Power Systems Analysis
Power Systems Analysis
Power Systems Analysis
1.0 INTRODUCTION
Power Systems Engineering is the oldest and most traditional of the various areas within
electrical engineering, yet it’s undergoing a more dramatic revolution in both technology and
industry structure. Its analysis is fundamental in the planning, design and operation of power
systems engineering. It is a complex interconnected network that is essentially divided into four
major parts, namely, generation, transmission and sub-transmission, distribution and loads.
1.1 POWER SYSTEM NETWORK REPRESENTATION
A balanced three-phase system can be conveniently solved on single-phase basis. It can be
represented by a single-phase system having one line and a neutral. The first step in fault
calculation is normally drawing a single line diagram. In this diagram the components of the
system are usually drawn in the form of their symbols. The neutral earthing is indicated. Figure 1
shows single line diagram of a simple system.
The next step is to draw impedance or reactance diagram or positive sequence network. In
impedance diagram each component is represented by its equivalent circuit. Figure 2 represents
the impedance diagram of the network in Figure 1.
1
diagram reduces to simplified reactance diagram as shown in Figure 3. Reactance diagram is also
called positive sequence network.
( )
( new) 2
( new) (old ) S B V (Bold )
Z pu = Z pu ( old ) (new )
SB VB
(4)
( new)
(old ) S B
Z(pu
new)
= Z pu
Or S(Bold )
(if voltage bases are the same) (5)
ADVANTAGES OF PER-UNIT SYSTEM
1. The per-unit system gives us a clear idea of relative magnitudes of various quantities such
as voltage, current, power and impedance.
2. The per-unit values of impedance, voltage and current of a transformer are the same
regardless of whether they are referred to the primary or the secondary side.
3. The per-unit systems are ideal for the computerized analysis and simulation of complex
power system problems.
4. Machine reactances given in per-unit, give a basis for comparison.
2
5. The per-unit impedance of equipment of the same general type based on their own ratings
fall in a narrow range regardless of the rating of the equipment.
6. The circuit laws are valid in per-unit systems, and power and voltage equations are
simplified since the factors of √ 3 and 3 are eliminated in the per-unit system.
TUTORIALS
1. What is the base impedance if base voltage = 1, 100 volts and base kVA = 106?
Solution:
Base kVA = 106
Base kV = 1,100V or 1.1kV
Base current,
4. A 30 MVA, 11kV generator has a reactance of 0.2p.u referred to its ratings as bases.
Determine the per unit reactance when referred to base kVA of 50,000kVA and base kV
of 33kV.
Solution:
Old base kVA = 30,000 kVA
Old base kV = 11kV
Old per unit impedance,
New base kVA = 50,000 kVA
New base kV = 33kV
3
New per unit impedance,
Generator 1:
Generator 2:
Generator 3:
6. Draw the p.u impedance diagram for the power system shown in Figure Q6. Neglect
resistance and use a base of 100 MVA, 220kV in 50Ω line. The rating of the generator,
motor and transformers are:
Generator: 40 MVA, 25kV, x″ = 20%
Motor: 50 MVA, 11kV, x″ = 30%
Y-Y Transformer: 40 MVA, 33Y-220Y kV, x = 15%
Y-Δ Transformer: 30 MVA, 11Δ-220Y kV, x = 15%
4
NOTE: To convert the following power system components values to per unit values on a
common base:
(a) Generators and transformers – percent impedance on rating is given by
5
Faults are classified as symmetrical faults and unsymmetrical faults. Symmetrical faults are
three phase faults that can be solved on per phase basis while unsymmetrical faults are solved by
using the method of symmetrical components.
6
3.1.1 The Symmetrical Components of an Unbalanced Three Phase System
A balanced set of components are: -
(1) Positive-sequence components consisting of three phasors equal in magnitude, displaced
from each other by 1200 in phase and having the same phase sequence as the original
phasors.
(2) Negative-sequence components consisting of three phasors equal in magnitude, displaced
from each other by 1200 in phase, and having the phase sequence opposite to that of the
original phasors.
(3) Zero-sequence components consisting of three phasors equal in magnitude and with zero
phase displacement from each other.
Please, note that the subscript 1, 2 and 0 denote positive, negative and zero sequence entities
respectively.
Since each of the original unbalanced phasors is the sum of its components, the original phasors
expressed in terms of their components are
V a = V a 1 + V a2 + V a 0 (85)
V b = V b1 + V b2 + V b0 (86)
V c = V c1 + V c2 + V c0 (87)
From equations (85) to (87) we note that the number of all known quantities can be reduced by
expressing each component of Vb and Vc as the product of some functions of the operator ‘a’ and
a component of Va. With reference to the set of symmetrical components.
V b 1 = a 2 V a1 ; V c1 = aV a1
V b 2 = aV a 2 ; V c 2 = a2 V a2
V b 0 = V a0 ; V c0 = V a 0 (88)
Upon substitution of equation (88) into equations (85) to (87)
7
V a = V a 1 + V a2 + V a 0 (89)
2
V b = a V a 1 + aV a 2 + V a 0 (90)
2
V c = aV a 1 + a V a 2 + V a 0 (91)
Representing in matrix form, we have
(91a)
Adding equations (89), (90) and (91) gives
V a + V b + V c = ( 1 + a + a 2 ) V a 1 + ( 1 + a + a 2 ) V a 2 + 3V a0 (92)
2
and since 1 + a + a = 0
1
∴ V a 0 = (V a + V b + V c )
3 (93)
Equation (93) enables us to find the zero-sequence component of three unsymmetrical phasors.
