- Email: [email protected]
Experimental investigation on the minimum ignition temperature of hybrid mixtures of dusts and gases or solvents
Journal of Hazardous Materials 301 (2016) 314–326
Contents lists available at ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Experimental investigation on the minimum ignition temperature of hybrid mixtures of dusts and gases or solvents Emmanuel Kwasi Addai ∗ , Dieter Gabel, Ulrich Krause Otto-von-Guericke-University, Institute of Instrumental and Environmental Technology Department of Systems Engineering and Plant Safety, Universitätsplatz 2, 39106 Magdeburg, Germany
h i g h l i g h t s • • • • •
Lower explosion limit of gases decreases upon addition of dust which is itself not ignitable. Minimum ignition temperature of gases decrease upon addition of dust which is itself not ignitable. Minimum ignition temperature of hybrid mixtures are lower than that of single substance. Minimum explosion concentration of dust decreases upon addition of gas which is itself not ignitable. Lower explosion limits of hybrid mixtures are lower than that of single substance.
a r t i c l e
i n f o
Article history: Received 9 July 2015 Received in revised form 30 July 2015 Accepted 2 September 2015 Available online 7 September 2015 Keywords: MIT Hybrid mixture explosion Dust explosion Gas explosion MIT of hybrid mixture explosion
a b s t r a c t Investigations on the minimum ignition temperatures (MIT) of hybrid mixtures of dusts with gases or solvents were performed in the modified Godbert–Greenwald (GG) furnace. Five combustible dusts and six flammable gases (three ideal and three real) were used. The test protocol was according to EN 502812-1 for dust–air mixtures whereas in the case of gases, solvents and hybrid mixtures this standard was used with slight modification. The experimental results demonstrated a significant decrease of the MIT of gas, solvent or dust and an increase in the likelihood of explosion when a small amount of dust, which was either below the minimum explosion concentration or not ignitable by itself, was mixed with gas and vice versa. For example, the MIT of toluene decreased from 540 ◦ C to 455 ◦ C when small amount of lycopodium was added. It was also confirmed that a hybrid mixture explosion is possible even when both dust and vapour or gas concentrations are respectively lower than their minimum explosion concentration (MEC) and lower explosion limit (LEL). Another example is CN4, the MEC of which of 304 g/m3 decreased to 37 g/m3 when propane was added, even though the concentrations of the gas was below its LEL. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Hybrid mixture explosions can occur in a wide range of industries such as mining, agriculture, pharmaceutical and petrochemical industries that involve handling or processing of flammable gases or solvents and combustible dusts. The mixtures consist of two or more combustible substances with different aggregate states. The minimum ignition temperature (MIT) and lower explosion limit (LEL) or minimum explosible concentration (MEC) are critical parameters when conducting hazard assessment for processes involving hybrid mixtures. In the present paper we
∗ Corresponding author. Fax: +49 391 67 11128. E-mail addresses: [email protected], [email protected] (E.K. Addai). http://dx.doi.org/10.1016/j.jhazmat.2015.09.006 0304-3894/© 2015 Elsevier B.V. All rights reserved.
use the term lower explosion limit in accordance with the standard EN 1127-1 [1] for the lowest volume fraction in% of a flammable gas mixed with air, for which flame propagation is still possible. In contrast, for a dust–air mixture the minimum amount of fuel necessary for self-sustained flame propagation is usually given in mass of dust per volume of air. We therefore in the present paper prefer the term minimum explosible concentration which has the unit g/m3 . Hot surfaces capable of igniting dust clouds exist in a number of situations in industry (furnaces, burners and dryers of various kinds). In addition, hot surfaces can be generated accidentally by overheated bearings and other mechanical parts. If an explosible dust cloud is generated in some uncontrolled way in the proximity of a hot surface with a temperature above the actual minimum ignition temperature, a dust explosion can result [2,3].
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
315
Fig. 1. Particle size distribution of dust.
Table 1 Preparatory analysis of the dust used. Toner
CN4
Wood
Lycopodium
Starch
Molecular formula Media diameter (m) Volatile content (% mass) Moisture content (% mass) Heat of combustion (kJ/kg)
C7.17 H7.75 O0.33
C6.70 H1.31 O0.88 N0.02
C4.19 H6.26 O2.71 S0.004
C5.77 H9.59 O1.23 S0.001 N0.08
C3.69 H6.34 O3.06 S0.01
13.4 90.18 0.92 35792
60 17.08 0.23 26630
307 84.38 0.2 16446
31.6 91.06 0.35 28447
29.2 93.77 0.5 15302
Elemental analysis C H O S N
86.05 7.72 5.23 1 0
80.37 1.31 14.01 3.02 0.34
50.33 6.26 43.28 0.13 0
69.26 9.59 19.62 0.38 1.15
44.34 6.34 48.94 0.38 0
Unlike solitary dust, gas or solvent explosions, which have been widely studied in the past decades, data on explosion characteristics of hybrid mixtures are comparatively sparse. The minimum ignition temperature (MIT) is an explosion characteristic which is used for dust–air mixtures as well as for gas–air mixtures or for mixtures of vapours of flammable liquids with air. However, while for dusts the GG furnace or the BAM oven are used to determine the MIT [4], for gases or vapours the MIT is determined in the Erlenmeyer flask [5]. In case of the MIT of hybrid mixtures the Erlenmeyer flask is not suitable because it does not allow the dispersion of dust particles. Consequently, for the determination of the MIT of hybrid mixtures the GG furnace has to be modified in such a way that besides the generation of a dust–air mixture also a flammable gas or vapour can be added to the mixture. The set-up used for this purpose is described in Section 2. Consequently, in the prevention and mitigation of dust explosions, it is important to know the minimum ignition temperature (MIT), minimum explosible dust concentration (MEC) and lower explosion limit (LEL) of dusts in order to take adequate precautions to ensure that hot surface temperature does not reach this value. For pure dusts, the MIT depends on the material properties, such as particle size distribution of the dust. Furthermore, the environments conditions have also to be taken into account, e.g. oxygen availability, environmental temperature and so forth [6–10]. For hybrid mixtures, the conditions become more complex, as the MIT of hybrid mixtures cannot be predicted by simply overlapping the
effects of the single substances. Moreover, there is no systematic study that is able to quantify the roles of dust and gas in driving the explosion. The main conclusions of previous studies [11–20] on the explosion properties of hybrid mixtures could be summarized non-exhaustively by the following assertions, that the ignition sensitivity of the powder can be strongly increased by the addition of a few percent of combustible gases or vapours, even with contents lower than the LEL. It has notably been shown that hybrid mixtures can also be explosible when the concentrations of the dust and the gas are both below their respective MEC and LEL, respectively. However, in the case of minimum ignition temperatures of hybrid mixtures data are nearly not available. Hence, the work of this paper is to investigate on the MIT and LEL/MEC of hybrid mixtures of dust and gas or solvent in vapour state. In order to achieve this, combinations of five dust samples and six gases or vapours were tested. All these tests were performed in the modified Godbert–Greenwald furnace. 2. Materials and experimental work 2.1. Tested materials Five different dusts namely cornstarch, lycopodium, toner powder, wood dust and CN4 as well as each three perfect gases (methane, propane and hydrogen) and three real gases (vapour of ethanol, toluene and isopropanol) were used in the present study.
316
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
Fig. 2. SEM images for various dust with different magnifications.
The particle size distribution is one of the most important properties that affect the minimum ignition temperature of dust clouds. In order to determine the particle size distribution, all the dust samples underwent CAMSIZER probing. Fig. 1 shows the particle size distributions for the dusts used. Furthermore, in Fig. 2 scanning electron microscopy (SEM) images for all the dusts used with different magnification in order to reveal the surface structure of the particles are presented. Table 1 summarizes the data of the elemental analysis, median particle size values, mass content of volatiles, moisture mass content, heat of combustion and molecular formula of the dusts which were calculated from elemental analysis. Table 2 gives the basic, thermodynamic, ignition or combustion properties of the gases used [21–23].
Fig. 3. Image of experimental setup.
2.2. Experimental determination of MIT for dust and hybrid mixtures The experimental setup consists of a Godbert–Greenwald furnace (GG furnace) as it is used to determine the MIT of dust clouds. In contrast to the EN 50281-2-1 [24] the GG furnace used for the present study was of double length of the reaction cylinder (42 cm). For the test with only dust, the EN 50281-2-1 testing procedure was used. In case of the tests with gases, solvent or hybrid mixtures, modifications were done on the equipment. Fig. 3 shows the image of the experimental setup while Fig. 4 displays the schematic diagram for the experimental setup for the GG furnace used in the work. In order to ensure the repeatability and reproducibility of the experiments, a precise protocol was followed. All experiments were carried out under the same initial conditions according to the methods used.
