IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 18, NO. 4, NOVEMBER 2003 1275

Analysis of Energy Pricing in Urban Energy ServiceSystems Considering a Multiobjective Problem of

Environmental and Economic ImpactHirohisa Aki, Member, IEEE, Tsutomu Oyama, Member, IEEE, and Kiichiro Tsuji, Member, IEEE

AbstractIntroduction of an integrated energy service systemin an urban area is assumed. An energy supply plant is installed inthe area to provide integrated energy service. It supplies electricity,gas, cooling, and heating to consumers.

To consider CO2

emission constraint in the area, analyses ofenergy pricing, economic impact on energy consumers, and oper-ation of the system under the constraint of CO2

emission wereperformed.CO

2

emission and economic impact on the consumers andthe supplier of various energy pricing scenarios were calculatedusing linear programming models. The solution that satisfied theCO

2

emission restriction, and was economically optimal to theconsumers was chosen from the results. The chosen solutionsare Pareto optimum solutions of a multiobjective problem thatconcerns both CO2

emission and cost to the consumers.

Index TermsCO2

mitigation, economic impact, energypricing, energy system, multiobjective problem.

NOMENCLATURE

Season (middle, summer, winter).Hour.Cost of equipment.Cost of energy.Depreciation cost of equipment.Maintenance cost of equipment.Capacity of equipment.Base charge for energy.Meter charge for energy.Maximum energy supply from EP.Annual energy supply from EP.EP electricity supply.EP gas supply.EP cooling supply.EP heating supply.EP heating supply for space heating.EP heating supply for water heating.Electricity demand.

Manuscript received February 25, 2003.H. Aki is with the Energy Network Group, Energy Electronics Institute,

National Institute of Advanced Industrial Science and Technology (AIST),Tsukuba 305-8568, Japan.

T. Oyama is with the Department of Electrical and Computer Engineering,Faculty of Engineering, Yokohama National University, Yokohama 240-8501,Japan.

K. Tsuji is with the Department of Electrical Engineering, Graduate Schoolof Engineering, Osaka University, Osaka 565-0871, Japan.

Digital Object Identifier 10.1109/TPWRS.2003.818599

Gas demand.Cooling demand.Heating demand.Hot water demand.Cooking demand.EP total electricity supply.EP total gas supply.EP total cooling supply.EP total heating supply.EP electricity purchases from outside.EP gas purchases from outside.MGT generation.MGT heat recovery.MGT heat recovery for heating.MGT heat recovery for water heating.GE generation.GE heat recovery.Gas turbine generation.Gas turbine heat recovery.Heat pump cooling output.Heat pump heating output.Gas heater output.Gas boiler output.Gas stove output.Absorption refrigerator output.Electric turbo refrigerator output.Steam turbo refrigerator output.COP of heat pump (cooling operation).COP of heat pump (heating operation).Efficiency of MGT (generation).Efficiency of MGT (heat recovery).Efficiency of gas engine (generation).Efficiency of gas engine (heat recovery).Efficiency of gas turbine (generation).Efficiency of gas turbine (heat recovery).COP of absorption refrigerator.COP of electric turbo refrigerator.COP of steam turbo refrigerator.

I. INTRODUCTION

GLOBAL environmental issues such as the greenhouseeffect resulting from energy consumption are attractingconsiderable attention. On the other hand, the movement

0885-8950/03$17.00 2003 IEEE

1276 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 18, NO. 4, NOVEMBER 2003

toward deregulation is accelerating restructuring in the energybusiness field.

In this paper, the introduction of an integrated energy servicesystem into an urban area of Japan is assumed. Introduction ofsuch alternative systems offers the possibility of mitigation ofenvironmental impact including emissions [1].

An energy supply plant (EP) is assumed to be installed in thecentral part of the area, and has the responsibility and obliga-tion of energy supply in the area. It supplies electricity, gas,cooling, and heating by operating a combined heat and power(CHP) system. Energy consumers can combine energy supplyfrom the plant and their own energy equipment such as CHP, airconditioners, etc.

Discussions on the influence of economic factors on the be-havior of energy suppliers and consumers are essential to eval-uate the effects of introduction of such energy service systems[2][4]. Energy pricing is considered to be an economic factorfor evaluating the effects of introducing and operating an inte-grated energy service system.

