1-s2.0-S0306261913007307-main

Applied Energy 113 (2014) 1301–1314

Contents lists available at ScienceDirect

Applied Energy

journal homepage: www.elsevier .com/locate /apenergy

Thermodynamic and economic analysis of polygeneration systemintegrating atmospheric pressure coal pyrolysis technology withcirculating fluidized bed power plant

0306-2619/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.apenergy.2013.08.086

⇑ Corresponding author. Tel.: +86 571 87952802; fax: +86 571 87951616.E-mail addresses: [email protected] (Z. Guo), [email protected]

(Q. Wang), [email protected] (M. Fang), [email protected] (Z. Luo), [email protected] (K. Cen).

Zhihang Guo, Qinhui Wang ⇑, Mengxiang Fang, Zhongyang Luo, Kefa CenState Key Laboratory of Clean Energy Utilization, Zhejiang University, Hangzhou 310027, PR China

h i g h l i g h t s

� A lignite pyrolysis-based polygeneration plant was proposed and modeled.� Polygeneration plant has a 9.04% point higher efficiency than CFB power plant.� Polygeneration plant increases ca. 14% point of IRR based on CFB power plant.� Electricity price rise makes polygeneration plant less competitive.

a r t i c l e i n f o

Article history:Received 16 April 2013Received in revised form 27 August 2013Accepted 30 August 2013Available online 27 September 2013

Keywords:Polygeneration systemCoal pyrolysisCirculating fluidized bed power plantMethanolThermodynamic and economic analysis

a b s t r a c t

Lignite-based polygeneration system has been considered as a feasible technology to realize clean andefficient utilization of coal resources. A newly polygeneration system has been proposed, featuring thecombination of a 2 � 300 MW circulating fluidized bed (CFB) power plant and atmospheric pressure flu-idized bed pyrolyzers. Xiaolongtan lignite is pyrolyzed in pyrolyzers. Pyrolyzed volatiles are further uti-lized for the co-generation of methanol, oil, and electricity, while char residues are fired in CFB boilers tomaintain the full load condition of boilers. Detailed system models were built, and the optimum opera-tion parameters of the polygeneration plant were sought. Technical and economic performances of opti-mum design of the polygeneration plant were analyzed and compared with those of the conventional CFBpower plant based on the evaluation of energy and exergy efficiency, internal rate of return (IRR), andpayback period. Results revealed that system efficiency and the IRR of the polygeneration plant are ca.9% and 14% points higher than those of the power plant, respectively. The study also analyzed the effectsof market fluctuations on the economic condition of the polygeneration plant, and found that prices offuel, material, and products have great impacts on the economic characteristics of the polygenerationplant. Polygeneration plant is more economic than CFB power plant even when prices fluctuate withina wide range. This paper provides a thorough evaluation of the polygeneration plant, and the study indi-cates that the proposed polygeneration plant has a bright prospect.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

China relies heavily on coal, which accounts for 81.3% and 71.9%of the energy supply and energy consumption respectively in 2010[1]. However, the majority of coal is directly fired for power gener-ation, which has low efficiency and gives rise to serious pollution.The exploration of novel efficient and clean coal utilization tech-nologies has drawn much attraction in recent years. Coal-basedpolygeneration technology offering synthetic fuels, chemical

products, and electricity, has been regarded as a promisingalternative to tackle current energy problems. To date, variouspolygeneration schemes of different configurations depending onthe feedstock and products have been proposed [2–17]. Yi et al.[2,3] reported a polygeneration system using coke-oven gas andcoal gasified gas and established a newly configuration with CO2

emission control. On the basis of integrated gasification combinedcycle (IGCC) technology, Liu et al. [5,6,16] developed coal-basedpolygeneration systems with different configurations, which re-vealed better overall performance and more flexibility than IGCC.Kreutz et al. [9] and Chiesa et al. [8] put forward a polygenerationtechnology using commercially ready technologies to producehydrogen and electricity with CO2 capture and storage. Yu et al.[10] investigated the performance of polygeneration processes

Nomenclature

A last period with a negative cumulative cash flowAj domestic factor of facility jB absolute value of cumulative cash flow at the end of the

period Abj scale factor of facility jBt net cash flow in year t of period AC annual net cash flow during the next period after ACF annual fuel cost, 106 $Cj capital investment of facility j, 106 $CM annual material cost, 106 $Cp annual product sales income, 106 $CRF ratio of annual investmentCt annual cash flow of the year tCVCoal caloric heat of coal, MJ/kgCVMeOH caloric heat of methanol, MJ/kgCVOil caloric heat of oil, MJ/kgEXcoal exergy of coal, MWEXElectricity exergy of electricity, MWEXMeOH exergy of methanol, MWEXOil exergy of oil, MWFCoal flow rate of coal, kg/sFj installation factor of facility jFMeOH flow rate of methanol, kg/sFOil flow rate of oil, kg/si discount rate, %Ir,j purchase cost of facility j in the reference scale, 106 $j facilitym total number of facilitiesn plant lifetimeO&M ratio of annual operating expenditures and manage-

ment cost to FCIr recycling tail gas ratio, %Sj present scale of facility jSr,j scale of facility j in the reference scalet yearWElectricity electricity energy, MW

Greek symbolsa interest rate during construction, %g system energy efficiency, %e system exergy efficiency, %

AbbreviationsBRL boiler rated loadCFB circulating fluidized bedCPD chemical percolation devolatilizationDAEM distributed activation energy modelDEPG dimethyl ether of polyethylene glycolDPP discounted payback periodFCI fixed capital investment, 106 $FG-DVC functional group-depolymerisation vaporisation cross-

linkingGE General Electric CompanyGTCC gas turbine combined cycleH/C (H2 � CO2)/(CO + CO2)HP high pressureHRSG heat recovery steam generatorIGCC integrated gasification combined cycleIP intermediate pressureIRR internal rate of return, %LHHW Langmuir–Hinshelwood–Hougen–WatsonLP low pressureMeOH methanolNMP N-Methyl-2-pyrrolidonePC propylene carbonatePSA pressure swing absorptionSPP simple payback periodTIT turbine inlet temperatureWGS water gas shiftXLT Xiaolongtan

1302 Z. Guo et al. / Applied Energy 113 (2014) 1301–1314

converting coal to liquid fuels and electricity with CO2 sequestra-tion. Normann et al. [12] represented a CO2 neutral oxy-fuel poly-generation system co-producing transportation fuel and electricityby co-firing coal and biomass.

Fluidized bed technology has been widely used in variousapplications, such as coal/biomass gasifier [18–22], coal/biomasspyrolyzer [23–25], catalytic reactor [26,27] and circulatingfluidized bed (CFB) boiler [28–32]. A novel polygeneration facilitycoupling an atmospheric pressure fluidized bed pyrolyzer with aCFB boiler for the simultaneous production of gas, tar, electricityand steam was proposed by our research group [33], as shown inFig. 1. Coal is firstly fed into a pyrolyzer (550–750 �C) andpyrolyzed into gas, tar, and char. The heat needed for pyrolysis isprovided by hot circulating solid, which is separated in the cycloneand collected in the Loop seal of a CFB boiler. Volatiles, i.e., coal gasand tar, are separated from fly ash in cyclones. The char residueand fly ash are sent into the boiler via Loop seals and burnt at about850–950 �C. Steam, which is generated in the boiler and back-passheat exchangers, can drive steam turbines (not shown here) togenerate electric power. Coal gas and tar can be cooled and sepa-rated in gas cooler and clean-up facilities. After purification, gasis split into two parts; one part is pumped back into the pyrolyzerto fluidize the bed material while the other part cannot only beused as domestic gas and gaseous fuel, but also be synthesized toliquid fuel after further clean-up. Tar, chiefly comprised of aliphatichydrocarbon, phenols, and aromatics, can be refined to chemicals

or be upgraded to liquid fuels. The most outstanding feature ofthe polygeneration technology is that it can easily fulfill the stagedconversion of coal resources and also maintain the boiler run underrated load (BRL) condition.

Over the past decade, on the basis of our fundamental research,a 1 MW bench scale facility and a 12 MW pilot scale facility havebeen successfully established and continuously operated [34–36].The next phase of our project is the building and operating of anindustrial scale facility. Ahead of the industrialization, we needto evaluate the feasibility of a large-scale polygeneration plant,and this is the task of this paper.