The MIT for the individual substance were initially tested and the process flow diagram for these tests is shown in Fig. 5. The furnace tube was heated and fixed at the desired temperature and the weighed amount of dust was placed in the dust chamber. The air reservoir was filled with air up to the desired dispersion pressure and the dust sample was then dispersed through the furnace tube by a blast of air. The criterion for indicating an explosion was an observation of a flame at the bottom open mouth of the furnace with the help of a mirror. Both the pressure (0.1–0.5 bar above atmospheric pressure) and the mass of dust (0.1–0.5 g) were varied until a vigorous explosion was obtained. The conditions at which the vigorous explosion was obtained were taken as the “best” explosion region. This region was maintained, the furnace temperature was lowered and testing continued until no flame was observed in ten tests. The difference in temperatures between explosion and
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
317
Table 2 Properties of gases and solvents used. Properties
Methane
Propane
Hydrogen
Ethanol
Isopropanol
Toluene
Molecular formula Purity (%) Density (g/cm3 ) Molecular weight Explosible range (vol.%) Melting point (◦ C) Specific heat capacity (J/mol K) Boiling point (◦ C) Heat of vaporization (kJ/mol) Maximum explosion pressure (bar) MIT (◦ C) MSG (mm) Temperature class Explosion group LOC (vol.%)
CH4 99.87 6.6E-4 16 4.4–17 −161 35.69 −182.52 −74.87 8.1 595 1.14 T1 IIA 12.0
C3 H8 99.00 4.93E-4 44.1 1.7–10.8 −187 73.60 −42.1 −103.80 9.8 490 0.92 T1 IIA 9.4
H2 99.99 8.99E-4 2 4.0–77 −259 28.80 −253 0.00 8.3 570 0.29 T1 IIC 4.3
C2 H6 O 96.90 0.79 46.07 3.3–19 −114 112.40 78.0 38.56 8.4 400 0.89 T2 IIB 8.5
C3 H8 O 99.9 0.78 60.1 2.0–13.4 −88 246.00 82 44.0 8.2 425 0.99 T2 IIA 8.7
C7 H8 99.00 0.87 92.1 1.1–7.8 −95 155.96 111 38.06 7.7 535 1.06 T1 IIA 9.6
no explosion was 5 K. The lowest temperature at which an ignition with a flame occurred was taken as the minimum ignition temperature. For the tests with gases, the GG furnace was modified by introducing a gas feed line to the air reservoir as shown in Fig. 4. The same experimental principle as explained for dust was used for the gas test. The only difference in this case was that the air was premixed with the combustible gas in the air reservoir and the dust chamber was left empty. The composition of the gas mixtures was determined based on partial pressures. The chosen pressures were between 0.1 and 0.5 bar above atmospheric pressure and the concentrations were also within the explosible range of individual substances. Fig. 6 explains how each test for the MIT was obtained. For solvent testing, the same procedure as explained for only gas testing was followed but in this case, another modification was done on the equipment to allow an input of solvent. The solvent chamber was heated with a heating filament to allow vaporization of the solvent before being dispersed into the furnace. The required amount of solvent to be tested was measured with a syringe, placed in the solvent chamber and allowed for some time in order to obtain a
complete vaporization. The required amount of air is pressurized into the air reservoir and upon opening the valve the pressurized air pushes the vaporized solvent into the hot furnace. If the concentration of the air–solvent mixtures are within the explosible range and the temperature either at the MIT or above, explosion will be obtained. For the test with hybrid mixtures, the same experimental principle as explained before was used. In this case, it was just the combination of the test method with pure dust and pure gas or solvent. The proceeding steps follow the same test principle for pure dust. Fig. 7 explains the detail flow diagram on how the MIT of hybrids were obtained. As soon as the MIT was obtained further tests were performed at 5 K below the MIT by varying both pressure and concentration to check if ignition will occur or not. 2.3. Experimental determination of MEC/LEL for dust and hybrid mixtures The same procedure used in the determination of MIT of single components was also used for the determination of MEC/LEL of the single components but in this case, at the best explosion region,
Fig. 4. Schematic diagram for experimental setup.
318
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
Fig. 5. MIT & MEC/LFL test of single component flow chart.
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
Fig. 6. An example of how either the MIT or LEL was obtained.
Fig. 7. MIT test of hybrid mixture flow chart.
319
320
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
the temperature was kept constant and the concentration of the dusts were varied until a point where no ignition was obtained. At the concentration where the last ignition was obtained is termed as the MEC dusts or the LEL of gases For the test with hybrid mixtures, the same experimental principle as explained before was used. In this case, it was just the combination of the test methods with pure dust and pure gas or solvent. The proceeding steps followed the same test principle for pure substance. Fig. 8 explains the detail flow diagram on how the MEC/LEL of hybrids were obtained. After obtaining the MEC/LEL of hybrid mixture, further tests were performed below that concentration by varying the temperature and pressure to check if ignition will be obtained or not.
ethanol, isopropanol and toluene. Two main cases where considered in this work: a MIT of dust, gas and hybrid mixture. b MEC/LEL of dust, gas and hybrid mixtures. In order to prove the validity of our experimental procedure for the first case, the MIT for gas–air mixtures were initially tested and the results were compared with literature values. From Fig. 9 it can be seen that the experimental results are in satisfactory agreement with the data published by Brandes and Möller [23] obtained according to the standard test method. The maximum deviation occurred for toluene and isopropanol was 15 K. Most of the MIT of gases from our experimental results were 5–15 K higher than that reported from standard test. This is because during turbulence the heat transfer inside the gas is intensified so that local overheating of the gas is prevented [25]. This has proven that even if the method used here differs from the standard procedure, there is no cause for alarm to use this method to test for MIT of hybrid mixtures. It must be noted that the method used does not seek to replace the standard procedure for gas–air mixtures, but can be used to test for the MIT of mixtures consisting of components in different states of aggregation.
3. Results and discussion 3.1. MIT of single-fuel and hybrid mixtures The minimum ignition temperature of single-fuel and hybrid mixtures were investigated in the modified GG furnace as explained in Chapter 2. Five dusts, namely; starch, lycopodium, toner, wood and CN4 were mixed with methane, propane and hydrogen,
START
HYBRID LFL/MEC Test (D)
Conc.gas is taken @ LFL gas
Yes
MITgas > MITdust
Gas on MEC of Dust
Effect of
Dust on LFL of Gas
MITdust > MITgas
No Test LFL of Gas at temperature use for MEC Dust (Test B)
Yes
Conc.dust is taken @MEC of dust
No
Dustconc. = Constant Use condion of LFL of pure substance (TEST B)
Gasconc. = Constant Use condion of MEC of pure substance (TEST B)
Test MEC of Dust at temperature use for LFL of gas (Test B)
Conc.gas = 0.8 x LFLgas@Tdust,MEC
Conc.dust = 0.8 x MECdust@TgasLFL No effect of Gas/ Dust on LFL/MEC
No
Ignion Yes Temperature, T = fixed Conc, C = Fixed Pressure, P = VariedNote 2
Decrease the conc. Pressure & Temp. is fixed
Yes
Ignion
No
Yes
Ignion
No
LFL/MEC = min conc. that explosion occur
Fig. 8. LLLEL/MEC test of hybrid mixture flow chart.
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
321
650
MIT (°C)
600
MIT Experiment
Brandes et al
550
500
450
400
methane
propane
hydrogen
ethanol
isopropanol
toluene
Fig. 9. Comparison of the experimental MIT of six gases and vapours with the GG furnace and according to standard procedure by Brandes and Möller [23].