In this paper, energy pricing is analyzed considering environ-mental impact and economic impact. Optimum energy pricing isobtained as a Pareto solution by a multiobjective model consid-ering both emissions and economic impact on consumers.

To analyze the relationships between energy pricing and eval-uation indexes (environmental and economic impact), a modelfor calculation including linear programming models was con-structed. Applying these models, a multiobjective model forminimization of emissions and cost to consumers was alsoconstructed.

II. ENERGY SERVICE SYSTEM

A. Energy Service System of the AreaIn conventional urban energy systems, consumers are sup-

plied electricity and gas from electricity and gas utilities, re-spectively, and they operate their own energy equipment suchas heat pumps to satisfy their demand in Japan.

In this paper, introduction of the integrated energy servicesystem shown in Fig. 1 is assumed. An energy supply plant (EP)is installed in the central part of the area and supplies electricity,gas, cooling, and heating to consumers.

The consumers combine energy supply from the EP and en-ergy equipment to satisfy their final energy demand.

The EP purchases electricity and gas from outside of the area,and produces electricity, cooling, and heating for supply in thearea.

B. System ConfigurationEnergy equipment owned by the consumers and the EP is

shown in Table I. The abbreviations used for the equipment arelisted below the table.

Two types of system configurations (residential dwellingsand business facilities) are assumed as the consumers. The EPsupplies electricity, gas, cooling, and heating to them. Both theconsumers and the EP have distributed generations (DGs) used

Fig. 1. Energy service system.

TABLE ISYSTEM CONFIGURATION

in the CHP mode. So, many DGs (CHPs)1 are installed andoperated in the area.

Residential dwellings have microgas turbines (MGTs)as CHPs. Generated electricity is consumed not only forelectricity demand such as lighting but also by electric heatpumps. Reverse flow (selling of electricity) is permitted. Anelectric heat pump supplies both cooling and heating. It is analternative to cooling and heating supply from the EP. Exhaustheat is recovered and supplied to the absorption refrigerator forcooling or consumed for heating demand and hot water.

Business facilities also have CHPs. Restaurants and shopshave MGTs as CHPs, and other facilities have gas engines(GEs). These system configurations are typical in Japan.Generated electricity is consumed for electricity demand andelectric turbo refrigerators. Recovered heat is consumed forabsorption refrigerators, heating demand, and water heating.

1Distributed generation (DG) is a generating plant serving a customer on-siteor providing support to a distribution network, connected to the grid at distribu-tion-level voltages. It includes gas engines, micro-gas turbines, fuel cells, pho-tovoltaic systems, etc. [5]. CHP is the simultaneous production and delivery ofelectricity and heat [6].

AKI et al.: ANALYSIS OF ENERGY PRICING IN URBAN ENERGY SERVICE SYSTEMS 1277

(a)

(b)Fig. 2. Main model for calculation. (a) Main model for calculation; (b)Auxiliary drawing of main model.

The EP also has a CHP and generated electricity is supplied tothe consumers. Recovered heat is combined with the output ofthe gas boilers and consumed for turbo refrigerators and heatingdemand.

III. MODEL FOR CALCULATIONAn energy system model including linear programming

models (Figs. 2 and 3) was constructed to describe assumedenergy service systems, and various types of analyses wereperformed for energy pricing, environmental impact, andeconomic impact on the consumers and the supplier. The modelconsists of a main model (Fig. 2) and a submodel (Fig. 3).

A. Main ModelFig. 2(a) describes flow to obtain a Pareto optimum solution.

Fig. 2(b) demonstrates auxiliary explanation.Calculation flow of the main model to obtain a Pareto op-

timum solution is described below. Roman numerals correspondto numbers in Fig. 2.

i) Various energy pricing scenarios are assumed and inputto the submodel. emission , consumers cost

Fig. 3. Submodel for calculation.

