Safety S torage tanks and other vessels containing fammable and/or explosive chemicals and mixtures are prevalent in chemical process industries (CPI) plants. Flame arresters are installed on such equipment to prevent potential hazards associated with fammable and explosive materials. For example, fame arresters can stop the spread of a fre, limit the spread of an explosive event, protect potentially explosive mixtures from igniting, and confne a fre within an enclosed, controlled location. A fame arrester works by forcing a fame front through very narrow channels. Gas travels through the device, but the passages are so narrow that the fame can no longer be maintained. This article provides an overview of the use of fame arresters in the CPI. It defnes some key terminology, describes the different types of fame arresters, explains how fame arresters are tested and certifed for specifc process conditions, discusses the concept of layers of protection and how a fame arrester can serve as a layer, and offers guidance on choosing a fame arrester. The article also explains why conservation vents do not provide reliable protection against atmospheric explosions (sidebar, p. 18), and why fame arrester elements made of polytetrafuoroethylene (PTFE) can extinguish a fame rather than melt or burn (sidebar, p. 20). Key terminology International standards (10) defne the combustion processes relevant to fame arresters. Explosion is the generic term for an abrupt oxida- tion or decomposition reaction that produces an increase A ame arrester may be installed to keep an external ame out of a tank or vessel or to ensure that a ame burning within a pipeline does not propagate to other equipment. Use the information presented here to determine what type of device is appropriate for your application. Michael Davies Thomas Heidermann Braunschweiger Flammenlter GmbH PROTEGO Protect Your Process with the Proper Flame Arresters p Figure 1. An atmospheric deagration occurs in the open air. p Figure 2. A pre-volume deagration takes place within a conned volume. p Figure 3. Stabilized burning is an even, steady ame that is stabilized at or near the ame arrester element. Copyright 2013 American Institute of Chemical Engineers (AIChE) CEP December 2013 www.aiche.org/cep 17 in temperature, pressure, or both simultaneously. Defagration is an explosion that propagates at subsonic velocity. The geometric shape of the combustion volume determines whether it is: an atmospheric defagration, which occurs in open air without a noticeable increase in pressure (Figure 1) a pre-volume defagration, which is initiated by an internal ignition source and occurs in a confned volume, such as within a vessel (Figure 2) an inline defagration, which is an accelerated explo- sion within a pipe that travels along the axis of the pipe at the fame-propagation speed. Stabilized burning is the even, steady burning of a fame that is stabilized at or near the fame arrester element (Fig- ure 3). A distinction is drawn between short-term burning (stabilized burning for a specifc time period) and endurance burning (stabilized burning for an unlimited time). Detonation is an explosion that propagates at supersonic velocity and is characterized by a shock wave. A distinction is drawn between a stable detonation, which has no signif- cant velocity or pressure variation, and an unstable detona- tion, which represents the transition of a combustion process from a defagration into a stable detonation. Figure 4 illustrates the progression of an inline explo- sion. A pipe flled with an explosive gas is ignited at one end, and the fame propagates from the ignition source to the other end of the pipe. Initially, in the frst section of the pipe, the fame-front velocity and the explosion pressure are low; this is a defagration. The velocity of the combustion wave and the explosion pressure increase rapidly and the defa- gration becomes an unstable detonation. When the velocity and pressure decline, the event becomes a stable detonation. Independent of the length of the pipe, the stable detona- tion propagates at supersonic velocities through the pipe. Flame arrester location in the process Flame arresters are classifed according to their location relative to the equipment they are designed to protect. End-of-line fame arresters (Figure 5a) are located directly on a vessel or tank vent nozzle, or on the end of a vent line from the vent nozzle. They are usually defa- gration fame arresters, and are commonly installed on atmospheric-pressure storage tanks, process vessels, and transportation containers. If the vented vapors are ignited, for example by lightning, the fame arrester will