This work has the following objectives: (1) to propose a feasiblepolygeneration system integrating a 2 � 300 MW CFB power plantwith atmospheric pressure fluidized bed pyrolyzers for liquid fuelsand power generation; (2) to find out optimal operation conditionsof the polygeneration plant and to analyze its thermodynamic andeconomic feasibility when compared with a conventional2 � 300 MW CFB power plant. As a basis for comparison, we ensurethe capacity of CFB boilers in both plants is the same.

2. Description of plants

2.1. Conventional CFB power plant

A subcritical CFB power plant is chosen as the reference plant.This plant has a total installed power capacity of 2 � 300 MW. It

Fig. 1. Principle chart of the polygeneration facility.

Z. Guo et al. / Applied Energy 113 (2014) 1301–1314 1303

consists of 2 units, and each unit is 300 MW-scale. A simplified dia-gram of each unit can be found in Fig. 2. Coal particles are fluidizedand combusted with hot air in the boiler at about 900 �C. Lime-stone particles are used to capture SOx in the furnace. Flue gas isseparated from fly ash in cyclones and then introduced into backpass heat exchangers for air and water preheating. Feed water isheated and evaporated into subcritical steam through the back-pass and the furnace. Superheated steam at 16.7 MPa and 537 �Cis expanded in the high-pressure turbine to an intermediate pres-sure of 3.8 MPa. This IP steam is reheated in the reheaters to 537 �Cand is then expanded in the IP steam turbine. Finally, the exhaustfrom the IP steam turbine is expanded in the LP (low pressure) tur-bine to 10 kPa and enters the condenser. The condensate water issent to a series of low-pressure feed heaters. The heated water issent to a deaerator to remove dissolved gases. Deaerated water ispassed through the high-pressure water heaters and is then fedto the economizer of the boiler [37].

2.2. Polygeneration plant

To tackle the shortage of oil supply in China, many polygenerationtechnologies were designed as liquid fuels-oriented [2–4,10,11,13–15,17]. Similarly, we established a polygeneration system producingelectricity and liquid fuels simultaneously (Fig. 3). The polygenera-tion system is based on the facility proposed by Zhejiang University(cf. Fig. 1).

Lignite is ground, mixed with hot ash from boilers, and pyro-lyzed in pyrolyzer 1 and pyrolyzer 2. Pyrolyzed char as well asash particles then flows from each pyrolyzer into the corre-sponding boiler where char combustion occurs. The operatingtemperatures of each pyrolyzer and each boiler are 700 �C and900 �C, respectively. Flue gas flows out of boilers, throughback-pass heat exchangers, and then into the environment.Steam generated in the furnace and back-pass is used forelectricity generation. The installed capacity of each set of steamturbine is 300 MW.

Coal gas, separated from char particles in cyclones, contains taraerosols, moisture, particles, as well as sulfuric and nitrogenouscompounds which can block or poison the catalyst in reformingand synthesis units [38]. Therefore, raw gas has to be processed be-fore its utilization. Raw gas from each pyrolyzer is mixed togetherand the mixed stream is cooled down during flowing through aheat recovery unit. After the cooling process, the majority of tarmolecules as well as water-soluble contaminants, such as NH3

and chlorides, are condensed and then removed from the gas.Tar aerosols are captured in tar trap units (including gas cooling

unit and precipitator), then purified, and collected in tar tanks. Tarhas some chemical and physical defects such as high viscosity,acidity, and instability, so upgrading processes, such as cracking,reforming, and hydrogenation are necessary before its application[39]. Since polygeneration technologies are liquid fuels-oriented,we prefer to use tar hydrogenation technology to produce oil. Atwo-stage tar hydrogenation technology developed by Edwardset al. [40,41] is adopted. Tar is firstly heated to 150 �C and thenpumped into the reactors. Tar is first hydrotreated in the first stageunder pressurized hydrogen, using a fixed-bed of disposable cata-lyst of sulfided steelwood, and then further upgraded under pres-surized hydrogen in the second fixed-bed reactor, using nickel/molybdenum (Ni/Mo) catalyst. Reactors are operated under13.8 MPa and �420 �C. The main characteristics of synthetic crudeoil are: (1) distillate fractions: 20 vol% gasoline, 39 vol% middle dis-tillate, 40 vol% residual oil; (2) specific gravity (15 �C/15 �C): 0.93;(3) nitrogen: 0.2 wt%; sulfur: 0.01 wt%; Conradson carbon:1.8 wt%; carbon aromaticity: 0.39. Feed water is used to carry theextra heat producing during reactions. The pressure of the gaseousproduct is released to approximately atmospheric pressure in aflash tower. Waste gas is separated in the top of the flash towerand mixed with raw gas prior to gas cleanup.

Afterwards, a part of gas is sent back to pyrolyzers as thefluidized medium. It has to be emphasized here that the flow rateof fluidized medium depends on the instantaneous bed height ofpyrolyzer and is varying all the time. For simplification, the 50%

Fig. 2. Schematic diagram of a unit of the CFB power plant.

1304 Z. Guo et al. / Applied Energy 113 (2014) 1301–1314

gas is used as the fluidized medium according to the results of a40 t/h feedrate-scale pyrolyzer. The rest 50% gas is desulfurizedto prevent adverse effect on catalyst downstream. Physical sol-vents and chemical solvents have been used for H2S removal fordecades. Physical solvents tend to be superior to chemical solventsfor removing high concentration of H2S. In addition, physical sol-vents have lower energy consumptions than chemical solvents[42]. Propylene carbonate (PC), dimethyl ether of polyethylene gly-col (DEPG), N-Methyl-2-pyrrolidone (NMP), and methanol (MeOH)are the major physical solvents [43]. The Selexol process [44] usingDEPG as solvents has been widely applied for selective H2S re-moval, and is preferred here. The captured H2S is converted to ele-mental sulfur in a Claus plant, while clean-up of flash gas isprocessed in a SCOT plant [8].

Escaping from Selexol plant, the purified gas flows through apressure swing absorption (PSA) unit. In this plant, hydrogen isabundant in pyrolyzed gas (H2 source) and tar upgrading process(H2 sink) needs hydrogen. Extracting H2 from pyrolyzed gas totar upgrading process enhances the energy integration. PSA is amature technology and is widely used in hydrogen plant to sepa-rate up to 85–90% of high purity (99.999%) hydrogen from coalgas [8]. Therefore, we use the technology to separate the requiredamount of H2 for tar hydrogenation process.

Coal gas has a rather low H/C ratio value but high CH4 content,therefore, we prefer a methane reformer to a water gas shift reac-tors ahead of a synthesis tower. There are three main methanereforming technologies, namely, partial oxidation [45], steamreforming [46], dry reforming [27]. Considering the high CO2 con-tent in coal gas, dry reforming is preferred. Dry reformer utilizesCO2 in the gas, and obtains rather high H/C ratio and low CO2 emis-sion. The methane reforming reactor uses a bubbling bed of highlyactive Ni/La/a-Al2O3 at about 700–800 �C [27]. The main equilib-rium reactions [27,47] are:

CH4 þ CO2 ¼ 2COþ 2H2; DRH0298K ¼ þ247:9 kJ=mol ð1Þ

COþH2O ¼ CO2 þH2; DRH0298K ¼ �41:16 kJ=mol ð2Þ

Methane reforming is an endothermic process. The required heat isprovided by the combustion of partial synthesis tail gas (see Fig. 3).

The reformed gas can be used for synthesizing liquid fuels, e.g.,gasoline, diesel, methanol, etc. Methanol, whose octane number(108.7) is even higher than gasoline, is a promising alternative fuel.Commercially, there are various available technologies for metha-nol synthesis [48]. The Lurgi gas phase technology is adopted here.The reformed gas is firstly cooled down, then compressed to thereactor pressure, mixed with recycling tail gas, and finally fed intomethanol synthesis reactor where methanol synthesis takes placeover Cu/ZnO/Al2O3 catalyst under 6.97 MPa and 180–260 �C. Thetemperature of the synthesis process is kept nearly constant bytransferring heat to the coolant (i.e., water). The main reactions in-volved in the reactor are as follows [48,49]:

CO2 þ 3H2 ¼ CH3OHþH2O; DRH0298K ¼ �49:5 kJ=mol ð3Þ

COþ 2H2 ¼ CH3OH; DRH0298K ¼ �90:7 kJ=mol ð4Þ

CO2 þH2 ¼ COþH2O; DRH0298K ¼ 41:16 kJ=mol ð5Þ

The outlet gases from the synthesis reactor is cooled down to40 �C by heating inlet gases and coolant, and then expanded to1.5 MPa through a flash drum, where flash gas flows out of thetop and crude methanol exits from the bottom and flows into a dis-tillation unit for further purification. The distillation unit includesthree towers, namely, flash tower, light tower, and heavy tower.The flashed tail gas can be divided into the recycling tail gas, whichis compressed and recycled back to the synthesis reactor, and theun-recycling tail gas, which is sent to downstream facilities.

