EEE 5103 POWER SYSTEMS ANALYSIS I

Lecturer: Peter Yegon

Prerequisites
EEE 4210 Power Systems II
Expected Learning Outcomes
By the end of this course, the learner should be able to;

• Apply load flow analysis to an electrical power network and interpret the results of the
analysis
Course Outline
Power flow: basic considerations, formula of the system array, non-linear algebraic equations,
Gauss-Seidel method, Newton-Raphson method, Simplification of the Newton-Raphson method;
the fast decoupled power flow.
Computer simulation of large power systems: Stott-Alsac fast decoupled algorithms.
Load forecasting techniques for short-term planning.
Course Assessment
Cats 10%
Assignments 5%
Labs 15%
Exam 70%

Total 100%

Core Text Book

Nagsarkar,T. (2007). Power System Analysis. India: Oxford University Press
Reference Text Books
1. Stevenson, William D. (1982). Elements of Power Systems Analysis (4th ed.).
Auckland: MacGraw-Hill.
Course Journals
1. IEEE Transactions on Power systems, ISSN: 0885-8950
Contents
1 INTRODUCTION: .................................................................................................................. 1
1.1. STRUCTURE AND REPRESENTATION OF A POWER SYSTEM .......................... 1
1.2. Single Line Representation of Power Systems ............................................................... 2
1.3. The Per Unit System ....................................................................................................... 4
1.3.1. Per Unit Representation of Transmission Lines ..................................................... 7
1.3.2. System Analysis in Per Unit System ...................................................................... 8
2 Load Flow Analysis ............................................................................................................... 12
2.1. Introduction ................................................................................................................... 12
2.1.1. Importance of Power (load) flow studies .............................................................. 12
2.1.2. Load Flow: Objectives and applications ............................................................... 12
2.2. Power/Load flow problem ............................................................................................ 12
2.2.1. Formulation of power-flow study ......................................................................... 13
2.2.2. Formulation of the Bus Admittance Matrix .......................................................... 14
2.2.3. Y-Bus Matrix Building Rules ............................................................................... 19
2.3. Exercise ......................................................................................................................... 20

ii
1 INTRODUCTION:
1.1. STRUCTURE AND REPRESENTATION OF A POWER SYSTEM
A typical power system consists of different parts. These are generation, transmission and
distribution sub-systems.

Fig 1.1. Structure of a Power System

In practice, several such power systems are interconnected to form regional or national grids. In
these grids, the generating stations, transmission and distribution networks and demand centres
are interconnected in such a manner as to improve system reliability and economy.

1
1.2. Single Line Representation of Power Systems
A single-line diagram of a balanced three-phase power system shows the main components as
well as the connections between them. A component is only given in the diagram if it is of
interest for the analysis. Figure 1.2 shows the single-line diagram of a simple balanced three-
phase power system. The system consists of four buses (or nodes) numbering from one to four,
two generators G1 and G2, two transformers T1 and T2, two loads LD1 and LD2, and a
transmission line between bus2 and bus3.

Figure 1.2. Single-line diagram of a small power system

Here-onward, if not otherwise explicitly stated, the following is valid in this compendium:

• all system quantities (power, voltage, current, impedances and admittances) are given in
the complex form,
• power is given as three-phase power in MVA, MW and/or MVAr,
• for the phasor voltage 𝑈𝑈‾ = 𝑈𝑈∠𝜃𝜃, the magnitude 𝑈𝑈 is a line-to-line voltage given in kV,
however the phase angle 𝜃𝜃 is the phase angle of a line-to-neutral voltage,
• currents (given in kA), impedances (given in Ω ) and admittances (given in S) are per
phase quantities.

Consider again the system shown in Figure 1.2. A typical system data can be given as follows:

• Generator G1 : 𝑆𝑆𝑛𝑛𝑛𝑛 = 30MVA, 𝑈𝑈𝑛𝑛𝑛𝑛 = 10kV, 𝑥𝑥𝑔𝑔 = 10%

• Generator G2 : 𝑆𝑆𝑛𝑛𝑛𝑛 = 15MVA, 𝑈𝑈𝑛𝑛𝑛𝑛 = 6kV, 𝑥𝑥𝑔𝑔 = 8%
𝑈𝑈1𝑛𝑛 10𝑘𝑘𝑘𝑘
• Transformer T1 : 𝑆𝑆𝑛𝑛𝑛𝑛 = 15MVA, = , 𝑥𝑥𝑡𝑡 = 10%
𝑈𝑈2𝑛𝑛 30𝑘𝑘𝑘𝑘
𝑈𝑈1𝑛𝑛 30𝑘𝑘𝑘𝑘
• Transformer T2 : 𝑆𝑆𝑛𝑛𝑛𝑛 = 15MVA, = , 𝑥𝑥𝑡𝑡 = 10%
𝑈𝑈2𝑛𝑛 6kV
• Line : 𝑟𝑟 = 0.17Ω/km, 𝑥𝑥 = 0.3Ω/km, 𝑏𝑏𝑐𝑐 = 3.2 × 10−6 𝑆𝑆/km and ℒ = 10 km
• Load LD1 : impedance load, 𝑃𝑃𝐿𝐿𝐿𝐿 = 15MW, 𝑈𝑈𝑛𝑛 = 30kV, cos 𝜙𝜙 = 0.9 inductive
• Load LD2 : impedance load, 𝑃𝑃𝐿𝐿𝐿𝐿 = 40MW, 𝑈𝑈𝑛𝑛 = 6kV, cos 𝜙𝜙 = 0.8 inductive

Comments:

2
𝑆𝑆𝑛𝑛𝑛𝑛 is the generator three-phase rating, 𝑈𝑈𝑛𝑛𝑛𝑛 is the generator rated (or nominal) line-to-line
voltage and 𝑥𝑥𝑔𝑔 is the generator reactance given as a percent based on the generator rated values.
The actual value of the generator reactance can be determined by
2
𝑥𝑥𝑔𝑔 𝑈𝑈𝑛𝑛𝑛𝑛
𝑋𝑋𝑔𝑔 = Ω and 𝑍𝑍‾𝑔𝑔 = 𝑗𝑗𝑋𝑋𝑔𝑔
100 𝑆𝑆𝑛𝑛𝑛𝑛

In a similar way the actual value of the transformer leakage reactance can be determined,
however, depending on which side of the transformer it will be calculated. Having the reactance
on the primary side, then it is determined by
2
𝑥𝑥𝑡𝑡 𝑈𝑈1𝑛𝑛
𝑋𝑋𝑡𝑡𝑡𝑡 = Ω and 𝑍𝑍‾𝑡𝑡𝑡𝑡 = 𝑗𝑗𝑋𝑋𝑡𝑡𝑡𝑡
100 𝑆𝑆𝑛𝑛𝑛𝑛

Having the reactance on the secondary side, then it is determined by

2
𝑥𝑥𝑡𝑡 𝑈𝑈2𝑛𝑛
𝑋𝑋𝑡𝑡𝑡𝑡 = Ω and 𝑍𝑍‾𝑡𝑡𝑡𝑡 = 𝑗𝑗𝑋𝑋𝑡𝑡𝑡𝑡
100 𝑆𝑆𝑛𝑛𝑛𝑛

For the line, using the model shown in Figure 4.6, we have

𝑍𝑍‾12 = ℒ(𝑟𝑟 + 𝑗𝑗𝑗𝑗) Ω and 𝑌𝑌‾𝑠𝑠ℎ−12 = 𝑗𝑗𝑏𝑏𝑐𝑐 ℒ S

For the load, 𝑃𝑃 is the consumed three-phase active power with the power factor cos 𝜙𝜙 at the
nominal (or rated) voltage 𝑈𝑈𝑛𝑛 . Thus, the impedance load can be determined by

𝑈𝑈𝑛𝑛2 𝑈𝑈𝑛𝑛2 𝑈𝑈𝑛𝑛2 𝑃𝑃𝐿𝐿𝐿𝐿

𝑍𝑍‾𝐿𝐿𝐿𝐿 = = = (cos 𝜙𝜙 + 𝑗𝑗sin 𝜙𝜙) where 𝑆𝑆𝐿𝐿𝐿𝐿 =
𝑆𝑆‾𝐿𝐿𝐿𝐿
∗ 𝑆𝑆𝐿𝐿𝐿𝐿 (cos 𝜙𝜙 − 𝑗𝑗sin 𝜙𝜙) 𝑆𝑆𝐿𝐿𝐿𝐿 cos 𝜙𝜙

Figure 1.3 shows the single-phase impedance diagram corresponding to the single-line diagram
shown in Figure 1.2.

