International Journal of Engineering Research & Technology (IJERT)

ISSN: 2278-0181
Vol. 4 Issue 06, June-2015

Optimal Load Flow Analysis in Interconnected

Network using MATLAB
Ram Kishore Pal (M.Tech Student) Rakesh Singh Lodhi
Electrical & Electronics Engineering Electrical & Electronics Engineering
Oriental University Oriental University
Indore, M.P., India Indore, M.P., India

Abstract- Load flow analysis is the basis of power system Depending upon the type of algebraic equations developed in
planning design and operation. The main objective of this paper the study, numerical methods are selected to solve them. The
is to present an optimal load flow method to solve the solution of mathematical expressions gives the parametric
Interconnected Distribution System... The load flow test case values which are helpful to decide whether the system is
problem has been successfully solved for interconnected
stable or unstable, controllable or uncontrollable.
distribution systems with different methods like Newton-
Raphson, Gauss-Seidel and Fast-Decoupled. The comparison of The energy is produced at the locations where necessary
results for the test case of IEEE 30 bus network clearly shows raw material is available economically and safely i.e. pithead
that the Newton-Raphson method is indeed capable of obtaining power station. This resulted in evolution of large central
optimum solution efficiently for Load flow problems. The power energy generating station with elaborate transmission and
loss has also been reduced on getting the solution and distribution systems. Due to large energy requirements an
implementation of remedial actions through MATLAB interconnected power systems are preferred over isolated
programming. power system.
An interconnected system gives better handling of load and
Keywords- Load flow, Newton-Raphson, Gauss-Seidel, Fast-
operation is economical as compared to the isolated system
decoupled method, MATLAB, Interconnected system
with given security and reliability constraints.

I. INTRODUCTION II. LOAD FLOW ANALYSIS

The history of civilization shows a close relationship One of the most common computational procedures used in
between the consumption of energy and progress of mankind. power system analysis is the load flow calculation.
The per capita electricity consumption shows a measure of Load flow studies determine the voltage, current, active, and
prosperity and progress of any country. Energy consciousness reactive power, power factor and power loss in the system.
in the people has created an interest in them to tap new Load flow studies are an excellent tool for system planning.
sources of energy from time to time. A number of operating stages can be analyzed, including
Amongst the various forms of energy, electricity has played a contingency conditions, such as the loss of a generator, a
very vital role for a devolved country. To fetch the electricity transmission line, a transformer, or a load and also to
as per the requirement of various categories of users a huge determine the size and location of capacitors for power factor
interconnected network is required Therefore a number of improvement.
studies need to be conducted on the system for its operation Following three methods are mostly used for the solution of a
control. Power system network is the biggest man made Load Flow Problem.
system in the world. This network is very vast and it is very • Newton-Raphson method (N-R)
difficult to understand its behavior. For getting values of • Fast-decoupled Power Flow method and
system parameters, a deep study of this system network is • Gauss-Seidel method (G-S)
required. A system may be under steady state or transient
state. When a power system is in operation, it frequently III. LOAD FLOW EQUATION
switches over from one state to another. Therefore the system
conditions need to be analyzed under both the conditions for For the formulation of the active and reactive power
entering a bus, we need to define the following quantities. Let

Vi  Vi  i  Vi cos  i  j sin  i 
better operation and control.
The study can be covered in to three stages: the voltage at the ith bus be denoted by
Network modeling, Mathematical modeling and Solution .. (1)

Yii  Yii ii  Yii cosii  j sinii   Gii  jBii

In the network modeling stage, the equipments are Self admittance at bus-i as
represented as an equivalent electrical circuit component.
In mathematical modeling stage, with the help of applicable

Yij  Yij ij  Yij cosij  j sinij   Gij  jBij

laws, the network model is converted in to mathematical .. (2)
model as algebraic equations in study state and .differential Mutual admittance between the buses i and j can be written as
equations are developed in dynamical studies. Solution of the
equations is obtained through Matlab programming.

