Kinetic study of thermocatalytic decomposition of methane over nickel supported catalyst in a fluidized bed reactor

(cid:1) Thermocatalytic decomposition of methane was performed in a ﬂuidized bed reactor. (cid:1) The effects of temperature, concentrations and space velocity were evaluated. (cid:1) The kinetics and deactivation of the reaction was modelled. (cid:1) Better understanding of the growth of the catalyst particles in a ﬂuidized bed reactor. (cid:1) The characteristics of the carbon nanomaterial were studied.


Introduction
The destructive consequences of the climate change crisis followed by ongoing efforts toward emission-free technologies have instigated a growing interest in low CO 2 and CO 2 -free hydrogen production (Amin et al., 2011;Borghei et al., 2010;Ashik et al., 2017;Nezam et al., 2021;Ra et al., 2020).Various approaches such as chemical looping reforming, steam methane reforming integrated with Carbon Capture and Sequestration processes (CCS), water splitting, thermal and thermocatalytic decomposition of methane have been studied for this purpose.Among those, thermocatalytic decomposition of methane (TCD) is one the most promising (Hadian et al., 2021).A major advantage of TCD is the potential capability of producing highly valuable carbon nanomaterials instead of CO 2 , next to hydrogen.The intrinsic characteristics of carbon nanomaterials make them suitable for many industrial applications such as building materials, semiconductors, catalytic materials and energy storage (Ashik et al., 2015;Ashik et al., 2017;Douven et al., 2011;Saraswat and Pant, 2013).In addition, TCD requires less complex down stream purification or separation units than conventional processes.These advantages make TCD an environmentally and economically attractive approach for CO 2 -free hydrogen production (Hadian et al., 2021).
Methane is thermally decomposed to solid carbon and gaseous hydrogen in the absence of a catalyst or oxidizing agents at temperatures above 1300 C (reaction 1).Alternatively, in TCD, a catalyst facilitates the same reaction at a much lower temperature (500 °C-950 °C) with formation of nano-structured carbon materials.The structure of this material depends on operating conditions and foremost on the catalyst properties that are employed (Hadian et al., 2021).Nickel, iron, copper and carbon are the most studied active sites of the catalyst and among them nickel on silica support, Ni-SiO 2 , showed the highest methane decomposition activity (Pudukudy et al., 2016;Wang et al., 2000;Reshetenko et al., 2003;Avdeeva et al., 1996;Li et al., 2000;Guevara et al., 2010).
CH 4 ðgÞÀ > CðsÞ þ 2H 2 ðgÞ DH ð298KÞ ¼ þ74:52 kJ=mol ð1Þ A considerable amount of literature has been published on the preparation of single or bimetallic or carbonaceous catalysts.Their performance in very small lab-scale units and under mild reaction conditions have been established.Srilatha et al. (2017) and Ashik et al. (2015Ashik et al. ( , 2017) ) reviewed and compared these studies using carbonaceous and metallic catalysts, most of which have been employed in small-scale fixed bed reactors with up to 0.5g of catalyst at limited space velocities and low concentrations of methane.
Since the size of the catalyst particle in TCD is increasing over time due to carbon build-up, fixed bed reactors suffer from serious drawbacks such as a high probability of clogging, particle crushing, increasing pressure drop and fracturing the body of the reactor.Therefore, fluidized bed reactors are preferred over fixed bed reactors for large-scale TCD.Indeed, there have been few studies that use fluidized bed reactors or high space velocities (ratio of the total flow rate at normal conditions per gram of catalyst initially loaded), SV.For instance Torres et al. (2012) performed experiments with 20g of fine catalyst particles in a fluidized bed; however, the SV did not exceed 0.2L n / g cat / min.Suelves et al. (2009) used higher SV (2L n / g cat / min) in a fixed bed reactor that contained no more than 0.05g of catalyst.Alongside experimental parametric studies on the performance of the reaction, kinetic studies on TCD in a fixed bed reactor and mild conditions and the mechanism investigations of reaction 1 over metallic catalysts have been performed.These studies revealed that the actual rate of TCD is not constant over time and can be described by Eq. 2, where r 0 is the maximum reaction rate and aðtÞ is defined as the deactivation factor (Borghei et al., 2010;Amin et al., 2011;Douven et al., 2011;Latorre et al., 2010).
In an earlier contribution, the authors summarized the kinetic studies and the proposed kinetic models, including the maximum reaction rate and deactivation factor of the catalyst (Hadian et al., 2021).Several researchers (Amin et al., 2011;Saraswat et al., 2016) proposed a mechanism based on the molecular adsorption of methane followed by step-by-step dehydrogenation reactions until separate adsorbed atoms of carbon and hydrogen are obtained.The first dehydrogenation reaction was found to be the rate-limiting step.The remaining carbon atom of methane on the surface of metal active site, passes through the metal by diffusion and forms nano layers of carbon on the other side.If the decomposition step occurs faster than diffusion and construction rate of carbon nano-structures, carbon atoms accumulate on the surface of metal active site and deactivate it by encapsulation (Toebes et al., 2002;Henao et al., 2021).The maximum reaction rate is modelled by a Langmuir-Hinshelwood type equation that accounts for the thermodynamic equilibrium and the competition between hydrogen and methane adsorption over the active sites, as represented by Eq. 3. The semi-empirical deactivation factor expression is obtained from a species balance on the active sites of the catalyst, resulting in Eq. 4.
Although extensive research has been performed on catalyst preparation and reaction mechanism of TCD, very little is currently known about the feasibility of the TCD process in large scale industrial fluidized beds at the harsh operating conditions encountered.The performance of the catalyst and the fluidized bed reactor need to be thoroughly investigated by performing experimental and numerical studies.This performance can be expressed in terms of the maximum reaction rate, r 0 , the life-time of the catalyst, LT, and carbon yield (the ratio of the mass of produced carbon to the initial mass of catalyst used, Eq. 5), CY.The outline of this paper is as follows: in Section 2 we describe the experimental setup, materials as well as the adopted procedures.In Section 3 we intro- carbon yield ðCYÞ ¼ mass of produced carbon ðgÞ initial mass of catalyst ðgÞ ð5Þ Experimental materials and methods