prevent the fame from spreading from the outside atmosphere to the inside of the vessel. Inline fame arresters are installed in piping systems to protect downstream equipment from either defagrations (Figure 5b) or detonations (Figure 5c). The choice between b c a p Figure 4. An inline explosion begins as a deagration, turns into an unstable detonation, and then becomes a stable detonation. L L Accelerating Deflagration Unstable Detonation Transition Zone D Stable Detonation VP Velocity of the Flame Front Pressure Buildup Caused by Volume Expansion of the Burned Mixture u Figure 5. Flame arresters are classied as end-of-line (a) or inline based on the location relative to the equipment they protect, and they may protect against deagrations (b) or detonations (c). Copyright 2013 American Institute of Chemical Engineers (AIChE) 18 www.aiche.org/cep December 2013 CEP Safety A conservation vent is a pressure/vacuum relief valve mounted on the roof of a storage tank to allow for tank breathing, thereby minimizing emissions and conserving product. A weighted or spring-loaded pallet within the valve keeps the vent closed as long as the pressure in the tank remains below the valves setpoint. An increase in the pres- sure or vacuum causes the pallet to begin to lift off its seat, relieving the excess pressure (by allowing a small amount of vapor to exit the tank) or vacuum (by allowing air to enter the tank). An overpressure of 1060% is usually required to achieve full lift of the pallet. For several decades, government agencies and engi- neering societies have published guidelines for the design and safe management of storage tanks. Even though these guidelines are based on the latest research and engineering practices at the time they were developed, conicts exist among some of the standards. For example, the 5th edition of API Standard 2000, Venting Atmospheric and Low-Pressure Storage Tanks (1), states that a ame arrester is not considered necessary for use in conjunction with a pressure/vacuum valve venting to atmosphere, because ame velocities are lower than vapor velocities across the seat of the pressure/vacuum valve. On the other hand, the German TRbF 20 standard, Technical Regulations for Flammable Liquids (2), calls for ame arrest- ers if the tank contains liquids that can create an explosive atmosphere in other words, any liquid that is ammable. According to the United Nations Recommendations on the Transport of Dangerous Goods, Model Regulations (3) and the Globally Harmonized System of Classication and Label- ing of Chemicals (GHS) (4), a liquid with a ashpoint lower than 60C is considered a potential source of ammable vapor inside a tank. Nevertheless, vents are a likely place of ignition, and it is a best practice to install ame arresters to prevent tank explosions (5). These inconsistencies among glob- ally recognized standards and publica- tions prompted research to determine whether a conservation vent can truly assure ame transmission through the vent pallet to prevent ashback and thus explosions. During the development of ISO 28300: Petroleum, Petrochemical, and Natural Gas Industries Venting of Atmospheric and Low-Pressure Storage Tanks (6), vents were tested according to the ISO 16852 ame arrester standard (7), and the results are incorporated into ISO 28300, as well as its equivalent, API 2000, 6th Ed. (8). Two different tests were conducted: an atmospheric defagration test, to investigate whether a conservation vent would fail if a vapor cloud in the vicinity of the vent is ignited, for example by a lightning strike a high-velocity discharge test, to investigate whether the theoretical basis of engineering guidelines such as API 2210, Flame Arresters for Vents of Tanks Storing Petro- leum Products (9), is correct in stating that ashback through the conservation vent is not possible because the discharge velocity is always above the ame velocity. For the atmospheric deagration test, an explosionproof vessel is lled with a ammable air/fuel mixture and vented through a 4-in. (DN 100) conservation vent into a plastic bag until the bag is completely full. A chemical igniter with a low ignition energy is used to ignite the air/fuel mixture approxi- mately 1 m above the point where the valve is connected to the vessel. If the conservation vent cannot prevent ash- back, the ame will propagate through the pressure/vacuum valve; an explosion will occur inside the vessel, the vessel diaphragm will burst, and ames will propagate out of the vessel. Five manufacturers conservation vents, each set at +10 mbar (+4.0 inH 2 O) and 2 mbar (0.8 inH 2 O), were tested using propane (with explosive limits of LEL = 