IJERTV4IS060086 www.ijert.org 91
(This work is licensed under a Creative Commons Attribution 4.0 International License.)
 International Journal of Engineering Research & Technology (IJERT)
ISSN: 2278-0181
Vol. 4 Issue 06, June-2015

..(3) The steps of a computational algorithm are given below:

Let the power system contains a total number of n buses. The Step-1: With the load profile known at each bus i.e. P D, QD

I i  Yi1V1  Yi 2V2    YinVn

injected current at bus-i is given as are known, allocate Pca and Qca to all generating stations.
Step-2: Assembly of bus admittance matrix YBUS with the

  YikVk
line and shunt admittance data stored in the computer, YBUS is
n .. (4) assembled by using the rule for self and mutual admittances.
Step-3: Iterative computation of bus voltages (Vij= 2, 3 ….,
k 1
n): to start the iterations a set of initial voltage values is

Pi  jQi  Vi I i  Vi  Yik Vk

The complex power at bus-i
n
assumed.
Step-4: Computation of slack bus power: substitution of all
k 1
bus voltages computed in step 3 along with Vi yields S*= P-

 Vi cos  i  j sin  i  [ Yik Vk cos  ik  j sin  ik cos  k  j sin  k ]

n jQ.
Step-5: Computation of line flows and line losses: this is the
k 1

  [ Yik ViVk cos  i  j sin  i cos  ik  j sin  ik cos  k  j sin  k ]

last step in the load flow analysis wherein the power flows on
n the various lines of the network are computed. Consider the
lines connecting buses i and k.
k 1
B Newton-Raphson Method

cos  i  j sin  i cos  ik  j sin  ik cos  k  j sin  k 

(5)
Note that The first method (G-S) is simpler but the second (NR) is

 cos  i  j sin  i cos ik   k   j sin  ik   k 

reported to have better convergence characteristics and is
faster than (G-S) method.

 cos ik   k   i   j sin  ik   k   i 
The Newton-Raphson procedure is as follows:

V(0) of all np load buses and n  1 angles (k) of the voltages

Step1: Choose the initial values of the voltage magnitudes

Step-2: Use the estimated V(k) and (k) to calculate a total n 

of all the buses except the slack bus.

Pi   YikViVk cos ik   k   i 
Therefore substituting in (5) we get,
n
1 number of injected real power P calc(k) and equal number of
(6) real power mismatch P(k).
Step-3: Use the estimated V(k) and (k) to calculate a total np
k 1

Qi   YikViVk sin  ik   k   i 
n
number of injected reactive power Qcalc(k) and equal number
of reactive power mismatch Q(k).
(7)
k 1
This is load flow equation. Step-3: Use the estimated V(k) and (k) to formulate the

  2 
Jacobian matrix J(k).
Pi ,inj  PGi  PLi
Then the total real power injected in bus-i is
   P2 
 
(8)
 
  n

   for (k) and
Let the injected power calculated by the load flow program
 V 

 Pn 
Pi  Pi ,inj  Pi ,calc  PGi  PLi  Pi ,calc (9)
Step-4: Solve
J  
be Pi,calc. Then the mismatch

 V2   Q2
2


   
   

Similarly, the mismatch between the reactive power and  
    
Qi  Qi ,inj  Qi ,calc  QGi  QLi  Qi ,calc (10)  
calculated values is given by V1 n p
 V 
Q1 n p


 1 n p 

The purpose of the load flow is to minimize the above two V(k)V(k).
mismatches. It is to be noted that (6) and (7) are used for the Step-5: Obtain the updates from
calculation of real and reactive power in (9 to10).

 1   0    0 
 V k 

IV TECHNICAL SOLUTION (11)

k 1 k 
V 1  k 

 
Because of the nonlinearity and the difficulty involved in V (12)
the analytical expressions for the above power flow V
equations, numerical iterative techniques are applied such as:
Step-6: Check if all the mismatches are below a small
A Gauss-Seidel Method number. Terminate the process if yes. Otherwise go back to
step-1 to start the next iteration with the updates given by
The Gauss-Seidel Method is another iterative technique for (11) and (12).
solving the load flow problem, by successive estimation of
the node voltages.

