constant
swarup@ee.iitm.ac.in. τ and magnitude (I0 ) of the decaying DC component:
Department of Electrical Engineering, Univ. College of Engg., Osmania University, Hyderabad, A.P, INDIA.
 16th NATIONAL POWER SYSTEMS CONFERENCE, 15th-17th DECEMBER, 2010 654
2

Data sampling
∆t
τ =− (5)
ln Z(t+∆t)
Z(t) Dc time constant calculation

Z(t)
I0 = − −t T
(6) Dc value calculation (1 cycle)
−τ.e (1 − e τ )
τ Sampling buffer (1 cycle)

In 5 and 6, to obtain Z(t) and Z(t + ∆t), we should only Dc value subtraction (1 cycle)
calculate the integral of the measured fault current at t and
t + ∆t.
DFT calculation

III. I MPLEMENTATION OF THE A LGORITHM

Fundamental component
In a practical digital relaying scheme, all of the calculations
are performed in a discrete time base using sampled data
and should be completed in each sampling period. For the Fig. 1. Procedure for the fundamental component calculation
practical application of the algorithm, fast calculation time is
required. equation magnitude1, it is clear that the decaying DC IV. T EST R ESULTS
component can be Mathematically calculated. However, since
the calculation of the natural logarithm should be performed A. Static Simulation Test
for every sample, it can be a computational burden from the In order to verify the performance of the proposed algo-
practical perspective. In order to reduce the computational rithm, two types of simulation tests were performed with
load, the time constant can be obtained by using a Taylor PSCAD/EMTDC. The first simulation test was a static test.
series expansion in (5) In this test, several sampled signals, which contained a DC
−∆t 1 1 Z(t + ∆t) component, were applied to verify the performance of the DC
e τ = 1 + (−∆t) + (−∆t)2 2 + .... = (7) component. The calculated time constants and fundamental
τ τ Z(t)
components were compared to the applied signals.
In equation 7 ∆t is the sampling period, so ∆t is much lower
than, the time constant of the power system. Thus, we can B. Estimated Time Constants for Change in Magnitude
use only the first two terms to calculate the time constant and
A static simulation test was performed to evaluate the per-
simplify the time constant to be
formance of the DC magnitude and time constant calculation
−∆t 1 Z(t + ∆t) algorithm. Test signals consisted of a fundamental component
e τ = 1 + (−∆t) = (8)
τ Z(t) and a DC component with different magnitudes and time
1 Z(t + ∆t) 1 constants. The ratio of the magnitude of the fundamental
= (1 − ) (9)
τ Z(t) ∆t component and the decaying DC component was set to 0.2,
Also, the magnitude of the decaying DC component can be 0.4, 0.6, 0.8, and 1.0 p.u. The time constants used for the
calculated by applying t = T to 6. Therefore, we acquire the performance evaluation were 5, 25, 50, 100, 150, and 200
following equation: ms. Also, the sampling rate was set to 64 samples per cycle.
Table I shows the estimated time constants using the proposed
Z(T )
I0 = − −T (10)
−τ.(e τ − 1) algorithm for the applied time constant and ratio of magnitude
But the calculation of the magnitude of the DC component changes. It can be seen from Table I that the estimated time
is not necessary to remove the DC component from the fault constants of the test signal exhibited good agreement with the
current signal. If we know the time constant, we can calculate applied value. In the case of τ =5 ms and ratio 0.2, the error
the DC value directly at a sampling instant. From (6), the DC between the applied and estimated time constant was 3.14%.
value at time is calculated with a time constant and integral of
the fault current. The DC value for the next sampling instant TABLE I
T HE TIME CONSTANT OF INPUT SIGNAL ( MS )
can be calculated by multiplying previous DC values by an
exponential increment as shown in If 1
IDC
5 25 50 100 150 200
−t Z(t) 0.2 5.157 25.173 50.195 100.225 150.235 200.223
I0 e τ = T (11)
τ.(e τ − 1) 0.4 5.143 25.148 50.155 100.162 150.138 200.144
0.6 5.138 25.139 50.141 100.140 150.133 200.118
A. Fundamental Component Calculation 0.8 5.136 25.135 50.134 100.130 150.120 200.105
1.0 5.135 25.132 50.130 100.123 150.112 200.097
The proposed algorithm can estimate a DC value after one 1.The ratio of the magnitude of the fundamental component and the
cycle, by subtracting the calculated DC value from each of the decaying DC component.
sampled data in buffers which contain one cycle of sample
data, and applying these results to the DFT, we can extract
the fundamental component without any DC components. C. Estimated Time Constants For OFF Nominal Frequency
Overall, the calculation procedure to extract the fundamental Table II shows the estimated time constants using the pro-
component of sampled data is shown in Fig. 1. posed algorithm for the applied time constant and frequency
Department of Electrical Engineering, Univ. College of Engg., Osmania University, Hyderabad, A.P, INDIA.
 16th NATIONAL POWER SYSTEMS CONFERENCE, 15th-17th DECEMBER, 2010 655
3