We discover that no zero-sequence component exist if the sum of the unbalanced phasors is zero.
Since the sum of the line-to-line voltage phasors in a three-phase system is always zero, zero
sequence components are never present in the line voltages, regardless of the amount of
unbalance. The sum of the three line-to-neutral voltage phasors is not necessarily zero, and
voltages to neutral may contain zero sequence components. The positive sequence component of
three unsymmetrical phasors may be found by manipulation of equations (90) and (91), i.e.,
multiplying equation (90) by ‘a’ and equation (91) by ‘a 2’ gives after substitution of 1 for ‘a 3’ and
a for ‘a4’
aV b = V a 1 + a2 V a 2 + aV a 0 (94)
2 2
a V c = V a 1 + aV a 2 + a V a 0 (95)
Adding equations (94), (95) to equation (89) gives
V a + aV b + a2 V c = 3V a1 + ( 1 + a + a2 ) V a2 + ( 1 + a + a2 ) V a0 (96)
Since 1 + a + a2 = 0
1
3( a
∴ V a1 = V + aV b + a2 V c )
(97)
The negative sequence component may be found in a similar manner. First multiply equation
(90) by ‘a2’ and equation (91) by ‘a’. Adding the resulting equation to equation (89) and which
V + a2 V + aV = 3V + ( 1 + a + a2 ) V + ( 1 + a + a2 ) V
gives a b c a2 a1 a0 (98)
2
and since 1 + a + a = 0,
1
∴ V a 2 = ( V a + a2 V b + aV c )
3 (99)
Equations (93), (97) and (99) enables us to find the complete set of symmetrical components of a
given set of three phasors, since we can find the component of V b and Vc from the component of
Va by the equation given in equation (88). They may be solved either analytically or graphically.
Similarly, for current,
8
Ia = Ia1 + Ia2 + Ia0
I b = a 2 I a1 + aI a 2 + I a 0
I c = aI a 1 + a2 I a 2 + I a 0
1
I a1 =
3
( I a + aI b + a 2 I c )
(100)
1
I a 2 = ( I a + a2 I b + aI c )
3 (101)
1
Ia0 = (Ia + Ib + Ic )
3 (102)
In a three-phase system the sum of the line current is equal to the current, I n in the return path
through the neutral. Thus, Ia + Ib + Ic = In (103)
Comparing equations (102) and (103) gives, In = 3Ia0 (104)
In the absence of a path through the neutral of a three-phase system, I n is zero, and the line
currents contain no zero sequence components. A delta-connected load provides no path to
neutral, and the line currents flowing to a delta-connected load can contain no zero sequence
components.
Determination of phase voltage from unsymmetrical line voltage when unsymmetrical three
phase line voltages are applied to the terminals of a balanced star-connected load, the method of
symmetrical component provide one means of determining the voltage to neutral of each phase if
no zero sequence currents flow in the impedances. The voltages to neutral caused by the positive
sequence current would be of a positive sequence only because each component of each voltage
to neutral will be displaced equally in phase from the current component causing it.
Similarly, negative sequence current will cause only negative sequence voltage drop in a
balanced system. If no zero sequence currents are present, there can be no zero-sequence voltage
drops in a balanced star-connected impedances.
1
V an1 = V ∠−30°
√ 3 ab1 (105)
1
V an2 = V ab2 ∠ 30°
√3 (106)
We find Van as the sum of its components. So,
9
V an = V an1 + V an2 (107)
The other voltages to neutral are found by obtaining their components from V an1 and Van2 from
equation (88). If the voltage to neutral are in p.u. referred to the base voltage to neutral and the
1
line voltages are in p.u. referred to the base from line to line, the √ 3 factors must be omitted in
equations (105) and (106). If both voltages are referred to the same base or are in actual volts the
equations are correct as given.
Example 8: Three identical resistances are star-connected and are rated 2300V, 500kVA as a
three-phase unit. The neutral point is not available. The resistance unit is connected to
unsymmetrical three phase system whose line voltages are measured and found to be |V ab| = 1840
volts, |Vbc| = 2760 volts and |Vca| = 2300 volts. Find the current in each line by the method of
symmetrical component.
Solution: Select as base for the circuit 2300V, 500kVA so that each resistance in the three-phase
load has a resistance of 1.0 p.u. Assuming an angle of 180˚ for Vca.
500000
= 125.5 A
The base current is √ 3 × 2300
1840 2760 2300
= 0.8 p.u. = 1 .2 p .u . = 1 . 0 p . u.
and |Vab| = 2300 ; |Vbc| = 2300 ; |Vca| = 2300
Since the sum of the line voltage of a three-phase circuit is zero, the triangle formed by the
phasors of the line voltages can be solved to find the angle of each phasor. The triangle of each
phasor is as shown – the line voltage phasor.
