UNIVERSITY OF EL SALVADOR

FACULTY OF ENGINEERING AND ARCHITECTURE

SCHOOL OF ELECTRICAL ENGINEERING
CYCLE II - 2018

Cristian Antonio Aguilar Quintanilla.

Load flow exercises.

Exercise 1 (Based on Example 9.2 and Example 9.3 from the book Power System Analysis)
by Grainger and Stevenson Jr.). Consider the following four-bar system, the power
The base three-phase power is 100 MVA, and the base line-to-line voltage is 230 kV.

The series impedances and susceptances are given in per unit for the equivalent circuits.
nominal π of the calculations of the four identified lines that end in the bars.

1
̅ ̅ = Admittances in parallel
̅
Line, from bar to bar
̅
Rseries Xseries Gseries Bseries Total load MVAR
2
1-2 0.01008 0.05040 3.815629 -19.078144 10.25 0.05125
1-3 0.00744 0.03720 5.169561 -25.847809 7.75 0.03875
2-4 0.00744 0.03720 5.169561 -25.847809 7.75 0.03875
3-4 0.01272 0.06360 3.023705 -15.118528 12.75 0.06375
̅ 2
Note: This was done to calculate it. = 3V√ I e = / √3it is obtained that = .
2
(10.25 MVAR)/230(kV
For the first piece of data 2 times 10-4Sthe
system base impedance
equals)1.9376
it will be = 2/ 3 2
(230kV )/(100 MVA) = 529 Ω, therefore [pu ] = × =
0.1025 puthe shunt admittance will be: ó [pu] = [pu]/2 = 0.05125 pu.

The following table lists the values of P, Q, and V for each bar.

Generation Load Tension

Bar
P [MW]
1 ----- ----- 50 30.99 1.00 0°
2 0 0 170 105.35 1.00 0°
3 0 0 200 123.94 1.00 0°
4 318 ----- 80 49.58 1.02 0°
Note: The reactive load powers at the bars are calculated based on the values of the active powers.
considering a power factor of 0.85 lagging.

Since buses 2 and 3 have no generation, they are pure load buses (PQ buses). Bus 4
it has load and generation, but the reactive power generated is not specified (it is a
variable), so it is a controlled tension bar (PV bar, meaning fixed tension in magnitude,
but with adjustable bar tension angle). Bar 1, due to the unknown powers
generated, is assigned as the reference bar.

Perform three iterations to find , , 4̅ , 2 ̅ 3, y 4

, through:
1 1

a) Gauss-Seidel method.
b) Gauss-Seidel method with over-relaxation of 1.6 (acceleration of convergence).

For part a), complete the entire Gauss-Seidel algorithm, for part b) the results will be obtained.
equal results. Some apply the acceleration of convergence when they finish calculating the
non-accelerated values of an iteration, that form is not mathematically correct, although
It yields satisfactory results, but more iterations are needed; the best thing is to accelerate each value.
calculated and use it in the following state variable to calculate.

Solution. From step 3, the difference between the method without acceleration and with
acceleration. The values that have no units are in pu.
Step 1. Establish the initial injections + in all the system bars:

It cannot be determined yet

0
[ − ] −1.7
[ ] = =[ ]
3 −2
2.38
 It cannot be determined yet
0
[ − ] −1.054
[ ]= =[ ]
3 −1.239
It cannot be determined yet.

Step 2. Construction of the bus admittance matrix.

̅ + 13
12 ̅ +̅ + ̅ 120 130 − ̅ 12 − ̅ 13 0
− ̅ 21 ̅21+ 24
̅ + ̅ + ̅ 210 240 0 − ̅ 24
[̅ ]=[ ]
− ̅ 31 0 ̅ 31 ̅ + 310
+ 34 ̅ + ̅ 340 − ̅ 34
0 − ̅ 42 − ̅ 43 ̅ + 43
42 ̅ + ̅ + ̅ 420 430
8.985− 44.836−3.816 + j19.078 −5.17 + j25.848 0
−3.816 + j19.078 8.985− 44.836 0 −5.17 + j25.848
[̅ ]=[ ]
−5.17 + j25.848 0 8.193− 40.864−3.024+ j15.119
0 −5.17 + j25.848−3.024+ j15.119 8.193− 40.864
Step 3. Literal a), Gauss-Seidel method without convergence acceleration.

First iteration.
Second iteration.
Third iteration.

To continue with the following steps, iterations up to 14 were carried out, obtaining:
Step 4. Calculation of the power injected into the reference bus and the generated powers
and demands from it and the one generated in the controlled tension bar, after the
iteration 14.

Step 5. Calculation of the line flows.

The losses on the lines will be:
Simulation in Power World
Step 3. Literal b), Gauss-Seidel method with convergence acceleration.

First iteration.
Second iteration.
Third iteration.
Exercise 2. Consider the following system of three bars. A table is presented with the
generated and/or demanded power at each bus. Apply the Gauss-Seidel method with a
acceleration factor of 1.4 and a global tolerance of 0.02, find the tensions in the
bars 2 and 3 and resolve the load flow for the two conditions shown.

