A MULTI-OBJECTIVE OPTIMIZATION

Overview:
Today, by increasing the environmental awareness and restrictions, the environmental constraints must be taken into
account for modeling dispatch problems. The economic optimization and environmental dispatch problems have two
different trends. While the economic dispatch reduces the operation costs of the system, the rate of environmental
pollutions are increasing and reverse. Based on these reverse trends, it is necessary to model economic and emission
objective functions simultaneously due to find an operating point that strikes a balance between operation costs and
pollutions. The idea behind the combined economic and emission dispatch (CEED) is to compute the optimal
generation for individual units of the power system by minimizing the fuel cost and emission levels simultaneously
subject to various system constraints. Different methods have been used to solve combined economic and
environmental dispatch problem: Lagrange multiplier methods and multi-objective stochastic search technique are
used in order to solve combined problems. Hesli and Hobbs [6] presented a model for evaluating the cost and the
employment impact due to dispatching and fuel switching as a means of reducing emissions. Granelli et al. [6]
proposed an emission constrained dynamic dispatch algorithm. It minimizes fuel cost during
a pre-selected time horizon and thoroughly takes into account the environmental
constraints. A price penalty factor for solving the problem which combines the emission
costs with the normal fuel costs was proposed by Kulkarni et al. [8]. Wong [9] proposed a
cooling mutation technique in evolutionary algorithm to solve the problem for a 9-unit
system. Evolutionary algorithms vastly used for solving multi-objective problems; Genetic algorithm and
simulated are very popular. Xuebin [1] has proposed a hybrid approach to solve the problem. Some researchers
proposed methods to convert a multi-objective problem into a single objective problem by assigning different
weights to each objective. This allows a simpler minimization process but does require the knowledge of the relative
importance of each objective.This paper presents a method to consider the fuel costs and environmental emissions
simultaneously. The elasticity of cost with respect to the emission (say, emission elasticity) has been estimated. It is
invaluable indicator in multi-objective circumstance. The operator can compare emission sensitivity in different fuel
system while the problem is optimized subject to both operation costs and environmental pollutants to fulfill
exogenous demand.
Method:
Problems with multiple objectives and criteria are generally known as multiple criteria
optimization or multiple criteria decision-making (MCDM) problems. The area of multiple
criteria decision making has developed rapidly. For electric power generation and
dispatching problems, absolute minimum cost is not any more the only criterion to be met.
Environmental considerations have become one of the major management concerns. A more
flexible formulation of the problem can be accomplished by stating the problem in a manner
of multi-objective optimization. More formally, a multi-objective nonlinear programming
model for the ECED is given as follows:

Minimize F j ( TC j , E j ) for j=1 M

Subject to:
M

PD + P L = P j
j

P jmin P j P jmax

Where

PD is the total system demand (MW),

minimum power output limit of

PL is the total transmission network loss (MW),

j th generator (MW) and

P jmin is

P jmax is the maximum power output limit of

th generator (MW). The word minimize means that we want to minimize all the objective functions, namely the
total fuel cost (

TC j ) and environmental emissions ( E j ) simultaneously. In this paper, the

-constraint

technique is utilized for optimization problem to obtain Pareto front (or, Pareto optimal solutions). The method
generates sub-problems, called

-constraint problems, by transforming objectives into constraints. The upper

bounds of these constraints are given by

-vector and by varying the

-vector, the Pareto front can

theoretically be generated.
The elasticity of cost with respect to emission measures the percentage change in fuel costs caused by a percent
change in particular emission. Mathematically the elasticity of cost with respect to emission would be as follows:

ln( E j )
ln ( TC j )

Cj =
For j=1 M
The above elasticity is calculated over all Pareto front. The elasticity for an operating point can determine the
sensitivity of the power system to variation of equilibrium point.
This information is invaluable, due to contradiction nature of objectives in combined economic and emission
problem.
Results:
Here, based on the above problem formulation, the combined economic/environmental problem for regional
electricity market in Iran has been implemented. First of all, we have estimated cost function by regression analysis
based on the above formulation where the input is in Mcl/hr and the output is in MW. The output data indicates
gross hourly production in MWh and the inputs are the amount of fuel consumption in Mega Calorie (MCl) in each
hour. All data is provided by Isfahan Regional Electricity Company. Similarly, we have estimate emission function
for power plants. All data have been provided by Iranian Power Generation and Transmission (TAVANIR) holding
Company. The output data indicates the amount of pollutant (in ton) released by the generator and the inputs are the
amount of fuel consumption in Mega Calorie (MCl). The elasticity of cost respect to emission is shown in Figures 1.

25.00
20.00
15.00

Elasticity of cost 10.00

5.00
0.00
1,600

1,650

1,700

Emission (kg/hr)

1,750

Based on the results, in average, one percent decrease in emission released from natural gas, increases fuel cost by 5
percent. Moreover, elasticity of costs respect to emission for wide range of emission is less than 5. It shows that
reducing one present pollutant emissions increases cost less than 5 percent. The elasticity of costs provides
additional information for planners.
Conclusion:
In this paper, elasticity of cost as a practical indicator is developed by formulating combined economic and
environmental problem. A

-constraint technique is used to generate Pareto optimal solutions.

The power system operator needs particular information to adjust the operation point regarding environmental and
economic considerations. The proposed model facilitates policy making consideration based on the system
restrictions.
The proposed model is used for Isfahan regional electricity market including four power plants as a real used case.
The results show that, decision maker should use two different strategies to achieve satisfactory consequences. In
addition, the elasticity of cost analysis determines the preferred solution regarding cost and emission trade-off.