10
V ab = 0.8∠82.8° p.u
V bc = 1.2∠ 41. 4° p.u
V ca = 1.0∠180° p.u
1
V ab1 = ( 0.8∠82 .8° + 1.2∠(120° + 41 .4° ) + 1.0∠(240° + 180° ))
3
= − 0.179 + j 0.681 = 0.704∠104.7° p.u
1
V ab2 = ( 0.8∠82 .8° + 1.2∠(240° + 41.4° ) + 1.0∠(120° + 180° ))
3
= 0.279 − j0.416 = 0.5∠−56 .2° p.u
V ab = V ab1 + V ab2
= (−0.179 + j0.681 ) + ( 0.279 − j 0.416 )
= 0.1 + j 0.265
The absence of the neutral connection means that zero sequence currents are not present.
Therefore, the base voltage at the load contains positive and negative sequence components only.
1
The phase voltages are found from equations (105) and (106) with factor √3 omitted since the
line voltages are expressed in terms of the base voltage from line to line and the phase voltage is
desired in p.u. of the base voltage to neutral. Thus,
Moving the diagram of a delta connected load and phasor diagrams of the positive and negative
sequence components of the line and phase unit, inspection of the diagram shows the relationship
between the components of Ia and Ib.
1
I ab 1 = I ∠30°
√3 a1 (108)
1
I ab2 = I a2 ∠−30°
√3 (109)
Example 9: The line current entering a balanced system (delta-connected) are I a = 5A; Ib = 4A; Ic
= 3A. Find the current in phase ab of the load by the method of the symmetrical component. If
the current in phase c is taken as Ia = - 3 – j4, Ib = 0 – j4, Ic = 3 + j0.
Solution: The symmetrical components of the line current are
1
I a1 = (−3− j 4+4 ∠(−90 °+120 °) + 3∠ 240° )
3
= −0 .345− j 1.532 = 1. 57 ∠−102. 7° A
1
I a 2 = (−3− j 4+4 ∠(−90 °+240 ° ) + 3∠120° )
3
= −2.655+ j 0. 199 = 2. 66 ∠175 . 7° A
From equations (108) and (109)
12
1 .57
I ab 1 = ∠−72 .7 ° = 0 . 906 ∠−72 .7 ° = 0 . 269− j0 . 865 A
√3
2 .66
I ab 2 = ∠145 . 7 ° = 1. 536 ∠145. 7 ° = −1 . 269+ j 0 . 866 A
√3
If the load contains no e.m.f to induce zero sequence current, I ab is composed of positive and
negative sequence current only.
I ab = I ab 1 + I ab2
= ( 0 . 269− j0 . 865 ) + (−1. 269+ j0 . 866 )
= −1+ j 0 .001
= 1 ∠180° A
Example 10: In a 3-phase, 4-wire system, the currents in the R, Y, and B lines under abnormal
conditions of loading were as follows: Calculate the
positive, negative and zero-phase sequence currents in the R-line and the return current in the
neutral conductor.
Solution: The positive-sequence components of current in the R-line is
Now,
13
The neutral current is
Example 11: A 3-phase, 4-wire system, supplies loads which are unequally distributed on the
three phases. An analysis of the currents flowing in the direction of the loads in the R, Y and B
lines shows that in the R-line, the positive phase sequence current is and the negative
phase sequence current is . The total observed current flowing back to the supply in
the neutral conductor is . Calculate the currents in phase and magnitude in the three
lines.
Solution: For R-phase,
For Y-phase,
For B-phase,
14
An unloaded grounded generator is as shown. The e.m.f of the phases are E a, Eb and Ec
respectively. When the faults occur at the terminals of a generator, currents, I a, Ib and Ic flow in
the lines. If the fault involves ground, the current flowing into the neutral of the generator is
designated In. One or two of the line current may be zero but the current can be resolved into a
symmetrical component regardless of how unbalanced they may be. Current of a given phase
sequence in a balanced system produce voltage drops of the same sequence only. In any part of
the circuit the voltage drops caused by current of certain sequence depends on the impedance of
that part of the circuit to current of the sequence. The impedance of any section of a balanced
network to current of one sequence may be different from the impedance of another sequence.
The impedance of a circuit when positive sequence currents only are flowing is called the
positive sequence impedance. Similarly, we have negative sequence impedance and zero
sequence impedance. The analysis of unsymmetrical faults on a symmetrical system consists of
finding the symmetrical components of the balanced currents which are flowing. Since the
component currents of one phase sequence cause voltage drops of like sequence only and are
independent of currents of other sequences in a balanced system, currents of any one sequence
may be considered to flow in an independent network composed of impedances to the current of
that sequence only. Generator voltages are of the positive sequence since generator is designed to
supply balanced three phase voltages. Therefore, the positive sequence network is composed of
an e.m.f in series with the positive sequence impedance of the generator. The negative- and zero-
sequence networks contain no e.m.f but include impedances of the generator to negative- and
zero-sequence currents respectively.
15
For the zero sequence,
Voltage drop =
−I ao Z go − 3 I ao Z n
16
∴ Z 0 = 3 Z n + Z go (110)
where Zgo is the zero-sequence impedance per phase of the generator. From the sequence
networks,
V a 1 = E a − I a1 Z 1 (111)
V a 2 = − I a 2 Z2 (112)
V a 0 = − I a0 Z 0 (113)
where Ea is the positive sequence no-load voltage to neutral, Z 1 and Z2 are the positive and
negative sequence impedances of the generator and Z0 is as defined in equation (110).