, economic impact, etc. are calculated from the en-ergy pricing and other input data by the submodel. A setof calculation results is obtained by inputting various en-ergy pricings to the submodel. The set is plotted as shownin Fig. 2(b)-(i).

ii) Cases that fulfill the assumed emission constraint(i.e., ) are selected from the set of calculationresults.

iii) Optimum cases are chosen from the sets as the final step.The objective function in this study is minimization ofthe consumers cost. Therefore, the case, having the min-imum value for consumers cost is chosen for each

constraint.iv) ii) and iii) above are repeated as changing values for the

emission constraint . As the result, sets that cor-respond to the constraint values are obtained as closedcircles in Fig. 2(b)-(iv). The line in Fig. 2(b)-(iv) showsthat Pareto optimality2 is achieved by this procedure.

B. SubmodelThe submodel calculates the behavior of both the consumers

and the EP from the energy pricing and other input data. Theconsumers and the EP are assumed to behave rationally de-pending only on their economic impact. Uncertainties such aschange in demands, pricing, and other factors in the future arenot considered in this paper.

The calculation in the submodel consists of two steps (theconsumers and the EP).

In the first step, energy prices and the consumers energydemand are input to linear programming models for the con-sumers. The total energy demand that is supplied by the EP, totalcapacity, and operational strategies of energy equipment ownedby the consumers, and annual disbursement of the consumersare calculated.

The hourly and seasonal end-use energy demand of each fa-cility is assumed as the daily load curve per floor area. Seasonalvariations are represented by three typical days (summer, winter,and middle).

2The Pareto optimality is defined as a state, where it is impossible to improvean objective function without making other objective functions worse off. [7].

1278 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 18, NO. 4, NOVEMBER 2003

The energy pricing assumptions are described later(Table IV).

The total energy demand of the area is calculated by multi-plying and summing the calculation result for each facility andits floor area. This gives the energy demand that the EP has tosupply.

At the second step, the total energy demand of the area is inputto the linear programming model for the EP. Energy purchases(electricity and gas) from outside of the area, and profit and lossof the EP are calculated by the model.

By summarizing these calculation results, the emissionand economic indexes are obtained.

C. Linear Programming Models

Linear programming models have been developed for resi-dential dwellings, business facilities, and the EP.

Their objective function is minimization of annual cost. Theannual cost consists of energy cost and equipment cost, which isin proportion to the capacity of the equipment. The constraintsconsist of energy supply and demand, capacity of the equip-ment, and purchase of energy.

In general, not only cost minimization but also maximizationof profit can be considered to be the objective function of the EP.The EP is assumed to be obligated to supply all of the energydemand of the consumers. The income of the EP depends on theconsumers. Therefore, minimizing cost is the only measure toincrease profit for the EP.

Details of the models (equations) are attached as theAppendix.

IV. APPLICATION TO THE MODEL AREA

A. Description of the AreaA middle-sized urban area that includes two types of residen-

tial dwellings and five types of business facilities is assumed asthe model area. The configuration of the energy consumers isshown in Table II.

Daily load curves of the end-use demand of each type of fa-cility are also assumed [8], [9]. The load curves are various, de-pending on each consumer, and constantly change. However, itis appropriate to apply an average curve as a representative foranalyses that involve a large number of consumers, as in thisstudy.

Consumer demand is characterized by future uncertainty. Itis related to economic conditions, climate conditions, and con-sumers lifestyles. A marked increase or decrease of demand isnot expected in either residential dwellings or business facilitiesin Japan.

B. Energy Supply Equipment

Efficiencies (COP: coefficient of performance) and annualcosts of energy supply equipment are shown in Table III [10].The costs consist of depreciation and periodic maintenancecosts. It is assumed that they depend on the capacity of theequipment in this study.

TABLE IICONFIGURATION OF ENERGY CONSUMERS IN MODEL AREA

TABLE IIIEFFICIENCY (COP) AND ANNUAL COST OF EQUIPMENT

TABLE IVENERGY PRICES

C. Energy PricingAssumed energy prices are shown in Table IV. Each energy

price, except gas, consists of a base charge that is based on peakdemand through the year, and a meter charge that is in propor-tion to purchased energy.

The actual tariffs of an electricity utility [11] and a gas utility[12] in Japan were referenced for the assumptions.

The reverse charge (selling price) is less than half the metercharge (buying price) in Japan. However, the reverse charge forthe consumers is assumed to be the same as their buying price,

AKI et al.: ANALYSIS OF ENERGY PRICING IN URBAN ENERGY SERVICE SYSTEMS 1279

because reverse flow of surplus electricity from the consumer tothe EP does not reduce energy over the whole area.

