Xun Tao, Fan Zhou, Xinlei Yu, Songling Guo, Yunfei Gao*, Lu Ding, Guangsuo Yu, Zhenghua Dai,Fuchen Wang*

Shanghai Engineering Research Center of Coal Gasification, East China University of Science and Technology, Shanghai 200237, China

Keywords:

ABSTRACT

1. Introduction

H2S is widely present in crude natural gas from the gas or oil wells with diverse concentrations,and is also an important byproduct in various processing industries like oil refinery,coal gasification or any other desulfurization of petroleum stocks[1–3].It is not an acceptable option that H2S is directly discharged without any treatment since this will cause the serious environmental pollution and health. Currently, Claus process is widely used to recover sulfur from H2S feed under Claus condition[4–7],and mainly contains the thermal and catalytic section. The thermal reaction was conducted in the Claus furnace,that typically was with a furnace temperature of about 1273 K in the industrial filed. In the thermal section,one-third of H2S was oxidized to SO2(Eq.(1))under Claus conditions, and SO2formed further reacted with unconverted H2S to form S2species (Eq. (2)). Typically, acid gas combustion in the reaction furnace was conducted in a coaxial jet double channel burner, and mainly impacted by the operational conditions, variation in the composition of acid gas feed and rector design configuration such as residence time, checker wall and burner structure[8]. These factors further contributed to the difference between the flame temperatures in various industrial conditions.According to the previous literature, the flame temperature in the industrial field was commonly in the range of 1250–1450 K [9–11]. In the catalytic section, the S8species was further yielded through (Eq.(3)) over the catalytic bed in the presence of alumina or titanium catalysts at 473–573 K.

The thermal section is considered to be the most important part in Claus process since it completes about 60% of sulfur recovery[12,13]. Generally, the acid gas mainly contains H2S and CO2with different concentrations,and CO2in the lean acid gas is even more than 50% [14]. However, CO2tended to form the unwanted byproducts such as carbonyl sulfide(COS)and CS2via a series of side reactions [15,16], which will reduce the performance of catalytic section. The sulfur emission involving COS and CS2account for 20%–50% of the total sulfur emission in Claus process [17]. Therefore, the decrease of CO2in the acid gas feed is significant to improve the sulfur recovery and reduce sulfur emission.

To date, tremendous efforts have been invested in the carbon capture, utilization and storage (CCUS) to utilize fossil fuel resources in an environmentally friendly manner for reducing CO2emission. Oxy-fuel combustion has become one of the most promising technologies for dealing with CCUS, which requires the pure oxygen rather than air[18–21].However,a little attention has been paid to the application of oxy-fuel combustion to the Claus process for reducing CO2emission.

Several researchers have experimentally studied the combustion of H2S/CO2mixtures in air/O2atmosphere to understand the chemical reactivity of CO2during the H2S combustion. Selim et al. [22] studied the effect of CO2on the H2S combustion in pure oxygen atmosphere, and found that the presence of CO2slightly decreased the peak temperature in the combustion region,whereas increased downstream reactor.This result was consistent with the result that studied by Ibrahim et al. [23], and found the presence of CO2promoted the formation of COS.The similar results were also found in previous studies related to the H2S-CO2-air diffusion flame, and also found that CO2addition improved the conversion of H2S and the formation of SO2[24]. Moreover, Li et al.[25] experimentally studied the acid gas combustion in air atmosphere, and concluded that a significant amount of CO was detected near the burner,which was tightly related to the CO2concentration in the feed, and the presence of CO2and CO triggered the formation of COS.

Several researchers have investigated the conversion between CO2and COS through the lab-scale experiments and simulation analysis to grasp the connection between CO2and COS.Karan et al. [26] studied the formation of COS in (CO2+ H2S) and(CO + H2S) mixtures, and found that the formation of COS was mainly through the reaction of CO with sulfur species. This result was consistent with the studies by Ibrahim et al. [12], and found the reaction of CO with sulfur species triggered the formation of COS. Clark et al. [27] also found that CO2could react with sulfur species to form COS. Li et al. [28] studied the partial oxidation of H2S and CO2mixtures using a flow plug reactor through an experimental and kinetic modeling,and found the formation of COS was through the reaction of CO and CO2with sulfur species,which also can be informed from the reaction pathway. This was consistent with the kinetic simulation results of acid gas (H2S and CO2)decomposition studied by Ibrahim et al. [16], and also found the reaction of CO and sulfur was accounted as the major pathways for COS formation.

Actually, in the Claus furnace COS was mainly formed in the flame region through a series of side reactions. Currently, COS released from the acid gas combustion was mainly studied in air or oxygen enriched air atmosphere[23,25,29,30].The H2S combustion in pure oxygen atmosphere has been reported in previous studies,whereas we noted that the information related to COS conversion was relatively limited [22,31,32]. To fill this gap, the present work is to study the effect of CO2on oxy-fuel combustion of H2S with particular focus on the formation and distribution of COS under the different CO2amount. Moreover, to better understand the contribution of relevant reactions to the target products,kinetic model is utilized to analyze the production rate of gas products and the reaction pathway of acid gas combustion, which was not reported in the previous studies.

The experiments were conducted in a coaxial jet double channel burner under stoichiometric condition, where CO2addition in the H2S feed varied up to 40% with an increment of 20%. The flame temperature and the volume fractions of H2S, SO2, H2, CO and COS were detected along central axial (R = 0.0) and axial line at about 3.0 mm(R=0.75)in radial,respectively.The diffusion or premixed opposed-flow flame model in Chemkin-Pro software was applied to analyze the production rate of H2S, SO2, H2and COS in the H2S feed without or with 40% CO2. A simplified reaction pathway of oxy-fuel combustion of acid gas was derived from the detailed mechanism.

2. Experimental

2.1. Experimental setup

Fig. 1 presents the schematic diagram of the experimental setup, and mainly includes the gas supply section, gas sampling,gas composition analysis, temperature measurement, combustion reactor and tail gas treatment. The combustion reactor is a quartz tube with inner diameter of 94.0 mm, length of 350.0 mm and thickness of 3.0 mm. A double concentric tubular burner was located at the bottom center of the reactor. According to the work of Selim et al.[33],the bluff body is designed and located at the top of the burner center tube to stabilize the flame since it can induce the radial diffusion of the oxygen into fuel stream. The flow rates were controlled by mass flow controller (CS200, Beijing Sevenstar Electronics Co., Ltd, China). H2S (99.5%, purity) and CO2(99.5%)are premixed in the gas pipeline and introduced into the reactor via the center channel of burner. The O2(99.9%, purity) is fed into the reactor via the annular channel of burner.

This setup is also equipped with some other apparatus to collect the gas product and flame temperature such as programmingcontrolled traversing device, temperature controller, transfer pump,gas sampling probe,K-type thermocouple,two quartz tubes and three gas collection bottles with a volume of 60.0 ml. The gas sampling probe is a quartz tube with inner diameter of 4.0 mm,length of 500.0 mm and thickness of 0.5 mm. Two quartz tubes have an inner diameter of 10.0 mm, a length of 200.0 mm and a thickness of 2.0 mm.

The gas product is analyzed by a gas chromatography(GC-2060,Ruiming Instrument Corp. China), which was equipped with thermal conductivity detector (TCD), flame ionization detector (FID)and flame photometric detector (FPD). H2and H2S are analyzed by TCD, CO by FID (Parapak Q column, 6 m × 3 mm), COS and SO2by FPD (GDX-303 column, 4 m × 3 mm).

2.2. Experimental procedure

In experiment start-up stage, acid gas (H2S and CO2) cannot be directly combusted because H2S with high concentration is highly toxic and difficult to handle.To solve this problem,CH4was introduced from the center channel and flamed to create a‘‘pre-flame”.After forming a stable flame, CO2and H2S were slowly introduced into the reactor, until a ‘‘blue-colored” flame was observed, indicating that H2S has been flamed. Then, CH4was stopped, and CO2and H2S were tuned to the desired flow rates according to experimental conditions.

In the experimental operation stage,a programming-controlled traversing device was applied to control the K-type thermocouple and the gas sampling probe for collecting flame temperature and stable end-products at any desired position along the axial and radial of the reactor, respectively. The flame temperature data was recorded by temperature controller in real time. Two quartz tubes were used to quench the gas sample, and placed in series in an ice-water mixing cooling unit during the gas collection process.Three gas collection bottles are also equipped with fine quartz wool and desiccant for removing sulfur and water from the sample gas.The gas sample was collected into PVF(polyvinyl fluoride)collecting bags(TDL-31,Dalian Hade Technologies,China).Moreover,it is worth noting that the transfer pump has a constant pumping rate of 100.0 ml?min-1and it takes approximately 4.0 min for gas-purging before reliable results are fully collected.

Fig. 1. Figure diagram of experimental setup.

The experiment was stopped after the sampling and temperature measurement. As a safe shutdown procedure, CH4was first injected to replace H2S while maintaining the flame. We note that a maintained flame could effectively convert H2S and avoid H2S over-leaking into the hood. Then, H2S was stopped and eventually purged with N2to eliminate residual H2S in the pipeline. Finally,CH4was stopped, and flame was extinguished.

The experimental exhaust gas was vented into a bucket containing saturated NaOH solution for treatment and was discharged to the ventilation system of the laboratory. The experiments were performed in a fume hood.

2.3. Parameter definition

Dimensionless axial (A) and radial (R) distances are used to show the distribution of temperature and gas products [33],defined by A = axial distance/Djetand R = radial distance/Djet. Djetdenotes the jet diameter, axial distance denotes the longitudinal axis distance above the center channel of burner, and corresponding A = 0 denotes the central channel exit, and the inner diameter of central channel is 4.0 mm.

2.4. Experimental conditions

The experiments are to study the effect of CO2on the oxy-fuel combustion of H2S under stoichiometric condition (equivalence ratio, ER = 1). The flow rates of H2S and O2maintain constant for all the experiments. Moreover, the flow rate of CO2is varied to change its concentration in the H2S feed. According to the acid gas feed components in the industrial field [11], the peak amount of CO2injected into the H2S feed is 40%with 20%increment,where H2S feed without CO2was regarded as the baseline case. Table 1 lists the flow rates of H2S, CO2and O2for each experiment.

Table 1 Gas flow rates for each experimental conditions

2.5. Uncertainties analysis

According to the information provided by the the manufacturer,the accuracy of K type thermocouple is ±0.4% within the temperature range of 648–1273 K. The accuracy of mass flow controller is±1% when the flow rate is within the measurement range of 30%,and ±0.3% when the flow rate exceeds 30%. For the gas chromatograph, the accuracy is ±2%.

The distribution of temperature was measured using K type thermocouple. The flame temperature is corrected for radiation losses according to the Refs. [34,35].

where Nu is Nusselt number,T(K)is actual measured temperature,TTC(K) is thermocouple temperature, Tsurr(K) is surroundings temperature that measured at about 100.0 mm above burner,dTC(mm)is thermocouple bead diameter, the value is 1.5, ε is emissivity of thermocouple bead,the value is 0.37 according to Perry’s Chemical Engineers’Handbook(8th edition)[36],λ is thermal conductivity of hot gases (W?m-1?K-1), σ is Stefan-Boltzmann constant(5.67 × 10-8W?m-2?K-4).

According to Whitaker’s [37] modified equation for the airflow around the periphery of the sphere, the Nusselt number (Nu)between the combustion gas and the coupled beads can be determined as:

where Re and Pr denote Reynolds number and Prandtl number respectively;μ and μsis the gas viscosity at wall temperature and galvanic bead temperature,respectively.Re and Pr can be expressed as:

Taking the radiation heat transfer error at the peak flame temperature as an example. In this experiment, the peak temperature is 1094.0 K in the H2S feed without CO2, and corresponding the thermal properties(λ,ρ,Cpand μ)at 0.1 MPa are 0.052 W?m-1?K-1,0.38 kg?m-3, 1388 J?kg-1?K-1and 4.479 × 10-5Pa?s, respectively.Because of the total gas flow, airflow expansion and burner size,the gas velocity at the peak temperature is about 1.71 m?s-1,Re = 21.76 and Nu = 4.065. The gas temperature (T) calculated is(1299.2±4.0) K via Eq. (8). Therefore, the radiation heat transfer error is approximately (15.79 ± 0.23)% using the following equation.

Utilizing the same calculation method, the gas temperature (T)at R = 0.0 are about (1198.9±6.1) K and (1155.7±7.2) K for the H2S feed with 20% and 40% CO2, respectively, and corresponding the radiation heat transfer error values are about (12.42±0.11)% and(11.65±0.14)%, respectively. For the gas temperature (T) at R = 0.75, the values are about （1281.8±6.2) K, （1190.9±6.4) K and (1143.8±6.2) K for the H2S feed with 0%, 20% and 40% CO2,respectively, and corresponding the radiation heat transfer error values are about (15.51±0.06)%, (12.08±0.25)% and (11.43±0.09)%,respectively.

3. Numerical Approach

3.1. Model description

Fig.2. Geometry of the axisymmetric counter flow diffusion flame.

The oxy-fuel combustion of acid gas is numerically simulated using the diffusion or premixed opposed-flow flame model in Chemkin-Pro software. This model has been successfully verified in previous studies related to the premixed flames and the laminar flow diffusion flames[38,39].Fig.2 presents the geometry of counter flow diffusion flame mainly involving two concentric circular nozzles pointing towards each other.Fuel and oxidizer are injected from two opposing nozzles to form an axis-symmetric impinging flow field, and a stagnation plane forms between the two nozzles.The impinging flow field can be represented by the axial coordinate(x)and the radial coordinate(r)since the flame is a axis-symmetric structure. For the counter flow diffusion flame model, the radial velocity can be assumed as varying linearly along the radial direction. The combustion flame can be regarded as a one-dimensional flame flow through the simplification of the two-dimensional flame flow [40]. The temperature, gas concentration and axial velocity along the axial distance are independent of the radial distance. Therefore, the fluid properties can be considered as a function of the axial distance, and the governing equations can be simplified to the differential equations as follows [41–43].

The boundary conditions for the fuel (F) and oxidizer (O)streams at the nozzles are shown as follows:

3.2. Reaction mechanism

A detailed kinetic mechanism has been developed involving 90 species and 596 elementary reactions in the previous studies[28].The mechanism mainly contains the H2S decomposition, H2S and H2oxidation,C/H/O system and CS2/COS oxidation.It has been successfully verified through the decomposition and partial oxidation of H2S with or without CO2under the different experimental conditions using the Chemkin-Pro. The present work adopts this mechanism to analyze the oxy-fuel combustion of H2S with particular focus on the rate of production (ROP) analysis and reaction pathway.For some active species in this mechanism,over the decades, several researchers have been preformed the study on the capture and measurement of active species in H2S flame. Azatyan et al. [44] studied the low pressure flames of a number of compounds containing sulfur by the electron spin resonance (ESR)method, and found the high concentrations of atomic H, O, and SO and OH radicals were discovered in the H2S flame. Selim et al.[45]studied the spectroscopic examination of the emission spectra of excited species in H2S/O2flame, and detected the some species such as H*, OH*, SH*, SO* and SO3*, and also provided the corresponding wavelength. Zhou et al. [46] found that in the H2S flame the SH self-reaction involved HSS and HSSH radicals.Table 2 given some major elementary reactions for the conversion of H2S/CO2.

4. Results and Discussion

4.1. Flame temperature distribution

Fig. 3 presents the distribution of flame temperature in three cases along the reactor, where Fig. 3(a) shows the temperature at R = 0.0 and Fig. 3(b) is at R = 0.75. The temperature increased to a peak value, followed by decreased with the axial distance increases since the heat lost from the reactor walls [25,30]. The CO2addition reduced the peak temperature, and corresponding the values at R = 0.0 are 1094.0 K 1050.0 K and 1021.0 K, respectively. This could be due to CO2diluted the concentration of H2S and weakened the intensity of H2S combustion. Some heat energy was used to heat CO2that also reduced the temperature during the combustion process. The temperature maintained in a relatively stable region compared with H2S feed without CO2since the specific heat capacity of CO2was improved within the temperature range of 300–1200 K,and CO2shown a buffering effect on the temperature [25]. For H2S feed with 40% CO2, the temperature downstream reactor (A > 20) was slightly higher than the other two cases, which was consistent with previous studies [22,30]. This was attributed to that OH radical, released from the reaction of CO2with H radical, improved the oxidizing medium amount in reaction system, and further reacted with combustible gas to release heat [25,31,32]. The distribution of flame temperature in Fig.3(b)was similar to that in Fig.3(a),whereas the peak temperature was relatively lower, and the peak values are 1083.0 K,1047.0 K and 1013.0 K, respectively.

4.2. Combustion products analysis

Fig.4 depicts the distribution of H2S along the reactor.It should be noted that H2S was not detected at R = 0.75 because the complete combustion occurred. Some literature have been confirmed that H2S combustion experienced the thermal and chemical decomposition, H2S and H2oxidation [33,54].The volume fraction of H2S decreased sharply at the first sampling position, then reduced slowly to zero. Guldal et al. [55] studied the catalytic decomposition of H2S, and found about 4% of H2S converted at 1073 K without catalysts through the thermal decomposition.Therefore, in the present study the thermal decomposition hardly proceeded with an observable rate because of the relatively low flame temperature. According to the distribution of H2and SO2near the burner exit,it can be initially inferred that the H2S mainly occurred the chemical decomposition under the role of H and S radicals (Eqs. (16) and (17)), where H2as the main product[34,48,56]. As the H2S combustion proceeded, the volume fraction SO2rapidly increased to a peak value, which indicated that H2S occurred the oxidation reaction.The previous studies reported that H2S oxidation mainly occurred under the role of O and OH radicals(Eqs.(18)–(20))[50–52].According to the distribution of H2S,it can be inferred that the chemical decomposition was the main pathway of H2S conversion.

Table 2Some major elementary reactions for H2S/CO2/O2 mechanism (1 cal?mol-1=4.18 J?mol-1)

Fig. 3. Flame temperature versus CO2 addition along the axial reactor: (a) R = 0.0, (b) R = 0.75.

Fig. 4. Volume fraction of H2S versus the axial distance (R = 0.0).

Fig. 5 depicts the distribution of H2along the reactor. The H2volume fraction at R = 0.0 was shown in Fig. 5(a), which shown the similar trend that increased to a peak and then decreased sharply to a relative stable range with the axial distance increases.This indicated that the H2S oxidation was accompanied by the formation and consumption of H2.It should be noted that H2amount also probably derived from H2O.Ibrahim et al.[57]studied the effects of H2O on the sulfur production,and found the H2also released from the chemical decomposition of H2O under the role of SH, O and H radicals.The effect of CO2on the distribution of H2was primarily in the flame region,where CO2decreased the peak value of H2volume fraction,particularly in H2S feed with 40%CO2.It was attributed to that CO2possibly diluted the concentration of H2and also more H2occurred the oxidation reaction under the role of O and OH radicals(Eqs.(21)and(22))[52,57].Some studies also reported that the H2occurred the oxidative competition with H2S,and was inhibited by the oxidation of H2S [25,33].

The distribution of H2at R = 0.75 was shown in Fig. 5(b). The volume fraction of H2was detected with a peak value at the first sampling position, which was the main difference compared with Fig.5(a).Moreover,the volume fraction of H2at R=0.75 was lower than that at R = 0.0 possibly since some H2occurred the oxidation reaction.

Fig. 6 presents the distribution of SO2along the reactor. SO2at R = 0.0 was detected with a low volume fraction at the first sampling position in Fig. 6(a) since the chemical decomposition occurred, and then increased to a steady value. It can be initially inferred that H2S was potentially converted to S2, can be represented by Eq. (23) [25].

Fig. 5. Volume fraction of H2 versus the axial distance: (a) R = 0.0, (b) R = 0.75.

Fig. 6. Volume fraction of SO2 versus the axial distance: (a) R = 0.0, (b) R = 0.75.

Some studies has reported that H2S was oxidize to SO2with the presence of O2, rather than sulfur during the acid gas combustion[25,30]. Presence of CO2increased the oxidizing medium in the reaction system, which promoted the H2S oxidation. Selim et al.[22] reported that the increase of CO2in the H2S feed increased SO2amount. However, in the present study the SO2amount released from H2S feed with CO2did not increase to a relatively high value. This was possibly attributed to that CO2addition diluted the product concentration. Selim et al. [22,23] found that SO2released from H2S oxidation mainly experienced two stages,H2S was firstly oxidized to intermediate species such as SO, SH,and HSO radicals(Eqs.(18)–(20),(24)and(25)),and then the intermediate species were further recombined or oxidized to SO2(Eqs.(26)–(29)).

The volume fraction of SO2at R = 0.75 was shown in Fig. 6(b),and compared with the volume fraction of SO2detected at R=0.0 in Fig.6(a),SO2essentially reached a higher value at the first sampling location.This was due to that H2S occurred the oxidation reaction at R=0.75.It can be seen that the difference in SO2distribution was still in the flame region compared with the SO2distribution in Fig. 6(a).

Fig.7 presents the distribution of CO along the reactor.It should be noted that CO was not released from the H2S feed in the absence of CO2.The CO was detected with a low volume fraction at the first sampling position in Fig. 7(a), and increased to a peak value with the increase of axial distance, then decreased to a relative stable value. The presence of CO2triggered the formation of CO, and the CO amount was closely related to the CO2in the feed. For the formation of CO, Selim et al. [22,23,25] found that CO was mainly through the decomposition of CO2(Eq. (30)) and the reaction of CO2with H radical (Eq. (31)).

The oxidizing medium (O and OH radicals) released from these two reactions were involved in the oxidation of H2S and other combustible species. However, in the present study the flame temperature is not sufficient for CO2decomposition since the carbon–oxygen double bonds in CO2are strong,it takes higher temperature to break them. For the consumption of CO, Li et al. [25] reported that the decrease of CO mainly preformed the oxidation reactions under the role of O and OH radicals. Selim et al. found that the decrease of CO was attributed to the conversion of CO into SO and CO2(Eq. (32)) [22]. Ibrahim et al. [23,32] reported that CO enhanced the SO2decomposition (Eq. (33)), and also reacted with sulfur species to produce COS and CS2. However, according to the distribution of SO2and COS in the present study,it can be inferred that the decrease of CO was mainly attributed to the oxidation reactions.

The distribution of CO at R = 0.75 was depicted in Fig. 7(b),which shown the same trend as H2at R = 0.75. The peak value was detected in the flame region,and still lower than that detected at R = 0.0 due to the oxidation.

Fig. 8 presents the distribution of COS along the reactor. It should be noted that COS was not detected in the H2S feed without CO2. The volume fraction of COS in Fig. 8(a) increased to a peak value, and then decreased with the axial distance increases. It can be seen that the COS volume fraction did not increased with the increase of CO2amount, mainly in the oxidation stage. COS released from the H2S feed with 40%CO2was relatively lower than that released from the H2S feed with 20% CO2, this was initially inferred that COS possibly occurred the oxidation reaction rather than CO2dilution since CO2could serve as an oxidizing medium.Moreover,this inference was also reinforced by the higher volume fraction of COS downstream reactor in the presence of 40% CO2.

Fig. 7. Volume fraction of CO versus the axial distance: (a) R = 0.0, (b) R = 0.75.

Fig. 8. Volume fraction of COS versus the axial distance: (a) R = 0.0, (b) R = 0.75.

For the formation of COS, Karan et al. [26] found that the peak yield of COS released from (CO2and H2S, <2% (vol)) feed was less than 0.3%, which can be neglected compared with COS released from(CO and H2S, <2% (vol)) feed.In this study,COS was detected with a lower value of 0.3% at the first sampling position, which can be inferred that COS was mainly through the reaction of CO2with sulfur species such as H2S, S and SH radicals (Eqs. (34)–(36)) [16,50,58,59]. As the axial distance increases, COS volume fraction increased to a peak value possibly due to the increase of flame temperature and CO amount,indicated that COS formation was mainly through the reaction of CO with sulfur species (Eqs.(37) and (38)) [16]. Moreover, small amount of COS possibly released from the reaction of CO2with sulfur species in this process.

For the consumption of COS, The previous studies has reported that the decrease of COS was mainly oxidized under the role of OH and O radicals (reverse Eqs. (35)-(37)) [25,31]. COS hydrolysis(reverse Eq. (34)) could be considered as another channel for its consumption, whereas this was not the dominant channel.

Fig. 8(b) presents the distribution of COS at R = 0.75 along the reactor. The peak value of COS volume fraction was relatively low compared with COS at R = 0.0 because of the oxidation reaction.The COS was detected with a peak value at the first sampling position,and decreased with the axial distance increases.It can be initially inferred that COS formation was possibly through the reaction of CO with sulfur species. Moreover, the distribution of COS shown the same trend as that of CO at R = 0.75, which also indicated that COS formation was closely related to CO.

4.3. ROP analysis

The ROP analysis was applied to analyze the the oxy-fuel combustion of H2S without or with 40% CO2. ROP analysis provided a better channel for understanding the contribution of each reaction to the main target products in the viewpoint of net production or consumption. This section mainly discussed the ROP distribution of H2S, H2, SO2and COS under stoichiometric condition. It should be noted that a positive ROP value represented that this reaction promoted the formation of target products, whereas a negative value represented the consumption of target products [60]. In this study some reactions were not listed because these reactions had a low contribution to the target products compared with the major reactions, and even could be neglected. Moreover, the ROP distribution of CO and CO2were also not exhibited since the conversion between CO and CO2was mainly through CO2+ H = OH + CO.

The ROP distribution of H2S was shown in Fig.9,The ROP results indicated that the chemical decomposition of H2S was mainly through the r2. H2S + H = SH + H2, and the oxidation of H2S was mainly through the r28. H2S + OH = SH + H2O (r2 and r28 denote the 2nd and 28th reaction in the detail mechanism). It can be clearly observed that the ROP value of r2 was higher,this indicated that the conversion H2S was mainly through the chemical decomposition.This was consistent with the present experimental result at R = 0.0.

Some differences can be observed that H2S also consumed through the reverse r11 with a low ROP value in Fig.9(a),whereas the r11 did not proceeded with an observable rate in Fig. 9(b).Moreover, the ROP of r28 in the H2S system increased rapidly to a peak value and then decreased to a certain value slightly above zero. Whereas in the H2S/CO2system the ROP of r28 decreased to a certain value slightly below zero, indicated that the oxidation of H2S essentially proceeded with a low reaction rate since the OH radical released continuously from CO2+ H = CO + OH.

The ROP distribution of H2was presented in Fig.10.The formation of H2was depicted as r2.H2S+H=SH+H2,r5.S+H2=SH+H and r248.OH+H2=H2O+H.The O and OH radicals played a dominant role in the H2consumption.

The ROP results of H2in the H2S system was shown in Fig.10(a).The formation of H2was mainly through the r2 in the initial stage,and the ROP of r2 increased to a peak value at about 1.24 cm,then decreased and until it reached to zero. The partial H2occurred the oxidation reaction through the r248 at a lower ROP value in this stage.However,at an axial distance over about 1.27 cm the inverse r248 replaced the r2 as the dominant pathway for H2formation,and the ROP increased to a peak value at about 1.33 cm. Meanwhile, the total ROP value also reached its peak, and then decreased to zero at about 1.35 cm indicating the end of H2formation. For the consumption of H2, the ROP increased rapidly to a peak value, and then decreased to zero, which indicated the end of H2oxidation. The oxidation of H2was mainly under the role of O and OH radicals, this was consistent with the previous studies[22,30].

The ROP results of H2in the H2S/COS system was shown in Fig.10(b).Presence of CO2increased the total ROP value compared with the ROP value in Fig.10(a).The addition of CO2was basically unchanged the ROP peak value of r2, whereas overall reduced the reaction rate of r2 since CO2possibly weakened the H2S decomposition.Moreover,compared with the ROP value of r248,CO2did not have a great impact on the ROP value of r5 and r250, whereas increased the reaction rate of H2formation from the inverse r248, and decreased the oxidation rate of H2in the initial stage.The reaction of CO2with H radical released more OH radical without any restriction from H radical since the ROP value of r2 decreased.Therefore,the peak rate of H2released from the inverse r248 in Fig. 10(b) was significantly higher.

The ROP distribution of SO2was depicted in Fig. 11. The r115 and r138 merely contributed to the formation of SO2in the whole reaction system. Other reactions could be regarded as the reversible reaction and contributed to the formation or consumption of SO2at a certain region, which indicated that the conversion of SO2was complex.The ROP value in Fig.11(b)was higher than that in Fig. 11(a) since CO2was regarded as an oxidizing medium.

The formation of SO2was in a relatively low ROP value compared with H2in the initial stage, which indicated that the SO2had a small amount. This was initially inferred that H2S mainly proceeded the chemical decomposition. As the axial distance increases, the ROP of SO2increased rapidly to a peak value,whereas the ROP of H2decreased to a lower value and also had a negative value.This indicated that H2S mainly occurred the oxidation reaction and accompanied by the H2oxidation. This was also consistent with the present experimental result at R = 0.0.

In the chemical decomposition stage,the SO2consumption was mainly through the r236 in Fig. 11(a), whereas in Fig. 11(b) the total ROP value of SO2was about zero, indicating that the formation and consumption of SO2were essentially balanced.Moreover,in this stage some reactions exhibited the different functions in both systems. The r169 had no contribution to the conversion of SO2in the H2S system, and the r238 consumed the SO2with a low reaction rate. However, in the H2S/CO2system the r238 promoted the formation of SO2and the r169 consumed SO2. This was attributed to that the OH radical, released from the reaction of CO2with H radical, was not conducive to the formation of S2.

As the axial distance increases,the ROP of SO2increased rapidly to a peak value, particular in the H2S/CO2system it has a higher value, where the r115 played a dominant role in both systems.For the H2S system, the r117, r135, r138 and r236 also promoted the formation of SO2with the different reaction rates, whereas the r117 in the H2S/CO2system had no significant effect on the SO2formation.Moreover,in the H2S system the r169 played a role in the formation and consumption of SO2at a certain distance,which probably implied that r169 was temperature- dependent.Whereas the role of r169 to the H2S/CO2system could be neglected in this process. Meanwhile, some reactions also consumed SO2such as inverse r133,r135 and r238,where the inverse r133 played a dominant role.

Fig. 9. The ROP distribution of H2S versus axial distance.

Fig. 10. The ROP distribution of H2 versus axial distance.

Fig. 11. The ROP distribution of SO2 versus axial distance.

The ROP distribution of COS was depicted in Fig. 12. The ROP value of COS in the H2S/CO2system was much lower than the other three species, which implied that a small amount of COS was released from the acid gas. This was consistent with the experimental results in the present work. The inverse r583(COS + H2O = H2S + CO2) can be considered as the most swift pathway for COS formation. The consumption of COS in the H2S oxidation stage was mainly through the oxidation reaction r585(COS + O = CO + SO) and hydrolysis reaction (r583).

Fig. 12. The ROP distribution of COS versus axial distance.

4.4. Reaction pathway analysis

According to the ROP results,the combustion of acid gas mainly experienced the H2S chemical decomposition, H2S and H2oxidation. The section mainly analyzed the reaction pathway of each stage during the acid gas (H2S and CO2) combustion, and corresponding the results were presented in Figs. 13, 14 and 15,respectively.

The reaction pathway in the H2S chemical decomposition stage was depicted in Fig. 13. This stage was also accompanied by the partial oxidation of H2S, H2S was mainly converted into S2species and a small amount of SO2.According to the Fig.13,it can be summarized that the formation pathway of S2species could be depicted as H2S → SH → S → S2, and SO2was H2S →SH →HSS →HOSO →SO* →SO →SO2. The SH radical was a significant specie for the conversion of H2S to S2and SO2since it bridged the conversion pathways between sulfur species.For the CO2conversion,COS bridged the relationship between carbon and sulfur species,complicating the H2S conversion.The reaction pathway initially verified the inference of above experiments analysis about the formation of COS in the H2S chemical decomposition stage.

Fig.13. Reaction pathway for the chemical decomposition and partical oxidation of H2S.

Fig. 14. Reaction pathway of H2S oxidation.

Fig. 14 shown the reaction pathway in the H2S oxidation stage.It can be observed that the reaction pathway of H2S conversion was mainly through four pathways, indicating that this stage was more complex compared with the chemical decomposition stage.The formation pathways of SO2could be depicted as H2S →SH →HSS →S2→SO →SO2, H2S →SH →HSS →S2→SO → SO* → SO2, H2S → SH → S → SO → SO2and H2-S →SH →S →SO →SO* →SO2. H2S was directly oxidized to SH radical under the role of O and OH radicals, which was consistent with the experimental analysis.The S and S2species were oxidized to SO, and SO was further oxidized to SO2, which varied the viewpoint that H2S was oxidized to SO2in the presence of O2rather than S2species, and the formation of SO2was mainly through the recombination or oxidation of intermediate species.Moreover,the formation of S2was mainly through the reaction of HSS with H and OH radicals, rather than through the reaction of SH radical with S, which was different from the decomposition stage. It should be noted that the SH radical also played an important role in the formation of H2. SH radical was directly converted to H2under the role of H radical, and also could be indirectly converted to H2through its reaction products such as HSS,HSSH radicals and H2O. For the CO2conversion in the H2S oxidation stage, the presence of CO2indirectly triggered the COS formation through the reaction of CO with S and SO, which was considered as another essential pathway. Moreover, CO2also reacted with S species to form COS.The consumption of COS was mainly through the oxidation reaction under role of O and OH radicals.

Fig. 15. Reaction pathway of H2 oxidation.

Fig.15 depicted the reaction pathway in the H2oxidation stage.The H2formed is directly oxidized to H2O and OH radical under the role of OH and O radicals, respectively. Because the oxidation of H2S accompanied by the oxidation of H2, so the reaction pathway of SO2released from H2S could be summarized as:H2S →SH →S →SO →SO* →SO2and H2S →SH →S →S2→SO →SO* →SO2. It should be noted that the conversion pathway of COS in this stage was the same as that in the H2S oxidation stage,thus the conversion of CO2into COS was not listed.

5. Conclusions

The effect of CO2addition (0%, 20% and 40%) on the oxy-fuel combustion of H2S was conducted in a coaxial jet double channel burner.The distribution of the flame temperature and gas products along the axial at R=0.0 and R=0.75 were analyzed,respectively.Chemkin-Pro software was utilized to analyze the ROP for gas products and the reaction pathway of acid gas combustion.

H2S combustion at R=0.0 involved the H2S chemical decomposition, H2S and H2oxidation, whereas at R = 0.75 experienced the complete oxidation. Presence of CO2and CO triggered the formation of COS at R = 0.0, whereas at R = 0.75 COS formation was mainly through the reaction of CO with sulfur species. The COS consumption mainly occurred the oxidation reaction in the presence of O and OH radicals.

Both experimental and ROP results indicated the chemical decomposition was the main conversion pathway of H2S. H2S occurred the oxidation reaction,accompanied by the H2oxidation.The formation of H2was mainly through H2S + H = H2+ SH in the chemical decomposition stage, whereas in the oxidation stage H2mainly released from H2O + H = H2+ OH. The formation of SO2was mainly through the sulfur species oxidation and the reaction between sulfur species.COS was directly formed by the reaction of H2S with CO2, and oxidized by the O radical.

The H, OH and SH radicals were involved in each stage of H2S combustion, CO2complicated the combustion of H2S, particularly in the H2S oxidation stage. The reaction pathway results further verified the viewpoint that H2S was oxidized to SO2in the presence of oxygen rather than sulfur, and SO2was formed by the reaction between intermediate species.

The study related to the reaction mechanism of acid gas combustion has further improved and optimized the formation and consumption pathways of each species, providing an important theoretical basis for the conversion and utilization of H2S.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The project was supported by the National Natural Science Foundation of China (21978092), Chenguang Program by Educational Administration of Shanghai (21CGA35), and Yangfan Program by Scientifical Administration of Shanghai (22YF1410300).