Lines, from bar to bar ̅ ̅ pu

in puY/2in
1-2 j0.01 0
2-3 j0.02 0
1-3 j0.02 0

Load condition 1:

Bar Generation Load Tension

P [pu] Q [pu] P [pu] Q [pu] V [pu] δ [degrees]
1 ----- ----- 0 0 1.0 0°
2 1.2 0.74 0 0 1.0 0°
3 0 0 3.64 2.5 1.0 0°

Loading condition 2:

Bar Generation Load Tension Power limits

P [pu] Q [pu] P [pu] Q [pu] V [pu] δ [degrees] QGminQGmax[pu]
1 ----- ----- 0 0 1.03 0.2° ----- -----
2 1.2 ------ 0 0 1.03 0° 0.87 1.6
3 0 0 3.64 2.5 1.0 0° ----- -----

Solution. For load condition 1, bar 1 is the reference, bar 3 is the load and
bar 2 when specifying its generated powers (both active and reactive), its voltage of
The magnitude of the bar is not fixed; it is considered as a loading bar (due to conditions of
classification). For load condition 2, now bar 2 is a tension-controlled bar
with restrictions on the levels of reactive power generated, if those limits are violated, it is
It is necessary to restrict the limited power value to a value given by the limits and to handle it
bar like PQ, and since there are no restrictions for tension in that bar it cannot be
return to being PV bar.
Load condition 1:

Step 1. Bar injections.

Step 2. Construction of bus admittance matrix.

Step 3.
First iteration.

Second iteration.
Third iteration.

Fourth iteration.

It stops because the tolerance requirement for both voltages is met.

Step 4. Determination of power in the reference bar.

Step 5. Calculation of line flows.

Simulation in Power World assuming voltage levels of 138 kV, 100 MVA.
Load Condition 2:

Step 1.

Step 2. The same matrix as in condition 1.

Step 3.

First iteration.
Second iteration.
Third iteration.
Fourth iteration. It could have stopped at the end of the previous iteration, but to decrease
the mistake became one more.

Stop the iteration.

Step 4. Calculation of the power injected at the reference bus.

Step 5. Calculation of line flows.

Note: In case of violating reactive power, do 2= 2 á − 2 , and the bar will be PQ

now. Since the tension should not be kept constant on bar 2 and as there are no
tension limits in that bar, it will remain as PQ, meaning there is no possibility that
return to being PV. If the case arises that the bar becomes PQ, it can be established 2 = 1 pu,
instead of the given value when it was considered PV, but it is not necessary, since the same
the iterative process will correct it.
Exercise 3. (Based on problem 6.36 from Duncan's book). The admittance matrix of
The bar in the power system shown in the figure is as follows:

3− j9 −2 + j6 −1 + j3 0
−2 + j6 3.666 − j11−0.666 + j2−1 + j3
[̅ ]=[ ]
−1 + j3 −0.666 + j2 3.666− j11−2 + j6
0 −1 + j3 −2 + j6 3− j9

The complex powers are shown in bars 2, 3, and 4 in the same figure. Determine
the values of the tensions in those bars produced for the first and second iteration,
using the methods:

a) Gauss-Seidel.
b) Newton-Raphson.

Assume constant initial tensions of1.0∠0° puin bars 2, 3 and 4.

Solution. Bars 2, 3, and 4 are pure load bars (they have no generators connected, or
in the case of bars 2 and 4, their injected active power is positive Pi=Pgi-Pdi, so
they could be generators that send constant active power and consume reactive power
constant, that is why their injections are negative, meaning they are synchronous generators
sub-excited, with the field circuits and mechanical axis adjusted for those values
shown), bar 3 can be a simple RC load.
 a) Gauss-Seidel method. Step 2 is omitted because they already provide the admittance matrix.
bar.

Step 1.
Step 3.

First iteration.
Second iteration.

Kothari's book provides a procedure to make the calculations in a 'simpler' way.

doing them in parts:
First iteration.
Second iteration.
 b) Newton-Raphson method.

Step 1. Establish the initial voltages and calculate the bus admittance matrix.
First iteration.

Step 2. Calculation of the for bars PQ and PV, and for the PQ bars, starting from the
power balance equations ’= − = − , ’’ = − =
− ) = − , = − .

The scheduled powers are those of the diagram, the shown ones, which would be =
− y = − .
Step 3. The residual vector is established and the Jacobian matrix is calculated.

The vector 'r' was only calculated to indicate the order of the terms, the solution is worked on.
with the vector 'R'.
Step 4. Calculation of the corrections.

Step 5. Update of angles and tension magnitudes.

The vectors containing the angles, magnitudes, and voltage phasors are updated in the
first iteration.
Second iteration.

Step 2.
Step 3.
Step 4.

Step 5.