Dire-Dawa University, Institute of Technology

School of Electrical and Computer Engineering

Power stream
Course Name: Computer Application in Power System

Prepared by Chalachew M.
 OUT-LINE

 Introduction
 Network modelling
 Basic nodal method
 Power system matrix
 Fast decouple power flow

2
 Lecture 2: Load flow analysis
2.1. Introduction
 Under normal conditions electrical transmission systems
operate in their steady-state mode and the basic calculation
required to determine the characteristics of this state is termed
load flow (or power flow).
 The object of load-flow calculations is to determine the
steady-state operating characteristics of the power
generation/transmission system for a given set of busbar loads.
 The solution is expected to provide information of voltage
magnitudes and angles, active and reactive power flows in the
individual transmission units, losses and the
reactive power generated or absorbed at voltage-controlled
buses
3
 Cont’d

For load flow analysis methods five main properties are

required.
 High computational speed.
 Low computer storage.
 Reliability of solution.
 Versatility and
 Simplicity

4
 2.2 NETWORK MODELLING
Transmission plant components are modeled by their
equivalent circuits in terms of inductance, capacitance and
resistance.
 Each unit constitutes an electric network in its own right
and their interconnection constitutes the transmission
system.
Among the many alternative ways of describing
transmission systems to comply with Kirchhoff's laws mesh
and nodal analysis-are normally used.
But, nodal analysis is suitable for digital computer work
and for routine network calculations. WHY? Because the
nodal approach has the following advantages.

5
 Cont’d

i. The numbering of nodes, performed directly from a

system diagram, is very simple.
ii. Data preparation is easy.
iii. The number of variables and equations is usually less
than with the mesh method for power networks.
iv. Network crossover branches present no difficulty.
v. Parallel branches do not increase the number of
variables or equations.
vi. Node voltages are available directly from the solution,
and branch currents are easily calculated.
vii. Off-nominal transformer taps can easily be represented

6
 2.2.1. Transmission line
• The transmission lines are generally of medium length
or of long length.
• The total resistance and inductive reactance of the line is
included in the series arm of the equivalent-pi and the
total capacitance to neutral is divided between its shunt
arms.
• Figure as shown below equivalent model of the
transmission line.

7
 2.2.2 transformer model

A. Transformer on Nominal Ratio:

From figure 2.1 we can develop

equation 2.0 in matrix form.
Where ysc and yoc are short
and open circuit admittance
respectively.
8
 B. Off-nominal Transformer Tap Settings

A transformer with turns ratio a interconnecting two

nodes i, k can be represented by an ideal transformer in
series with the nominal transformer leakage admittance
as shown in Fig. 2.2(a).
 If the transformer is on nominal tap (a= 1), the nodal
equations for the network branch in the per unit system are

In this case Iik = - I ki.

But, for an off-nominal tap setting and letting the voltage on
the k side of the ideal transformer be Vt we can write,
9
 Cont’d

Eliminating Vt between equations 2.3 and 2.4 we obtain,

A simple equivalent-pi circuit can be deduced from

equations (2.6) and (2.7) This circuit is illustrated in Fig. 2.2
(b) as shown next page.

10
 Cont’d

2.3. BASIC NODAL METHOD

o In the nodal method as applied to power system
networks, the variables are the complex node (bus-bar)
voltages and currents, for which some reference must be
designated.

11
 Cont’d

o In fact, two different references are normally chosen: for

voltage magnitudes the reference is ground, and for voltage
angles the reference is chosen as one of the bus-bar voltage
angles, which is fixed at the value zero (usually).
o In the nodal method, it is convenient to use branch
admittances rather than impedances.
o Denoting the voltages of nodes k and i as Ek and Ei
respectively, and
o the admittance of the branch between them as Yki then
the current flowing in this branch from node k to node i is
given by,

12
 Cont’d

o Let the nodes in the network be numbered 0, 1, ... , n,

where 0 designates the reference node (ground).
o By Kirchhoff's current law, the injected current Ik must
be equal to the sum of the currents leaving node k, hence

13
 2.4. Power system matrix

o If 2.10 equation is written for all the nodes except the

reference, i.e. for all bus-bar in the case of a power system
network, then a complete set of equations defining the
network is obtained in matrix form as,

14
 Cont’d
In shorthand matrix notation, equation (2.11) is simply .

Y-admittance matrix has the following properties.