1.7 vol% and UEL = 10.8 vol%) at three concentrations (4.2 vol%, 5.5 vol%, and 6.0 vol%) as the fuel. In all of the tests, all of the conservation vents failed to prevent ame propagation into the vessel, the bursting diaphragm ruptured, and a large re ball propagated out of the vessel. The high-velocity discharge test is also conducted in an explosionproof vessel with a conservation vent installed on top of the vessel. A stoichiometric fuel/air mixture (4.2 vol% propane in air) is fed into the vessel and allowed to escape through the pressure side of the conservation vent. A pilot burner installed near the vent serves as an ignition source. The rst tests were performed using a volumetric ow- rate of 85 m/h of the explosive gas mixture. Upon ignition of the vapor cloud, the ame stabilized at the valve seat; after a few seconds, the ame propa- gated through the gap between the seat and the pallet, causing an explosion inside the vessel; the vessels rupture panel broke, and a re ball propagated to the outside of the vessel. The second series of tests was conducted at a higher volumetric owrate (100 m/h). Again, just a few seconds after ignition, a ash- back was detected. These tests demonstrated that conservation vents do not act as ame arresters. These pressure/vacuum relief valves are not able to stop a ame. The ow of a typical pressure valve is not high, nor is it stable enough to stop the propagation of a ame. A conservation valve should not be used as a ame arrester, unless the device is approved according to ame arrester standards. A Conservation Vent is not a Flame Arrester Copyright 2013 American Institute of Chemical Engineers (AIChE) CEP December 2013 www.aiche.org/cep 19 a defagration arrester and a detonation arrester depends on the distance between the potential ignition source and the arrester (called the run-up distance, and designated L), because this length rela- tive to the pipe diameter (L/D) infuences the transition of a defagration to a stable detonation. Inline defagration fame arresters should be used only where the L/D ratio is less than 50 for hydro- carbon/air mixtures and less than 30 for hydrogen/air mixtures; detonation fame arresters are used with longer lengths or where the location of the ignition source is not known. A third type of arrester, the pre- volume defagration fame arrester, is used in some applications to protect the surroundings against the negative impacts of an internal vessel explosion. Assessing the application Flame arresters are designed for particular operating conditions and should be selected for the intended use. To ensure that an effective fame arrester is selected, the conditions under which it will be used must be evaluated: What should be protected? What kind of explosive atmosphere can occur? Where may an explosive atmosphere arise? Where are the potential ignition sources? What are the operating parameters of the process (tem- perature, pressure, etc.)? Maximum experimental safe gap If a fammable gas or mixture enters (or is ignited in) a narrow gap between two plates, the fame propagates through the gap in the direction of the unburned gas (Fig- ure 6). Heat is transferred from the fame front to the walls of the gap; the narrower and the longer the gap, the larger the infuence of the cold wall on the fame profle. With an appropriately sized gap, this heat transfer extinguishes the fame and prevents fashback. This is a fame arresters principle of operation. The appropriate gap size depends on the reactivity of the gas/air mixture. Each gas or vapor is associated with a maxi- mum fame-quenching gap, called the maximum experi- mental safe gap (MESG). The MESG is an experimentally determined property of the gas mixture. It is defned as the maximum clearance between two parallel, 25-mm-long metal surfaces that has been found, under specifed test con- ditions, to prevent an explosion in a test chamber from being propagated to a secondary chamber containing the same gas or vapor at the same concentration. Figure 7 provides MESG values for common gases. Explosion groups MESG values are one means of categorizing gases based on their capacity to produce a fashback. (The other, minimum ignition current ratio relative to methane [MIC] is not discussed here.) Table 1 shows the National Electrical T W T F Heat Transfer from Flame to Wall Flame Front Unburned Gas D T W << T F p Figure 6. A ame arrester transfers heat from the ame front to the walls through the boundary layer. M a x i m u m
E x p e r i m e n t a l
S a f e
G a p
( M E S G ) ,