Catalyst and reactants
In the experiments a commercial catalyst made by BASF (Ni 5256 E RS) was used.The catalyst is originally designed as a hydrogenation fixed bed catalyst that contains 56% nickel on silica support and was supplied as extrudates and in reduced and passivated state.Table 1 shows the characteristics of the fresh catalyst.All of the gases used in this study were at least 99.995% pure, supplied by Linde.

Experimental setup and procedure
The experiments were performed in a cylindrical quartz fluidized bed reactor equipped with a spherical free-board section, see Fig. 1.The inner diameter and the height of the cylindrical part are 1cm and 10cm, respectively and the diameter of the free-board is 7.5cm.The spherical free-board reduces the chance of entrainment by lowering the gas velocity and at the same time acts as an expansion space for the growing catalyst particles.The reactor was placed in an electric oven and the desired feed gas composition and flow rate were adjusted by calibrated Bronkhorst mass flow controllers.The local temperature in the reactor can be measured with the help of thermocouples.The outlet gas is transferred to a SICK gas analyzer model GMS815P (three measuring modules: Thermor, Oxor-P and Multor) for gas composition measurement after cooling down, Fig. 2.
In each experiment 1g of crushed and sieved (500-600lm) catalyst was loaded into the reactor, unless otherwise stated (to examine the effect of the parameter).Prior to activation of the catalyst, the air in the porous catalyst particle was extracted by flowing 2L n / g cat / min nitrogen to the reactor for 30min.Subsequently the temperature of the reactor was increased to 250 °C by a ramp of 5°C/min using 10vol.%H 2 in the feed to actuate the catalyst.The catalyst was further reduced in 100vol.%H 2 with the same ramp of temperature up to 500 °C.Afterwards, the catalyst was heated up further in nitrogen.Once the reaction temperature was reached, 4.5L n / g cat / min of the predefined gas composition of methane, hydrogen and the inert gas (nitrogen) was fed to the reactor, and the outlet composition was measured typically until the catalyst was fully deactivated.Finally, the reactor was cooled down and the product was collected and weighed.

Product characterization methods
BET surface area and pore volume measurements have been carried out for the used catalyst using Thermo Fisher Scientific Analyzer model Surfer.Oxidation temperature of the produced carbon was measured in air via Thermo Gravimetric Analysis (TGA) to characterize the products.Surface composition of the fresh and used catalyst was analyzed by X-ray Photoelectron Spectroscopy (XPS) measurements on Thermo Scientific K-Alpha XPS with an Al source (1486.6 eV).The structure of carbon nanomaterials were obtained by performing Transmission Electron Microscopy (TEM) imaging of the samples of the products on a FEI cryo TEM TITAN 300 kV.

Reactor Model
The plug flow reactor model The composition of the gas entering the reactor is known and the same as the predefined values for each experiment.However, as the gas passes through the reactor methane is consumed and hydrogen is produced.Therefore, the composition of the gas, and consequently the reaction rate is varying along the height of the bed whereas only the outlet conversion can be compared with the experimental data.This conversion is calculated by a Plug-Flow Reactor (PFR) model, Eq. 6. See Fig. 3-a.where X is conversion, F A 0 is the molar flow rate of methane to the reactor, and dw is a fraction of the bed, such that the reaction can be considered constant over the fraction.r 0 is replaced by Eq. 3.This differential equation is numerically integrated by Runge-Kutta fourth-order method (RK4).The equilibrium constant at temperature of reaction, K p in Eq. 3, is calculated by Eq. 7 proposed by Kuvshinov et al. ( 1998), Zavarukhin and Kuvshinov (2004).
The local partial pressures of methane and hydrogen are updated by Eq. 8 and 9. Finally, the kinetic parameter of Eq. 3 was fitted by comparing the conversion of the last section of the fluidized bed with the maximum conversion obtained from each of the experiments.
Dynamic model Over time as the carbon products are being formed, the catalyst particles grow in size and weight with different rates.This growth can reach a point where some of the particles become too heavy to be fluidized by the available gas flow rate.Therefore, they settle down at the bottom of the reactor.These segregated particles are exposed to the fresh feed entering the reactor with a higher concentration of methane and lower concentration of hydrogen compared to the upper parts of the reactor.As a result, they grow faster and they also deactivate quicker than the rest of the particles.Over time more and more particles are segregated and deactivated until the whole bed is deactivated.
In order to predict the deactivation of the catalyst and determine the parameters of Eq. 4 it is crucial to model this complex behavior over time.In order to obtain a predictive model for the deactivation, the model described in Section 3.1 is run for each time step starting from the beginning of the reaction until full deactivation.At the end of each time step, the total consumed methane and produced hydrogen and solid carbon are calculated and used to update the particle size.Then the minimum fluidization velocity of the particles is calculated for fine and coarse particles with Eqs. 10 and 11, respectively (Kunii and Levenspiel, 2013).
where Re p;mf is the Reynolds number of the particles, Eq. 12, and Ar is the Archimedes number calculated by Eq. 13.
Re p;mf ¼ q g u mf d p l By growing the particles the minimum fluidization velocity increases and the ratio of gas velocity to the minimum fluidization velocity, u=u mf decreases.The moment that u=u mf is not sufficient to maintain fluidization (u=u mf < 1:2 this ratio is dependent on the reactor and particles properties), the bottom part of the bed is separated from the rest of the reactor (Fig. 3-b), and the particles are not mixed with the top part anymore.The inflow of gas to the top part is higher due to production of 2 mol hydrogen for each mole of consumed methane in the segregated section.Therefore, the ratio u=u mf can be high enough for fluidization of the particles in the upper sections of the bed.Due to further growth of the particles, the segregated zone propagates along the reactor and eventually the entire bed becomes segregated as shown in Fig. 3-c.This is a phenomenological 1D model representing the evolving reactor and therefore radial difference, wall effect, channelization, and bubble formation are neglected.Before the particles in the first section start segregating, all the particles are fluidized and well mixed in the reactor.Therefore, the particles grow with the same rate at this stage.Segregation only occurs if the carbon yield is high enough (mostly in cases with lower temperature or if hydrogen is added to the feed).Due to segregation, particles are not mixed any more.The growth rate is higher at the bottom of the reactor but there the deactivation starts earlier.