Figure 16. A Single line to ground fault Figure 17: Connection of the sequence
on phase ‘a’ at the terminals of an unloaded networks of an unloaded generator for
generator whose neutral is grounded single line to ground fault on phase ‘a’
at through a reactance. the terminals of the generator.
19
Figure 18: Circuit diagram for a line-to-line fault between phases b and c terminals of an
unloaded generator whose neutral is grounded through a reactance.
The conditions of the fault are expressed by the following equations, Vb = Vc ; Ia = 0 ; Ib = -Ic.
Substituting the relations given by equations (100) to (102) gives
1 I
I a1 =
3
( 0+aI b −a2 I b ) = j b
√3
1 I
I a 2 = ( 0+a2 I b −aI b ) = − j b
3 √3
1
I a 0 = ( 0+I b −I b ) = 0
3
Therefore, Ia1 = -Ia2 (116)
Then from equations (90) and (91)
V a 2 = V a 1 = E a−I a 1 Z 1
which becomes the equations (112) and (116)
−I a 2 Z 2 = I a 1 Z 2 = E a−I a 1 Z 1
Ea
I a1 =
Solving for Ia1 gives Z 1 +Z 2 (117)
Equations (116) and (117) are the special equations for line-to-line fault. They are used with
equations (111) and (113) and symmetrical component relations to determine all the voltages and
currents at the fault. Since Z0 does not enter into the equations, the zero-sequence network is not
used. The positive and negative sequence networks must be in parallel since V a1 = Va2 and this
gives Ia1 = - Ia2. Since there is no ground at the fault, there is only one ground in the circuit at the
generator neutral and no current can flow to the ground, I n = 0. In the derivation of the relations
of a line-to-line fault, Ia0 = 0. This is consistent with the fact that no ground current can flow since
the ground current, In = 3Ia0. The presence or absence of a ground neutral at the generator does
not affect the fault current. If the generator neutral is not ground Z 0 is infinite and Va0 is
indeterminate but line to line voltages may still be found since they contain no zero sequence
components.
Example 11: Find the sub-transient current and line to line voltages of the fault under sub-
transient conditions. When the line-to-line fault occurs at the terminals of the generator under the
given conditions in example 10. Assume the generator is unloaded and operating at rated terminal
voltage when the fault occurs. Neglect resistance.
Solution:
1.0
I a1 = = − j 1. 667 p .u
j 0. 25+ j0 . 35
I a 2 = −I a 1 = j 1. 667 p .u
Ia0 = 0
I a = I a 1 +I a2 +I a0 = − j1 .667 + j 1. 667+0 = 0
2
I b = a I a1 +aI a 2 +I a 0
= − j1 .667 (−0 .5− j 0 .866 ) + j 1. 667 (−0 . 5+ j 0 .866 )
= −2. 892+ j0 p .u
I c =−I b = 2. 892 p . u
20
As in example 10, base current is 836.7A.
So, Ia = 0
I b = −2.892×836 .7 = 2420∠180 ° A
I c = 2. 892×836 .7 = 2420 ∠0 ° A
The symmetrical components of the voltages from a to ground are
V a 1 = V a 2 = 1−(− j1 . 667 ) ( j 0 .25 ) = 1−0 . 416 = 0 .584 p . u
V a 0 = 0 ( Neutral of the generator or ground )
Line to ground voltages are
V a = V a 1 +V a 2 +V a 0 = 0. 584+0. 584+0 = 1. 168∠0 ° p .u
V b = a2 V a 1 +aV a 2 +V a 0
V c = V b =−0. 584 p . u
Line to line voltages are
V ab = V a −V b = 1.168+0.584 = 1.752 ∠0° p.u
V bc = V b −V c = −0.584 +0.584 = 0 p.u
V ca = V c−V a = −0.584−1.168 = 1.752 ∠180 ° p.u
Expressed in volts the line-to-line voltages, we have
13.8
V ab = 1.752× = 13.95∠0°kV
√3
V bc = 0kV
13.8
V ca = −1.752× = 13.95∠180° kV
√3
The faulted phases are b and c, the conditions at the faults are expressed as follows.
21
Vb = Vc = 0
Ia = 0
Substituting, Vb = 0 and Vc = 0 in equations (93), (97) and (99) gives
1 Va
V a 1 = ( V a +0+ 0 ) =
3 3
1 Va
V a 2 = ( V a +0+ 0 ) =
3 3
1 Va
V a 0 = ( V a + 0+0 ) =
3 3
Therefore, a1
V =V =V
a2 a0 (118)
Solving equations (112) and (113) for Ia2 and Ia0 and substituting Va1 for Va2 and Va0, we obtain
V a2 V
Ia2 = − = − a1
Z2 Z2
V V
I a 0 =− a 0 = − a 1
Z0 Z0
Replacing Va1 by Ea – Ia1Z1 gives
E a +I a 1 Z 1
Ia2 = − and
Z2
E +I Z
I a 0 =− a a 1 1
Z0
Since Ia = 0, Ia1 + Ia2 + Ia0 = 0
E +I Z E +I Z
I a1 − a a 1 1 − a a1 1 = 0
Z2 Z0
I a1 Z 2 Z 0 −E a Z 0 + I a1 Z 1 Z 0 −E a Z 2 + I a1 Z 1 Z 2 = 0
E a ( Z 2+ Z 0)
I a1 =
Z 1 Z 2 +Z 1 Z 0 + Z 2 Z 0
Ea
=
Z2 Z 0
Z 1+
( Z 2+ Z 0 ) (119)
Equations (118) and (119) are special equations for a double line to ground fault. They are used
with equations (111) and (113) and the symmetrical component relations to determine all the
voltages and currents at the fault. Equation (119) indicates that sequence network should be
connected in parallel as shown in Figure 19 above.