The un-recycling tail gas is split into two parts. One part is burntwith the air in the burner of the methane reformer. The combus-tion heat of this part of gas is transferred into the methane

Fig. 3. Flow diagram of the polygeneration plant.

Table 1Main properties of XLT lignite.

Ultimate analysis (wt%, ar) Proximate analysis (wt%, ar)

Carbon 34.78 Moisture 34.79Hydrogen 2.30 Ash 15.04Nitrogen 0.88 Volatile matter 27.22Sulfur 1.09 Fixed carbon 22.95Oxygen 11.11 LHV (MJ/kg, ar) 12.15

Z. Guo et al. / Applied Energy 113 (2014) 1301–1314 1305

reformer. The other part of gas is firstly saturated with steam tocontrol NOx formation, and then burnt with compressed air inthe combustion chamber of gas turbine [50]. GE 6FA gas turbine(turbine inlet temperature of 1288 �C, expansion ratio of 15.5) isemployed in the paper, which has wide application, includingintegrated gasification combined cycle (IGCC), cogeneration, sim-ple cycle, and combined cycle [51]. Exhaust gas exiting from gasturbines flows into a heat recovery steam generator (HRSG). Partof feed water is heated and evaporated into high pressure (HP,12.5 MPa), intermediate pressure (IP, 2 MPa), and low pressure(LP, 0.24 MPa) steam in the HRSG. Part of feed water is pumpedinto tar hydrogenation reactor and methanol synthesis tower.These reactors transfer released heat to the feed water, andproduce IP steam (cf. steam B and steam from tar hydrogenationsection in Fig. 3). The other part of feed water is pumped to12.5 MPa, and evaporated into HP steam (cf. steam A in Fig. 3) byabsorbing the sensible heat of raw gas in the heat recovery unit.

HP steam from the HRSG and the heat recovery unit are expandedin HP steam turbine. The expanded steam from HP turbine is splitinto two parts, one is reheated in the HRSG and the other is re-heated in the heat recovery unit. After reheated, both parts ofsteam mix with other IP steam flows and drive IP and LP turbinesto generate electric energy.

3. Process simulation

This section gives the details of the modeling of main facilitiesin each plant mentioned in Section 2. Aspen Plus is selected tomodel both plants. This simulation software is a steady statechemical process simulator and has been widely used in modelingpolygeneration systems [2–17]. Xiaolongtan (XLT) lignite is se-lected as the raw material. The typical characteristics of the coalare reported in Table 1. The details of unit modeling parametersare presented in Table 2.

3.1. CFB boiler

Comprehensive models of coal combustion in a CFB boiler havebeen built in Aspen Plus [53,54]. These models integrating bothhydrodynamic and combustion models have been proved to pre-dict combustion process well. However, these models were rathercomplex so that they were time-consuming and heavy computa-tion. To improve computing speed, some researchers neglected

Table 2Modeling details in Aspen Plus.

CFB power plantGrinding Crusher – modeling the breaking of coal and limestone particles; coal particle size <10 mm, limestone particle size <1 mmCFB boiler Ryield – Yield reactor, modeling coal decomposition by specifying reaction yields, components include H2, O2, C, S, N2, Ash,

H2O. Temperature = 900 �CRgibbs – Chemical equilibria, modeling components combustion, possible products include H2, O2, C, SO2, SO3, N2, NO, NO2,N2O, Ash, H2O, CO, CO2, assuming ash is inert, carbon conversion = 99.8%, heat loss = 0.4%. Temperature = 900 �CRstoic – Stoichiometric reactor, modeling in-furnace desulfurization process, Reactions: CaCO3(s) = CaO + CO2,CaO(s)+0.5O2 + SO2 = CaSO4(s), assuming desulfurization efficiency = 95%. Temperature = 900 �CCyclone – Cyclone separator, modeling flue gas and fly ash separation, assuming separation efficiency = 99.5%, cyclonenumber = 4, pressure loss = 1%

Back-pass Heater – modeling superheater, reheater, economizer, and air-preheater

Steam cycles Including HP, IP and LP turbines, 3 HP heaters, 1 deaerator, 4 LP heaters, 2 pumps, and a condenserCompr – modeling HP, IP, and LP turbines; Isentropic efficiency = 0.90, Mechanical efficiency = 0.98. Inlet pressure:HP turbine = 16.7 MPa, IP turbine = 3.42 MPa, LP turbine = 0.75 MPa; exhaust pressure = 10 kPaHeater – modeling HP heaters, LP heaters, condenser, pressure loss = 1%, feed water temperature = 279 �CFlash2 – modeling deaerator

Pumps Pump – modeling pumps, global efficiency = 0.8, discharge pressure: condenser pump = 1.72 MPa, feed waterpump = 20.5 MPa

Fans Pump – modeling primary air fan, secondary air fan, induced draft fan; global efficiency = 0.8, pressure increase: primary airfan = 30.2 kPa, secondary air fan = 1.46 kPa, induced draft fan = 7.8 kPa, primary air/secondary air = 6:4

Polygeneration plantGrinding, CFB boiler, Back-pass, Steam

cycle, Fans, PumpsThe same as CFB power plant

Pyrolyzer Ryield – Yield reactor, modeling lignite pyrolysis by known products yield. Products includes: H2, CO, CO2, CH4, H2O, Char,H2S, NH3, C2H4, and TarTar is represented by five model compounds: phenol, hexadecane, quinoline, naphthalene, and dibenzothiophene.Temperature = 700 �C

Heat recovery unit Heater – modeling gas cooling process, pressure loss = 1%Tar trap unit Sep – separating inlet streams, assuming tar separation efficiency = 100%Tar hydrogenation Ryield – Yield reactor, possible products: oil, CH4, C2H6, C3H8, C4H10, Coke, H2O, CO, CO2, NH3, H2S. Temperature = 420 �C,

pressure = 13.8 MPaSelexol plant Sep – separating gas into acid gas, flash gas, and clean gas; pressure: acid gas 0.1 MPa, flash gas 0.69 MPa, clean gas

3.11 MPa; assumption: H2S absorption efficiency = 99%Claus/SCOT plant Rstoic – Stoichiometric reactor, main reactions: H2S + 1.5O2 = SO2 + H2O, 2H2S + SO2 = 2H2O + 3S(s),

SO2 + 3H2 = H2S + 2H2O, 2H2S + O2 = 2S(s)+2H2OPSA Sep – assuming H2 separation efficiency = 85–90%, purity = 99.999%Reformer Rplug – Plug flow reactor, LHHW kinetic expression, main reactions: CH4 + CO2 = 2CO + 2H2, CO + H2O = CO2 + H2

Length: 7.8 m, Diameter: 4 m, Adjusting tube numbers to achieve 0.1 m/s gas velocity, catalyst loading 1512 tonneaccording to contact time of 31.8 g s ml�1 [27]

Burner Rgibbs – chemical equilibria, modeling gas combustionSynthesis tower Rplug – Plug flow reactor, LHHW kinetic expression, main reactions: CO2 + 3H2 = CH3OH + H2O, CO2 + H2 = CO + H2O

Diameter: 0.04 m, Length: 7 m, tube numbers: 3240 [52]Methanol distillation section Flash2 – phase equilibrium, modeling flash drum, flash pressure = 0.15 MPa

Radfrac – multistage vapor–liquid fractionation, modeling methanol distillation process. Three towers adopted: flashtower, light tower, and heavy towerMethanol purity = 99.85%

Gas turbine Compr – modeling air compressors and gas turbinesIsentropic efficiency: compressor = 0.80, turbines = 0.922Mechanical efficiency: compressor = 0.98, turbines = 0.98Rgibbs – modeling gas combustion in combustion chamber, adjusting air flow rate to achieve TIT = 1288 �C

HRSG and steam cycle Heater – modeling HP heat exchanger, IP heat exchanger, LP heat exchanger, economizer, and reheater, assuming pinchtemperature = 10 �C, approach temperature = 10 �C, flue gas temperature = 130 �C. Inlet pressure: HP turbine = 12.5 MPa, IPturbine = 2 MPa, LP turbine = 0.24 MPa, exhaust pressure = 4.6 kPa, steam turbine: isentropic efficiency = 0.90, mechanicalefficiency = 0.98

1306 Z. Guo et al. / Applied Energy 113 (2014) 1301–1314

the hydrodynamic complexity of CFB boiler, and established amuch simpler model [55]. Their results indicated that the simplemodel can also accurately predict coal combustion in a boiler.Therefore, we adopt this model to simulate a CFB boiler. Coal, char,and ash are considered as nonconventional components, and char-acterized by their ultimate and proximate analyses. Ryield, Rgibbs,and Rstoic models are applied to simulate coal decomposition, charand volatile combustion and in-furnace desulfurization processesin the CFB boiler, respectively. Cyclone modules are used to modelgas solid separation in cyclones with designed separation effi-ciency. The details of CFB boiler models can be seen in Table 2.