Table 3 MEC of dusts as well as the concentrations of dust used in the hybrid mixtures test. Dust
MEC (g/m3 )
Cd (g/m3 )
Starch Lycopodium Wood Toner CN4
145 108 217 87 304
82.29 84.91 124.57 30.40 122.85
Table 4 LEL of gases as well as the various concentrations gases used in the hybrid mixtures test. Gases
LEL (vol.%)
Cg (vol.%)
Methane Propane Hydrogen Toluene Ethanol Isopropanol
4 2 5 1.1 3 2
2 1 3 0.6 1.8 1
A similar behaviour was also seen for the effect of dust on the MIT of solvents as shown in Fig. 11. The MIT of toluene decreased from 540 ◦ C to 525 ◦ C, 505 ◦ C, 475 ◦ C and 455 ◦ C when small amounts of CN4, starch, lycopodium ad wood dust were added. Analogously, the effect of addition of flammable gases on the MIT of dusts was also studied by adding a volume fractions of gases which were below the lower explosion limits to the dust. Fig. 12 shows the results for the MIT of hybrid mixtures of dust and gas (i.e. the effect of gas on the MIT of dust). With the exception of lycopodium and toner, the MIT of dusts (starch, wood and CN4) decreased when the amount of gas according to Table 4 was added. For example the MIT of wood dust of 460 ◦ C decreased to 420 ◦ C and 430 ◦ C when methane and propane were added, respectively. A similar behaviour was also noticed for the effect of solvent on the MIT of dust as shown in Fig. 13. The MIT of CN4 of 640 ◦ C decreased to 570 ◦ C, 590 ◦ C and 600 ◦ C when toluene, ethanol and isopropanol in volume fractions as indicated in Table 4 were added, respectively. 3.2. LEL/MEC for single and hybrid mixtures
Based on the results from the single-fuel mixtures the MIT of hybrid mixtures were determined. Two different test series were considered which reflected the effect of dust on the MIT of gases/vapours and the effect of gases on the MIT of dusts. Starting from the MIT of the single-fuel air mixture, the temperature of the GG furnace was further decreased to check if an addition of a small amount of combustible dust would decrease the MIT of gas or not. The mass of dust was selected such that the dust itself if mixed with air would not have formed an ignitable mixture in the GG furnace. Similarly, the volume fractions of the flammable gases added to the dust–air mixtures would not have reached their lower explosion limit. The compositions of the hybrid mixtures under investigation are given in Tables 3 and 4 where Cd and Cg are the various concentration of dust or gas used in the hybrid mixtures test respectively. Fig. 10 shows the results for the MIT of hybrid mixtures of dust and gas. The red squares indicate the MIT of the gas–air mixtures while the other symbols indicate the MIT of hybrid mixtures. It can be seen that the MIT of hybrid mixtures were in all cases under investigation lower than those for the gas–air mixtures. For example, the MIT of methane of 600 ◦ C decreased to 530 ◦ C and 560 ◦ C when small amounts of toner and wood dust which itself does not form explosion atmosphere (shown in Table 3) were added, respectively.
The second test case considered in this paper was the LEL/MEC for single and hybrid mixtures. This was done to verify if the addition of a flammable dust at a concentration below the MEC could decrease the LEL of gases and vice versa. Various studies have been done by different authors on the LEL/MEC of hybrid mixtures [26–31] finding that the addition of small concentrations of either gas or dust (below the individual LEL or MEC) decreased the LEL or MEC of gas or dust, respectively. Unlike the standard 20 l-sphere which is normally used to determine the LEL or MEC of single and hybrid mixtures, in this work the GG furnace was used instead. The LEL/MEC of single substances was measured at furnace temperatures 20 K above the MIT of the substances. In order to prove the validity of our experimental procedure, the LEL for single-fuel gas–air mixtures were initially tested and the results were compared with literature values. Fig. 14 presents a comparison between the experimental results for the LEL obtained with the GG furnace with data published by Brandes and Möller [23]. Only in case of hydrogen the deviation was higher than 0.3 % in volume fraction. In a first test series, the temperature at which the LEL of the single-fuel gas–air mixture was obtained was kept constant and the amount of dust indicated in Table 3 was added. The LEL of the gas–air mixture decreased when dust at a concentration below its MEC was added. This behaviour can be seen in Fig. 15. For example the LEL of methane of 4% in volume decreased to 1.5% in volume, 1.3% in volume and 1% in volume when small
322
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
MIT of hybrid mixtures (°C)
610 590 570 550 530 510 490 470
methane
propane
MIT of gas (without dust) effect of lycopodium on the Mit of gaes effect of wood on the MIT of gases
hydrogen
effect of starch on MIT of gases effect of toner on the MIT of gases
Fig. 10. MIT of hybrid mixtures of dust and gas (effect of dust on the MIT of gases). (For interpretation of the references to colour in the text, the reader is referred to the web version of this article.)
MIT of hybrid mixtures (°C)
550
500
450
400
350
ethanol
isopropanol
MIT of solvent (vapor without dust) effect of lycopodium on the Mit of gaes effect of wood on the MIT of gases
toluene
effect of starch on MIT of gases effect of toner on the MIT of gases effect of CN4 on gases
Fig. 11. MIT of hybrid mixtures of dust and solvent in vapour state (effect of dust on the MIT of solvent).
650
MIT of Hybrid mixture (°C)
MIT of dust (without gas) effect of methane on dust
600
effect of propane on dust 550
effect of hydrogen on dust
500
450
400
350
starch
lycopodium
toner
wood
CN4
Fig. 12. MIT of hybrid mixtures of dust and gas (effect of gases on the MIT of dust).
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
323
MIT of hybrid mixtures (°C)
650
600
MIT of dust (without solvent) effect of ethanol on dust effect of isopropanol on dust effect of toluene on dust
550
500
450
400
350
starch
lycopodium
toner
wood
CN4
Fig. 13. MIT of hybrid mixtures of dust and solvent in vapour state (effect of solvents on the MIT of dust).
5.5
Experimental LEL
5
Brandes et al
LEL (vol. %)
4.5 4 3.5 3 2.5 2 1.5 1
methane
propane
hydrogen
ethanol
isopropanol touleune
Fig. 14. Comparison of the LEL for six different gases obtained in the GG furnace with data from Brandes and Möller [23].
LEL of hybrid mixtures (vol. %)
5.4
4.4
3.4
2.4
1.4
0.4
methane
propane
LEL of gases (without dust) effect of Lycopodium on the LEL of gases effect of toner on the LEL of gases
hydrogen
effect of starch on the LEL of gases effect of wood on the LEL of gases effect of CN4 on the LEL of gases
Fig. 15. LEL of hybrid mixture of dust and gas (effect of dust on the LEL of gases).
324
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
LEL of hybrid mixtures (vol. %)
3.5 3 2.5 2 1.5 1 0.5 0
ethanol
isopropanol
LEL of solvent- vapor (without dust) effect of Lycopodium on the LEL of gases effect of toner on the LEL of gases
toluene
effect of starch on the LEL of gases effect of wood on the LEL of gases
Fig. 16. LEL of hybrid mixtures of dust and solvents in vapour state (effect of dust on the LEL of solvent).
MEC of hybrid mixtures (g/m3)
310
MEC of dust (without gas) effect of propane on dust
effect of methane of dust effect of hydrogen on dust
starch
wood
260
210
160
110
60
10
lycopodium
toner
CN4
Fig. 17. Effect of addition of combustible gases on the MEC of different combustible dusts.
concentrations of starch, CN4 and lycopodium were added, respectively. Similar behaviour was also noticed for the effect of dusts on the LEL of solvent vapours. It can be seen from Fig. 16 that the LEL of the solvents decreased when dust concentrations below the MECs are added. A typical example is toluene where the LEL decreased from 1.1% in volume to 0.1% in volume when a concentration of wood dust below its MEC was added. The next test series on the dust, gas or vapour of hybrid mixtures was about the effect of combustible gas on the MEC of dusts. This was achieved by adding a concentration of gas or solvent at a concentration below their LEL to the dust. The concentration of the dust was below its MEC in these experiments. Fig. 17 presents the results for the effect of gas on the MEC of dust. It can be seen that the MEC of the dusts decreased when gases with concentrations below their LEL were added. For example, the MEC of CN4 of 304 g/m3 decreased to 37 g/m3 , 130 g/m3 and 217 g/m3 when propane, hydrogen and methane were added. In the case of adding flammable solvents to dusts, a similar behaviour was noticed as shown in Fig. 18. The various reasons contributing to the decrease of the MIT or LEL or MEC of either dust, solvent or gas could be briefly explained
by the mechanism of ignition of both dust and hybrid mixtures with respect to the test in the GG oven. With respect to organic dust (i.e. the dust used in this work), as soon as the dust is introduced to the furnace, three main stages could be applied before ignition occurs. The first, which is the pre-heating stage, the moisture content in the substance is vaporized. Usually, dust with higher moisture content requires higher ignition temperature because evaporation and heating of water represents an inert heat sink [4,32]. Eckhoff [4] further added that the water vapour mixes with the pyrolysis gases in the preheating zone of the combustion wave and makes the gas mixture less reactive. The next stage is the pyrolysis and devolatilization step. This step is considered as the first step in combustion process. During this stage, the organic particle is heated, producing volatile matters or combustible gases. The combustible gases are then mix with air in the space between the particle. The final stage is the oxidation of pyrolysis gases, where the gas phase combustion of premixed volatile-air takes place or in other words, the oxidation of homogenous gas take place [33]. In the case of hybrid mixture explosions, where flammable gas or vaporizing solvent are added to the dust, it can be envisaged that this addition could promote the combustion by bypassing the
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
325
MEC of hybrid mixtures (g/m3)
350 300
MEC of dust (without solvent) effect of isopropanol on dust
effect of ethanol of dust effect of toluene on dust
250 200 150 100 50 0 starch
lycopodium
wood
toner
CN4
Fig. 18. Effect of addition of combustible vapours on the MEC of different combustible dusts.