Figure 1.3. Impedance network of a small power system.

The simple system shown in Figure 1.3. has three different voltage levels (6,10 and 30kV ). The
analysis of the system can be carried out by transferring all impedances to a single voltage level.
This method gives often quite extensive calculations, especially dealing with large systems with

3
several different voltage levels. To overcome this difficulty, the so called per-unit system was
developed, and it will be presented in the next section.

1.3. The Per Unit System

A common method to express voltages, currents, powers and impedances in an electric network
is in per-unit (or percent) of a certain base or reference value. The per-unit value of a certain
quantity is defined as

true value
Per-unit value = (1.1)
base value of the quantity

The per-unit method is very suitable for power systems with several voltage levels and
transformers. In a three-phase system, the per-unit value can be calculated using the
corresponding base quantity. By using the base voltage

𝑈𝑈𝑏𝑏 = base voltage, kV (line-to-line voltage) (1.2)

and a base power,

𝑆𝑆𝑏𝑏 = three-phase base power, MVA (1.3)

the base current

𝑆𝑆𝑏𝑏
𝐼𝐼𝑏𝑏 = = base current /phase, kA (1.4)
√3𝑈𝑈𝑏𝑏

as well as a base impedance

𝑈𝑈𝑏𝑏2 𝑈𝑈𝑏𝑏
𝑍𝑍𝑏𝑏 = = = base impedance, Ω (1.5)
𝑆𝑆𝑏𝑏 √3𝐼𝐼𝑏𝑏

can be calculated. In expressions given above, the units kV and MVA have been assumed, which
imply units in kA and Ω. Of course, different combinations of units can be used, e.g. V, VA, A, Ω
or kV, kVA, A, kΩ.

There are several reasons for using a per-unit system:

• The percentage voltage drop is directly given in the per-unit voltage.

• It is possible to analyze power systems having different voltage levels in a more efficient
way.
• When having different voltage levels, the relative importance of different impedances is
directly given by the per-unit value.
• When having large systems, numerical values of the same magnitude are obtained which
increase the numerical accuracy of the analysis.

4
 • Use of the constant √3 is reduced in three-phase calculations.

Figure 1.4 shows the single-phase impedance diagram of a symmetrical three-phase transformer.
In Figure 1.4 a), the transformer leakage impedance is given on the primary side, and in Figure
1.4 b), the transformer leakage impedance is given on the secondary side. Furthermore, 𝛼𝛼 is the
ratio of rated line-to-line voltages. Thus, based on transformer properties we have

𝑈𝑈1𝑛𝑛 1
= and 𝐼𝐼‾1 𝐼𝐼‾�2 = 𝛼𝛼 (1.6)
𝑈𝑈2𝑛𝑛 𝛼𝛼

Let the base power be 𝑆𝑆𝑏𝑏 . Note that 𝑆𝑆𝑏𝑏 is a global base value, i.e. it is the same in all different
voltages levels. Let also 𝑈𝑈1𝑏𝑏 and 𝑈𝑈2𝑏𝑏 be the base voltages on the primary side and secondary
side, respectively. The base voltages have been chosen such that they have the same ratio as the
ratio of the transformer, i.e.

𝑈𝑈1𝑏𝑏 1
= (1.7)
𝑈𝑈2𝑏𝑏 𝛼𝛼

Furthermore, since 𝑆𝑆𝑏𝑏 = √3𝑈𝑈1𝑏𝑏 𝐼𝐼1𝑏𝑏 = √3𝑈𝑈2𝑏𝑏 𝐼𝐼2𝑏𝑏 , by virtue of equation (1.7) we find that

𝐼𝐼1𝑏𝑏
= 𝛼𝛼 (1.8)
𝐼𝐼2𝑏𝑏

where, 𝐼𝐼1𝑏𝑏 and 𝐼𝐼2𝑏𝑏 are the base currents on the primary side and secondary side, respectively. The
base impedances on both sides are given by

The base impedances on both sides are given by

2 2
𝑈𝑈1𝑏𝑏 𝑈𝑈1𝑏𝑏 𝑈𝑈2𝑏𝑏 𝑈𝑈2𝑏𝑏
𝑍𝑍1𝑏𝑏 = = and 𝑍𝑍2𝑏𝑏 = = (1.9)
𝑆𝑆𝑏𝑏 √3𝐼𝐼1𝑏𝑏 𝑆𝑆𝑏𝑏 √3𝐼𝐼2𝑏𝑏

a) b)

Figure 1.4. single-phase impedance diagram of a symmetrical three-phase transformer.

Now consider the circuit shown in Figure 1.4 a). The voltage equation is given by

5
 𝑈𝑈‾2
𝑈𝑈‾1 = √3𝐼𝐼‾1 𝑍𝑍‾𝑡𝑡𝑡𝑡 + (1.10)
𝛼𝛼
In per-unit (pu), we have

𝑈𝑈‾1 √3𝐼𝐼‾1 𝑍𝑍‾𝑡𝑡𝑡𝑡 𝑈𝑈‾2 𝐼𝐼‾1 𝑍𝑍‾𝑡𝑡𝑡𝑡 𝑈𝑈‾2

= + = + ⇒ 𝑈𝑈‾1𝑝𝑝𝑝𝑝 = 𝐼𝐼‾1𝑝𝑝𝑝𝑝 𝑍𝑍‾𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 + 𝑈𝑈‾2𝑝𝑝𝑝𝑝 (1.11)
𝑈𝑈1𝑏𝑏 √3𝐼𝐼1𝑏𝑏 𝑍𝑍1𝑏𝑏 𝛼𝛼𝑈𝑈1𝑏𝑏 𝐼𝐼1𝑏𝑏 𝑍𝑍1𝑏𝑏 𝑈𝑈2𝑏𝑏

Next, consider the circuit shown in Figure 1.4 b). The voltage equation is given by

𝛼𝛼𝑈𝑈‾1 = √3𝐼𝐼‾2 𝑍𝑍‾𝑡𝑡𝑡𝑡 + 𝑈𝑈‾2 (1.12)

In per-unit (pu), we have

𝛼𝛼𝑈𝑈‾1 𝛼𝛼𝑈𝑈‾1 √3𝐼𝐼‾2 𝑍𝑍‾𝑡𝑡𝑡𝑡 𝑈𝑈‾2 𝐼𝐼‾2 𝑍𝑍‾𝑡𝑡𝑡𝑡 𝑈𝑈‾2

= = + = + ⇒ 𝑈𝑈‾1𝑝𝑝𝑝𝑝 = 𝐼𝐼‾2𝑝𝑝𝑝𝑝 𝑍𝑍‾𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 + 𝑈𝑈‾2𝑝𝑝𝑝𝑝 (1.13)
𝑈𝑈2𝑏𝑏 𝛼𝛼𝑈𝑈1𝑏𝑏 √3𝐼𝐼2𝑏𝑏 𝑍𝑍2𝑏𝑏 𝑈𝑈2𝑏𝑏 𝐼𝐼2𝑏𝑏 𝑍𝑍2𝑏𝑏 𝑈𝑈2𝑏𝑏

By virtue of equations (1.11) and (1.13), we find that

𝐼𝐼‾1𝑝𝑝𝑝𝑝 𝑍𝑍‾𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 = 𝐼𝐼‾2𝑝𝑝𝑝𝑝 𝑍𝑍‾𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡

Furthermore, based on equations (1.6) and (1.8) it can be shown that 𝐼𝐼‾1𝑝𝑝𝑝𝑝 = 𝐼𝐼‾2𝑝𝑝𝑝𝑝 (show that).
Thus,

𝑍𝑍‾𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 = 𝑍𝑍‾𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 (1.14)

Equation (1.14) implies that the per-unit impedance diagram of a transformer is the same
regardless of whether the actual impedance is determined on the primary side or on the
secondary side.