IJERTV4IS060086 www.ijert.org 92
(This work is licensed under a Creative Commons Attribution 4.0 International License.)
 International Journal of Engineering Research & Technology (IJERT)
ISSN: 2278-0181
Vol. 4 Issue 06, June-2015

C. Fast-Decoupled Load Flow method: TABLE-1

IEEE-30 Bus Data comparison of different load flow methods
Bus Status
Power-flow Programming Steps
Step 1: Input raw data: Input the electrical grid data in the S.
Data N-R G-S
Fast
standard data structure and estimate start points of the No. Decoupled
remaining undefined bus magnitude and phase angle.
1 Iteration 4 83 15
Step 2: Data regeneration: Line data and bus data are
specified to change along with bus numbers simultaneously. 2 Error 1.2e-005 0.000951 0.0009195
Step 3: Evaluate Ybus: Evaluate bus admittance matrix.
Step 4: Evaluate power mismatch: Evaluate ∆Pi and ∆Qi Total Load
respectively from real power mismatch equation (9) and (10). 3 Bus active 283.400 283.400 283.400
power in MW
Step 5: Evaluate Jacobian elements. Total Load
Step 6: Evaluate increments. Evaluate increments of bus Bus Reactive
voltage magnitude ∆V and increments of bus angle degree 4
power in
126.200 126.200 126.200
∆δ by power flow matrix with Gauss Elimination method. MVAr
Step 7: Update variables. Update new bus voltage magnitude Total
Generation
and new bus phase angle by using equations as follows. And 5 300.998 300.727 300.998
Bus active
then returns to Step5 with new values to replace the precious power in MW
values. Total
V +ΔVk
V(k+1)= k
(13) Generation
6 Bus Reactive 125.144 125.098 125.145
θ = θ(k)+Δθ(k)
(k+1)
(14) power in
Step 8: Examine reactive power of each PV Bus. MVAr
Step 9: Termination conditions. The process is continued Line Losses
until the residuals ∆P(k) and ∆Q(k) are less than tolerance 1
Total MW
17.599 17.573 17.598
Line Losses
accuracy as follows.
Total MVAr
Max|∆P(k) |< € 2
Line losses
22.244 22.139 22.245
Max|∆Q(k)|< €
Note the € are the small positive constant, such as 0.001, The rate of convergence of the GS method is slow as
0.00001, and the smaller numbers are pretty more accurate. compare to other two methods. In GS method no of iteration
Step 10: Evaluate power of the swing bus and each PV bus: are in increases trend in proportion to the no of buses in
Evaluate reactive power of each PV bus and evaluate active network whereas in NR and FDLF method no of iteration are
power and reactive power of the swing bus eq. lower side despite increase of no of buses in network.
Step 11: Reconfigure bus numbers: Return each bus number The chief advantage in GS method is the easy programming
to the original bus number in the bus data matrix and the line and most efficient utilization of core memory. In FDLF,
data matrix. storage requirements are around 60% to that of the NR
Step 12: Evaluate line flow and loss: Evaluate real and method, but slightly more than the GS method. The
reactive power flows in transmission lines and transformers, programming of NR method is complex and it has the
as well as equipment losses. disadvantage of requiring a large computer memory. But
Because of high accuracy and lowest no of iterations, the NR
V. OPTIMAL POWER FLOW PROBLEMS-SOLUTION method is preferable for load flow and stability studies. The
TECHNIQUES: method can be extended for transformer tap-changing
operations, variable constraints on bus voltage, optimal real
A. Problem – I (IEEE – 30 bus system) and reactive power scheduling. The GS method is useful
only for smaller bus system. NR method is reliable small and
One case is to make use of it to gives the optimal solution large system also. The FDLF method is comparatively more
of load flow techniques on the IEEE- 30 bus test system. The reliable than GS method and of course the NR method is
case study is evaluating various analytical methods and most reliable for any system whether small or larger one.
computer program for the solution by Gauss-seidel, Newton-
Raphson and Fast Decouple load flow method. Bus 1 is taken VI. CONTROL TECHNIQES
as a slack bus with its voltage adjusted to 1.06∠0˚ pu. The
data for the voltage controlled buses are given in Appendix-I A. Load sharing
and II. In the given data, the active load on bus no.-5 is 94.2 MW
and line losses is 17.599 MW/22.244MVAr as shown in table
B. Test Results 1.Corrective step:-
Load flow analysis is carried out in IEEE 30 bus test system. From the highly loaded bus no 5, the 10% load is shared by
Output Voltage magnitude and Voltage Angle Load Bus bus no 7 with the modified loading condition of bus no 5 as
active and reactive power, generation bus active and reactive 84.8MW and on bus no 7 as 32.2MW.
bus and line losses values from N-R, G-S and FDLF. Now the results simulation process is obtained by way of
reduced line losses to the extent of 17.536MW/21.976 MVAr
as show in table2

