Emtp Lab
Emtp Lab
8.1 AIM
(i) To study and understand the electromagnetic transient phenomena in power
systems caused due to switching and faults by using Electromagnetic Transients
Program (EMTP).
(ii) To become proficient in the usage of EMTP to address problems in the areas
of overvoltage protection and mitigation and insulation coordination of EHV
systems.
8.2 OBJECTIVES
(i) To study thetransients due to energization of a single-phase and three-phase load
"!#$ %&(') * +-,./0&1
from a non-
(ii) To study the transients due to energization of a single-phase and three-phase load
from a non-ideal source and line represented by distributed parameters.
(iii) To study the transient over voltages due to faults for a SLG fault at far end of a
line.
(iv) To study the Transient Recovery Voltage (TRV) associated with a breaker for a
three-phase fault.
Intentional and inadvertent switching operations in EHV systems initiate over voltages,
which might attain dangerous values resulting in destruction of apparatus. Accurate
computation of these over voltages is essential for proper sizing, coordination of
insulation of various equipment’s and specification of protective devices. Meaningful
design of EHV systems is dependent on modelling philosophy built into a computer
program. The models of equipment’s must be detailed enough to reproduce actual
conditions successfully – an important aspect where a general purpose digital computer
program scores over transient network analysers.
8-1
the explanations simple, however, single phase network elements will be used, rather
than the more complex multiphase network elements.
2 3
o o
R L
i12 i13
5 Loss less distributed line 1 C 4
o o
Since node voltages are used as state variables, it is necessary to express the branch
currents, i12 etc., as functions of node voltages. For the resistance;
i12 (t) = 1/R (v1 (t) – v2 (t)) (8.2)
8-2
i13(t) = ¦a§¨N©ª^«&¬ 1(t)-v3(t-a®&¯[¯U°±Wm®n²N³´^¯Dµ 13(t- U®W¯°¶ 13(t- a®&¯
i13(t) = a®
²c³´·±Wµ 1(t) – v3(t)) + I13(t-®W¯ (8.3a)
where
I13(t-U®&¯¸ ±[U®n²&³´w¯ µ 13(t-a®N¯[°U¶ 13(t- U®W¯ ¹]®º:»¼#½4¾¿®º¶n¾¿®&ÀÁWÂî&»"Á[Ä
with i13 computed from the values of the preceding time step;
The derivation for the branch equation of the capacitance is analogous, and leads to
i14(t) = 2C/U®[±&µ 1(t) – v4(t)) + I14(t-U®W¯ (8.4a)
For the transmission line between nodes 1 and 5, losses are first ignored, and then re-
introduced later on. Then the wave equations
- ÆÇÉÈvÊÌË
ÆÍ
ÎÏ ÎÐ
- ÑÒÓ ÔÖÕ
Ñ×
ÑØ ÑÙ
(L’, C’ = inductance and capacitance per unit length, x=distance from sending end)
have the well-known solution due to d’ Alembert;
where F(x-ct) and f(x+ct) are functions of the composite variables x-ct and x+c, z is the
surge impedance, and c is the velocity of propagation;
z= Ú
ÛÜÝ&ÞÖÜßàáÌâ%Ý&Ú
ÛÌÜnÞÖÜNßãä]åNæÃà4äçè¿åNéçå
If the current in equation (8.5a) is multiplied by z and added to the voltage, then
v+zi = 2zF(x-ct) (8.5b)
Note that the composite expression v+zi does not change if x-ct does not change.
Imagine a fictitious observer travelling on the line with wave velocity c. The distance
travelled is x = x0 + ct (x0 = location of starting point), or x-ct = constant. If x-ct is
constant, then the value v+zi, measured by the observer, must also remain constant.