pyrolysis step. Dufaud et al. [34] analysed the minimum explosible concentration of dust for combustible gas/dust hybrid mixtures to stress the impact of the introduction of highly flammable compounds on the explosivity of dust clouds. They observed that the dust explosivity is strongly promoted by the addition of combustible gas. However, the authors further highlighted that this effect is more significant below 50% of dust to gas ratio. Therefore, it is assumed that dust still plays a significant role in the combustion kinetics down to this limit; for greater amounts of gases, the specific behaviour of the combustible gases is clearly predominant. Such addition, as low as 1 vol.% induces changes in the rate-limiting step of the combustion reaction, from devolatilization to homogeneous gas phase reaction, and implies a drastic decrease of the MIT, LEL or MEC for hybrid mixtures. 4. Conclusions The effects of flammable vapour or gas on minimum ignition temperature and explosion limit of combustible dust, and vice versa have been studied using the Godbert–Greenwald furnace. Various types of dusts, gases and solvents as vapour were used in these experiments. This present work has affirmed that both gases/vapours and dust have influence on MIT, and LEL and MEC of hybrid mixture even though the added concentration was below their respective explosible limit. The addition of combustible gases can be seen as replacement for volatiles released from the dust during pyrolysis and hence significantly affect its MIT, LEL of gases and MEC of dust. Based on our findings, the following conclusions can be made. • The MIT and LEL of gases decrease if a small amount of dust below its own MEC is added. • The MIT and MEC of dusts decrease if a small amount of gas below its own LEL is added. References [1] EN 1127-1, Determination of explosion limits of gases and vapours, (2010). [2] P. Amyotte, An Introduction to Dust Explosions: Understanding the Myths and Realities of Dust Explosions for a Safer Workplace, Elsevier, 2013, p. 46. [3] A. Di Benedetto, P. Russo, On the determination of the minimum ignition temperature for dust/air mixtures, J. Loss Prev. Process Ind. 20 (2007) 303–309. [4] R. Eckhoff, Dust Explosions in the Process Industries, 3rd ed., Gulf Professional Publishing, Elsevier Science, 2003, pp. 25–141, 251–424. [5] EN 14522, Determination of the minimum ignition temperature of gases and vapours, CEN (2005).
[6] R.H. Essenhigh, M.K. Misra, D.W. Shaw, Ignition of coal particles: a review, Combust. Flame 77 (1989) 3–30. [7] H.M. Cassel, I. Liebman, The cooperative mechanism in the ignition of dust dispersions, Combust. Flame 3 (1959) 467–475. [8] X. Du, K. Animalia, The transient ignition of isolated coal particles, Combust. Flame 97 (1994) 339–354. [9] M. Chen, L.S. Fan, R.H. Essenhigh, Prediction and measurement of ignition temperatures of coal particles, in: Proceedings of the 20th International Symposium on Combustion, The Combustion Institute, Pittsburgh, PA, 1984, pp. 1513–1521. [10] S. Bandyopadhyay, D.B. Bhaduri, Prediction of ignition temperature of a single coal particle, Combust. Flame 18 (1972) 411–415. [11] D. Zhang, T.F. Wall, An analysis of the ignition of coal dust clouds, Combust. Flame 92 (1993) 475–480. [12] C. Engler, Beitrage zur Kenntniss der Staubexplosionen, Chemische Industrie (1885) 171–173. [13] K. Chatrathi, Dust and hybrid explosibility in a 1 m3 spherical chamber, Process Saf. Prog. 13 (1994) 183–189. [14] R. Pilão, E. Ramalho, C. Pinho, Explosibility of cork dust in methane/air mixtures, J. Loss Prev. Process Ind. 19 (2006) 17–23. [15] Y. Liu, J. Sun, D. Chen, Flame propagation in hybrid mixture of coal dust and methane, J. Loss Prev. Process Ind. 20 (2007) 691–697. [16] P.R. Amyotte, K.J. Mintz, M.J. Pegg, Y.H. Sun, The ignitability of coal dust–air and methane–coal dust–air mixtures, Fuel 72 (1993) 671–679. [17] P. Amyotte, M. Lindsay, R. Domaratzki, N. Marchand, A. Di Benedetto, P. Russo, Prevention and mitigation of dust and hybrid mixture explosions, Process Saf. Prog. 29 (2010) 17–21. [18] M. Nifuku, H. Tsujita, K. Fujino, K. Takaichi, C. Barre, E. Paya, et al., Ignitability assessment of shredder dusts of refrigerator and the prevention of the dust explosion, J. Loss Preve. Process Ind. 19 (2–3) (2006) 181–186. [19] O. Dufaud, L. Perrin, M. Traore, S. Chazelet, D. Thomas, Explosions of vapour/dust hybrid mixtures: a particular class, Powder Technol. 190 (2009) 269–273. [20] W.E. Baker, M.J. Tang, Gas, Dust and Hybrid Explosions, Elsevier, 1991, ISBN: 978-0-444-88150-2 http://www.elsevier.com/books/gas-dust-and-hybridexplosions/bakerandampdagger/978-0-444-88150-2. [21] W. Nazaroff, R. Harley, Air Quality Engineering, University of California, Berkeley, 2007, CE 218A. [22] Standard Thermodynamic Properties of Chemical Substances by CRC Press llc (2000). [23] E. Brandes, W. Möller, Sicherheitstechnische Kenngroßn, Brand1:Brennbare Flüssigkeiten und Gase, PTB (2003). [24] EN 50281-2-1, European standard test method for the minimum ignition temperature for dust cloud (2007). [25] M. Hattwig, H. Steen, Handbook of Explosion Prevention and Protection, Wiley–VCH Verlag GMbH & Co. CGAC, 2004, pp. 165, ISBN-13: 978-3527307180, Section 2.3.3.2. [26] M. Encinar, F.J. Beltra, J.F. Gonza, M.J. Moreno, Pyrolysis of maize, sunflower, grape and tobacco residues, J. Chem. Tech. Biotechnol. (1997) 400–410. [27] I. Khalili, O. Dufaud, M. Poupeau, N. Cuervo-Rodriguez, L. Perrin, Ignition sensitivity of gas–vapour/dust hybrid mixtures, Powder Technol. 217 (2012) 199–206. [28] A. Agreda, A. Di Benedetto, P. Russo, E. Salzano, R. Sanchiro, Dust/gas mixtures explosion regimes, Powder Technol. 205 (2011) 81–86, http://dx.doi.org/10. 1016/j.powtec.2010.08.069.205. [29] E.K. Addai, D. Gabel, U. Krause, Explosion characteristics of three component hybrid mixtures, Process Saf. Environ. Prot. 9 (8) (2015) 72–81.
326
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
[30] E.K. Addai, D. Gabel, U. Krause, Lower explosion limit of hybrid mixtures of burnable gas and dust, J. Loss Prev. Process Ind. (2015), http://dx.doi.org/10. 1016/j.jlp.2015.02.014. [31] E.A. Addai, D. Gabel, U. Krause, Ignition and explosion behavior of hybrid mixtures of two and three components in: 4. Magdeburger Brand- und Explosionsschutztag, HS- Magdeburg Stendal: Magdeburg, 2015, ISBN 978-3-00-048960-0, 26. und 27. März 2015, Datenträger. [32] M. Hertzberg, I.A. Zlochower, R.S. Conti, K.L. Cashdollar, Thermo-kinetic transport control and structural microscopic; realities in coal and polymer
pyrolysis and devolatilization: their dominant role in dust explosion, Am. Chem. Soc. Div. Fuel Chem. 32 (3) (1987) 24–41. [33] Kwasi Emmanuel Addai, Dieter Gabel, Ulrich Krause, Minimum ignition temperature of hybrid mixtures of burnable dusts and gases, in: Proceedings of the 25th International Colloquium on the Dynamics of Explosions and Reactive Systems, 2–7 August, Leeds, UK, ICDERS, 2015, insges. 6 S. [34] O. Dufaud, L. Perrin, M. Traoré, Dust/vapour explosions: hybrid behaviours? J. Loss Prev. Process Ind. 21 (4) (2008) 481–484.
