Experimental study of partially decoupled oxidation of ethane for producing ethylene and acetylene

Jiajia Luo,Jinfu Wang,Tiefeng Wang*

Beijing Key Laboratory of Green Reaction Engineering and Technology,Department of Chemical Engineering,Tsinghua University,Beijing 100084,China

Keywords:Partial oxidation Experimental validation Pyrolysis Ethane conversion Partially decoupled process(PDP)Jet-in-cross- flow(JICF)reactor

A B S T R A C T With increasing amount of unconventional natural gas,the production of ethane,propane and other low alkanes continues to increase.In our previous works,a partially decoupled process(PDP)was proposed for conversion of ethane based on numerical simulations,which showed higher acetylene and ethylene selectivities than the original partial oxidation process.In the current work,the PDP of ethane for producing acetylene and ethylene was studied experimentally to verify the PDP concept.In the PDP of ethane,coke-oven gas or other cheap gas combusts with stoichiometric oxygen as heat carrier,and ethane is mixed with the heat carrier and undergoes pyrolysis at high temperatures.The jet-in-cross- flow(JICF)reactor was designed and manufactured to realize the PDP.Apositioning device of 0.1 mm accuracy and a mass spectrometer were used to measure the spatial profiles of the species concentrations.The maximum combined yield(52.7%)of acetylene and ethylene was obtained even at the condition of heat loss,confirming that the PDP of ethane was advantage ousover the partial oxidation process and at least comparable to the steam cracking process.

1.Introduction

With the successful application of horizontal well and fracturing technology,the production of shale gas continues to increase[1–3].Different from conventional natural gas,the shale gas contains a considerable amount of hydrocarbons heavier than methane,such as ethane,propane and butane.Meanwhile,the exploitation of natural gas hydrate has developed rapidly in recent years.Similar to shale gas,terrestrial natural gas hydrate also contains hydrocarbons heavier than methane,which can exceed 40%[4–6]and can be separated by condensation.Most ethane is used as the feedstock of steam cracking to produce ethylene[7,8].

In industry,the steam cracking process is the main approach to convert ethane into ethylene,in which the ethylene yield is around 52%with 80%selectivity,and the major by-product is methane[9].Although the ethylene selectivity of the steam cracking process is high,it has the following disadvantages:(1)the indirect heating manner causes a complex reactor structure and high investment cost;(2)the relatively long residence time requires a large reactor size;(3)the severe coking needs periodic decoking operation,which reduces the productivity of equipment[10–12].Other methods of utilizing ethane include catalytic partial oxidation[13,14]and oxidative dehydrogenation[15,16].Using V2O5/Al2O3[17],Pt and Rh coated on monolith[18]and MoVTeNbO mixed metal oxide[19]as catalysts,the yield of ethylene changed in the range of 40%–75%.

The partial oxidation(POX)of methane to produce acetylene is an important natural gas chemical process.In the POX process,a part of methane combusts with oxygen to provide heat for the pyrolysis of the remaining methane.There are two underlying assumptions in the POX process:the oxidation reactions are much faster than the pyrolysis reaction,and the pyrolysis reactions are independent of the oxidation reactions.However,analysis of the POX process based on detailed mechanism shows that the oxidation reactions are coupled with pyrolysis reactions,and the pyrolysis products such as acetylene can be further oxidized,leading to a low acetylene yield.

In our previous works[20,21],the partially decoupled process(PDP)was proposed to enhance the C2H2yield from CH4and the C2(C2H2+C2H4)yield from C2H6.In the PDP of ethane,coke-oven gas or other cheap gas combusts with stoichiometric oxygen as heat carrier,and ethane is mixed with heat carrier and undergoes pyrolysis at a high temperature.The reaction rate in the PDP is much faster than that in the steam cracking process due to the higher concentrations of oxidizing species and higher pyrolysis temperature in the PDP.The oxidation reactions and pyrolysis reactions proceed separately in different physical domains,but they are not separated from each other completely as in the steam cracking process.The original POX and the PDP of ethane have been simulated by computational fluid dynamics(CFD)coupled with a detailed mechanism,the modified GRI 3.0[22,23].For the original POX of ethane,the results showed that the maximum yield of C2H2was 37.5%and the maximum yield of C2(C2H2+C2H4)was 42%;for the PDP of ethane,the maximum yield of C2(C2H2+C2H4)was 69%with 72.5%selectivity,the maximum yield of C2H4was 50.5%with 53%selectivity,and the maximum yield of C2H4was 54.9%with 58%selectivity[20].

Fig.1.Schematic of experimental apparatus.

To further confirm the concept of the PDP process,experimental validation is needed.Ktalkherman et al.did a series of partially decoupled experiments[24],in which hydrogen combusted with oxygen as heat carrier and the feedstock included naphtha[25],propane[26]and liquefied petroleum[27].The yield of ethylene in the PDP was higher than that in the steam cracking process.How ever,these experiments used excessive pure hydrogen as fuel and heavier hydrocarbons as feed stocks,which are different for that in the PDP of ethane under the present consideration.

In this work,the PDP of ethane experiments was carried out in a jetin-cross- flow reactor.The spatial profiles of species concentrations in both axial and radial directions were sampled with apositioning system and measured by mass spectrometry(MS).The conversion of C2H6and the yield and selectivity of major products were analyzed and compared with those of the ethane steam cracking process and the simulation results of the ethane PDP.The results showed that the ethane PDP was advantageous over the POX process and at least comparable to the steam cracking process.

2.Experimental

2.1.Experimental apparatus

The schematic of the experimental apparatus for ethane PDP is shown in Fig.1.CH4,H2and CO were mixed to simulate the cokeoven gas,Ar was used to cool the combustion chamber,and N2was used to dilute O2during the ignition process.The gas purity was 99.5%for C2H6and 99.9%for other gases.All the gas flow s were controlled by mass flowmeters,which were calibrated by a wet flowmeter.In the experiments,the coke-oven gas containing 65%H2,25%CH4and 10%CO combusted with stoichiometric oxygen as heat carrier,and then ethane was mixed with heat carrier and undergone pyrolysis at high temperatures.A corundum sampling tube was fixed on the positioning system that could move in both axial and radial directions with 0.1 mm accuracy to sample the reaction products for analysis by a quadrupole MS(RGA200,SRS,America).

Fig.2.Structure of combustion chamber:(a)longitudinal profile;(b)top view.

Table 1Operating conditions

The reaction experiments were carried out in a jet-in-cross- flow reactor shown in Fig.2.The premixed coke-oven gas was jetted into the combustion chamber from the top,and the oxygen was jetted into the annular cavity and further jetted into the combustion chamber through small holes,as shown in Fig.2(a).The distributions of the small holes and the central fuel gas nozzle were shown in Fig.2(b).The azimuthal angle between two small holes was 20°,and one of the small holes was used to install the sparking needle.

The combustion chamber was made of high temperature resistant stainless steel,with an inner diameter of 20 mm and an outer diameter of 55 mm.The thick wall was designed to reduce the heat loss and the deformation of the combustion chamber.In order to cool the combustion chamber,Ar was injected into the jacket welded outside the combustion chamber.

2.2.Measurement methods

2.2.1.Temperature measurement

The temperature in the combustion chamber was far beyond the tolerance of thermocouples.The current work focused on the pyrolysis process of ethane rather than the combustion of coke-oven gas.Therefore,the temperature was measured only after the mixing of ethane and heat carrier.The temperature was measured by a K type thermocouple,which could be safely used below 1400°C.The measurement errors of the thermocouple mainly included thermal conduction error,thermal radiation error and the error caused by coke at the detecting tip.According to the literature,the error of temperature measurement was±100 °C[28].In order to reduce the coking error,the thermocouple was taken out from the reactor every a few minutes to clean up the coke.The effect of thermal radiation was considered to be the major source of error in thermocouple measurement,which was corrected as follow s[29,30]:

w here Tgwas the real gas temperature,Ttwas the temperature measured by thermocouple,εRwas the emissivity of the wire,λ was the thermal conductivity of gas,d was the diameter of welded point,and σ was the Stefan–Boltzmann constant(5.6704× 10?8W·m?2·K?4).Assuming the welding point was a spherical bead,the Nusselt number is

Fig.3.(a)Repeatability of two parallel experiments;(b)radial profiles of concentrations at h=200 mm;(b)element balance at the centerline;(d)composition of heat carrier.(Fuel jet velocity=2.0 m·s?1,equivalence ratio φ =1 and ethane mixing ratio f=0.366.)

w here Re is the Reynolds number and Pr is the Prandtl number.

2.2.2.Gas sampling and analysis

The gas products at different axial and radial positions were sampled online with a probe and analyzed by MS.The sampling probe was made of a corundum tube of 2.5 mm o.d.and 1.5 mm i.d.,with a 250 μm diameter orifice at the tip.The other end of the corundum sampling probe was connected with a mechanical vacuum pump to decrease the pressure below 1000 Pa.When the gas passed through the sampling probe,the temperature decreased by several hundred degrees due to the decrease of pressure.The decrease in both the pressure and temperature quenched the reactions,therefore the measured results represented the in-situ information of the sampling point inside the reactor.The temperature of sampling tubing was kept above 100°C using heater band to prevent condensation of water vapor.

All radicals were annihilated on the wall of the probe before detected by MS and only stable species were measured.The measured species included C2H2,C2H4,C2H6,CO,H2,O2,CH4,H2O and CO2.The relative concentration was determined by peak intensity divided by its response coefficient,and the absolute concentration was determined by normalization of the sum of species concentrations to 1.0.The calibration was carried out by measuring the peak intensity ratio of a series of mixture with known compositions.Argon was used as a reference to obtain the response coefficient.

Fig.4.Axial profiles of the mole fractions along the centerline under different fuel jet velocities(φ=1 and f=0.366):(a)C2H6;(b)C2H4;(c)C2H2;(d)CO.

2.3.Operating conditions

The ratios of H2,CH4and CO were fixed according to the composition of presumed coke-oven gas.The fuel gas jet velocity,which was defined as the cross-sectional average velocity at the fuel inlet of 6 mm in diameter,varied within 1.8–2.2 m·s?1.The equivalence ratio φ,defined as the molar ratio of stoichiometric oxygen to actual oxygen,varied within 0.9–1.1.The mixing ratio f,defined as the molar ratio of ethane to fuel,varied within 0.21–0.44.Note that the volumetric flow rates were defined at atmospheric pressure and 20°C.The Reynolds number in the main channel varied within 2900–4000.All of the operating conditions are summarized and listed in Table 1.

2.4.Reliability verification

The experiment repeatability was studied with two parallel experiments.As shown in Fig.3(a),the mole fractions of C2H2,C2H4and C2H6were almost the same for the two parallel experiments at the axial position of h＞50 mm,which was defined as the distance from the ethane nozzles.Due to the large concentration gradient near the nozzle,the mole fraction of C2H6did not repeat well because a small position error could cause a large concentration error.The conversion,yield and selectivity were calculated from the data after an axial position of 100 mm,thus the results had a good repeatability.

Fig.3(b)showed the radial profiles of the mole fraction distributions of C2H2,C2H4and C2H6at h=200 mm.The radial gradient could be neglected,thus the concentrations at the centerline were used as the cross-sectional average values.The experiment reliability was also checked by element balance.At axial positions of h=200 mm,the radial profile of the concentrations became uniform and the element balance was 99.1%,99.7%and 103.8%for carbon,hydrogen and oxygen,respectively,as shown in Fig.3(c).In contrast,at axial positions of h＜50 mm the two ethane steams converged in the central region,leading to non-uniform radial profiles of species concentration,thus a high carbon element concentration at the centerline.

A key point of the PDP was to physically decouple the oxidation and pyrolysis reactions so that the concentrations of strongly oxidizing species were low in the pyrolysis process.The composition of heat carrier is shown in Fig.3(d).As expected,the fuel was almost completely combusted.As a result,the major species were H2O and CO2and the concentrations of strongly oxidizing species were low.

3.Results and Discussion

The effects of the fuel jet velocity,mixing ratio,and equivalence ratio in fuel combustion were investigated.The mole fractions and temperature were directly measured,while the conversion of C2H6and the yield and selectivity of major products were calculated from the mole fraction data by

where Fiwas the mole flow of species i in product,ziwas the number of carbon atoms in molecule,and FC2H6,0was the mole flow of C2H6in feed.It is important to note that the yield and selectivity were calculated based on carbon atoms in C2H6.The carbon atoms in the fuel were not considered in the calculations of yield and selectivity since they were almost converted into CO2during combustion.

Fig.5.Axial profiles of conversion,selectivity and yields along centerline under different fuel jet velocities(φ=1 and f=0.366):(a)C2H6 conversion;(b)C2H4 yield;(c)C2H2 yield;(d)selectivity of C2(C2H4+C2H2).

3.1.Effect of fuel jet velocity

The fuel gas jet velocity determined the fuel gas flow rate for a specified size of reactor.The experimental reactor had significant heat loss due to its small size and large specific surface area.A low fuel jet velocity could not provide enough heat for the pyrolysis of ethane,while a high fuel jet velocity would lengthen the flame,which in turn caused uncomplete consumption of oxygen before mixing with ethane.Different fuel jet velocities(1.8 m·s?1,2.0 m·s?1and 2.2 m·s?1)were investigated at f=0.366 and φ=1.0.When the fuel jet velocity exceeded 2.2 m·s?1,the reaction tail gas combusted outside the reactor with air and caused an abnormal operation of equipment.

Fig.4 shows the axial profiles of the molar concentrations of C2H6,C2H4,C2H2and CO at the centerline.As shown in Fig.4(a),the C2H6mole fraction had a peak near the nozzle and then decreased sharply.After the axial position of 100 mm,the C2H6mole fraction decreased slightly or became stabilized,because ethane was mainly converted in the high-temperature region near the nozzle.As shown in Fig.4(b)and(c),the C2H4mole fraction firstly increased and then decreased,while the C2H2mole fraction was almost unchanged at h＞50 mm,because C2H4was an intermediate product that is unstable at this temperature.With the increase of the fuel jet velocity,the C2H6mole fraction decreased and the mole fractions of C2H4,C2H2and CO increased.At a high fuel jet velocity,the heat loss was weakened relative to the increase of heat release,which enhanced the conversion of C2H6and increased the mole fractions of C2H4and C2H2.

The C2H6conversion,C2H4yield,C2H2yield and C2selectivity were calculated from the mole fractions measured at the centerline based on carbon atoms in C2H6.Since the mole fractions were not uniform near the nozzle,the data at centerline could represent the cross-sectional average values only when the axial position exceeded 100 mm.As the fuel jet velocity increased from 1.8 m·s?1to 2.0 m·s?1,the conversion of C2H6increased from 40%to 48%,the yield of C2H4increased from 21%to 24%,and the yield of C2H2increased from 5.8%to 8.5%,indicating that a high temperature favored the formation of C2H2.How ever,when the fuel jet velocity further increased to 2.2 m·s?1,the C2H4conversion and C2yield only changed slightly.Fig.5(d)shows the effect of the fuel jet velocity on the C2selectivity.The highest selectivity was obtained at 2.0 m·s?1fuel jet velocity.Generally,the conversion and yield were relatively high and the selectivity was the highest at 2.0 m·s?1fuel jet velocity,thus the fuel jet velocity of 2.0 m·s?1was used in the following experiments.

3.2.Effect of equivalence ratio in fuel combustion

The main advantage of the PDP over the original POX process is that the fuel combusts with oxygen prior to the addition of ethane to reduce the concentrations of strongly oxidizing species in the pyrolysis reactions.The concentrations of oxidizing species depend on the equivalence ratio in the fuel combustion.This effect was studied at equivalence ratios of 0.9,1.0 and 1.1.

Fig.6 shows the axial profiles of the mole fractions of C2H6,C2H4,C2H2and CO along the centerline under different equivalence ratios.The peak value of C2H6increased sharply with the increase of equivalence ratio,indicating that the oxidizing species greatly enhanced the conversion of C2H6,as shown in Fig.6(a).The mole fractions of C2H4and C2H2increased with the increase of C2H6conversion,as shown in Fig.6(b)and(c).As can be seen from Fig.6(d),the mole fraction of CO was almost unchanged.

The C2H6conversion,C2H4yield,C2H2yield and C2selectivity are shown in Fig.7.When the equivalence ratio decreased from 1.0 to 0.9,the C2H6conversion significantly increased from 31%to 60%,the yield of C2H4increased from 19%to 25%,and the yield of C2H2increased from 4%to 10%.How ever,the C2selectivity decreased sharply from 73%to 59%.Generally speaking,the increase of equivalence ratio promotes the conversion of C2H6and sharply lowers the C2selectivity,thus the equivalence ratio should be higher than 1.0 to keep a low concentration of oxidizing species.

3.3.Effect of mixing ratio

Numerical simulation results showed that f had a great influence on the ethane conversion and C2selectivity[20].In the experiments,the mixing ratio was changed by changing the ethane jet velocity at a fixed fuel jet velocity.It must be pointed out that changing the ethane jet velocity would also change the momentum ratio of ethane to heat carrier.In this experiment,the mixing ratio changed from 0.214 to 0.441,which was much lower than that in numerical simulations,where f in the range of 0.83–2.1 was used[20].

Fig.8 shows the axial temperature profiles along centerline under different mixing ratios.When there was no ethane,the maximum temperature near the nozzle was about 1100°C,and the temperature sharply decreased to 430°C at the outlet of the reactor.Although the heat loss in the pyrolysis section was severe,the conversion of C2H6mainly occurs at high temperatures above 800°C,which fits the requirement in industry.With increasing ethane flow rate,the mixing temperature continued to decrease.When the mixing ratio exceeded 0.366,the temperature had a maximum value in the axial direction.Because two ethane streams of low temperature crossed over the heat carrier stream and converged at the centerline,the temperature at centerline is relatively low.With the increase of h,the high-temperature heat carrier near the wall mixed with the gas in the central region,causing an increase of the temperature at the centerline.Compared with heat loss,the endothermic heat of pyrolysis reactions was much smaller,thus the outlet temperatures were similar in different conditions.

Fig.6.Axial profiles of mole fractions along the centerline at different equivalence ratios(fuel jet velocity=2.0 m·s?1 and f=0.366):(a)C2H6;(b)C2H4;(c)C2H2;(d)CO.

The conversion of C2H6was almost completed within 50 mm axial distance,as shown in Fig.4(a),with a corresponding residence time of about 0.02 s.The residence time of ethane in the steam cracking process is about 0.25 s,which is several times longer than that in the PDP.The oxidizing species,such as OH radical,O radical and residual O2,and the high temperature during the mixing process greatly accelerate the pyrolysis of ethane,which can significantly reduce the reaction time and reactor size.

Fig.8.Axial profiles of temperature along centerline at different mixing ratios.(Fuel jet velocity=2.0 m·s?1,φ =1.)

Fig.9 shows the C2H6conversion and the product selectivities and yields at different mixing ratios.The effect of mixing ratio was investigated from 0.214 to 0.441 to determine the optimal value.At a mixing ratio below 0.214,the concentrations were non-uniform at the outlet of reactor and the data at centerline could not represent the cross sectional average values,thus the low est mixing ratio investigated in experiment was 0.214.The yields of C2H2and C2H4are shown in Fig.9(a).The C2yield reached its maximum(52.7%)when the mixing ratio was 0.214.The concentration ratio of C2H2to C2H4in the product increased with the mixing ratio,because the high temperature favored the formation of C2H2.

Fig.9(b)shows the C2H6conversion and C2selectivity at different mixing ratios.As expected,the conversion of C2H6decreased with increasing mixing ratio,showing a maximum conversion of 73%and a minimum conversion of 35%.The minimum selectivity was 60%at a mixing ratio of 0.34.When the C2yield reached its maximum,the corresponding C2selectivity and C2H6conversion were 72%and 73%,respectively.Fig.9(c)and(d)shows the yields of C2H2and C2H4near the outlet of reactor.The yields were also unchanged in the axial direction in this region.Thus,it was reasonable to use the data at 200 mm axial position to compare in Fig.9(a)and(b).

Fig.9.(a)C2H4 and C2H2 yields(h=200 mm),(b)C2H6 conversion and C2 selectivity(h=200 mm),(c)axial profile of C2H4 yield,and(d)axial profile of C2H2 yield.(Fuel jet velocity=2.0 m·s?1,φ =1.)

The experimental maximum C2yield in the PDP was much higher than that in the simulations of the original ethane POX,confirming the superiority of the PDP over the POX process.Our previous CFD simulations of the ethane PDP showed that the maximum C2yield was 69%with 72.5%selectivity(fuel jet velocity=100 m·s?1,f=1.64,φ =1 and adiabatic condition)[20].The measured C2selectivity at the maximum C2yield was 73%,in consistence with the simulation results.Note that the simulated C2yield was higher than the experimental results because of the heat loss in the experimental apparatus.The effect of scale up includes a lower mixing efficiency and less heat loss.On the one hand,the lower mixing efficiency will decrease the C2selectivity,but this unfavorable effect can be eliminated by changing the con figuration of the reactor.On the other hand,the less heat loss will enhance the C2yield,since the ethane conversion could be enhanced from 72%in the experimental reactor to 95%in the numerical simulation at adiabatic conditions.

In the industrial steam cracking process,the single-pass ethylene yield was about 52%with a selectivity of 80%[9].In the current experiments,the maximum C2yield was 52.7%,which was very close to the C2H4yield in the steam cracking process.Because the conversion of C2H6in the experiment was higher than that in the steam cracking,the selectivity to C2(C2H4+C2H2)was 10%lower than that in the ethane steam cracking.

Fig.10.Distributions of(a)carbon atom and(b)oxygen atom in products under different mixing ratios.(Fuel jet velocity=2.0 m·s?1,φ =1.)

Different from the steam cracking process,the PDP directly mixes ethane with heat carrier,therefore the effect of oxidizing species in heat carrier on the pyrolysis reactions needs to be analyzed.The distributions of carbon and oxygen atoms in the products were analyzed,as shown in Fig.10.Based on carbon atoms in C2H6,the ratio of carbon atoms in C2H4and C2H2was about 70%and the ratio of carbon atoms in CO was about 50%,as shown in Fig.10(a).Note that the total ratio was higher than 100%,because some CO2in the heat carrier was converted into CO.These results indicated that the PDP reduced the emissions of CO2and enhanced the use of carbon sources.In contrast,the major by-product of the ethane steam cracking was CH4.

The distribution of oxygen atoms in the product is an indicator for the extent of decoupling of the oxidization and pyrolysis reactions.If all oxygen atoms are contained in H2O and CO2,the oxidation and pyrolysis are completely decoupled.As shown in Fig.10(b),the oxygen atoms in H2O and CO2accounted for more than 90%of all oxygen atoms,and less than 10%oxygen atoms were converted into CO.In the original POX of methane,industrial results show that 56.6%oxygen atoms are contained in H2O and CO2and 42.6%oxygen atoms are contained in CO[22].Compared with the original POX process,the PDP significantly prevents the formation of CO and enhances the yield of C2.

4.Conclusions

The new process called PDP for ethane conversion was studied experimentally in this work.The jet-in-cross- flow reactor was designed to realize the PDP.The experimental results were compared with the steam cracking process and the simulations of POX and PDP.The main conclusions are as follow s:

(1)The mixing ratio and equivalence ratio are important operating parameters that affect mixing temperature near the nozzle.Decreasing equivalence ratio greatly enhances the ethane conversion but decreases the C2selectivity.

(2)The maximum C2(C2H4+C2H2)yield is52.7%in the experiment,which is very close to the C2H4yield of 52%in the ethane steam cracking and much higher than the C2(C2H4+C2H2)yield in the original POX process,showing that the PDP is superior to POX and at least comparable to the steam cracking process.

(3)For the PDP of ethane,the measured highest C2selectivity(73%)is similar to the simulated value(72.5%),while the measured ethane conversion(72%)is much lower than simulated value(95%)due to the low reaction temperature.

(4)Compared with the steam cracking of ethane,the oxidizing species in the PDP greatly accelerates the conversion of ethane within 0.02 s;the major by-product of PDP is CO,and a part of CO2in heat carrier is converted into CO,which reduces the emissions of CO2and enhances the use of carbon sources.