The effect of operating conditions
Many experiments were conducted by systematically altering the settings of operating temperature, gas concentrations, catalyst particle size and WHGV.All results were confirmed with at least one duplicate experiment.In these experiments depending on the settings lifetime varied from 5min to longer than 12h where the obtained carbon yield ranged between 0g C /g cat to more than 70g C /g cat (at 550 °C and 70vol.%CH 4 -5vol.%H 2 ).
The max. conversion of the reactor was limited to about 20% because of the very high SV.It was observed that although at lower SV fluidization occurs with fresh catalyst particles (u mf % 0:1m=s), the heavier and larger particles including the produced carbon (u mf depends on the CY and in can exceed 2m=s for the largest particles) cannot be fluidized and therefore leads to breaking the reactor.Therefore, the gas flow rate (and as a result SV) is chosen to be high enough for mobilization of the grown catalyst particles even after hours of accumulation of carbon on them.

Operating temperature
Since the diameter of the reactor is relatively small and the consumed heat by reaction is small compared to the heat supplied by the oven, temperature drop along the reactor was limited to a maximum of 17 °C (at maximum reaction rate at 600 °C and with a feed of 100% CH 4 ).Fig. 4 shows that the maximum reaction rate increases as the temperature is increased as expected.On the other hand, as can be seen in Fig. 5-bottom, a high temperature has a negative effect on the lifetime of the catalyst.These findings are in agreement with literature findings (Hadian et al., 2021;Amin et al., 2011).The carbon yield is a parameter that integrates both effects of maximum reaction rate and the lifetime of the catalyst.Therefore, as shown in Fig. 5-top from 550 °C to 650 °C a shorter lifetime can overcome the higher reaction rate and carbon yield is significantly lower.However, at lower temperatures the carbon yield is more affected by lower maximum reaction rate and there is an optimum temperature for carbon yield between 500-550 °C, balancing initial reaction rate and lifetime of the catalyst.

Concentration of methane
Fig. 6 shows that the maximum reaction rate is directly dependent on the volumetric concentration of methane as the only reactant of the reaction.What stands out in this figure is that at 550 °C and lower, the maximum reaction rate slightly declines with an increase in methane concentration to 90vol.%.A possible explana-Fig.4. The effect of temperature on the maximum reaction rate of the reaction.
tion for this might be that adsorption of methane is the dominating step at this temperature and saturation of the active sites allow less neighboring active sites to facilitate the detachment of hydrogen molecules.
One unanticipated finding was that the lifetime of the catalyst is shorter when the concentration of methane (and therefore the reaction rate) is lower, Fig. 7, while some other researchers observed the opposite behavior (Latorre et al., 2010;Henao et al., 2021).This effect is stronger at lower concentrations or at higher temperatures.We believe that the key difference is the scale of the reactor.Small reactors can be considered as a differential reactor and all the catalyst particles are in an environment with the same concentrations as the feed, while in the larger reactors such as in this work, except for the small portion next to the gas inlet, catalyst particles are in contact with a gas mixture containing the produced hydrogen.Another difference is that the methane concentration range in this study is much higher than most of the literature studies where only mild conditions were tested (max.methane concentration of 7.5% and 42.9% was tested by Latorre et al. ( 2010) and Henao et al. ( 2021) respectively).High concentrations lead to higher reaction rate and larger temperature drop along the reactor that is in favor of longer lifetime.At higher methane concentrations, larger amounts of hydrogen are also produced and are present in the reactor.As mentioned in Section 4.1.3,the addition of hydrogen changes the chemical potential and enhances the reverse reaction and converts the accumulated carbon on the surface of active sites back to methane.This phenomenon prevents encapsulation of the active sites and renews them, which boosts the lifetime of the catalyst significantly.Fig. 7 also reveals that carbon yield is decreased as the methane concentration is lowered by dilution with nitrogen.This was confirmed by using argon instead of nitrogen that led to almost the same effect.Analyzing the gas outlet with a mass spectrometer and also the solid products with XPS tests confirmed that neither nitrogen nor argon are involved in any reaction and are indeed inert.

Concentration of hydrogen
Adding a small fraction of hydrogen to the feed decreases the maximum reaction rate by promoting the reverse reaction by Le Chatelier's principle, Fig. 8.This leads to higher refresh rate of the surface of active sites and as a result a higher lifetime of the catalyst.Fig. 9 shows that also the carbon yield is improved by    introducing small amounts of hydrogen to the reactor.This behavior was also observed by other researchers such as Latorre et al. (2010).

Particle size
Three different catalyst particle sizes were tested to investigate the importance of mass transfer limitation in TCD process.It can be seen from Fig. 10 that the effect of changing the particle size from the average diameter of 350lm to 800lm on the maximum reaction rate is very small.This confirms previous findings in the literature (Hadian et al., 2021;Saraswat et al., 2016;Borghei et al., 2010).However, over time the catalyst particle become larger and the effect of mass transfer limitation becomes more important by decreasing both diffusion length scale and pore volume due to carbon formation (See Table 1).These findings are also confirmed by Weisz-Prater criterion, see the appendix.Fig. 11 shows that the carbon yield is directly affected by the initial size of the catalyst particle.Because, mass transfer limitation keeps the concentra-tions of methane and hydrogen inside the particles compared with smaller particles lower and higher respectively.Therefore, the lifetime of the catalyst is boosted and as the result carbon yield is also increased.

Space velocity (SV)
SV was changed in the experiments while maintaining the volumetric flow rate at 4.5L n /min by changing the amount of the catalyst in the reactor.Lowering the volumetric flow rate would change the fluidization regime and can lead to breaking the quartz reactor due to defluidization of the grown catalyst particles and increasing it would turn the reactor at the beginning of the reaction into a pneumatic riser.Fig. 12 shows that the reaction rate does not linearly increase with SV because the local reaction rate decreases along the height of the reactor.An increase in the reaction rate is due to larger amounts of methane and smaller amounts of hydrogen available per unit of catalyst.This also explains the shorter lifetime and lower carbon yield of the catalyst at higher SV, Fig. 13.

Characterization of the products
BET measurement results are presented in Table 1 and reveal a sharp decrease in the specific surface area and pore volume compared with the initial catalyst, suggesting the pores are filled up with carbon.The material density of the produced carbon including the catalyst is much lower than the fresh catalyst.As a result, since carbon is less dense than the catalyst material containing nickel the bulk density is also reduced by about 32%.XPS analyses of the deactivated catalyst showed that all the nickel was in the metallic form and covered by carbon graphene layers.It suggests that deactivation is happening due to encapsulation of the active sites which makes them inaccessible for the methane molecules.It was found that even in the case of less pure gases (99.5%) no byproducts (either solid or gas) were formed.
The fraction of carbon that is produced in the form of desired graphene layer structures can be determined by evaluating the Derivative ThermoGravimetric (DTG) curve of oxidation temperature.The DTG is defined as the derivative of a TGA curve of the corresponding oxidation temperature.The highest temperature limit for oxidation of amorphous carbon reported in literature is 410°C (Luxembourg et al., 2005).However, Hu et al. (2003) and Li and Zhang (2005) reported 365°C and 350°C respectively for the oxidation of amorphous carbon and the carbon that is oxidized in temperatures above these limits is generally accepted to be nanostructured carbon.Fig. 14 illustrates the TGA and DTG curve of the carbon produced from a feed of 100% CH 4 at 550 °C.Even with considering 410 °C as the limit temperature of oxidation of amorphous carbon, the lowest purity of the different tested samples was about 96% nano-structured carbon.
Fig. 15 shows a TEM image of a cluster of produced carbon nanofibers, CNFs, in 100% CH 4 at 550 °C.The diameters are in the range of 15-80nm.CNFs produced at the different operating conditions were in the form of stacked cones, see Fig. 16.Stacked cones are also called a fish bone structure and were obtained also in the literature on the nickel catalyst supported by silica (Toebes et al., 2002;Lehman et al., 2011).

Initial reaction rate
The lifetime of the catalyst for some of the conditions at 650 °C is so short that the maximum reaction rate could not be measured reliably.At 500 °C the reaction was limited by thermodynamic equilibrium in some of the operating conditions.Therefore, only the experimental data obtained at 550 °C, 575 °C and 600 °C were used to find the exact kinetic parameters by the model described in Section 3.1.Table 2 presents the kinetic parameters of the TCD reaction rate, Eq. 3. The average and maximum relative error of Eq. 3, using parameters from Table 2, do not exceed 11% and 22% respectively.

Deactivation factor
The deactivation factor was fitted to the experimental data at 550 °C, 575 °C and 600 °C and Table 3 shows the obtained values.The total carbon yield obtained from each experiment and the model were compared.The average and maximum relative difference were 13.0% and 28.7% at the highest.Figs.17,18 illustrate three examples of the performance of the catalyst over time in the experiments and predicted by the model at different operating   conditions.Segregation of the reactor as it is described in Section 3.2 is clearly visible by drops of the deactivation factor.

Conclusions and outlook
In this study, the thermocatalytic decomposition of methane in a fluidized bed reactor was studied and the corresponding reaction kinetics were established.A commercial nickel on silica catalyst was used in this study and carbon yields of up to and in excess of 70g C /g cat were obtained.The carbon produced was mainly in the form of carbon nano fibers.Its purity was characterized by TEM, TGA and XPS tests.The produced carbon had at least 96% purity of fish bone structures.
The effect of operating conditions has been investigated and it was found that at lower temperature, a larger amount of carbon was produced despite that the maximum reaction rate was lower.This was due to the delayed deactivation of the catalyst at lower temperature.Lowering the concentration of methane lowered the maximum reaction rate, lifetime and carbon yield.Our study has also revealed that although the presence of hydrogen decreases the maximum reaction rate, a higher carbon yield is achieved due to longer lifetime of the catalyst.
A kinetic model describing the maximum reaction rate and the deactivation factor was developed.This model describes the reaction rate of TCD as a function of time in the temperature range of 550-600 °C with a reasonable accuracy and averaged error in initial kinetic rate of 10% and deactivation factors up to 17 %.This model together with the corresponding commercially available catalyst can be used for further study of TCD and reproduction of data.This kinetic model provides the basis to simulate a fluidized bed reactor for TCD with more detailed (i.e.CFD-based) models to facilitate reactor development and optimization.
A very high SV was chosen in this study to facilitate the mobility of catalysts that have grown due to large depositions of carbon, and to prevent breaking the reactor's wall.As a result, the conversion of    the gas was limited.In an industrial scale, this can be overcome with a proper continuous reactor design and partial recycle of the gas stream.

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.

Fig. 1 .
Fig. 1.The fluidized bed reactor with a spherical free-board section.

Fig. 3 .
Fig. 3.The schematic of (a) the initial plug flow reactor model, (b) the reactor model when only the bottom section is segregated, (c) the reactor model when it is completely segregated.

Fig. 5 .
Fig. 5.The effect of temperature on the carbon yield (top) and lifetime (bottom).

Fig. 6 .
Fig. 6.The effect of concentration of methane (diluted by nitrogen or argon) on the maximum reaction rate of the reaction.

Fig. 7 .
Fig. 7.The effect of concentration of methane (diluted by nitrogen or argon) on the carbon yield (top) and lifetime (bottom).

Fig. 8 .
Fig.8.The effect of concentration of hydrogen on the maximum reaction rate of the reaction at 600 °C.

Fig. 9 .
Fig. 9.The effect of concentration of hydrogen on the carbon yield (top) and lifetime (bottom) at 600 °C.

Fig. 11 .
Fig. 11.The effect of size of the catalyst particle on the carbon yield (top) and lifetime (bottom) in 100vol.%CH 4 and 600 °C.

Fig. 12 .
Fig.12.The effect of SV on the maximum reaction rate of the reaction.

Fig. 13 .
Fig. 13.The effect of SV on the carbon yield (top) and lifetime (bottom).

Fig. 14 .
Fig. 14.Thermogravimetric and derivative thermogravimetric of product obtained in a feed of 100% CH 4 at 550 °C.

Fig. 16 .
Fig. 16.TEM image of Two of the produced CNFs in 100% CH 4 at 500 °C including an active site covered by carbon.

Fig. 17 .
Fig. 17.Comparison of the deactivation factor of experiments and the model in a feed of 50% CH 4 at 550 °C.

Fig. 18 .
Fig. 18.Comparison of the deactivation factor of experiments and the model in a feed of 50% CH 4 and 10% H 2 at 600 °C.
DH CH 4 ; DH H 2 ; DH d;C ; DH d;CH 4 ; DH d;H 2 Heat of adsorption of components kJ/mol duce the reactor model used for the interpretation of the experiments, whereas the results and discussion is given in Section 4. Finally in Section 5 the conclusions are presented.

Table 1
Characteristics of the fresh catalyst and one sample of the used catalyst.
* at 550 °C and SV of 4.5L n / g cat / min of 100% CH 4

Table 3
Parameters of Eq. 4

Table 4
Parameters used to calculate CWP