3.2. Pyrolyzer

Coal pyrolysis is a complex process and heavily depends on coaltype, operating conditions and pyrolysis devices. Many researchers

proposed different models to calculate pyrolysis process, e.g. CPD[56], DAEM [57], FG-DVC [58], FlashChain [59], etc. However, mostmodels were linked to coal microscope structure parameters andtoo complex to utilize for a wide range of coal. Maffei et al. [60–62] proposed multi-step models to simulate the pyrolysis processeasily and accurately only by elemental analysis. Here, the multi-step models are used to predict the pyrolysis of XLT lignite viaChemkin software. Since the composition of tar is very complex,tar is generally assumed to be a simple hydrocarbon CmHn, suchas C14H10 [63]. However, this method is inaccurate to evaluatethe property of coal tar only by one compound. To better representthe characteristics of tar, we choose five different model com-pounds, i.e., phenol, hexadecane, quinoline, naphthalene, anddibenzothiophene. A pyrolyzer is simulated using Ryield reactormodel in Aspen Plus. The simulated pyrolysis results of Chemkinsoftware is input into the model as product yield. Since the

Z. Guo et al. / Applied Energy 113 (2014) 1301–1314 1307

combination of a boiler and a pyrolyzer is a dual-furnace design, itis difficult to achieve convergence for circulating solid particles be-tween two furnaces. To solve this problem, we disconnect the ashstream from the boiler to the pyrolyzer, and introduce a heat flowfrom models of CFB boiler to the pyrolyzer instead.

3.3. Gas conditioning

Gas cooling unit is simply modeled using Heater block in AspenPlus, and tar capture process is simulated by Sep model, assuming100% tar is captured. Coal gas from the pyrolyzer is under atmo-spheric pressure, coal gas has to be pressurized before entering Sel-exol plant. Since Selexol plant is a complex process, it is alwayssimply modeled with a Sep block in Aspen Plus [50,64]. Sep blocksplits the coal gas into three streams, namely, clean gas, acid gas,and flash gas (cf. Table 2). Clean gas, is H2S-free and has almostthe same pressure as the pressurized coal gas. Acid gas, which isrich in H2S, is vented into Claus plant model. Flash gas has a med-ium pressure, is sent into SCOT plant. Claus, and SCOT plants adoptmodels in the literature [50]. Similar to Selexol process, PSA unit isalso simply simulated using Sep block, assuming that H2 of99.95 mol% purity is separated [65].

Methane reforming process is controlled by kinetics. Plug reac-tor (Rplug) module is used to simulate the reformer. Kinetic mod-els of methane reforming adopts Langmuir–Hinshelwood reactionrates expressions [27,47]. Rates coefficients are taken from Wurzelet al. [47]. The burner which combusts gases to provide the refor-mer external heat resource is simulated using a chemical equilibriareactor model (Rgibbs).

3.4. Tar hydrogenation

Since the detailed kinetic mechanism of tar hydrogenation pro-cesses is unclear so far, it is difficult to exactly predict the hydroge-nation process by kinetics. Yield reactor module (Ryield) is used tosimulate hydrogenation process by inputting Edwards et al.’sexperimental results [40] into the reactor as the yields distribution.Pump and Mcompr modules simulate tar and hydrogen compres-sion processes, respectively.

3.5. Methanol synthesis

Rplug model is used to simulate Lurgi synthesis technology, cal-culating with Langmuir–Hinshelwood–Hougen–Watson (LHHW)kinetics [52]. Flash towers are simply simulated with vapor–liquidseparator module (Flash2), and distillation towers are modeledusing RadFrac column [3]. Column conditions are adjusted to re-cover methanol at the required purity (99.85%).

3.6. Gas turbine combined cycle

Gas combustion with air in the combustion chamber is simu-lated with chemical equilibria reactor module (Rgibbs). The airsupply is adjusted to make the turbine inlet temperature (TIT)reach the set value 1288 �C. Compr blocks are applied to simulatethe air compressor and gas turbine. Blades in each stage of gas tur-bine are cooled by splitting air from different stages of air compres-sor. HRSG is modeled by a series of Heater units. Generally, thepinch temperature rage is from 8 to 20 �C, and the approach tem-perature range from 5 to 20 �C. So we set the pinch temperatureand the approach temperature as 10 �C. Steam turbine is modeledusing compressor and expander module (Compr). Details can befound in Table 2.

3.7. Additional information

Peng Robinson cubic equation of state with the Boston-Mathiasalpha function (PR-BM) physical property method is used in coaland gas processing units, including combustion, pyrolysis, PSA,reforming, gas turbine, and Selexol/Claus plant. The Soave–Redlich–Kwong cubic equation of state (SRK) physical propertymethod is used in high pressure reactors, such as methanol synthe-sis tower and tar hydrogenation reactor. STEAM-TA, which usesASME 1967 steam tables, is used to calculate all thermodynamicproperties for pure water and steam, and is applied to pumps,steam turbines, and steam side of heat exchangers.

4. Methodology

The present section describes a thermodynamic and economic-based analysis approach that we use to evaluate the feasibility ofpolygeneration plant. The schematic of the analysis approach isshown in Fig. 4. It includes four main layers: (1) Optimize operat-ing conditions of some facilities, find the optimal conditions; (2)calculate the thermodynamic performance of optimum design,and compare it with that of CFB power plant; (3) calculate the eco-nomic performance of optimum design, and compare it with thatof CFB power plant; (4) analyze the price factors that influencethe economic performance.

4.1. Optimization method

Optimization process adopts simple variable method (cf. Fig. 4).Briefly, vary only one parameter and fix other parameter valueseach time till to find an optimal value to replace its initial value.Then, vary the second parameter and find its optimal value to takethe place of the initial one. Take the same method for the restparameters till all parameters have been optimized.

4.2. Thermodynamic evaluation criteria

Energy efficiency and exergy efficiency are generally used toevaluate thermodynamic performance of a polygeneration plant[2,3,10,11,66,67]. Energy efficiency, based on the first law of ther-modynamics, is the conversion efficiency of coal energy to the de-sired energy in a system. Exergy is the maximum work that can beobtained from the system, when its state is brought to the refer-ence or ‘‘dead state’’ (standard atmospheric conditions). Exergyefficiency calculates conversion efficiency of the exergy in a systemaccording the second law of thermodynamics. In the polygenera-tion system, the output is oil, methanol, and electricity, and the in-put is lignite coal. Therefore, the calculations of overall energy andexergy efficiencies of the system are defined as

g ¼ FMeOHCVMeOH þWElectricity þ FOilCVOil

FCoalCVCoalð6Þ

and

e ¼ EXMeOH þ EXElectricity þ EXOil

EXCoal; ð7Þ

where g and e denote the overall energy efficiency and exergy effi-ciency of the polygeneration system, respectively. FMeOH, FOil, andFCoal denote the flow rate of methanol, oil, and coal, respectively.CVMeOH, CVOil, and CVCoal represent the caloric heat of methanol,oil, and coal, respectively. WElectricity, EXMeOH, EXElectricity, EXOil, andEXCoal represent the electricity energy and exergy of methanol, elec-tricity, oil, and coal, respectively. The calculation of fuels’ exergy canbe found elsewhere [68].

Fig. 4. Feasibility analysis methodology.

