FLAMMABLE MIXTURE ANALYSIS FOR HAZARDOUS AREA CLASSIFICATION Allan Bozek P.Eng, MBA Vince Rowe, P.Eng Member IEEE EngWorks Inc. 1620, 49th Avenue SW Calgary, Canada T2T2T7 Member IEEE Marex Canada Ltd. 1550, 555 - 4th Avenue SW Calgary, Canada T2P 3E7 Abstract - The properties of flammable mixtures as they apply to a hazardous area classification analysis are discussed. Mathematical formula and application rules of thumb are provided to help estimate the relative density, group classification, autoignition temperature, material flashpoint and the flammability of a mixture. Application guidelines are provided on how to apply the results in the context of a hazardous area classification analysis. Index Terms – Hazardous area classification, flammable mixture analysis, density, group classification, autoignition, flashpoint, flammability. I. INTRODUCTION The vast majority of flammable materials handled in oil, gas and petrochemical facilities consist of flammable mixtures of various compositions. Performing a hazardous area classification risk assessment requires an understanding of the behavior of flammable mixtures and how they might influence the extent, group classification and autoignition temperature of the classified area. Unfortunately, very little guidance is given on mixture analysis by the various industry recognized standards and recommended practices on area classification. The search for the proper chemical relationships and formula can be frustrating. The mixture analysis must be performed with help of chemical textbooks and other references not specifically written for the purpose of a hazardous area classification analysis. The purpose of this paper is to provide guidance on how to analyze the properties of flammable mixtures as they apply to a hazardous area classification. The key flammable material parameters and the test methods used to determine the properties of pure flammable materials are discussed. In the absence of actual test results, a series of mathematical relationships and application rules of thumb are provided for estimating the properties of flammable mixtures. Application guidelines are also provided to assist in documenting the results within the context of a hazardous area classification analysis. II. HAZARDOUS AREA CLASSIFICATION OVERVIEW A hazardous area classification analysis is a structured risk assessment process used to determine the probability of a flammable atmosphere being present during normal or abnormal operations. The analysis divides the facility layout into zones of risk which are then used to determine the appropriate equipment and wiring methods for safe operation. The probability that a flammable atmosphere exists is dependent on the chemical properties of the flammable materials present, as well as the grade and rate of release. The extent of the hazard will be influenced by the behavior of the flammable material under normal ambient pressure and temperature conditions, the degree of ventilation within an area, the geometry and velocity of the release and the physical topography of an area. There are several methods for classifying an area once the flammable materials and physical parameters are identified. The most common way is to use the direct example approach where the facility is classified using diagrams that depict typical situations. This method of classification is incorporated into most industry recommended practices for hazardous area classification. A second method uses the point source approach to classify facilities. Potential release sources are identified and a formal calculation is performed to determine the extent of a classified area. The point source approach is more rigorous and thus requires more detailed information on the nature of the flammable materials to obtain a reasonable conclusion. III. PROPERTIES OF FLAMMABLE MATERIALS To assess the risk and the extent of the hazard, several key properties of the flammable material must be determined. They include: A. Vapor Pressure and Flashpoint The hazard posed by flammable materials is influenced by how easily it will mix with air and potentially form a flammable mixture. This is a function of vapor pressure and temperature. Flammable gases have a very low vapor pressure and as such are always in a gaseous state at ambient temperatures. Flammable liquids in contrast, are in a state of transition between a liquid and a gas. The hazard posed by a flammable liquid is a function of how easily it changes state from a liquid to a vapor at ambient temperatures. The