Upper Flammability Limits of Hydrogen and Light Hydrocarbons in Air

Article pubs.acs.org/IECR

Upper Flammability Limits of Hydrogen and Light Hydrocarbons in Air at Subatmospheric Pressures Hai Le, Subramanya Nayak, and M. Sam Mannan* Mary Kay O’Connor Process Safety Center, Artie McFerrin Department of Chemical Engineering, Texas A&M University System, College Station, Texas 77843-3122, United States ABSTRACT: The upper flammability limits (UFL) of hydrogen−air, methane−air, ethane−air, n-butane−air, and ethylene−air were determined experimentally at room temperature (20 °C) and initial pressure of 1.0, 0.7, 0.5, 0.3, 0.1, and 0.05 atm. Experiments were conducted in a closed cylindrical stainless steel vessel (i.d. 10.22 cm, length 100 cm) with upward flame propagation. The UFL of hydrogen was observed to be inversely proportional to the initial pressure in the range from 1.0 to 0.3 atm and proportional to the initial pressure from 0.3 to 0.05 atm. In contrast, the UFLs of the lower alkanes and ethylene decreased with the initial pressure. The average flame propagation velocities at UFL concentrations of hydrogen, methane, ethane, n-butane, and ethylene in air at reduced pressures were also examined. It was found that the flame propagation velocity of hydrogen was larger than those of the hydrocarbons, increased when the initial pressure decreased from 1.0 to 0.3 atm, and then decreased with further decrease of pressure. Flame propagation velocities at UFL concentrations of the hydrocarbons decreased with the initial pressure. Finally, based on the behavior of the UFLs and flame propagation velocities, the relative risk and hazards of ignition and flame escalation of hydrogen and the light hydrocarbons at subatmospheric pressures were discussed.

1. INTRODUCTION Hydrogen is produced and used in various industrial processes ranging from the oil and gas industry to food manufacturing.1 Hydrogen is also increasingly explored as a promising alternative to traditional fossil fuels primarily due to its environmental benefits. Except for some levels of NOx, combustion of hydrogen emits no toxic substances and pollutants such as CO, CO2, SOx, or soot.2 In addition, hydrogen energy is regarded as renewable and abundant since hydrogen can be produced from water. The amount of hydrogen generated worldwide is estimated in the magnitude of a million tons a year and worth billions of U.S. dollars annually.1 The high volume of hydrogen produced and increasing presence of hydrogen in industry require that hazards and risk associated with hydrogen be carefully assessed and prevented. It is recommended that great precautions should be taken when handling hydrogen whether in its pure state or in mixtures with other chemicals since hydrogen poses a unique risk of fire and explosion due to its high probability of ignition compared to most hydrocarbons.3 For example, the flammable range of hydrogen in air at atmospheric condition is 4−75%,4 which is much wider than those of most hydrocarbon fuels. The minimum ignition energy of hydrogen in air is extremely low, 0.018 mJ, compared to those of common fuels,3 and the deflagration index of hydrogen is about 10 times the value of methane and 5 times that of gasoline at atmospheric condition.3,5 The ARIA (Analysis, Research, and Information on Accidents) database showed that approximately 84% of the studied accidents involving hydrogen resulted in fires and/or explosion, and the consequences of the accidents were tremendously serious with 25% of the cases resulting in death or serious injuries.5 Therefore, knowledge of the flammability characteristics, especially flammability limits, of hydrogen is © 2012 American Chemical Society

extremely important in preventing accidents when handling, using, storing, and transporting hydrogen. Reference to the literature shows that there is limited data on hydrogen flammability limits, especially upper flammability limits, at nonatmospheric conditions. Specifically, when the influence of pressure was studied, there was an apparent tendency to examine the flammability limit at high pressures while subatmospheric pressure condition was almost uninvestigated. For example, data reported by Coward and Jones4 shows that the flammability limit of hydrogen, both the upper flammability limit (UFL) and the lower flammability limit (LFL), first narrowed when the initial pressure increased up to 20 atm and then steadily widened at higher pressures. However, it is unclear how the flammability limit of hydrogen, particularly UFL, behaves at subatmospheric pressures. The increasing presence of hydrogen in various laboratory and industrial processes operating at different conditions, including subatmospheric pressure condition such as vacuum drying and vacuum distillation, also justifies the need to study the flammability limits of hydrogen at reduced pressures.6 In addition to the necessity to investigate the flammability limits of hydrogen at subatmospheric pressures, it is essential to understand the low-pressure flammability limits of the hydrocarbons commonly found in mixtures with hydrogen, such as methane, ethane, n-butane, and ethylene, in order to predict the flammability limits of the mixtures at low pressures.7 However, there is very limited study on the low-pressure flammability limits, particularly UFLs, of these hydrocarbons in Received: Revised: Accepted: Published: 9396

January 30, 2012 May 14, 2012 June 15, 2012 June 15, 2012 dx.doi.org/10.1021/ie300268x | Ind. Eng. Chem. Res. 2012, 51, 9396−9402

Industrial & Engineering Chemistry Research

Article

Figure 1. Experimental setup.