In the absence of a ground connection at the generator, no current can flow into the ground at the
fault. In this case Z0 will be infinite and Ia0 = 0. So, as current is concerned the result will be the
Ea
I a1 →
same as line-to-line fault as Z0 → ∞ and Z 1 + Z 2 (comes from equation (119)).
Example 12: Find the sub-transient current and the line-to-line voltages at the fault on sub-
transient condition when double line to ground fault occurs at the terminals of the generator
22
described in example 10. Assume that the generator is unloaded and operating at rated voltage
when the fault occurs. Neglect the resistance.
Solution:
Ea
I a 1=
Z2 Z0
Z 1+
( Z2 +Z 0 )
1.0
=
j0 . 35× j 0. 1
j 0. 25+
j0 . 35+ j 0 . 1
1
= = − j3 . 05 p . u
j 0. 3277
V a 1 = V a 2 = V a 0 = E a −I a1 Z1
= 1−(− j 3. 05 )( j0 . 25 )
= 1−0 .763
= 0 . 237 p. u
V a2 0 . 237
Ia2 = − =− = j 0 . 68 p . u
Z2 j0 . 35
Va0 0 .237
I a 0 =− =− = j 2. 37 p. u
Z0 j 0 .1
I a = I a 1 +I a2 +I a0 = 0
I b = a 2 I a1 +aI a 2 +I a 0
= (−0 . 5− j0 . 866 ) (− j 3 .05 )+ (−0 .5+ j 0. 866 )( j0 . 68 ) + j 2 .37
= −3. 230+ j3 . 555
= 4 .80 ∠132 .3 ° p . u
I c = aI a 1 +a 2 I a2 +I a 0
= (−0 . 5+ j 0 .866 ) (− j3 . 05 ) + (−0 . 5− j 0. 866 )( j0 . 68 ) + j 2 .37
= 3 .230+ j 3. 555
= 4 .80 ∠ 47 .7 ° p . u
I n = 3 I a0 = 3× j2 .37 = j7 . 11 p .u
I n = I b +I c = −3 .230+ j3 . 555+3 . 230+ j 3 .555 = j7 . 11 p . u
V a = V a 1 +V a 2 +V a 0 = 3 V a1 = 3×0 .237 = 0. 711 p .u
Vb = Vc = 0
V ab = V a −V b = 0 .711 p . u
V bc = 0
V ca = V c−V a = −0. 711 p .u
Expressed in amperes and volts,
23
Ia = 0
I b = 837×4.80 ∠132 .3° = 4017 ∠132 .3° A
I c = 837×4.80 ∠47 .7° = 4017 ∠ 47.7° A
I n = 837×7.11∠ 90° = 5951 ∠90 ° A
13 .8
V ab = 0.711× = 5.66 ∠ 0° kV
√3
V bc = 0
13 .8
V ca = −0.711× = 5.66 ∠180 ° kV
√3
24
2. Generator bus: The generator bus requires the availability of real power generated and the
voltage magnitude of the bus; the reactive power and angle should be calculated. Generator bus is
also known as P-V bus.
3. Slack, Swing or Floating bus: This is a special generator bus serving as the reference bus for
the entire power system. The voltage magnitude and angle are specified (e.g. 1 ∠0 ° pu ) and
assumed to be fixed. This bus provides uncontrolled real and reactive powers necessary to
stabilize the system which remain unknown until the completion of the load flow study. Hence, it
is called infinite bus.
( )
X ij
θij = tan−1 and V i =|V i|e jδi
Rij
where (50)
Y sh
I1 = V 1 + ( V 1 − V 2 ) Y 12
So, from Figure 7, 2 (51)
¿
The complex power S is given by: S=V I (52)
25
S1
∴ I1 = ¿
V1 (53)
where V1* is the complex conjugate of V1, therefore
V 1 Y sh
I1 = + ( V 1 − V 2 ) Y 12
but 2
S1 V 1 Y sh
∴ ¿ = + ( V 1 − V 2 ) Y 12
V1 2 (54)
S2 V 2 Y sh
¿ = + ( V 2 − V 1 ) Y 21
Similarly, V 2
2 (55)
Substituting values of S1 and S2 from equation (47) into equation (54):
jδ 1
( Pg 1 − Pd 1 ) + j ( Q g 1 − Q d 1 ) |V 1|e
Y sh + (|V 1|e )|Y 12|e
jδ 1 jδ 2 − jθ 12
− jδ 1
= j − |V 2|e
|V 1|e 2
(56)
Multiplying both sides of equation (56) by
(57)
where
(59)
Recall that , and substituting in equation (59)
(61)
(63)
Adding equations (60) and (62) to obtain the real power balance equation:
P g1 + Pg 2 = P d 1 + Pd 2 + Y 12 cosθ12 {|V 1|2 + |V 2|2 − 2|V 1||V 2| cos ( δ 1 − δ 2 ) }
= d1 d2 P +P +P
L (64)
Also adding equations (58) and (60) to obtain the reactive power balance equation:
Y
{
Q g1 + Q g2 = Q d 1 + Qd 2 + Y 21 sin θ21 |V 1|2 + |V 2|2 − 2|V 1||V 2| cos ( δ 1 − δ 2 ) + (|V 1|2 + |V 2|2 ) sh
2 }
26
= Qd 1 + Q d 2 + Q shL + Q L (62)
where PL, QL and QshL are the real power loss, reactive power loss and reactive power loss due to the
shunt admittance which is a function of the voltage parameters only. There are twelve unknown
variables in the above non-linear algebraic equations relating voltages and powers which must be
reduced to four to match equations (57) to (60).
