The Gauss-Seidel Method-Procedure & Example
Consider a power system with a total of n buses, out of which G is
the number of generator buses. Bus number 1 is taken as the slack
(swing) bus. The steps for the Gauss-Seidel algorithm for load
flow studies are summarized in the following.
1. Form the network YBUS (excluding the generator-transformer
circuit) in pu system.
2. Assume the following initial values,
δi=0 i = 2,3, ……………. G
Vi=1, δi=0 i = G+1, G+2, ……… n
3. Calculate Q2 by the equation, use the value of subscript i=2
n
(7)
Q i = − Im[(Yii Vi + ∑ Yij Vj )Vi* ] , j≠i
j=1
4. Calculate V2=V2c∠δ2c by the equation (c stands for calculated
value), use i=2
1 Pi − jQ i n (8)
Vi = [ − ∑ Yij Vj ], j ≠ i
Yii Vi* j=1
Set V2=V2s∠δ2c where, V2s is the specified generator voltage at bus
2.
5. Repeat steps 3 and 4 for k = 3, 4, …G
6. Calculate VG+1 by equation (8). Note this is a load bus and the
sign of Pi and Qi should be reversed.
7. Repeat step 6 for all the load buses k = G+2, G+3…. n
8. Calculate ∆V and ∆δ for all the buses comparing with values
from the previous step by the relationship,
∆Vi = Vi1 − Vi0 , ∆δ i = δ i1 − δ i0 i=2,3, ….n
9. This is the end of the iteration. Check convergence of the
solution for all the bus voltages. If the maximum value is
within the specified tolerance, a solution has been reached.
Otherwise, repeat steps 3 though 8 until the solution is
reached.
10. Calculate P1 and Q1 from
n (9)
S1 = V1[( ∑ Y1 j Vj )]*
j=1
The transmission line flows can be computed from
(10)
Sij = Vi [(Vi − Vj ) y ij ]*
where, yij is the admittance of line i-j.
Convergence can be accelerated by using an acceleration
factor (between 1.3 and 1.6) in steps 4 though 7. The new
value is calculated as,
Vikα+1 = Vik + α (Vik +1 − Vik )
k +1
where, α is the acceleration factor. Viα is the accelerated
value which is used in the next step instead of Vik+1
Example: The Gauss-Seidel Method
The one line diagram of a power system, with the generator reactances
omitted, is given below. The reactances indicated are on 100 MVA base.
1 j0.5 2
j0.2 j0.25
The power output of generator 2 is 60 MW and the voltage is 1 pu. The load
on bus 3 is P=80 MW, Q=60 MVAR. Considering the voltage of slack
generator 1 to be, obtain a power flow solution using a tolerance of 0.01 on
the voltage magnitudes.
Summary of bus information provided
Bus 1 : slack (swing) V 1= 1∠0 0 P1 = Q 1=
Bus 2 : Generator (PV) V2=1 pu, δ2= P2= 60 MW (0.6 pu) Q 2=
Bus 3 : Load (PQ) V 3= δ3= P3=80 MW Q3=60MVAR
6 known quantities, 6 unknown quantities.
Follow the steps as given on the procedure:
7 − 2 − 5
1.
Y = − j − 2 6 − 4
BUS
− 5 − 4 9
2. Select the unknown quantities as,
δ2 = 0 │V3 │= 1 δ3 = 0
3. Iteration #1
Calculate,
Q2= -Im[(Y22 V2 + Y21 V1 + Y23 V3) V2*]
= -Im[ (-j6 x 1∠0 0 +j2 x 1∠0 0 + j4 x1∠0 0)]
=0
1 P2 − jQ 2
4. V2 = [ − Y21V1 − Y23 V3 ]
Y22 V2*
1 0.6 + j0
= [ − j2 x1∠0 − j4 x (1∠0)]
- j6 V2*
= 1 + j0.1 = 1.00498∠5.710
Set, V2= 1.00∠5.71 = 0.999+j0.0995
0
5. Does not apply
6.
1 P3 − jQ 3
V3 = [ − Y31V1 − Y32 V2 ]
Y33 V3*
1 − 0.8 + j0.6
= [ − j5x1∠0 − j4 x (1∠5.71)]
- j9 V3*
= 0.93112 - j0.0446 = 0.9321∠ - 2.74 0
7. Does not apply
8. Check:
∆V2 = V2(1) − V2( 0 ) = (0.995 + j0.0995) − (1 + j0) = 0.0996
∆V3 = V3(1) − V3( 0 ) = (0.93112 − j0.0446) − (1 + j0) = 0.082
MaximumVariation = 0.0996
9. Iteration #2
Q 2 = − Im{[− j6 x1∠5.710 + j2 x1∠0 0 + j4 x 0.932∠ − 2.74 0 ]x (1∠ − 5.710 )}
= 0.3216 pu
1 0.6 − j0.3216
V2 == [ *
− j2 x1∠0 − j4 x (0.932∠ − 2.74 0 )]
- j6 V2
= 1.0003∠4.3030
Set, V2 = 1∠4.3030
1 − 0.8 + j0.6
V3 = [ *
− j5 x1∠0 − j4(1∠ 4 . 303 0
)]
- j9 V3
= 0.9227 − j0.058 = 0.9245∠ − 3.6250
Check:
∆V2 = 1∠4.3030 − 1∠5.710 = 0.0245
=
∆V3 0.9245∠ − 3.6250 − 0.9321∠ − 2.74
= 0
0.016
Magnitude of maximum change is 0.0245
Iteration #3
V2 = 0.9978 + j0.065 = 1∠3.72 0
V3 = 0.921 − j0.062 = 0.923∠ − 3.880
∆V2 = 0.01004
∆V3 = 0.0043
The desired tolerance is acheived.
The Line flows, Line losses, and Slack Generation
Flow in line 1-2
S12 = V1I12
*
= V1 [(V1 − V2 ) y12 ]* xBaseMVA
= −13.0043MW + j0.4232 MVAR
Similarly, flow in line 2-1
S21 = V2 I*21
= V2 [(V2 − V1 ) y 21 ]* xBaseMVA
= 13.0043MW + j0.4232 MVAR
Loss in line 1-2
Sloss(1-2)= j0.8464 MVAR
or, Ploss(1-2) =0; Qloss(1-2) = 0.8464 MVAR (reactive)
Flow in line 1-3
S13 = V1I13
*
= V1 [(V1 − V3 ) y13 ]* xBaseMVA
= 31.2438MW + j39.489 MVAR
Calculate the other line flows.
Slack Generation
S1= S12 + S13
=18.239 MW + j39.9128 MVAR
The line flows are shown in the figure below.
V1 = 1∠0 0 V2 = 1∠3.72 0
18.24+j39.9 60+j34.4
1 2
-13.0+j.42 13.0+j.42
31.24+j39.48 48.89 +j33.99
-31.24-j34.41 3 -48.89-j25.12
S3=80+j60