Table 1. Specifications of Chemicals Used in the Experimentsa purity a

H2

CH4

C2H6

C4H10

C2H4

air

99.999%

99.99%

99.995%

99.98%

99.995% (H2O < 1 ppm)

Ultra zero certified (H2O < 2 ppm)

Supplier: Matheson Tri Gas.

is shown schematically in Figure 1 and briefly discussed in the following section. Ultra-high-purity (UHP) grade fuels and UHP-grade air from pressurized cylinders were loaded into the mixing vessel. The specifications of the fuels and air are provided in Table 1. The fuel/air quantities were determined on a partial pressure basis. Great care was taken to make sure that the desired fuel concentration was achieved. The concentration step size, step change made in the fuel concentration between experiments, was 0.05 mol %. The uncertainty associated with the measured UFLs was estimated using the combined standard uncertainty method15 or the law of propagation of uncertainty method,15b which is based on random errors associated with the calibration and measurement of the amount of test substance (fuel) in the test mixture (0.10 mol %) and the uncertainty associated with the sampling interval and concentration step size. The obtained maximum uncertainty on the upper flammability limit was 0.15 mol %. The mixing vessel was a stainless steel cylinder which contained a cylindrical Teflon block that glided along the length of the vessel when the vessel was rotated lengthwise. Gases moving between the Teflon block and the vessel wall created highly turbulent zones in front of and behind the moving block; these zones facilitated fast and complete mixing of the gas components. For each mixture, the mixing vessel was rotated for 5 min, approximately 300 inversions. The test mixture was then permitted to flow into the test vessel, which was initially evacuated. The design of the test vessel was similar to that used by the U.S. Bureau of Mines and the more recent European standard EN 1839 (T). The test vessel was a stainless steel (SS 316) closed cylinder 11.43 cm o.d., 10.22 cm i.d., 100 cm long with a clean and smooth inner surface. The pressure inside the test vessel was monitored by a dynamic pressure transducer (Omega DPX 101, 0−250 psig pressure rise range) mounted on the top. The temperature change inside the vessel was

the literature; for example, there is almost no data about the UFLs of n-butane and ethane at reduced pressures.4 In order to address the aforementioned limitations and further enhance the understanding of the UFL of hydrogen, as well as those of the above hydrocarbons at subatmospheric pressures, we performed UFL measurements in a closed cylindrical steel vessel with upward flame propagation. Experiments were carried out at room temperature (20 °C), and the initial pressure ranged from 1.0 atm to as low as 0.05 atm. The behavior of the UFL of hydrogen at low pressures was analyzed and compared with those of the hydrocarbons. Besides the flammability limit, the flame propagation velocity, or spatial velocity of the flame front,8 is an important indicator of the hazard and risk of fire/explosion of a combustible mixture.9 The flame propagation velocity is also used to calculate the fundamental laminar burning velocity,8−10 which represents the reactivity and exothermicity of a combustible mixture,11 and employed in numerical modeling studies to predict and simulate the dynamics of combustion of mixtures.9,11a While there are studies about the flame propagation velocities of hydrogen and hydrocarbons at atmospheric pressure,10b,11,12 there is almost no data on their velocities at UFL concentration and subatmospheric pressures available in the literature. Therefore, to enhance the understanding of the flame propagation velocity and the hazard/risk of hydrogen and the hydrocarbons, particularly at UFL concentration and subatmospheric pressure, the average flame propagation velocities at the UFL concentrations of hydrogen, methane, ethane, n-butane, and ethylene in air at subatmospheric pressures were also determined and discussed.

2. EXPERIMENTAL SETUP 2.1. Test Apparatus and Procedure. The methodology and instruments used in this study were similar to those reported by Wong13 and Zhao et al.14 The experimental setup 9397

dx.doi.org/10.1021/ie300268x | Ind. Eng. Chem. Res. 2012, 51, 9396−9402


Article

Table 2. Upper Flammability Limit of Hydrogen in Air at Atmospheric Condition

a

fuel

this work (mol %)

previous work (mol %)

apparatus type

FL criteria

H2

75.73 ± 0.15

75.00a 74.70a 75.10a 75.80 ± 0.20 76.60 ± 0.20

vertical glass tube (i.d. 7.5 cm, L 150 cm)4 vertical stainless steel tube (i.d. 5 cm, L 100 cm)18 glass flask, V = 5 dm3, ASTM E681-0117 vertical glass tube (i.d. 6 cm, L 30 cm), DIN 51649−117 glass cylinder (i.d. 8 cm, L 30 cm), EN 1839 (T)17

visual thermal visual visual visual

No information about the uncertainty range.