4.4 Load Flow Solution Methods
We shall examine four solution methods as enumerated below:
(a) DC technique
(b) Gauss-seidel technique
(c) Newton-Raphson method
(d) Fast-Decoupled power flow method.
[ ]
sp sp
1 Pk − jQ k
Vk = − ( Y k 1 V 1 + Y k 2 V 2 + .. . + Y km V m + . .. + Y kN V N )
Y kk V ¿k
(65)
¿ m ≠ k ¿¿ ¿ ¿¿¿
N (66)
where * is the conjugate function.
Computation of Reactive Power injected into Bus k
k = − Im [ V k ( Y k 1 V 1 + Y k2 V 2 + ... + Y km V m + ... + Y kN V N ) ]
¿
Qcalc .
(67)
27
Im[ ] indicates the imaginary part of the expression in the brackets in equation (67). The
computed value of the reactive power Q kcalc. is used to replace Qksp in equation (65) and Vk is
recalculated. Then the magnitude of V k is reset to its specified value V ksp, but the new value of its
phase angle is retained.
The Gauss-seidel method has the attractions of simplicity, comparatively good performance
and non-storage of previous values. It has a very reliable convergence characteristic, but the rate
of convergence is quite slow. The rate of convergence is improved somewhat with the use of an
accelerating factor, α. The best value of acceleration factor to use is usually different for different
systems. A value of 1.6 for α is generally considered a good choice.
[ V (ki + 1) ] mod = V (ki ) + αΔV (ki + 1 ) (68)
ΔV (ki + 1 ) = V (ki + 1 ) − V (ki ) (69)
The real power and reactive power losses of any transformer or transmission line are derived
from the line power flows computed by using equation (70) below.
( )
¿
Vk − V m Y
Skm = Pkm + jQ km = V k + Vk
Z 2 (70)
where Z = series impedance and Y = shunt admittance; subscripts m and k are buses.
In summary, the solution steps of Gauss-seidel technique (a.c. technique) are as follows:
V ∠δ = 1.0∠0°
(a) Assume m m for all load bus bars m and δj = 0 0 for all generator bus bars j.
These are the initial guesses and known as the flat voltage profile.
(b) If bus bar k is a load bus bar, from equation (66), P k, Qk are known hence Vk can be computed
using the known values of Vm (obtained either from the initial voltage profile or from the best
estimate obtained earlier).
(c) If bus bar k is a generator bus bar, rearranging equation (71) below,
( )
n ¿
Pk + jQ k
= ∑ Y km V m
Vk m=1
(71)
[ ]
n
Qk = − Im aginary V k
¿
∑ Y km V m
m=1 (72)
Example 4: Determine the value of the V2 at bus 2 that is produced by the first iteration of the
Gauss-seidel method from Figure 8. Bus number 1 is taken as swing bus. Find the bus admittance
matrix, Ybus.
28
Solution: The bus admittance matrix is thus formed,
1 1
Y 12 = Y 21 = = =2 − j6
Z 12 0.05 + j0.15
1 1
Y 13 = Y 31 = = =1 − j3
Z 13 0.10 + j 0.30
Y 14 = Y 41 = 0
1 1
Y 23 = Y 32 = = = 0.5 − j 1.5
Z 23 0.20 + j 0.60
1 1
Y 24 = Y 42 = = =1− j3
Z 24 0.10 + j0.30
1 1
Y 34 = Y 43 = = = 2 − j6
Z34 0.05 + j 0.15
Note: Yii = sum of the admittances incident on bus i, therefore
Y 11 = ( 2 − j 6 ) + ( 1 − j3 ) + 0 = 3 − j 9
Y 22 = ( 2 − j6 ) + ( 0 . 5 − j 1. 5 ) + ( 1 − j 3 ) = 3 .5 − j10 . 5
Y 33 = ( 1 − j3 ) + ( 0 .5 − j1 . 5 ) + ( 2 − j 6 ) = 3 .5 − j10 . 5
Y 44 = ( 1 − j 3 ) + ( 2 − j6 ) = 3 − j 9
Y 12 = Y 21 = − Y 12 = −2 + j 6; Y 13 = Y 31 = − Y 13 = − 1 + j 3; Y 14 = Y 41 = 0
Y 23 = Y 32 = − Y 23 = −0.5 + j 1.5; Y 24 = Y 42 = − Y 24 = − 1 + j 3
Y 34 = Y 43 = − Y 34 = − 2 + j6
Therefore, the bus admittance matrix is thus given below.
