Load Flow Studies

By
Dr. Lalit Chandra Saikia
Professor
NIT Silchar
Introduction:
• Load flow solution is a solution of the network under steady
state condition subject to certain inequality constraints under
which the system operates.
• These constraints can be in the form of load nodal voltages,
reactive power generation of the generators, the tap settings
of a tap changing under load transformer etc.
• The load flow solution gives the nodal voltages and phase
angles and, hence the power injection at all the buses and
power flows through interconnecting power channels
(transmission lines).
• Load flow solution is essential for designing a new power system and
for planning extension of the existing one for increased load demand.
• These analyses require the calculation of numerous load flows under
both normal and abnormal (outage of transmission lines, or outage of
some generating source) operating conditions.
• Load flow solution also gives the initial conditions of the system when
the transient behaviour of the system is to be studied.
• Load flow solution for power network can be worked out both ways
according as it is operating under (i) balanced, or (ii) unbalanced
conditions.
• The following treatment will be for a system operating under balanced
conditions only.
• For such a system a single phase representation is adequate.
A load flow solution of the power system requires mainly the following
steps:
(i) Formulation or the network equations.
(ii) Suitable mathematical technique for solution of the equations.

• Since we are studying the system under steady state conditions the
network equations will be in the form of simple algebraic equations.
• The load and hence generation are continually changing in a real
power system.
• We will assume here that loads and hence generation are fixed at a
particular value over a suitable period of time, e.g., half an hour or
so.
Bus classification
• In a power system, each bus or node is associated with four quantities,
real and reactive powers, bus voltage magnitude, and its phase angle.
• In a load flow solution, two out of the four quantities are specified and
the remaining two are required to be obtained through the solution of
the equations.
• Depending upon which quantities have been specified, the buses are
classified in the following three categories:

❖ Load Bus
❖ Generator Bus or Voltage Controlled Bus
❖ Slack Bus
1. Load Bus: At this bus the real and reactive components or power are
specified. It is desired to find out the voltage magnitude and phase angle
through the load flow solution. It is required to specify only PD and QD at
such a bus as at a load bus voltage can be allowed to vary within the
permissible values e.g., 5%. Also phase angle of the voltage is not very
important for the load.
2. Generator Bus or Voltage Controlled Bus: Here the voltage magnitude
corresponding to the generation voltage and real power PG corresponding
to its ratings are specified. It is required to find out the reactive power
generation QG and the phase angle of the bus voltage.
3. Slack, Swing or Reference Bus:
• In a power system there are mainly two types of buses: load and
generator buses.
• For these buses, we have specified the real power P injections.
• The losses remain unknown until the load flow solution is complete.
• It is for this reason that generally one of the generator buses is made
to take the additional real and reactive power to supply transmission
losses.
• That is why this type of bus is also known as the slack or swing bus.
• At this bus, the voltage magnitude V and phase angle δ are specified
whereas real and reactive powers PG and QG are obtained through the
load flow solution
Table Showing summary of the buses
• The phase angle of the voltage at the slack bus is usually taken as the
reference.
• In the following analysis the real and reactive components of voltage at a
bus are taken as the independent variables for the load flow equations i.e
where ei and fi are the real and reactive components of
voltage at the ith bus.

• There are various other formulations wherein either voltage or current or both are
taken as the independent variables.
• The load flow equations can be formulated using either the loop or bus frame of
reference.
• However, from the viewpoint of computer time and memory, the nodal admittance
formulation, using the nodal voltages as the independent variables is the most
economic.
NODAL ADMITANCE MATRIX
• The load flow equations, using nodal admittance formulation for a three-
bus system (Fig.), are developed first and then they are generalized for an
n-bus system.
At node 1

Here y11 is the shunt charging admittance at bus 1 and ground

• Similarly, nodal current equations for the other nodes can be written as
follows:

These equations can be written in a matrix form as follows:

or in compact form these equations can be written as

From this we now write nodal current equation for an n-bus system where each node is
connected to all other nodes

or in matrix form
It can be shown that the nodal admittance matrix is a sparse matrix (a few number
of elements are non-zero) for an actual power system. Consider Fig. 18.2.