Contents lists available at ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Experimental investigation on the minimum ignition temperature of hybrid mixtures of dusts and gases or solvents Emmanuel Kwasi Addai ∗ , Dieter Gabel, Ulrich Krause Otto-von-Guericke-University, Institute of Instrumental and Environmental Technology Department of Systems Engineering and Plant Safety, Universitätsplatz 2, 39106 Magdeburg, Germany
h i g h l i g h t s • • • • •
Lower explosion limit of gases decreases upon addition of dust which is itself not ignitable. Minimum ignition temperature of gases decrease upon addition of dust which is itself not ignitable. Minimum ignition temperature of hybrid mixtures are lower than that of single substance. Minimum explosion concentration of dust decreases upon addition of gas which is itself not ignitable. Lower explosion limits of hybrid mixtures are lower than that of single substance.
a r t i c l e
i n f o
Article history: Received 9 July 2015 Received in revised form 30 July 2015 Accepted 2 September 2015 Available online 7 September 2015 Keywords: MIT Hybrid mixture explosion Dust explosion Gas explosion MIT of hybrid mixture explosion
a b s t r a c t Investigations on the minimum ignition temperatures (MIT) of hybrid mixtures of dusts with gases or solvents were performed in the modified Godbert–Greenwald (GG) furnace. Five combustible dusts and six flammable gases (three ideal and three real) were used. The test protocol was according to EN 502812-1 for dust–air mixtures whereas in the case of gases, solvents and hybrid mixtures this standard was used with slight modification. The experimental results demonstrated a significant decrease of the MIT of gas, solvent or dust and an increase in the likelihood of explosion when a small amount of dust, which was either below the minimum explosion concentration or not ignitable by itself, was mixed with gas and vice versa. For example, the MIT of toluene decreased from 540 ◦ C to 455 ◦ C when small amount of lycopodium was added. It was also confirmed that a hybrid mixture explosion is possible even when both dust and vapour or gas concentrations are respectively lower than their minimum explosion concentration (MEC) and lower explosion limit (LEL). Another example is CN4, the MEC of which of 304 g/m3 decreased to 37 g/m3 when propane was added, even though the concentrations of the gas was below its LEL. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Hybrid mixture explosions can occur in a wide range of industries such as mining, agriculture, pharmaceutical and petrochemical industries that involve handling or processing of flammable gases or solvents and combustible dusts. The mixtures consist of two or more combustible substances with different aggregate states. The minimum ignition temperature (MIT) and lower explosion limit (LEL) or minimum explosible concentration (MEC) are critical parameters when conducting hazard assessment for processes involving hybrid mixtures. In the present paper we
∗ Corresponding author. Fax: +49 391 67 11128. E-mail addresses: [email protected], [email protected] (E.K. Addai). http://dx.doi.org/10.1016/j.jhazmat.2015.09.006 0304-3894/© 2015 Elsevier B.V. All rights reserved.
use the term lower explosion limit in accordance with the standard EN 1127-1 [1] for the lowest volume fraction in% of a flammable gas mixed with air, for which flame propagation is still possible. In contrast, for a dust–air mixture the minimum amount of fuel necessary for self-sustained flame propagation is usually given in mass of dust per volume of air. We therefore in the present paper prefer the term minimum explosible concentration which has the unit g/m3 . Hot surfaces capable of igniting dust clouds exist in a number of situations in industry (furnaces, burners and dryers of various kinds). In addition, hot surfaces can be generated accidentally by overheated bearings and other mechanical parts. If an explosible dust cloud is generated in some uncontrolled way in the proximity of a hot surface with a temperature above the actual minimum ignition temperature, a dust explosion can result [2,3].
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
315
Fig. 1. Particle size distribution of dust.
Table 1 Preparatory analysis of the dust used. Toner
CN4
Wood
Lycopodium
Starch
Molecular formula Media diameter (m) Volatile content (% mass) Moisture content (% mass) Heat of combustion (kJ/kg)
C7.17 H7.75 O0.33
C6.70 H1.31 O0.88 N0.02
C4.19 H6.26 O2.71 S0.004
C5.77 H9.59 O1.23 S0.001 N0.08
C3.69 H6.34 O3.06 S0.01
13.4 90.18 0.92 35792
60 17.08 0.23 26630
307 84.38 0.2 16446
31.6 91.06 0.35 28447
29.2 93.77 0.5 15302
Elemental analysis C H O S N
86.05 7.72 5.23 1 0
80.37 1.31 14.01 3.02 0.34
50.33 6.26 43.28 0.13 0
69.26 9.59 19.62 0.38 1.15
44.34 6.34 48.94 0.38 0
Unlike solitary dust, gas or solvent explosions, which have been widely studied in the past decades, data on explosion characteristics of hybrid mixtures are comparatively sparse. The minimum ignition temperature (MIT) is an explosion characteristic which is used for dust–air mixtures as well as for gas–air mixtures or for mixtures of vapours of flammable liquids with air. However, while for dusts the GG furnace or the BAM oven are used to determine the MIT [4], for gases or vapours the MIT is determined in the Erlenmeyer flask [5]. In case of the MIT of hybrid mixtures the Erlenmeyer flask is not suitable because it does not allow the dispersion of dust particles. Consequently, for the determination of the MIT of hybrid mixtures the GG furnace has to be modified in such a way that besides the generation of a dust–air mixture also a flammable gas or vapour can be added to the mixture. The set-up used for this purpose is described in Section 2. Consequently, in the prevention and mitigation of dust explosions, it is important to know the minimum ignition temperature (MIT), minimum explosible dust concentration (MEC) and lower explosion limit (LEL) of dusts in order to take adequate precautions to ensure that hot surface temperature does not reach this value. For pure dusts, the MIT depends on the material properties, such as particle size distribution of the dust. Furthermore, the environments conditions have also to be taken into account, e.g. oxygen availability, environmental temperature and so forth [6–10]. For hybrid mixtures, the conditions become more complex, as the MIT of hybrid mixtures cannot be predicted by simply overlapping the
effects of the single substances. Moreover, there is no systematic study that is able to quantify the roles of dust and gas in driving the explosion. The main conclusions of previous studies [11–20] on the explosion properties of hybrid mixtures could be summarized non-exhaustively by the following assertions, that the ignition sensitivity of the powder can be strongly increased by the addition of a few percent of combustible gases or vapours, even with contents lower than the LEL. It has notably been shown that hybrid mixtures can also be explosible when the concentrations of the dust and the gas are both below their respective MEC and LEL, respectively. However, in the case of minimum ignition temperatures of hybrid mixtures data are nearly not available. Hence, the work of this paper is to investigate on the MIT and LEL/MEC of hybrid mixtures of dust and gas or solvent in vapour state. In order to achieve this, combinations of five dust samples and six gases or vapours were tested. All these tests were performed in the modified Godbert–Greenwald furnace. 2. Materials and experimental work 2.1. Tested materials Five different dusts namely cornstarch, lycopodium, toner powder, wood dust and CN4 as well as each three perfect gases (methane, propane and hydrogen) and three real gases (vapour of ethanol, toluene and isopropanol) were used in the present study.
316
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
Fig. 2. SEM images for various dust with different magnifications.
The particle size distribution is one of the most important properties that affect the minimum ignition temperature of dust clouds. In order to determine the particle size distribution, all the dust samples underwent CAMSIZER probing. Fig. 1 shows the particle size distributions for the dusts used. Furthermore, in Fig. 2 scanning electron microscopy (SEM) images for all the dusts used with different magnification in order to reveal the surface structure of the particles are presented. Table 1 summarizes the data of the elemental analysis, median particle size values, mass content of volatiles, moisture mass content, heat of combustion and molecular formula of the dusts which were calculated from elemental analysis. Table 2 gives the basic, thermodynamic, ignition or combustion properties of the gases used [21–23].
Fig. 3. Image of experimental setup.
2.2. Experimental determination of MIT for dust and hybrid mixtures The experimental setup consists of a Godbert–Greenwald furnace (GG furnace) as it is used to determine the MIT of dust clouds. In contrast to the EN 50281-2-1 [24] the GG furnace used for the present study was of double length of the reaction cylinder (42 cm). For the test with only dust, the EN 50281-2-1 testing procedure was used. In case of the tests with gases, solvent or hybrid mixtures, modifications were done on the equipment. Fig. 3 shows the image of the experimental setup while Fig. 4 displays the schematic diagram for the experimental setup for the GG furnace used in the work. In order to ensure the repeatability and reproducibility of the experiments, a precise protocol was followed. All experiments were carried out under the same initial conditions according to the methods used.