Ybus =¿[ 3− j9 −2 + j6 −1 + j3 0¿][−2 + j6 3.5 − j10.5 −0.5 + j1.5 −1 + j3¿][−1 + j3 −0.5 + j1.5 3.5 − j10.5 −2 + j6¿] ¿
The specified real and reactive powers at bus 2 are
¿
29
Psp
2 = −1. 00 p .u
Q2sp = −0 .75 p .u
The initial values of the bus voltages are
V 1 = 1.05∠ 0° p.u; V 2 = 1.00∠0° p.u
V 3 = 1.02∠0° p.u; V 4 = 1.00∠0° p.u
The voltage at bus 2, for the first iteration, is calculated as follows:
[ ]
sp sp
1 P2 − jQ 2
V2 = − Y 21 V 1 − Y 23 V 3 − Y 24 V 4
Y 22 V ¿2
Then substituting the initial values of the bus voltages and the elements of the admittance matrix.
V2 =
1
[
−1.00 + j0.75
3.5 − j10.5 1.00∠0°
− (−2 + j6 )( 1.05∠ 0° ) − (−0.5 + j1.5 )( 1.02∠0° ) − (−1 + j3 )( 1.00∠0° ) ]
= ( 0.0286 + j 0.0857 ) [ (−1 + j 0.75 ) − (− 2.1 + j6.3 ) − (−0.51 + j 1.53 ) − (−1 + j3 ) ]
= ( 0.0286 + j 0.0857 ) ( 2.61 − j 10.08 )
= 0.9385 − j0.0646 or 0.94∠−3.94° p.u
TUTORIALS
1. A two-bus power system has a bus admittance matrix given by
30
4.4.3 THE NEWTON-RAPHSON METHOD
The generalized Newton-Raphson method is an iterative technique for solving a set of
simultaneous non-linear equations involving the same number of unknown variables.
[ H N ¿]¿¿¿
¿ (73)
Equation (73) is solved directly for the incremental voltage angles and incremental voltage
magnitudes that are used as corrections to the current estimate of the solution. Any known
solution technique for a system of linear algebraic equations may be used to solve it. These
computed corrections are used to determine new estimates of the voltage phase angles and
voltage magnitudes as follows:
δ new = δ old + Δδ new
V new = V old + ΔV new (74)
The iterative process is said to have converged when the bus real power mismatch vector ∆P and
reactive power mismatch vector ∆Q which can be computed by using equations (75) and (76)
below:
N
ΔP k = V k ∑ V m [ G km cos ( δ k − δ m) + Bkm sin ( δ k − δ m )] − Pspk = 0
m=1 (75)
N
ΔQ k = V k ∑ V m [ Gkm sin (δ k − δm ) − Bkm cos (δ k − δm ) ] − Q spk = 0
m=1 (76)
The Newton-Raphson method converges most rapidly of any of the power flow solution
techniques, especially when the solution point is close. It has a quadratic convergence
31
characteristic. It is reliable and has minimal sensitivity to factors that cause poor convergence,
such as the choice of swing bus and the presence of series capacitors in the network which could
cause problems in the Gauss-seidel method.
Example 5:
Ybus =¿[ 3−j 9 −2+j 6 −1+j3 ¿] [−2+j 6 2.5 −j 7.5 −0.5 +j 1.5 ¿] ¿¿
¿
(a) For each bus k, specify the bus type, and determine which of the variables V , δ , P and k k k
Qk are input data and which are unknowns.
(b) Assume an initial estimate of V2 = 1.0∠0 ° and δ3 = 0˚, and calculate the bus real and
reactive power mismatches to be used in the first iteration of the Newton-Raphson power
flow method.
(c) Set up the linearized system of equations what are solved at each iteration of the Newton-
Raphson power flow method.
Solution:
(a)
Bus Number Bus Type Input Data Unknowns
1 Swing V1, δ1 P1, Q1
2 Load P2, Q2 V2, δ2
3 Generator P3, V3 δ3, Q3
V = 1.00∠ 0°
(b) Using the initial estimates 2
and δ 3 = 0.0 , the real and reactive bus power
mismatches are calculated using equations (75) and (76) with m ( k
δ = δ −δ m )
as follows:
ΔP 2 = V 2 [ V 1 ( G21 cos δ 21 + B21 sin δ 21) + V 2 G22 + V 3 ( G23 cosδ 23 + B23 sin δ23 ) ] − P2sp
= 1. 00 [ 1. 05 (−2 cos0 + 6. 0 sin 0 ) + 1. 00×2. 5 + 1 . 02 (−0. 5 cos0 + 1.5 sin 0 ) ] − (−1.5 )
= 1 .39 p .u
32
ΔP 3 = V 3 [ V 1 ( G31 cosδ 31 + B 31 sin δ 31 ) + V 3 G33 + V 2 ( G32 cosδ 32 + B 32 sin δ 32 ) ] − Psp3
= 1. 02 [ 1. 05 (−1. 0 cos 0 + 3. 0 sin 0 ) + 1. 02×1 .5 + 1 .00 (−0. 5 cos0 + 1.5 sin 0 ) ] − ( 0 . 6 )
= −0. 62 p. u
ΔQ 2 = V 2 [ V 1 ( G21 sin δ21 − B21 cosδ 21 ) − V 2 B 22 + V 3 ( G23 sin δ 23 − B23 cos δ23 ) ] − Q2sp
= 1.00 [ 1.05 (−2.0 sin 0 − 6.0 cos0 ) − 1.00×−7.5 + 1.02 (−0.5 sin 0 + 1.5 cos0 ) ] − (−1.00 )
= 0.67 p.u
(c) The real and reactive power mismatches calculated above are used in the linearized system of
equations to find corrections or improvements in the values of the voltage magnitudes and angles
as shown in the following equations.