Example 1.1

Assume that a 15MVA transformer has a voltage ratio of 6kV/30kV and a leakage reactance of
8%. Calculate the pu-impedance when the base power of the system is 20MVA and the base
voltage on the 90kV-side is 99kV.

Solution

Based on given data, 𝑆𝑆𝑛𝑛𝑛𝑛 = 15MVA, 𝑈𝑈1𝑛𝑛 /𝑈𝑈2𝑛𝑛 = 6/30, 𝑥𝑥𝑡𝑡 = 8% and 𝑈𝑈2𝑏𝑏 = 33kV. We first
calculate the transformer impedance in ohm on the 30kV-side and after that, the per-unit value.

6
 𝑍𝑍‾% 𝑍𝑍‾% 𝑈𝑈2𝑛𝑛
2
𝑗𝑗8 ⋅ 302
𝑍𝑍‾30𝑘𝑘𝑘𝑘 = 𝑍𝑍 = = = 𝑗𝑗4.8Ω
100 𝑡𝑡𝑡𝑡30 100 𝑆𝑆𝑛𝑛𝑛𝑛 100 ⋅ 15
𝑍𝑍‾Φ𝐸𝐸
𝑍𝑍𝑡𝑡𝑡𝑡30 𝑍𝑍‾30𝑘𝑘𝑘𝑘 𝑍𝑍‾30𝑘𝑘𝑘𝑘 ⋅ 𝑆𝑆𝑏𝑏 𝑗𝑗4.8 ⋅ 20
𝑍𝑍‾𝑡𝑡𝑡𝑡𝑡𝑡 = 100 = = 2 = = 𝑗𝑗0.088pu
𝑍𝑍2𝑏𝑏 𝑍𝑍2𝑏𝑏 𝑈𝑈2𝑏𝑏 332

The given leakage reactance in percent can be considered as the per unit value of reactance based
on the transformer ratings, i.e. 𝑍𝑍𝑡𝑡𝑡𝑡 . To convert this per unit value to the system per unit value, we
may apply the following equation
2
𝑈𝑈𝑏𝑏− given 𝑆𝑆𝑏𝑏− new
𝑍𝑍‾tpu-new = 𝑍𝑍‾tpu-given 2
𝑆𝑆𝑏𝑏− given 𝑈𝑈𝑏𝑏− new

In our case, 𝑍𝑍‾𝑡𝑡𝑡𝑡𝑡𝑡−𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔 = 𝑗𝑗8/100, 𝑈𝑈𝑏𝑏−𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔 = 𝑈𝑈2𝑛𝑛 = 30, 𝑆𝑆𝑏𝑏−𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔 = 𝑆𝑆𝑛𝑛𝑛𝑛 = 15, 𝑆𝑆𝑏𝑏− new = 20,
and 𝑈𝑈𝑏𝑏− new = 33.

Thus,

𝑗𝑗8 302 20
𝑍𝑍‾𝑡𝑡𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛𝑛𝑛 = = 𝑗𝑗0.088pu (1.15)
100 15 332
The pu-value of the reactance can be also determined based on the base values on the primary
side. From equation (1.7), we have

𝑈𝑈1𝑏𝑏 1 6 6 6
= = ⇒ 𝑈𝑈1𝑏𝑏 = 𝑈𝑈2𝑏𝑏 = 33
𝑈𝑈2𝑏𝑏 𝛼𝛼 30 30 30
2
𝑈𝑈1𝑏𝑏 1 2 𝑈𝑈2𝑏𝑏
2
1 2
𝑍𝑍1𝑏𝑏 = =� � = � � 𝑍𝑍2𝑏𝑏
𝑆𝑆𝑏𝑏 𝛼𝛼 𝑆𝑆𝑏𝑏 𝛼𝛼

Thus,

𝑍𝑍‾% 𝑍𝑍‾% 𝑈𝑈1𝑛𝑛

2
𝑗𝑗8 62
𝑍𝑍‾6𝑘𝑘𝑘𝑘 = 𝑍𝑍 = =
100 𝑡𝑡𝑡𝑡6 100 𝑆𝑆𝑛𝑛𝑛𝑛 100 15
𝑍𝑍‾6𝑘𝑘𝑘𝑘 𝑗𝑗8 62 30 2 20 𝑗𝑗8 302 20
𝑍𝑍‾𝑡𝑡𝑡𝑡𝑡𝑡 = = � � = = 𝑗𝑗0.088pu
𝑍𝑍1𝑏𝑏 100 15 6 332 100 15 332

1.3.1. Per Unit Representation of Transmission Lines

Figure 1.5 shows the 𝜋𝜋-equivalent model of a line, where 𝑦𝑦‾𝑠𝑠ℎ−𝑘𝑘𝑘𝑘 = 𝑌𝑌‾𝑠𝑠ℎ−𝑘𝑘𝑘𝑘 /2.

7
Figure 1.5. 𝜋𝜋-equivalent model of a line.

The voltage at bus 𝑘𝑘 in kV is given by

𝑈𝑈‾𝑘𝑘
𝑈𝑈‾𝑘𝑘 = √3𝑍𝑍‾𝑘𝑘𝑘𝑘 𝐼𝐼‾ + 𝑈𝑈‾𝑗𝑗 , where 𝐼𝐼‾ = 𝐼𝐼‾𝑘𝑘 − 𝐼𝐼‾𝑠𝑠ℎ = 𝐼𝐼‾𝑘𝑘 − 𝑦𝑦‾𝑠𝑠ℎ−𝑘𝑘𝑘𝑘
√3

Let 𝑆𝑆𝑏𝑏 , 𝑈𝑈𝑏𝑏 , 𝐼𝐼𝑏𝑏 and 𝑍𝑍𝑏𝑏 be the base values for the line. Note that the base admittance is given by
𝑌𝑌𝑏𝑏 = 1/𝑍𝑍𝑏𝑏 . Then, the above equations i per unit are given by

𝑈𝑈‾𝑘𝑘 √3𝑍𝑍‾𝑘𝑘𝑘𝑘 𝐼𝐼‾ 𝑈𝑈‾𝑗𝑗

= + ⇒ 𝑈𝑈‾𝑘𝑘𝑘𝑘𝑘𝑘 = 𝑍𝑍‾𝑘𝑘𝑘𝑘𝑘𝑘𝑘𝑘 𝐼𝐼‾𝑝𝑝𝑝𝑝 + 𝑈𝑈‾𝑗𝑗𝑗𝑗𝑗𝑗
𝑈𝑈𝑏𝑏 √3𝑍𝑍𝑏𝑏 𝐼𝐼𝑏𝑏 𝑈𝑈𝑏𝑏

Where;