1308 Z. Guo et al. / Applied Energy 113 (2014) 1301–1314

4.3. Economic evaluation criteria

4.3.1. Fixed capital investment (FCI) calculation methodThe fixed capital investment (FCI) can be estimated by scaling

up method [44], which can be expressed by

FCI ¼Xm

j¼0

Cj ¼Xm

j¼0

Aj � Fj � Ir;j �Sj

Sr;j

� �bj" #

ð8Þ

where Cj, Aj, Fj, Sj, and bj denote capital investment, domestic factor,installation factor, scale, and scale factor of facility j in the presentscale, respectively. m is the total number of facilities. Ir,j and Sr,j rep-resent purchase cost and scale of facility j in the reference scale,respectively. Aj is equal to the ratio of the price of domestic pro-duced equipment to that of fully imported equipment. Since ourstudy is based on energy market of China, we consider domestic fac-tor of China. Facilities in China are much cheaper than those inAmerica or Euro. Here, we take domestic factor as 0.65. The capitalcost data of facilities with basic scale can be found elsewhere, whichare illustrated in Table 3. Prices are based on different currencies indifferent year, so all capital costs are updated to costs of the year2011 in million dollars (106 $ for short) to calculate on the same

Table 3Basic capital cost data of facilities.

Component Basic cost (106 $) Basic sc

CFB power planta 268.733 300Fluidized bed pyrolyzer and gas cleanup unitb 27.864 1196.91Heat exchangerc 13.138 138H2 PSAc 17.519 16616Selexold 46.195 81Claus/SCOTd 31.485 81Gas turbined 114.114 276Steam turbine and cyclec 55.474 275HRSGc 49.635 355Synthesis reactorc 3.927 25.25Methanol upgradingc 1.907 25.25Methane reformere 0.249 0.021Coal gas compressorc 5.839 10Pumpe 0.020 250Tar hydrogenationb 2.322 200000

a Static investments of 300 MW CFB power plant built in China are taken [70].b Costs are taken from feasibility study reports of chemical engineering projects in Ch

cooling and purifying unit is according to the feasibility study report of the 40t/h feedrac The costs data of facilities unless otherwise specified are from Meerman et al. [72], th

factor item means the capital cost includes installation cost.d Costs are based on Kreutz [9], assuming the capital cost per MW of GE 6FA gas turbe Cost data of methane reformers and pumps are according to Xue [73] and Huijgen e

basis according to cost indexes method, which is based on Marshall& Swift equipment cost index [67]. The yearly Chinese Yuan/US dol-lar and Euro/US dollar exchange rates of 2011 are 0.1548 and1.3936, respectively [69].

4.3.2. Internal rate of return (IRR)IRR is commonly used as an economic criterion to evaluate the

feasibility of a project [3]. The calculation of IRR is according to thefollowing equation [3]:

Xn

t¼0

Ctð1þ IRRÞ�t ¼ 0 ð9Þ

where Ct denotes the annual cash flow of the year t. n representsplant lifetime. The annual cash flow can be calculated as follows [4]:

Ct ¼ Cp � ðFCI� ðCRF� ð1þaÞ þ O&MÞ þ CF þ CMÞ ð10Þ

where Cp, CF, and CM denote annual product sales income, fuel cost,and material cost, respectively. FCI represents fixed capitalinvestment, while O&M denotes the ratio of annual operating andmanagement cost to FCI. a is the interest rate during construction.

ale Scaling factor Scale unit Overall installation factor

0.74 MW –0.7 GJ/h input –0.6 MWth 1.9950.65 H2 kmol/h 2.560.67 t-sulfur input/day –0.67 t-sulfur input/day –0.75 net MW –0.67 gross MW 1.47321 MWth exchanged 1.47320.6 t-methanol/h –0.65 t-methanol/h –0.75 m3/s CO2 –1 MW 1.53120.14 m3/h 1.9950.7 t-crude oil/y 1.995

ina [71]. Specifically, cost of pyrolyzer and gas cleanup unit including cyclones, gaste polygeneration plant.e default installation factor is considered as 1.995. The symbol ‘–’ in the installation

ine close to that of GE 7H gas turbine.t al. [74], respectively.

Z. Guo et al. / Applied Energy 113 (2014) 1301–1314 1309

CRF means the ratio of annual average investment, and can be cal-culated with discount rate (i) and plant lifetime [3,66]:

CRF ¼ i1� ð1þ iÞ�n ð11Þ

4.3.3. Payback periodAdditionally, payback period is taken to evaluate the time re-

quired for the return of total investment, including two methods,i.e., simple payback period (SPP) and discounted payback period(DPP). DPP considers the effect of discount rate during plant life-time. Specifically, DPP can be calculated as follows:

DPP ¼ AþPA

t¼0Btð1þ iÞ�t

Cð12Þ

where A denotes last period with a negative cumulative cash flow,and Bt represents the net cash flow in year t. C means the annualnet cash flow during the next period after A.

Fig. 5. Comparison of modeling and experimental pyrolysis results (left column:experiment, right column: simulation).

Fig. 6. Effects of methane r

5. Results and discussion

5.1. Evaluation of the pyrolyzer model

The results calculated by multi-step pyrolysis models are com-pared with our experimental results. The experiment was carriedout on a 40 t/h XLT lignite-fueled pyrolyzer. The pyrolyzer wasoperated at 700 �C and fluidized by circulating gas. All parametersin the experiment are the same with those used in our modelingexcept that the scale of our model is larger than the experiment’sscale. The comparison between modeling data and experimentalresults is shown in Fig. 5 (left column: experimental, right column:modeling). Obviously, modeling and experimental results showfairly good agreement. Furthermore, according to the heat balance,we can evaluate the required circulating ash amount, which isapproximately 2389.5 t/h. According to our knowledge of300 MW scale CFB boiler, this value can be achieved by conven-tional operation of CFB boiler. Therefore, these models can be usedin our study.

5.2. Operation parameters optimization

Prior to the optimization process, initial values of operationparameters are given: (1) methane reforming temperature:900 �C; (2) methanol synthesis temperature: 230 �C; (3) recyclingratio of synthesis tail gas (r): 0.

5.2.1. The temperature of reforming unitFig. 6 demonstrates the influence of reforming temperature.

Clearly, g rises with reforming temperature until 800 �C and there-after it seems to stabilize. This is related to the approaching of theequilibrium state of the methane reforming reactions when tem-perature reaches 800 �C [2]. With the increase of reforming tem-perature, methanol yield continuously increases and reachesapproximately 13.5 kg/s at 800 �C. Further increasing of reformingtemperature affects little methanol production, due to reaching theequilibrium of reforming reactions. Net power generation showsconcave trends and bottoms out at 800–900 �C. This is because alarge quantity of gases is converted to methanol in 800 �C. Overall,800 �C is the optimal value of reforming temperature.

eforming temperature.

Fig. 7. Effects of methanol synthesis temperature.

Fig. 8. Effects of recycling tail gas ratio.

1310 Z. Guo et al. / Applied Energy 113 (2014) 1301–1314

5.2.2. The temperature of synthesis unitAs seen in Fig. 7, synthesis temperature has great influence on

methanol yield, which peaks at about 14 kg/s under the tempera-ture of 210 �C. Increasing temperature improves the reaction rate.However, since the reactions are exothermic, lower temperature fa-vors the equilibrium shifting rightward and thus an upward trendappears at a low temperature. When temperature rises, a decreas-ing tendency arises. The up-down trends are similar with Chenet al. [52] On the other hand, power generation decreases with syn-thesis temperature until 210 �C and thereafter it increases steadily.We can clearly see that system efficiency reaches their peaks at210 �C. Hence, 210 �C is the optimal synthesis temperature.

5.2.3. Recycling tail gas ratioFig. 8 displays the influence of recycling ratio of the tail gas on

system performance. Total methanol yield increases linearly fromca. 14.4 kg/s at the recycling ratio of 0% to ca. 17.5 kg/s at 80%.More tail gas recycling into the reactor will facilitate methanol pro-duction. With the increase of recycling ratio from 0% to 60%, a ten-dency of increase for exergy efficiency can be observed. However,exergy efficiency is gradually lower with higher recycling ratioafterwards. This is referred to the result from a sharp decrease of

electricity generation. Consequently, 60% is the desirable factionof recycling tail gas.

Based on the optimization, optimal system parameters arefound: (1) reforming temperature: 800 �C, (2) synthesis tempera-ture: 210 �C, (3) recycling tail gas ratio: 60%. Simulation data ofmain streams in the proposed polygeneration plant are listed inTable 4.

