SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A SHORT LITERATURE REVIEW
  5. APPARATUS AND EXPERIMENTAL METHODS
  6. EXPERIMENTS AND RESULTS
  7. CONCLUSIONS
  8. LITERATURE CITED

Electrical apparatus in industry for use in the presence of explosive gases must be specially designed to prevent the apparatus from igniting the gas. In the case of flameproof enclosures, any holes and gaps in the enclosure wall must be sufficiently long and narrow to prevent hot combustion products expelled by a gas explosion inside the enclosure from igniting an external explosive cloud. There are tight requirements as to the maximum permissible gap surface roughness. This article describes an experimental study of the influence of severe mechanical and corrosive damage of flame gap surfaces on gap performance for IIB and IIC gases in air, using ethylene and hydrogen as test gases. In agreement with previously published findings for propane (IIA gas), it appeared that even with ethylene and hydrogen gap surfaces can suffer considerable damage without this causing any reduction of gap performance. © 2013 American Institute of Chemical Engineers Process Saf Prog 33: 49–55, 2014


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A SHORT LITERATURE REVIEW
  5. APPARATUS AND EXPERIMENTAL METHODS
  6. EXPERIMENTS AND RESULTS
  7. CONCLUSIONS
  8. LITERATURE CITED

The present article summarizes the work described in the master thesis of Ringdal [1] and Steiner [2]. An unpublished preliminary version of this article was presented by Arntzen et al. [3]. The work is a direct continuation of the quite extensive investigation on IIA gases (propane) summarized by Eckhoff et al. [4] on the basis of the earlier articles by Opsvik et al. [5], Grov et al. [6], and Solheim et al. [7].

As described by Eckhoff [8], several principles are available for design of potentially incendiary electrical apparatuses for use in explosive gas atmospheres to ensure that the apparatuses do not ignite the explosive gas. Flameproof design is one of the options.

The purpose of the present investigation has been to study the influence of various kinds of extensive damage (severe rusting and milled mechanical grooves) of flame-gap surfaces on the ability of the gap to prevent flame transmission, using both ethylene/air (IIB gas) and hydrogen/air (IIC gas) in the experiments. A significant reduction of Maximum Experimental Safe Gap (MESG) compared with that obtained with a standard undamaged surface (roughness <6.3 µm) would mean that the type of damage tested had reduced the gap efficiency significantly, whereas a significant increase of “MESG” would mean that the type of damage tested had in fact significantly improved the gap efficiency.

A SHORT LITERATURE REVIEW

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A SHORT LITERATURE REVIEW
  5. APPARATUS AND EXPERIMENTAL METHODS
  6. EXPERIMENTS AND RESULTS
  7. CONCLUSIONS
  8. LITERATURE CITED

The MESG is the largest width that a gap in an enclosure wall can have to prevent the hot combustion products from an internal explosion igniting an external explosive atmosphere. Phillips [9, 10] developed a theory for predicting MESGs. Phillips considered a spherical chamber consisting of two halves joined by two flanges between which there is a narrow plane circular gap. The chamber is filled with and surrounded by a given explosive gas mixture. The gas mixture inside the chamber is ignited at some point by a small ignition source. The concern is then to identify the maximum gap width between the two flange plates that prevents the hot combustion products expelled through the gap, from igniting the external explosive gas cloud. The sequence of events is illustrated in Figure 1.

image

Figure 1. Two-dimensional model of plane jet of hot combustion gases emerging from a plane flange gap. From Ref [ [12], based on a similar figure in Ref [ [10].

Download figure to PowerPoint

Due to the pressure rise generated by the explosion, an explosion having a narrow gap in its wall will cause a hot jet of combustion products to be expelled through the gap and into the explosive atmosphere surrounding the primary chamber. If the gap width is larger than a certain maximum permissible value, the hot gas jet will ignite the explosive atmosphere outside the chamber.

In this work, ethylene/air and hydrogen/air mixtures were used as test gases because these two fuel gases represent group IIB and IIC gases, respectively.