Fundamental signal without harmonics

changes to investigate the off nominal frequency system op- 3
eration in the power system.
2
TABLE II
T HE FREQUENCY OF INPUT SIGNAL *. (H Z )
1

Magnitude
τ (ms) 59.8 59.9 60.0 60.1 60.2
0
25 24.844 25.024 25.139 25.171 25.223
50 29.454 49.908 50.141 50.138 49.302
−1

I c
∗. IDf C =0.6 pu
−2

The Fig. 2 shows the applied signal and estimated DC value

−3
in the time-domain simulation. It took one cycle to estimate 0 0.02 0.04 0.06 0.08 0.1 0.12
Time(sec)
the DC value from the first appearance of the DC decaying
component. Fig. 4. Calculated fundamental value(τ =25ms,ratio=0.4)

Fault signal without harmonics

5
Faultsignal The removal of the DC offset is performed before applying
4
the current signal to the digital filter used for phasor esti-
3 mation.After removing the DC component from fault current
2
signal, the fundamental waveform is obtained as shown in fig 4
The removal of decaying dc offset from current signals,
Magnitude

1 sinusoidal signal has a zero average over a full cycle or mul-

0 tiples of the full cycle of its fundamental frequency, whereas
an exponential signal has a nonzero average over that same
−1
interval. The fig ?? shows the applied signal and estimated dc
−2 value in the time domain simulation with harmonics.It took
one cycle to estimate the DC value with harmonics from the
−3
0 0.02 0.04 0.06 0.08 0.1 0.12 first appearance of the DC decaying component.
Time(sec)

Fig. 2. Applied signal and calculated DC value(τ =25ms,ratio=0.4) Time constant of 25ms
2.5
Conventional DFT
Proposed algorithm
During normal operating conditions, the voltage and current
2
signals are close to pure sine waveforms of nominal fre-
quency. However, when faults and disturbances occur in power
Magnitude of current

1.5
systems, many transient components are generated and the
nominal frequency current and/or voltage signals are distorted.
These components include decaying DC offset, harmonics and 1

off-nominal frequency components. The proposed dc offset

removal method is applied to the fault current waveform and 0.5

the corrected signal obtained is shown in Fig 3

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
Estimated DC without harmonics Time in sec
4
Estimated DC Time constant of 150ms
3
3.5 Actual DC

3 2.5

2.5
2
Magnitude

2
Magnitude

1.5 1.5

1
1
Proposed algorithm
0.5 Conventional DFT
0.5
0

−0.5 0
0 0.02 0.04 0.06 0.08 0.1 0.12 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
Time(sec) Time(sec)

Fig. 3. Calculated DC value(τ =25ms,ratio=0.4) Fig. 5. Time domain responses of the proposed algorithm and conventional
DFT for different time constants.