With travel time
ê áìëíç:îïëî"ç:ð]åæ
c
an observer leaving node 5 at time t – êñ íëNë
òî4éè¿ó#ô$îõå&æ:îÖö#é]ëó#îö 5(t- ê ÷^øúùû 51(t- ü ÷$ýþ"ÿ
ÿìþ û
:þ
Uþaÿ Wþ
nþ $þ
&û!" ü ÷ ý$#Ãû%&'(þ ) *+
:þ +
(t) –
1
zi15 (t) (negative sign because i15 has opposite direction of i51 ). But the value as
8-3
measured by the observer must remain constant, so these values must be equal, giving,
after rewriting,
i15(t) = (1/z) v1(t) + I15(t- , -/. /0 1243657-
where the term I15 is again known from previously computed values
I15(t - 89-;: - (1/z) v5(t- 89- – i51(t- 89-*. (8.6b)
<>=$?@BACDEGFD9H IKJLNM9OOQP(R%STVUW)X7S
UGYZW\[YZ^] _N`badcfeGg'hViGakjlGmn9oqp&iGrcts/ue4v6wVxypzo>{n7rQ|j+c}jjr
that the “~
/>$&*
G
^^%+/^6V7"6q!"Q)~$&&76
&¡¢¤£¥¦§¢¨©¥ªV¢« ¡¬¨}¢®
¬G¯^°§6
¢7±²³B¨/ ¢°$¨´¢7µ ¬6ªV¢6¶¸·K¹º¢¥ªK¹º6 ¢ª*°¬¦&¥9 ¹&¬Gº¹¨§¨}¢7±B¹%µ;»«¹¨¼º¬G ¥º
¹!º6
¢7½6¢ª«¯Q§$¦! q¹°¦¢"¬µ¿¾K *¶ÀÁy¶;Â/öÅÄ6Ƥ¹%¨+¥ºÇ¢È¥ t¨)¬¦§6 ¹%¬6ºÉµ ¬6ª¿ ¡¢Ê¦%¬¨b¨+¦¢¨b¨¦¹!º¢G¶·¬¨b¨¢¸¨
can
often be represented with sufficient accuracy by inserting lumped resistance in a few
places along the line. The program described here lumps R/4 at both ends and R/2 in the
middle. For a more sophisticated treatment of losses, especially with frequency
dependent parameters, see ref. 1.
If Eq. (8.2), (8.3a), (8.4a) and (8.6a) are inserted in to Eq. (8.1), then the node
equation becomes
ËÍÌfÎ}ÏÐtÑ>ÒÓÌqÔKÕ%ÏqÖ6×Ñ;ÒÓÌqÖØ©ÏyÔ ÙÚ>ÛÜ9Ý}Þ&ß7ÚÍàtá Üqâ¼Ù%Þã6äKÚåá Üqãæ©ÞâKÙÚ>á
1 (t) – (1/R) v2 (t) – 3 (t) - 4 (t) =
âÙyÚ âKÙÚ Ú;è
i1 (t) – I13 (t- – I14 (t- – I15 (t- ç (8.7)
which is simply a linear, algebraic equation in unknown voltages, with the right-hand
side known from the values of preceding steps.
For any type of network with n-nodes, a system of n such equations can be formed*;
[G] [v(t)] = [i(t)] – [I] (8.8a)
with [G] = nodal conductance matrix
[v(t)] = vector of n node voltages
[i(t)] = vector of current sources, and
[I] = vector of “past history” terms.
Normally, some nodes have known voltage sources connected. In this case Eq. (8.8a)
can be partitioned into nodes “A” , with unknown voltages, and nodes “B”, with known
voltages. The unknown voltages are found by solving
for [vA (t)]. The actual computation in the transients program proceeds as follows: the
matrixes [GAA] and [GAB] are built and [GAA] is triangularised with ordered elimination,
and exploitation of sparsity. In each time step, the vector on the right- hand side of Eq.
(8.86) is “assembled” from past history terms, and current and voltage sources. Then
the system of linear equations is solved for [vA(t)], using the information contained in
the triangularised conductance matrix. Before proceeding to the next time step, the
terms I of Eq. (8.3b), (8.4b) and (8.6b) are updated for use in future time steps.
* Brackets are used to indicate matrix and vector quantities.
8-4
8.5. EXCERCISES
8.5.1 Prepare the data for the network given in the Annexure 8.1 and run EMTP.
Obtain the plots of source voltage, load bus voltage and load current following
the energisation of a single-phase load. Comment on the results. Double the
source inductance and obtain the plots of the variables mentioned earlier.
Comment on the effect of doubling the source inductance.
8.5.2. Prepare the data for the network given in the Annexure 8.2 and run EMTP.
Obtain the plots of voltages of phases a, b, c at the load bus and switch A current
of phase a following energisation of the three-phase load by closing the switches
simultaneously. Repeat the same assuming following switch closing sequence:
a-c-b with a delay between them equal to 3 millisecond (pole discrepancy).
Comment on the results obtained for simultaneous and asynchronous closing.
8.5.3. Prepare the data for the network given in the Annexure 8.3 and run EMTP.
Obtain the plots of voltages at source, Bus 1 and Bus 12 following the
energisation of the single phase open ended line represented by distributed
parameters. Obtain the plot of voltage at Bus 12 by expanding the time scale by
a factor of ten, i.e, plot the voltage for the first 2.5 millisecond. Comment on the
results obtained by comparing with those obtained from 8.5.1. Double the line
length and repeat the exercise. What are the effects of doubling the length?
8.5.4 Prepare the data for energisation of a three-phase load fed by a three-phase
distributed parameter line as given in the Annexure 8.4 and run EMTP. Obtain
the plots of voltages at source, Bus 1 and phase a voltage at Bus 12 following
the energisation by simultaneous closing of all the three phases. Comment on
the results. Repeat the exercise by assuming pole discrepancy as given in 8.5.2
and comment on the results.