temperature at which a liquid provides sufficient vapor to form a flammable mixture on the surface of the liquid is termed its flashpoint. This is the key parameter used to classify the potential hazard associated with flammable liquids. The flashpoint of a flammable material is determined experimentally using two recognized methods as illustrated in Fig. 1 and 2. The open cup flashpoint test procedure requires heating the flammable material in an open cup and periodically using an open flame to ignite the surface vapor. When there is sufficient vapor to sustain combustion, the flashpoint temperature is recorded. The closed cup flashpoint test procedure is similar to the open cup flashpoint test except the flammable material is sealed within a closed cup environment. The closed cup flashpoint test typically results in a lower flashpoint temperature than observed with the open cup test. The flashpoint temperatures most often quoted for flammable materials in various publications are derived using the closed cup method. Flashpoint is used as a basis for categorizing the volatility of flammable liquids. NFPA 30 [1] categorizes flammable materials into three classes based on their flashpoint and boiling point vapor pressure (Table I). Class I materials are those that readily flash to atmosphere at ambient temperatures and pressures. They are considered highly volatile and require due consideration with respect to an area classification. Class II materials typically do not flash at temperatures that would normally be considered “ambient”. From an area classification perspective, Class II flammable materials are considered when they are stored or handled at temperatures above their flashpoint. Class III materials are only a consideration when they are exposed to very high process temperatures and typically do not factor into a hazardous area classification. In all cases, the properties of the flammable material and the process conditions must be assessed prior to defining the degree and extent of a classified area. Part 15 (IP15) of the British Institute of Petroleum Model Code of Safe Practice in the Petroleum Industry has a similar classification for petroleum liquids based on flashpoint. The petroleum classes defined in Table II are used in conjunction with a series of typical diagrams that define the degree and extent of a classified area. IP15 takes into account the process conditions and adjusts the extent of the classification based on if the material is handled above or below its flashpoint temperature. It also makes reference to situations where the fluid cannot be easily classified by flashpoint and addresses the potential hazards associated with flammable mists and sprays. Under these circumstances, a judgment call must be made as to the fluid category and the characteristics of the potential hazard. The point source method of area classification is normally used in these situations [2]. Thermometer Flame Bunsen Burner Fig. 1 Open Cup Flashpoint Test Apparatus Test Cup Thermometer Agitator Bath Thermometer Flammable Liquid Flame Liquid Bath Heating Vessel Fig. 2 Closed Cup Flashpoint Test Apparatus TABLE I NFPA CLASSIFICATION OF FLAMMABLE/COMBUSTIBLE MATERIALS NFPA Flashpoint (ºC) Examples Class BP = Boiling Point IA < 22.8, (BP < 37.8) Ethyl Ether, Pentane IB <22.8, (BP > 37.8) Acetone, Gasoline IC ≥22.8 and < 37.8 Naphtha, Xylene II >37.8 and < 60 Fuel Oil, Kerosene IIIA ≥ 60 and < 93 Ethylene Glycol IIIB ≥ 93 Asphalt, Transformer Oil Open Cup with Flammable Liquid TABLE II IP Class 0 I II(1) II(2) III(1) III(2) Unclassified B. IP15 PETROLEUM CLASSES Flashpoint (ºC) Application NA <21 ≥21 and ≤ 55 ≥21 and ≤ 55 >55 to 100 >55 to 100 ≥ 100 LPGs Flam. at ambient conditions Handled below FP Handled above FP Handled below FP Handled above FP Relative Vapor Density The relative vapor density of a flammable gas/vapor is a key parameter when using the direct example approach method of area classification. The diagrams selected for a particular situation are dependent on the relative density of the gas/vapor as compared to air under standard ambient temperature-pressure conditions. For practical applications, a gas/vapor mixture that has a relative vapor density of 0.8 is regarded as lighter-than-air and a release would rapidly rise and collect in the upper levels of a confined area. Gas/vapor mixtures that have a relative vapor density greater than 1.2 are regarded as heavier-than-air and will collect at ground level. In theory, heavier-than-air flammable mixtures can travel long distances and result in large areas of potential hazard. Gas/vapor mixtures with a relative vapor density of between 0.8 and 1.2 can exhibit properties of both lighter-than-air and heavier-than-air behavioral characteristics so both possibilities should be considered [3] [4]. While gas/vapor density may have an influence on the extent of a classification in enclosed areas under ideal conditions, recent experiments using dispersion modeling have shown that the relative density has little influence on the extent of a hazard in a pressurized release. In these circumstances, the angle of release and how close the release point is to the ground has a larger effect. The extent of a flammable region is determined more by the direction and the velocity of the release than by the density of the gas or vapor released [5]. C. Group Classification of Flammable Materials The group classification assigned to a hazardous location determines the explosionproof or flameproof enclosure design requirements and the performance specifications for intrinsically safe circuits. The design of an explosionproof or flameproof enclosures is dependent on flamepath tolerances which are in-turn based on the Maximum Experimental Safe Gap (MESG) distance defined for a given flammable gas or vapor. The smaller the MESG defined for a material, the longer the flamepath required to cool the hot gases discharged from an explosionproof assembly during an ignition event. Typically, the longer the flamepath required, the more expensive the enclosure. The group classifications for flammable gases/vapors are determined experimentally based on the MESG determined under test conditions. The NEC group classifications for flammable materials are based on tests performed by Underwriter Laboratories using a Westerberg Explosion Test Vessel (WETV). The WETV consists of two chambers, one within the other, and each filled with a stochiometric explosive mixture. The mixtures are separated by an adjustable 25mm wide joint gap assembly. The flammable atmosphere is ignited in the inner chamber and allowed to propagate to the outside chamber. The opening of the joint gap assembly is reduced until an internal chamber ignition does not propagate to the external chamber. The opening of the 25mm wide joint gap at that point is the MESG. The group classification for the flammable material is then assigned based on Table III. The IEC group classifications are performed in a similar fashion. The size and shape of the explosion chamber differs from the WETV but the results are statistically similar. The IEC group classification of materials as it relates to MESG is also provided in Table III. TABLE III GROUP CLASSIFICATION CRITERIA NEC Gas Grouping IEC Gas Grouping Group MESG MIC Group MESG MIC (mm) ratio (mm) Ratio A Acetylene IIC ≤0.50 ≤0.45 B ≤0.45 ≤0.40 >0.45 >0.40 >0.50 >0.45 C IIB ≤0.75 ≤0.80 ≤0.90 ≤0.80 D >0.75 >0.80 IIA >0.90 >0.80 The group classification for a flammable material can also be based on the minimum ignition current (MIC) required to ignite the material under stochiometric conditions within a specified test apparatus. This is derived experimentally with the results often expressed as a ratio to the minimum ignition energy required for methane. Table III provides the group classifications based on the minimum ignition current ratio (MIC ratio) as compared to methane. D. Lower and Upper Flammable Limits Flammable gases and vapors are flammable only when they are between their lower flammable limits (LFL) and their upper flammable limits (UFL). For area classification purposes, the LFL is of greater concern as the hazard can be reduced by controlling the level of ventilation to dilute the flammable gas to a concentration well below the LFL. The LFL is a key parameter required for all hazardous area classification ventilation and point source calculations. Air Flammable Gas Mixture Manifold Hot Air Ignition Vessel Air Electrode Manifold Fig. 3 Flammable Limits Test Apparatus The %LFL and %UFL values for pure substances are derived experimentally using the apparatus described in ASTM E 681-04 and illustrated in Fig. 3. The concentration of the flammable material is gradually increased until an ignition is observed. The % volume to air is recorded as the lower flammable limit. The concentration is further increased and ignited until the mixture no longer ignites. The % concentration is then recorded as the upper flammable limit. E. Autoignition Temperature (AIT) The autoignition temperature of a flammable material is required to determine maximum safe