The MIT for the individual substance were initially tested and the process flow diagram for these tests is shown in Fig. 5. The furnace tube was heated and fixed at the desired temperature and the weighed amount of dust was placed in the dust chamber. The air reservoir was filled with air up to the desired dispersion pressure and the dust sample was then dispersed through the furnace tube by a blast of air. The criterion for indicating an explosion was an observation of a flame at the bottom open mouth of the furnace with the help of a mirror. Both the pressure (0.1–0.5 bar above atmospheric pressure) and the mass of dust (0.1–0.5 g) were varied until a vigorous explosion was obtained. The conditions at which the vigorous explosion was obtained were taken as the “best” explosion region. This region was maintained, the furnace temperature was lowered and testing continued until no flame was observed in ten tests. The difference in temperatures between explosion and
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
317
Table 2 Properties of gases and solvents used. Properties
Methane
Propane
Hydrogen
Ethanol
Isopropanol
Toluene
Molecular formula Purity (%) Density (g/cm3 ) Molecular weight Explosible range (vol.%) Melting point (◦ C) Specific heat capacity (J/mol K) Boiling point (◦ C) Heat of vaporization (kJ/mol) Maximum explosion pressure (bar) MIT (◦ C) MSG (mm) Temperature class Explosion group LOC (vol.%)
CH4 99.87 6.6E-4 16 4.4–17 −161 35.69 −182.52 −74.87 8.1 595 1.14 T1 IIA 12.0
C3 H8 99.00 4.93E-4 44.1 1.7–10.8 −187 73.60 −42.1 −103.80 9.8 490 0.92 T1 IIA 9.4
H2 99.99 8.99E-4 2 4.0–77 −259 28.80 −253 0.00 8.3 570 0.29 T1 IIC 4.3
C2 H6 O 96.90 0.79 46.07 3.3–19 −114 112.40 78.0 38.56 8.4 400 0.89 T2 IIB 8.5
C3 H8 O 99.9 0.78 60.1 2.0–13.4 −88 246.00 82 44.0 8.2 425 0.99 T2 IIA 8.7
C7 H8 99.00 0.87 92.1 1.1–7.8 −95 155.96 111 38.06 7.7 535 1.06 T1 IIA 9.6
no explosion was 5 K. The lowest temperature at which an ignition with a flame occurred was taken as the minimum ignition temperature. For the tests with gases, the GG furnace was modified by introducing a gas feed line to the air reservoir as shown in Fig. 4. The same experimental principle as explained for dust was used for the gas test. The only difference in this case was that the air was premixed with the combustible gas in the air reservoir and the dust chamber was left empty. The composition of the gas mixtures was determined based on partial pressures. The chosen pressures were between 0.1 and 0.5 bar above atmospheric pressure and the concentrations were also within the explosible range of individual substances. Fig. 6 explains how each test for the MIT was obtained. For solvent testing, the same procedure as explained for only gas testing was followed but in this case, another modification was done on the equipment to allow an input of solvent. The solvent chamber was heated with a heating filament to allow vaporization of the solvent before being dispersed into the furnace. The required amount of solvent to be tested was measured with a syringe, placed in the solvent chamber and allowed for some time in order to obtain a
complete vaporization. The required amount of air is pressurized into the air reservoir and upon opening the valve the pressurized air pushes the vaporized solvent into the hot furnace. If the concentration of the air–solvent mixtures are within the explosible range and the temperature either at the MIT or above, explosion will be obtained. For the test with hybrid mixtures, the same experimental principle as explained before was used. In this case, it was just the combination of the test method with pure dust and pure gas or solvent. The proceeding steps follow the same test principle for pure dust. Fig. 7 explains the detail flow diagram on how the MIT of hybrids were obtained. As soon as the MIT was obtained further tests were performed at 5 K below the MIT by varying both pressure and concentration to check if ignition will occur or not. 2.3. Experimental determination of MEC/LEL for dust and hybrid mixtures The same procedure used in the determination of MIT of single components was also used for the determination of MEC/LEL of the single components but in this case, at the best explosion region,
Fig. 4. Schematic diagram for experimental setup.
318
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
Fig. 5. MIT & MEC/LFL test of single component flow chart.
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
Fig. 6. An example of how either the MIT or LEL was obtained.
Fig. 7. MIT test of hybrid mixture flow chart.
319
320
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
the temperature was kept constant and the concentration of the dusts were varied until a point where no ignition was obtained. At the concentration where the last ignition was obtained is termed as the MEC dusts or the LEL of gases For the test with hybrid mixtures, the same experimental principle as explained before was used. In this case, it was just the combination of the test methods with pure dust and pure gas or solvent. The proceeding steps followed the same test principle for pure substance. Fig. 8 explains the detail flow diagram on how the MEC/LEL of hybrids were obtained. After obtaining the MEC/LEL of hybrid mixture, further tests were performed below that concentration by varying the temperature and pressure to check if ignition will be obtained or not.
ethanol, isopropanol and toluene. Two main cases where considered in this work: a MIT of dust, gas and hybrid mixture. b MEC/LEL of dust, gas and hybrid mixtures. In order to prove the validity of our experimental procedure for the first case, the MIT for gas–air mixtures were initially tested and the results were compared with literature values. From Fig. 9 it can be seen that the experimental results are in satisfactory agreement with the data published by Brandes and Möller [23] obtained according to the standard test method. The maximum deviation occurred for toluene and isopropanol was 15 K. Most of the MIT of gases from our experimental results were 5–15 K higher than that reported from standard test. This is because during turbulence the heat transfer inside the gas is intensified so that local overheating of the gas is prevented [25]. This has proven that even if the method used here differs from the standard procedure, there is no cause for alarm to use this method to test for MIT of hybrid mixtures. It must be noted that the method used does not seek to replace the standard procedure for gas–air mixtures, but can be used to test for the MIT of mixtures consisting of components in different states of aggregation.
3. Results and discussion 3.1. MIT of single-fuel and hybrid mixtures The minimum ignition temperature of single-fuel and hybrid mixtures were investigated in the modified GG furnace as explained in Chapter 2. Five dusts, namely; starch, lycopodium, toner, wood and CN4 were mixed with methane, propane and hydrogen,
START
HYBRID LFL/MEC Test (D)
Conc.gas is taken @ LFL gas
Yes
MITgas > MITdust
Gas on MEC of Dust
Effect of
Dust on LFL of Gas
MITdust > MITgas
No Test LFL of Gas at temperature use for MEC Dust (Test B)
Yes
Conc.dust is taken @MEC of dust
No
Dustconc. = Constant Use condion of LFL of pure substance (TEST B)
Gasconc. = Constant Use condion of MEC of pure substance (TEST B)
Test MEC of Dust at temperature use for LFL of gas (Test B)
Conc.gas = 0.8 x LFLgas@Tdust,MEC
Conc.dust = 0.8 x MECdust@TgasLFL No effect of Gas/ Dust on LFL/MEC
No
Ignion Yes Temperature, T = fixed Conc, C = Fixed Pressure, P = VariedNote 2
Decrease the conc. Pressure & Temp. is fixed
Yes
Ignion
No
Yes
Ignion
No
LFL/MEC = min conc. that explosion occur
Fig. 8. LLLEL/MEC test of hybrid mixture flow chart.
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
321
650
MIT (°C)
600
MIT Experiment
Brandes et al
550
500
450
400
methane
propane
hydrogen
ethanol
isopropanol
toluene
Fig. 9. Comparison of the experimental MIT of six gases and vapours with the GG furnace and according to standard procedure by Brandes and Möller [23].