Electricity and gas utilities have recently begun to reduceprices, reflecting deregulation of the energy market in Japan3[5], [13]. The prices shown in Table IV may therefore changein the future. However, the possibility that violent fluctuationswill occur is considered to be small.

The energy prices shown in Table IV are the standard prices.Assumptions of the meter charges paid by consumers to theEP are varied from 20% to 20% in increments of 5% (i.e.,

, standard price, ) to analyzethe effect on energy pricing.

D. Price ElasticityThe influence of price elasticity is also considered in this

study.In general, there are two types of price elasticity regarding the

relationship between goods and prices. These factors are oftenused in macroeconomic models.

The macroeconomic type calculation is applied only to self-elasticity in this study, although calculations are performed ba-sically by bottom-up models.

The cross-elasticity factor is not explicitly treated in thisstudy. However, optimization is performed by combining fourtypes of energy, which realizes application of the concept ofcross-elasticity.

E. Emission Intensityemission is calculated by multiplying purchases of

energy from outside of the area by each emission intensity.Several parameters were considered for estimating theemission intensity of electricity from outside. The intensitydepends on the generation mix (thermal, nuclear, hydro, etc.),and the generation mix constantly changes. It is difficult todefine how much is reduced when 1 kWh of electricity issaved [14], [15]. The following three parameters were thereforeconsidered.

1) average intensity through a year;2) intensity of thermal plants;3) daytime intensity of thermal plants, and nighttime inten-

sity of nuclear plants.We calculated each of the above three cases. The result of

case 3) is described as a typical case in this paper. The values ofemission intensity used in this study are shown in Table V.

F. Indexes for Conventional SystemCalculation was performed for the conventional system to

provide an evaluation standard as a reference.The system configuration assumed is shown in Table VI. The

calculation results are shown in Table VII.

3For example, Tokyo Electric, which is the largest private electric utility in theworld, reduced their electric prices 5.1% for residential dwellings (not deregu-lated market) and about 1214% for large-scale buildings and hotels (deregu-lated market) at the beginning of fiscal year 2002, against penetration of onsitegenerations. Refer to [13] for more details.

TABLE VCO EMISSION INTENSITY

TABLE VICONFIGURATION OF CONVENTIONAL SYSTEM

TABLE VIICALCULATION RESULTS FOR CONVENTIONAL SYSTEM

V. ANALYSIS AND EVALUATION

The calculation results are shown in Figs. 4 to 6. The hori-zontal axis shows the emission allowance. The values indi-cated are the deviations from the values of the conventional case.As the values move toward the left on the horizontal axis, the

constraint becomes stricter (the emission allowancebecomes smaller).A. Optimum Energy Pricing

Optimum energy pricing is shown in Fig. 4. This is the Paretooptimum solution of the multiobjective problem that considers

emission and consumers cost. The consumers cost is alsoplotted with a solid line for reference.

The vertical axis shows the energy price (deviation fromthe standard price: %) and consumers cost (deviation from thevalues of the conventional case: %).

For example, in the case in which the emission al-lowance is 0%, electricity is found at the 5% energy pricewhile gas and heat are found at 5% and 20%, respectively.Optimum energy pricing for the case in which the con-straint is 0% is therefore the set of 5% for electricity, 5%for gas, and 20% for heat.

These solutions are summarized in Table VIII. Generally, ex-pensive energy pricing is chosen for a strict constraint.This creates an economic disadvantage for the consumer. Theelectricity price rises as the constraint become stricter.The gas price also rises, but the trend is not so clear. The heat(cooling and heating) price remains at a low level unless the

constraint becomes very strict. The effects of the prices ofelectricity and heat on emission and the consumers costare larger than those of the gas price.

1280 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 18, NO. 4, NOVEMBER 2003

Fig. 4. Optimum energy pricing.

Fig. 5. Economic evaluation.

B. Economic EvaluationThe economic indexes are shown in Fig. 5. The EPs profit

and social welfare are described as deviations from the case inwhich the standard price is applied.