[ ][ ]
¿
∂P2 ∂P2 ∂P2 ∂P3 ∂P3 ∂P3
¿
∂δ 2 ∂δ 3 ∂V 2 ∂δ2 ∂δ3 ∂V 2
¿ ¿ ¿¿
Because of the negligible value of the phase angle of the bus voltages and the substantial values
of the transformer and line reactances, the following assumptions are made.
cos ( δ k − δ m ) ≃ 1 G km << Bkm
(79)
sin ( δ k − δ m ) ≃ 0 Q k << Bkk V 2k (80)
So, good approximations of equations (77) and (78) above are given by
33
N
∑ V k (−B'km) V m Δδ m = − ΔPk
m=1 (81)
N
∑ V k (−B'km' ) ΔV m = − ΔQ k
m=1 (82)
Thus, the Fast-Decoupled power flow equations may be expressed as
B' Δδ =
ΔP k
Vk [ ] (83)
'
B' ' ΔV =
''
[ ]ΔQk
Vk
(84)
Example 6: Form the B and B matrices that are used in the Fast-Decoupled power flow solution
from this four-bus power system in Figure 8.
'
Solution: The B matrix is formed using the same procedure as in building the bus admittance
matrix with the resistances of the lines neglected.
'
The elements of B are found as follows:
1 .0
jB 12 = jB 21 = − = j 6 .667
j0 . 15
1 .0
jB 13 = jB 31 = − = j 3. 333
j 0 . 30
jB 14 = jB 41 = 0
jB 11 = − j 6 . 667 − j 3. 333 = j 10
1. 0
jB 23 = jB 32 =− = j1 .667
j 0. 60
34
1. 0
jB 24 = jB 42 = − = j3 .333
j 0. 30
jB 22 = − j6. 667 − j 1 .667 − j3 . 333 = − j11. 667
1. 0
jB 34 = jB 43 =− = j6 .667
j 0. 15
jB 33 = − j3. 333 − j1. 667 − j 1.667 = − j 11.667
jB 44 = − j3. 333 − j 6. 667 = − j10
'
Thus, the B matrix is given by
[B =¿ −10.0 0 6.6 7 3.3 3 0.0 0¿][ 6.6 7 −11.6 7 1.6 7 3.3 3¿][ 3.3 3 1.6 7 −11.6 7 6.6 7¿]¿¿
'
Although these words, forecasting and prediction are often used interchangeably, these two
words differ in meaning.
Prediction is generally a statement or a guess of something which will happen regardless of
whether the people want it or not and is something beyond one’s control, e.g. one might predict
rain or bad weather.
On the other hand, forecasting is a statement of what will happen if certain conditions or trends
continue and it assumes that the causes of the events are under human control. It has been
extensively used in areas such as electrical load forecasting, economic trends and market surveys.
Load forecasting can be divided into two categories, namely, long term load forecasting and
short-term load forecasting.
5.1 LONG TERM LOAD FORECASTING: It takes a pretty long time to plan, install and
commission additional generating capacity. Generation system expansion planning starts with a
forecast of anticipated future load requirements. Proper long-term forecasting is necessary for
optimal generation capacity expansion.
35
Various techniques are used for long term load forecasting. One of the techniques is
extrapolation technique which involves fitting trend curves to basic historical data adjusted to
reflect the growth trend itself. Once the trend curve is known, the forecast is found by evaluating
the trend curve function at the desired future point. And some of the functions used in trend curve
fitting are straight line, parabola, polynomial, exponential, etc.
Another technique for load forecasting is correlation. This technique relates system loads to
various demographic and economic factors. Correlation is a measure of how well a line or curve
fits a given set of data. It shows how one variable is related to another.
Regression analysis is a technique that studied the behavior of a time series or a process in the
past and its mathematical modelling so that future behavior can be extrapolated from it.
Regression analysis can further be discussed under the following regression curves used in power
system forecasting:
(1) Linear, i.e. y = A + Bx, where x = T − 1 + n ; T = Number of years for which statistical
trend is studied,
n = Number of years for which forecast is required.
(2) Exponential, i.e. y = A ( 1+B )x
y = Ax B
(3) Power, i.e.
2
(4) Polynomial, i.e. y = A + Bx + Cx
However, partial differentiation with respect to the regressions coefficients and the equations set
to zero, yield these two simultaneous equations:
∑ y = an + b ∑ x (50)
∑ xy = a ∑ x + b ∑ x
2
(51)
where n = number of pairs of figures.
This is referred to as Least Square Line and could be said that it is the regression of y on x.
Example: The maximum demand pattern of power system in Abuja from 2007 to 2013 in MW
are given in the table shown below. Using the least square regression method, predict the demand
for 2014 and 2015.
5.2 SHORT TERM LOAD FORECASTING: Short term load forecasting is essentially used for
monitoring and controlling power system operation. The load forecast can be done hourly or
weekly. The pattern representing the hour of the day effect or the day of the week effect on load.
37