detected by five thermistors positioned at the center of the vessel. The location of each thermistor relative to each other and to the test vessel is shown in Figure 1. The thermistors were fast-response thermistors which could detect flame front in real time and locate self-sustained flame propagation distance when the test mixture was ignited from the bottom and burnt upwardly along the test vessel. The igniter system is similar to that described by ASTM E 918-83, which is capable of providing 10 J of energy with a consistent power delivery. The ignition source is a 1 cm piece of AWG 40 tinned copper wire, vaporized by a 500 VA isolation transformer at 115 V AC switched on with a zero-crossing solid state relay. Signals from the pressure transducer and five thermistors were obtained by a Keithley data acquisition card (Keithley KPCI-3102, 225 signals/s at 0.05% accuracy) installed in a desktop computer, and the experimental process was controlled by a LabView program. After each experiment, the test vessel was purged with nitrogen to remove all combustion products. After purging, the vessel was evacuated and flushed with the test mixture to as many as 3 evacuation−flush cycles to ensure that there were no impurities left. Then the vessel was evacuated to a pressure as low as 0.02 psia (0.001 atm) to be ready for the new experiment. The initial temperature and pressure of the test mixture inside the test vessel were strictly controlled. Experiments were performed at room temperature (20 °C) and initial pressures of 1.0, 0.7, 0.5, 0.3, 0.1, and 0.05 atm. 2.2. Flammability Limit Determination and Data Analysis. Wong13 and Zhao et al.14 reported that there were five combustion types in the test vessel depending on the fuel concentration of the test mixture: nonpropagation, flash combustion, discontinuous flame propagation, temperately continuous flame propagation, and continuous flame propagation. A mixture was considered flammable if it displayed continuous flame propagation behavior in which upon ignition a flame was formed and able to propagate to the top thermistor. The flame propagation for mixtures near the flammability limits has a probabilistic nature16 due to random errors in composition, fluctuations in mixture conditions (turbulence, pressure, temperature), and variation in ignition energy or power;13 thus, multiple experiments with mixtures with the same composition may yield different results. In this work, the upper flammability limits were generated using the same method described by Wong13 and Zhao et al.14 in which 10 repetitive experiments were performed at each fuel concentration, and the flammability limit was the concentration which has the probability of continuous flame propagation equal to or less than 50% compared to the nearest concentration having more than 50% of flame propagation.13

obtained UFL of hydrogen was similar to those generated with apparatuses whose configurations were consistent with that developed by the U.S. Bureau of Mines and the apparatuses established by European standards DIN 51649 and EN 1839 (T) (Table 2). The higher values of the upper flammability limits reported by the European methods DIN 51649 and EN 1839 (T) can be explained by their measurement criteria and flammability limit definition. According to DIN 51649 and EN 1839 (T), a mixture is considered flammable if upon ignition the resulting flame can propagate a distance of at least 10 cm from the electrodes,17 whereas our method and the similar to that of the Bureau of Mines require a flame propagation distance of at least 75 cm to be considered flammable. DIN 51649 and EN 1839 (T) define the flammability limit as the concentration where the last nonpropagation point occurs, while our method marks the flammability as the concentration which lies between the nonpropagation and the propagation points. 3.2. Upper Flammability Limit of Hydrogen at Subatmospheric Pressures. The UFLs of hydrogen at subatmospheric pressures are presented in Table 3 and Table 3. UFL of Hydrogen in Air at Subatmospheric Pressures and Room Temperature initial pressure (atm) UFLa (mol %) a

1.00

0.70

75.73

75.88

0.50 77.3

0.30 77.8

0.10

0.05

76.95

73.92

Uncertainty: 0.15 mol %.

illustrated in Figure 2. When the initial pressure decreased below atmospheric pressure, the UFL of hydrogen initially increased; the UFL kept on increasing until the initial pressure was lowered to about 0.3 atm; then the UFL started to decrease; the UFL was still larger than the value at atmospheric

3. RESULTS AND DISCUSSION 3.1. Upper Flammability Limit of Hydrogen at Atmospheric Pressure. At atmospheric pressure, the

Figure 2. Upper flammability limit of hydrogen in air at subatmospheric pressures and room temperature. 9398