m m Gas or Vapor Concentration in Air, vol% 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 n-Hexane 2 Methyl isobutyl ketone 3 Ethyl acetate 4 Ethane 5 Acetaldehyde 6 Methane 7 Hydrogen sulfide 8 Propylene oxide 9 Dioxane 10 Ethylene 11 Acetylene 12 Carbon disulfide 1 2 3 4 5 6 7 8 9 10 11 12 p Figure 7. Each gas has a maximum experimental safe gap (MESG). Table 1. Gases are classied into explosion groups based on their maximum experimental safe gap (MESG). Maximum Experimental Safe Gap, mm IEC Group NEC Group Reference Material MESG = 1.14 I Methane > 0.9 IIA D Propane 0.50.9 IIB C Ethylene < 0.5 IIC B Hydrogen Copyright 2013 American Institute of Chemical Engineers (AIChE) 20 www.aiche.org/cep December 2013 CEP Safety Code (NEC) and International Electrotechnical Commission (IEC) groups as a function of MESG. These groups form the basis for selecting fame arresters. Flame arresters are tested with mixtures of vapors/gases in the different explosion groups, and based on the results of those tests are approved for a specifc vapor group IIA (NEC Group D), IIB (NEC Group C), or IIC (NEC Group B). The engineer then chooses a fame arrester with the appropriate approval level. For example, hexane has an MESG of 0.93 mm, making it a Group IIA vapor, so a Group IIA-certifed defagration arrester should be used. Arresters approved for one vapor group can be used for a lower group (i.e., mixtures with a higher MESG), but arresters approved for a lower group are not suitable for a higher vapor group. For instance, a device certifed for Group IIC can be used for Group IIB or IIA mixtures as well, but not the reverse. Common practice is to select fame arresters that have been approved based on MESG tests at standard atmo- spheric conditions, without regard for the effects of pres- sure, temperature, and oxygen concentration on the reactiv- ity of the explosive gas mixture. Flame arresters tested at ambient conditions can be used at operational temperatures up to 60C (140F) and operational pressures up to 1.1 bar absolute (15.95 psia). If the working conditions are above these values, fame arresters that have been specifcally tested for the actual conditions are needed. Figures 8 and 9 illustrate the effect of pressure and temperature on MESG. Layers of protection CPI plant safety relies on layers of protection i.e., independent measures that reduce the likelihood of an adverse event or the consequences of such an event should it occur, by control, prevention, or mitigation. Layers of protection include basic process monitoring and automation systems, alarms to trigger operator intervention, emergency shutdown systems, relief valves and rupture disks, contain- ment dikes, and emergency response procedures. Flame p Figure 8. As pressure increases, the maximum safe gap gets smaller. H 2 C 2 H 4 C 6 H 14 CH 4
G a p
W i d t h ,
m m 0.8 1.6 2.4 3.2 3.6 2.4 1.2 0 1/P, bar 1
p Figure 9. As temperature increases, the maximum safe gap gets smaller. Mixture Temperature, C G a p
W i d t h ,
m m 1.1 1.0 0.9 0.8 0.65 0.55 0.35 0.25 20 60 100 140 180 220 260 CS 2 H 2 C 2 H 4 C 6 H 14 CH 4
How can Flame Arrester Elements Made of PTFE Extinguish a Flame? S tatic ame arresters operate by cooling a ame as it ows through the arrester elements to such an extent that the temperature of the gas mixture falls below the igni- tion point on the side being protected. In other words, the rate of heat diverted through the boundary layer to the ame arrester element is much greater than the heat added by the combustion reaction. But the ame arrester element itself experiences very little warming, because it is subjected to a high temperature for an extremely short period of time a few microseconds. Heat transfer is almost exclusively by convection/diffusion. Conduction through the element occurs only after the ame has been dissipated. This occurs more slowly in polytetra- uoroethylene (PTFE) than in stainless steel because of the formers lower thermal conductivity. Thus, regardless of the material of construction PTFE or stainless steel the ame arrester element does not experience any serious warming. However, PTFE does provide protection against corrosion and plugging due to polymerization, making ame arrest- ers made of PTFE a suitable alternative in applications subject to such issues. Copyright 2013 American Institute of Chemical Engineers (AIChE) CEP December 2013 www.aiche.org/cep 21 arresters are commonly employed as one of these layers of protection. Hazardous areas in a CPI plant are clas- sifed into zones according to the frequency and duration of the presence of an explosive atmosphere: Zone 0 is an area where an explosive atmosphere is always present, or is frequently present for long periods of time. Zone 1 is an area where an explosive atmosphere is sometimes