𝐼𝐼‾𝑘𝑘 𝑈𝑈‾𝑘𝑘 √3𝑍𝑍𝑏𝑏 𝑈𝑈‾𝑘𝑘

𝐼𝐼‾𝑝𝑝𝑝𝑝 = − 𝑦𝑦‾𝑠𝑠ℎ−𝑘𝑘𝑘𝑘 = 𝐼𝐼‾𝑘𝑘𝑘𝑘𝑘𝑘 − 𝑦𝑦‾𝑠𝑠ℎ−𝑘𝑘𝑘𝑘 𝑍𝑍𝑏𝑏 = 𝐼𝐼‾𝑘𝑘𝑘𝑘𝑘𝑘 − 𝑦𝑦‾𝑠𝑠ℎ−𝑘𝑘𝑘𝑘𝑘𝑘𝑘𝑘 𝑈𝑈‾𝑘𝑘𝑘𝑘𝑘𝑘
𝐼𝐼𝑏𝑏 √3 𝑈𝑈𝑏𝑏 𝑈𝑈𝑏𝑏

Figure 1.6 shows the per-unit impedance diagram of a transmission line.

Figure 1.6. Per-unit impedance diagram of a transmission line.

1.3.2. System Analysis in Per Unit System

To analysis a three-phase power system, it is more convenient and effective to convert the
physical quantities into the per-unit system as follows:

1. Choose a suitable base power for the system. It should be in the same range as the rated
power of the installed system equipments.
2. Choose a base voltage at one section (or voltage level) of the system. The system is
divided into different sections (or voltage levels) by the transformers.

8
 3. Calculate the base voltages in all sections of the system by using the transformer ratios.
4. Calculate all per-unit values of all system components that are connected.
5. Draw the per-unit impedance diagram of the system.
6. Perform the system analysis (in the per-unit system).
7. Convert the per-unit results back to the physical values.

Example 1.2

Consider the power system shown in Figure 1.7, where a load is fed by a generator via a
transmission line and two transformers. Based on the given system data below, calculate the load
voltage as well as the active power of the load.

Figure 1.7. Single-line diagram of the system for Example 1.2.

System data:

Generator 𝐆𝐆: 𝑈𝑈𝑔𝑔 = 13.8kV,

𝑈𝑈1𝑛𝑛 13.8kV
Transformer T1: 𝑆𝑆𝑛𝑛𝑛𝑛 = 10MVA, = , 𝑋𝑋𝑡𝑡𝑡𝑡 = 1.524Ω (on 13.8kV-side),
𝑈𝑈2𝑛𝑛 69kV

𝑈𝑈1𝑛𝑛 66kV
Transformer T2: 𝑆𝑆𝑛𝑛𝑛𝑛 = 5MVA, = , 𝑥𝑥𝑡𝑡 = 8%,
𝑈𝑈2𝑛𝑛 13.2kV

Line : 𝑥𝑥 = 0.8Ω/km and ℒ = 10 km

Load LD : impedance load, 𝑃𝑃𝐿𝐿𝐿𝐿 = 4MW, 𝑈𝑈𝑛𝑛 = 13.2kV, cos 𝜙𝜙 = 0.8 inductive.

Solution

1. Let the base power be 𝑆𝑆𝑏𝑏 = 10MVA.

2. Let the base voltage at the generator be 𝑈𝑈1𝑏𝑏 = 13.8kV.
3. The transformer ratio gives the base voltage 𝑈𝑈2𝑏𝑏 = 69kV for the line and 𝑈𝑈3𝑏𝑏 = 69 ⋅
13.2/66 = 13.8kV for the load.

In Figure 1.8, the different sections of the system are given.

9
Figure 1.8. Different sections of the system given in Example 1.2.

4. Calculate the per-unit values of the system components.

𝑈𝑈𝑔𝑔 13.8
G: 𝑈𝑈𝑔𝑔𝑔𝑔𝑔𝑔 = 𝑈𝑈1𝑝𝑝𝑝𝑝 = = = 1.0pu
𝑈𝑈1𝑏𝑏 13.8
𝑍𝑍‾𝑡𝑡𝑡𝑡 10
T1: 𝑍𝑍‾𝑡𝑡1𝑝𝑝𝑝𝑝 = = 𝑗𝑗1.524 = 𝑗𝑗0.080pu
𝑍𝑍1𝑏𝑏 13.82

8 13.22 1 8 13.22 10
T2: 𝑍𝑍‾𝑡𝑡2𝑝𝑝𝑝𝑝 = 𝑗𝑗 = 𝑗𝑗 = 0.146pu
100 5 𝑍𝑍3𝑏𝑏 100 5 13.82

ℒ𝑍𝑍 ‾ 10
Line: 𝑍𝑍‾23𝑝𝑝𝑝𝑝 = 23 = 10 ⋅ 𝑗𝑗0.8 = 𝑗𝑗0.017𝑝𝑝𝑝𝑝
𝑍𝑍2𝑏𝑏 692

𝑈𝑈𝑛𝑛2 𝑈𝑈𝑛𝑛2 𝑈𝑈𝑛𝑛2

LD: 𝑍𝑍‾𝐿𝐿𝐿𝐿 = ‾ ∗ = = (cos 𝜙𝜙 + 𝑗𝑗sin 𝜙𝜙) =
𝑆𝑆𝐿𝐿𝐿𝐿 𝑆𝑆𝐿𝐿𝐿𝐿 (cos 𝜙𝜙−𝑗𝑗sin 𝜙𝜙) 𝑆𝑆𝐿𝐿𝐿𝐿

13.22
= (0.8 + 𝑗𝑗0.6) = 27.88 + 𝑗𝑗20.91Ω
4/0.8

𝑍𝑍‾𝐿𝐿𝐿𝐿 10
𝑍𝑍‾𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 = = (27.88 + 𝑗𝑗20.91) ⋅ = 1.464 + 𝑗𝑗1.098𝑝𝑝𝑝𝑝
𝑍𝑍3𝑏𝑏 13.82

5. By using these values, an impedance diagram can be drawn as shown in Figure 1.9.

Figure 1.9. Impedance network in per-unit

6. The current through the network can be calculated as

10
 1 + 𝑗𝑗0
𝐼𝐼‾𝑝𝑝𝑝𝑝 = = 0.3714 − 𝑗𝑗0.3402𝑝𝑝𝑝𝑝 (1.16)
𝑗𝑗0.08 + 𝑗𝑗0.017 + 𝑗𝑗0.146 + 1.464 + 𝑗𝑗1.098

The load voltage is

𝑈𝑈‾4𝑝𝑝𝑝𝑝 = 𝑈𝑈‾𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 = 𝐼𝐼‾𝑝𝑝𝑝𝑝 𝑍𝑍‾𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 = 0.9173 − 𝑗𝑗0.0903𝑝𝑝𝑝𝑝 (1.17)

The load power is

𝑆𝑆‾𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 = 𝑈𝑈‾𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 𝐼𝐼‾𝑝𝑝𝑝𝑝

∗
= 0.3714 + 0.2785𝑝𝑝𝑝𝑝 (1.18)

7. The load voltage and active load power in physical units can be obtained by multiplying
the per-unit values with corresponding base quantities.

𝑈𝑈𝐿𝐿𝐿𝐿 = 𝑈𝑈𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 𝑈𝑈3𝑏𝑏 = �0.91732 + 0.09032 13.8 = 12.72kV (1.19)

𝑃𝑃𝐿𝐿𝐿𝐿 = Real �𝑆𝑆‾𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 �𝑆𝑆𝑏𝑏 = 0.371 ⋅ 10 = 3.71MW (1.20)

Note that the 𝑃𝑃𝐿𝐿𝐿𝐿 given in the system data (i.e. 𝑃𝑃𝐿𝐿𝐿𝐿 = 4MW ) is the consumed active power at
the rated (or nominal) voltage 𝑈𝑈𝑛𝑛 = 13.2kV. However, the actual voltage at bus 4 is 12.72 kV.
Therefore, the actual consumed power is 3.71MW.