Article

pressure until the initial pressure was reduced to 0.05 atm (Figure 2). This means for hydrogen, a mixture that cannot propagate flame at atmospheric pressure may be able to do so at subatmospheric pressures; in other words, hydrogen poses a higher risk of ignition at reduced pressures. However, the risk is not significantly higher since the increase was relatively small, within 2.1% of the value at atmospheric condition (Figure 5). Previous UFL studies19 with hydrogen at initial pressures larger than 1.0 atm showed that initially the UFL decreased when the initial pressure increased; however, until certain pressure, the trend changed where the UFL started to steadily increase. This behavior of hydrogen UFL at high pressures coupled with that at subatmospheric pressures described in this study can be explained by the combustion reaction mechanism of hydrogen in air. In general, combustion reaction involves chain reactions consisting of multiple reaction steps and free radicals.8 Although combustion of hydrogen in air involves only two elements, H and O, the chemical reaction mechanism is quite complex with more than 50 elementary reactions including the initiation, chain propagating, chain branching, and chain termination steps.20 When the initial pressure increases, the amount of reactants and concentration of free radicals increase. This higher density of reactants and free radicals leads to an increase of the overall reaction rate which results in promotion of the combustion process and widening of the flammability range. Thus, we found an increase of the hydrogen UFL when the initial pressure was raised from 0.05 to 0.3 atm (Figure 2). However, when the pressure was further increased, the hydrogen UFL started to decrease, which may be explained by the involvement of a three-body reaction8,20 H + O2 + M → HO2 + M

Table 4. UFLs of the Hydrocarbons in Air at Subatmospheric Pressures and Room Temperature UFLa (mol %)

a

initial pressure (atm)

methane

ethane

n-butane

ethylene

1.0 0.7 0.5 0.3 0.1

15.40 14.85 14.65 14.50 14.35

14.00 13.64 12.86 12.37 11.76

8.46 8.33 8.18 8.10 8.08

30.61 29.49 27.50 23.39 19.26

Uncertainty: 0.15 mol %.

Figure 3. Upper flammability limits of methane−air, ethane−air, and n-butane−air at subatmospheric pressures and room temperature.

(1)

M can be any third molecule which acts as a stabilizer for the combination of H and O2. The relatively unstable hydroperoxy molecule HO2 diffuses to the wall and is consumed there by the following reactions8 HO2 →

1 H 2 + O2 2

(2)

HO2 →

3 1 H 2O + O2 2 4

(3)

Figure 4. Upper flammability limit of ethylene−air at subatmospheric pressures and room temperature.

Reaction 1 was considered the dominant chain termination reaction for hydrogen at the UFL concentration.21 As the initial pressure increases, the probability of the three-body collision in reaction 1 increases, which results in a decreased rate of overall reaction as well as narrowing of the flammability range.21a Therefore, we found that the UFL decreased when the initial pressure was raised from 0.3 atm (Figure 2) to as much as 5 atm.19 However, when the initial pressure increases further, HO2 can react with H2 to form H2O2, H, and OH radicals, which enhance the chain-branching step,8,22 thus increasing the overall reaction rate and widening the flammability zone. Therefore, the UFL started to increase when the initial pressure was further raised to higher than 5 atm.19 3.3. Upper Flammability Limits of Methane, Ethane, n-Butane, and Ethylene in Air at Subatmospheric Pressures. The UFLs of the hydrocarbons in air at subatmospheric pressures are presented in Table 4 and illustrated in Figures 3 and 4. In contrast to the behavior of the UFL of hydrogen, the UFL of methane, ethane, n-butane, and ethylene decreased when the initial pressure decreased below atmospheric pressure. This means the above hydro-

carbons pose a lower risk of fire and explosion when the initial pressure is reduced. In addition, the decrease of the UFLs of the hydrocarbons is larger than the increase of the UFL of hydrogen at reduced pressures (Figure 5), which suggests that these hydrocarbons present a greater reduction of fire risk compared with the increasing risk of hydrogen at subatmospheric pressures. The decrease in the UFLs of the hydrocarbons at reduced pressures is expected since previous studies also showed similar results. For example, Mason and Wheeler23 observed a decrease in the UFL of methane at pressures less than 1.0 atm in a tube (2 cm diameter, 50 cm length) with downward propagation. Another study with methane in a tube (5 cm diameter, 50 cm length) with upward, horizontal, and downward propagation also found a decrease in UFL at low pressures.4 In other words, the UFLs of the hydrocarbons increased with pressure; similar behavior was observed at elevated pressures in published studies.4,19b,24 There was not a decrease in the UFLs of the 9399



Article

Figure 5. Percentage deviation of UFLs of hydrogen−air, methane− air, n-butane−air, and ethylene−air at subatmospheric pressures from the UFLs at 1.0 atm.

