lable at ScienceDirect Energy 103 (2016) 715e724 Contents lists avai Energy journal homepage: www.elsevier .com/locate/energy Carbon exergy tax applied to biomass integrated gasification combined cycle in sugarcane industry Valdi Freire da Fonseca Filho a, Jos�e Alexandre Matelli b, Jos�e Antonio Perrella Balestieri b, * a Embraer S.A. e S~ao Jos�e dos Campos (Matriz), Avenida Brigadeiro Faria Lima, 2.170, S~ao Jos�e dos Campos, SP, Brazil b Univ Estadual Paulista, Campus of Guaratinguet�a, Department of Energy, Avenida Dr. Ariberto Pereira da Cunha, 333, Guaratinguet�a, SP, Brazil a r t i c l e i n f o Article history: Received 28 July 2015 Received in revised form 25 February 2016 Accepted 28 February 2016 Available online 29 March 2016 Keywords: Gasification Biomass IGCC Carbon exergy tax CO2 emission * Corresponding author. E-mail addresses: valdi.filho@embraer.com.br (V.F unesp.br (J.A. Matelli), perrella@feg.unesp.br (J.A. Perr http://dx.doi.org/10.1016/j.energy.2016.02.161 0360-5442/© 2016 Elsevier Ltd. All rights reserved. a b s t r a c t The development of technologies based on energy renewable sources is increasing worldwide in order to diversify the energy mix and satisfy the rigorous environmental legislation and international agreements to reduce pollutant emission. Considering specific characteristics of biofuels available in Brazil, studies regarding such technologies should be carried out aiming energy mix diversification. Several technol- ogies for power generation from biomass have been presented in the technical literature, and plants with BIGCC (biomass integrated gasification combined cycle) emerge as a major technological innovation. By obtaining a fuel rich in hydrogen from solid biomass gasification, BIGCC presents higher overall process efficiency than direct burning of the solid fuel in conventional boilers. The objective of this paper is to develop a thermodynamic and chemical equilibrium model of a BIGCC configuration for sugarcane bagasse. The model embodies exergetic cost and CO2 emission analyses through the method of CET (carbon exergy tax). An exergetic penalty comparison between the BIGCC technology (with and without CO2 capture and sequestration), a natural gas combined cycle and the traditional steam cycle of sugar- cane sector is then presented. It is verified that the BIGCC configuration with CO2 capture and seques- tration presents technical and environmental advantages when compared to traditional technology. © 2016 Elsevier Ltd. All rights reserved. 1. Introduction The need to develop technologies based on renewable energy sources, such as biomass, growsworldwide. This development aims energy mix diversification and also meeting rigorous environ- mental legislation and international agreements to reduce pollutant emission. Biomass had a bad reputation for a long time. People who are not familiar with the opportunities and benefits from the use of biomass for energy and who have only little knowledge about biomass conversion technologies tend to have prejudices. People transfer such experience to new biomass energy plants and tend to think that the techniques for the use of biomass for energy are out- of-date, i.e., old fashioned, no high technology and low efficiency Ref. [28]. However, newmethodologies to estimate its potential as a feasible energy source, new high efficiency energy conversion technologies presented in demonstration plants and the biomass . Fonseca Filho), matelli@feg. ella Balestieri). renewability contributed to change this unfavorable image in recent years. Furthermore, availability may be very high since some industrial sectors generates large amount of biomass as by- products. Biomass technical and economic feasibility depends on new energy conversion processes and technological improvement of traditional processes because, from a commercial perspective, there are still no high reliability technologies for small scale gen- eration at competitive costs [12]. BIGCC (Biomass integrated gasi- fication combined cycle) is a promising technology that may contribute to a rational and efficient biomass use, but biomass di- versity in terms of physical characteristics and chemical composi- tion (for instance, black liquor and sugarcane bagasse are very different biomass) are still barriers to overcome. These difficulties, along with biomass advantages such as renewability, low sulfur emissions and neutral carbon emissions justify studies in BIGCC technology and its potential to reduce emissions. According to [4], Rankine-based cogeneration cycle is the foundation for energy generation in Brazilian sugar/ethanol in- dustry. Traditionally, backpressure steam turbines are used in a typical configuration, but more advanced technological routes are Delta:1_given name Delta:1_surname Delta:1_given name mailto:valdi.filho@embraer.com.br mailto:matelli@feg.unesp.br mailto:matelli@feg.unesp.br mailto:perrella@feg.unesp.br http://crossmark.crossref.org/dialog/?doi=10.1016/j.energy.2016.02.161&domain=pdf www.sciencedirect.com/science/journal/03605442 http://www.elsevier.com/locate/energy http://dx.doi.org/10.1016/j.energy.2016.02.161 http://dx.doi.org/10.1016/j.energy.2016.02.161 http://dx.doi.org/10.1016/j.energy.2016.02.161 Table 1 Ultimate and proximate analysis for sugarcane bagasse [8]. Ultimate analysis (dry basis) Carbon 44.6% Hydrogen 5.80% Nitrogen 0.60% Oxygen 44.5% Sulfur 0.10% Chlorine 0.02% Mineral oxides 4.38% Proximate analysis Moisture 50.2% Ashesa 2.1% Carbona 18.0% Volatilea 79.9% a Dry basis. V.F. Fonseca Filho et al. / Energy 103 (2016) 715e724716 considered nowadays due to the changes in electricity market regulation that allow exceeding power selling to the grid. The au- thors analyzed a steam condensing cycle and a BIGCC under different cost scenarios and concluded that BIGCC requires 48% cost reduction in order to be competitive with conventional bagasse burning plants. Coal gasification is a dominated technology, and biomass gasi- fication is still under development; their integration with CCS (CO2 capture and storage) industry is not yet adequately established because the components do not currently function together in the manner required for large-scale CO2 reduction [27]. Gasification process involves biomass devolatilization and chemical degrada- tion in order to produce a low heating value fuel