11
2 Load Flow Analysis
2.1. Introduction
A power flow study (load-flow study) is a steady-state analysis whose target is to determine the
voltages, currents, and real and reactive power flows in a system under a given load conditions.

The power system is assumed to be operating under balanced condition and can be represented
by a single line diagram. The power system network contains hundreds of buses and branches
with impedances specified in per-unit on a common MVA base. Power flow studies, commonly
referred to as load flow, are essential of power system analysis and design.

The purpose of power flow studies is to plan ahead and account for various hypothetical
situations. For example, if a transmission line is to be taken off line for maintenance, can the
remaining lines in the system handle the required loads without exceeding their rated values.

2.1.1. Importance of Power (load) flow studies

• They are used in planning studies to determine if and when specific elements will become
overloaded.
• Major investment decisions begin with reinforcement strategies based on load-flow analysis.
• In operating studies, load-flow analysis is used to ensure that each generator runs at the
optimum operating point; demand will be met without overloading facilities; and
maintenance plans can proceed without undermining the security of the system.

2.1.2. Load Flow: Objectives and applications

• Optimum system running conditions and load distribution

• Generator scheduling and reactive scheduling to minimize losses
• Optimum rating and tap-range of transformers
• Optimal placement of reactive compensation
• Improvements from change of conductor size and system voltage
• Starting point of other studies such as fault analysis and stability analysis

2.2. Power/Load flow problem

In power system analysis, the load-flow (or power-flow) problem involves determining the
voltages, power flows, and power losses in an electrical power system under a given load
condition

At any bus (or node) in the power system, four main quantities of interest are considered:

• |V|: Magnitude of the bus voltage.

12
 • θ: Phase angle of the bus voltage.
• P: Active (real) power.
• Q: Reactive power.

To solve the load-flow problem, certain quantities must be specified at each bus, and the others
must be calculated based on these specifications. The general rule is:

• If any two of these four quantities are specified, the other two must be calculated.

Specifying more than two quantities at a bus lead to over-specification, resulting in more
unknowns than available equations to solve them.

Because records enable the real and reactive power to be accurately estimated at loads, 𝑃𝑃 and 𝑄𝑄
are specified quantities at loads, which are called 𝑷𝑷𝑷𝑷 buses.

2.2.1. Formulation of power-flow study

Each bus in a power system can be classified as one of three types:

1 Load bus ( 𝑃𝑃 − 𝑄𝑄 bus) - a buss at which the real and reactive power are specified, and for
which the bus voltage will be calculated. All busses having no generators are load busses.
In here, 𝑉𝑉 and 𝛿𝛿 are unknown.
2 Generator bus ( 𝑃𝑃 − 𝑉𝑉 bus) - a bus at which the magnitude of the voltage is defined and is
kept constant by adjusting the field current of a synchronous generator. We also assign
real power generation for each generator according to the economic dispatch. In here, 𝑄𝑄
and 𝛿𝛿 are unknown
3 Slack bus (swing bus) - a special generator bus serving as the reference bus. Its voltage is
assumed to be fixed in both magnitude and phase (for instance, 1∠0∘ pu ). In here, 𝐏𝐏 and
𝐐𝐐 are unknown.

A bus is called Swing bus when the magnitude and phase of the voltage are specified for it.
Swing bus is the reference bus for load flow solution and it is required for accounting line losses.
The slack bus is needed in power systems analysis to account for transmission losses.

The following table summarises the above discussion:

Specified Unknown
Bus type
quantities quantities

Slack bus |𝑉𝑉|, 𝛿𝛿 𝑃𝑃, 𝑄𝑄

Load bus Voltage

𝑃𝑃, 𝑄𝑄 ∣ 𝑉𝑉, 𝛿𝛿
controlled bus

13
Note that the power flow equations are non-linear, thus cannot be solved analytically. A
numerical iterative algorithm is required to solve such equations. A standard procedure follows:

1 Create a bus admittance matrix 𝑌𝑌bus for the power system;

2 Make an initial estimate for the voltages (both magnitude and phase angle) at each bus in
the system;
3 Substitute in the power flow equations and determine the deviations from the solution.
4 Update the estimated voltages based on some commonly known numerical algorithms
(e.g., Newton-Raphson or Gauss-Seidel).
5 Repeat the above process until the deviations from the solution are minimal.

2.2.2. Formulation of the Bus Admittance Matrix

The first step in developing the mathematical model describing the power flow in the network is
the formulation of the bus admittance matrix.

The bus admittance matrix is an 𝑛𝑛×𝑛𝑛 matrix (where 𝑛𝑛n is the number of buses in the system)
constructed from the admittances of the equivalent circuit elements of the segments making up
the power system.

Most system segments are represented by a combination of shunt elements (connected between a
bus and the reference node) and series elements (connected between two system buses).

Formulation of the bus admittance matrix follows two simple rules:

1. The admittance of elements connected between node k and reference is added to the (k, k)
entry of the admittance matrix

2. The admittance of elements connected between nodes j and k is added to the (j, j) and (k, k)
entries of the admittance matrix.

The negative of the admittance is added to the (j, k) and (k, j) entries of the admittance matrix.

Consider a three-bus transmission system below;

14
The line impedances joining buses 1,2 and 3 are denoted by z12, , z23 and z31 respectively.

The corresponding line admittances are y12 , y23 and y31

The total capacitive susceptance at the buses are represented by y10 , y20 and yзо. Applying
Kirchhoff's current law at each bus:

𝐼𝐼1 = 𝑉𝑉1 𝑦𝑦10 + (𝑉𝑉1 − 𝑉𝑉2 )𝑦𝑦12 + (𝑉𝑉1 − 𝑉𝑉3 )𝑦𝑦13 𝐼𝐼1 = (𝑦𝑦10 + 𝑦𝑦12 + 𝑦𝑦13 )𝑉𝑉1 − 𝑦𝑦12 𝑉𝑉2 − 𝑦𝑦13 𝑉𝑉3

𝐼𝐼2 = 𝑉𝑉2 𝑦𝑦20 + (𝑉𝑉2 − 𝑉𝑉1 )𝑦𝑦21 + (𝑉𝑉2 − 𝑉𝑉3 )𝑦𝑦23 𝐼𝐼2 = −𝑦𝑦12 𝑉𝑉1 + (𝑦𝑦20 + 𝑦𝑦12 + 𝑦𝑦23 )𝑉𝑉2 − 𝑦𝑦23 𝑉𝑉3

𝐼𝐼3 = 𝑉𝑉3 𝑦𝑦30 + (𝑉𝑉3 − 𝑉𝑉1 )𝑦𝑦31 + (𝑉𝑉3 − 𝑉𝑉2 )𝑦𝑦32 𝐼𝐼3 = −𝑦𝑦31 𝑉𝑉1 − 𝑦𝑦32 𝑉𝑉2 + (𝑦𝑦30 + 𝑦𝑦31 + 𝑦𝑦32 )𝑉𝑉3

In Matrix form

𝐼𝐼1 𝑦𝑦10 + 𝑦𝑦12 + 𝑦𝑦13 −𝑦𝑦12 −𝑦𝑦13 𝑉𝑉1

𝐼𝐼
� 2� = � −𝑦𝑦12 𝑦𝑦20 + 𝑦𝑦12 + 𝑦𝑦23 −𝑦𝑦13 𝑉𝑉
� × � 2�
𝐼𝐼3 −𝑦𝑦13 −𝑦𝑦23 𝑦𝑦30 + 𝑦𝑦13 + 𝑦𝑦23 𝑉𝑉3