Department of Electrical Engineering, Univ. College of Engg., Osmania University, Hyderabad, A.P, INDIA.
 16th NATIONAL POWER SYSTEMS CONFERENCE, 15th-17th DECEMBER, 2010 656
4

Fig 5 shows the time domain responses of the fundamental Fig. 7 shows the time-domain response of the current signal
value of the applied signal using the proposed algorithm, using the proposed algorithm at the relay when a three phase
conventional DFT, In this case, the time constants of 25 fault occurred at the middle of a transmission line for a fault
ms and 150 ms with a 0.6 magnitude ratio were applied. resistance of 0.001 and 10 .
In Fig 5 the conventional DFT had an oscillation in the
fundamental component and required more times to obtain Scenario’A’
0.9
a stable output. But the proposed algorithm extracted the
0.8 Proposed algorithm
fundamental component without any undesired oscillations in Conventional DFT
0.7
one cycle and delay to obtain a stable output. From the results, 0.6
the magnitude and time constant of a decaying DC component

Magnitude
0.5
can be accurately calculated from the input signal by using
0.4
simple mathematical expressions. Also, the performance of 0.3
the proposed algorithm can meet the requirement of extensive 0.2
relay studies for various systems and fault conditions. 0.1

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
V. DYNAMIC S IMULATION T EST Time(sec)

1) In order to demonstrate the effectiveness of the proposed
algorithm used for the distance relay study, an extensive
simulation on PSCAD/EMTDC was performed.
2) As mentioned before, various conditions, such as fault
locations and fault resistance, were considered in the test
studies.
3) A set of simulation tests was verified using the con-
figuration of the power system shown in Figure. 6 The
simulated system was a 230kV, 100km transmission line
with sources at both terminals.
4) The model of the distance relay using the proposed
algorithm was embedded into the model of the power
system for the relay at bus 1. Sample numbers were set
to 64 per period and the impedance characteristic was
an mho type. The Zone 1 setting was to about 85%.
5) To examine the robustness of the proposed algorithm,
two scenarios were defined and are shown in Table III.
All scenarios assumed that the fault was initiated at 0.2
s and
IEEE cleared at ON
TRANSACTIONS 0.3POWER
s. DELIVERY, VOL. 24, NO. 1, JANUARY 2009
Fig. 7. Time responses of the current signal of the algorithm for A-G with
50km location
Fig. 7 illustrates the time-domain response of the current
signal for various fault locations of 50 and 80 km. Fig 8.
depicts the comparative results of the current estimation using
fault types, such as single phase fault and three phase faults.
In order to obtain a clearer picture of the performance of the
two techniques, the apparent impedance as seen by the relay
located at bus 1 for scenario A is shown in Fig. 7 when a
A-B-C-to-ground fault occurred at a fault distance of 80%, In
contrast, The fault was set between bus 1 and the generator
terminal. As shown in these figures, a distance relay adopting
the proposed algorithm had no problems with the operations.
Scenario ’B’
0.8
Proposed algorithm
0.7 Conventional DFT