The maximum gap width for preventing ignition of a mixture of a given fuel gas and mixture depends on the fuel/air ratio. Figure 2 shows some data presented by Beyer [12] based on a report by Redeker [13]. The figure indicates that for ethylene the MESG was found to be about 0.65 mm at a fuel concentration of about 6.5 vol%. For hydrogen, the MESG was about 0.3 mm at a fuel concentration of about 28 vol%.

image

Figure 2. Maximum experimental safe gap S as a function of vol % fuel CB in mixtures of the different fuel gases with air. The minima of the S curves were taken as the MESG. From Ref [ [13].

Download figure to PowerPoint

APPARATUS AND EXPERIMENTAL METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A SHORT LITERATURE REVIEW
  5. APPARATUS AND EXPERIMENTAL METHODS
  6. EXPERIMENTS AND RESULTS
  7. CONCLUSIONS
  8. LITERATURE CITED

The plane-rectangular-slit apparatus (PRSA) was used in all the experiments. This apparatus was first described by Grov et al. [6]. A cross section is shown in Figure 3. The series of flame gaps with seven parallel crosswise-milled grooves used in the present experiments is shown in Figure 4. The desired gap width was obtained by inserting two sets of shims of the desired thickness as spacers between undamaged parts of the two gap surfaces, before clamping them together. In the case of the experiments with rusted gaps, a whole series of permanently assembled gaps with different gap widths was produced and tested for flame transmission both before and after heavy rusting at the sea-side. A gap assembly before rusting is shown in Figure 5. Further details can be found in Opsvik et al. [5], Grov et al. [6], Solheim et al. [7] and Eckhoff et al. [4].

image

Figure 3. Cross section of the plane-rectangular-slit apparatus. Consist of 1 liter primary chamber, a plane flame gap with 25 mm width to a secondary chamber of 3 liters. From Refs [ [6] and [7].

Download figure to PowerPoint

image

Figure 4. Rectangular carbon steel flame gap surfaces with parallel crosswise rectangular grooves milled into both surfaces. Seven parallel grooves of widths and depths as indicated in the figure. From Ref [ [15].

Download figure to PowerPoint

image

Figure 5. A plane rectangular flame gap assembled permanently by two screws. Definitions of gap width, gap length and gap breadth indicated from Ref [2]. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com].

Download figure to PowerPoint

EXPERIMENTS AND RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A SHORT LITERATURE REVIEW
  5. APPARATUS AND EXPERIMENTAL METHODS
  6. EXPERIMENTS AND RESULTS
  7. CONCLUSIONS
  8. LITERATURE CITED

Optimal Fuel Concentrations for External Ignition

Experiments with undamaged gaps were conducted both with ethylene/air and hydrogen/air to determine the most favorable fuel concentrations and distances from the gaps for ignition of the mixture outside the test apparatus. Based on introductory experiments with stoichiometric fuel/air mixtures to find the optimal ignition distances, the distances used in these experiments were 25 mm upstream of the gap entrance for ethylene/air, and 20 mm for hydrogen/air.

Ethylene/Air

Experiments were conducted with a gap width of 0.69 mm, which is slightly larger than MESG for undamaged gaps (0.67 mm). The results are shown in Figure 6. For each fuel concentration tested, 10 successive apparently identical experiments were carried out. The numbers of ignitions obtained for each concentration are given as columns in the diagram. As can be seen, the largest number of reignitions was obtained with a mixture of 6.7 vol% ethylene in air, which is in close agreement with the optimal concentration for ethylene shown in Figure 2. This concentration was therefore used in all subsequent experiments with ethylene/air.

image

Figure 6. Results from determination of optimal ethylene concentration for reignition. From Ref [ [2]. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com].