Where;

𝑌𝑌11 = 𝑦𝑦10 + 𝑦𝑦12 + 𝑦𝑦13

𝑌𝑌22 = 𝑦𝑦20 + 𝑦𝑦12 + 𝑦𝑦23
𝑌𝑌33 = 𝑦𝑦30 + 𝑦𝑦13 + 𝑦𝑦23

We get

𝐼𝐼1 𝑌𝑌11 𝑌𝑌12 𝑌𝑌13 𝑉𝑉1

�𝐼𝐼2 � = �𝑌𝑌21 𝑌𝑌22 𝑌𝑌23 � �𝑉𝑉2 �
𝐼𝐼3 𝑌𝑌31 𝑌𝑌32 𝑌𝑌33 𝑉𝑉3
15
Note

𝑌𝑌12 = 𝑌𝑌21 = −𝑦𝑦12

𝑌𝑌13 = 𝑌𝑌31 = −𝑌𝑌13

𝑌𝑌23 = 𝑌𝑌32 = −𝑦𝑦23

And

Ibus = Ybus Vbus

Example 2.1

Solution

Convert the impedances to admittances.

1 1
𝑦𝑦𝑖𝑖𝑖𝑖 = =
𝑧𝑧𝑖𝑖𝑖𝑖 𝑟𝑟𝑖𝑖𝑖𝑖 + 𝑗𝑗𝑥𝑥𝑖𝑖𝑖𝑖

16
Applying KCL to the independent nodes 1 through 4 results in

𝐼𝐼1 = 𝑦𝑦10 𝑉𝑉1 + 𝑦𝑦12 (𝑉𝑉1 − 𝑉𝑉2 ) + 𝑦𝑦13 (𝑉𝑉1 − 𝑉𝑉3 )

𝐼𝐼2 = 𝑦𝑦20 𝑉𝑉2 + 𝑦𝑦12 (𝑉𝑉2 − 𝑉𝑉1 ) + 𝑦𝑦23 (𝑉𝑉2 − 𝑉𝑉3 )
0 = 𝑦𝑦23 (𝑉𝑉3 − 𝑉𝑉2 ) + 𝑦𝑦13 (𝑉𝑉3 − 𝑉𝑉1 ) + 𝑦𝑦34 (𝑉𝑉3 − 𝑉𝑉4 )
0 = 𝑦𝑦34 (𝑉𝑉4 − 𝑉𝑉3 )

Rearranging these equations yields

17
 𝐼𝐼1 = (𝑦𝑦10 + 𝑦𝑦12 + 𝑦𝑦13 )𝑉𝑉1 − 𝑦𝑦12 𝑉𝑉2 − 𝑦𝑦13 𝑉𝑉3
𝐼𝐼2 = −𝑦𝑦12 𝑉𝑉1 + (𝑦𝑦20 + 𝑦𝑦12 + 𝑦𝑦23 )𝑉𝑉2 − 𝑦𝑦23 𝑉𝑉3
0 = −𝑦𝑦13 𝑉𝑉1 − 𝑦𝑦23 𝑉𝑉2 + (𝑦𝑦13 + 𝑦𝑦23 + 𝑦𝑦34 )𝑉𝑉3 − 𝑦𝑦34 𝑉𝑉4
0 = −𝑦𝑦34 𝑉𝑉3 + 𝑦𝑦34 𝑉𝑉4

Matrix Formation of the Equations

𝐼𝐼1 (𝑦𝑦 + 𝑦𝑦12 + 𝑦𝑦13 ) −𝑦𝑦12 −𝑦𝑦13 0 𝑉𝑉1

⎡ 10 ⎤
𝐼𝐼 −𝑦𝑦21 (𝑦𝑦20 + 𝑦𝑦21 + 𝑦𝑦23 ) −𝑦𝑦23 0 ⎥ 𝑉𝑉2
� 2� = ⎢ ⋅� �
0 ⎢ −𝑦𝑦31 −𝑦𝑦32 (𝑦𝑦31 + 𝑦𝑦32 + 𝑦𝑦34 ) −𝑦𝑦34 ⎥ 𝑉𝑉3
0 ⎣ 0 0 −𝑦𝑦43 𝑦𝑦43 ⎦ 𝑉𝑉4

Completed Matrix Equation

𝑌𝑌11 = (𝑦𝑦10 + 𝑦𝑦12 + 𝑦𝑦13 ) = −𝑗𝑗8.50 𝑌𝑌23 = 𝑌𝑌32 = −𝑦𝑦23 = 𝑗𝑗5.00

𝑌𝑌12 = 𝑌𝑌21 = −𝑦𝑦12 = 𝑗𝑗2.50 𝑌𝑌33 = (𝑦𝑦31 + 𝑦𝑦32 + 𝑦𝑦34 ) = −𝑗𝑗22.50
𝑌𝑌13 = 𝑌𝑌31 = −𝑦𝑦13 = 𝑗𝑗5.00 𝑌𝑌34 = 𝑌𝑌43 = −𝑦𝑦34 = 𝑗𝑗12.50
𝑌𝑌22 = (𝑦𝑦20 + 𝑦𝑦21 + 𝑦𝑦23 ) = −𝑗𝑗8.75 𝑌𝑌44 = 𝑦𝑦34 = −𝑗𝑗12.50

𝐼𝐼1 −𝑗𝑗8.50 𝑗𝑗2.50 𝑗𝑗5.00 0 𝑉𝑉1

𝐼𝐼 𝑗𝑗2.50 −𝑗𝑗8.75 𝑗𝑗5.00 0 𝑉𝑉
� 2� = � � ⋅ � 2�
0 𝑗𝑗5.00 𝑗𝑗5.00 −𝑗𝑗22.50 𝑗𝑗12.50 𝑉𝑉3
0 0 0 𝑗𝑗12.50 −𝑗𝑗12.50 𝑉𝑉4

In the above network, since there is no connection between bus 1 and 4, Y14 = Y41 = 0; similarly
Y24 = Y42 = 0

18
 • Extending the above relation to an n bus system, the node-voltage equation in matrix
form is

𝐼𝐼1 𝑌𝑌11 𝑌𝑌12 ⋯ 𝑌𝑌1𝑖𝑖 ⋯ 𝑌𝑌1𝑛𝑛 𝑉𝑉1

⎡ 𝐼𝐼 ⎤ ⎡𝑌𝑌 𝑌𝑌22 ⋯ 𝑌𝑌2𝑖𝑖 ⋯ 𝑌𝑌2𝑛𝑛 ⎤ ⎡𝑉𝑉2 ⎤
⎢ 2 ⎥ ⎢ 21 ⎥⎢ ⎥
⎢⋮⎥=⎢ ⋮ ⋮ ⋮ ⋮ ⎥⎢ ⋮ ⎥
⎢ 𝐼𝐼𝑖𝑖 ⎥ ⎢ 𝑌𝑌𝑖𝑖1 𝑌𝑌𝑖𝑖2 ⋯ 𝑌𝑌𝑖𝑖𝑖𝑖 ⋯ 𝑌𝑌in ⎥ ⎢ 𝑉𝑉𝑖𝑖 ⎥
⎢⋮⎥ ⎢ ⋮ ⋮ ⋮ ⋮ ⎥⎢ ⋮ ⎥
⎣𝐼𝐼𝑛𝑛 ⎦ ⎣𝑌𝑌𝑛𝑛1 𝑌𝑌𝑛𝑛2 ⋯ 𝑌𝑌ni ⋯ 𝑌𝑌𝑛𝑛𝑛𝑛 ⎦ ⎣𝑉𝑉𝑛𝑛 ⎦