• The nodal admittance matrix for

the above system is as follows:

(18.7)
If the interconnection between the various nodes for a given system, and
the admittance value for each interconnecting circuit are known, the
admittance matrix may be assembled as follows:
(i) The diagonal element of each node is the sum of the admittances
connected to it.
(ii) The off-diagonal element is the negated admittance between the nodes.

• However, it can be seen that the sum of the elements in each column of
the admittance matrix summates to zero which means that nodal
admittance matrix is a singular matrix and hence the rows of the matrix
are linearly dependent.
• We have already justified the necessity of selecting one of the buses as
the slack or reference bus based on the power balance in the system.
• Based on admittance matrix approach, it can be said mathematically
that one of the buses should be taken as slack, otherwise the nodal
matrix is singular and cannot be handled.
• By taking one of the buses as reference, the corresponding row and
column are deleted from the nodal admittance matrix. Hence, the
reduced matrix becomes non-singular, which can be handled very easily.
Shunt admittance:
• Shunt admittances, are added to the diagonal elements corresponding to
the nodes at which they are connected. The off-diagonal elements are
unaffected.
• The addition of a shunt element in this way may either strengthen or
weaken the Y matrix diagonal depending on whether the additional
elements add to or subtract from the diagonal terms, representing series
admittance summations.
• Shunt inductances strengthen the diagonal while shunt capacitances
weaken it.
• With this also the singularity of the admittance matrix can be avoided.
Tap changing transformer
• The tapped transformers operating at off-nominal tap positions provide
means of exchanging reactive power between networks operating at
different voltages, and between generators and the network system to
which they are connected, it is required to reflect this into power balance
equations.
• Refer to Fig. 18.3. Here t is the per unit off-nominal tap position, either
positive or negative, and the sending end of the equivalent circuit
corresponds to the node to which the transformer is connected.

• From the equivalent circuit,

where Z is the per unit

impedance of the transformer.
From the above two equations
eliminating E we have
• Equations (18.10) and (18.11) for Ir
and Is have been rearranged so as
to suit the equivalent circuit as
shown in Fig. 18.4.
• The equivalent circuit can be
used to include the presence of
tap changing transformer while
formulating the nodal
admittance matrix.
• The shunt admittances of the
equivalent circuit are added to
the Y matrix diagonal
elements at the two nodes,
and the negated series
admittance is the off-diagonal
element corresponding to the
transformer interconnection.
• If the transformer connection is included in a transmission circuit as
shown in Fig. 18.5, the equivalent circuit representation in the above
equation (18.10–18.11) will hold good except that the branch impedance
is the sum of the transformer impedance and the circuit impedance.
• If Z1 is the line impedance, the expressions for Ir and Is become

• If the transformer is
connected between
nodes p and q the
equivalent circuit for
nodal admittance
formulation will be
as shown in Fig.
18.6.
• With this we are able to take into account the shunt
branches and the effect of tap changing under load
transformers on the power flow in the interconnected
network.
• Once these components are included in the nodal
admittance matrix we need not bother about them as their
presence will be taken care of by the admittance matrix.
Development of load flow equations
• The nodal current equations derived earlier are rewritten below for a n-
bus system in equation (18.6).
• Ip has been substituted by the real and reactive powers because
normally in a power system these quantities are specified.
Iterative Methods
• Equations (18.18) are the load flow equations where bus voltages are
the variables.
• It can be seen that the load flow equations are nonlinear and they can
be solved by an iterative method. The iterative methods are
1. Gauss’s method,
2. Gauss-Seidel method.
• Before these methods
are explained we take
a specific example and
apply these methods
for the solution of the
load flow equations.
Refer to Fig. 18.2.
• The five bus system has two
generators at buses 1 and 3
and three load buses 2, 4 and
5. The nodal admittance matrix
is also given along with Fig.
18.2.
• By using this nodal admittance
matrix along with the equation
(18.18), the load flow
equations for the five bus
system are written as follows.
Assuming bus 1 as the slack
bus,
The nodal admittance matrix for the above system
• By using this nodal
admittance matrix along
with the equation
(18.18), the load flow
equations for the five
bus system are written
as follows. Assuming
bus 1 as the slack bus,
➢To understand the procedure for solution of these equations we make
a simplifying assumption at this stage that all the other buses except
bus No. 1 are load buses i.e., buses where P and Q are specified.
➢We will include the presence of PV buses is the next article.
➢The admittances and voltages as used in these equations are complex
quantities and the number of nonlinear equations is (n – 1) where n is
the total number of buses in the system.
➢The following is the Gauss iterative procedure for solving the
equations.
➢The flow chart for the procedure is given in Fig. 18.7.
The bus voltage equations (18.22) can also be solved by the Gauss-Seidel method.
In this method the new calculated voltage ViK+1 immediately replaces ViK and is used
in the solution of the subsequent equations. The equations (18.19) in this case
become
The general load flow equation resultant from Gauss-Seidel method will be as given below:

• The second term on the r.h.s. of the above equation is clear because
the voltage prior to bus p should correspond to the value as calculated
during the current iteration.
• The procedure for solution of these equations is outlined below and
takes into account the presence of voltage controlled buses in addition
to the load buses.
Before we proceed to solve these equations (18.24) the following points
must be kept in mind:
1. The voltages and admittances are complex quantities.
2. Buses 2, 4 and 5 are load buses, where P and Q are known quantities
whereas bus 3 is a voltage controlled bus where only P is known and Q is
not known. Therefore, for solving equation corresponding to a voltage
controlled bus Q must first be calculated and is given as
• After Q at the voltage-controlled bus has been calculated, it should be
checked whether this value violates the reactive power generation at the
bus i.e., whether it is less than the minimum reactive power generation
permissible (because of stability problem) or it is more than the maximum
reactive power generation permissible (because of rotor heating).
• If it violates, the voltage-controlled bus is made to act as a load bus for that
iteration only and the reactive power to be substituted in the expression
for that iteration will correspond to the limit it has violated, i.e., if it is less
than Qmin, the value to be substituted will be Qmin.
• If the calculated value of Q lies within the limits, this value of Q will be
substituted in the expression.
• The violation of reactive power limit could be due to the specified
voltage either being too low or too high.
• Since | V |spec can be obtained only by controlling reactive power,
therefore, it is possible that we have specified | V | beyond the
capability of the reactive power generation of the generator.
• If the value of Q for a generator bus is calculated after assuming
magnitude equal to | V |sp for that bus and phase angle corresponding
to the value in that iteration and if Q so obtained violates the limit, the
value of voltage should be taken corresponding to the voltage in that
iteration and not Vsp .
• Otherwise the voltage should correspond to the specified voltage and
phase angle as in that iteration.
3. The number of nonlinear equation is (n – 1), where n is the number of buses
in the system.

The following is Gauss-Seidel iterative procedure for solving equations:

1. Assume a flat voltage profile 1 + j0.0 for all nodal voltages except the slack bus 1.
Assume a suitable value of  the convergence criterion i.e., if the absolute value of
the maximum change in voltage between any two consecutive iterations is less
than a prespecified tolerance ε, the convergence is achieved and the iterative
procedure is terminated.
1 (a) Set iteration count k = 0
1 (b) Set bus count p = 1
1 (c) Check for slack bus. If it is a slack bus, go to step 4 (a), otherwise go to next step
2. Check which of the buses are voltage controlled and which are load buses. For voltage
controlled buses go to next statement, otherwise go to step 4.
3.
• Replace the value of the voltage magnitude of voltage controlled bus in that
iteration by the specified value.
• Keep the phase angle same as in that iteration.
• Calculate Q for the generator bus. If Q lies within the lower and upper bounds
calculate the term (P – jQ)/V for this bus.
• Repeat this for all voltage controlled buses and calculate this term and substitute
this term (P – jQ)/V in the load flow equation corresponding to the voltage
controlled bus.
• Calculate the new value of voltage for the bus.
• It is to be noted that if there are more than one generator buses, the voltage
magnitude of that bus only is replaced by its specified value, while calculating P
and Q of a particular bus.
• The voltage of other generator buses will be corresponding to the
value in that iteration.
• In case any or all the voltage controlled buses violate the reactive
power generation, the bus will be treated as a load bus and the
magnitude of the reactive power at this bus will correspond to the
limit it has violated, as explained earlier, and the value of the
magnitude of voltage will correspond to the value in that iteration
(not corresponding to specified voltage), and then go to next step.