gas. Air or steam is typically used in biomass gasification, resulting in a heating value around 5.5 MJ/m3 (n). The use of pure oxygen instead of air can provide a fuel gas with heating value up to 20 MJ/m3 (n). However, the costs are quite high and the use of pure oxygen is only rec- ommended to produce syngas [18]. Uddi and Barreto [24] estimated the CO2 mitigation costs of biomass-fired cogeneration technologies with CCS considering BIGCC and steam condensing cycle. A cogeneration system based on natural gas combined cycle without CO2 capture was taken as the reference system. Results shows that BIGCC with CO2 capture and storage was found very energy and emission efficient and cost competitive when compared to other conversion systems. The cost-effectiveness of imposing a carbon tax for reducing greenhouse gas emissions is discussed by Ref. [22]. New bioenergy technologies for the year 2030 are then considered, including BIGCC with and without carbon capture and storage. Results indi- cate that a carbon tax on fossil fuels performs cost-effectively regarded the considered policy targets (greenhouse gas emission reduction and fossil fuel substitution) if bioenergy systems with carbon capture and storage are not available. Klein et al. [11] considered IGCC with CO2 capture an important alternative to mitigate emissions. However, costs are high because the cycles are highly complex, especially regarding CO2 capture and liquefying. Thus, these systems are not cost-competitive against conventional technology using coal, natural gas or even direct-firing biomass. According to Rhodes et al. (2005) [29], the power cost generated by an IGCC may be attractive if the cost of emitted CO2 is internalized. Recent published studies related to CO2 capture discuss the best available technologies, mainly when coal is the fuel to be gasified [14,25], stating the appropriateness of absorption methods. Advanced concepts, as the integration of fuel cells [5] and of un- derground coal gasification [17] into IGCC with CO2 capture, has been recently proposed. A proper mechanism for taxing CO2 emissions should take into account the plant inefficiency, so that more inefficient plants should be penalized. Exergy destruction and exergy lost are the basis for the CET (carbon exergy tax), a CO2 taxing method pro- posed in the works of [20,19,1,2]. CET method relates the CO2 emissions to the efficient use of exergetic resources and, conse- quently, to the plant efficiency. This work presents a comparative analysis of thermal cycles e a traditional one, based on CST (condensing steam turbine), two advanced plants based on biomass gasification combined cycle with (BIGCC-CCS) and without (BIGCC-nCCS) carbon capture and sequestration, respectively, and a NGCC (natural gas-fired com- bined cycle) e by using the CET (carbon exergy tax). For applying such method of comparison to the configurations, it was needed to develop a rule for the original CETmethod to compare fossil and renewable fuels, as well as the CCS. A thermodynamic and chemical equilibrium model of a BIGCC configuration for sugarcane bagasse, considering gasification with pure oxygen, was then developed. Themodel also embodies exergetic cost and CO2 emission analyses through the method of CET (carbon exergy tax). The main contributions of this work are: i) concept of a BIGCC with CCS, using pure oxygen instead of air in the biomass gasifi- cation process; ii) application of CET method to renewable ther- moelectric power plants, which was not considered in the original works of [1,2]; iii) setting how biomass CO2 emission can be treated in CET method; iv) confirming CET method as an instrument for renewable energy policies regarding biomass-fired power plants. 2. Methodology 2.1. Biomass gasification model In this section, the biomass gasification is modeled. First, it is considered gasification with air and the model is validated against experimental results found in the literature. After validation, gasi- fication with pure oxygen is then considered. Biomass chemical composition can be determined through ul- timate and proximate analysis according standard tests (e.g. ASTM E870). Hassuani et al. [8] presented typical results from ultimate and proximate analysis of sugarcane bagasse, as shown in Table 1. For simplification, chlorine andmineral oxides are not considered, so that bagasse empirical formula results CH15.6N0.011O0.75S0.00083, with molecular weight equal to Mb ¼ 25.6 kg/kmol. The biomass is considered briquette-shaped with moisture content w* ¼ 5.31% [8]. Gasification process is modeled according to the following hypothe- sis: i) steady state; ii) gasification products considered ideal gases; iii) products and reactants are in chemical equilibrium; iv) reaction takes place in an isothermal fluidized bed. The model is based in one global gasification reaction and three chemical equilibrium reactions. As a result, syngas chemical composition and its lower heating value are obtained. This syngas is then considered the prime mover fuel in the IGCC model described in Section 3.2. Equation (1) shows the global biomass-air gasification reaction, in which w is the number of moles of water in the bagasse, m is the required number of moles of oxygen and ai is the stoichiometric co- efficient of the i-th product. Since gasification is basically a sub- stoichiometric combustion, there is no oxygen in the reactionproducts. CH1:56N0:011O0:75S0:00083 þwH2OþmðO2 þ 3:76N2Þ/ /a1COþ a2CO2 þ a3H2 þ a4CH4 þ a5H2Oþ a6SO2 þ a7N2 þa8C2H4 (1) Three other reactions in chemical equilibrium are considered: carbon monoxide-water shifting (Eq. (2)), ethylene decomposition (Eq. (3)) and methane-water shifting (Eq. (4)). Gas turbine (adapted for syngas): Steam turbine: Siemens SGT5-8000H Siemens SST 900 Power output (ISO): 375 MW Power output: 250 MW Pressure ratio: 19.2 Inlet steam pressure: 16.5 MPa Exhaust gases flow: 820 kg/s Inlet steam temperature: 585 �C Exhaust gases temperature: 625 �C Extraction pressure: 0.5 MPa V.F. Fonseca Filho et al. / Energy 103 (2016) 715e724 717 COþ H2O4CO2 þ H2 (2) C2H4 þ 2H2O4CH4 þ CO2 þ 2H2 (3) CH4 þ H2O4COþ 3H2 (4) Bagasse moisture content w� is the ratio between the mass of water in the bagasse and themass of wet bagasse. It can be related to the number of moles of water in the bagasse w according to Eq. (5). w ¼ Mbw� MH2oð1�w�Þ (5) The stoichiometric coefficients in Eq. (1) are determined through chemical species balances (Eqs. (6)e(10)) combined with chemical equilibrium constant from the three equilibrium reactions (Eqs. (11)e(13)), in which Xi represents the molar fraction of the i- th product (Eq. (14)). Equilibrium constants in Eqs. (11)e(13) are K1 ¼ 1:136, K2 ¼ 1:810� 106 and K3 ¼ 707 [13]. a1 þ a2 þ a4 þ 0:5a8 ¼ 1 (6) 2a3 þ 4a4 þ 2a5 þ 4a8 ¼ 1:56þ 2w (7) 2a7 ¼ 0:011þ 7:56m (8) a1 þ 2a2 þ a5 þ 2a6 ¼ 0:75þwþ 2m (9) a6 ¼ 0:00083 (10) K1 ¼ XCO2 XH2 XCOXH2O (11) K2 ¼ XCH4 XCO2 X2 H2 XC2H4 X2 H2O � P P0 ��1 (12) K3 ¼ XCOX 3 H2 XCH4 XH2O (13) Xi ¼ aiP ai (14) Once syngas composition is known, it is easy to calculate syngas thermodynamic properties in molar (or volume) basis, such as lower heating value (Eq. (15)) and molar mass (Eq. (16)). LHVs ¼ X i XiLHVi (15) Ms ¼ X i XiMi (16) 2.2. BIGCC-CCS model The BIGCC-CCS1 configuration proposed in the present work is based on a previous configuration presented by Ref. [3]; as shown in Fig. 1. The gasification section of this configuration is similar to the actual gasifier presented in Ref. [8]; so that their experimental data can be used to validate the model. 1 Biomass integrated gasification combined cycle with carbon capture and storage. In order to establish a proper and more accurate thermody- namic model, the BIGCC-CCS configuration is detailed bellow and depicted in Fig. 2. -ASU (Air separation unit) is based on a N2/O2 double separation column; -Circulating fluidized bed gasifier, pressurized, with pure oxy- gen instead of air; -Pre-treated biomass supplied to the gasifier; -Syngas treatment involves drying and acid gases removal; -CO2 removal through physical absorption; -CO2 cryogenic liquefaction through mechanical compression with intercooler; Proposed BIGCC-CCS plant is divided into four main sections: air separation, gasification, power generation and CO2 removal. In or- der to provide pure oxygen for the plant gasifier, an ASU (air sep- aration unit) is required. An intercooled compressor demands 31.5MWof power to feed the ASUwith air at 0.48MPa (stream 2) at a specific consumption of 62 kWh/ton. ASU cryogenic cycle de- mands around 25.4 MW of power (equivalent to 50 kWh/ton), so that the air is liquefied for N2/O2 separation in a double column. After vaporization, ASU vents nitrogen to atmosphere (stream 3) and oxygen is compressed at 2.5 MPa to feed the gasifier (stream 5). The oxygen compressor demands around 8.7 MW of power (73 kWh/ton). Oxygen from air separation section reacts with pre-treated biomass (stream 6) in the gasification section in order to produce syngas in a pressurized, circulating fluidized bed gasifier. Gasifica- tion section also involves syngas cooling, drying and acid gases removal. Part of steam required in the power generation section (stream 28) is produced by cooling the syngas from gasifier (stream 7). Power generation section of the proposed BIGCC-CCS plant is based on a combined cycle designed to produce part of the demanded power by generating it by a gas turbine (stream 31) and a steam turbine (stream 34). Plant ancillary power related to pumps and compressors are provided by these turbines. More specifically, power from gas turbine drives O2 compressor (stream 40) and ASU refrigeration unit (stream 41); power from steam turbine drives steam cycle pumps (streams 37 and 38), air compressor (stream 39) and CO2 compressor (stream 39). Plant also demands water and ancillary thermal energy, mainly for equipment cooling purposes. An extraction able to provide steam at 0.5 MPa is available in the steam turbine casing, which makes the plant suitable for an eventual cogeneration application. Turbine specifications are given as follows: Finally, CO2 removal section separates carbon dioxide from the exhaust gases through physical absorption, as described by Ref. [3]. It must be considered that there is less experience with large-scale CCS biomass plants compared with CCS coal plants. After separa- tion, CO2 is compressed, liquefied and either stored through some carbon-capture technology or distributed as raw material for other industrial processes. Power required in this section is about 100 MW. acid gases O2 syngas syngas rich of H2 and CO2 N2 sugar cane bagasse air CO2 steam steam ASU physical absorption + liquefaction of CO2 gasifier process cleaning and cooling Fig. 1. BIGCC-CCS configuration according to [3]. 8 CO2 liquefaction sector 15 (H 2O) Gasification sector Air separation sector 49 21 23 24 22 2954 1928 27 26 20 1310 7Bagasse 6 5 4 (O2) 3 (N2) 2 1 (air) 9 11 12 14 18 50 25 36 16 (N2) 17 Syngas cleaning/cooling Power generation sector 39 40 38 4155 42 ASU 56 47 53 58 59 57 45 44 CO2 48 37 35 Process51 46 30 31 40 41 Gasifier 34 37 38 39 48 43 CO2 removal and capture 32 52 33 Fig. 2. Proposed BIGCC-CCS plant. V.F. Fonseca Filho et al. / Energy 103 (2016) 715e724718 BIGCC-CCS performance is predicted through a thermody- namic model based on mass (Eq. (17)), energy (Eq. (18)) and exergy (Eq. (19)) balances in each plant component, consid- ering steady state condition. Exergy calculation is necessary for the CET (carbon exergy tax) analysis described in the next section. Physical and chemical exergy equations were obtained from literature, being fuels and mixture of gases (synthesis gas and exhaust gas) calculated according to specific models that take into account lower heating value and molar fraction of each component, respectively [15]; reference state is stated by P0 ¼ 101 kPa and T0 ¼ 298 K. Technical data from equipment and design parameters either adopted (e.g. ambient tempera- ture) or taken from literature (e.g. gasification pressure) pro- vides additional information to the model, which results in a set of equation solved with EES (Engineering Equation Solver) software [7]. Results from BIGCC model are shown ahead in Section 3.2. X _mo ¼ X _mi (17) _Q � _W ¼ X _moho � X _mihi (18) � 1� T0 T � _Q � _W ¼ X _mobo � X _mibi þ _ED (19) 2.3. CET-based penalty analysis From contributions presented in the literature for pollutant emissions internalization, CET (carbon exergy tax) is considered in the analysis presented here. For [1], the aim of such a procedure is to assign a fee related to CO2 emissions of energy plants, so that the V.F. Fonseca Filho et al. / Energy 103 (2016) 715e724 719 fee is based only on thermodynamic analysis (efficiency and exergy) rather than policy guidelines.2 A proper mechanism for taxing CO2 emissions should take into account the plant inefficiency, so that more inefficient plants should be penalized. Exergy destruction and exergy lost are the basis for CET model in the works of [20,19]; which relates CO2 emissions to the efficient use of exergetic resources and, conse- quently, to the plant efficiency. The method requires exergy costs calculation of all streams, which is done through the TEC (Theory of Exergetic Cost) by Ref. [26]. For the sake of brevity, just a quick overview of the main equations of both CET and TEC are presented in this work, since a comprehensive presentation of both methods can be found in the literature. CO2 cost emission is composed of the following elements: -Destroyed exergy cost (Eq. (20)) is related to the sum of destroyed exergy of all j-th (BD;j) plant components and their respective unitary cost (cD;j). CD ¼ X cD;jBD;j (20) -Residual exergy cost (Eq. (21)) is related to the exergy rejected from pollutant streams to the environment [20,19]. Here, the unitary cost of the j-th component cR;j can be seen as an in- efficiency index associated to residual exergy from the exhaust gases, since this exergy is a potential product. Considering an utilization factor fc, both destroyed and residual exergy cost determine the inefficiency penalty Pε (Eq. (22)). CR ¼ X cR;jBR;j (21) Pε ¼ fcðCD þ CRÞ (22) -CO2 emission index (Eq. (23)) is a non-dimensional ratio be- tween plant CO2 emission and exergy related to the plant products Bp, also taking into account a reference CO2 emission index I0. Santarelli et al. [19] conceived I�C as a dimensionless parameter. In order to keep it non-dimensional, they established an arbitrary reference value I0 ¼ 1 kgCO2 =kWh because they used kWh as energy/exergy unit throughout their work. In the present work it is chosen kJ as energy/exergy unit, so that I0 is arbitrarily set to 1 kgCO2 =kJ. The higher I�C the higher the impacts related to the plant CO2 emissions. It can also be expressed in terms of plant exergetic efficiency ε and fuel lower heating value. Thus, three major factors penalize plant performance related to CO2 emission: high gross emission, low exergetic ef- ficiency and poor fuel. I�C ¼ _mCO2 BPI0 ¼ _mCO2 εLHV _mFI0 (23) -CO2 emission cost (Eq. (24)) is the product of CO2 emission index and inefficiency penalty. Borchiellini et al. [1,2], developed an approximated equation useful for preliminary analysis (Eq. (25)) from basic fuel parameters: flow ( _mF), heating value and carbon fraction (C). 2 However, CET can be seen as an instrument of energy policy that promotes the utilization of advanced energy systems, as discussed by Ref. [21] when he applied the procedure to four conventional and one advanced energy schemes. CCO2 ¼ PεI�C (24) CCO2 z C _mFðBD þ BRÞ LHV _mF � ðBD þ BRÞ (25) - CET: carbon exergy tax (Eq. (26)) is the resulting cost per mass unit of CO2 emitted. Borchiellini et al. [1] related it to a penalty (or cost increasing) for both power (Eq. (27)) and useful thermal energy (Eq. (28)) generated by a plant. CET ¼ CCO2 _mCO2 (26) DCe ¼ CET _mCO2 _We (27) DCt ¼ CET _mCO2 _Qt (28) In order to demonstrate CET as an instrument of energy policy that promotes the utilization of advanced energy systems [21], applied the procedure to four power plants: three conventional ones (simple gas turbine; cogeneration with regenerated gas tur- bine; two-pressure levels combined cycle) and an advanced one (pressurized internal reforming solid oxide fuel cell with gas tur- bine). The author found that just simple gas turbine plant presented power cost higher than the advanced plant for every value of car- bon tax (USD=tonCO2 ). On the other hand, advanced plant power cost is higher than values from regenerated gas turbine cycle and two-pressure levels combined cycle for no and lower carbon tax values, but they are very close for higher carbon tax. In this work, four configurations are compared. Two of them represent conventional technologies: a CST (condensing steam turbine) Rankine cycle burning sugarcane bagasse (Fig. 3) and a NGCC (natural gas combined cycle, Fig. 4). The other two represents advanced technologies: a biomass (sugarcane bagasse) gasification combined cycle with carbon capture and sequestration (BIGCC-CCS, Fig. 2) and a biomass gasification combined cycle with no carbon capture and sequestration (BIGCC-nCCS, Fig. 5). As far as possible, the same technical and economic conditions were assumed for all plant configurations: -For biomass-based configurations (BIGCC-CCS, BIGCC-nCCS and CST), fuel properties are determined as presented in Sec- tion 3.1; for the NGCC configuration, natural gas properties are taken from methane for simplicity; -All configurations are proposed as thermal power plants, so that net electric power is assumed to be about 340 MW or higher; -It is assumed 8000 h/y of operation in all cases; biomass and natural gas prices are taken according to recent Brazilian eco- nomic scenario (natural gas price is lower than biomass price in USD/kJ); environmental conditions are P ¼ 101 kPa and 303 K; -It is assumed an annual interest rate equal I ¼ 14% and n ¼ 20 years investment, resulting in a capital recovery factor3 equal to CFR ¼ 0.1510 (year�1). TPC (Total power cost) is calculated by summing the CO2 penalty to the LCC (levelized capital cost), as shown in Eq. (29). 3 CRF ¼ ið1þ iÞn=½ð1þ iÞn � 1�. 31 (air) 3 (bagasse) 4 20 13 9 16 17 235 2 1 10 36 21 28 25 7 14 15 8 6 12 11 18 Process26 27 22 19 24 Fig. 3. CST plant. 49 21 23 24 22 2954 20 1310 9 11 12 14 18 50 25 36 Power generation sector 59 57 45 37 35 Process51 46 30 31 34 37 43 32 33 Fig. 4. NGCC plant. V.F. Fonseca Filho et al. / Energy 103 (2016) 715e724720 LCC is related to CRF (capital recovery factor), CC (capital cost) (CC) and CF (capacity factor) (CF) according to Eq. (30). TPC ¼ LCCþ DCe (29) LCC ¼ CC$CRF 8760$CF (30) In the previous works, CET was applied to fossil fuels plants (natural gas and coal), so the generated CO2 was anthropogenic e this is the case of NGCC in the present paper. However, traditional CST technology and innovative BIGCC-nCCS, both based on sug- arcane bagasse, just emit the CO2 that was previously recovered by sugarcane during its growing, in a cyclic non-anthropogenic process. For BIGCC-CCS, sugarcane bagasse is gasified and non- anthropogenic CO2 is captured and sequestered. Fig. 6 illus- trates the described structure for fuels, technologies and CO2 emissions. 3. Results and discussion 3.1. Biomass gasification Stoichiometric number of moles of oxygen (Eq. (1)) is found mg ¼ 1.016. The fuel-air equivalence ratio was varied from 1 to 3, and for F ¼ mg/m ¼ 2.23, gasification model solving results m ¼ 0.455 as the required number of moles of oxygen to perform the gasification reaction. Syngas composition result from biomass- air gasification model is presented in Table 2 and is validated against experimental results found by Ref. [8]. The model is then solved considering m ¼ 0.308 mol of pure oxygen instead of air e this is the required number of moles of pure oxygen that result the same adiabatic flame temperature of biomass air-gasification. Such a temperature is considered a proper design parameter for an isothermal gasifier. Syngas composition obtained in this case is also presented in Table 2. Results from biomass-air model differ up to 11% from experi- mental results by Ref. [8]; but the model is highly inaccurate for methane and ethylene. One of the hypotheses adopted in the model, chemical equilibrium is not observed in an actual gasifier. Thus, a kinetic model with more intermediate reactions should be adopted instead. However, methane and ethylene represent only 4% of syngas composition and errors related to their concentration do not affect significantly the lower heating value. Indeed, lower heating value found from model is 4.04 MJ/m3 (n) against 4.10 MJ/ m3 (n) presented in Ref. [8]; corresponding to a difference of 1.46%. 3.2. BIGCC-CCS model Results from BIGCC-CCS model presented in Section 3.2 are presented in Table 3. All streams presented in Table 3 refer to those depicted in Fig. 2. It must be advised that non-material streams (Table 3c) just identifies the occurrence of irreversible processes inside the equipment due to their inefficiencies. The model presented in Section 3.2 applied to the CST (Fig. 3), NGCC (Fig. 4) and BIGCC-nCCS (Fig. 5) generates similar results to those presented in Table 3. These results are required for the CET- based penalty model, but they are not presented for the sake of brevity. Results referring to performance of the different cycles are summarized in Table 4, in which total net power output considers the generated power in gas and steam turbines minus the power used for driving pumps and compressors. 3.3. CET-based penalty Results from CET model are shown in Table 5. Parameter A ex- presses the mass of CO2 emitted from the combustion of 1 kg of carbon. According to the proposal of Fig. 6, A ¼ 0 represents a non- anthropogenic CO2 cycle for BIGCC-nCCS and CST technologies. For 8 Gasification sector Air separation sector 49 21 23 24 22 2954 1928 27 26 20 1310 7Bagasse 6 5 4 (O2) 3 (N2) 2 1 (air) 9 11 12 14 18 50 25 36 Syngas cleaning/cooling Power generation sector 39 40 38 4155 42 ASU 56 47 53 58 59 57 45 37 35 Process51 46 30 31 40 41 Gasifier 34 37 38 39 43 32 52 33 Fig. 5. BIGCC-nCCS plant. Fig. 6. Proposed structure for fuels, technologies and CO2 emissions. Table 2 Results from gasification models. i Xi (%, dry basis) % Error relative to [8] Biomass-oxygen model Biomass-air model Hassuani et al. [8] CO 44.8 18.1 18.1 0.0 CO2 15.0 12.6 13.8 �8.2 H2 39.8 16.4 14.8 10.8 CH4 0.027 9.2 � 10�4 3.2 e N2 0.33 52.9 49.0 7.8 SO2 0.05 0.025 N/A e C2H4 1.1 � 10�10 6.4 � 10�13 0.7 e V.F. Fonseca Filho et al. / Energy 103 (2016) 715e724 721 NGCC, A>0 represents anthropogenic CO2 emission; for BIGCC- CCS, A<0 indicates sequestration of non-anthropogenic carbon dioxide. Thus, for BIGCC-CCS technology, CO2 gross emission _mCO2 ; CO2 emission index I�C (Eq. (23)); and CO2 emission cost CCO2 (Eq. (24)) are all negative. On the other hand, CET (Eq. (26)) is always positive regardless the technology considered. Plant exergetic efficiency is strongly related to the inefficiency penalty parameterPε (Eq. (22)). Low exergetic efficiencymeans the plant presents high exergy loss and destruction, resulting in a high inefficiency penalty parameter Pε. Thus, the higher Pε, the lower the exergetic efficiency. Unlike NGCC high efficiency technology, CST is heavily penalized due to its very low exergetic efficiency. However, this is not a fair comparison because the fuels in each technology are different. Now comparing Pε of the biomass-fired plants, it is clear that both advanced cycles based on gasification (BIGCC-CCS and BIGCC-nCCS) are far more efficient than the traditional one (CST). Despite emitting anthropogenic CO2 due to fossil fuel burning, NGCC technology presents the lowest CET because both exergetic efficiency and fuel heating value are high enough to compensate its high CO2 gross emission. On the other hand, CET is very high for CST technology because it is strongly penalized by its low exergetic efficiency and fuel heating value, despite its low CO2 gross emis- sion. BIGCC-nCCS technology presents lower CET than CST because CO2 gross emission is lower and exergetic efficiency is higher. When compared to NGCC technology, BIGCC-nCCS presents a much Table3a BIGCC-CCS model results: material streams. i Description _m (kg/s) P (MPa) T (K) h (kJ/kg) s (kJ/kg-K) b (kJ/kg) 1 Air 141.4 0.10 303.0 303.20 6.87 0.04328 2 Air 141.4 0.48 521.1 524.90 6.98 199.3 3 Nitrogen 108.5 0.48 288.0 297.80 6.34 141.2 4 Oxygen 32.95 0.48 288.0 �10.62 �0.44 123.9 5 Oxygen 32.95 2.50 565.2 252.70 �0.23 331.5 6 Bagasse 90.62 303.0 e e 20,635 7 Syngas 123.6 2.20 1123.0 e e 10,792 8 Acid gases 13.01 2.00 673.0 e e 787.3 9 Syngas 110.6 2.00 673.0 e e 10,235 10 Air 1397 0.10 298.0 298.2 6.86 0.0 11 Air 1397 1.95 910.0 945.6 7.18 571.5 12 Combustion gases 1508 1.85 1573 e e 1316 13 Combustion gases 1508 0.12 898.0 e e 387.9 14 Exhaust gases 1508 0.11 375.0 e e 68.83 15 Water 82.59 0.11 306.0 137.7 0.48 0.4501 16 Nitrogen 1122 0.11 388.0 403.0 7.09 19.18 17 Carbon dioxide 303.1 11.00 398.0 26.40 �0.75 526.9 18 Water 304.1 12.50 307.0 153.1 0.49 12.96 19 Steam 304.1 10.72 623.0 2896 5.88 1149 20 Steam 334.4 10.72 623.0 2896 5.88 1149 21 Steama 319.4 0.01 306.0 2439 8.00 60.75 22 Steam 15.00 5.00 593.0 2984 6.31 1108 23 Water 5868 0.30 298.0 104.4 0.36 0.2006 24 Water 5868 0.10 328.0 229.7 0.77 5.921 25 Waterb 319.4 0.01 306.0 137.7 0.48 0.3482 26 Water 30.31 0.10 305.5 135.7 0.47 0.3916 27 Water 30.31 12.50 306.5 150.9 0.48 12.90 28 Water 30.31 10.72 623.0 2896 5.88 1150 36 Water 15.00 0.01 305.5 135.6 0.47 0.2947 49 Water 334.4 0.01 305.5 135.6 0.47 0.3457 50 Water 304.1 0.01 305.5 135.6 0.47 0.2950 52 Syngas 123.6 2.11 673.0 e e 10,259 a Saturated, quality x ¼ 0.95. b Saturated, quality x ¼ 0. Table 3b BIGCC-CCS model results: non-material streams exergy e power and heat. i Description B (MW) 29 Gas turbine net shaft work 389.5 31 Gas turbine net power 335.9 32 Steam turbine net shaft work 144.5 34 Steam turbine net power 137.3 35 Process heata 9.560 37 Steam cycle pump power 5.320 38 Syngas cooling pump power 0.4618 39 Air compressor power 31.50 40 Oxygen compressor power 8.677 41 ASU cryogenic cycle power 25.39 48 Carbon dioxide compressor power 100.0 54 Gas turbine gross shaft power 1254 a T ¼ 383.9 K. Table 3c BIGCC-CCS model results: non-material streams e destroyed exergy. i Description B (MW) 30 Gas turbine generator 19.50 33 Steam turbine generator 7.226 42 ASU cryogenic cycle 34.17 43 Heat recovery steam generator 135.5 44 Carbon dioxide compressor 22.55 45 Gas turbine 157.5 46 Steam turbine 203.9 47 Gasifier 547.2 51 Process 7.054 53 Syngas cooling 157.1 55 Air compressor 3.327 56 Oxygen compressor 1.836 57 Gas turbine set compressor 66.25 58 Syngas treatment 125.9 59 Gas turbine set combustion chamber 54.95 V.F. Fonseca Filho et al. / Energy 103 (2016) 715e724722 lower CO2 gross emission, but it is penalized by its lower exergetic efficiency and fuel heating value. Application of CET model for all technologies results no penalty charge (DCe ¼ 0) for cyclic non- anthropogenic cycles, as expected. On the other hand, anthropo- genic CO2 emissions penalize NGCC plant (DCe >0), whereas a negative penalty (DCe <0) charge results for BIGCC-CCS plant due to non-anthropogenic CO2 sequestration. Power costs with CO2 tax internalized are shown in Table 6, considering different CC (capital cost) and CF (capacity factor) found in the literature [10]. LCC (Levelized capital cost) and TPC (total power cost)4 are estimated according to Eqs. (30) and (29), 4 TCP ¼ LCC � DCe. respectively. NGCC technology presents the lowest LCC because: i) its capital cost is very competitive; ii) its operating cost related to fuel consumption is low due to its high efficiency. However, NGCC penalty charge DCe (Eq. (27)) is high because its emitted CO2 is anthropogenic and not captured thereafter. It is quite interesting to note that the penalty charge DCe is negative for BIGCC-CCS tech- nology. Despite BIGCC-CCS presents the highest levelized capital cost (due to its very high capital cost), its TPC (total power cost) is lower than NGCC because the negative penalty charge represents a monetary bonus related to the non-anthropogenic carbon capture. Table 5 CET-based penalty model results. Parameter BIGCC-CCS BIGCC-nCCS CST NGCC A ðkgCO2 =kgCÞ �3.67 0 0 3.67 _mf ðkg=sÞ 90.62 90.62 98.03 24.88 CD ðUSD=sÞ 5.39 5.24 30.87 0.21 CR ðUSD=sÞ 0.13 0 0.67 0.00 Pε ðUSD=sÞ 5.03 4.79 28.80 0.20 _mCO2 ðkg=sÞ �76.49 0.02 0.02 68.48 BP ðMWÞ 345.5 345.5 350.0 351.2 I�C (�) �0.80 0.00 0.00 0.71 CCO2 ðUSD=sÞ �4.01 0.00 0.00 0.14 CET ðUSD=tonCO2 Þ 52.46 49.89 296.28 2.01 DCe ðUSD=MWhÞ �43.00 0.00 0.00 1.40 -30.00 -20.00 -10.00 0.00 10.00 20.00 30.00 40.00 50.00 60.00 0% 5% 10% 15% 20% 25% BIGCC-CCS BIGCC-nCCS CST NGCC Fig. 7. TPC variation with interest rate. Table 4 Cycle performance results. BIGCC-CCS BIGCC-nCCS CST NGCC Efficiency (1st law) 0.2915 0.2946 0.1919 0.4138 Fuel heat input (MW) 1244.0 1244.0 1774.0 1244.0 Power output (MW) Total net 362.3 366.5 340.4 514.8 Gas turbine set 370.0 370.0 e 389.5 Steam turbine 137.3 42.3 2 � 175.0 130.7 V.F. Fonseca Filho et al. / Energy 103 (2016) 715e724 723 Penalty charge for BIGCC-nCCs is near zero, which means its TPC is neither penalized nor rewarded. Thus, its TCP is not as competitive because its CC is high. TPC variation with interest rate is shown in Fig. 7, on which only the most favorable scenario shown in Table 6 is considered for each technology. Annual interest rate varies in the range 0e25% for 20 years life (values from Table 6 are calculated for i ¼ 14%). It is observed that BIGCC-nCCS and CST plants present similar TPC for lower interest rate, although CST is slightly more attractive than BIGCC-nCCS for any interest rate. NGCC plant present TPC comparatively favorable relatively to the previous plants due to its high efficiency, low capital cost and rich fuel. When carbon capture and sequestration is included in the BIGCC cycle, a CO2 emission tax charging equal to DCe would make BIGCC-CCS the most attractive technology, with the lowest TPC in all the range considered. More interesting, TPC is negative for i(13%. It means that BIGCC-CCS would be fully rewarded for capturing and sequestrating carbon, regardless the power selling price. It would be an innovative business model for the power market, on which BIGCC-CCS is not a power plant that seques- trates carbon which is fully rewarded by selling power. Instead, it is a non-anthropogenic carbon sequestrating plant on which po- wer is a by-product. The previous discussion evinces CET methodology as a proper toll to establish energy policies that promote the use of high efficiency technologies and renewable fuels, in accordance to Table 6 Power generated costs with CO2 tax internalized. Parameter BIGCC-CCS BIGCC-nCCS CC (USD/W) 2.13a 2.13d 2.56a 1.47a 1.49d CF (e) 0.67b 0.83d 0.67b 0.67b 0.83 LCC (USD/MWh) 54.70 44.30 65.90 37.90 30.90 4 TPC (USD/MWh) 11.70 1.30 22.90 37.90 31.00 4 a Source [23]. b Source [16]. c Source [9]. d Source [6]. Santarelli (1999) [19]. Such policies could, for instance, impose carbon taxing, provide subsides for efficient plants and stimulate investments in R&D to reduce the technology costs. 4. Conclusion The development of technologies based on energy renew- able sources must be technically adequate and satisfy the rigorous environmental legislation agreements to reduce pollutant emission. Biomass integrated gasification combined cycle emerge as a major technological innovation and was considered in this work in two different schemes: with and without carbon capture and sequestration. Two other tradi- tional technologies were also considered: natural gas combined cycle and biomass steam cycle traditionally employed in Bra- zilian ethanol industry. Carbon exergy tax method was applied in all four considered technologies. It revealed that BIGCC without carbon capture and sequestration is not economically attractive in terms of its total power cost. However, when assuming a tax related to CO2 emis- sions, BIGCC with carbon capture and sequestration becomes an interesting choice because it removes non-anthropogenic carbon from the atmosphere. The CO2 tax is assumed as an instrument of energy policy that promotes the use of advanced energy systems. Moreover, under considered economic scenario, BIGG with carbon capture and sequestration would be fully rewarded for capturing and sequestrating carbon, regardless the power selling price. It would be an innovative business model for the power market, on which BIGCC-CCS is a non-anthropogenic carbon sequestrating plant that generates power as a by-product instead a power plant that sequestrates carbon that is fully rewarded by selling power. CST NGCC 1.86a 1.07d 1.08b 0.60a 0.65a 0.92d 0.90c 0.67b 0.83d 0.67b 0.60b 0.60b 0.87d 0.60b 7.80 38.80 27.90 17.20 18.70 18.20 25.90 7.90 38.90 28.00 21.40 22.80 22.30 30.00 V.F. Fonseca Filho et al. / Energy 103 (2016) 715e724724 Acknowledgments The last author is indebted to the National Council for Scientific and Technological Development (CNPq) for his productivity grant (process 303350/2014-8) and to the S~ao Paulo Research Foundation (FAPESP, process 2013/07287-3). References [1] Borchiellini R, Massardo AF, Santarelli M. An analytical procedure for the carbon tax evaluation. Energy Convers Manag J 2000a;41:1509e31. [2] Borchiellini R, Massardo AF, Santarelli M. A carbon tax evaluation based on the efficient use of energy resources: a case study. Journalof Applied Thermody- namics 2000b;3:129e45. [3] Chiesa P, Lozza G. CO2 emission abatement in IGCC power plants by semi- closed cycles: Part B e with air-blown combustion and CO2 physical absorp- tion. J Eng Gas Turbines Power 1999;121:642e8. [4] Dantas GA, Legey LFL, Mazzone A. Energy from sugarcane bagasse in Brazil: an assessment of the productivity and cost of different technological routes. Renew Sustain Energy Rev 2013;21:356e64. [5] Duan L, Zhu J, Yue L, Yang Y. Study of a gas-steam combined cycle system with CO2 capture by integratingmolten carbonate fuel cell. Energy 2014;74:417e27. [6] EIA. US Energy Information Administration. Updated capital cost estimates for utility scale electricitygenerating plants. 2013. Available at: http://www.eia. gov/forecasts/capitalcost/pdf/updated_capcost.pdf [accessed 04.03.15]. [7] Fchart. Enginering equation solver e academic profissional V.9.705. 2014. www.fchart.com. Available at:. [8] Hassuani SJ, Leal MRLV, Macedo IC. Biomass power generation: sugar cane bagasse and trash. Piracicaba: PNUD-CTC (S�erie Caminhos Para Sustentabili- dade); 2005. ISBN:85-99371-01-0. [9] Ihiabe D. Assessing biomass-fired gas turbine power plants: a techno- economic and environmental perspective [PhD thesis]. School of Engineer- ing: Cranfield University; 2013. [10] IPCC. Intergovernmental Panel on Climate Change. Carbon dioxide capture and storage -Chapter 8: cost and economic potential. 2005. Available at: https://www.ipcc.ch/pdf/special-reports/srccs/srccs_chapter8.pdf [accessed 12.02.16]. [11] Klein D, Bauer N, Bodirsky B, Dietrich JP, Popp A. Bio-IGCC with CCS as a long- term mitigation option in a coupled energy-system and land-use model. En- ergy Procedia 2011;4:2933e40. [12] Lora ES, Andrade RV. Biomass as energy source in Brazil. Renew Sustain En- ergy Rev 2009;13:777e88. [13] Mas V, Kipreos R, Amadeo N, Laborde M. Thermodynamic analysis of ethanol/ water system with the stoichiometric method. Int J Hydrogen Energy 2006;31:21e8. [14] Mondal MK, Balsora HK, Varshney P. Progress and trends in CO2 capture/ separation technologies: a review. Energy 2012;46:431e41. [15] Oliveira Jr S. Exergy: production, cost and renewability. London: Springer- Verlag; 2013. [16] Pereira Jr AO, Costa RC, Costa CV, Marreco JM, La Rovere EL. Perspectives for the expansion of new renewable energy sources in Brazil. Renew Sustain Energy Rev 2013;23:49e59. [17] Prabu V, Geeta K. CO2 enhanced in-situ oxy-coal gasification based carbon- neutral conventional power generating systems. Energy 2015;84:672e83. [18] Rosillo-Calle F, Bajay SV, Rothman H. Uso da biomassa para produç~ao de energia na indústria brasileira. Editora da Unicamp. Campinas-SP: Uni- versidade de Campinas; 2005. [19] Santarelli M, Borchiellini R, Massardo AF. Environomic optimization for combined plants including CO2 influence. Indianapolis: ASME Paper 99-GT-5; 1999. [20] Santarelli M. Thermodynamic, economic and environmental considerations in the analysis and optimization of energy systems [PhD thesis]. Torino, Italy (in Italian): DipartimentodiEnergetica, Politecnicodi Torino; 1998. [21] Santarelli MGL. Carbon exergy tax: a thermos-economic method to increase the efficient use of exergy resources. Energy Policy 2004;32:413e27. [22] Schmidt J, Leduc S, Dotzauer E, Schmid E. Cost-effective policy instruments for greenhouse gas emission reduction and fossil fuel substitution through bio- energy production in Austria. Energy Policy 2011;39:3261e80. [23] Seabra JEA, Macedo IC. Comparative analysis for power generation and ethanol production from sugarcane residual biomass in Brazil. Energy Policy 2011;39:421e8. [24] Uddi Sk N, Barreto L. Biomass-fired cogeneration systems with CO2 capture and storage. Renew Energy 2007;32:1006e19. [25] Urech J, Tock L, Harkin T, Hoadley A, Mar�echal F. An assessment of different solvent-based capture technologies within an IGCC-CCS power plant. Energy 2014;64:268e76. [26] Valero A, Lozano MA, Serra L, Torres C. Application of the exergetic cost theory to the CGAM problem. Energy 1994;19:365e81. [27] Zhang X, Gundersen T, Roussanaly S, Brunsvold AL, Zhang S. Carbon chain analysis on a coal IGCC: CCS system with flexible multi-products. Fuel Process Technol 2013;108:146e53. [28] R€oscha C, Kaltschmitta M. Energy from biomassddo non-technical barriers prevent an increased use? Biomass Bioenergy 1999;16:347e56. [29] Rhodes JS, Keith DW. Engineering economic analysis of biomass IGCC with carbon capture and storage. Biomass Bioenergy 2005;29:440e50. http://refhub.elsevier.com/S0360-5442(16)30223-7/sref1 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref1 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref1 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref2 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref2 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref2 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref2 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref3 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref3 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref3 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref3 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref3 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref3 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref3 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref4 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref4 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref4 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref4 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref5 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref5 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref5 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref5 http://www.eia.gov/forecasts/capitalcost/pdf/updated_capcost.pdf http://www.eia.gov/forecasts/capitalcost/pdf/updated_capcost.pdf http://www.fchart.com http://refhub.elsevier.com/S0360-5442(16)30223-7/sref8 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref8 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref8 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref8 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref9 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref9 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref9 https://www.ipcc.ch/pdf/special-reports/srccs/srccs_chapter8.pdf http://refhub.elsevier.com/S0360-5442(16)30223-7/sref11 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref11 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref11 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref11 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref12 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref12 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref12 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref13 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref13 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref13 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref13 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref14 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref14 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref14 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref14 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref15 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref15 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref16 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref16 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref16 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref16 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref17 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref17 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref17 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref17 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref18 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref18 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref18 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref18 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref19 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref19 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref19 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref19 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref20 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref20 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref20 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref21 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref21 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref21 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref22 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref22 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref22 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref22 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref23 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref23 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref23 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref23 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref24 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref24 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref24 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref24 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref25 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref25 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref25 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref25 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref25 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref26 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref26 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref26 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref27 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref27 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref27 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref27 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref298 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref298 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref298 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref298 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref298 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref29 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref29 http://refhub.elsevier.com/S0360-5442(16)30223-7/sref29 Carbon exergy tax applied to biomass integrated gasification combined cycle in sugarcane industry 1. Introduction 2. Methodology 2.1. Biomass gasification model 2.2. BIGCC-CCS model 2.3. CET-based penalty analysis 3. Results and discussion 3.1. Biomass gasification 3.2. BIGCC-CCS model 3.3. CET-based penalty 4. Conclusion Acknowledgments References