Figure 6. UFLs of methane−air, ethane−air, and n-butane−air expressed as equivalence ratios at subatmospheric pressures and room temperature.

hydrocarbons when the pressure increased as observed with the UFL of hydrogen. One possible reason is that the dominant termination reactions of the hydrocarbons at UFL concentration are two-body reactions instead of the three-body reaction in eq 1 as concluded by Law and Egolfopoulos.21a Two-body termination reactions exert no advantage over the chain branching reactions when the pressure increases as opposed to the three-body reaction in eq 1 which becomes more efficient with increasing pressures (as observed with hydrogen). For the lower alkanes (methane, ethane, n-butane), the UFL decreased linearly with pressure (Figure 3). At initial pressures higher than atmospheric pressure, the UFLs of these lower alkanes increased also linearly as observed by some previous studies.19b,24 For example, an experimental study with methane using a spherical vessel (7.6 cm diameter) with central ignition at room temperature and initial pressures higher than 1.0 atm found that the UFL of methane increased rapidly and linearly with pressure.19b Van den Schoor and Verplaetsen,24 who experimented on ethane and n-butane with a spherical vessel (20 cm diameter) and central ignition, discovered that the UFLs of these hydrocarbons increased linearly when the initial pressure was raised up to 20 (for ethane) and 10 bar (for nbutane). Another study4 with ethane and n-butane at elevated initial pressures using a closed tube (20 cm diameter, 40 cm length) also showed a linear increase of the UFLs. It was found in this study that that when the UFLs of the lower alkanes were expressed as equivalence ratios (the actual fuel/air ratio divided by the stoichiometric fuel/air ratio), the higher the carbon number in the series of the observed alkanes, the higher the UFLs (Figure 6). A similar observation was also made by Van den Schoor and Verplaetsen24 at elevated pressures (up to 55 bar) and elevated temperature (200 °C). The decrease in the UFL of ethylene at subatmospheric pressures had two distinct features. First, the decrease was not linear but more like a logarithmic decrease with pressure (Figure 4). The increase in the UFL of ethylene at elevated pressures was also nonlinear as determined by Craven and Foster25 and Hashiguchi et al.26 Second, the UFL change at reduced pressures is much greater for ethylene compared with those of hydrogen and the lower alkanes, particularly when the initial pressure was reduced below 0.5 atm (Figure 5). A larger degree of change in the UFL of ethylene at elevated pressures compared with those of the lower alkanes was also observed by several researchers;19a,24−26 for example, Berl and Werner

showed that the UFL of ethylene in air increased sharply from 16% at 1.0 atm to 68% at 90 atm.19a The pronounced effect of pressure on the UFL of ethylene can be attributed to the sensitivity of the combustion reaction mechanism of ethylene to changes in the pressure. It was shown by Carriere et al.27 that the dominant ethylene consumption pathway and route to the final oxidation products of the combustion of ethylene changed greatly with pressure. For example, when the pressure increased, destruction of ethylene changed from an abstraction reaction forming C2H3 to an addition reaction forming C2H5; consequently, the pathway to formation of final products via oxygenated species appeared and became more important.27 Some reactions on this pathway were pressure dependent in a way that an increase in pressure further enhanced the rates of these reactions,27 which promoted flame propagation and resulted in a large increase of the UFL as observed in this study and previous research.19a,24−26 3.4. Average Flame Propagation Velocities at UFL Concentrations of Hydrogen, Methane, Ethane, and nButane in Air at Subatmospheric Pressures. The average flame propagation velocity (the actual flame speed) was determined by dividing the distance between the bottom and the top thermistors (60 cm) by the response time difference between these two thermistors. In this study, the average flame propagation velocities were measured at the UFL concentrations of the fuels at room temperature and initial pressures ranging from 1.0 to 0.05 atm as shown in Figure 7. These flame propagation velocities can be used to calculate fundamental laminar burning velocities,8,10a,28 which are employed in numerical modeling studies to predict and simulate the dynamics of combustion at subatmospheric pressures9,11a such as estimation of flame stability and acceleration, calculation of combustion reaction rate and flame thickness, validation of combustion reaction mechanism. Details about calculation of the burning velocity from the flame propagation velocity depend on the geometry of the flame front, the pressure change inside the reaction vessel, and the coefficient of expansion of the combustion products, which will be the subject of future studies. This study aims at (i) examining the actual flame speed of hydrogen compared with those of the lower alkanes and ethylene at initial pressures smaller than atmospheric pressure and (ii) evaluating the relative risk and hazard of the flame escalation of hydrogen in air and the hydrocarbons in air at reduced pressures based on the actual flame propagation velocity. 9400



Article

Table 5. Calculated Equilibrium Flame Temperatures Using the CHEMKIN Package31 at the UFL Concentrations of Methane, Ethane, n-Butane, and Ethylene flame temperature (K) pressure (atm)

CH4

C2H6

C4H10

C2H4

1.0 0.7 0.5 0.3 0.1

2137 2193 2213 2226 2232

1528 1580 1694 1766 1856

1439 1470 1505 1524 1529

1550 1563 1593 1663 1745

flame propagation velocity of ethylene did not change much at reduced pressures and was not so different from the velocities of the alkanes.