Table 3 MEC of dusts as well as the concentrations of dust used in the hybrid mixtures test. Dust
MEC (g/m3 )
Cd (g/m3 )
Starch Lycopodium Wood Toner CN4
145 108 217 87 304
82.29 84.91 124.57 30.40 122.85
Table 4 LEL of gases as well as the various concentrations gases used in the hybrid mixtures test. Gases
LEL (vol.%)
Cg (vol.%)
Methane Propane Hydrogen Toluene Ethanol Isopropanol
4 2 5 1.1 3 2
2 1 3 0.6 1.8 1
A similar behaviour was also seen for the effect of dust on the MIT of solvents as shown in Fig. 11. The MIT of toluene decreased from 540 ◦ C to 525 ◦ C, 505 ◦ C, 475 ◦ C and 455 ◦ C when small amounts of CN4, starch, lycopodium ad wood dust were added. Analogously, the effect of addition of flammable gases on the MIT of dusts was also studied by adding a volume fractions of gases which were below the lower explosion limits to the dust. Fig. 12 shows the results for the MIT of hybrid mixtures of dust and gas (i.e. the effect of gas on the MIT of dust). With the exception of lycopodium and toner, the MIT of dusts (starch, wood and CN4) decreased when the amount of gas according to Table 4 was added. For example the MIT of wood dust of 460 ◦ C decreased to 420 ◦ C and 430 ◦ C when methane and propane were added, respectively. A similar behaviour was also noticed for the effect of solvent on the MIT of dust as shown in Fig. 13. The MIT of CN4 of 640 ◦ C decreased to 570 ◦ C, 590 ◦ C and 600 ◦ C when toluene, ethanol and isopropanol in volume fractions as indicated in Table 4 were added, respectively. 3.2. LEL/MEC for single and hybrid mixtures
Based on the results from the single-fuel mixtures the MIT of hybrid mixtures were determined. Two different test series were considered which reflected the effect of dust on the MIT of gases/vapours and the effect of gases on the MIT of dusts. Starting from the MIT of the single-fuel air mixture, the temperature of the GG furnace was further decreased to check if an addition of a small amount of combustible dust would decrease the MIT of gas or not. The mass of dust was selected such that the dust itself if mixed with air would not have formed an ignitable mixture in the GG furnace. Similarly, the volume fractions of the flammable gases added to the dust–air mixtures would not have reached their lower explosion limit. The compositions of the hybrid mixtures under investigation are given in Tables 3 and 4 where Cd and Cg are the various concentration of dust or gas used in the hybrid mixtures test respectively. Fig. 10 shows the results for the MIT of hybrid mixtures of dust and gas. The red squares indicate the MIT of the gas–air mixtures while the other symbols indicate the MIT of hybrid mixtures. It can be seen that the MIT of hybrid mixtures were in all cases under investigation lower than those for the gas–air mixtures. For example, the MIT of methane of 600 ◦ C decreased to 530 ◦ C and 560 ◦ C when small amounts of toner and wood dust which itself does not form explosion atmosphere (shown in Table 3) were added, respectively.
The second test case considered in this paper was the LEL/MEC for single and hybrid mixtures. This was done to verify if the addition of a flammable dust at a concentration below the MEC could decrease the LEL of gases and vice versa. Various studies have been done by different authors on the LEL/MEC of hybrid mixtures [26–31] finding that the addition of small concentrations of either gas or dust (below the individual LEL or MEC) decreased the LEL or MEC of gas or dust, respectively. Unlike the standard 20 l-sphere which is normally used to determine the LEL or MEC of single and hybrid mixtures, in this work the GG furnace was used instead. The LEL/MEC of single substances was measured at furnace temperatures 20 K above the MIT of the substances. In order to prove the validity of our experimental procedure, the LEL for single-fuel gas–air mixtures were initially tested and the results were compared with literature values. Fig. 14 presents a comparison between the experimental results for the LEL obtained with the GG furnace with data published by Brandes and Möller [23]. Only in case of hydrogen the deviation was higher than 0.3 % in volume fraction. In a first test series, the temperature at which the LEL of the single-fuel gas–air mixture was obtained was kept constant and the amount of dust indicated in Table 3 was added. The LEL of the gas–air mixture decreased when dust at a concentration below its MEC was added. This behaviour can be seen in Fig. 15. For example the LEL of methane of 4% in volume decreased to 1.5% in volume, 1.3% in volume and 1% in volume when small
322
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
MIT of hybrid mixtures (°C)
610 590 570 550 530 510 490 470
methane
propane
MIT of gas (without dust) effect of lycopodium on the Mit of gaes effect of wood on the MIT of gases
hydrogen
effect of starch on MIT of gases effect of toner on the MIT of gases
Fig. 10. MIT of hybrid mixtures of dust and gas (effect of dust on the MIT of gases). (For interpretation of the references to colour in the text, the reader is referred to the web version of this article.)
MIT of hybrid mixtures (°C)
550
500
450
400
350
ethanol
isopropanol
MIT of solvent (vapor without dust) effect of lycopodium on the Mit of gaes effect of wood on the MIT of gases
toluene
effect of starch on MIT of gases effect of toner on the MIT of gases effect of CN4 on gases
Fig. 11. MIT of hybrid mixtures of dust and solvent in vapour state (effect of dust on the MIT of solvent).
650
MIT of Hybrid mixture (°C)
MIT of dust (without gas) effect of methane on dust
600
effect of propane on dust 550
effect of hydrogen on dust
500
450
400
350
starch
lycopodium
toner
wood
CN4
Fig. 12. MIT of hybrid mixtures of dust and gas (effect of gases on the MIT of dust).
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
323
MIT of hybrid mixtures (°C)
650
600
MIT of dust (without solvent) effect of ethanol on dust effect of isopropanol on dust effect of toluene on dust
550
500
450
400
350
starch
lycopodium
toner
wood
CN4
Fig. 13. MIT of hybrid mixtures of dust and solvent in vapour state (effect of solvents on the MIT of dust).
5.5
Experimental LEL
5
Brandes et al
LEL (vol. %)
4.5 4 3.5 3 2.5 2 1.5 1
methane
propane
hydrogen
ethanol
isopropanol touleune
Fig. 14. Comparison of the LEL for six different gases obtained in the GG furnace with data from Brandes and Möller [23].
LEL of hybrid mixtures (vol. %)
5.4
4.4
3.4
2.4
1.4
0.4
methane
propane
LEL of gases (without dust) effect of Lycopodium on the LEL of gases effect of toner on the LEL of gases
hydrogen
effect of starch on the LEL of gases effect of wood on the LEL of gases effect of CN4 on the LEL of gases
Fig. 15. LEL of hybrid mixture of dust and gas (effect of dust on the LEL of gases).
324
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
LEL of hybrid mixtures (vol. %)
3.5 3 2.5 2 1.5 1 0.5 0
ethanol
isopropanol
LEL of solvent- vapor (without dust) effect of Lycopodium on the LEL of gases effect of toner on the LEL of gases
toluene
effect of starch on the LEL of gases effect of wood on the LEL of gases
Fig. 16. LEL of hybrid mixtures of dust and solvents in vapour state (effect of dust on the LEL of solvent).
MEC of hybrid mixtures (g/m3)
310
MEC of dust (without gas) effect of propane on dust
effect of methane of dust effect of hydrogen on dust
starch
wood
260
210
160
110
60
10
lycopodium
toner
CN4
Fig. 17. Effect of addition of combustible gases on the MEC of different combustible dusts.
concentrations of starch, CN4 and lycopodium were added, respectively. Similar behaviour was also noticed for the effect of dusts on the LEL of solvent vapours. It can be seen from Fig. 16 that the LEL of the solvents decreased when dust concentrations below the MECs are added. A typical example is toluene where the LEL decreased from 1.1% in volume to 0.1% in volume when a concentration of wood dust below its MEC was added. The next test series on the dust, gas or vapour of hybrid mixtures was about the effect of combustible gas on the MEC of dusts. This was achieved by adding a concentration of gas or solvent at a concentration below their LEL to the dust. The concentration of the dust was below its MEC in these experiments. Fig. 17 presents the results for the effect of gas on the MEC of dust. It can be seen that the MEC of the dusts decreased when gases with concentrations below their LEL were added. For example, the MEC of CN4 of 304 g/m3 decreased to 37 g/m3 , 130 g/m3 and 217 g/m3 when propane, hydrogen and methane were added. In the case of adding flammable solvents to dusts, a similar behaviour was noticed as shown in Fig. 18. The various reasons contributing to the decrease of the MIT or LEL or MEC of either dust, solvent or gas could be briefly explained
by the mechanism of ignition of both dust and hybrid mixtures with respect to the test in the GG oven. With respect to organic dust (i.e. the dust used in this work), as soon as the dust is introduced to the furnace, three main stages could be applied before ignition occurs. The first, which is the pre-heating stage, the moisture content in the substance is vaporized. Usually, dust with higher moisture content requires higher ignition temperature because evaporation and heating of water represents an inert heat sink [4,32]. Eckhoff [4] further added that the water vapour mixes with the pyrolysis gases in the preheating zone of the combustion wave and makes the gas mixture less reactive. The next stage is the pyrolysis and devolatilization step. This step is considered as the first step in combustion process. During this stage, the organic particle is heated, producing volatile matters or combustible gases. The combustible gases are then mix with air in the space between the particle. The final stage is the oxidation of pyrolysis gases, where the gas phase combustion of premixed volatile-air takes place or in other words, the oxidation of homogenous gas take place [33]. In the case of hybrid mixture explosions, where flammable gas or vaporizing solvent are added to the dust, it can be envisaged that this addition could promote the combustion by bypassing the
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
325
MEC of hybrid mixtures (g/m3)
350 300
MEC of dust (without solvent) effect of isopropanol on dust
effect of ethanol of dust effect of toluene on dust
250 200 150 100 50 0 starch
lycopodium
wood
toner
CN4
Fig. 18. Effect of addition of combustible vapours on the MEC of different combustible dusts.