4 (a) Advance the bus count by 1 and check if all the buses have been taken into
account. If yes, go to next step, otherwise go to step 1 (c).
5. Find out the largest of the absolute value of the change in voltage. If this
is less than a prespecified tolerance move on to next step, otherwise go
back to step 2.
6. Calculate the injected powers and the line flows using the nodal
voltages.
The flow chart for Gauss-Seidel method is given in Fig. 18.8. The flow chart
includes the presence of voltage controlled buses and the simplification of
the computational procedure.
It is found in practice that the process of convergence due to G-S method is slow i.e., it
requires a large number of iterations before a solution is obtained. The process of
convergence can be speeded up if the voltage correction during consecutive iteration is
modified to

where α is known as acceleration factor and is a real number. A suitable value of α for a
particular system can be obtained by running trial load flows. α = 1.6 is a general
recommended value for most of the systems.
However, it may be noted that a wrong selection of α may result in slower convergence
and sometimes even result in divergence from the solution.
Newton Rapson Method
• Development of load flow equations: The load flow problem can also
be solved by using the Newton-Raphson method. The equations for
the method are derived as follows:
• These three sets of equations are the load flow equations and it can be
seen that they are non-linear equations in terms of the real and
imaginary components of nodal voltages.
• Here the left-hand quantities i.e., Pp , Qp (for a load bus), and Pp and |Vp| for a
generator bus are specified and ep and fp are unknown quantities.
• For an n-bus system, the number of unknowns are 2(n – 1) because the voltage at
the slack bus is known and is kept fixed both in magnitude and phase.
• Therefore, if bus 1 is taken as the slack, the unknown variables are (e2, e3, ..., en–1,
en, f2, f3, ..., fn–1, fn).
• Thus, to solve the problem for 2(n – 1) variables, we need to solve 2(n – 1) set of
equations.
Newton-Raphson method is an iterative method that approximates the set of
nonlinear simultaneous equations to a set of linear simultaneous equations using
Taylor’s series expansion and the terms are limited to first approximation.

The mathematical background of this method is explained as follows: Let the

unknown variables be (x1, x2, ..., xn) and the specified quantities y1, y2, ..., yn.
These are related by the set of non-linear equations:
• To solve these equations we start with an approximate solution

• Here superscript zero means the zeroth iteration in the process of solving the above
non-linear equations (18.28).
• It is to be noted that the initial solution for the equations should not be very far from
the actual solution.
• Otherwise, there are chances of the solution diverging rather than converging and it
may not be possible to achieve a solution whatever be the computer time utilized.
• At first glance it may appear to be a great drawback for the Newton-Raphson
technique but the problem of initial guess for a power system is not at all difficult.
• A flat voltage profile i.e., Vp = 1.0 + j0.0 for p = 1, 2, ..., n except the slack bus has
been found to be satisfactory for almost all practical systems.
• The equations are linearized about the initial guess. We will expand first equation
y1 = f1 and the result for the other equations will follow.
When referred to a power system problem (assuming there is only one generator bus
which is taken as slack bus and all other buses are load buses), the above set of
linearized equations become
In short form it can be written as

In case the system contains all types of buses, the set of equations can be written as

The elements of the Jacobian matrix can be derived from the three load flow equations
(18.25) to (18.27).
where the superscript zero means the value calculated corresponding to initial guess
i.e., zeroth iteration.
 • These values of voltages will be used in
the next iteration.
• The process will be repeated and in
general the new better estimates for bus
voltages will be
• The process is repeated till the magnitude of the largest element in the
residual column vector is less than the prespecified value.
• The sequence of steps for the solution of load flow problem using Newton-
Raphson method is explained as follows (flow chart in Fig. 18.9):
Example 18.4: The load flow data for the sample power system are given below. The
voltage magnitude at bus 2 is to be maintained at 1.04 p.u. The maximum and minimum
reactive power limits of the generator at bus 2 are 0.35 and 0.0 p.u. respectively.
Determine the set of load flow equations at the end of first iteration by using Newton-
Raphson method.
Similarly we find out the partial derivatives of the reactive power. Diagonal
elements: