InternationalJournalofEnergyScience(IJES)Volume4Issue1,February2014

doi:10.14355/ijes.2014.0401.01

www.ijesci.org

OptimumReplacementPoliciesof
ConventionalEnergySourcesbyCleanOnes
SubjecttoCurrentElectricityDemand
MaryamParsa1,NasirUddinAhmed*2,MustaphaC.E.Yagoub3
SchoolofElectricalEngineeringandComputerScience,UniversityofOttawa,ON,Canada
mparsa@purdue.edu;*2ahmed@site.uottawa.ca;3myagoub@site.uottawa.ca

Abstract
Developing optimum decision policy for gradual
replacementofconventionalenergysourcesbycleanonesis
animportantfieldofcurrentresearchandthemainconcern
ofthisarticle.Eightmainenergysourcesareconsideredand
the decision policies are formulated with the objective of
minimizing the implementation and environmental costs
whilemeetingtheelectricitydemandduringtheentireplan
period. The selected energy sources are Coal, Petroleum,
Natural Gas, Hydropower, Wind, Solar, Geothermal and
Biomass.
AnonlineardynamicLotkaVolterramodel,firstintroduced
in Miah, Ahmed and Chowdhury [1] and later extended in
Parsa, Ahmed and Yagoub [2], is used for modeling the
dynamic changes in the level of electricity generation from
each of the eight available energy sources. Optimal control
theory is used to find the optimum decision policies for
integration of renewable energy sources into the national
powergirdofanycountry.
As a case study for our numerical results, official released
data of United States Energy Information Administration
website is used for the level of electricity generation from
each of the energy sources mentioned above. Different
scenarios are considered for the electricity demand. These
range from U.S prediction for twenty years plan period to
twopercentannualgrowthratefordifferentplanperiodsof
twenty and thirty years. The proposed methodology is
generalenoughandhenceappliestootherenergyproblems
withslightmodificationsbasedontheplannersobjectives.
Keywords
Mathematical Models for Energy Systems; Optimum Decision
Policy;ElectricityDemandandEnvironmentalConcerns

Introduction
Withtheincreaseofworldpopulation,andgrowthof
the industrial complex in developing countries,
electricity demand continues to rise. Associated with
therisingdemand,isasignificantincreaseinthelevel
of Green House Gas emissions due to electricity

generationfrompollutingenergysourcessuchasCoal
and Petroleum thereby threatening health of the
generalpopulation.
In this paper, we consider electricity generation from
all the eight main energy sources and use the
Pontryagin Minimum Principle to develop the
optimumdecisionpoliciesforthegradualreplacement
of conventional energy sources by clean ones. The
decision policies for this transition are found while
satisfying the electricity demand during the plan
period and keeping the implementation and
environmental costs as low as possible. Coal,
Petroleum, Natural Gas, Hydropower, Wind, Solar,
Geothermal, Wood and Biomass are the eight main
energysourcesthatareconsideredinourmodel.
In a recent paper of the authors [2], a similar model
and method was proposed for different energy
problems, satisfying the desired levels of electricity
generation from each of the energy sources while
minimizing the environmental and investment costs.
In contrast, in this paper integration of renewable
energy sources into the power grid is accomplished
while the demand during the plan period is met and
the environmental and implementation costs are kept
toaminimum.
Use of the LotkaVolterra model in energy systems
andoptimalcontroltheorywasproposedforthefirst
time in [1]. There energy sources were classified into
two categories: polluting and nonpolluting. Here we
consideralltheeightprincipalenergysourcessomeof
which are conventional and some renewable and
present optimal policy of electricity generation from
eachofthesources.
This article is mainly a continuation of our previous
paper [2], with emphasis on satisfying the electricity
demandduringeachtimeintervaloftheplanperiod.
The official released data of U.S. Energy Information

www.ijesci.orgInternationalJournalofEnergyScience(IJES)Volume4Issue1,February2014

Administration website [3] is used to validate the

model.Asetofnonlineardifferentialequationsbased
on LotkaVolterra type interaction is presented ([4],
p24)asthedynamicmodelforourenergygeneration
and regulation problem. Optimal policies are
constructed using the Pontryagin Minimum Principle
ofOptimalControlTheory([1,2,4]).

compared to demand, environmental cost (in terms

health and climate) and the cost of investment for
programimplementation.Alltheseconcernsaretaken
intoaccountinthefollowing(cost)function(2):

tf

Restofthepaperisorganizedasfollows.Thedynamic
model of the system and the proposed cost function
are presented in section II followed by the system
identification in section III. The problem of
determining the optimum decision policies and the
corresponding results for different scenarios are
discussedinsectionIV.Conclusionandfutureworkof
thepaperarepresentedinsectionsVandVI.

The nonlinear dynamic model based on Lotka

Volterrasystemisutilizedinthisarticle.Thismodelis
very similar to the one presented in [2]. For
convenience of the reader, this model is also
reproducedherebriefly.

j1, ji

x x j , i 1, 2,...,8

ij i

(1)

F(xi (t)) vi xi (t) , v 0

where xi denotes the level of electricity generation

from energy source i, ui represents the percentage
growth rate of energy source i, and the ij stands for
the interactions between the level of electricity
generationfromenergysourceiandj.Inotherwords,
ij denotes the impact of the level of electricity
generation from energy source j on the level of
electricitygenerationfromenergysourcei.

(3)

According to experimental data given in [5], for the

same level of electricity generation, Petroleum
produces1.23timesmoreCO2thanCoal.Thisrelation
is used for the weights i in our numerical results
section. The factor i (> 0) denotes the penalty
(proportional to the cost of public health and climate
degradation) associated with the production of
electricityusingthepollutingsourcei.

In this model, the variables ui are considered as the

decision policies, that take values from the set U={u:
|ui| 1, i = 1,, 8}, where ui = 1 means a hundred
percent rate of increase and ui = 1 means a hundred
percent rate of decrease in the level of electricity
generationfromenergysourcei.

The expression (2), as shown above, consist of three

components. The first component is dependent on
mismatchbetweendemandandsupplyofelectricity aimed
at minimizing the difference between the level of
electricitygeneratedandtheactualelectricitydemand
during the plan period. The second component
representing {Environmental Cost} is aimed at
minimizing the electricity production from the
polluting energy sources (Coal and Petroleum). The
thirdtermrepresents{ImplementationCost},andthisis

CostFunction
The cost function must represent all the concerns of
the planner. In this paper the planner is concerned
with the cost of insufficient production of energy

(2)

In general, F is any suitable monotonically increasing

function of its argument representing the damage
caused to the environment due to production of
greenhouse gasses. For our numerical computations
wehavechosenFasfollows:

The dynamic system is given by a system of eight

nonlinear interconnected differential equations as
follows:

tf

HereD(t)standsfortheelectricitydemandduringthe
plan period and the quantity given by the sum
represents the total energy actually produced. In this
article we consider two different profiles for the
electricity demand. One is based on the U.S. EIA [3]
predictionandtheotherisbasedontheassumptionof
two percent annual growth rate. In fact the planner
canchooseanydemandfunctionbasedonprojections
ofpopulationandindustrialgrowth.InEquation(2),
(>0)istheweight(orimportance)giventoamismatch
between the supply and demand. The factor qi (> 0)
represents the investment cost per unit change of
existinginfrastructurebynewones.

NonLinearDynamicLotkaVolterraModel

1
1
{ F(xi (t))} dt { qi ui (t))2 } dt

2 t0 i1
2 t0 i1

Problem Description

xi ui xi

8
1 f
{(D(t)

xi (t))2 } dt

2 t0
i1
t

J(u)

InternationalJournalofEnergyScience(IJES)Volume4Issue1,February2014

Optimum Decision Policy

aimed at keeping implementation costs as low as

possiblegiventhelimitedresources.

Consider the system Equation (1), where all the fifty

sixunknownparametersareidentifiedandgivenin
Table 2. For this system, we introduced the objective
functiongivenbyEquation(2).Optimalcontroltheory,
inparticular,thePontryaginMinimumPrinciple[1,2,
4,7],isusedtodeterminetheoptimumdecisionpolicy.
The minimum principle minimizes the cost function
given by the expression (2) subject to the dynamic
constraint(1).

Theobjectiveistodeterminethedecisionpoliciesui(t),
that minimizes the cost function J(u), expressed by
Equation(2).
System Identification
Therearefiftysixunknownparameters={i,j}inthe
model expressed by Equation (1). Finding the
optimum decision policies would be impossible
without precisely identifying the system model.
Therefore, performing the system identification is the
firststep.Thisidentificationcanalsobecarriedoutby
useofoptimalcontroltheory.UsingthebasicOptimal
Control Theory [4] system identification was carried
out in our paper [2]. The required historicaldata was
taken from U.S. energy Administration website [6] as
presented in Table 1. All the fiftysix unknown
parameterswereidentifiedandgiveninTable2.

Details of the mathematical procedure and the

methodology for computation of optimum decision
policyaregivenin[2].Herewehaveconsideredthree
differentscenariosfortheelectricitydemandfunction
D(t). In first scenario we use the electricity demand
predictedbyU.S.EIA[3]fortwentyyearsplanperiod.
Inthesecondandthirdscenarioweassume2%annual
growth rate for the electricity demand for the twenty
yearsplanperiod,andforthethirtyyearsplanperiod.
In the computational scheme developed here the
planner can easily choose any given trend for
electricitydemandandthedurationoftheplanperiod
asnecessary.

TABLE1GENERATIONLEVELOFELECTRICPOWERSECTORS(TRILLION

KILOWATTHOURS)HISTORICALDATA[6]

Year2010

Year2015

Coal(x1)

1.831

1.562

Petroleum(x2)

0.034

0.026

NaturalGas(x3)

0.898

1.028

ConventionalHydropower(x4)

0.25532

0.29543

Wind(x5)

0.09449

0.15097

Solar(x6)

0.00128

0.00647

Geothermal(x7)

0.01567

0.01868

WoodandBiomass(x8)

0.01151

0.02128

TABLE2COMPUTEDVALUESOFPARAMETERS

The level of electricity generation is given by xi, i =

1,,8andthedecisionpolicyisgivenbyui,i=1,,8
for eight energy sources of Coal, Petroleum, Natural
Gas, Hydropower, Wind, Solar, Geothermal and
Biomass,respectively.Theweightingparametersqi,i
andaregiveninTable3.
TABLE3VALUESUSEDFORTHEWEIGHTS

AFTER10,000

ITERATIONS
ij

Value

ij

Value

ij

Value

ij

www.ijesci.org

Value

INEXPRESSION(2)

qi

Value

qi

Value

Value

Value

q1

10

q5

150

q2

q6

0.59

5*1.23

q3

q7

1.8

q4

q8

0.68

12

0.00226

23

0.00495

35

0.00676

48

0.00024

21

0.00971

32

0.00109

53

0.08901

84

0.02516

13

0.02771

24

0.00356

36

3.55e6

56

0.00010

Scenario1

31

0.02508

42

0.00072

63

0.03664

65

0.00385

14

0.00944

25

0.00046

37

4.35e5

57

0.00027

41

0.03931

52

0.00176

73

0.00750

75

0.00078

15

0.00460

26

9.28e5

38

3.20e5

58

0.00093

51

0.14887

62

0.00238

83

0.08850

85

0.00931

16

0.00089

27

1.20e5

45

0.00202

67

0.00063

61

0.07470

72

0.00038

54

0.02075

76

1.07e5

17

0.00073

28

3.53e5

46

2.75e5

68

0.00046

Inthisscenario,weconsideredtheelectricitydemand
basedontheU.S.prediction[3]asshowninFigure1.
The value of in Equation (2) is assumed as 150 and
given in Table 3. This assumption was based on the
high priority of meeting the electricity demand. The
plannermayeasilyreducethisvalueif,fromhispoint
of view, the priority of implementation and
environmental costs is higher than the electricity
demand.

71

0.01529

82

0.00735

64

0.01041

86

0.00012

18

0.00407

34

0.00070

47

0.00033

78

0.00009

81

0.18045

43

0.01928

74

0.00213

87

0.00054

The state trajectories representing the level of

electricity generation from each of the energy sources
are shown in Figure 1. As expected, the level of

www.ijesci.orgInternationalJournalofEnergyScience(IJES)Volume4Issue1,February2014

electricity production from polluted energy sources

(Coal and Petroleum) is decreased by the end of the
plan period. It is interesting to note that to meet the
demandduringearlyperiodsoftheplanninghorizon
the level of electricity production from Coal is
increased.Thisisbecausethenewsourcesarenotyet
developed and the demand is met by producing
sufficient energy from the retiring sources. It is
expected that the total energy produced from all the
eight sources meet the demand over the plan period.
ThisisshowninTable4.

decreasing as shown (Figure 1). Correspondingly the

decision policy (u1) is initially positive and then
negative as shown in Figure 2. The value of the cost
functionafter100,000iterationsisshowninFigure6a.
Figure 6b shows the converging trend of this cost
functionfromacloserviewinthefirst1000iterations.

(a)

(b)

FIGURE2.SCENARIO1,20YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)COALAND(b)PETROLEUMENERGYSOURCES

(a)

(b)

FIGURE1.SCENARIO1,20YEARSPLANPERIOD,(a)DEMANDTREND
ANDSTATETRAJECTORIESFORCOAL,NATURALGAS,HYDRO,WIND
AND(b)STATETRAJECTORIESFORPETROLEUM,SOLAR,GEOTHERMAL
ANDBIOMASS
TABLE4SCENARIO1,20YEARSPLANPERIOD,COMPARISONBETWEEN

THEELECTRICITYDEMANDANDTHEELECTRICITYGENERATEDFROM
EACHENERGYSOURCEDURINGTHEPLANPERIOD(TRILLIONKWH)

xi

Year
2015

Year
2020

Year
2025

Year
2030

Year
2035

x1

1.5620

0.7909

0.3833

0.2095

0.1211

x2

0.0260

0.0180

0.0127

0.0090

0.0065

x3

1.0280

2.7430

3.1580

3.0980

2.7992

x4

0.2954

0.4181

0.5171

0.6563

0.9110

x5

0.1510

0.1760

0.2264

0.3301

0.5632

x6

0.0065

0.0118

0.0235

0.0495

0.1098

x7

0.0187

0.0255

0.0348

0.0481

0.0669

x8

0.0213

0.0295

0.0512

0.1039

0.2461

3.1089

4.2128

4.4070

4.5044

4.8238

3.9000

4.2100

4.4050

4.5040

4.7000

Demand

(a)

(b)

FIGURE3.SCENARIO1,20YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)NATURALGASAND(b)HYDROPOWERENERGYSOURCES

(a)

(b)

FIGURE4.SCENARIO1,20YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)WINDAND(b)SOLARENERGYSOURCES

Decision policies for generation levels of the eight

energy sources during the plan period are shown in
Figures 25. For example, energy production from
Coal given by the trajectory (x1) is increasing during
the early periods of the plan period and then

(a)

(b)

FIGURE5.SCENARIO1,20YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)GEOTHERMALAND(b)WOODANDBIOMASSENERGYSOURCES

InternationalJournalofEnergyScience(IJES)Volume4Issue1,February2014

x5
x6
x7
x8

Demand

(a)

www.ijesci.org

0.1510
0.0065
0.0187
0.0213

0.1996
0.0162
0.0301
0.0394

0.2844
0.0448
0.0498
0.0874

0.4475
0.1287
0.0835
0.2118

0.9262
0.4239
0.1461
0.6746

3.1089

4.3016

4.7556

5.2045

5.8525

3.8580

4.2980

4.7500

5.2020

5.6660

Figures811showthecorrespondingdecisionpolicies.
Comparing this with scenario 1, we observe that for
some of the energy sources larger efforts (in terms of
decision policies) are required to satisfy the greater
demand(inscenario2).ThisisseeninFigure5(b)for
scenario 1, where for the biomass energy source the
growth rate is around 50% and this is changed to
around 58% in scenario 2 as shown in Figure 11(b).
Similarly, for geothermal energy source, the growth
rateinscenario1isaround10%whileitisaround15%
in scenario 2. The converging trend of the cost as a
functionofiterationisshowninFigure12.

(b)

FIGURE6.SCENARIO1,20YEARSPLANPERIOD,(a)COSTFUNCTION
AFTER100,000ITERATIONSAND(b)ZOOMEDCOSTFUNCTIONAFTER
FIRST1000ITERATIONS

Scenario2
Let us consider the case where the annual growth of
electricitydemandis2%.Inthiscase,weexpectlarger
values (greater efforts) for the decision policies to
satisfy this increased demand, compared to the
previousscenario,wherethemeanannualgrowthrate
ofdemandwasonly0.958%.Thevaluesoftheweights
{qi ,i = 1,,8; i , i =1,2; } for the cost function
(Equation(2))aregiveninTable3.
Figure 7 shows the electricity demand and the state
trajectories for all energy sources during the twenty
years plan period. This demand is satisfied by
increasing the level of electricity generated from the
renewable energy sources due to lower
implementation costs as shown in Table 3. Table 5
showsthatduringtheplanperiodoftwentyyears,the
sum of all the eight energy sources satisfies the total
demandasrecordedinthelasttworows.

(a)

FIGURE8.SCENARIO2,20YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)COALAND(b)PETROLEUMENERGYSOURCES

(a)
(a)

(b)

(b)

(b)

FIGURE9.SCENARIO2,20YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)NATURALGASAND(b)HYDROPOWERENERGYSOURCES

FIGURE7.SCENARIO2,20YEARSPLANPERIOD,(a)DEMANDTREND
ANDSTATETRAJECTORIESFORCOAL,NATURALGAS,HYDRO,WIND
AND(b)STATETRAJECTORIESFORPETROLEUM,SOLAR,GEOTHERMAL
ANDBIOMASS
TABLE5SCENARIO2,20YEARSPLANPERIOD,COMPARISONBETWEEN

THEELECTRICITYDEMANDANDTHEELECTRICITYGENERATEDFROM
EACHENERGYSOURCEDURINGTHEPLANPERIOD(TRILLIONKWH)
xi
x1
x2
x3
x4

Year
2015
1.5620
0.0260
1.0280
0.2954

Year
2020
0.9266
0.0183
2.7020
0.3694

Year
2025
0.4772
0.0128
3.3700
0.4292

Year
2030
0.2738
0.0089
3.5240
0.5263

Year
2035
0.1443
0.0063
2.8355
0.6956

(a)

(b)

FIGURE10.SCENARIO2,20YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)WINDAND(b)SOLARENERGYSOURCES

www.ijesci.orgInternationalJournalofEnergyScience(IJES)Volume4Issue1,February2014

(a)

(a)

(b)

FIGURE11.SCENARIO2,20YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)GEOTHERMALAND(b)WOODANDBIOMASSENERGYSOURCES

(b)

FIGURE14.SCENARIO3,30YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)COALAND(b)PETROLEUMENERGYSOURCES
TABLE6SCENARIO3,30YEARSPLANPERIOD,COMPARISONBETWEEN

THEELECTRICITYDEMANDANDTHEELECTRICITYGENERATEDFROM
EACHENERGYSOURCEDURINGTHEPLANPERIOD(TRILLIONKWH)

(a)

xi

Year
2015

Year
2020

Year
2025

Year
2030

Year
2035

Year
2040

Year
2045

x1

1.5620

1.0130

0.5352

0.3134

0.1847

0.1112

0.0579

x2

0.0260

0.0182

0.0125

0.0085

0.0058

0.0042

0.0029

x3

1.0280

2.7410

3.6200

4.1590

4.4530

4.3980

3.4192

x4

0.2954

0.2808

0.2583

0.2534

0.2792

0.3220

0.4300

x5

0.1510

0.1697

0.2017

0.2534

0.3321

0.4573

0.8745

x6

0.0065

0.0127

0.0270

0.0596

0.1325

0.2734

0.7289

x7

0.0187

0.0267

0.0393

0.0596

0.0921

0.1372

0.2250

x8

0.0213

0.0341

0.0626

0.1188

0.2229

0.4035

0.0555

3.1089

4.2962

4.7566

5.2257

5.7023

6.1068

6.7944

Deman
d

3.8580

4.2930

4.7500

5.2250

5.6990

6.1060

6.5690

(b)

FIGURE12.SCENARIO2,20YEARSPLANPERIOD,(a)COSTFUNCTION
AFTER100,000ITERATIONSAND(b)ZOOMEDCOSTFUNCTIONAFTER
FIRST1000ITERATIONS

Scenario3

In this scenario, we extend the plan period to thirty

years. In this case, again we consider 2% annual
growth rate of electricity demand. The demand trend
and the state trajectories for all the eight energy
sources are shown in Figure 13. It is interesting to
observethatthedemandisverywellsatisfiedduring
the entire plan period as shown in Table 6. The
corresponding optimum decision policies for all the
eight energy sources are shown in Figures 1418
(includingtheconvergenceofthecost).
It may be interesting to note that by increasing the
plan period from twenty years (scenario 2) to thirty
years (scenario 3), the annual investment (or
implementation)costisreducedasshowninTable7.

(a)

(b)

FIGURE15.SCENARIO3,30YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)NATURALGASAND(b)HYDROPOWERENERGYSOURCES

(a)

(b)

FIGURE13.SCENARIO3,30YEARSPLANPERIOD,(a)DEMANDTREND
ANDSTATETRAJECTORIESFORCOAL,NATURALGAS,HYDRO,WIND
AND(b)STATETRAJECTORIESFORPETROLEUM,SOLAR,GEOTHERMAL
ANDBIOMASS

(a)

(b)

FIGURE16.SCENARIO3,30YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)WINDAND(b)SOLARENERGYSOURCES

InternationalJournalofEnergyScience(IJES)Volume4Issue1,February2014

www.ijesci.org

resolved. The results presented are completely

compatible with what is expected. The level of
electricity generation from polluting energy sources
such as Coal and Petroleum is reduced without
compromisingdemand.