The thermodynamic characteristics of the optimum design ofpolygeneration plant are compared with those of conventional2 � 300 MW CFB power plant, and summarized in Table 5. Itshould be emphasized here that CFB boilers in both plants arein the full load condition to ensure 600 MW power generatingcapacity in boilers side. Only about 40% of lignite is converted intochar (to be fire in boilers), and the results cause a much larger lig-nite feed rate (239.50 kg/s) in the polygeneration plant comparingwith conventional power plant (143.16 kg/s). The increase of lig-nite in the polygeneration plant is efficiently converted into extrapower (172.16 MW), methanol (16.86 kg/s), and oil (2.95 kg/s).With the help of cascade energy utilization and heat integration,polygeneration plant shows higher energy and exergy efficiency(43.2% and 43.24%) than conventional CFB power plant (34.9%and 34.2%).

Table 4Simulation data of main streams.

Stream in Fig. 1 2 3 4 5 6 7 8

Temperature,�C

130 700 37.8 37.8 30.9 800 210 580.6

Pressure, bar 1.10 1.00 31.10 31.0 1.01 1.01 66.70 1.05Mole flow,

kmol/s32.19 10.02 2.87 2.652 2.178 2.58 2.27 6.296

Mole fractionH2 0.0000 0.2437 0.5014 0.5423 0.4378 0.4745 0.1126 0.0000O2 0.0305 0.0000 0.0001 0.0001 0.0002 0.0000 0.0000 0.1340N2 0.7796 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.6968H2O 0.0277 0.4788 0.0327 0.0212 0.0259 0.0766 0.0002 0.1031CO2 0.1619 0.1141 0.2024 0.1758 0.2147 0.0455 0.1114 0.0651CO 0.0000 0.0814 0.1444 0.1550 0.1893 0.3750 0.4468 0.0000SO2 0.0002 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000SO3 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000NO2 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000NO 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000N2O 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000C2H4 0.0000 0.0130 0.0231 0.0250 0.0305 0.0257 0.0668 0.0000CH4 0.0000 0.0426 0.0764 0.0792 0.0968 0.0011 0.0030 0.0000NH3 0.0000 0.0123 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000H2S 0.0000 0.0100 0.0180 0.0000 0.0000 0.0000 0.0000 0.0000C2H6 0.0000 0.0002 0.0006 0.0007 0.0008 0.0007 0.0018 0.0000C3H8 0.0000 0.0001 0.0004 0.0004 0.0005 0.0004 0.0010 0.0000C4H10 0.0000 0.0001 0.0002 0.0002 0.0003 0.0003 0.0005 0.0000Tar 0.0000 0.0036 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000MeOH 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.2557 0.0000

Table 5Thermodynamic performances of CFB power plant and polygeneration plant.

Items CFB powerplant

Polygenerationplant

Energy inputCoal input (kg/s) 143.16 239.50Coal energy input (MW) 1739.39 2909.93Coal exergy input (MW) 1776.66 2972.12

Energy outputMethanol yields (kg/s) 0.00 16.86Oil yields (kg/s) 0.00 2.95Methanol energy output (MW) 0.00 349.50Methanol exergy output (MW) 0.00 377.56Oil energy output (MW) 0.00 127.76Oil exergy output (MW) 0.00 127.70Power generation in boiler side (MW) 649.73 649.73Power generation in gas turbine

(MW)0.00 66.16

Power generation in steam turbine(MW)

0.00 180.93

Auxiliary power consumptionGrinding (MW) 0.04 0.07Tar pump (kW) 0.00 64.54GTCC steam cycle pumps (MW) 0.00 1.31Boiler side steam cycle pumps (MW) 15.59 15.58Primary air fan (MW) 14.13 17.54Secondary air fan (MW) 0.46 0.56Induced draft fan (MW) 11.78 10.12Circulating gas fan (MW) 0.00 1.78Gas compressor before Selexol (MW) 0.00 31.72Gas compressor after reformer (MW) 0.00 17.8Hydrogen compressor (MW) 0.00 2.26Recycling tail gas compressor (MW) 0.00 17.54Claus compressors (MW) 0.00 1.01Total consumption (MW) 41.93 117.35

SummaryNet power output (MW) 607.80 779.96Total energy output (MW) 607.80 1257.22Total exergy output (MW) 607.80 1294.36System energy efficiency (%) 34.9 43.20System exergy efficiency (%) 34.2 43.24

Z. Guo et al. / Applied Energy 113 (2014) 1301–1314 1311

5.3. Economic analysis

5.3.1. Basic evaluationThe assumptions of basic economic data are listed in Table 6.

Crude oil and methanol prices are dependent on the political factorand regional situation and fluctuate over years. China’s average im-port price of crude oil in 2011, and the average prices of methanol,limestone, sulfur, electricity, and water are preferred in the basiccase.

Table 7Comparisons of economic results.

Items CFBpowerplant

Polygenerationplant

Fuel and material consumption (million tonne/year)Coal 3.71 6.21Water 10.97 15.68Limestone 0.22 0.16

Product (tonne/year)Sulfur 0.00 41,741.36Methanol (99.85% purity) 0.00 437,025.16Crude oil 0.00 76,412.16Electricity (kW h/year) 4.38 � 109 5.62 � 109

Capital cost (106 $)Coal power plant(including boilers, steam

turbines, fans, grinding and other facilities)542.63 541.42

Fluidized bed pyrolyzer and gas cleanup – 63.89Heat exchanger – 27.78H2 PSA – 12.63Selexol – 43.49Claus/SCOT – 29.64Gas turbine – 28.06Steam turbine and cycle – 40.16HRSG – 15.49Methanol reactor – 7.18Product upgrading – 3.66Methane reformer – 19.28Coal gas compressors – 21.85Pumps – 0.07Tar hydrogenation – 2.36Fixed capital cost 542.63 856.97Interest during installation 53.18 83.98Total capital cost 595.81 940.95O&M (106 $/year) 21.71 34.28

SummaryMaterial and fuel cost (106 $/year) 147.48 245.44Annual output value (106 $/year) 243.88 565.74Annual profit (106 $/year) 74.69 286.02IRR (%) 10.63 24.07SPP (years) 9.98 5.29DPP (years) 14.49 6.36

Table 6Basic assumptions and prices for economic calculation.

Items Value Price Value

Plant economic lifetimea 30 years XLTlignite

38.7 $/t

Facility available days per year 300 Water 0.310 $/tO&Ma 4.0% of the

capital costElectricity 0.056 $/

kW hDiscount ratea 8% Crude oil 774 $/tOverall interest rate during

constructionb9.8% Methanolc 417.96 $/t

Construction period 3 years Limestone 2.121 $/tSulfur 263.16 $/t

a Plant economic lifetime, O&M, and discount rate are according to literature[66].

b Taken from literature data [4].c Methanol price is according to the average value reported by Yang[75].

1312 Z. Guo et al. / Applied Energy 113 (2014) 1301–1314

According to the economic evaluation method (cf. Section 4),the basic economic features of both plants are evaluated andshown in Table 7. Higher total capital cost for polygeneration plant(940.95 � 106 $) compared with CFB power plant should be attrib-uted to the extra facilities in polygeneration plant. It is obvious thatpyrolyzers, gas clean-up and tar utilization facilities account foronly 36% of the total fixed capital cost of the polygeneration plant.Boilers and the conventional power generation unit entail majorinvestment for building a polygeneration plant. In other words, itwould not cost too much if we establish a polygeneration plantbased on an existing CFB power plant. On the other hand, the an-nual profit of polygeneration plant quadruples that of CFB powerplant. Furthermore, IRR values for both plants are higher than dis-count rate, suggesting their feasibility. The IRR value of polygener-ation plant is nearly 14% points higher than that of CFB powerplant. Meanwhile, polygeneration plant cuts down approximately8 years of the DPP value compared with the CFB power plant.Comparing the DPP with the SPP, it can be noticed that discountrate has trifling impact on the polygeneration plant while discountrate substantially prolongs the period of investment return of CFBpower plant. Overall, at the present market condition, building apolygeneration plant is much attractive to the government andinvestors.