Figure 7. Average flame propagation velocities at UFL concentrations of hydrogen, methane, ethane, n-butane, and ethylene in air at subatmospheric pressures and room temperature.

4. CONCLUSION The upper flammability limits (UFL) and average flame propagation velocities at UFL concentrations of hydrogen, methane, ethane, and n-butane in air were measured at room temperature (20 °C) and initial pressures ranging from 1.0 to 0.05 atm. The following results were found. Initially the UFL of hydrogen increased when the pressure decreased. When the initial pressure was further reduced to 0.3 atm, the UFL started to decrease but was still larger than the value at 1.0 atm until the initial pressure was reduced to 0.05 atm. The maximum increase of the hydrogen UFL at subatmospheric pressure was 2.1% higher than the UFL at atmospheric pressure. The increased UFL at initial pressure less than 1.0 atm suggests that hydrogen poses a higher risk of ignition at subatmospheric pressure condition. In contrast to the behavior of the UFL of hydrogen, the UFLs of methane, ethane, n-butane, and ethylene decreased when the initial pressure decreased. The UFL of the lower alkanes decreased linearly with pressure, while the UFL of ethylene decreased logarithmically. The change of the UFL of ethylene at subatmospheric pressures was larger than those of hydrogen and the lower alkanes, particularly at a low-pressure regime (less than 0.5 atm), which indicates that pressure has a greater impact on the UFL of ethylene than on those of hydrogen and the lower alkanes at reduced pressures. The average flame propagation velocity at the UFL concentration of hydrogen in air increased when the initial pressure decreased below 1.0 atm. The maximum increase of the velocity was at 0.3 atm, which was 12% of the velocity at atmospheric pressure. The increased flame propagation velocity of hydrogen at subatmospheric pressures suggests that the burning of hydrogen poses a higher risk of flame propagation and escalation at low pressure. The flame propagation velocities at the UFL concentrations of methane, ethane, n-butane, and ethylene in air decreased with initial pressure. Overall, the flame propagation velocities at the UFL concentrations of the hydrocarbons were smaller than that of hydrogen. Among the studied hydrocarbons, at UFL concentration, methane had the largest flame propagation velocity at all initial pressures. The flame propagation velocity of ethylene did not change much at reduced pressures and was not so different from the velocities of the alkanes.

Overall, at UFL concentration, the flame propagation velocity of hydrogen was higher than those of the hydrocarbons at all initial pressures studied. This is expected as hydrogen is known to have larger flame velocity in air compared with most hydrocarbons including those in this study.2a,29 This is due to the fact that combustion of hydrogen in air involves much lighter free radicals which diffuse more easily to the reaction zone, thus enhancing the flame propagation velocity. The high flame propagation velocity of hydrogen poses a high risk of fire escalation to explosion, especially in confined space. Therefore, hydrogen should be handled or processed in well-ventilated areas with a reliable gas detection system, and storage of hydrogen should be outdoors as much as possible. When the initial pressure decreased, the flame propagation velocity of hydrogen in air increased (Figure 7). When the initial pressure was about 0.3 atm, the velocity started to decrease but the velocity was still larger than the value at atmospheric pressure. This behavior is similar to that of the UFL of hydrogen at subatmospheric pressures. The maximum increase of the velocity at subatmospheric pressure was 12% of the velocity at atmospheric pressure. This suggests that the burning of hydrogen at vacuum condition poses a higher risk of flame propagation and escalation; thus, more care should be taken when handling hydrogen at subatmospheric pressure conditions. In contrast to the behavior of the flame propagation velocity of hydrogen, at UFL concentration, the actual flame speeds of the studied hydrocarbons decreased with the initial pressure (Figure 7), especially for n-butane and methane whose velocities decreased more rapidly with the maximum decrease of 31.0% and 21.5% accordingly. This indicates that methane, ethane, n-butane, and ethylene pose a lower risk of flame propagation and escalation at reduced pressures. Among the hydrocarbons, at UFL concentration, methane had the largest flame propagation velocity at all initial pressures. This may be due to methane’s higher flame temperature compared to those of the other hydrocarbons at the UFL concentrations. Higher flame temperature promotes dissociation reactions which introduce more free radicals into the flame, thus enhancing the overall combustion reaction rate and increasing the flame propagation velocity.8,30 As shown in Table 5, at UFL concentration, methane burns at a much higher flame temperature than ethane, n-butane, and ethylene do; this explains the larger observed flame propagation velocity of methane compared to those of the other hydrocarbons. The