present. Zone 2 is an area where an explosive atmosphere is rare or is present only for short periods of time. Similarly, an ignition source may be present always, sometimes, rarely, or never. Taken together, these two factors deter- mine how many layers of protection must be provided to guard against fame transmission, as shown in Table 2. The selection of those safety measures depends on the specifc process. Figure 10 depicts a waste-gas incinerator classifed as Zone 0, because the combustor operates around the clock and is a permanent ignition source. Thus, according to Table 2, three independent protection measures are required: 1. The frst layer of protection is fow control on the air/nitrogen feed lines to the burner inlet nozzle. This main- tains fow to the burner in case the waste gas fowrate drops below the low-fow setpoint. 2. The second layer of protection is an inline defagra- tion fame arrester installed as close as possible to the ignition source. This should be a temperature-monitored defagration fame arrester that can detect a stabilized fame on the fame-arresting element. 3. The third layer of protection is a detonation fame arrester installed in line with the potential ignition source. Monitoring of the oxygen concentration in the incoming waste gas (not shown) could be added as an option. These three independent measures ensure a high level of safety. Choosing an effective ame arrester To select the appropriate fame arrester for your application, provide the following information to the manufacturer: 1. Service. Briefy describe the intended use for the fame arrester. 2. Analysis of gases or vapors. Supply details of all fammable and nonfammable components so that the explo- sion group can be identifed and the correct fame arrester design and materials of construction can be selected. 3. Molecular weight or density of the gas or vapor. This will allow an equivalent air fowrate to be calculated for pressure-drop determination. 4. Flowrate. This should be stated in volumetric terms, or suffcient information should be provided to allow a volumetric fowrate to be calculated. For storage tank applications, give the inbreathing and outbreathing require- ments, or suffcient information on the tank type, pressure resistance shape, dimensions, and flling and emptying rates to allow these parameters to be calculated. 5. Temperature ranges. Provide the maximum and mini- mum temperatures for both design and operating conditions to allow the correct fame arrester element and mechanical design of the arresters housing to be determined. 6. Pressure ranges. Supply the maximum and minimum pressures for both design and operating conditions. This information is also used (along with the temperature ranges) to determine the correct fame arrester element and mechan- ical design of the housing. The maximum pressure at which a fammable mixture can ignite in the process should be highlighted if this is different from the normal operating pressure. For storage tank applications, give both pressure and vacuum requirements. Figure 10. A Zone 0 waste-gas incineration process requires three layers of protection. Inert Gas Motive Gas Control Unit FIS-A TIS+A+ Waste Gas Zone 0 Detonation Arrester Deflagration Arrester End-of-Line Endurance Burn Arrester Combustor Gas Jet Pump Layer 2 Layer 1 Layer 3 Table 2. The likelihood of an explosive atmosphere and an ignition source being present determine the number of layers of protection that need to be provided. Ignition Source Explosive Atmosphere Always (Zone 0) Sometimes (Zone 1) Rarely (Zone 2) Never Always 3 2 1 0 Sometimes 2 1 0 0 Rarely 1 0 0 0 Never 0 0 0 0 Article continues on next page Copyright 2013 American Institute of Chemical Engineers (AIChE) 22 www.aiche.org/cep December 2013 CEP Safety 7. Allowable pressure drop. This is determined from the volumetric fowrate, and will enable the correct fame arrester confguration to be selected. 8. Type. Specify whether an inline, end-of-line, or pre- volume arrester is needed, whether the device will experi- ence short-term or endurance burning, and whether an unstable or stable detonation would be expected. For inline arresters, provide details of the piping between the fame arrester and the possible source of ignition in the form of a dimensioned sketch or isometric drawing. 9. Orientation. Note the intended orientation of the fame arrester vertical or horizontal. 10. Pipe size. State the nominal size of the pipework connected to the fame arrester. 11. Connection type. Provide details of the fanged or screwed connections. 