5.3.2. Price factor analysisThe economic situation of a plant is always changing since

prices of fuel, material, and products fluctuate in the market econ-omy. The influences of coal price, methanol price, oil price, and

Fig. 9. Effects of prices o

electricity price are shown in Fig. 9. The abscissas refer to theprices, and the ordinates are the difference between the IRR ofpolygeneration plant and CFB power plant (DIRR). The DIRR in-creases with the rise in prices of coal, methanol, and oil, but showsa decrease with the decrease of electricity price. When coal pricebecomes larger than 50$/t, the CFB power plant will encounterpoor economy. Therefore, the upper bound of coal price is no morethan 50$/t. The results show that the IRR difference is getting lar-ger with the increase of coal price, which means that higher coalprice makes polygeneration plant more feasible than CFB powerplant. Also, the DIRR is proportional to the prices of methanoland oil. The phenomena are because methanol and oil are onlyproduced in the polygeneration plant. The prices’ rise increasesthe annual profits of the polygeneration plant, but CFB power plantdoes not benefit from the rise. It is interesting to note that increas-ing electricity price from 0.04$/kW h to 0.16$/kW h makes theDIRR drop from 17% to 2%. The DIRR tends to be narrowed byincreasing electricity price from 0.04$/kW h to 0.16$/kW h. Weeasily expect that DIRR will approach zero and even enter theminus zone when electricity price keeps growing. This is relatedto the fact that the profitability of a power plant relies heavilyon the price of its only product, electricity. When electricity be-comes expensive, the power plant will benefit more than the poly-generation plant. Furthermore, we can easily find the order ofprice factors’ effects: electricity > methanol > oil > coal. The resultsalso reveal that polygeneration plant shows a higher IRR valuethan CFB power plant even when they are subject to substantialprice fluctuations.

n economic feature.

Z. Guo et al. / Applied Energy 113 (2014) 1301–1314 1313

4. Conclusions

This paper presents a new coal-based polygeneration systemco-producing methanol, oil, and electricity by integrating a2 � 300 MW CFB power plant with atmospheric pressure fluidizedbed pyrolyzers. Detailed steady state models for polygenerationplant and CFB power plant are separately established. The opti-mum design of the polygeneration plant is found by optimizingoperating parameters, and its thermodynamic and economic per-formances are compared with those of the CFB power plant. It isfound that polygeneration plant with higher energy and exergyefficiency is more efficient than CFB power plant. Polygenerationplant, which has larger IRR (24.07%) and shorter DPP (6.36 years)and SPP (5.29 years), shows more profitable than CFB power plantat the present prices. Price fluctuations have great influence on theeconomic condition of the polygeneration plant. The results ofprice factor analysis show that the rises of coal, oil, and methanolprices sharpen the competitive edge of the polygeneration plant,while electricity price rise weakens its competitiveness. The resultsalso indicate that the polygeneration plant has a good economicfeature within a wide range of price fluctuations. Overall, thepolygeneration system proposed in this paper is an efficient andeconomic technology.

Acknowledgements

We would like to gratefully acknowledge the support from theNational High Technology Research & Development Program ofChina (No. 20136AA051203) and the collaboration project ofCERC-ACTC (2010DFA72730-201) and Program for New CenturyExcellent Talents in University (NCET-09-0696).

References

[1] Department of Energy Statistics, National Bureau of Statistics, People’sRepublic of China. China energy statistical yearbook 2011. Beijing: ChinaStatistics Press; 2011.

[2] Yi Q, Feng J, Li WY. Optimization and efficiency analysis of polygenerationsystem with coke-oven gas and coal gasified gas by Aspen Plus. Fuel2012;96:131–40.

[3] Yi Q, Lu B, Feng J, Wu Y, Li W. Evaluation of newly designed polygenerationsystem with CO2 recycle. Energy Fuels 2012;26:1459–69.

[4] Lin H, Jin H, Gao L, Han W. Techno-economic evaluation of coal-basedpolygeneration systems of synthetic fuel and power with CO2 recovery. EnergyConvers Manage 2011;52:274–83.

[5] Liu P, Georgiadis MC, Pistikopoulos EN. Advances in energy systemsengineering. Ind Eng Chem Res 2010;50:4915–26.

[6] Liu P, Pistikopoulos EN, Li Z. A multi-objective optimization approach topolygeneration energy systems design. AIChE J 2010;56:1218–34.

[7] Liu P, Gerogiorgis DI, Pistikopoulos EN. Modeling and optimization ofpolygeneration energy systems. Catal Today 2007;127:347–59.

[8] Chiesa P, Consonni S, Kreutz T, Williams R. Co-production of hydrogen,electricity and CO2 from coal with commercially ready technology. Part A:performance and emissions. Int J Hydrogen Energy 2005;30:747–67.

[9] Kreutz T, Williams R, Consonni S, Chiesa P. Co-production of hydrogen,electricity and CO2 from coal with commercially ready technology. Part B:economic analysis. Int J Hydrogen Energy 2005;30:769–84.

[10] Yu G, Xu Y, Hao X, Li Y, Liu G. Process analysis for polygeneration of Fischer–Tropsch liquids and power with CO2 capture based on coal gasification. Fuel2010;89:1070–6.

[11] Zhou L, Hu S, Li Y, Zhou Q. Study on co-feed and co-production system basedon coal and natural gas for producing DME and electricity. Chem Eng J2008;136:31–40.

[12] Normann F, Thunman H, Johnsson F. Process analysis of an oxygen lean oxy-fuel power plant with co-production of synthesis gas. Energy Convers Manage2009;50:279–86.

[13] Li H, Hong H, Jin H, Cai R. Analysis of a feasible polygeneration system forpower and methanol production taking natural gas and biomass as materials.Appl Energy 2010;87:2846–53.

[14] Qian Y, Liu J, Huang Z, Kraslawski A, Cui J, Huang Y. Conceptual design andsystem analysis of a poly-generation system for power and olefin productionfrom natural gas. Appl Energy 2009;86:2088–95.

[15] Pellegrini LA, Soave G, Gamba S, Langè S. Economic analysis of a combinedenergy–methanol production plant. Appl Energy 2011;88:4891–7.

[16] Liu G-j, Li Z, Wang M-h, Ni W-d. Energy savings by co-production: a methanol/electricity case study. Appl Energy 2010;87:2854–9.

[17] Gao L, Li H, Chen B, Jin H, Lin R, Hong H. Proposal of a natural gas-basedpolygeneration system for power and methanol production. Energy2008;33:206–12.

[18] Zhang J, Zhao Z, Zhang G, Xi Z, Zhao F, Dong L, et al. Pilot study on jetting pre-oxidation fluidized bed gasification adapting to caking coal. Appl Energy2013;110:276–84.

[19] Ngo SI, Nguyen TDB, Lim Y-I, Song B-H, Lee U-D, Choi Y-T, et al. Performanceevaluation for dual circulating fluidized-bed steam gasifier of biomass usingquasi-equilibrium three-stage gasification model. Appl Energy2011;88:5208–20.

[20] Cordiner S, De Simone G, Mulone V. Experimental–numerical design of abiomass bubbling fluidized bed gasifier for paper sludge energy recovery. ApplEnergy 2012;97:532–42.

[21] Link S, Arvelakis S, Paist A, Martin A, Liliedahl T, Sjöström K. Atmosphericfluidized bed gasification of untreated and leached olive residue, and co-gasification of olive residue, reed, pine pellets and Douglas fir wood chips. ApplEnergy 2012;94:89–97.

[22] Kim YD, Yang CW, Kim BJ, Kim KS, Lee JW, Moon JH, et al. Air-blowngasification of woody biomass in a bubbling fluidized bed gasifier. Appl Energy2013.

[23] Boateng AA, Daugaard DE, Goldberg NM, Hicks KB. Bench-scale fluidized-bedpyrolysis of switchgrass for bio-oil production�. Ind Eng Chem Res2007;46:1891–7.

[24] Tyler RJ. Flash pyrolysis of coals. 1. Devolatilization of a Victorian brown coalin a small fluidized-bed reactor. Fuel 1979;58:680–6.

[25] Stiles HN, Kandiyoti R. Secondary reactions of flash pyrolysis tars measured ina fluidized bed pyrolysis reactor with some novel design features. Fuel1989;68:275–82.

[26] Kato K, Wen CY. Bubble assemblage model for fluidized bed catalytic reactors.Chem Eng Sci 1969;24:1351–69.

[27] Olsbye U, Wurzel T, Mleczko L. Kinetic and reaction engineering studies of dryreforming of methane over a Ni/La/Al2O3 catalyst. Ind Eng Chem Res1997;36:5180–8.

[28] Basu P. Combustion of coal in circulating fluidized-bed boilers: a review. ChemEng Sci 1999;54:5547–57.

[29] Yue G, Yang H, Lu J, Zhang H. Latest development of CFB boilers in China. In:Proceedings of the 20th international conference on fluidized bedcombustion. Xi’an, China: Springer; 2010. p. 3–12.

[30] Duan L, Liu D, Chen X, Zhao C. Fly ash recirculation by bottom feeding on acirculating fluidized bed boiler co-burning coal sludge and coal. Appl Energy2012;95:295–9.

[31] Liukkonen M, Hälikkä E, Hiltunen T, Hiltunen Y. Dynamic soft sensors for NOxemissions in a circulating fluidized bed boiler. Appl Energy 2012;97:483–90.

[32] Wu Y, Wang C, Tan Y, Jia L, Anthony EJ. Characterization of ashes from a100kWth pilot-scale circulating fluidized bed with oxy-fuel combustion. ApplEnergy 2011;88:2940–8.

[33] Fang MX, Luo ZY, Li XT, Wang QH, Ni MJ, Cen KF. A multi-product cogenerationsystem using combined coal gasification and combustion. Energy1998;23:203–12.

[34] Wang Q, Luo Z, Fang M, Ni M, Cen K. Design of a 12 MW tri-cogenerationsystem of gas, heat and power. In: Donald WG, Geiling PE, editors. Proceedingsof the 16th international conference on fluidized bed combustion. Reno,Nevada: ASME; 2001. p. 2001.

[35] Wang Q, Luo Z, Fang M, Ni M, Cen K. Development of a 12 MW multi-generation of gas, steam and power. J Fuel Chem Technol 2002;30:141–6.

[36] Fang M, Cen J, Wang Q, Shi Z, Luo Z, Cen K. 25 MW circulating fluidizedbed heat-power-coal gas poly-generation installation. J Power Eng2007;27:635–9.

[37] Lv G, Wang H, Ma W, Yu C. Energy and exergy analysis for 300MW thermalsystem of Xiaolongtan power plant. In: International conference on computerdistributed control and intelligent environmental monitoring (CDCIEM); 2011.p. 180–4.

[38] Bartholomew CH. Mechanisms of catalyst deactivation. Appl Catal A2001;212:17–60.

[39] Li CS, Suzuki K. Resources, properties and utilization of tar. Resour ConservRecycl 2010;54:905–15.

[40] Edwards JH, Schluter K, Tyler RJ. Upgrading of flash pyrolysis tars to syntheticcrude oil: 2. Second-stage hydrotreatment using nickel/molybdenumcatalysts. Fuel 1986;65:202–7.

[41] Edwards JH, Schluter K, Tyler RJ. Upgrading of flash pyrolysis tars to syntheticcrude oil: 1. First stage hydrotreatment using a disposable catalyst. Fuel1985;64:594–9.

[42] Kohl AL, Nielsen R. Gas purification. Gulf Professional Publishing; 1997.[43] Burr B, Lyddon L. A comparison of physical solvents for acid gas removal. Gas

Processors’ Association Convention, Grapevine, TX2008.[44] Sweny J. Synthetic fuel gas purification by the SELEXOL process. Prepr Am

Chem Soc Div Fuel Chem. 1976;18.[45] Vernon PD, Green ML, Cheetham AK, Ashcroft AT. Partial oxidation of methane

to synthesis gas. Catal Lett 1990;6:181–6.[46] Van Hook JP. Methane-steam reforming. Catal Rev 1980;21:1–51.[47] Wurzel T, Malcus S, Mleczko L. Reaction engineering investigations of CO2

reforming in a fluidized-bed reactor. Chem Eng Sci 2000;55:3955–66.[48] Tijm PJA, Waller FJ, Brown DM. Methanol technology developments for the

new millennium. Appl Catal A 2001;221:275–82.

1314 Z. Guo et al. / Applied Energy 113 (2014) 1301–1314

[49] Bussche KMV, Froment GF. A steady-state kinetic model for methanolsynthesis and the water gas shift reaction on a commercial Cu/ZnO/Al2O3

catalyst. J Catal 1996;161:1–10.[50] Zhu Y. Evaluation of gas turbine and gasifier-based power generation system.

North Carolina State University; 2004.[51] GE Energy. Heavy duty gas turbine products, 2009.http://www.ge.com/

mining/docs/2981884_1346774901_GE_Heavy_Duty_Gas_Turbines.pdf[52] Chen L, Jiang Q, Song Z, Posarac D. Optimization of methanol yield from a Lurgi

reactor. Chem Eng Technol 2011;34:817–22.[53] Sotudeh-Gharebaagh R, Legros R, Chaouki J, Paris J. Simulation of circulating

fluidized bed reactors using ASPEN PLUS. Fuel 1998;77:327–37.[54] Liu B, Yang X, Song W, Lin W. Process simulation development of coal

combustion in a circulating fluidized bed combustor based on Aspen Plus.Energy Fuels 2011;25:1721–30.

[55] Wang B, Dong L, Wang Y, Matsuzawa Y, Xu G. Process analysis of lignitecirculating fluidized bed boiler coupled with pyrolysis topping. In: Yue G,Zhang H, Zhao C, Luo Z, editors. Proceedings of the 20th internationalconference on fluidized bed combustion. Berlin, Heidelberg: Springer; 2010. p.706–11.

[56] Grant DM, Pugmire RJ, Fletcher TH, Kerstein AR. Chemical model of coaldevolatilization using percolation lattice statistics. Energy Fuels1989;3:175–86.

[57] Anthony DB, Howard JB. Coal devolatilization and hydrogastification. AIChE J1976;22:625–56.

[58] Solomon PR, Hamblen DG, Carangelo RM, Serio MA, Deshpande GV. Generalmodel of coal devolatilization. Energy Fuels 1988;2:405–22.

[59] Niksa S, Kerstein AR. Flashchain theory for rapid coal devolatilization kinetics.1. Formulation. Energy Fuels 1991;5:647–65.

[60] Maffei T, Sommariva S, Ranzi E, Faravelli T. A predictive kinetic model of sulfurrelease from coal. Fuel 2011.

[61] Maffei T, Ranzi E, Frassoldati A, Faravelli T. One step kinetic model of coalpyrolysis for CFD applications. In: XXXIV meeting of the Italian Section of theCombustion Institute, Rome; 2011.

[62] Sommariva S, Maffei T, Migliavacca G, Faravelli T, Ranzi E. A predictive multi-step kinetic model of coal devolatilization. Fuel 2010;89:318–28.

[63] Hamelinck CN, Faaij APC, den Uil H, Boerrigter H. Production of FTtransportation fuels from biomass; technical options, process analysis andoptimisation, and development potential. Energy 2004;29:1743–71.

[64] Chen C. A technical and economic assessment of CO2 capture technology forIGCC power plants. Carnegie Mellon University; 2005.

[65] Tarun CB, Croiset E, Douglas PL, Gupta M, Chowdhury MHM. Techno-economicstudy of CO2 capture from natural gas based hydrogen plants. Int J GreenhouseGas Control 2007;1:55–61.

[66] Li S, Gao L, Zhang X, Lin H, Jin H. Evaluation of cost reduction potential for acoal based polygeneration system with CO2 capture. Energy 2012;45:101–6.

[67] Peters MS, Timmerhaus KD, West RE, Timmerhaus K, West R. Plant design andeconomics for chemical engineers. NY: McGraw-Hill; 1968.

[68] Govin O, Diky V, Kabo G, Blokhin A. Evaluation of the chemical exergy of fuelsand petroleum fractions. J Therm Anal Calorim 2000;62:123–33.

[69] Wikipedia. List of Renminbi exchange rates.[70] Lu X, Yu Z, Wu L, Yu J, Chen G, Fan M. Policy study on development and

utilization of clean coal technology in China. Fuel Process Technol2008;89:475–84.

[71] Feasibility study report of a 200000 tonne-scale coal tar hydrogenrationproject; 2011 [in Chinsese].

[72] Meerman JC, Ramírez A, Turkenburg WC, Faaij APC. Performance of simulatedflexible integrated gasification polygeneration facilities, Part B: economicevaluation. Renew Sustain Energy Rev 2012;16:6083–102.

[73] Xue B. Economic evaluation of dual-gas polygeneration system. Taiyuan:Taiyuan University of Technology; 2010.

[74] Huijgen WJJ, Comans RNJ, Witkamp G-J. Cost evaluation of CO2 sequestrationby aqueous mineral carbonation. Energy Convers Manage 2007;48:1923–35.

[75] Yang C-J, Jackson RB. China’s growing methanol economy and its implicationsfor energy and the environment. Energ Policy 2012;41:878–84.