■

AUTHOR INFORMATION

Corresponding Author

*Tel.: +1 979-862-3985. Fax: +1 979-845-6446. E-mail: [email protected]. 9401



Article

Notes

(16) Wierzba, I.; Karim, G. A.; Cheng, H. The Flammability of Rich Gaseous Fuel Mixtures Including Those Containing Propane in Air. J. Hazard. Mater. 1988, 20, 303−312. (17) Schr̈oder, V.; Molnarne, M. Flammability of gas mixtures Part 1: Fire potential. J. Hazard. Mater. 2005, 121 (1−3), 37−44. (18) Wierzba, I.; Wang, Q. The flammability limits of H2-CO-CH4 mixtures in air at elevated temperatures. Int. J. Hydrogen Energy 2006, 31 (4), 485−489. (19) (a) Berl, E.; Werner, G. The limit of combustibility of flammable gas and vapor air mixtures at high pressure. Z. Angew. Chem. 1927, 40, 245−250. (b) Bone, W. A.; Newitt, D. M.; Smith, C. M. Gaseous combustion at high pressures IX. The influence of pressure upon the explosion limits of inflammable gas-air, etc. mixtures. Proc. R. Soc. 1928, 117, 553−576. (c) Terres, E.; Plenz, F. Influence of pressure on the burning of explosive gas mixtures. J. Gasbeleucht. Verw. Beleuchtungsarten Wasserversorg. 1914, 57, 990−995, 1001−1007, 1016−1019, 1025−1027. (20) Dougherty, E. P.; Rabitz, H. Computational Kinetics and Sensitivity Analysis of Hydrogen-Oxygen Combustion. J. Chem. Phys. 1980, 72 (12), 65−71. (21) (a) Law, C. K.; Egolfopoulos, F. N. Kinetic criterion of flammability limits: The C-H-O-Inert system. Twenty-Third Symposium (International) on Combustion; The Combustion Institute: Pittsburgh, PA, 1990; pp 413−421. (b) Shebeko, Y. N.; Tsarichenko, S. G.; Korolchenko, Y. A.; Trunev, A. V.; Navzenya, Y. V.; Papkov, S. N.; Zaitzev, A. A. Burning velocities and flammability limits of gaseous mixtures at elevated temperatures and pressures. Combust. Flame 1995, 102 (4), 427−437. (22) Sohn, C. H.; Chung, S. H. Effect of Pressure on the Extinction, Acoustic Pressure Response, and NO Formation in Diluted Hydrogen−Air Diffusion Flames. Combust. Flame 2000, 121 (1−2), 288− 300. (23) Mason, W.; Wheeler, R. V. Effect of Temperature and of Pressure on the Limits of Inflammability of Mixtures of Methane in Air. J. Chem. Soc. 1918, 113, 45−57. (24) Van den Schoor, F.; Verplaetsen, F. The upper explosion limit of lower alkanes and alkenes in air at elevated pressures and temperatures. J. Hazard. Mater. 2006, 128 (1), 1−9. (25) Craven, A. D.; Foster, M. G. The limits of flammability of ethylene in oxygen, air and air −nitrogen mixtures at elevated temperatures and pressures. Combust. Flame 1966, 10 (2), 95−100. (26) Hashiguchi, Y.; Ogahara, T.; Iwasaka, M.; Ozawa, K. Effect of pressure on the detonation limit of ethylene. Int. Chem. Eng. 1966, 6, 737−743. (27) Carriere, T.; Westmoreland, P. R.; Kazakov, A.; Stein, Y. S.; Dryer, F. L. Modeling ethylene combustion from low to high pressure. Proc. Combust. Inst. 2002, 29 (1), 1257−1266. (28) Van den Schoor, F.; Verplaetsen, F.; Berghmans, J. Calculation of the upper flammability limit of methane/hydrogen/air mixtures at elevated pressures and temperatures. J. Hazard. Mater. 2008, 153 (4), 1301−1307. (29) Jo, Y.-D.; Crowl, D. A. Explosion characteristics of hydrogen −air mixture in a spherical vessels. Process Safety Progress 2010, 29 (3), 216−223. (30) Sivashinsky, G. I. Instabilities, pattern formation, and turbulence in flames. Ann. Rev. Fluid Mech. 1983, 15 (1), 79−99. (31) (a) Kee, R. J.; Dixon-Lewis, G.; Warnatz, J.; Coltrin, M. E.; Miller, J. A. A FORTRAN Computer Code Package for the Evaluation of Gas-phase, Multicomponent Transport Properties; SAND86-8246; Sandia National Labs: Albuquerque, NM, 1992; (b) Kee, R. J.; Rupley, F. M.; Miller, J. A. The CHEMKIN Thermodynamic Data Base; SAND878215B; Sandia National Labs: Albuquerque, NM, 1992; (c) Smith, G. P.; Golden, D. M.; Frenklach, M.; Moriarty, N. W.; Eiteneer, B.; Goldenberg, M.; Bowman, C. T.; Hanson, R. K.; Song, S.; William C. Gardiner, J.; Lissianski, V. V.; Qin, Z. http://www.me.berkeley.edu/ gri_mech/.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was sponsored by the Mary Kay O’Connor Process Safety Center, Artie McFerrin Department of Chemical Engineering, Texas A&M University.