0.6
Magnitude of current

0.5

0.4

0.3

Fig. 4. One-line diagram of a test system. 0.2

Fig. 6. Single line diagram of a test system
0.1

a 230-kV, 100-km transmission line with sources at both termi- 0

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
nals. The used:-
Specifications model of the distance relay using the proposed algo- Time in sec
rithm was embedded into the model of the power system for the
Transmission line:-
relay at bus 1. Sample numbers were set to 64 per period and the
Lenth:100Km
impedance characteristic was an type. The Zone 1 setting Fig. 8. Time responses of the current signal of the algorithm for A-B-C-G
was to aboutat85%.
Voltage:230KV 60Hz with 80km location
To examine the robustness of the proposed algorithm, four
Termination style:Direct connection
scenarios were defined and are shown in Table III. All scenarios
assumed that the fault was initiated at 0.2 s and cleared at 0.3 s.
SourceFig.Generator:-
5(a) and (b) shows the time-domain response of the current
signal using the proposed algorithm at the relay when a three-
Voltage:230KV
phase fault with ramp
occurred at theup timeof of
middle 0.05 sec line for a
a transmission
Base MVA:100 MVA
fault resistance of 0.001 and 10 . TABLE III
D ESCRIPTION OF THE SCENARIOS
R-L:9.186Fig. 5(a) and
ohms 138(c)milli
illustrates the time-domain response of the
Henry
current signal for various fault
6 locations of 50 and 80 km. Fig.
Breaker5(c)
Resistance:-0pen 10 ohmsresults
and (d) depicts the comparative and close 0.1 ohms
of the current es- Scenario Fault type Location(km) Resistance
Fault Resistance:
timation usingON two 0.01 ohmssuch
fault types, and OFF 106 faults
as three-phase ohms. and A A-G 50 0.001
single-line-to-ground faults. B A-B-C-G 80 10
In order to obtain a clearer picture of the performance of the
Department of Electrical
two techniques, Engineering,
the apparent impedance Univ.
as seenCollege of Engg., Osmania University, Hyderabad, A.P, INDIA.
by the relay
located at bus 1 for scenario A is shown in Fig. 6, when a
phase-A-to-ground fault occurred at a fault distance of 80%. In
contrast, Fig. 7 depicts the relay response for an external fault,
 16th NATIONAL POWER SYSTEMS CONFERENCE, 15th-17th DECEMBER, 2010 657
5

VI. C ONCLUSION
A new algorithm based on decaying DC component estima-
tion and elimination for digital relaying has been presented.
The proposed algorithm can obtain the DC magnitude and the
time constant calculation which was estimated by integrating
fault rents during one cycle. Thus, it can be used for extraction
of the fundamental frequency component.
A comprehensive set of simulation results through static and
dynamic tests has shown that the proposed algorithm has faster
convergence and better accuracy than the conventional DFT
and Furthermore, a distance relay employing the proposed
algorithm provides a correct relay response. The proposed
algorithm can be easily implemented and applied to fast digital
distance relaying for transmission lines.

R EFERENCES
[1] A.G Phadke, and J.S Thorp, ”Computer Relaying for Power Systems.
New York: Wiley, 2002.
[2] G. Benmouyal, ”Removal of dc-offset in current waveforms using digital
mimic filtering,” IEEE Trans Power Syst, vol. 10,no.2, pp. 621-630 Apr.
1995.
[3] J.Z Yang and C.W. Liu, ”Complete elimination of dc offset in current
signals for relaying applications,”in ProcIEEE Power Engg.soc, Winter
Meeting, pp. 1933-1938, Jan 2000.
[4] C.S. Chen,C.W. Liu and J.A Jiang, ”Application of combined adaptive
fourier filtering technique and fault detector to fast distance protection,”
IEEE Trans Power Del., vol. 21,no.2 pp. 619-626, Apr 2006.
[5] J.C Gu and S.L Yu, ”Removal of dc offset in current and voltage
signals using a novel fourier filter alorithm ” IEEE Trans power Deliver,
vol. 15,no.1, pp. 73-79, Jan 2000.
[6] T.S Sindu and X. Zhang, F. Albasri and M.S Sachdev,”Discrete fourier
transform based technique for removal of decaying dc offset from phasor
estimates” Proc.Inst.Electr.Eng., vol. 150, pp. 745-752, Nov 2003.
[7] T.S Sindu and X. Zhang and V.Balamourougan, ”A new half-cycle phasor
estimation algorithm,” Electr Power Syst, vol. 76, pp. 194-199, Jan 2006.

Department of Electrical Engineering, Univ. College of Engg., Osmania University, Hyderabad, A.P, INDIA.