Download figure to PowerPoint

Hydrogen/Air

Also hydrogen/air experiments were conducted with a gap width slightly larger than MESG for undamaged gaps, in this case 0.31 mm. The results are shown in Figure 7. Again, 10 successive apparently identical experiments were carried out for each fuel concentration tested. The numbers of external ignitions obtained for each concentration are given as columns in the diagram. As can be seen, the largest number of ignitions was obtained with a mixture of 30.5 vol% hydrogen in air, which is in close agreement with the optimal concentration for hydrogen shown in Figure 2. This concentration was therefore used in all subsequent experiments with hydrogen/air.

image

Figure 7. Results from determination of optimal hydrogen concentration for reignition. From Ref [ [1]. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com].

Download figure to PowerPoint

Optimal Ignition Point Location for External Ignition

Ethylene/Air

The results are summarized in Figure 8. With undamaged flame gaps the optimal ignition position in the primary chamber for obtaining ignition in the secondary chamber was 25 mm upstream of the gap entrance. The corresponding maximum gap width for preventing ignition was 0.67 mm, which was then taken as the MESG for ethylene/air as obtained in the apparatus illustrated in Figure 2. However, in the case of the gap surfaces with seven milled crosswise grooves shown in Figure 4, the optimal ignition position for ignition was found to be only 5 mm upstream the gap entrance, as also indicated in Figure 7. This is considerably shorter than the distance of 25 mm found for the undamaged gap. A comprehensive explanation for this difference has not yet been found.

image

Figure 8. Maximum experimental safe gaps, that is, the largest gaps that gave no reignitions in 10 repeated apparently identical experiments, for various distances between ignition point and gap entrance (ignition distances). Comparison of results obtained with undamaged gaps and with gaps with milled grooves of configuration PH-7.2.3, that is, seven crosswise rectangular grooves of breaths 2 mm and depths 3 mm. The data for propane are from Refs [ [12] and [15]. Adjusted figure from Ref [ [2]. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com].

Download figure to PowerPoint

Hydrogen/Air

The results for undamaged gaps with hydrogen/air are shown in Figure 9. A mixture of 30.5 vol% hydrogen and 69.5 vol% air was used in all the experiments. As Figure 9 shows, the optimal ignition point in the primary chamber for having ignition in the secondary chamber was 20 mm upstream of the flame gap entrance. As also shown by Ringdal [1], the optimal ignition distance for ignition of hydrogen/air was 20 mm even with the gap surfaces with seven multiple crosswise grooves shown in Figure 4. This is contrary to the findings with ethylene/air, for which the optimal ignition distance with such grooves in the gap surfaces were considerably shorter than for undamaged surfaces, as shown in Figure 8.

image

Figure 9. Reignition frequency as a function of ignition distance (distance from ignition point to entrance of flame gap) for 30.5 vol% hydrogen in air. 10 apparently identical experiments per ignition distance. Undamaged gap surfaces. Figure from Ref [ [1]. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com].

Download figure to PowerPoint

MESG for Gap Surfaces with 7 Crosswise Grooves of Different Depths

Ethylene

The results for ethylene/air are given in Table 1, which gives experimentally determined MESGs with gap surfaces having seven parallel grooves milled into both surfaces.

Table 1. Experimentally determined MESGs for premixed 6.7 vol% ethylene in air, with gap surfaces having seven parallel grooves milled into both surfaces (Figure 4).
Groove Depth (mm)Ignition Point Location (mm)MESG (mm)Pmax (bar, g)
  1. Pmax is the arithmetic mean of the maximum explosion pressures recorded in each series of 10 successive apparently identical experiments with the gap equal to MESG. From Ref [ [2].

350,703,70
250,713,55
150,703,47
0.550,673,56
Undamaged50,713,29
Undamaged250,673,11

The MESG values are the largest gap widths that did not give any external ignitions in a series of 10 successive apparently identical experiments. A photo of the entire selection of gap assemblies used in the experiments is given in Figure 4. The groove width was 2 mm throughout, whereas the groove depths varied between 0.5 and 3 mm. Table 1 shows that none of the MESG values obtained with milled grooves in the gap surfaces were smaller than the smallest opening of 0.67 mm found with an undamaged gap and with ignition at the optimal point 25 mm upstream of the gap entrance. It can be concluded, therefore, that none of the groove configurations tested deteriorated the ability of the flame gap to prevent ignition of the ethyle/air mixture downstream of the gap. The majority of the results in Table 1 in fact confirm that the grooves improved the performance of the gaps even for ethylene/air (IIB gases). This is in accordance with the findings by Grov et al. [6] and Solheim et al. [7] for propane/air (IIA gases).

Table 1 also shows that the maximum explosion pressure in the primary chamber, Pmax, was significantly lower with undamaged gaps than with gaps with grooves in the surfaces. Again this is in agreement with the findings of Grov et al. [6] and Solheim et al. [7] for propane/air. A likely reason is that the flow resistance through the gap increased, and hence, the venting efficiency of the gap decreased, when this type of grooves had been milled into the gap surfaces.

Table 2 shows the gas temperature at 20 mm downstream of flame gap exit following ethylene/air explosions in the primary explosion vessel. The reason for the high temperature obtained with a groove depth of 0.5 mm is that with the particular ignition distance used, this groove depth gave ignition downstream of the gap, whereas with the other depths this was not the case. Table 2 clearly shows that, in the absence of ignition, all the gaps with grooves gave a lower mean gas temperature at the measurement point, as compared to the undamaged gap.

Table 2. Means of max. gas temperatures at 20 mm downstream of gap exit in ten apparently identical experiments. Premixed 6.7 vol% ethylene in air. From Ref [2].
Groove Depth (mm)Mean Gas Temp. Outside Gap Exit (oC)
  1. Seven parallel crosswise rectangular grooves of widths 2 mm and depths as given in table, in both surfaces. Gap widths 0.70 mm throughout between undamaged parts of surfaces. From Ref [2].

382
280
1100
0,5294
Undamaged gap170
Hydrogen

Table 3 gives experimentally determined MESGs with gap surfaces having seven parallel grooves milled into both surfaces, as shown in Figure 4. The ignition point distance was kept at 20 mm throughout. The MESG values are the largest gap widths that did not give any external ignitions in a series of 10 successive apparently identical experiments. Table 3 shows that all the MESG values obtained with milled grooves in the gap surfaces were larger than the MESG of 0.29 mm found with an undamaged gap. Therefore, it can be concluded that all the groove configurations tested in fact improved the ability of the flame gap to prevent reignition downstream of the gap. This is in accordance with the findings by Grov et al. [6] and Solheim et al. [7] for propane/air (IIA gases).

Table 3. Experimentally determined MESGs for premixed 30.5 vol% hydrogen in air. From Ref [1].
Groove Depth (mm)MESG (mm)Pmax (bar, g)
  1. Seven parallel rectangular grooves of widths 2 mm milled into both surfaces. Pmax is the arithmetic mean of the maximum explosion pressures recorded in each series of ten successive apparently identical experiments with the gap equal to MESG. From Ref [1].

30,332,49
20,322,55
10,332,23
0,50,342,62
No grooves (undamaged gap)0,292,52

Table 3 also gives the maximum explosion pressures in the primary chamber, Pmax. The reason for the comparatively low value found for 1 mm groove depth has not yet been determined. The other maximum pressures measured were rather equal, which indicates that with such small gap widths, the venting of the explosion in the primary chamber through the gap is of minor importance for the maximum pressure obtained. Cooling of the combustion products by the vessel wall, rather than venting, is the main reason for the maximum pressures being lower than the adiabatic constant-volume explosion pressure.

As in the case of ethylene/air, the temperature of the combustion gases exiting the flame gap were measured 20 mm downstream of the gap exit. Even with hydrogen/air it was found that the temperature was systematically lower with multiple crosswise grooves in the gap surfaces, as compared to undamaged surfaces.

