Solving Economic Dispatch Problem Using Mipower

by Lambda Iteration Method

Gaurav Chauhan, Anjali Jain, Neelam Verma
Department of Electrical and Electronics Engineering
Amity University Uttar Pradesh

Abstract — Economic dispatch is considered as one of the cost of a power plant depends on various factors like
critical advancement errand and operational choice which gives availability of fuel, cost of fuel, transportation, cost of land,
a financial condition to control framework. The principle running cost etc. The purpose of economic load dispatch
emphasis of economic dispatch is the means by which a power (ELD) is to allocate the share of total load on each generator
system network can be arranged with the goal that we have
in order to minimize the overall operating cost, taking care
minimum cost of electric generation to take care of the load
demand, while meeting all the balance and imbalance that all the constraints are met. Thus, ELD is formulated as a
constraints. Economic load dispatch arrangements are found by problem of allocating generation among committed unit such
explaining the customary techniques, for example lambda that the objective function of total generation cost can be
iteration technique, gradient search technique, linear minimized wherein all equality and inequality constraints are
programming technique, dynamic programming technique and met [1]. The basic equality constraint is that total output
so on. Out of every single traditional technique the convergence power of generating units must be equal to demand and losses
rate of lambda iteration method is very fast and it gives accurate in the power system. Losses can be calculated by Kron’s
results and precise outcomes. This work has been made by formula [2] or load flow analysis.
utilizing MiPower version9 programming to distinguish the best
Conventionally, the minimization function of optimizing
scheduling of generator. Lambda iteration method is utilized to
take care of economic dispatch issue for 6 generating unit the fuel cost of a generating unit is represented by a single
framework with and without transmission losses. quadratic function [3] because each unit used one fuel and no
inequality constraints are taken in to consideration. Classical
Keywords— Economic load dispatch; B coefficient; MiPower . methods such as analytical method, lambda iteration method,
dynamic programming, gradient search method and linear
I. INTRODUCTION programming method [1] are used to solve linear ELD
A power plant needs to take into account load conditions problems. In all these traditional techniques, the presumption
all for the duration of the day, for summer or winter. It is of monotonically expanding cost work is made. For practical
accordingly strange to expect that a similar level of energy power system problem, characteristics of generating units are
must be created at unsurpassed. The power age must change nonlinear. Consequently traditional strategies confront a few
as per the load pattern, which may thusly shift with season. In troubles in taking care of ELD issue for the practical power
this way the economic operation must consider the load framework. For taking care of an advancement issue utilizing
condition constantly. In addition once the economic this techniques the choice of appropriate beginning stage is
generation condition has been computed, the turbine vital. On the off chance that wrong starting focuses were
representative must be controlled such that this generation chosen then the divergence or convergence of the algorithm
condition is kept up. to some local point may occur.
In the starting era of finding solution by economic Here MiPower programming is utilized to analyze the
dispatch, the solution found was to choose supply from the economic load dispatch issue of 6 producing unit framework
most proficient plant at light load conditions. The most with and without transmission misfortunes. MiPower is a very
proficient plant will be utilized to its fullest as the load is intelligent, easy to use windows based power framework
expanded, and if still the load has not been met the power is examination bundle. It incorporates an arrangement of
supplied by next most proficient plant and so on. With modules for playing out an extensive variety of energy
additionally increment in load, the process will go on till the framework outline and examination consider. MiPower
generation is not meeting the demand and losses completely. features include a top notch Windows GUI. Economic load
Along these lines the power would be provided by the most dispatch examination can be performed with most precision
productive to the minimum effective plant to achieve the and resistance. It utilizes the lambda iteration method to
pinnacle request. Unfortunately in any case, this strategy streamline the ELD issue.
neglected to limit the aggregate cost of electricity generation The paper flow under consideration is as follows: Section
We should hence scan for alternative technique which II describes the background and description of ELD problem
considers the aggregate cost generation of the considerable related to power system. Section III contains detail of the
number of units of a plant that is providing a load. The total proposed method. To validate our proposed method, the six
unit power system network has been taken and two cases are

978-1-5090-4264-7/17/$31.00 ©2017 IEEE

95
compared that is with and without transmission losses and the For the most part, the cost of work, supply and upkeep are
results obtained are discussed in fourth section. After that, settled. Pmin is the generation level underneath which is
conclusion has been drawn in fifth section. uneconomical or in fact infeasible to work the units. Pmax is the
greatest generation control constraint. For detailing the
II. ECONOMIC LOAD DISPATCH
dispatch issue, fuel costs are typically spoken to as a quadratic
Economic Load dispatch (ELD) plays a vital role in the capacity of yield control, as demonstrated as follows.
economic operation of power plant, whose purpose is to = + + (1)
allocate power generation to different committed units to Problem Formulation:
match load demand at minimum possible cost while satisfying Total fuel cost of operating N generators is formulated by,
all the equality and inequality constraints in power system.
Economic dispatch can be a linear or non-linear problem. = ( ) (2)
Linear ELD problems are those whose input-output
Ignoring transmission losses, add up to generation should
characteristic is piecewise linear and monotonically
meet the aggregate load. Thus, the equality constraint can be
increasing. Simple quadratic equation is used in linear
written as,,
problems [1]. Here lambda iteration is used to solve the linear
∑ = load (3)
quadratic equation to economize the power system network.
Here, economic dispatch problem is solved considering no In light of the maximum and minimum power breaking
points of the generators, following imbalance or inequality
losses and with losses.
requirements can be forced:
A. Economic dispatch problem-without losses Pi, min ≤Pi ≤ Pi, max;∀i=1,2,...,N. (4)
To begin with we define the ELD issue where the losses This is an constrained optimization issue that can be
incurred because of transmission is neglected.. The case is illuminated by Lagrange multiplier strategy. The Lagrange
legitimized by considering all the generators are connected to technique is figured as,
a specific busbar, as on account of individual generating units L=F T +λ ϕ (5)
in a power plant, or when they are physically found near each Where ϕ = P load –∑
other. This guarantees the transmission losses can be ignored represents the equity constraint ; λ is the Lagrange Multiplier.
because of the short distance included. Contribution to every The fundamental condition for Fi to be least is that the
unit is is expressed in terms of cost rate. The aggregate cost derivative of Lagrange work as for every free factor is zero.
rate of generation is the sum of cost rates of individual Consequently the essential conditions for the streamlining
generating units. The fundamental working limitation is that issue are:
the aggregate of the power generation must be equivalent to = {∑ Fi(Pi) + λ(Pload − ∑ Pi) }
the load
The solution of economic dispatch issue assures that the
level of generation by different generating units aims to = −λ=0; ∀i=1,2,...,N. (6)
minimize the total cost of generation for a defined level of =ϕ=0. (7)
load. Presently, for thermal generating units, the cost of fuel Rewriting (6)
per unit power output changes essentially with the power
output of the unit. Consequently one need to consider the fuel =λ; ∀i=1,2,...,N. (8)
cost attributes of the 6 generators while finding their ideal Equation (8) states that, the fuel cost is minimized if all the
genuine power generation. incremental fuel costs same. Equation (8), along with (3) and
(4) are called the coordination equations for economic load
dispatch without considering network losses.
Note: Using (1), characteristics of fuel cost of all the
generators are formulated as under,
Fi=aiPi2+biPI+cI; ∀i=1,2,...,N. (9)
Using (8), the necessary conditions for the optimal solutions
are given by,
=2aiPi+bi=λ; ∀i=1,2,...,N.. (10)
= ; ∀i=1,2,...,N. (11)
Substituting Pi from above in (3),
= Pload
∑ ( )
λ= (12)
∑ ( )
Hence λ can be calculated by (10), and then Pi , I =
Fig. 1. Cost curve of generator without losses 1,2,...,N can be calculated by eq. 9

96
B. Economic dispatch problem-with losses III. PROPOSED METHOD
So far we have considered the economic operation of a Here MiPower programming is utilized to analyze the
single plant in which we have discussed how a particular economic load dispatch issue, MiPower utilizes lambda
amount of load is shared between the different units of a iteration method to take care of the ELD issue (with and
plant. In that case, the transmission line losses are not without transmission misfortunes). Lambda iteration method
considered and assumed that the losses were a part of the load is a standout amongst the most mainstream customary
supplied. However in this case, transmission line losses are strategies to take care of ELD issue for limiting the cost of
considered. Load is distributed between the different plants producing unit. In spite of the fact that lambda emphasis
that are joined by transmission lines, then the line losses have system computational technique is unpredictable yet it joins
to be explicitly included in the economic dispatch problem. In quick for this sort of advancement issue.
this section we shall discuss this problem. The detailed algorithm for lambda iteration method [4,5]
Let us consider a practical power system consist of k- for ELD problem is given below:
thermal plants in it, load center is always far away from 1. Read given information, for instance cost coefficients
thermal plants hence power flows to load by using (ai, bi, ci), B-coefficients, power limits and power
transmission lines so losses are more, hence it is needed to demand.
economize the entire system by including the losses. Each and 2. Assume the beginning estimations of λ and Δλ.
every thermal plant will have fuel cost as well as incremental
fuel cost. Incremental fuel cost of plant k: 3. Compute generated power Pgi from each unit.
4. Check generation limits for every unit.
= a k +b k P k (13) If Pgi > Pgimax set Pgi = Pgimax
If Pgi < Pgimin set Pgi = Pgimin
Constraints taken into account:
5. Calculate total generated power.
PD+PL=PG (14) 6. Calculate mismatch in power which is given by
following equation
PD+PL-PG=0=φ (15)
= −
This problem is solved by using Lagrange function, from
equation (5) and the solution is given by;
7. If ΔP < ε, then stop estimation and compute the
generation cost. Otherwise go to following step.
=0 ;i=1,2,...,N. (16)
8. If ΔP > 0, then λ = λ-Δλ.

That is; If ΔP < 0, then λ = λ+Δλ.

9. Repeat the procedure from step 3. The above steps
can be condensed in the flow chart of figure 2.
0=[0+0+…..+ ]+λ[ - {0+0+…..+1}]
IV. RESULT AND DISCUSSION
+ λ[ - 1] ; i=1,2,...,N. The Lambda iteration method is implemented in six
generating units. The results are compared for two different
cases with and without losses. In first case generator
= λ[1 - ] ; i=1,2,...,N. constraints are considered along with the lossless system and
in second case generator constraints are considered with the
losses. All the work has been done on MiPower software.
λ=[ ] ; i=1,2,...,N. In this case, six unit thermal power plants is considered
( )
which is solved for two different cases with and without
losses. The cost coefficients and generator power limits are
λ= Li ; i=1,2,...,N. (17) shown by Table I. The B-coefficients for six unit system is
given in Table II. The negative value of B- coefficients are
Where Li is penalty factor associated with i’th plant . For not taken by the Mipower software. Hence the values are
economic load operation the product of incremental fuel cost considered to be zero for the purpose of running the software.
and penalty factor of all plants are equal [1,2]. The results are shown in Table III and Table IV. The paper
gives the insight of using Mipower software for
understanding economic load dispatch and solving the same
with the help of lambda iteration method.

97
 V. CONCLUSIONS Considering the Generator Constraints." IOSR J Electron Electr
Eng (IOSR-JEEE) e-ISSN: 2278-1676.
This paper tackles the economic load dispatch issue is by a [5] Suman, M., M. Venu Gopala Rao, A. Hanumaiah, and K. Rajesh.
well-known conventional system known as lambda iteration "Solution of economic load dispatch problem in power system
method. A six unit framework is illuminated utilizing lambda using Lambda Iteration and back propagation neural network
methods." International Journal on Electrical Engineering and
cycle technique for two distinct cases. In first case ELD issue Informatics 8, no. 2 (2016): 347.
has been settled by considering generator constraints without [6] Krishnamurthy, Senthil, and Raynitchka Tzoneva. "Investigation
transmission losses and in second case ELD issue has been on the impact of the penalty factors over solution of the dispatch
solved by considering generator constraints with transmission optimization problem." In Industrial Technology (ICIT), 2013
losses. It has been inferred that losses have critical impact on IEEE International Conference on, pp. 851-860. IEEE, 2013.
generation cost. It can be concluded from Table III and Table
TABLE I GENERATING UNIT CAPACITY AND COST-COEFFICIENTS [16]
IV that with demand the effect of losses is also increases as
depicted in Table V. Unit Pgimin Pgimax ai bi ci