pyrolysis step. Dufaud et al. [34] analysed the minimum explosible concentration of dust for combustible gas/dust hybrid mixtures to stress the impact of the introduction of highly flammable compounds on the explosivity of dust clouds. They observed that the dust explosivity is strongly promoted by the addition of combustible gas. However, the authors further highlighted that this effect is more significant below 50% of dust to gas ratio. Therefore, it is assumed that dust still plays a significant role in the combustion kinetics down to this limit; for greater amounts of gases, the specific behaviour of the combustible gases is clearly predominant. Such addition, as low as 1 vol.% induces changes in the rate-limiting step of the combustion reaction, from devolatilization to homogeneous gas phase reaction, and implies a drastic decrease of the MIT, LEL or MEC for hybrid mixtures. 4. Conclusions The effects of flammable vapour or gas on minimum ignition temperature and explosion limit of combustible dust, and vice versa have been studied using the Godbert–Greenwald furnace. Various types of dusts, gases and solvents as vapour were used in these experiments. This present work has affirmed that both gases/vapours and dust have influence on MIT, and LEL and MEC of hybrid mixture even though the added concentration was below their respective explosible limit. The addition of combustible gases can be seen as replacement for volatiles released from the dust during pyrolysis and hence significantly affect its MIT, LEL of gases and MEC of dust. Based on our findings, the following conclusions can be made. • The MIT and LEL of gases decrease if a small amount of dust below its own MEC is added. • The MIT and MEC of dusts decrease if a small amount of gas below its own LEL is added. References [1] EN 1127-1, Determination of explosion limits of gases and vapours, (2010). [2] P. Amyotte, An Introduction to Dust Explosions: Understanding the Myths and Realities of Dust Explosions for a Safer Workplace, Elsevier, 2013, p. 46. [3] A. Di Benedetto, P. Russo, On the determination of the minimum ignition temperature for dust/air mixtures, J. Loss Prev. Process Ind. 20 (2007) 303–309. [4] R. Eckhoff, Dust Explosions in the Process Industries, 3rd ed., Gulf Professional Publishing, Elsevier Science, 2003, pp. 25–141, 251–424. [5] EN 14522, Determination of the minimum ignition temperature of gases and vapours, CEN (2005).
[6] R.H. Essenhigh, M.K. Misra, D.W. Shaw, Ignition of coal particles: a review, Combust. Flame 77 (1989) 3–30. [7] H.M. Cassel, I. Liebman, The cooperative mechanism in the ignition of dust dispersions, Combust. Flame 3 (1959) 467–475. [8] X. Du, K. Animalia, The transient ignition of isolated coal particles, Combust. Flame 97 (1994) 339–354. [9] M. Chen, L.S. Fan, R.H. Essenhigh, Prediction and measurement of ignition temperatures of coal particles, in: Proceedings of the 20th International Symposium on Combustion, The Combustion Institute, Pittsburgh, PA, 1984, pp. 1513–1521. [10] S. Bandyopadhyay, D.B. Bhaduri, Prediction of ignition temperature of a single coal particle, Combust. Flame 18 (1972) 411–415. [11] D. Zhang, T.F. Wall, An analysis of the ignition of coal dust clouds, Combust. Flame 92 (1993) 475–480. [12] C. Engler, Beitrage zur Kenntniss der Staubexplosionen, Chemische Industrie (1885) 171–173. [13] K. Chatrathi, Dust and hybrid explosibility in a 1 m3 spherical chamber, Process Saf. Prog. 13 (1994) 183–189. [14] R. Pilão, E. Ramalho, C. Pinho, Explosibility of cork dust in methane/air mixtures, J. Loss Prev. Process Ind. 19 (2006) 17–23. [15] Y. Liu, J. Sun, D. Chen, Flame propagation in hybrid mixture of coal dust and methane, J. Loss Prev. Process Ind. 20 (2007) 691–697. [16] P.R. Amyotte, K.J. Mintz, M.J. Pegg, Y.H. Sun, The ignitability of coal dust–air and methane–coal dust–air mixtures, Fuel 72 (1993) 671–679. [17] P. Amyotte, M. Lindsay, R. Domaratzki, N. Marchand, A. Di Benedetto, P. Russo, Prevention and mitigation of dust and hybrid mixture explosions, Process Saf. Prog. 29 (2010) 17–21. [18] M. Nifuku, H. Tsujita, K. Fujino, K. Takaichi, C. Barre, E. Paya, et al., Ignitability assessment of shredder dusts of refrigerator and the prevention of the dust explosion, J. Loss Preve. Process Ind. 19 (2–3) (2006) 181–186. [19] O. Dufaud, L. Perrin, M. Traore, S. Chazelet, D. Thomas, Explosions of vapour/dust hybrid mixtures: a particular class, Powder Technol. 190 (2009) 269–273. [20] W.E. Baker, M.J. Tang, Gas, Dust and Hybrid Explosions, Elsevier, 1991, ISBN: 978-0-444-88150-2 http://www.elsevier.com/books/gas-dust-and-hybridexplosions/bakerandampdagger/978-0-444-88150-2. [21] W. Nazaroff, R. Harley, Air Quality Engineering, University of California, Berkeley, 2007, CE 218A. [22] Standard Thermodynamic Properties of Chemical Substances by CRC Press llc (2000). [23] E. Brandes, W. Möller, Sicherheitstechnische Kenngroßn, Brand1:Brennbare Flüssigkeiten und Gase, PTB (2003). [24] EN 50281-2-1, European standard test method for the minimum ignition temperature for dust cloud (2007). [25] M. Hattwig, H. Steen, Handbook of Explosion Prevention and Protection, Wiley–VCH Verlag GMbH & Co. CGAC, 2004, pp. 165, ISBN-13: 978-3527307180, Section 2.3.3.2. [26] M. Encinar, F.J. Beltra, J.F. Gonza, M.J. Moreno, Pyrolysis of maize, sunflower, grape and tobacco residues, J. Chem. Tech. Biotechnol. (1997) 400–410. [27] I. Khalili, O. Dufaud, M. Poupeau, N. Cuervo-Rodriguez, L. Perrin, Ignition sensitivity of gas–vapour/dust hybrid mixtures, Powder Technol. 217 (2012) 199–206. [28] A. Agreda, A. Di Benedetto, P. Russo, E. Salzano, R. Sanchiro, Dust/gas mixtures explosion regimes, Powder Technol. 205 (2011) 81–86, http://dx.doi.org/10. 1016/j.powtec.2010.08.069.205. [29] E.K. Addai, D. Gabel, U. Krause, Explosion characteristics of three component hybrid mixtures, Process Saf. Environ. Prot. 9 (8) (2015) 72–81.
326
E.K. Addai et al. / Journal of Hazardous Materials 301 (2016) 314–326
[30] E.K. Addai, D. Gabel, U. Krause, Lower explosion limit of hybrid mixtures of burnable gas and dust, J. Loss Prev. Process Ind. (2015), http://dx.doi.org/10. 1016/j.jlp.2015.02.014. [31] E.A. Addai, D. Gabel, U. Krause, Ignition and explosion behavior of hybrid mixtures of two and three components in: 4. Magdeburger Brand- und Explosionsschutztag, HS- Magdeburg Stendal: Magdeburg, 2015, ISBN 978-3-00-048960-0, 26. und 27. März 2015, Datenträger. [32] M. Hertzberg, I.A. Zlochower, R.S. Conti, K.L. Cashdollar, Thermo-kinetic transport control and structural microscopic; realities in coal and polymer
pyrolysis and devolatilization: their dominant role in dust explosion, Am. Chem. Soc. Div. Fuel Chem. 32 (3) (1987) 24–41. [33] Kwasi Emmanuel Addai, Dieter Gabel, Ulrich Krause, Minimum ignition temperature of hybrid mixtures of burnable dusts and gases, in: Proceedings of the 25th International Colloquium on the Dynamics of Explosions and Reactive Systems, 2–7 August, Leeds, UK, ICDERS, 2015, insges. 6 S. [34] O. Dufaud, L. Perrin, M. Traoré, Dust/vapour explosions: hybrid behaviours? J. Loss Prev. Process Ind. 21 (4) (2008) 481–484.