IJERTV4IS060086 www.ijert.org 93
(This work is licensed under a Creative Commons Attribution 4.0 International License.)
 International Journal of Engineering Research & Technology (IJERT)
ISSN: 2278-0181
Vol. 4 Issue 06, June-2015

TABLE-3
TABLE-2 IEEE-30 Bus Data comparison after 10% MVAr power
IEEE-30 Bus Data comparison after 10% load sharing from bus no 5 to 7
injected at bus no 30
Bus Status
S. Fast
Data N-R G-S Bus Status
No. Decoupled
Iteration S. Data Fast
1 4 83 15 N-R G-S
No. Decoupled
2 Error 7.75e-007 0.0009625 0.0009271 Iteration
1 4 83 15
Total Load Bus
3 active power in 283.400 283.400 283.400 2 Error 7.69e-007 0.0009300 0.0009720
MW Total Load Bus
Total Load Bus 3 active power in 283.400 283.400 283.400
4 Reactive power 126.200 126.200 126.200 MW
in MVAr Total Load Bus
Total 4 Reactive power 126.200 126.200 126.200
Generation Bus in MVAr
5 300.936 300.661 300.936 Total
active power in
MW Generation Bus
5 300.978 300.693 300.962
Total active power in
Generation Bus MW
6 124.876 124.829 125.877 Total
Reactive power
in MVAr Generation Bus
6 124.040 114.650 124.041
Line Losses Reactive power
1 Total MW Line in MVAr
17.536 17.510 17.536 Line Losses
Losses
2 Total MVAr 1 Total MW Line
21.976 21.869 21.977 17.578 17.536 17.578
Line losses Losses
2 Total MVAr
21.141 21.69 22.141
B. Control by VAR generators Line losses

The control the voltage profiles of an interconnected system

using positive VAR power injects the load buses. Load flow C. Changing the tap position of the transformer
solution gives the voltage levels at the load buses. In bus no In the given data, the transformer tap position feeding bus
30 voltage is less as compare to other buses voltages, it is no 4 is 0.932 as shown in table 1. And the line losses are to
indicate of the fact that the reactive power flow capacity of the extent of 17.599MW/22.244MVAr.
the transmission lines for specifies voltage limits cannot meet Corrective step:-The transformer tap position is put on 1
the reactive load demand. This situation can be remedied by instead of 0.932 and the results are obtained through the
installing VAR generators at the load bus no 30. After the simulation process by way of reduced line losses to the extent
positive VAR inject. of 17.542MW/22.625MVAr as shown in table-4.
Reduction in line losses by injecting 10% MVAr power in the
system:- TABLE-4
IEEE-30 Bus Data comparison after changing the tap position 0.932 to 1.0 at
In the given data, PU voltage magnitude is less than the rated bus no 4.
to the extend of 0.995 and the line losses are to the extend of Bus Status
17.599MW/22.244MVAr as shown in table 1. S. Data
N-R G-S
Fast
Corrective step:- No. Decoupled
1 Iteration 4 83 15
From the given data table 1, bus no 30 is identifies as the
2 Error 0.0004082 0.0009567 0.0009871
weak bus in terms of the rated voltage and therefore Total Load Bus
additional 10% reactive power 1.0MVAr is injected in bus 3 active power in 283.400 283.400 283.400
no 30. MW
Now the results are obtained through simulation process by Total Load Bus
4 Reactive power 126.200 126.200 126.200
way of reduced line losses to the extent of 17.578 in MVAr
MW/22.1415MVAr as shown in table 3. Total
Generation
5 300.939 300.686 300.928
Bus active
power in MW
Total
Generation
6 Bus Reactive 125.482 125.476 125.547
power in
MVAr
Line Losses
1 Total MW
17.542 17.531 17.528
Line Losses
2 Total MVAr
22.625 22.517 22.644
Line losses