 Square of order n x n.
 Symmetrical, since Yki = Yik
 Complex.
 Each off-diagonal element Yki is the negative of the
branch admittance between nodes k and i, and is frequently
of value zero.
 Each diagonal element Ykk is the sum of the admittance
of the branches which terminate on node k, including
branches to ground.
15
 2.4.1. Classification of Power System Buses

The complete definition of power flow requires knowledge

of four variables at each bus k in the system. These are
 Pk-real or active power
 Qk- reactive power
 Vk-voltage magnitude
ծk- voltage phase angle.
 Out of the four variables only two are known a priori to
solve the problem, and the aim of the load flow is to solve
the remaining two variables at a bus k.
 In power system there are three types of buses
corresponding to the known variables.

16
Cont’d

17
 2.4.2. Power Flow Iterative Methods

 The basic load flow equations represent a set of

simultaneous, non-linear, algebraic equations as such no
closed form, analytical solution for these equations exist.
 Hence, these equations can only be solved by using
suitable numerical iterative techniques. For solving the
load flow problem, various iterative methods are:
A. Gauss-Seidel LF technique
B. Newton-Raphson LF technique
C. Fast-Decoupled LF technique

18
 A. Gauss-Seidel LF technique
The GS method applied to power flow problem is discussed
in two phases as shown below.
Case (a): Systems with PQ buses only
 Initially assume all buses to be PQ type buses, except the
slack bus. This means that (n–1) complex bus voltages have
to be determined.
 For ease of programming, the slack bus is generally
numbered as bus-1and PQ buses are ordered next in
sequence.
Algorithm for GS method

19
Cont’d

20
Cont’d

21
Cont’d

22
Cont’d

23
 Case (b): Systems with PV buses also present:

At PV buses, the magnitude of voltage and active power are

specified but the reactive power is not specified. Hence it is
needed to first make an estimate of

24
Cont’d

25
 B. Newton-Raphson LF technique
Newton-Raphson (N-R) method is an iterative method
which approximates set of non-linear simultaneous
equations to a set of linear simultaneous equations using
Taylor’s expansion and the terms are limited to first
expansion.

26
 Cont’d
Like that of single equation we develop the update formula
for the case where we have n equations and n unknowns.
We call this the multidimensional case.
Let there is n equations in n unknown variables x1,x2, x3,
……,xn as given below,

27
 Cont’d

To solve the above equations, first we take an initial of the

solution and let the vector of initial guesses be denoted as
x(0).
The first order Taylor’s series expansion is carried out for
these equation around the initial guess of the solution as
follows:

28
 Cont’d
Equations 2.22 can be written in matrix form as,

In equation 2.23, the matrix containing the partial

derivative terms is known as the Jacobin matrix (J). It is an
n x n square matrix.
By rearranging and simplifying the above equation yields
29
 Cont’d

Equation 2.24 is the basic equation for solving the ‘n’

algebraic equations.
Application of NR to Power Flow Solution
To apply the N-R method to solution of load flow equation
we may choose to express bus voltage and line admittance
in polar or rectangular form. For this we choose polar form
and rewriting the power flow equations as shown in
equation 2.25.
30
 Cont’d

The Jacobian matrix of 2.23 has the form

31
 Cont’d

Table 2.0 n k
Elements of the
Jacobian matrix

32
 Cont’d

K, n = 2,3,4,5………N

33
 Cont’d

We now apply to the power-flow problem the four

Newton–Raphson steps outlined in Section 2.25, starting
with at the ith iteration.

STEP 1: Use 2.25 to compute

STEP 2: Use the equations in Table 2.0 to calculate the

Jacobian matrix.