Ibus = Ybus Vbus

The admittance matrix for the network is obtained by inspection is as follows

−𝑗𝑗8.50 𝑗𝑗2.50 𝑗𝑗5.00 0

𝑗𝑗2.50 −𝑗𝑗8.75 𝑗𝑗5.00 0
𝑌𝑌bus =� �
𝑗𝑗5.00 𝑗𝑗5.00 −𝑗𝑗22.50 𝑗𝑗12.50
0 0 𝑗𝑗12.50 −𝑗𝑗12.50

2.2.3. Y-Bus Matrix Building Rules

• Square matrix with dimensions equal to the number of buses

• Convert all network impedances into admittances
• Diagonal elements:
𝑛𝑛

𝑌𝑌𝑖𝑖𝑖𝑖 = � 𝑦𝑦𝑖𝑖𝑖𝑖 𝑗𝑗 ≠ 𝑖𝑖
𝑗𝑗=0

• Off-diagonal elements:
𝑌𝑌𝑖𝑖𝑖𝑖 = 𝑌𝑌𝑗𝑗𝑗𝑗 = −𝑦𝑦𝑖𝑖𝑖𝑖

• Matrix is symmetrical along the leading diagonal

19
2.2.4. Exercise

1. Explain the importance of load flow studies.

2. Discuss briefly the bus classification.
3. What is the need for a slack bus or reference bus? Explain.
4. Generate the bus admittance matrix for the following system

5. Obtain the bus admittance matrix for the systems below

20
Assignment
1. Find out the Y matrix of the sample power system as shown in Fig. 7.3. Data for this system
are given in Table

Fig. 3 bus sample power system.

Table: Per unit impedances and line charging for sample power system shown in Fig.

Bus code Impedance Line charging

′
𝑖𝑖 − 𝑘𝑘 𝑍𝑍ik 𝑦𝑦ik/2

1−2 0.02 + 𝑗𝑗0.06 𝑗𝑗0.03

1−3 0.08 + 𝑗𝑗0.24 𝑗𝑗0.025

2−3 0.06 + 𝑗𝑗0.18 𝑗𝑗0.020

2. Evaluate the admittance matrix from the following circuit diagram.

21
3. A power system network is shown in Figure below. The generators at buses 1 and 2 are
represented by their equivalent current sources with their reactance in per unit on a 110-MVA
base. The lines are represented by π model where series reactance and shunt reactance are also
expressed in per unit on a 110 MVA base. The loads at buses 3 and 4 are expressed in MW
and Mvar. Assuming a voltage magnitude of 1.0 per unit at buses 3 and 4, convert the loads to
per unit impedances. Convert network impedances to admittances and obtain the bus
admittance matrix by inspection.

22
2.3. Numerical Solution Techniques for Non-Linear Algebraic Equations
Power flow (PF) equations are non-linear and cannot be solved analytically. Therefore,
numerical iterative algorithms are employed to solve them. A standard procedure is as follows:
1. Create a bus admittance matrix for the power system.
2. Make an initial estimate for the voltages at each bus in the system.
3. Substitute in the PF equations and determine the deviations from the solution.
4. Update the estimated voltages based on commonly known numerical algorithms (e.g.,
Newton-Raphson or Gauss-Seidel).
5. Repeat the above steps until the deviations from the solution are minimal.

2.3.1. Gauss-Seidel Method

• A non-linear algebraic equation solver

• method of successive displacements

• iterative steps:

• take a function and rearrange it into the form 𝑥𝑥 = g(𝑥𝑥) {there are several possible
arrangements}

• make an initial estimate of the variable 𝑥𝑥: 𝑥𝑥 [0] = initial value

• find an iterative improvement of 𝑥𝑥 [𝑘𝑘] , that is: 𝑥𝑥 [𝑘𝑘+1] = g�𝑥𝑥 [𝑘𝑘] �

• a solution is reached when the difference between two iterations is less than a specified
accuracy: �𝑥𝑥 [𝑘𝑘+1] − 𝑥𝑥 [𝑘𝑘] � ≤ 𝜀𝜀

Gauss-Seidel Example
• Find the root of the equation: 𝑓𝑓(𝑥𝑥) = 𝑥𝑥 3 − 6𝑥𝑥 2 + 9𝑥𝑥 − 4 = 0
• Step 1. Cast the equation into the 𝑔𝑔(𝑥𝑥) form.
9𝑥𝑥 = −𝑥𝑥 3 + 6𝑥𝑥 2 + 4
1 6 4
𝑥𝑥 = − 𝑥𝑥 3 + 𝑥𝑥 2 + = 𝑔𝑔(𝑥𝑥)
9 9 9

• Step 2. Starting with an initial guess of x[0] = 2, several iterations are performed.

23
 1 6 4
𝑥𝑥 [1] = 𝑔𝑔�𝑥𝑥 [0] = 2� = − (2)3 + (2)2 + = 2.2222
9 9 9
1 6 4
𝑥𝑥 [2] = 𝑔𝑔�𝑥𝑥 [1] = 2.2222� = − (2.2222)3 + (2.2222)2 + = 2.5173
9 9 9
1 6 4
𝑥𝑥 [3] = 𝑔𝑔�𝑥𝑥 [2] = 2.5173� = − (2.5173)3 + (2.5173)2 + = 2.8966
9 9 9
𝑥𝑥 [4] = 3.3376
𝑥𝑥 [5] = 3.7398
𝑥𝑥 [6] = 3.9568
𝑥𝑥 [7] = 3.9988
𝑥𝑥 [8] = 4.0000

If initial estimate x is within convergent region, solution will converge in zigzag fashion to one
of the roots

If initial estimate x is outside convergent region, process will diverge, no solution found

• In some cases, an acceleration factor 𝛼𝛼 is added to improve the rate of convergence

• Acceleration factor should not be too large to produce overshoot

• Find the root of the equation: 𝑓𝑓(𝑥𝑥) = 𝑥𝑥 3 − 6𝑥𝑥 2 + 9𝑥𝑥 − 4 = 0 with an acceleration factor
of 1.25

• Additional iterations

• Starting with an initial guess of 𝑥𝑥 [0] = 2.

24
 𝑥𝑥 [0] = 2
1 6 4
𝑔𝑔(2) = − (2)3 + (2)2 + = 2.2222
9 9 9
𝑥𝑥 [1] = 2 + 1.25[2.2222 − 2] = 2.2778
1 6 4
𝑔𝑔(2.2778) = − (2.2778)3 + (2.2778)2 + = 2.5902
9 9 9
[2]
𝑥𝑥 = 2.2778 + 1.25[2.5902 − 2.2778] = 2.6683

𝑥𝑥 [3] = 3.0801
𝑥𝑥 [4] = 3.1831
𝑥𝑥 [5] = 3.7238
𝑥𝑥 [6] = 4.0084
𝑥𝑥 [7] = 3.9978
𝑥𝑥 [8] = 4.0005
Gauss-Seidel Advantages
• Each iteration is relatively fast (computational order is proportional to number of
branches & number of buses in the system)

• Relatively easy to program

Disadvantages
• Tends to converge relatively slowly, although this can be improved with acceleration

• Has to miss solutions particularly on large systems

• Tends to diverge on cases with negative branch reactance (common with compensated
lines)

• Need to program using complex numbers

2.3.2. Newton-Raphson PF Solution

• Quadratic convergence

• mathematically superior to Guass-Seidel method

• More efficient for large networks

• number of iterations required for solution is independent of system size

• The Newton-Raphson equations are cast in natural power system form

• solving for voltage magnitude and angle, given real and reactive power injections