IJERTV4IS060086 www.ijert.org 94
(This work is licensed under a Creative Commons Attribution 4.0 International License.)
 International Journal of Engineering Research & Technology (IJERT)
ISSN: 2278-0181
Vol. 4 Issue 06, June-2015

VII. CONCLUSION Appendix-I

30 Bus system Line Data
The load flow case problem has been successfully solved B/2
From Bus To Bus R (pu) X (pu) X'mer Tap
(pu)
for interconnected systems with different load flow methods
1 2 0.0192 0.0575 0.0264 1
and obtained results have been compared for all the three
methods like Newton-Raphson, Gauss-Seidel and Fast- 1 3 0.0452 0.1652 0.0204 1
Decouple. 2 4 0.057 0.1737 0.0184 1
The Newton-Raphson, Gauss-Seidel and Fast-Decoupled load
flow method were successfully designed and implemented to 3 4 0.0132 0.0379 0.0042 1
solve the Load flow problem. The comparison of results for 2 5 0.0472 0.1983 0.0209 1
the test case of IEEE 30 bus network clearly shows that the
2 6 0.0581 0.1763 0.0187 1
Newton-Raphson method is indeed capable of obtaining
optimum solution efficiently for Load flow problems.After 4 6 0.0119 0.0414 0.0045 1
getting the solution,. the line losses have also been reduced
5 7 0.046 0.116 0.0102 1
by taking the remedial actions like reactive power injection,
changing the transformer tap position and shifting of load and 6 7 0.0267 0.082 0.0085 1
after formulation of the programmed results obtained through 6 8 0.012 0.042 0.0045 1
running on mat lab software as shown in table 2and4. This
indicates the significance of the Newton Raphson method to 6 9 0 0.208 0 0.978

solve load flow problems interconnected power system 6 10 0 0.556 0 0.969

network.
9 11 0 0.208 0 1

REFERENCES 9 10 0 0.11 0 1

4 12 0 0.256 0 0.932
[1] IEEE Standard - 141 (1993) on “IEEE Recommended Practice for
Electric Power Distribution for Industrial Plants” 12 13 0 0.14 0 1
[2] IEEE Standard - 399 (1997) on “IEEE Recommended Practice for
Industrial and Commercial Power System Analysis” 12 14 0.1231 0.2559 0 1
[3] Gross C. A. and Melipoulos. S.P., “Per-Unit scaling in Electric Power
Systems”, IEEE transactions on Power Systems, Vol. 7, No. 2, May 12 15 0.0662 0.1304 0 1
1992.
12 16 0.0945 0.1987 0 1
[4] Jian Yang, Anderson M. D., “PowerGraf : An Educational Software
Package for Power Systems Analysis and Design”, IEEE Power 14 15 0.221 0.1997 0 1
Systems, Vol. 13, No. 4, November 1998
[5] Joong-Rin Shin, Wook-Hwa Lee, Dong-Hae Im, “A Windows based 16 17 0.0824 0.1923 0 1
Interactive and Graphic Package For the Education and Training of
Power System Analysis and Operation”, IEEE Power Systems, Vol. 14, 15 18 0.1073 0.2185 0 1
No. 4, November 1999.
18 19 0.0639 0.1292 0 1
[6] Marizan Sulaiman, Zainuddin Mat Isa, “A GUI based Per-Unit
Representation of Power System with Symmetrical Components : An 19 20 0.034 0.068 0 1
E-Learning Approach”, Student Conference on Research and
Development Proceedings, Malaysia, pp. 257-261, 2003 10 20 0.0936 0.209 0 1
[7] Ming-Tong Tsay, Shun-Yu Chan, “A Personal Computer Graphical
Environment for Industrial Distribution System Education, Design and 10 17 0.0324 0.0845 0 1
Analysis”, IEEE Power Systems, Vol. 15, No. 2, May 2000.
10 21 0.0348 0.0749 0 1
[8] A Sekar, Bhaskar Mahyavanshi,” Power System Simulation Software
for Use in Cyberspace”, report from Tennessee Technical University, 10 22 0.0727 0.1499 0 1
Cookeville, USA, pp. 230-233, 2005.
[9] Nagrath I. J. & Kothari D. P., “Modern Power System Analysis”, Tata 21 23 0.0116 0.0236 0 1
McGraw Hill, 1989.
[10] Stevenson & Grainger, “Power System Analysis”, Tata McGraw Hill, 15 23 0.1 0.202 0 1
1994.
[11] Kusik G.L., “Computer-Aided Power System Analysis”, Tata McGraw 22 24 0.115 0.179 0 1
Hill, 1992.
23 24 0.132 0.27 0 1
[12] H. Saadat, “Power System Analysis”, Tata McGraw Hill, 1999.
[13] Chapman S. J., “MATLAB programming for Engineers”, Thomson, 24 25 0.1885 0.3292 0 1
2002.
[14] Amos Gilat, “MATLAB : An Introduction with Applications”, John 25 26 0.2544 0.38 0 1
Wiley & Sons, 2003.
[15] IS : 3842(Part-I)-1967, Application Guide for Electrical Relays for 25 27 0.1093 0.2087 0 1
A.C. Systems. Part-I Overcurrent Relays for Feeders and Transformers.
28 27 0 0.396 0 0.968
[16] Jackson Prajapati, Virendra Patel, Hemal Patel, “Load Flow, Short
Circuit and stability Analysis Using Matlab”, 27 29 0.2198 0.4153 0 1
ICGCCEE.2014.6922365, 14665786/IEEE, P.g.1 – 5.
27 30 0.3202 0.6027 0 1