(a)

Different scenarios for electricity demand are

presented. One of these is based on the official
prediction for the electricity demand during 20 years
planperiodintheUnitedStates(around0.95%annual
growth). Another scenario is based on unexpected
increase in demand, for example, 2% annual growth.
Thisiscarriedoutfordifferentplanperiodssuchas20
yearsand30years. The methodologydeveloped here
canbeadaptedtoanygivensituationforanycountry.
The planner has to only change the basic data to suit
theobjectivesofthecountry.

(b)

FIGURE17.SCENARIO3,30YEARSPLANPERIOD,DECISIONPOLICIESFOR
(a)GEOTHERMALAND(b)WOODANDBIOMASSENERGYSOURCES

Future Work

(a)

(b)

The research presented in this article is based on the

U.S. prediction during the plan period of 20152035.
However,thispredictionmaychangethroughoutthis
periodduetouncertaintiessuchas,economicgrowth,
population increase, retirement of existing power
plants, etc. Such uncertainties could affect the
projection of electricity demand and in turn might
change the generation policies. This point could be
addressed as future work using Stochastic Optimal
ControlTheory.

FIGURE18.SCENARIO3,30YEARSPLANPERIOD,(a)COSTFUNCTION
AFTER100,000ITERATIONSAND(b)ZOOMEDCOSTFUNCTIONAFTER
FIRST1000ITERATIONS
TABLE7COMPARISONBETWEENTHECUMUATIVEANDANNUAL

IMPLEMENTATIONCOSTSOFSCENARIO2ANDSCENARIO3

Scenario2

Scenario3

PlanPeriod

20years

30years

CumulativeImplementationCost

70.0801

74.6538

AnnualImplementationCost

3.5040

2.4885

REFERENCES

Conclusion

[1] M.

Miah,

NasirUddin

Ahmed,

Monjur

Chowdhury, Optimum policy for integration of

The methodology for development of optimum

decision policy for integration of renewable energy
sourcesintothenationalpowergridofanycountryis
presented. As a case study we consider the United
States since in this case complete historical data are
availableinthepublicdomain.Alleightmainenergy
sourcesofCoal,Petroleum,NaturalGas,Hydropower,
Wind, Solar, Geothermal and Biomass are considered
in a single nonlinear dynamic model. Pontryagin
Minimum Principle, which is well known to control
theorists, is utilized to determine the optimum
decision policies. All the costs of program
implementation,environmentalcostsanddemandare
considered in a general objective function (cost
function).

renewable energy sources into the power generation

system, Energy Economics 34, 558567, 2012, doi:
10.1016/j.eneco.2011.08.002
[2] Maryam Parsa, NasirUddin Ahmed, Mustapha C.E.
Yagoub.OptimumDecisionPolicyForReplacementof
Conventional Energy Sources by Renewable Ones.
InternationalJournalofEnergyScience(IJES)3(5),311319,
2013.doi:10.14355/ijes.2013.0305.03
[3] U.S. Energy Information Administration Official
Website,www.eia.gov
[4] NasirUddin Ahmed, Elements of Finite Dimensional
System and Control Theory, Pitman Monographs and
Surveys in Pure and Applied Mathematics, Longman

Theproblemofmeetingtheelectricitydemandduring
theplanperiod,whileminimizingtheimplementation
costs and the environmental costs, is successfully

Suruz

ScientificandTechnicalwithJohnWiley,London,New
York,Vol.37,1988

www.ijesci.orgInternationalJournalofEnergyScience(IJES)Volume4Issue1,February2014

[5] U.S. Energy Information Administration, Environment

12).pdf, Released Date: June 2012, Accessed September

Analysis and Projections, U.S. Energy Related Carbon

Dioxide

Emissions,

12,2012

2011,

[7] Donald E. Kirk, Optimal Control Theory, An

http://www.eia.gov/environment/emissions/carbon/,

Introduction, Dover publications, Inc., Mineola, New

AccessedSeptember12,2012

York,2004

[6] AnnualEnergyOutlook2012withprojectionsto2035,
U.S.

Energy

Information

[8] Xuezhong Wang, Solving Optimal Control Problems

withMATLABIndirectMethods,November2009

Administration,

http://www.eia.gov/forecasts/archive/aeo12/pdf/0383(20