8.5.5 Prepare the data for the network given in the Annexure 8.5 and run EMTP.
Obtain the plots of voltages at source, Bus 1 and Bus 2 following a single line-
to-ground fault at the far end, Bus 2. Comment on the maximum phase voltage
attained.
8.5.6. Prepare the data for the network given in Annexure 8.6 and run EMTP. Obtain
the transient recovery voltage (TRV) in each phase for a three-phase fault at
Bus 1. The TRVs are the voltages across the switches between Bus1 and
BKR1. Comment on the results.
Repeat the above with
(a) Line 1-12 represented as open at the receiving end.
(b) Line 1-12 not represented.
Comment on the results
8.6 REFERENCES
1. R. Ramanujam, Computation and Analysis of Transients in Power Systems using
Electromagnetic Transients Program (EMTP), Power System Engineering Division,
School of Electrical and Electronics Engineering, College of Engineering, Guindy,
Anna University, Chennai-600 025, July 1998.
2. H.W Dommel and I.I. Dommel, Transients Program User’s Manual, Aug.1978
8-5
ANNEXURE – 8.1
Case Study – I
Energization of a single phase 0.95 pf load from a non ideal source and a more realistic
line representation (lumped R,L,C )
19.72 mH
0.81 µF 0.84 µF
56.34
cos ωt
~
22.61 Ω
(ii) Data
SRCBUS13L
8-6
ANNEXURE – 8.2
Case Study - II
56.34
cos ωt
~
56.34
cos ~
(ωt-120)
56.34
cos ~
(ωt+120)
8-7
(ii) Data
SRCA SRCB SRCC B13LA B13LB B13LC SRCA SRCB SRCC B13LA B13LB B13LC
8-8
ANNEXURE – 8.3
Energization of a 15 mile long line from an ideal voltage source, line representation by
distributed parameters
-
~ 187.79 cos ωt
(ii) Data
8-9
ANNEXURE – 8.4
Case Study – IV
Energisation of a 3 - phase 15 mile distributed parameter line connected to a transformer and RL
load, all three phases close simultaneously – Lumped and Distributed Parameter Modelling
24.14 Km Transposed
~ ~ ~ Distributed Line
R0 = 0.3167 Ω / Km
L0 = 3.222 mH / Km
C0 = 0.00787 µf / Km
R1 = 0.243 Ω / Km
L1 = 0.923 mH / Km
C1 = 0.0126 µf / Km
8-10
(ii) Data
BUS12A BUS1A
8-11
ANNEXURE – 8.5
Case Study – V
Transient Fault over voltages for a SLG fault at bus 2, phase a – Symmetrical PI
model
(coupled) And distributed line parameter
Fig.5 Calculation of transient fault overvoltages for a SLG fault at bus 2, phase a
(ii) Data
8-12
BUS1A BKR1A -.1E-2 .9999E+4 1
BUS1B BKR1B -.1E-2 .9999E+4 1
BUS1C BKR1C -.1E-2 .9999E+4 1
BUS2A .1E-2 .9999E+4 1
8-13
ANNEXURE – 8.6
Case Study – VI
(i) Circuit diagram
BKR1
Closed in steady state
(Represented by small
Closed in resistances 1.0E-6
Steady state
topen = 1ms
(ii) Data
TRV FOR BUS1 BRKR FOR A 3-PH FAULT AT BUS1 :EXPER NO:6 1
.1E-4 .25E-1 -1 1
1 THEVA BUS7A .0833 34.56
2 THEVB BUS7B .0233-5.427 .0833 34.56
3 THEVC BUS7C .0233-5.427 .0233-5.427 .0833 34.56
1 GEN3A BUS1A .0 131.0
2 GEN3B BUS1B .0 -23.4 .0 131.0
3 GEN3C BUS1C .0 -23.4 .0 -23.4 .0 131.0
-1 BUS7A BUS1A .3167 3.222 .0078 144.4
-2 BUS7B BUS1B .0243 .9236 .0126 144.4
-3 BUS7C BUS1C
-1 BKR1A BUS2A .3167 3.222 .0078 193.1
-2 BKR1B BUS2B .0243 .9236 .0126 193.1
-3 BKR1C BUS2C
-1 BUS1ABUS12A .3167 3.222 .0078 24.14
-2 BUS1BBUS12B .0243 .9236 .0126 24.14
-3 BUS1CBUS12C
BUS12ABUS13A .0E+0 70.16 .0
BUS12BBUS13B .0E+0 70.16 .0
BUS12CBUS13C .0E+0 70.16 .0
BUS13A .25E+3 219.1 .0
BUS13B .25E+3 219.1 .0
BUS13C .25E+3 219.1 .0
BKR1A .1E-5 .0 .0
8-14
BKR1B .1E-5 .0 .0
BKR1C .1E-5 .0 .0
8-15