Heavily Rusted Gap Surfaces

An example of a heavily rusted gap assembly is shown in Figure 10. In the case of heavy rusting, the method of assembling the gap by shims and clamping after the rusting of the gap surfaces had taken place, had to be abandoned due to the swelling and compressibility of the rust layers. Therefore, as discussed by Solheim [14] and Solheim at al. [7], all PRSA flame gaps to be exposed to rusting were assembled firmly and permanently, as shown in Figure 5, before rusting. The heavy rusting was obtained by hanging gap assemblies of various fixed gap widths outdoors at the sea-side (salt water), midway between high and low tide, for various lengths of time.

image

Figure 10. The same type of permanently assembled flame gap as shown in Figure 5 after 2 months of outdoor exposure to salt sea water. From Ref [1]. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com].

Download figure to PowerPoint

Ethylene

Six heavily rusted gap assemblies (Figure 10), with gap widths before rusting ranging from 0.68 to 0.75 mm, were used in flame transmission experiments with premixed ethylene/air in the apparatus shown in Figure 3. The concentration of ethylene in the mixture was 6.7 vol% throughout. None of the rusted assemblies that had given no ignitions in 10 repeated tests in the original unrusted state gave any ignitions in ten repeated tests. The largest gap width of 0.75 mm tested, which gave 10 ignitions prior to rusting, gave six ignitions after rusting. However, none of the rusted gaps, not even that of width 0.75 mm, gave reignition in the very first test in the series of 10. This is a significant finding, because this very first test is the most important one as regards practical relevance. The reason is that in practice any flame gap in an Ex-d enclosure in industry would only be exposed to just one single internal accidental gas explosion, if any at all, during its lifetime.

As shown in Figure 11, the maximum pressure obtained decreased with the number of experiments performed with the same rusted flame gap. This is because the rust inside the gap is partly porous and hence partly blocks the gap. During the first repeated experiments in the series of 10, the loose rust is blown away by the explosion, and the cross sectional area of the gap opening, that is, the explosion vent area, increases. An example is shown in Figure 12. The largest drop in maximum pressure occurred between explosion tests 1 and 2, presumably because most of the loose rust was blown out during the first experiment. Smaller amounts of rust were apparently also blown out in the subsequent five tests, whereas for the final four tests the explosion venting efficiency remained unchanged.

image

Figure 11. Maximum explosion pressure in primary chamber in 10 successive ethylene/air explosions with rusted gap assembly. Gap width before rusting 0.72 mm. From Ref [ [2].

Download figure to PowerPoint

image

Figure 12. Heavily rusted gap assembly, with 1.01 mm initial gap width before rusting, prior to the very first flame transmission experiment and after a series of 10 successive experiments. From Ref [ [15]. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com].

Download figure to PowerPoint

Hydrogen

A total of 12 permanently assembled flame gaps of the type shown in Figure 5, with various gap widths in the range from 0.25 mm to 0.35 mm were exposed to outdoor rusting conditions at the at sea-side (salt water); six of them for 1 month, and the other six for 2 months. A totality of about 300 explosion tests were performed with the 12 rusted gap assemblies. Premixed hydrogen/air with 30.5 vol% hydrogen and ignition at 20 mm upstream of the gap entrance was used throughout. The results are given in Table 4.

Table 4. Comparison of external ignition frequencies with gaps before rusting, after 1 month of rusting, and after 2 months of rusting.
Gap Width Before Rusting (mm)Number of Ignition Before RustingNumber of Ignitions 1 Month RustingNumber of Ignitions 2 MonthsRusting
  1. Premixed hydrogen/air with 30.5 vol% hydrogen. From Ref [1].

0.25000
0.28000
0.29000
0.30000
0.311000
0.351040

No external ignitions occurred with initial flame gaps between 0.25 mm and 0.30 mm. With the gaps before rusting, the MESG was found to be 0.30 mm. Both with gap width 0.31 mm and 0.35 mm, all 10 successive experiments gave ignitions. After 1 month of rusting, only the four last tests of the series of 10 with the largest gap width of 0.35 mm before rusting, gave ignitions. After 2 months of rusting, the corresponding number of ignitions was zero. Hence, even with this largest gap width before rusting, the first experiment in the series of 10 did not give flame transmission. This means that rusting in fact improved gap performance significantly. Furthermore, gap performance seemed to improve with the length of the period of rusting.

As shown in Figure 13, the maximum pressure obtained decreased with the number of experiments performed with the same rusted flame gap even for hydrogen/air. In this case, it may seem as, for some unknown reason, most of the rust is blown away in two stages; viz. in experiments 1, 2, and 3, and in experiments 9 and 10.

image

Figure 13. Maximum explosion pressure in primary chamber in 10 successive hydrogen/air explosions with rusted gap assembly. Gap width before rusting 0.29 mm. From Ref [ [1].

Download figure to PowerPoint

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A SHORT LITERATURE REVIEW
  5. APPARATUS AND EXPERIMENTAL METHODS
  6. EXPERIMENTS AND RESULTS
  7. CONCLUSIONS
  8. LITERATURE CITED
  1. The current IEC and European standards for flameproof apparatuses require that the mean roughness of the flame gap surfaces be less than 6.3 µm. Implicitly, the standards also require that the flame gaps be restored to their original quality if in some way significantly damaged. However, no guidance as to what extent of damage is to be regarded as significant is provided by the standards. Consequently, flameproof enclosures with only minor visible damage on the flame gap surfaces may be subject to overhaul or even replacement, without any documentation of the real necessity of this.
  2. In this study, premixed ethylene/air and hydrogen/air were used as the experimental gas mixtures throughout. Hence, the results obtained in the present investigation should be valid for gases and vapours belonging to gas groups IIB and IIC.
  3. Two main series of experiments were carried out, that is, experiments with milled multiple crosswise grooves in the gap surfaces (severe mechanical damage) and experiments with heavily rusted flame gap surfaces. A flexible “MESG” concept was adopted to assess whether the gap efficiency increased or decreased, due to the different types of damage of the gap surface. If the “MESG” had increased, gap performance had improved, and vice versa.
  4. For ethylene/air, the optimal distance between the ignition source in the primary chamber and the gap entrance, for ignition in the secondary chamber, was 25 mm for undamaged gap surfaces, and just 5 mm for severely damaged gaps (seven milled parallel crosswise grooves). For hydrogen/air, the optimal distance was 20 mm for both undamaged gap surfaces and for surfaces with seven milled parallel crosswise grooves.
  5. In the case of flame gaps with milled crosswise grooves in the gap surfaces, gap performance mostly improved significantly compared with that of undamaged gap surfaces. This is also in agreement with the corresponding results published earlier for propane/air (IIA gases).
  6. Severely rusted gap surfaces in fact prevented external ignition more effectively than the same gaps before rusting, for both ethylene/air and hydrogen/air.
  7. Both with ethylene/air and hydrogen/air the maximum explosion pressures in the primary chamber were significantly higher with crosswise grooves than with undamaged plain surfaces of roughness that was less than 6.3 µm. It is suggested that this is due to the increased flow resistance in the gap produced by the grooves.
  8. Both with ethylene/air and hydrogen/air the mean exhaust gas temperatures 20 mm downstream of the gap in 10 successive tests were significantly lower with crosswise grooves than with undamaged plain surfaces of roughness that was less than 6.3 µm. A possible reason could be more efficient entrainment of cold air by the hot gas jet emerging from the flame gap, due to increased jet turbulence.
  9. Perhaps the IEC could discuss whether the comparatively simple apparatus used in the present investigation may also be adapted to determination of standard MESG values for gases and vapours, along with the more complicated apparatus [15] currently used as an IEC standard for this purpose.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. A SHORT LITERATURE REVIEW
  5. APPARATUS AND EXPERIMENTAL METHODS
  6. EXPERIMENTS AND RESULTS
  7. CONCLUSIONS
  8. LITERATURE CITED
  • 1
    L. Ringdal, An Experimental Investigation of the Influence of Mechanical Damage and Rust on the Ability of Flame Gaps to Prevent Hydrogen Gas Explosion Transmission, Master thesis, Department of physics and technology, University of Bergen, Norway, June 2012.
  • 2
    M. Steiner, An Experimental Investigation of the Effect of Rust and Mechanical Damage on the Maximum Experimental Safe Gap for Ethylene Gas Explosions, Master thesis, Department of physics and technology, University of Bergen, Norway, June 2012.
  • 3
    B.J. Arntzen, L. Ringdal, M. Steiner, R.K. Eckhoff, Effects of Mechanical Damage and Rusting of Flame Gap Surfaces in Flameproof Electrical Apparatus for IIB and IIC Gases. Paper presented at the IX International Symposium on Hazards, Prevention and Mitigation of Industrial Explosions, Krakow, Poland, July 21–26, 2012, Available electronically from the organizer Central Mining Institute, Plac Gwarków 1, 40166 Katowice, Poland.
  • 4
    R.K. Eckhoff, B.J. Arntzen, H.E.Z Opsvik, A. Grov, F. Solheim, Is the safe performance of flame gaps in flameproof electrical apparatus deteriorated by rusting and mechanical damage? Part 1: Group IIA gases. Process Saf Prog, 32 (2013), 4956.
  • 5
    H.E.Z. Opsvik, A. Grov, R.K. Eckhoff, MESG for Propane/Air in Standard Circular-Flange Experiments. Influence of Sandblasting and Corrosion of Flame Gap Surfaces. Proceedings of 6th Intenational Seminar on Fire and Explosion Hazards, Leeds, UK, April 11–16, 2010.
  • 6
    A. Grov, H.E.Z. Opsvik, R.K. Eckhoff, Effects of significant damage of flame gap surfaces in Flameproof electrical apparatus on flame gap efficiency, J Loss Prev Process Ind 24 (2011), 552557.
  • 7
    F. Solheim, B.J. Arntzen, R.K. Eckhoff, Effect of rusting and mechanical damage of gap surfaces on efficiency of flame gaps in flameproof electrical apparatus, Process Saf Environ Protect 90 (2011), 317325.
  • 8
    R.K. Eckhoff, Explosion hazards in the process industries, Gulf Publishing Company, Houston, TX, USA, 2005. ISBN 0-9765113-4–7.
  • 9
    H. Phillips, The Mechanism of Flameproof Protection, Research Report No. 275, Safety in Mines Research Establishment (SMRE), UK, 1971.
  • 10
    H. Phillips, The Physics of the Maximum Experimental Gap, Proceedings of International Symposium On the Explosion Hazard Classification of Vapors, Gases and Dusts. National Materials Advisory Board, National Research Council. Publication NMAB-447, National Academy Press, Washington DC, 1987.
  • 11
    A. Grov, An experimental study of the influence of major damage of flame gap surfaces in flameproof apparatus on the ability of the gaps to prevent gas explosion transmission, Master thesis, Department of physics and technology, University of Bergen, Norway, June 2010.
  • 12
    M. Beyer, Über den Zünddurchschlag explodierender Gasgemische an Gehäusen der Zündschutzart, Druckfeste Kapselung, PhD thesis, Technical University Carolo-Wilhelmina, Braunschweig, Germany, VDI Fortshrittberichte, Reihe 21: Elektrotechnik, Nr. 228, VDI-Verlag GmbH, Düsseldorf, Germany, 1997.
  • 13
    T. Redeker, Classification of flammable gases and vapours by the flameproof safe-gap and the incendivity of electrical sparks, PTB-Bericht W-18, Physikalisch-Technische Bundesanstalt, Braunschweig, Germany, 1981.
  • 14
    F. Solheim, An experimental investigation of the influence of mechanical damage, rust and dust on the ability of flame gaps to prevent gas explosion transmission, Master thesis, Department of physics and technology, University of Bergen, Norway, November 2010.
  • 15
    International Standard IEC 60679-20-1, Explosive atmospheres – part 20-1: Material Characteristics for gas and vapour classification – Test Methods and Data, International Electrotechnical Commission, Geneva, 2010.