34
 Cont’d

STEP 3: Use Gauss elimination and back substitution to

solve

STEP 4: Compute

35
 Cont’d

Starting with initial value x(0), the procedure continues

until convergence is obtained or until the number of
iterations exceeds a specified maximum.
 Convergence criteria are often based on ծy(i) (called
power mismatches) rather than on ծx(i) (phase angle and
voltage magnitude mismatches).
N.B.
For each voltage-controlled bus, the magnitude Vk is
already known and the function Qk(x) is not needed.
 Therefore, we could omit Vk from the x vector and Qk
from the y vector.

36
 Cont’d
We could also omit from the Jacobian matrix the column
corresponding to partial derivatives with respect to Vk and
the row corresponding to partial derivatives of Qk(x).
 Alternatively, rows and corresponding columns for
voltage-controlled buses can be retained in the Jacobian
matrix.
 Then during each iteration, the voltage magnitude
Vk(i+1) of each voltage-controlled bus is reset to Vk, which
is input data for that bus.
 At the end of each iteration, we compute Qk(x) from
(2.25) and Qgk = Qk(x) + Qlk for each voltage-controlled
bus.
37
 Cont’d

 If the computed value of Qgk exceeds its limits, then the

bus type is changed to a load bus with Qgk set to
its limit value. The power-flow program also computes a
new value for Vk.

Simplified versions of Newton’s method for fast

computation
B.1. Decoupled Load Flow
 The change in real power is primarily governed by the
changes in the voltage angles, but not in voltage
magnitudes.
 On the other hand, the changes in the reactive power
38
 Cont’d
are primarily influenced by the changes in voltage
magnitudes, but not in the voltage angles. Hence in Eq.
(2.26), we can set the matrix J2 and J3 =0. then Eq. 2.26
reduce to,

From equation 2.26 the elements of jacobian matrixes are

defied as,

39
 Cont’d

B.2. Fast Decoupled Load Flow

In decoupled load flow, in spite of the memory savings, the
elements of the matrices J1 and J4 have to be calculated at
each iteration. With some engineering judgment, we can do
some approximations to make these matrices are constant.
These approximations are,
A. Under normal steady state operation, the voltage
magnitudes are all nearly equal to 1.0.

40
 Cont’d

B. As the transmission lines are mostly reactive, the

conductance's are quite small as compared to the
susceptance (Gij << Bij). Yij = Gij + Bij ; → lossless line (G
is neglected)
C. Under normal steady state operation the angular
differences among the bus voltages are quite small (θi − θj ≈
0 (within 50 − 100)).→cos θi − θj ≈ 1 & sin (θi − θj) ≈ (θi −
θj) rad ≈ 0.
D. The injected reactive power at any bus is always much
less than the reactive power consumed by the elements
connected to this bus when these elements are shorted to the
ground (Qi << BiiVi2).
41
 Cont’d
Based on the above assumptions, equation 2.31 and 2.32
can rewrite as follows.

42
 Cont’d

 Where, B’ is (n-1)*(n-1) constant matrix.

 The elements of matrix B is the negative of the imaginary
part of the element (i,k) of the Ybus matrix
Where i&k = 2,3,4….. n

43
 Cont’d

 Matrix B’’ is a constant matrix of (n-m)*(n-m)

 The elements of matrix B’’ is the negative of the
imaginary part of the the element (i,k) of the Y bus matrix
Where, i&k = m+1, m+2,…..n.
Due to, the matrix B’ and B’’ are constant, the inverse
matrix can be stored once and used in every iteration.

44
TECHNIQUES WHICH MAKE THE NEWTON-
RAPHSON METHOD COMPETITIVE IN LOAD FLOW
oA conventional matrix techniques were to be used for n
buses system, the storage and computing time is Directly
proportional to n2 and n3 respectively would be prohibitive.
o But, A typical power system has an average of fewer than
three lines connected to each bus.
o As such, each row of Ybus has an average of fewer than
four nonzero elements.
o Such a matrix, which has only a few nonzero elements, is
said to be sparse.
o Therefore, Newton–Raphson power-flow programs
employ sparse matrix techniques to reduce computer
storage and time requirements
45
 Cont’d
Among the techniques some are listed as follows.
I. Compact storage of Y-bus and J(i),
II. Reordering of buses to avoid fill-in of J(i) during Gauss
elimination steps and
III. Matrix (Triangular) factorization.