29 30 0.2399 0.4533 0 1

8 28 0.0636 0.2 0.0214 1

6 28 0.0169 0.0599 0.065 1

IJERTV4IS060086 www.ijert.org 95
(This work is licensed under a Creative Commons Attribution 4.0 International License.)
 International Journal of Engineering Research & Technology (IJERT)
ISSN: 2278-0181
Vol. 4 Issue 06, June-2015

Appendix-II
30 Bus System Bus Data
B Ty the P
V Qg Pi Ql Qmin Qmax
us pe ta g
1 1 1.06 0 0 0 0 0 0 0
1.04 4 21.
2 2 0 50 12.7 -40 50
3 0 7
3 3 1 0 0 0 2.4 1.2 0 0

4 3 1.06 0 0 0 7.6 1.6 0 0

94.
5 2 1.01 0 0 37 19 -40 40
2
6 3 1 0 0 0 0 0 0 0
22.
7 3 1 0 0 0 10.9 0 0
8
37.
8 2 1.01 0 0 30 30 -10 40
3
9 3 1 0 0 0 0 0 0 0

10 3 1 0 0 19 5.8 2 0 0
1.08 16.
11 2 0 0 0 0 -6 24
2 2
11.
12 3 1 0 0 0 7.5 0 0
2
1.07 10.
13 2 0 0 0 0 -6 24
1 6
14 3 1 0 0 0 6.2 1.6 0 0

15 3 1 0 0 0 8.2 2.5 0 0

16 3 1 0 0 0 3.5 1.8 0 0

17 3 1 0 0 0 9 5.8 0 0

18 3 1 0 0 0 3.2 0.9 0 0

19 3 1 0 0 0 9.5 3.4 0 0

20 3 1 0 0 0 2.2 0.7 0 0
17.
21 3 1 0 0 0 11.2 0 0
5
22 3 1 0 0 0 0 0 0 0

23 3 1 0 0 0 3.2 1.6 0 0

24 3 1 0 0 4.3 8.7 6.7 0 0

25 3 1 0 0 0 0 0 0 0

26 3 1 0 0 0 3.5 2.3 0 0

27 3 1 0 0 0 0 0 0 0

28 3 1 0 0 0 0 0 0 0

29 3 1 0 0 0 2.4 0.9 0 0
10.
30 3 1 0 0 0 1.9 0 0
6

IJERTV4IS060086 www.ijert.org 96
(This work is licensed under a Creative Commons Attribution 4.0 International License.)