REFERENCES 1 50 200 0.00375 2.00 0

[1] Wood, Allen J., and Bruce F. Wollenberg. Power generation, 2 20 80 0.01750 1.75 0
operation, and control. John Wiley & Sons, 2012..
[2] Dhillon, J. S., and D. P. Kothari. "Power system 3 15 50 0.06250 1.00 0
optimization." Preintce Hall of India Private Limited (2010).
[3] Asbfaq Husain, “Electrical power system,” CBS publishers and
4 10 35 0.00834 3.25 0
Distributers, Fifth Edition, 2007.
5 10 30 0.02500 3.00 0
[4] Dewangan, Susheel Kumar, Achala Jain, and A. P. Huddar. "A
Traditional Approach to Solve Economic Load Dispatch Problem 6 12 40 0.02500 3.00 0

TABLE II B COEFFICIENT OF SIX UNIT SYSTEM

0.000218 0.000103 0.000009 0.000000 0.000002 0.000027 

0.000103 0.000181 0.000005 0.000000 0.000002 0.000030

0.000009 0.000005 0.000417 0.000000 0.000000 0.000000
Bij =  
0.000000 0.000000 0.000000 0.000221 0.000094 0.000050
0.000002 0.000002 0.000000 0.000094 0.000243 0.000000
 
0.000027 0.000030 0.000000 0.000050 0.000000 0.000358

TABLE III RESULTS OF SIX UNI SYSTEM (WITH LOSSES)

S.NO. Load demand Pg1 Pg2 Pg3 Pg4 Pg5 Pg6 Total generation cost
(MW) (MW) (MW) (MW) (MW) (MW) (MW) (Rs)

1 200 121.20631 35.10384 16.38434 10.00000 10.00000 12.00000 519.0966

2 250 161.42421 44.73985 19.16718 10.91021 10.00000 12.00000 684.5681
3 300 185.92924 50.91993 21.28214 25.98939 13.88059 12.95436 864.4796
4 350 200.00000 61.35554 24.52391 35.00000 21.84534 20.63536 1059.3483
5 400 200.00000 80.00000 32.10045 35.00000 30.00000 38.00000 1288.2040

TABLE IV RESULTS OF SIX UNI SYSTEM (WITHOUT LOSSES)

S.NO. Load demand Pg1 Pg2 Pg3 Pg4 Pg5 Pg6 Total generation cost
(MW) (MW) (MW) (MW) (MW) (MW) (MW) (Rs)
1 200 119.95516 32.84753 15.19731 10 10 12 505.3012
2 250 159.19286 41.25561 17.55157 10 10 12 657.6415
3 300 195.94641 49.13137 19.75678 13.16535 10 12 824.5825
4 350 200 59.42004 22.63761 34.75428 16.59403 16.59403 1006.7968
5 400 200 77.98509 27.83583 35 29.58957 29.58957 1214.4470

TABLE V PERCENTAGE INCREASE IN COST WITH INCREASE IN LOAD

S.NO. Load demand Cost (With Losses) Cost (Without losses) Percentage increase in cost from without
(MW) losses to with losses
1 200 519.0966 32.84753 2.6
2 250 684.5681 41.25561 3.9
3 300 864.4796 49.13137 4.6
4 350 1059.3483 59.42004 4.9
5 400 1288.2040 77.98509 5.72

98
 Start

Read Ng,ai, bi, ci, Pd, Pgimin, Pgimax

Choose a suitable value of λ and Δλ

Set i=1

Calculate Pgi
No
Yes
Set
Is
Pgi>Pgimax Pgi=Pgimax

Yes
Is Set
Pgi<Pgimin Pgi=Pgimin

Yes
Set i=i+1

No
Is i=Ng

Calculate ΔP =∑Pgi-Pd

Yes
Print Generation
Is ΔP<ε
and Cost

No
No

λ=λ+Δλ Is ΔP>0

Yes
λ=λ-Δλ

Fig2: Flow chart for Lambda Iteration method

99