25
Newton-Raphson Method
• A method of successive approximation using Taylor's expansion
• Consider the function: 𝑓𝑓(𝑥𝑥) = 𝑐𝑐, where 𝑥𝑥 is unknown
• Let 𝑥𝑥 [0] be an initial estimate, then Δ𝑥𝑥 [0] is a small deviation from the correct solution
𝑓𝑓�𝑥𝑥 [0] + Δ𝑥𝑥 [0] � = 𝑐𝑐

• Expand the left-hand side into a Taylor's series about 𝑥𝑥 [0] yeilds

[0]
𝑑𝑑𝑑𝑑 1 𝑑𝑑2 𝑓𝑓 2
𝑓𝑓�𝑥𝑥 � + � � Δ𝑥𝑥 + � 2 � �Δ𝑥𝑥 [0] � + ⋯ = 𝑐𝑐
[0]
𝑑𝑑𝑑𝑑 2 𝑑𝑑𝑥𝑥

• Assuming the error, Δ𝑥𝑥 [0] , is small, the higher-order terms are neglected, resulting in
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑
𝑓𝑓�𝑥𝑥 [0] � + � � Δ𝑥𝑥 [0] ≈ 𝑐𝑐 ⇒ Δ𝑐𝑐 [0] ≈ � � Δ𝑥𝑥 [0]
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑

• Where;
Δ𝑐𝑐 [0] = 𝑐𝑐 − 𝑓𝑓�𝑥𝑥 [0] �

• rearranging the equations

Δ𝑐𝑐 [0]
Δ𝑥𝑥 [0] =
𝑑𝑑𝑑𝑑
� �
𝑑𝑑𝑑𝑑
𝑥𝑥 [1] = 𝑥𝑥 + Δ𝑥𝑥 [0]
[0]

Example
• Find the root of the equation: 𝑓𝑓(𝑥𝑥) = 𝑥𝑥 3 − 6𝑥𝑥 2 + 9𝑥𝑥 − 4 = 0
The analytical solution given by the Newton-Raphson algorithm is

26
 𝑑𝑑𝑑𝑑(𝑥𝑥)
= 3𝑥𝑥 2 − 12𝑥𝑥 + 9
𝑑𝑑𝑑𝑑
Δ𝑐𝑐 (0) = 𝑐𝑐 − 𝑓𝑓�𝑥𝑥 (0) � = 0 − [(6)3 − 6(6)2 + 9(6) − 4] = −50
𝑑𝑑𝑑𝑑 (0)
� � = 3(6)2 − 12(6) + 9 = 45
𝑑𝑑𝑑𝑑
Δ𝑐𝑐 (0) −50
Δ𝑥𝑥 (0) = (0)
= = −1.1111
𝑑𝑑𝑑𝑑 45
� �
𝑑𝑑𝑑𝑑
Therefore, the result at the end of the first iteration is
𝑥𝑥 (1) = 𝑥𝑥 (0) + Δ𝑥𝑥 (0) = 6 − 1.1111 = 4.8889
The subsequent iterations result in
13.4431
𝑥𝑥 (2) = 𝑥𝑥 (1) + Δ𝑥𝑥 (1) = 4.8889 − = 4.2789
22.037
2.9981
𝑥𝑥 (3) = 𝑥𝑥 (2) + Δ𝑥𝑥 (2) = 4.2789 − = 4.0405
12.5797
0.3748
𝑥𝑥 (4) = 𝑥𝑥 (3) + Δ𝑥𝑥 (3) = 4.0405 − = 4.0011
9.4914
0.0095
𝑥𝑥 (5) = 𝑥𝑥 (4) + Δ𝑥𝑥 (4) = 4.0011 − = 4.0000
9.0126

Newton Raphson method for solving 𝑛𝑛 variables

(0) (0) (0)
∂𝑓𝑓1 (0) (0) ∂𝑓𝑓1 (0) (0) ∂𝑓𝑓1 (0) (0)
𝑓𝑓1 �𝑥𝑥 + Δ𝑥𝑥 � = 𝑓𝑓1 �𝑥𝑥 � + � � Δ𝑥𝑥1 + � � Δ𝑥𝑥2 + ⋯ + � � Δ𝑥𝑥𝑛𝑛 = 𝑐𝑐1
∂𝑥𝑥1 ∂𝑥𝑥2 ∂𝑥𝑥𝑛𝑛
∂𝑓𝑓2 (0) (0) ∂𝑓𝑓2 (0) (0) ∂𝑓𝑓2 (0) (0)
𝑓𝑓2 �𝑥𝑥 (0) + Δ𝑥𝑥 (0) � = 𝑓𝑓2 �𝑥𝑥 (0) � + � � Δ𝑥𝑥1 + � � Δ𝑥𝑥2 + ⋯ + � � Δ𝑥𝑥𝑛𝑛 = 𝑐𝑐2
∂𝑥𝑥1 ∂𝑥𝑥2 ∂𝑥𝑥𝑛𝑛
(0) (0) (0)
∂𝑓𝑓𝑛𝑛 (0) (0) ∂𝑓𝑓𝑛𝑛 (0) (0) ∂𝑓𝑓𝑛𝑛 (0) (0)
𝑓𝑓𝑛𝑛 �𝑥𝑥 + Δ𝑥𝑥 � = 𝑓𝑓𝑛𝑛 �𝑥𝑥 � + � � Δ𝑥𝑥1 + � � Δ𝑥𝑥2 + ⋯ + � � Δ𝑥𝑥𝑛𝑛 = 𝑐𝑐𝑛𝑛
∂𝑥𝑥1 ∂𝑥𝑥2 ∂𝑥𝑥𝑛𝑛
Rearrange in matrix form
(0)
⎡� ∂𝑓𝑓1 � ∂𝑓𝑓1 (0) ∂𝑓𝑓1 (0) ⎤
� � ⋯ � �
(0) ⎢ ∂𝑥𝑥1 ∂𝑥𝑥2 ∂𝑥𝑥𝑛𝑛 ⎥ Δ𝑥𝑥 (0)
⎡ 𝑐𝑐1 − 𝑓𝑓1 ⎤ ⎢ (0) (0) (0) ⎥ ⎡ 1 ⎤
⎢ 𝑐𝑐2 − 𝑓𝑓 (0) ⎥ ⎢� ∂𝑓𝑓2 � �
∂𝑓𝑓2
� ⋯ �
∂𝑓𝑓2
� ⎥ Δ𝑥𝑥2 ⎥
⎢ (0)
⎢ 2
⎥ = ∂𝑥𝑥1 ∂𝑥𝑥2 ∂𝑥𝑥𝑛𝑛 ⎢ ⎥
⎢ ⋮ ⎥ ⎢ ⎥⎢ ⋮ ⎥
(0) ⋮ ⋮ ⋮ ⋮ (0)
⎣𝑐𝑐𝑛𝑛 − 𝑓𝑓𝑛𝑛 ⎦ ⎢ (0) (0) (0) ⎥ ⎣Δ𝑥𝑥𝑛𝑛 ⎦
⎢� ∂𝑓𝑓𝑛𝑛 � �
∂𝑓𝑓𝑛𝑛
� ⋯ �
∂𝑓𝑓𝑛𝑛
� ⎥
⎣ ∂𝑥𝑥1 ∂𝑥𝑥2 ∂𝑥𝑥𝑛𝑛 ⎦
The matrix can be written as
- Δ𝐶𝐶 (𝑘𝑘) = 𝐽𝐽(𝑘𝑘) Δ𝑋𝑋 (𝑘𝑘)

27
28