operating surface temperature for equipment installed in a hazardous location. The maximum surface operating temperature for electrical equipment is defined by a temperature code. The temperature code is used to determine if the equipment item is suitable for installation based on the AIT defined for the hazardous location. There are a number of definitions for autoignition temperature in current use. For example, API RP500 defines AIT as: “The minimum temperature required, at normal atmospheric pressure, to initiate or cause self sustained combustion (independent of any externally heated element).” [3] API RP505 defines AIT as: “The lowest temperature of a heated surface at which, under specified conditions, the ignition of a flammable substance in the form of a gas or vapor mixture with air will occur.” [4] The API RP505 definition is a modification of the definition in IEC 60079-4 which is: “Ignition temperature – The lowest temperature at which ignition occurs when the method prescribed in this standard is used.” [7] The test method used in IEC 60079-4 consists of heating a 200ml Erlenmeyer flask in a hot air furnace as illustrated in Fig. 4. The flask is heated to the desired temperature, at which point a liquid or gaseous sample is injected into the flask. If no flame is observed within 5 minutes, the test is repeated until the minimum temperature at which ignition occurs is determined. The process is repeated until a degree of correlation exists between several identical tests. This correlated value is then defined as the autoignition temperature for the flammable material. ASTM E659 Standard Test Method for Autoignition Temperature of Liquid Chemicals defines AIT as: “The minimum temperature at which Autoignition occurs under the specified conditions of test.” [8] The test method described in standard ATSM E659 is similar to the test procedure used in IEC 60079-4 with the exception of the volume and the shape of the flask used. Published AIT tables are usually based on the laboratory procedures outlined in IEC 60079-4 and ASTM E659. The prescribed methods are relatively simple and the results are repeatable, however, they are only accurate for the conditions under which the testing is performed. Autoignition temperatures can be influenced by the volume, shape and the material composition of the test apparatus, the method and rate of heating, the flame detection method as well as the percent gas/vapor/air composition of the sample [9]. The published AIT tables typically reference the lowest AIT observed under controlled laboratory conditions. Industry experience however indicates that the actual minimum temperature at which ignition occurs is much higher. API RP 2216 Ignition Risk of Hydrocarbon Liquids and Vapors by Hot Surfaces in Open Air concluded that the minimum temperature (referred to as the minimum hot surface ignition temperature) should not be assumed unless the surface temperature is approximately 360°F (182°C) above the published AIT [10]. It should be noted that there is no standardized test for hot surface ignition temperature. The temperature at which hot surface ignition occurs is not a fundamental fluid property and is influenced by a number of factors including ambient conditions as well as the size, geometry and properties of the hot surface itself. Hot surface ignition data cannot easily be extrapolated to different situations and the use of a general rule of thumb based on a minimum autoignition temperature can be very inaccurate [11]. Fig. 5 illustrates the relationship between vapor pressure, upper and lower flammability limits, autoignition and hot surface ignition temperature. Flame Arrestor Mirror Viewing Line Syringe 200 ml Erlenmeyer Inflatable Reservoir Thermocouple Heater Elements Electric Furnace To Instrument Monitors To Heater Controls Fig. 4 IEC 60079-4 Autoignition Temperature Test Apparatus V a p o r P r e s s u r e Mixture Vapor Pressure Curve Published Autoignition Temperature Upper Flammable Limit Autoignition Zone Flash Point Temperature Hot Surface Ignition Zone (Curve may shift based on less ideal conditions) Lower Flammable Limit Mixture/Surface Temperature Fig. 5 Autoignition and Hot Surface Ignition Temperature as a Function of Temperature and Vapor Pressure IV. INDUSTRY REFERENCES Most area classification standards and recommended practices do not incorporate a list of flammable material properties. A secondary publication must usually be referenced to obtain the necessary data. The one notable exception is NFPA 497 Recommended Practice for