The consumers cost increases monotonously as theconstraint become stricter. The increase rate is approximately78 600 yen/t- (288 000 yen/t-C).4 This is much higher thanthe emission cost currently being discussed in the contextof permits or a carbon tax.

The difference between the consumers costs for loose andstrict constraints is about 20%. This means that the con-sumers have to accept 20% higher payment to mitigateemission.

The other economic index for consumers is consumers sur-plus. The decrease of consumers surplus for a strict restrictionof emission shows that the consumers final energy de-mand is reduced by elasticity as the restriction becomesstricter. This is caused by expensive energy pricing to reduce

emission.It can be seen that some degree of economic disadvantage to

consumers is unavoidable to mitigate emission.The EPs profit also increases as the constraint becomes

stricter. Energy prices increase when the constraint be-4About U.S.$ 2 200/t-C (1 US$ = 130 yen).

Fig. 6. Other indexes. a) End-use demand; b) CO emission intensity.TABLE VIII

TYPICAL PRICING STRATEGY

comes stricter, increasing the income of the EP. On the otherhand, the change in the EPs expenditures is negligible. As aresult, the EPs profit increases as the constraint becomesstricter.

AKI et al.: ANALYSIS OF ENERGY PRICING IN URBAN ENERGY SERVICE SYSTEMS 1281

The EP has both an obligation and a monopoly with regardto energy supply. Therefore, the EP has the nature of a publicorganization. Considering this character of the EP, such an in-crease in profit should be returned to society, because the EPgains the profit without efforts.

Social welfare may be considered an index for evaluating thechange of economic impact on society to mitigate emis-sion.5 It is calculated as the sum of the consumers surplus andthe EPs profit. As shown in Fig. 5, it remains at almost zero.Only energy prices in the area are changed in this study. There-fore, cash flow in the subject area will change in the manner ofa zero-sum game.

C. Other IndexesThe direct factors for mitigation of emission are reduc-

tion of end-use demand [Fig. 6(a)] and improvement ofemission intensity [Fig. 6(b)].

Both indexes decrease as the emission constraint be-comes stricter.

Compared with the case of loose constraint, end-use demandis reduced by 9.6% and emission intensity is reduced by5.1% when the constraint is very strict.

As a measure for mitigation, the reduction of emis-sion intensity is more acceptable than the reduction of end-usedemand, because it involves no economic disadvantage to theconsumers. However, no economic disadvantage means no in-centive. Consumers will therefore unavoidably suffer some eco-nomic disadvantage.

VI. CONCLUSIONIn this paper, introduction of an integrated energy service

system into an urban area is assumed.Pareto optimum solutions for a multiobjective model that has

two objective functions, emission and consumers cost, areobtained.

The conclusions reached from the analysis are as follows.1) emission and economic impact greatly depend on

energy pricing. Therefore, not only the system configu-ration but also energy pricing should be considered whenintroducing alternative energy service systems.

2) The cost to consumers increases as the constraintbecomes strict. An appropriate strategy mix for mitigationof emission is necessary so as to avoid excessiveeconomic disadvantage to the consumers. The economicdisadvantage should be shared by all related individualsand organizations.

3) The increase in the EPs profit in the case of strictconstraint should be returned to society without spoilingthe incentives to mitigate emission.

APPENDIX

The mathematical formulations of the LP models used in thisstudy are as follows.

5As analyses were performed from the perspective of energy, utility and cashflow, external costs were not taken into consideration in this paper. So, the ben-efit gained by mitigating CO is not taken into account.

A. Residential DwellingsThe objective function is minimization of cost (1). The con-

straint consists of energy demand and supply (2)(9), capacityof equipment (10), and purchased energy (11).

Heating supply from the EP is not applicable to absorptionrefrigerators, because its temperature is not sufficiently high.Absorption refrigerators use only recovered heat of MGT (8).

Minimize Cost

(1)

Subject to

(2)

(3)

(4)

(5)

(6)(7)

(8)(9)

Capacity of Equip. Output of Equip.(s,t) (10)Max. of EP Supply EP Supply(s,t) (11)

B. Business FacilitiesThe objective function of business facilities is also cost min-

imization, as in the case of residential dwellings. Constraints(2)(9) are displaced by (12)(16). Equation (17) is applied, be-cause absorption refrigerators use only recovered heat or boileroutput as in (8).