I. Compact storage of Y-bus and J(i )

In this method, the non-zero elements, along with their
positions in the Y-bus matrix, are stored in only four vectors
as described below.
i. DIAG: This contains the diagonal elements of the Y-
bus in the order of the rows. The length of the DIAG is
equal to the NO-buses.
46
 Cont’d
ii) OFF-DIAG: This contains the non-zero off-diagonal
elements of the Y-bus matrix. The length of the OFF-DIAG
is equal to (n2-n-z).
iii) Y row: indicates the number of off-diagonal elements in
each row
iv) Y-Col: contains the column number of each off-diagonal
element. The length of column is equal to that of off-
diagonal.

47
 Cont’d

II. Ordering of Buses

The other sparsity technique is reorder of the buses.
One simple reordering method is to start with those buses
having the fewest connected branches and to end with
48
 Cont’d

those having the most connected branches.

Ordering of buses can be divided into two.
(a) Preordering.
In this case nodes are renumbered before triangulation or
eliminations once in sequence of increasing number of
connected lines.
Diagonal banding-nonzero elements are arranged about
either the major or minor diagonals of the matrix.
 No complicated programming or storage is required to
keep track of row and column interchanges.

49
 Cont’d
(b) Dynamic ordering.
 Ordering is effected at each row during the elimination.
 That means buses can be renumbered during each Gauss-
elimination step in order to account for changes during the
elimination process.
 The dynamic ordering is better than the preordering
conservation, number of arithmetic operations required,
ordering times and total solution time.
III. Matrix (Triangular) factorization.
If, Ax =b (2.38),
Then to solve x for different set of vectors b it will be
convenient to systemize the procedure in the preceding
section.
50
 Cont’d

One such method is to factorize the matrix A as,

A = LU (2.39)
The matrix L is lower triangular matrix with non-zero
diagonal element and the matrix U is an upper triangular
matrix with the diagonal is equal to 1. After factorization,
Eq. (2.37) becomes

2.40

ie., 2.41
Where, 2.42
51
 Cont’d
Since L is lower triangular matrix, we can solve for y1,
y2,…. yn. Successively from

2.43

Knowing Yi’s, we can solve for xi’s from

2.44

52
 Cont’d

An alternative version to solve eq. 2.38 is express A as

follows,
A = L’DU 2.45
Where,
L’ is lower triangular matrix with diagonal entries
equal to 1 by dividing each column by its diagonal value.
D is the diagonal matrix of the original matrix
U is the upper triangular matrix as before.
The advantage of this factorization is that if Ais symmetric,
then both L’ and U are transpose of each other.
Thus in computer, we need to store only L’ or U.
53
 Summery of the chapter

The basic information contained in the load-flow output is:

i) All bus voltage magnitudes and phase angles w.r.t the
slack bus.
ii) All bus active and reactive power injections.
iii) All line sending- and receiving-end complex power
flows.
iv) Individual line losses and
v) Total system losses can be deduced by summing
item iv).
NB. The most important information obtained from the
load-flow is the system voltage profile.
54
 Cont’d

Why we need of Load Flow Study?

Because load flow is a base to;

 Plan and design a power system

 Expansion of power system

 Provide guidelines for optimum operation of power

system

 Provide guidelines for various power system studies

55
 COMPARISON OF LOAD FLOW METHODS

The following are the main Criteria points to select or

compare best technique:
A. Will the iteration procedure converge to the unique
solution?
B. What is the convergence rate (how many iterations are
required)?
C. When using a digital computer, what are the computer
storage and time requirements?
D. Simplicity for programming and incorporating different
extra technique.

56
Cont’d

57
 Cont’d

THANK YOU!!!

58