the Classification of Flammable Liquids, Gases, or Vapors and of Hazardous (Classified) Locations for Electrical Installations in Chemical Process Areas which incorporates a list of flammable material properties within the context of the document [12]. Historically, API 500 and 505 has referenced NFPA 325 Guide to Fire Hazard Properties of Flammable Liquids, Gases and Volatile Solids as the recommended source of information for flammable material properties. The document was originally intended for fire prevention purposes and does not incorporate information on MESG, MIC ratio or minimum ignition energy data required for determining the group classification of a flammable material. As of 1998, NFPA 325 is no longer in print and has been superseded by the NFPA Fire Protection Guide to Hazardous Materials [13]. The document combines information from nine different NFPA reference documents on hazardous materials. With NFPA 325 no longer being published, the 2007 edition of API 500 and 505 now make reference to NFPA 497 as the recommended source of information for flammable material properties. NFPA 497 provides a complete list of chemical parameters including the NFPA flammable material classification, flashpoint, autoignition temperature, vapor pressure, density, group classification and the MESG and MIE/MIC for a list of commonly found flammable materials used in chemical process facilities. The list, although not comprehensive, includes most flammable materials commonly found in upstream production, transportation and refinery facilities. Area classifications performed using IEC 60079-10 [14] are referred to IEC 60079-20 Electrical Apparatus for Explosive Gas Atmospheres – Part 20 Data for Flammable Gases and Vapours [15] for a list of flammable material properties. IEC 60079-20 was created specifically for hazardous area classification purposes and includes a complete set of data for performing an analysis. There are minor differences between the data published in NFPA 497 and IEC 60079-20. The differences are related to the variation in test procedures and apparatus used. TABLE IV provides an overview of the differences between the reference publications for a cross sample of flammable materials. Other sources of flammable material property information include MSDS (Material Safety Data Sheets US OSHA) and WHMIS (Workplace Hazardous Materials Information System – Canada) material data sheets and a number of chemical reference texts that tabulate the properties of flammable materials. MSDS and WHMIS data can be valuable because they often provide information on chemicals and products not normally listed in industry chemical references. They however omit MIC and MESG information required for the group classification of a flammable material. TABLE IV FLAMMABLE PROPERTIES COMPARISON BETWEEN NFPA 497 and IEC 60079-20 Flammable AIT Ref. %LFL %UFL MESG Material ºC NFPA 5 15 630 1.12 Methane IEC 4.4 17 537 0.92 NFPA 1.5 7.8 243 0.93 Pentane IEC 1.4 7.8 258 0.93 NFPA 4 75 520 0.28 Hydrogen IEC 4 77 560 0.28 NFPA 4 44 260 0.90 Hydrogen Sulfide IEC 4 45.5 270 0.89 V. PROPERTIES OF MIXTURES In a perfect world, a hazardous area classification would be completed using the information provided in the various area classification standards and recommended practices and the properties listed in the hazardous material references. All process materials would be of a homogenous nature with their properties and flammable characteristics well defined. Unfortunately, real world process facilities combine process streams with different materials thus affecting the properties and characteristics of the combined mixture. This would not be an issue if the tests for flashpoint, MESG, %LEL and autoignition could be done for a mixture the same way it is done for a pure substance. This is however impractical in the context of plant design where the cost and effort associated simulating and testing each process stream does not justify the end result. Some mathematical means of estimating the properties of a flammable mixture is required to properly assess the nature of the hazard and define the extent of the classified zones. Fortunately there are some mathematical relationships that can be used to help predict the properties of flammable mixtures. A. Base assumptions To estimate the properties of a flammable mixture, a number of key assumptions are required. The first assumption is that the mixture is in a gas or vapor state upon release and it does not exist as mist or spray. The properties of flammable mists and