■

REFERENCES

(1) Winter, C. J. Hydrogen energy, Abundant, efficient, clean: A debate over the energy-system-of-change. Int. J. Hydrogen Energy 2009, 34 (14), S1−S52. (2) (a) Ogden, J. M. Hydrogen: The Fuel of the Future? Phys. Today 2002, 55 (4), 69. (b) Dunn-Rankin, D. Lean Combustion: Technology and Control; Academic Press: New York, 2008. (3) Crowl, D. A.; Jo, Y.-D. The hazards and risks of hydrogen. J. Loss Prev. Process Ind. 2007, 20 (2), 158−164. (4) Coward, H. F.; Jones, G. W. Limits of flammability of gases and vapors. Bureau of Mines Bulletin; Bureau of Mines: Washington, DC, 1952; Vol. 506. (5) ARIA Accidentology involving hydrogen: http://www.aria. developpement-durable.gouv.fr/ressources/etude_h2_gb.pdf. (6) Enloe, J. D.; Thomson, J. D. Flammability limits of hydrogenoxygen-nitrogen mixtures at low pressures. Combust. Flame 1966, 10, 393−394. (7) Kaldis, S. P.; Kapantaidakis, G. C.; Sakellaropoulos, G. P. Simulation of multi-component gas separation in a hollow fiber membrane by orthogonal collocation  hydrogen recovery from refinery gases. J. Membr. Sci. 2000, 173, 61−71. (8) Kuo, K. K. Principles of Combustion, 2nd ed.; John Wiley & Sons, Inc.: New York, 2005. (9) Molkov, V. Hydrogen safety research: state-of-the-art. Proceedings of the 5th International Seminar on Fire and Explosion Hazards, Edinburgh, U.K., 2007; Vols. 23−27, pp 28−43. (10) (a) Babkin, V. S.; Kozaehenko, L. S. Study of normal burning velocity in methane-air mixtures at high pressures. Combust., Explosion, Shock Waves 1966, 2 (3), 46−52. (b) Gerstein, M.; Levine, O.; Wong, E. L. Flame Propagation. II. The Determination of Fundamental Burning Velocities of Hydrocarbons by a Revised Tube Method. J. Am. Chem. Soc. 1951, 73 (1), 418−422. (c) Wagner, P.; Dugger, G. L. Flame Propagation. V. Structural Influences on Burning Velocity. Comparison of Measured and Calculated Burning Velocity. J. Am. Chem. Soc. 1955, 77 (1), 227−231. (d) Halter, F.; Chauveau, C.; Djebaïli-Chaumeix, N.; Gökalp, I. Characterization of the effects of pressure and hydrogen concentration on laminar burning velocities of methane−hydrogen−air mixtures. Proc. Combust. Inst. 2005, 30 (1), 201−208. (11) (a) Yu, G.; Law, C. K.; Wu, C. K. Laminar Flame Speeds of Hydrocarbon + Air Mixtures with Hydrogen Addition. Combust. Flame 1986, 63 (3), 339−347. (b) Bosschaart, K. J.; De Goey, L. P. H. The laminar burning velocity of flames propagating in mixtures of hydrocarbons and air measured with the heat flux method. Combust. Flame 2004, 136 (3), 261−269. (12) Tseng, L.-K.; Ismail, M. A.; Faeth, G. M. Laminar Burning Velocities and Markstein Numbers of Hydrocarbon/Air Flames. Combust. Flame 1993, 95 (4), 410−426. (13) Wong, W. K. Measurement of flammability in a closed cylindrical vessel with thermal criterion. Ph.D. Dissertation, 2006. (14) Zhao, F.; Rogers, W. J.; Mannan, M. S. Experimental measurement and numerical analysis of binary hydrocarbon mixture flammability limits. Process Saf. Environ. Prot. 2009, 87 (2), 94−104. (15) (a) Grosshandler, W.; Donnelly, M.; Womeldorf, C. Lean Flammability Limit as a Fundamental Refrigerant Property; DOE/CE/ 23810-98: ARTI MCLR Project No. 660−52401, 1998; (b) Taylor, B. N.; Kuyatt, C. E. Guidelines for Evaluating and Expressing the Uncertain of NISTMeasurement Results; NIST Technical Note 1297, National Institute of Standards and Technology: Gaithersburg, MD, 1994. 9402


Upper Flammability Limits of Hydrogen and Light Hydrocarbons in Air

Recommend Documents