(12)

(13)

(14)

1282 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 18, NO. 4, NOVEMBER 2003

(15)(16)(17)

C. EPThe objective function of the EP is also cost minimization.

Constraints (2)(9) or (12)(16) are displaced by (18)(21).

(18)

(19)

(20)

(21)

REFERENCES[1] K. Tsuji, M. Mizuno, E. Sugihara, K. Ito, R. Yokoyama, and T. Oyama,

Final Reports on Integrated Energy Service System, (in Japanese), ,Mar. 2002.

[2] H. Aki, T. Oyama, and K. Tsuji, Analysis on mitigation of environ-mental impact and its economic impact in the integrated energy supplysystems, in Proc. Int. Conf. Elect. Eng., July 2001, pp. 703707.

[3] , Analysis on mitigation effect of co emission by introduction ofalternative energy supply systems in urban area, in Proc. CIGRE Symp.,Brasilia, Brazil, May 2002.

[4] , Analysis on energy pricing, Co emission and economic impacton integrated energy supply systems in urban area (in Japanese), inProc. 12th Annu. Conf. Power and Energy Soc., Inst. Elect. Eng. Jpn.,Aug. 2001.

[5] Distributed Generation in Liberalised Electricity Markets: IEA, 2002.[6] Technology and Environmental Aspects of Advanced Co-Generation:

WEC, 1995, p. 2.[7] H. Nakayama and T. Taniya, Theory and Application of Multi-Objective

Programming (in Japanese): Corona, 1994.

[8] K. Tsuji, M. Mizuno, O. Saeki, S. Sano, and T. Ueno, Reports on theMonitoring of End-Use Demand for Residential Houses in the KansaiScience City (vol. 1), (in Japanese), , Mar. 2001.

[9] Natural Gas Co-generation Planning and Designing Manual 2000 (inJapanese): Japan Industrial Publishing, 2000.

[10] Price Data for Construction Cost Estimating (in Japanese): EconomicResearch Association, Dec. 1999.

[11] Tariff of the Kansai Electric Power Co. Inc.http://www.kepco.co.jp/ [On-line]

[12] Tariff of Osaka Gas Co., Ltd.http://www.osakagas.co.jp/gas-rate/menu/menu.htm [Online]

[13] Energy Forum vol. 5 2002 (in Japanese), 2002.[14] T. Ojima and T. Tanaka, Age of DSM (in Japanese), Japan: Waseda Univ.

Press, 1999, pp. 157158.[15] T. Kashiwagi, Micro Power Revolution (in Japanese): TBS Britannica,

2001, pp. 193195.

Hirohisa Aki (S00M02) received the B.S. andM.S. degrees in electrical engineering from OsakaUniversity, Japan, in 1994 and 1996, respectively,and the Ph.D. degree in electrical and computerengineering from Yokohama National University,Japan, in 2002.

Currently, he is with the National Institute ofAdvanced Industrial Science and Technology(AIST), Tskuba, Japan, in 2002. He was a PlantEngineer in electrical and instrumental design atMitsubishi Heavy Industries Co., Ltd., Yokohama,

Japan, from 1996 to 2001. His research interests include the analysis of urbanenergy systems and mitigation of environment impact.

Tsutomu Oyama (S78M84) received the B.S.,M.S., and Dr. Eng. degrees in electrical engineeringfrom the University of Tokyo, Tokyo, Japan, in 1978,1980, and 1983, respectively.

Currently, he is a Professor at Yokohama NationalUniversity, Japan, 1998.

Dr. Oyama is a member of CIGRE and the IEE ofJapan.

Kiichiro Tsuji (M73) is a Professor of electrical en-gineering at Osaka University, Japan. He received thePh.D. degree in systems engineering from the Grad-uate School of Engineering, Case Western ReserveUniversity, Cleveland, OH, in 1973.

His research interests include analysis, planning,and evaluation of urban energy systems, and controland analysis of electrical power systems.

Index:

CCC: 0-7803-5957-7/00/$10.00 2000 IEEE

ccc: 0-7803-5957-7/00/$10.00 2000 IEEE

cce: 0-7803-5957-7/00/$10.00 2000 IEEE

index:

INDEX:

ind: