Rate-based modelling of combined SO2 removal and NH3 recycling integrated with an aqueous NH3-based CO2 capture process

Applied Energy 148 (2015) 66–77

Contents lists available at ScienceDirect

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

Rate-based modelling of combined SO2 removal and NH3 recycling integrated with an aqueous NH3-based CO2 capture process Kangkang Li a,b, Hai Yu a,⇑, Guojie Qi c, Paul Feron a, Moses Tade b, Jingwen Yu c, Shujuan Wang c a

CSIRO Energy Flagship, 10 Murray Dwyer Circuit, Mayfield West, NSW 2304, Australia Department of Chemical Engineering, Curtin University of Technology Australia, GPO Box U1987, Perth, WA 6845, Australia c Department of Thermal Engineering, Tsinghua University, Beijing 100084, PR China b

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 A rigorous, rate-based model for an Vent gas

NH3–CO2–SO2–H2O system was developed.  Model predictions are in good agreement with pilot plant results.  >99.9% of SO2 was captured and >99.9% of slipped ammonia was reused.  The process is highly adaptable to the variations of SO2/NH3 level, temperatures.

Water makeup

Washing Chiller

NH3 exit Pump

Heater

Heat Exchanger

CO2 absorber

Pretreatment Flue gas

CO2 lean

NH3 rich

NH3 lean

SO2 Recovery

NH3 recycle

CO2 rich

Pump Coal-fired Power Staon

a r t i c l e

i n f o

Article history: Received 2 January 2015 Received in revised form 21 February 2015 Accepted 11 March 2015 Available online 28 March 2015 Keywords: CO2 capture SO2 removal NH3 recycling Aqueous NH3

⇑ Corresponding author. E-mail address: [email protected] (H. Yu). http://dx.doi.org/10.1016/j.apenergy.2015.03.060 0306-2619/Ó 2015 Elsevier Ltd. All rights reserved.

SO2 ferlizer SO2 Removal and NH3 Recycling Process

Typical CO2 Capture Process

a b s t r a c t To reduce the costs of controlling emissions from coal-fired power stations, we propose an advanced and effective process of combined SO2 removal and NH3 recycling, which can be integrated with the aqueous NH3-based CO2 capture process to simultaneously achieve SO2 and CO2 removal, NH3 recycling and flue gas cooling in one process. A rigorous, rate-based model for an NH3–CO2–SO2–H2O system was developed and used to simulate the proposed process. The model was thermodynamically and kinetically validated by experimental results from the open literature and pilot-plant trials, respectively. Under typical flue gas conditions, the proposed process has SO2 removal and NH3 reuse efficiencies of >99.9%. The process is strongly adaptable to different scenarios such as high SO2 levels in flue gas, high NH3 levels from the CO2 absorber and high flue gas temperatures, and has a low energy requirement. Because the process simplifies flue gas desulphurisation and resolves the problems of NH3 loss and SO2 removal, it could significantly reduce the cost of CO2 and SO2 capture by aqueous NH3. Ó 2015 Elsevier Ltd. All rights reserved.

K. Li et al. / Applied Energy 148 (2015) 66–77

1. Introduction The fossil fuel – coal is one of the world’s most important energy resources. According to International Energy Agency (IEA) [1], over 30% of the total energy supply and over 40% of the world electricity consumption come from coal. Driven by the economic growth of developing economies, coal will still continue to play an important role in powering the world’s economy in foreseeable future, particularly in those countries heavily dependent on coal for power generation [2]. For instance, in China more than 80% electricity is generated by coal combustion [3,4], while in Australia about 75% of power generation is coal based [5]. However, burning coal for power generation emits large amounts of carbon dioxide (CO2) leading to global climate change [2,6], as well as the gaseous pollutants sulphur dioxide (SO2) and oxides of nitrogen (NOx), which are important factors to environment pollution such as acid rain [7,8]. Traditional treatment to remove these pollutants is by a series of separate processes, including flue gas denitrification (deNOx) [9], flue gas desulphurisation (FGD) [10,11] and flue gas decarbonation (CO2 capture [12–16]). However, amine-based CO2 capture technologies require a flue gas desulphurisation system to reduce the SO2 concentration in the flue gas below 10 ppm before it enters the CO2 capture plant. This is because the SO2 is readily absorbed in aqueous amine solutions, leading to the formation of thermally stable, irreversible reaction products SO2 accumulates in amine solutions over time and reduces the solvent’s capacity to absorb CO2, which requires extra amine makeup and increases operating costs [17,18]. For a 500-MW power station, the capital cost of installing a wet flue gas desulphurisation system and a selective catalytic reduction system could be hundreds of millions each [19,20]. The cost for the CO2 capture system would be even higher, very likely into the billions. If each separate treatment process could be integrated into one process using one solvent, the costs to capture these pollutants could be significantly reduced. Aqueous ammonia (NH3) could be a suitable candidate for the capture of multi-acidic pollutants in flue gas and manufacture of value-added products such as ammonium sulphate and ammonium nitrate at the same time. The NH3-based wet flue gas desulphurisation technologies have been well documented [19] and drawn increasing attention in China due to its lower investment, higher removal efficiency, no secondary pollution, and value added byproducts [21]. Selective catalytic reduction of NO by NH3 gas is currently the most widely used method for removal of NO from flue gas in the coal fired power stations [9]. If NO in the flue gas can be oxidised to NO2 which is soluble in water, the NH3-based solvents can be used to remove NOx and SO2 simultaneously [22,23]. As an alternative to amine-based solvents for CO2 capture, aqueous NH3 has the advantages of low cost, low regeneration energy, and no absorbent degradation. Over the last few years, intensive research activities have been carried out to understand and improve the aqueous NH3-based CO2 capture technologies and assess their technical and economical feasibility [24–28]. The technical feasibility of NH3-based CO2 capture technology has been successfully proven in pilot-plant and demonstration trials by Alstom and CSIRO. Alstom’s chilled ammonia process (CAP) can achieve a CO2 capture efficiency of 90%, with a CO2 product purity of >99.5% and a potential heat requirement for CO2 regeneration of <2 MJ/kg CO2 [29–31]. In collaboration with Delta Electricity, CSIRO evaluated the NH3 process in pilot-plant trials using real flue gas at Munmorah Power Station in New South Wales, Australia. An 80–90% CO2 removal efficiency was obtained, with a purity of CO2 product >99% [32,33]. The rigorous, rate based models for CO2 capture by aqueous NH3 have been developed and validated using the pilot plant results in our previous work. [34–36].

67

The above review of the latest development on the aqueous NH3-based capture technologies suggests that aqueous NH3 has a potential of capturing multiple pollutants. However, there is little information available on the performance of the aqueous NH3-based multiple pollutants capture processes, in particular involving CO2 removal. Powerspan has developed an ‘ECO2’ process integrated with the ECO multi-pollutant control system. They claim that the process can use the vaporised NH3 to remove SO2 and produce ammonium sulphate fertiliser [37]. But no detailed reports on the performance of the process are available in the open literature. The SO2 concentration in the flue gas is varied dependent on the types of coal and boilers used, but is generally below 2000 ppm. This concentration only requires up to 4000 ppm NH3 to stoichiometrically produce ammonium sulphite/sulphate. However, the NH3 concentration in the flue gas leaving the absorber is vulnerable to the operating temperature and is generally >4000 ppm. This will make it difficult to maintain the material balance between SO2 and NH3 in the combined process. Dong et al. assessed the feasibility of a process that simultaneously removed the multi-acidic pollutants CO2, SO2 and NOx in the cement industry by NH3 scrubbing [38]. It should be pointed out that the process assessment was based on a thermodynamic model which did not consider the fact that the absorption processes are thermodynamically and kinetically controlled. In addition, the study did not report loss of NH3 in the process. It is well known that high NH3 loss, due to its intrinsic high volatility, still limits the economical feasibility of NH3-based capture technologies. An NH3 recovery unit, consisting of an NH3 scrubber and an NH3 stripper, is required to recycle the slipped NH3 in the system [39,40]. This process requires additional facilities and energy consumption, significantly increasing capital and operation costs. It is clear that more research is required to gain a detailed understanding of the combined removal technologies, address key issues related to NH3 loss and recovery and improve technical and economic feasibility of the technologies. In this study, we report an advanced process that can combine capture of SO2 and CO2 using aqueous NH3 and reduce the energy requirement for NH3 recovery. The development of this process is based on our recent studies. During pilot pilot-plant trials at Munmorah Power Station, we evaluated the feasibility of using NH3 for SO2 removal, and found that the SO2 removal efficiency by aqueous NH3 was >95% [32]. We also investigated an NH3 abatement and recycling process that uses the waste heat in flue gas to solve the problems of NH3 slip, NH3 makeup and flue gas cooling in the NH3-based CO2 capture process [34]. This study confirmed that the heat in flue gas can be used to regenerate NH3 for recycling, thus saving significant amounts of energy. The process investigated in the current study integrated NH3 recovery with CO2 and SO2 removal, which can potentially reduce the energy consumption of flue gas cooling and NH3 regeneration, and recover the SO2 to produce ammonium sulphite as a value-added product. To gain a detailed understanding of the process and evaluate its technical feasibility, we developed a rigorous, rate-based model of NH3–CO2–SO2–H2O system using Aspen Plus to understand simultaneous capture of CO2 and SO2 by aqueous NH3. The model was validated against the experimental results available in the literature and from pilot-plant trials conducted by CSIRO. Using a base-case scenario based on a typical pilot-plant experiment, we analysed the behaviours of SO2 removal and NH3 recycling in the proposed process configuration. The feasibility of the process under different scenarios, such as high SO2 concentration in the flue gas, high NH3 content from the CO2 absorber and high-temperature flue gas, was then discussed. To our knowledge, this is the first time that a combined capture of SO2 and CO2 using aqueous NH3 integrated with flue gas cooling and NH3 recycling in CO2 capture process has been presented and analysed in detail.

68

K. Li et al. / Applied Energy 148 (2015) 66–77

a heater heat up NH3-lean solvent to further enhance NH3 desorption and recycling.

2. Process description Fig. 1 shows the process flow-sheet diagram of the combined SO2 removal and NH3 recycling integrated with an aqueous NH3based CO2 capture unit (stripper not shown). The process consists of a pretreatment column, an NH3 wash column and a CO2 absorber. Initially, the stream (NH3 lean) was fed to the top of the NH3 absorber to scrub the slipped ammonia from CO2 absorber. After NH3 absorption, the vent gas will meet the set NH3 exhaust level (<25 ppmv) and discharged into environment; the lean solvent (NH3 lean) became rich solvent (NH3 rich) and was pumped to the pretreatment column, where it made direct contact with the high-temperature flue gas. At this stage, the pretreatment column acted as a cooler, an NH3 stripper and a SO2 scrubber: it cooled the high-temperature flue gas while also used the heat in the flue gas to release the captured NH3 and SO2 was absorbed by the NH3-rich solution. After pretreatment column, the rich NH3 stream (NH3 rich) returned to the lean NH3 stream (NH3 lean); the released NH3 was mixed with cooled flue gas and recycled to the CO2 absorber as ammonia make-up; the absorbed SO2 was formed as sulphite and remained in the solution. This completed one cycle. The second cycle started with the lean NH3 stream (NH3 Lean) undergoing the NH3 scrubbing in NH3 absorber and NH3 desorption and SO2 absorption in pretreatment column. The solution was continuously circulated between the pretreatment column and wash column, while absorbed SO2 accumulated in the circulated solution. The definition of number of cycles is the cycle index which counts the circulations of solution starting from wash column, going through the pretreatment column and returning to wash column. The use of number of cycles can represent the continuous operation of NH3 recycle and SO2 removal as a function of operating time. The operating time can refer to minutes, hours and days, depending on the real operating time of circulation solvent to complete one cycle in practical process. After a number of circulations, the entire system reached a semi-steady state in which the temperature in each stream remained constant. The solution looping ceased after the sulphur-containing electrolyte concentration reached the saturation point, and the saturated solution will be transferred to the ammonium sulphate production unit for further treatment (not included in this study). The water makeup was introduced into the wash column to maintain the water balance in the NH3 recycle and SO2 recovery system. A chiller was used to cool the NH3-lean solvent to achieve high NH3 capture efficiency, while

Vent gas

3. Pilot-plant trials of CO2 capture and SO2 removal CSIRO and Delta Electricity evaluated an NH3-based CO2 capture process under real flue gas conditions at Munmorah Power Station. The details about the CO2 capture process have been reported elsewhere [32,33]. To understand the characteristics of SO2 removal by NH3 and to collect additional results for model validation, an NH3dosing experiment was carried out in the pilot plant’s pretreatment column. The flue gas was introduced to the bottom of the column, initially without liquid circulation. After the introduction of flue gas, fresh water entered the column from the top and circulated between the column and the wash water tank. The NH3 was dosed into the wash water in the storage solvent tank at different flow rates. A Gasmet™ analyser (Fourier Transform Infrared Spectroscopy, CX-4000) equipped with a ZrO2 oxygen analyser allowed online identification and quantification of gas species, including CO2, SO2, NH3 and H2O, in the flue gas at the inlet and outlet of the pretreatment column. The pH of the solution at the outlet of the pretreatment was measured online using an industrial pH meter (Rosemount). The pretreatment column had an inner diameter of 0.5 m and was packed with 25-mm Pall rings to a height of 3 m. The detailed experimental activities and observations during the NH3-dosing experiment are listed in Table 1. 4. Model development The rigorous, rate-based process model was built within the RateFrac module in Aspen Plus. The process model consists of a thermodynamic model, a transport model and a rate-based model. 4.1. Thermodynamic and transport model The Pitzer property method embedded in Aspen Plus was employed to calculate the chemical and physical properties of the liquid phase, including the fugacity coefficient, entropy, enthalpy and Gibbs energy, while the Redlich–Kwong–Soave equation of state was used to calculate the vapour phase fugacity coefficient. NH3, CO2, SO2, N2, O2 were defined as Henry components and the Henry’s law constants of these species were retrieved from Electrolytes Expert System. The Pitzer parameters for the binary interactions in the NH3–CO2–H2O system were retrieved from the Aspen Plus databank, which have been regressed against the literature experimental data of vapour–liquid equilibrium, heat capacity

Water makeup

Table 1 Experimental activities and observations in the SO2 removal experiment. Flue gas flow-rate = 936 kg/h, CO2 flow-rate = 120 kg/h, SO2 concentration = ca. 200 ppmv, liquid flow-rate = 39 L/min, solvent inventory = ca. 300 L, gas inlet temperature = 35–38 °C, inlet wash water temperature = 25 °C.

Washing Chiller Pump Heater

Heat Exchanger Pretreatment Flue gas

CO2 lean

NH3 rich

NH3 lean

SO2 Recovery

CO2 absorber

CO2 rich

Stage

Time

Activities

pH

SO2 removal

CO2 removal

NH3 outlet

1

09:04

Flue gas on

No

No

No

2

09:53

Partial

No

No

3

11:40

Partial

No

No

4

12:30

Almost

No

No

5

13:33

Complete

Possible

Some

6

14:20

Water circulation Dosing NH3 at 0.2 kg/h Dosing NH3 at 0.5 kg/h Dosing NH3 at 1 kg/h Dosing NH3 at 1.8 kg/h

Not available Drop to 2.5 Increase to 2.7 Increase to 7.2 Increase to 8.2 Increase to 8.6

Complete

Possible

Sharp increase

Pump Coal-fired Power Staon

SO2 ferlizer SO2 Removal and NH3 Recycling Process

Typical CO2 Capture Process

Fig. 1. Combined SO2 removal and NH3 recycling process for CO2 capture by aqueous NH3.

69

K. Li et al. / Applied Energy 148 (2015) 66–77

and ion speciation [41,42]. For transport properties, the Jones–Dole electrolyte correction model, the Onsager–Samaras model and Riedel electrolyte correction model were used to calculate the liquid viscosity, liquid surface tension and thermal conductivity, respectively. The electrolyte solution characteristics and the liquid-vapour behaviours of the NH3–CO2–SO2–H2O system were modelled with an equilibrium chemistry package. Table 2 lists all the ionic reactions and their corresponding equilibrium constants included in the package. The chemical equilibrium constants of these reactions were expressed as:

ln K eq ¼ A þ B=T þ C lnðTÞ þ DT where Keq is the equilibrium constant of each reaction; T is the temperature, K; and constants A, B, C, D were adjustable parameters. The parameters for reactions 1–10 are available in the Aspen data2 bank and those for the reaction HSO 3 M S2O5 + H2O were obtained from Ermatchkov et al. [43]. 4.2. Rated-base model The rate-based model was developed using the RateFrac module of Aspen Plus. The module allows users to divide the column into stages and perform the mass and heat transfer, chemical reactions, hydraulic and interface behaviours at each stage. The twofilm theory was used to describe the mass and heat transfer resistance [44], and the Maxwell–Stefan theory was adopted to calculate multi-component mass and heat transfer between liquid and vapour phase [45]. The calculations of mass transfer and interfacial area were determined by the correlation method proposed by Onda et al. [46]. The interfacial area factor for the random packing material (25-mm Pall rings) was set as 1.2, based on our earlier experimental results [32]. The correlations proposed by ChiltonColburn [47] and Stichlmair et al. [48] were adopted for the calculation of heat transfer and liquid holdup, respectively. In the rate-based model, we assumed that the reactions involving SO2 were thermodynamically controlled due to the following factors: (1) fast SO2 absorption rate by the water [49] and (2) the generated acid H2SO3 is quickly dissociated and quickly neutralized by basic aqueous NH3. The reactions of CO2 with OH and NH3 were assumed to be kinetically controlled. The power law expressions were used to express the kinetically controlled reactions: n

E

r ¼ kT eRT

n Y

a

Ci i

i¼1

where r is the rate of reaction; k is the pre-exponential factor; n is the temperature exponent, which is chosen as zero for this simulation; E is the activation energy; R is the universal gas constant; T is the absolute temperature; Ci is the molarity concentration of

component i; and ai is the stoichiometric coefficient of component i in the reaction equation. The kinetic parameters k and E for kinetic reactions in Table 3 are derived from the work of Pinsent et al. [50] 5. Model validation 5.1. Thermodynamic model of NH3–SO2–CO2–H2O The validation of the thermodynamic model is one of the prerequisites and foundations of the development of a rigorous, rate-based model. We used the developed thermodynamic model of NH3–SO2–CO2–H2O to predict chemical and physical properties in the ternary systems of NH3–CO2–H2O (CO2 capture process) and NH3–SO2–H2O (SO2 capture process), and the quaternary system of NH3–SO2–CO2–H2O (combined CO2/SO2 removal and NH3 recycling). We then compared the model’s results with experimental data available in the literature. The present validation was proposed to evaluate whether the quaternary system can be independently applied for the prediction of a ternary system with the absence of SO2 or CO2. 5.1.1. Validation of NH3–CO2–H2O system The model for the NH3–CO2–SO2–H2O system in the absence of SO2 is identical to that of the NH3–CO2–H2O system. We validated the thermodynamic model for the NH3–CO2–H2O system under selective conditions in previous studies [34]. The present study extended the previous work and validated the model using additional experimental results, including CO2 partial pressure and liquid species distribution as a function of CO2 molarity. As shown in Fig. 2(a) and (b), the simulated CO2 partial pressure agrees very well with the experimental results. The results in Fig. 2(c) show that the model can satisfactorily predict the species concentration, including carbon- and nitrogen-containing species in the liquid phase. 5.1.2. Validation of NH3–SO2–H2O system Fig. 3 plots the simulated and experimental total pressures in the NH3–SO2–H2O system as a function of SO2 molality at two NH3 concentrations. The predicted total pressures agree reasonably well with the experimental measurements. Overestimations of the total pressure were observed at high SO2 concentrations (>6 mol/ kg H2O). This is likely to be caused by the limitation of Pitzer model, in which the electrolyte concentration in the solution should be no more than 6 mol/L ionic strength [41]. Another reason for the deviations might be the uncertainty of measurements at high temperatures and loadings. Considering that the proposed process was carried out in the electrolyte concentration range of 0–6 mol/L and at temperatures below 353 K, the model is suitable for the prediction of vapour–liquid equilibrium.

Table 2 Chemical reactions and equilibrium constants of the NH3–CO2–SO2–H2O system. No.

Reaction

1 2 3 4 5 6 7 8 9 10 11

2H2O M H3O+ + OH CO2 + 2H2O M H3O+ + HCO 3 2 + HCO 3 + H2O M CO3 + H3O +  NH3 + H2O M NH4 + OH  NH3 + HCO 3 M NH2COO + H2O 2H2O + SO2 M H3O+ + HSO 3 + 2 H2O + HSO 3 M H3O + SO3  2 HSO3 M S2O5 + H2O NH4HCO3(S) M NH+4 + HCO 3 (NH4)2SO3(S) M 2 NH+4 + SO2 3 + (NH4)2SO3(S)‘H2O M 2NH4 + SO2 3 +H2O

Equilibrium parameter A

B

C

D

132.899 231.465439 216.049 1.2566 4.583437 5.978673 25.290564 10.226 554.8181 920.3782 1297.041

13445.9 12092.1 12431.7 3335.7 2900 637.395996 1333.40002 2123.6 22442.53 44503.83 33465.89

22.4773 36.7816 35.4819 1.4971 0 0 0 0 89.00642 139.3449 224.2223

0 0 0 0.0370566 0 0.0151337 0 0 0.06473205 0.03619046 0.3515832

K. Li et al. / Applied Energy 148 (2015) 66–77

Table 3 The kinetic reactions and corresponding kinetic parameters in the NH3–CO2–SO2–H2O system. No.

Reaction

Parameters

1 2 3 4

CO2 + OH ? HCO 3  HCO 3 ? CO2 + OH NH3 + CO2 + H2O ? NH2COO + H3O+ NH2COO + H3O+ ? NH3 + CO2 + H2O

K

E (cal/mol)

4.32e+13 2.38e+17 1.35e+11 4.75e+20

13,249 29,451 11,585 16,529

In terms of liquid-phase validation, the solution pH reflects the 2 species distribution, e.g. HSO3, SO2 3 and S2O5 . As shown in Fig. 3 (c), the modelled pH values agree perfectly with the experimental results at electrolyte concentrations ranging from 0.001 to 6 mol/L. This implies that the proposed thermodynamic model can predict the ionic species concentration in the NH3–SO2–H2O system. 5.1.3. Validation of NH3–CO2–SO2–H2O Fig. 4 compares the predicted and experimental CO2 partial pressure in the SO2 loading range of 0–0.3 (mol SO2/mol NH3) and temperature range of 20–60 °C. These conditions are close to the actual conditions of the combined SO2 recovery and NH3 recycling process. The model’s prediction agrees reasonably well with the experimental results, although a small deviation was observed at CO2 molality >0.5 mol/kg H2O. Considering that SO2 removal and NH3 recycling will occur at low CO2 concentrations (<0.5 mol/kg H2O), the effect of the deviation can be ignored. In summary, the developed thermodynamic model can satisfactorily predict the vapour–liquid equilibrium and ion speciation for the ternary NH3–CO2–H2O and NH3–SO2–H2O systems, as well as the quaternary NH3–CO2–SO2–H2O system.

(a)

CO2 paral pressure/kPa

1250

Line: Model Point: Goppert et al

5.2.1. CO2 absorption by aqueous NH3 CSIRO has conducted pilot-scale trials of an NH3-based CO2 capture process and validated the rate-based model for CO2 absorption by aqueous NH3 using the pilot-plant results [32–34]. Table 4 summarises the experimental conditions and results, and compares the experimental results and model predictions. In general, the average relative error for the overall CO2 absorption rate in 30 trials was only ±6.0%, and for NH3 loss rate in 24 trials it was ±11.1%. This shows that the established rate-based model can satisfactorily predict the behaviour of the CO2 capture process by aqueous NH3 in the packed column. 5.2.2. SO2 removal by aqueous NH3 The results from the dosing experiments were used to validate the model for the absorption of SO2 by aqueous NH3. During process modelling, the pretreatment column was divided into 25 stages. The calculation methods of interfacial area, mass and heat transfer and liquid holdup were the same as those in the CO2 absorption model [34], which has been proven to accurately predict the behaviours of CO2 absorption by aqueous NH3 along the

NH3=2.186

500

(b)

8000

NH3=0.966

750

Compared to the equilibrium-based model, the rate-based model considers liquid- and vapour-phase mass and heat transfer, chemical reactions and kinetics, and hydraulic and interface properties of packed materials. It thus provides a more accurate description and characterisation of the SO2/CO2 absorption and NH3 recycling process. In this work, the rate-based model was validated against the Munmorah pilot-plant results. The validation work included CO2 absorption and SO2 removal by aqueous NH3 in packed columns.

T=333 K

NH3=0.721

1000

5.2. Rate-based model

NH3=3.837

250

CO2 paral pressure/kPa

70

Line: Model Point: Goppert et al T=353 K

7000 NH3=0.591

6000

NH3=2.006

5000 4000 3000

NH3=4.14

NH3=1.087

2000 1000

0

0

0

1

2

3

0.0

4

0.5

Mod, NH3

(c)

1.5

2.0

2.5

3.0

3.5

4.0

Expt, NH3

+

8

Species, mol/kg H2O

1.0

CO2 molality, mol/kg H2O

CO2 molality, mol/kg H2O

6

+

Mod, NH4

Expt, NH4

Mod, NH2COO-

Expt, NH2COO-

Mod, HCO3-

Expt, HCO3-

Mod, CO32-

Expt, CO32-

4 2 0 0

1

2

3

4

CO2, molality, mol/kg H2O Fig. 2. Experimental (point) and predicted (line) CO2 partial pressure and liquid species distribution as a function of CO2 molality at (a) various NH3 concentrations and T = 333 K; (b) various NH3 concentrations and T = 353 K; (c) NH3 concentration of 6.3 mol NH3/kg H2O and T = 313 K. Experimental CO2 partial pressure results are from Göppert et al. [51] and species results from Lichtfers [52].

71

K. Li et al. / Applied Energy 148 (2015) 66–77

4000 Mod.,313 K Mod.,333 K Mod.,353 K Mod.,373 K

3000 2500

(a)

Exp.,313 K Exp.,333 K Exp.,353 K Exp.,373 K

2000 1500 1000

Mod.,313 K Mod.,333 K Mod.,353 K Mod.,373 K

2000

Total pressure/kPa

Total pressure/kPa

3500

1500

(b)

Exp.,313 K Exp.,333 K Exp.,353 K Exp.,373 K

1000 500

500 0

0 0

2

4

6

8

0

SO2 molality, SO2 mol/kg H2O

2

4

13

8

10

(c)

Exp.

Mod.

12

6

SO2 molality, SO2 mol/kg H2O

N/S=49.7

11 N/S=5.66

10

pH

9 N/S=2.29

8 7

N/S=1.46

6 N/S=1.03

5 4

N/S=1.00

1E-3

0.01

0.1

1

10

Concentraon of NH3+SO2, mol/L Fig. 3. Total pressure of NH3–SO2–H2O for different temperatures at (a) mNH3 = 3.19 mol/kg H2O, (b) mNH3 = 6.08 mol/kg H2O with model data and experimental data from Rumpf et al. [53]; (c) pH of solutions as a function of (NH4)2SO3 concentrations at different molar NH3/SO2 (N/S) ratios and at 20 °C, with experimental data obtained from Scott and McCarthy [54].

(a)

100

10

1 T=293K 0 molSO2/molNH3 0.1 molSO2/molNH3 0.2 molSO2/molNH3 0.3 molSO2/molNH3

0.1

0.01 0.1

0.2

0.3

0.4

0.5

0.6

0.7

CO2 paral pressure/kPa

CO2 paral pressure/kPa

100

(b)

10

T=313K 1

0 molSO2/molNH3 0.1 molSO2/molNH3 0.2 molSO2/molNH3 0.3 molSO2/molNH3

0.1 0.1

0.8

0.2

CO2 molality, mol/kg H2O

CO2 paral pressure/kPa

100

0.3

0.4

0.5

0.6

0.7

0.8

CO2 molality, mol/kg H2O

(c)

10 T=333K 0 molSO2/molNH3 0.1 molSO2/molNH3 0.2 molSO2/molNH3 0.3 molSO2/molNH3

1

0.1 0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

CO2 molality, mol/kg H2O Fig. 4. Predicted and measured CO2 partial pressure of NH3–CO2–SO2–H2O system as a function of CO2 molalities at various SO2 loadings (molar ratio of SO2 to NH3) and the NH3 concentration of 5 wt%. (a) 293 K, (b) 313 K and (c) 333 K. Experimental data were obtained from Qi [55].

72

K. Li et al. / Applied Energy 148 (2015) 66–77

Table 4 Summary of pilot plant experimental conditions, and results and comparison of the experimental and modelling results. Pilot conditions

a b

which is generally expressed as

Results

Solvent flow-rate, L/ min Liquid inlet temperature, °C NH3, wt%

50–134

2–5

Experimental CO2 capture rate, kg/h Simulated CO2 capture rate, kg/h Relative error, %

15–28

on SO2 transportation from the gas phase to the liquid phase. 2 Second, the oxidation rate of SO2 3 to SO4 is kinetically controlled, SO2 4

60.4– 99.6 57.6– 94.8 0.5– 19.5 6.0

Lean CO2 loading, mol/mol Flue gas mass flowrate, kg/h Gas temperature, °C

0.2–0.4

Average relative error,a %

600– 1100 15–20

Flue gas inlet CO2, vol%

8.5–12

Experimental NH3 emission rate, kg/h Simulated NH3 emission rate, kg/h Relative error,%

Flue gas inlet H2O, vol%

1.2–3.5

Average relative errorb,%

1.9– 13.6 2.0– 12.6 0.3– 22.6 11.1

Based on 30 trials. Based on 24 trials.

packed column. The model settings for the pretreatment column were consistent with the conditions of the pilot-plant trials, including column size, packing materials and operating conditions. O2

During SO2 absorption, the oxidation process of SðIVÞ ! SðVIÞ inevitably occurs in the presence of oxygen. In our simulation, we assume that SO2 in the gas phase and SO2 3 in the aqueous solution are slightly oxidised by the oxygen during the capture process. This assumption was based on the following three considerations [49,56,57]. First, the oxidation of S(IV) ? S(VI) primarily takes place in the liquid phase after SO2 is dissolved into the solution, generating SO2 3 species. Thus, oxidation will have little influence

d½SO2 4  dt

¼ K½SO2 3 . This means the

is proportional to the concentration of rate of production of 2 SO2 3 while in the pilot SO2 removal process the SO3 concentration is very low (<0.011 mol/L). So the (NH4)2SO3 is the key compound removing SO2 from flue gas, which was also advocated by Gao et al. [21]. Third, the SO2 removal process was carried out in an oxygendeficit environment, due to the relatively low oxygen concentration of <8.0% in the flue gas, which does not favour oxidation. 2 Therefore, the SO2 3 will be partly oxidised to SO4 , but this minor sulphur oxidation will have little influence on the simulation of SO2 removal by aqueous NH3 in the pilot plant. This is verified by the excellent agreement between the model predictions and experimental results in terms of the solution pH, NH3, SO2 and CO2 concentration at the pretreatment column outlet (Fig. 5). However, we acknowledge that the sulphate will eventually dominate the sulphur species, as a result of the irreversible oxidation reaction after sulphite is exposed to oxygen. A detailed study of sulphur oxidation will be conducted in subsequent work. In terms of methodology of model validation against the operating time, a differential approach was proposed in which we divided one hour into 10  6 min and ran the simulation (SO2 absorption and NH3 recycle) 10 times at a time interval of 6 min. For each run, all experimental conditions were kept the same except the concentrations of species in the liquid at inlet which were varied due to SO2 absorption build-up and NH3 dosing. Considering the fact that the liquid flow in the whole system is far from a plug flow and the liquid leaving the column will mix with part of liquid in the system, the build-up SO2 products and dosed NH3 will be diluted by the rest of bulk solvent in the storage tank. Therefore, in the simulation a dilution factor was used and its value is affected by the solvent circulation rate, solvent inventory and extent of

12 Simu.

pH

8 6

1.8kg/hr NH3 0.2kg/hr NH3

1.0kg/hr NH3

4 2 0 08:24

250

Expt.

Water

SO2 concentraon/ppmv

(a) 10

10:48

12:00

13:12

Simu.

Water

200 0.5kg/hr NH3

150 100 1.0kg/hr NH3 1.8kg/hr NH3

50

14:24

15:36

08:24 09:36 10:48 12:00 13:12 14:24 15:36

Time

Time 500 Expt.

Simu.

120

(d)

400 1.8kg/hr NH3

300 1.0kg/hr NH3

200 100

0.5kg/hr NH3 0.2kg/hr NH3 Water

CO2 flowrate, Kg/hr

NH3 concentraon/ppmv

(c)

Expt.

0.2kg/hr NH3

0

0.5kg/hr NH3

09:36

(b)

100

Water

0.2kg/hr NH3 0.5kg/hr NH3

80

1.0kg/hr NH3 1.8kg/hr NH3

60 Simu.

0

Expt.

40 08:24 09:36 10:48 12:00 13:12 14:24 15:36

Time

08:24

09:36

10:48

12:00

13:12

14:24

15:36

Time

Fig. 5. Comparison of pilot-plant data with simulation results: (a) outlet solution pH from column; (b) gas SO2 concentration outlet from column; (c) gas NH3 concentration outlet from column; (d) gas CO2 flow-rate outlet from column.

73

K. Li et al. / Applied Energy 148 (2015) 66–77 Table 5 Properties of flue gas from power station and CO2 absorber. Source

Flow-rate, kg/h

Temperature, °C

Power station CO2 absorber outlet

760 656

120 16.9

solvent mixing in the system. In this study, a dilution factor of 10 was determined using trial and error approach to achieve good agreement between the experimental and simulation results (Fig. 5). Solution pH: Absorption of the acidic gas SO2 into the aqueous solution will decrease the solution pH, due to the low dissociation constant of H2SO3 (pKa = 1.81). As shown in Fig. 5(a), the solution pH dropped quickly as SO2 was absorbed by the circulating water, due to the increasing solution acidity. The pH value then increased step by step with the increasing NH3 dose rate. The simulation curve matched the experimental results very well, indicating that the model can predict pH variation during SO2 removal by aqueous NH3. As the solution pH is a very important indication of the species distribution, the validation of solution pH reflects that the model enables prediction of the solution species in the NH3– CO2–SO2–H2O system. Gas outlet SO2 concentration: The SO2 level in the outlet flue gas directly reflects the SO2 removal efficiency by aqueous NH3. As shown in Fig. 5(b), upon water circulation, the outlet SO2 concentration dropped rapidly, but quickly increased to the level close to the inlet SO2 concentration. This implies that fresh water has a relatively low SO2 removal capacity. When the NH3 dosing rate rose to >0.5 kg/h, the SO2 concentration dropped to a very low level. The simulation results agree reasonably well with the experimental data at these NH3 dosing stages. Gas outlet NH3 concentration: The NH3 introduced into the water was used to neutralise the acidic SO2 gas. If the dosed NH3 is excessive for SO2 absorption, some NH3 will slip to the flue gas, due to the high volatility of NH3 and the high pH of the wash water. As shown in Fig. 5(c), NH3 started to slip at a dosing rate of 1.0 kg/h NH3, above which NH3 evaporation increased dramatically. The trend of outlet NH3 concentration is consistent with the model results. Gas outlet CO2 flow-rate: As shown in Fig. 5(d), there is no appreciable CO2 removal in the experiment. This suggests that under the conditions studied, SO2 is absorbed in the solution in preference to CO2. In other words, SO2 can be selectively removed by wash water that contains a small amount of NH3. The simulation results were in good agreement with the experimental results.

25 NH3 recycled to CO2 absorber 20 80 15 60 10 40

Vent gas NH3 5

20 0

NH3 for SO2 capture

NH3 concentraon, ppmv

NH3 reuse efficiency, %

100

0 0

50

100

150

200

250

300

350

Number of cycles Fig. 6. NH3 reuse efficiency and vent gas NH3 emission concentration as a function of number of cycles.

Composition, vol/% CO2

H2O

O2

N2

NH3

SO2

10.7 3.23

6.0 1.83

7.8 9.0

75.5 84.2

– 1.20

200 ppmv –

The good agreement between the experimental results and the modelling suggests that the established model can predict SO2/CO2 absorption by aqueous NH3. This confirms that our assumptions were reasonable under the conditions studied. 6. Model application for combined SO2 removal and NH3 recycling After model validation, the rate-based model was reliable when used in practice to guide the process simulation of combined SO2 removal and NH3 recycling. The base case was first proposed based on conditions determined in our previous study [34]. These conditions were: 350 kg/h wash water circulation rate, 10 °C wash water inlet temperature, wash column size £0.5 m  h3.0 m with 16mm Pall rings, pretreatment column size £0.5 m  h3.0 m with 25-mm Pall rings, 5 °C temperature approach of the heat exchanger between hot inlet stream and cold outlet stream (hot side). The typical flue gas conditions from the Munmorah power station and the CO2 absorber outlet (test ID-32A) are shown in Table 5. The effect of SO2 content in flue gas, NH3 content from CO2 absorber and flue gas temperature were investigated to evaluate the technical adaptability and feasibility of the process. 6.1. Base-case scenario As described in Section 2, the combined SO2 removal and NH3 recycling process involved the continuous circulation of wash water between the pretreatment column and wash column. During the continuous circulation, SO2 was absorbed and accumulated in the wash water, while NH3 was absorbed in the wash column and desorbed in the pretreatment column. The amount of SO2 and NH3 present in the wash water and gas phase depended on the operation conditions and number of cycles (time of operation). In this study, a semi-bath simulation was conducted with 350 kg/h solvent circulation rate and 350 L solvent inventory. This indicated that the solvent leaving the pretreatment column will be sent back to the inlet of the column with the same composition, and that the operating time of solvent to complete one cycle was one hour, i.e. one cycle represented one hour. Using the developed model, we simulated the base-case scenario to investigate NH3 recycling, SO2 removal and column profiles as a function of the number of cycles, and determined the effectiveness of SO2 removal and NH3 recycling. 6.1.1. NH3 profiles In the combined SO2 removal and NH3 recycling system, the NH3 slipped from the CO2 absorber was either reused (recycled back to the CO2 absorber or retained in the wash water for SO2 removal) or emitted to the environment. Fig. 6 shows the NH3 reuse efficiency and vent gas NH3 emission concentration as a function of the number of cycles. NH3 reuse efficiency is defined as the ratio of NH3 used for recycling and SO2 capture to the total NH3 slipped from the CO2 absorber. At steady state, 96.62% of NH3 was recycled to the CO2 absorber, while 3.36% of NH3 was used for SO2 capture. The total NH3 reuse efficiency reached as high as 99.98%. The rest of the NH3 was discharged into the environment.

K. Li et al. / Applied Energy 148 (2015) 66–77

SO2 capture efficiency/%

60 80 SO2 capture efficiency Outlet SO2 concentraon

60

50 40 30

40

20 20

10

0 0

50

100

150

200

250

300

0 350

(b)

40 (NH4)2SO3

Mass fracon/%

(a) 70

100

Outlet SO2 concentraon/ppbv

74

30 20 10

(NH4)2S2O5 NH4HSO3

0 0

50

100

150

Number of cycles

200

250

300

350

Number of cycles

Fig. 7. (a) SO2 capture efficiency and SO2 emission level from pretreatment column and (b) concentration profiles of sulphur-containing species in SO2-rich solution as a function of number of cycles.

CO2 concentraon, vol% 2

4

Wash column height, m

3.0

6

8

SO2 concentraon, ppmv 10

12

(a)

2.5

Gas NH3 profile

2.0

Gas CO2 profile

1.5 1.0 0.5 0.0 0

2000

4000

6000

0

Pretreatment column height, m

0

25

75

3.0

100 125 150 175 200

(b)

2.5 2.0 Gas NH3 profile 1.5

Gas SO2 profile

1.0 0.5 0.0 0

8000 10000 12000

2000

4000

6000

8000 10000 12000

NH3 concentraon, ppmv

NH3 concentraon, ppmv Pretreatment coulmn height, m

50

(c)

3.0 2.5

Liquid temperature

2.0

Gas temperature Inlet gas temperature

1.5 1.0 0.5 0.0 40

50

60

70

80

90

100

110

120

Temperature, oC Fig. 8. (a) NH3 and CO2 profiles along the wash column height; (b) NH3 and SO2 profiles along the pretreatment column height; and (c) temperature profiles along the pretreatment column during 175 cycles (20 wt% (NH4)2SO3).

As NH3 vapour is a hazardous pollutant, its emission level must meet the NH3 emission standard. According to the United States National Institute of Occupational Safety and Health, NH3 concentrations in workplace air should not exceed 25 ppmv [58]. In the simulation, the emitted NH3 concentration was always below 15 ppmv at studied circulation times. This indicates that the clean vent gas can be directly discharged into the atmosphere without exceeding emission levels. 6.1.2. SO2 profiles As shown in Fig. 7(a), the proposed process achieved a high SO2 capture efficiency, which was constantly >99.98%. Trace levels of SO2 emissions varied from 10 to 30 ppbv. This high-efficiency removal is primarily attributed to the fast reaction between SO2

and H2O, and because the generated HSO3 was quickly neutralised by basic aqueous NH3. Fig. 7(b) shows the concentration profiles of sulphur-containing species in the wash water at the outlet of the pretreatment column as a function of number of cycles. SO2 accumulated in the solution in the forms of (NH4)2SO3, NH4HSO3 and (NH4)2S2O5 as cycle number increased. (NH4)2SO3 was persistently the dominant species in the wash water. Its concentration increased gradually to 35% (saturated concentration of (NH4)2SO3 at 10 °C), while the NH4HSO3 and (NH4)2S2O5 remained at relatively low levels. This is because the SO2 absorption process by aqueous NH3 was conducted at a pH >7, and the alkaline environment facilitated the generation of SO2 3 species. The concentrated SO2 solution is expected to undergo a further treatment, e.g. producing ammonium sulphite/sulphate fertilisers.

75

K. Li et al. / Applied Energy 148 (2015) 66–77 Table 6 Adaptability of the combined SO2 removal and NH3 recycling process to different SO2 concentrations in flue gas, NH3 levels from CO2 absorber, and flue gas temperature. Base case (200 ppmv SO2, 12,000 ppmv NH3, 120 °C flue gas)

Water circulation rate, kg/h Heater temperature, °C Chiller temperature, °C SO2 removal efficiency, % NH3 reuse for recycle, % NH3 reuse for SO2 capture, % Total NH3 reuse efficiency, % NH3 emission, ppmv SO2 emission, ppbv Electricity of heater, kJ/kg CO2a Electricity of chiller, kJ/kg CO2a a

350 – 10 >99.9 96.4 3.5 >99.9 <20 <30 0 68.7

SO2 concentration in flue gas, ppmv

NH3 level from CO2 absorber, ppmv

Flue gas temp, °C

0

2000

23,000

34,500

180

350 – 10 0 99.8 0 >99.8 <16 0 0 60.9

350 – 10 >99.9 71.3 28.6 >99.9 <5 <2.5 ppm 0 69.6

700 80 5 >99.9 97.9 2.0 >99.9 <80 <25 293.7 169.8

1200 90 5 >99.9 98.5 1.3 >99.6 <150 <5 658.0 220.8

350 – 10 >99.9 96.4 3.5 >99.9 <25 <100 0 68.8

The electricity usage of the heater and chiller is calculated on the basis of CO2 capture rate of 87.2 kg/h, coefficient of performance of 5.0 and 350 cycles.

6.1.3. Column profiles To gain insight into how NH3 was scrubbed and recycled and how SO2 was removed, we investigated the NH3 and CO2 profiles along the wash column and NH3 and SO2 profiles along the pretreatment column. As shown in Fig. 8(a), the absorption of NH3 primarily took place in the bottom half of wash column, where the NH3 absorption has a high driving force because of the high NH3 concentration in the flue gas. The vent gas after the wash column contained a low concentration of NH3, and can be discharged into the atmosphere without further treatment. As gaseous NH3 dissolved in the washing solution, a small amount of CO2 was absorbed in the solution in the wash column. This was due to the presence of free NH3 in the wash water and the high pH of the solution. The NH3-loaded solvent in the pretreatment column was heated by the hot flue gas to release the scrubbed NH3. As indicated in Fig. 8(b), the NH3 was vaporised rigorously, particularly in the bottom stages. During NH3 vapourisation, the SO2 in the flue gas was quickly removed by the NH3-loaded solution, and the SO2 absorption mainly occurred in the first 1 m from the bottom of the column. Accompanying the NH3 evaporation and SO2 absorption, the CO2 absorbed in the wash column was desorbed in the pretreatment column. Overall, virtually no CO2 absorption occurred in the combined capture process, which has been discussed in previous work [34]. Fig. 8(c) shows that the flue gas temperature decreased significantly (from 120 to 42 °C) along the pretreatment column after contact with the circulating solvent. This is of particular importance for saving some of the energy used to cool the flue gas. In summary, the pretreatment column in the proposed configuration plays three significant roles: (1) a chiller, to cool down the hot flue gas; (2) a heater, to strip and recycle almost all the slipped NH3; and (3) an efficient desulphurisation facility to remove SO2. 6.2. Process adaptability 6.2.1. Adaptability to SO2 concentration in flue gas SO2 concentrations are heavily dependent on the type and quality of the coal combusted in the power station. If poor quality coal is fired, the SO2 concentration will exceed 2000 ppmv in the flue gas. As shown in Table 6, both the NH3 reuse efficiency and SO2 removal efficiency were maintained at >99.9%, with the flue gas SO2 level ranging from 0 to 2000 ppm. This suggests that the advanced process adapts well to different SO2 concentrations. The high SO2 level was beneficial for reducing NH3 emissions, because the acid gas SO2 dissolved into the solution, decreasing the pH value and facilitating the absorption of alkaline NH3 gas.

However, the rising SO2 level increased effluent SO2 concentrations to the CO2 absorber, to close to 2.5 ppmv. The SO2 removal efficiency was still as high as 99.85%. This SO2 concentration was considered acceptable for entry into the CO2 absorber, because the SO2 content tolerance in the conventional MEA-based CO2 capture process is <10 ppmv. Moreover, the increasing SO2 concentration had no extra burden on the energy requirements of chilling duty, and the value was small compared to the reboiler duty for CO2 regeneration: in the range of 2000–3000 kJ/kg CO2 captured [59]. 6.2.2. Adaptability to NH3 concentration from CO2 absorber As shown in Table 6, the proposed process performed well at high NH3 concentrations up to 34,500 ppmv, achieving >99.6% NH3 reuse efficiency and 99.9% SO2 removal efficiency. This high adaptability allows the use of NH3 solvent with a higher NH3 concentration and a lower CO2 loading for CO2 absorption, and thus potentially reduces the size of the CO2 absorber. However, the strong adaptability to higher NH3 concentrations came at the expense of greater energy consumption. Specifically, a higher water circulation rate and a lower chilling temperature were required to ensure a higher NH3 capture efficiency, and a heater was required to raise the temperature of the NH3-rich solution for efficient NH3 recycling. This increased the energy penalty for chilling duty and heat duty, respectively. Compared with the large energy penalty of CO2 capture, the energy duties for heating and cooling were considered acceptable for the recovery of such high levels of NH3. Moreover, if low temperature steam (80–100 °C) is available in the power station, the energy consumption of the heater can be significantly reduced. 6.2.3. Adaptability to flue gas temperature The temperature of flue gas from coal-fired power stations is typically dependent on the type of coal combusted: ca.120 °C for black coal and ca.180 °C for brown coal in Australian power plants. The high-temperature flue gas provides more latent heat for NH3 desorption and recycling in the pretreatment column, which enables a greater saving of heat duty – especially for the case of high NH3 content from the CO2 absorber. Compared to the base case in Table 6, there is little difference between 120 °C and 180 °C flue gas in terms of energy consumption, SO2 removal and NH3 recycling efficiency, which implies that the proposed process is widely adaptable to a range of flue gas temperatures. Therefore, the proposed combined process of SO2 removal and NH3 recycling has proved strongly adaptable to extreme conditions, such as high SO2 content in flue gas, high NH3 content from CO2 absorber and high flue gas temperature. The principle behind the strong adaptabilities is adjusting the chilling temperature for

76

K. Li et al. / Applied Energy 148 (2015) 66–77

high NH3 scrubbing efficiency, and the heating temperature for high NH3 recycling efficiency. This can be used to guide other extreme scenarios, such as different gas compositions or flue gas temperatures. 7. Conclusions Burning coal for power generation causes emissions of acidic pollutants, such as NOx, SO2 and CO2. To reduce the costs of controlling these emissions, we developed a novel, effective NH3based capture process that simultaneously removes SO2 and CO2, and recycles slipped NH3. The NH3–CO2–SO2–H2O model for SO2 removal and NH3 recycling was developed using the commercial software Aspen Plus and validated by experimental results from pilot plant trials. We then employed the validated rate-based model to analyse the combined SO2 removal and NH3 recycling process. The results of a base-case scenario suggested that >99.9% of SO2 was captured and >99.9% of slipped NH3 was reused for NH3 recycling and SO2 capture, with a low energy requirement. SO2 continuously accumulated in the circulating solution until the (NH4)2SO3 reached saturation; this chemical could be used to produce sulphur-containing fertilisers. The proposed process can deal with a wide range of SO2 concentrations in the flue gas, NH3 content from the CO2 absorber and flue gas temperature. This suggests that the technology is very adaptable to the variable scenarios in real industrial applications. Further work on sulphur oxidation during the SO2 capture process will determine the behaviour of sulphur oxidation in solution in the presence of oxygen, and the effect of sulphur oxidation on SO2 equilibrium concentration in the gas phase. Acknowledgements The authors wish to acknowledge financial assistance provided through both CSIRO Energy Flagship and Australian National Low Emissions Coal Research and Development (ANLEC R&D). ANLEC R&D is supported by Australian Coal Association Low Emissions Technology Limited and the Australian Government through the Clean Energy Initiative. The views expressed herein are not necessarily the views of the Commonwealth, and the Commonwealth does not accept responsibility for any information or advice contained herein. Kangkang Li would like to thank the Australian IPRS-APA scholarship and CSIRO Top-up scholarship to support his research. References [1] International Energy Agency (IEA). 21st Century coal – Advanced Technology and Global Energy Solution; 2013. [2] International Energy Agency (IEA), World Energy Outlook 2014. OECD/IEA, London. [3] Zhou WJ, Zhu B, Fuss S, Szolgayová J, Obersteiner M, Fei WY. Uncertainty modeling of CCS investiment strategy in China’s power sector. Appl Energy 2010;87:2392–400. [4] China National Bureau of Statistics, China National Energy Administration. China energy statistics year book 2008. China Statistics Press; 2008. [5] CSIRO. Assessing post-combustion capture for coal-fired power stations in Asia-Pacific Partnership Countries. Final report to the Department of Resources, Energy and Tourism, April 2012. [6] Zhu L, Fan Y. A real options-based CCS investment evaluation model: case study of China’s power generation sector. Appl Energy 2011;88:4320–33. [7] Wang JS, Anthon EJ. Clean combustion of solid fuels. Appl Energy 2008;85:73–9. [8] Pikoñ K. Environmental impact of combustion. Appl Energy 2003;75:213–20. [9] Roy S, Hegde MS, Madras G. Catalysis for NOx abatement. Appl Energy 2009;86:2283–97. [10] Kaminski J. Technologies and costs of SO2-emissions reduction for the energy sector. Appl Energy 2003;75:165–72. [11] Nurrohim A, Sakugawa H. A fuel-based inventory of NOx and SO2 emissions from manufacturing industries in Hiroshima prefecture, Japan. Appl Energy 2008;78:355–69.

[12] Manzolini G, Fernandez ES, Rezvani S, Macchi E, Goetheer ELV, Vlugt TJH. Economic assessment of novel amine based CO2 capture technologies integrated in power plants based on European Benchmarking Task Force methodology. Appl Energy 2015;138:546–58. [13] Zhang M, Guo Y. Rate based modeling of absorption and regeneration for CO2 capture by aqueous ammonia solution. Appl Energy 2013;111:142–52. [14] Kunze C, Spliethoff H. Assessment of oxy-fuel, pre- and post-combustion based carbon capture for future IGCC plants. Appl Energy 2012;94:109–16. [15] Li H, Yan J, Yan J, Anheden M. Impurity impacts on the purification process in oxy-fuel combustion based CO2 capture and storage system. Appl Energy 2009;86:202–13. [16] Hu YK, Yan J, Li H. Effects of flue gas recycle on oxy-coal power generation systems. Appl Energy 2012;97:255–63. [17] Boot-Handford ME, Abanades JC, Anthony EJ, et al. Carbon capture and storage update. Energy Environ Sci 2014;7:130–89. [18] Markewitz P, Kuckshinrichs W, Leitner W, et al. Worldwide innovations in the development of carbon capture technologies and the utilization of CO2. Energy Environ Sci 2012;5:7281–305. [19] Kohl A, Nielsen R. Gas purification. 5th ed. Houston, Texas: Gulf Publishing Company; 1997. [20] Islas J, Grande G. Abatement costs of SO2-control options in the Mexican electric-power sector. Appl Energy 2008;85:80–94. [21] Gao X, Ding H, Du Z, Wu Z, Fang M, Luo Z, et al. Gas–liquid absorption reaction between (NH4)2SO3 solution and SO2 for ammonia-based wet flue gas desulfurization. Appl Energy 2010;87:2647–51. [22] Yu H, Zhu Q, Tan Z. Absorption of nitric oxide from simulated flue gas using different absorbents at room temperature and atmospheric pressure. Appl Energy 2012;93:53–8. [23] Guo R, Hao J, Pan W, Yu Y. Liquid phase oxidation and absorption of NO from flue gas: a review. Separ Sci Technol 2014. [24] Han K, Ahn CK, Lee MS, Rhee CH, Kim JY, Chun HD. Current status and challenges of the ammonia-based CO2 capture technologies toward commercialization. Int J Greenh Gas Con 2013;14:270–81. [25] Dave N, Do T, Puxty G, Rowland R, Feron P, Attalla M. CO2 capture by aqueous amines and aqueous ammonia – a comparison. Energy Procedia 2009;1:949–54. [26] Li KK, Yu H, Tade M, Feron P. Theoretical and experimental study of NH3 suppression by addition of Me(II) ions (Ni, Cu and Zn) in an ammonia-based CO2 capture process. Int J Greenh Gas Con 2014;24:54–63. [27] Strube R, Pellegrini G, Manfrida G. The environmental impact of postcombustion CO2 capture with MEA, with aqueous ammonia, and with an aqueous ammonia–ethanol mixture for a coal-fired power plant. Energy 2011;36:3763–70. [28] Ma S, Chen G, Guo M, Li Z, Han T, Zhu S. Path analysis on CO2 resource utilization based on carbon capture using ammonia method in coal-fired power plants. Renew Sust Energ Rev 2014;37:687–97. [29] Darde V, Thomsen K, van Well WJM, Stenby EH. Chilled ammonia process for CO2 capture. Int J Greenh Gas Con 2010;4:131–6. [30] Telikapalli V, Kozak F, Leandri JF, Sherrick B, Black J, Muraskin D, et al. CCS with the Alstom chilled ammonia process development program–Field pilot results. Energy Procedia 2011;4:273–81. [31] Kozak F, Petig A, Morris E, Rhudy R, Thimsen D. Chilled ammonia process for CO2 capture. Energy Procedia 2009;1:1419–26. [32] Yu H, Morgan S, Allport A, Cottrell A, Do T, Mcgregor J, et al. Results from trialling aqueous NH3 based post-combustion capture in a pilot plant at Munmorah power station: absorption. Chem Eng Res Des 2011;89:1204–15. [33] Yu H, Qi GJ, Wang SJ, Morgan S, Allport A, Cottrell A, et al. Results from trialling aqueous ammonia-based post-combustion capture in a pilot at Munmorah Power Station: gas purity and solid precipitation in the stripper. Int J Greenh Gas Con 2012;10:15–25. [34] Li KK, Yu H, Tade M, Feron P, Yu JW, Wang SJ. Process modeling of an advanced NH3 abatement and recycling technology in the ammonia-based CO2 capture process. Environ Sci Technol 2014;48:7179–86. [35] Qi GJ, Wang SJ, Yu H, Wardhaughb L, Feron P, Chen CH. Development of a ratebased model for CO2 absorption using aqueous NH3 in a packed column. Int J Greenh Gas Con 2013;17:450–61. [36] Yu J, Wang S, Yu H. Modeling analysis of solid precipitation in ammonia-based CO2 capture process. Int J Greenh Gas Con 2014;30:133–9. [37] McLarnon CR, Duncan JL. Testing of ammonia based CO2 capture with multipollutant control technology. Energy Procedia 2009;1:1027–34. [38] Dong R, Lu H, Yu Y, Zhang Z. A feasible process for simultaneous removal of CO2, SO2 and NOx in the cement industry by NH3 scrubbing. Appl Energy 2012;97:185–91. [39] Darde V, Mogensen BM, van Well WJM, Stenby EH, Thomsen K. Process simulations of CO2 capture with aqueous ammonia using the Extended UNIQUAC model. Int J Greenh Gas Con 2012;10:74–87. [40] Linnenberg S, Darde V, Oexmann J, Kather A, van Well JM, Thomsen K. Evaluating the impact of an ammonia-based post-combustion CO2 capture process on a steam power plant with different cooling water temperatures. Int J Greenh Gas Con 2012;10:1–14. [41] Aspen Technology. Aspen physical property system: physical property models. Burlington, MA, USA; 2010. [42] Aspen Technology. Rate-based model of the CO2 capture process by NH3 using Aspen Plus. Burlington, MA, USA; 2010. [43] Ermatchkov V, Kamps APS, Maurer G. The chemical reaction equilibrium constant and standard molar enthalpy change for the reaction

K. Li et al. / Applied Energy 148 (2015) 66–77

[44] [45]

[46] [47] [48]

[49] [50] [51]

2 {2HSO a spectroscopic and calorimetric 3 (aq) S2O5 (aq)+H2O(l)}: investigation. J Chem Thermodyn 2005;37:187–99. Whitman WG. Preliminary experimental confirmation of the two-film theory of gas absorption. Chem Metall Eng 1923;29:146–8. Alopaeus V, Aittamaa J, Nordén HV. Approximate high flux corrections for multicomponent mass transfer models and some explicit methods. Chem Eng Sci 1999;54(19):4267–71. Onda K, Takeuchi H, Okumuto Y. Mass transfer coefficients between gas and liquid phases in packed columns. J Chem Eng Jpn 1968;1:56–62. Chilton TH, Colburn AP. Mass transfer (absorption) coefficients prediction from data on heat transfer and fluid friction. Ind Eng Chem 1934;26:1183–7. Stichlmair J, Bravo JL, Fair JR. General model for prediction of pressure drop and capacity of countercurrent gas/liquid packed columns. Gas Sep Purif 1989;3:19–28. Miller JM, DePena R. Contribution of scavenged sulfur dioxide to the sulfate content of rain water. J Geophys Res 1972;30:5905–16. Pinsent BR, Pearson L, Roughton FJW. The kinetics of combination of carbon dioxide with hydroxide ions. Trans Farad Soc 1956;52:1512–20. GÖppert U, Maurer G. Vapor-liquid equilibria in aqueous solutions of ammonia and carbon dioxide at temperatures between 333 and 393 K and pressures up to 7 MPa. Fluid Phase Equilib 1988;41:153–85.

77

[52] Lichtfers U. Spektroskopische untersuchungen zur ermittlung von Speziesverteilungen im system ammoniak–kohlendioxid–wasser. Ph.D. thesis, Kaiserslautern University, Germany; 2000. [53] Rumpf B, Weyrich F, Maurer G. Simultaneous solubility of ammonia and sulfur dioxide in water at temperatures from 313.15 K to 373.15 K and pressures up to 2.2 MPa. Fluid Phase Equilibr 1993;83:253–60. [54] Scott WD, McCarthy L. The system sulfur–ammonia–water at 25 °C. Ind Eng Chem Res Fundament 1967;6(1):40–8. [55] Qi GJ, Wang SJ, Lu WY, Yu JW, Chen CH. Vapor–liquid equilibrium of CO2 in NH3–CO2–SO2–H2O system. Fluid Phase Equilibr 2015;25:47–55. [56] Hegg DA, Hobbs PV. Oxidation of sulfur dioxide in aqueous systems with particular reference to the atmosphere. Atmos Environ 1978;12:241–53. [57] McKay HAC. The atmospheric oxidation of sulphur dioxide in water droplets in presence of ammonia. Atmos Environ 1971;5:7–14. [58] Busca G, Pistarino C. Abatement of ammonia and amines from waste gases: a summary. J Loss Prevent Proc 2003;16:157–63. [59] Jilvero H, Normann F, Andersson K, Johnsson F. Heat requirement for regeneration of aqueous ammonia in post-combustion carbon dioxide capture. Int J Greenh Gas Con 2012;11:181–7.