12. Housing material. State the preferred material of construction for the fame arrester housing. The manufac- turer may verify the suitability of this material from an eval- uation of the mixture composition and operating conditions. 13. Element material. State the preferred material of construction for the fame arrester elements. The manu- facturer may verify the suitability of this material from an evaluation of the mixture composition and operating conditions. 14. Construction. Care should be taken when using materials such as aluminum or plastics, which, depending on the fuid and the operating conditions, may be subject to sparking and electrostatic charging. 15. Documentation. Tell the manufacturer what docu- mentation you require. This could include, for example, material certifcates or results of special tests. MICHAEL DAVIES is the Chief Executive Ofcer of Braunschweiger Flam- menlter GmbH PROTEGO (Braunschweig, Germany; Phone: + 49 5307-809-244; Email: [email protected]). Previously, he worked as a research scientist and in international project manage- ment. He has a masters degree in chemical engineering and a PhD in high-temperature thermodynamics from Technical Univ. of Clausthal. He serves on the standards committees for the National Fire Protec- tion Associations NPFA 68 (Standard on Explosion Protection by Deagration Venting) and NFPA 69 (Standard on Explosion Prevention Systems), as well as the American Petroleum Institutes API 2000 (Venting Atmospheric and Low-pressure Storage Tanks). THOMAS HEIDERMANN is Head of Research and Development at Braun- schweiger Flammenlter GmbH PROTEGO (Phone: + 49 5307-809- 205; Email: [email protected]), where since 2000 he has been responsible for the development of the PROTEGO line of ame arresters, valves, and tank equipment. In addition, he serves on several national and international committees as a recognized expert on explosion protection. Previously, he worked as a research fellow at the Max-Planck-Institut fr Strmungsforschung, a research manager at the Institute of Fluid Dynamics at the Univ. of Erlangen, and CEO of the engineering company Invent. He holds an MSc and a PhD in chem- istry from the Univ. of Gttingen, where he did his doctoral thesis on soot formation in high-pressure ames under Prof. Dr. H. Gg. Wagner in the Institute of Physical Chemistry. Literature Cited 1. American Petroleum Institute, Venting Atmospheric and Low-Pressure Storage Tanks; Nonrefrigerated and Refriger- ated, API Standard 2000, 5th Ed., API, Washington, DC (Apr. 1998). 2. German Government, Technische Regeln fr Brennbare Flssigkeiten (Technical Regulations for Flammable Liquids), TRbF 20, Lger, BArbBl Nr. 4/2001 S.60., Germany (2001). 3. United Nations, UN Recommendations on the Transport of Dangerous Goods, Model Regulations, www.unece.org/ fleadmin/DAM/trans/danger/publi/unrec/rev17/English/ Rev17_Volume1.pdf, United Nations Economic Commission for Europe (UNECE), New York, NY, and Geneva, Switzerland (2011). 4. United Nations, Globally Harmonized System of Classif- cation and Labeling of Chemicals (GHS), United Nations Economic Commission for Europe (UNECE), New York, NY, and Geneva, Switzerland (2003). 5. Chang, J. I., and C.-C. Lin, A Study of Storage Tank Acci- dents, Journal of Loss Prevention in the Process Industries, 19 (1), pp. 5159 (Jan. 2006). 6. International Organization for Standardization, ISO 28300: Petroleum, Petrochemical, and Natural Gas Industries Venting of Atmospheric and Low-Pressure Storage Tanks, 1st Ed., ISO, Geneva, Switzerland (2008). 7. International Organization for Standardization, ISO 16852:2008: Flame Arresters Performance Requirements, Test Methods, and Limits for Use, ISO, Geneva, Switzerland (2008). 8. American Petroleum Institute, Venting Atmospheric and Low-Pressure Storage Tanks, API Standard 2000, 6th Ed., API, Washington, DC (Nov. 2009). 9. American Petroleum Institute, Flame Arresters for Vents of Tanks Storing Petroleum Products, API Publication 2210, API, Washington, DC (2000). 10. British Standards Institution, Explosive Atmospheres Explosion Prevention and Protection Part 1: Basic Concepts and Methodology, EN 1127-1, BSI (Aug. 2011). Additional Resources Grossel, S. S., Defagration and Detonation Flame Arresters, Center for Chemical Process Safety, AIChE, New York, NY (2002). CEP ChemE on Demand Readers can explore this topic further by viewing author Davies webinar, Explosion Protection with Inline Flame Arresters, on ChemE on Demand (www.aiche.org/resources/chemeondemand/ webinars/explosion-protection-line-ame-arresters). AIChE undergraduate student members can view the webinar for free; nonstudent members can use a free credit. The price for nonmembers is $99. Copyright 2013 American Institute of Chemical Engineers (AIChE)