sprays are extremely difficult to model and must be treated as a unique case for area classification purposes [5]. The second assumption is that the flammable materials remain in mixture. The behavior of ideal vs. non-ideal mixtures will impact the accuracy of this assumption to a degree but not enough to render the approximation invalid. B. Estimating Mixture Relative Density As previously discussed, the relative density of a mixture must be estimated to determine if it is heavier-than-air, lighterthan-air or of neutral buoyancy. The information is necessary to determine what direct example approach diagram is appropriate for a particular situation. The relative density of a mixture is calculated based on the molar weight of the mixture as it compares to the molar weight of air. The molar weight of air is 28.96 grams/mole at 0°C at sea level. The following formula is the basis for the calculation: n RDair = where ∑ Mi,xi i (1) 28.96 RDair = Relative density of the mixture compared to air Mi = molar weight of the individual constituent xi = corresponding mole fraction The results are then compared to the values in TABLE V and then applied to the appropriate area classification diagram. TABLE V RELATIVE DENSITY CATEGORIZATION RDair < 0.8 0.8 ≥ RDair ≤ 1.2 RDair > 1.2 Lighter than air Neutral buoyancy Heavier than air 1 MESGmix= Consider the mixture illustrated in TABLE VI. Calculate the molar weight of the mixture and determine if it is lighter or heavier than air. TABLE VI EXAMPLE: ESTIMATING RELATIVE DENSITY Material Methane Ethane H2S Nitrogen Heptane Octane Total Mi 16.04 30.07 34.08 28.02 100.2 114.22 xi 50% 10% 15% 15% 5% 5% 100% Mi x i 8.02 3.0 5.11 4.2 5.0 5.71 31.04 The molar weight calculated for the mixture is 31.04. When divided by the molar weight of air (28.96) the mixture has a relative density of 1.07 indicating the mixture has neutral buoyancy. This information would then be used to judge the extent of the transient vapor zone associated with the appropriate area classification diagram. It should be noted that most industry references for flammable materials do not provide information on the molar weight of materials. A chemistry textbook must be referenced to source this information. C. system will counteract as far as possible, the effect of the disturbance on the system. This principle is applied to estimating the MESG and subsequent group classification of a flammable mixture. The mathematical relationship for estimating MESG using Le Chatelier’s principle is as follows: Estimating the Group Classification of Mixtures Determining the group classification of a mixture is essential for determining the appropriate hazardous location electrical equipment specifications. For most upstream production and downstream refinery facilities handling hydrocarbon mixtures, the group classification is typically group D under the division system and group IIA under the zone system of area classification. The two flammable materials that most often influence the group classification of a hydrocarbon processing facility are H2S and hydrogen. API RP500, RP505 and IP15 provide some guidance in determining the group classification of facilities handling H2S and hydrogen. Generally, a group C or IIB group classification is not justified unless H2S comprises at least 25% of the flammable mixture stream. A group B or IIC classification is not justified unless the hydrogen component of a flammable stream exceeds 30%. For most hydrocarbon processing facilities, the H2S and hydrogen rule of thumb guidelines are sufficient to determine the group classification. For chemical processing facilities handling flammable mixtures other than hydrocarbons, the group classification is more complex. In these situations, Le Chatelier’s principle can be applied. Le Chatelier studied the influence of pressure, temperature and concentration on systems at equilibrium. Le Chatelier postulated that when a system at equilibrium is disturbed, the ∑ i where (2) xi MESGi MESGmix = Estimated MESG of the mixture MESGi = MESG of the individual constituent xi = corresponding mole fraction Consider the example of a flammable mixture consisting of 50% Diethyl Ether, a Group C, (IIB) classified material as illustrated in TABLE VII. TABLE VII ESTIMATING MESG EXAMPLE Material Diethyl Ether Methanol Nitrogen Isopropyl Ether Methyl Ether MESG (mm) 0.83 0.92 0.94 0.84 xi 50% 20% 15% 10% 5% Group C, (IIB) D, (IIA) D, (IIA) C, (IIB) Applying Le Chatelier’s principle to the mixture: 1 MESGmix = 0.5 0.83 + 0.2 0.92 + 0.15 ∞ + 0.1 0.94 + 0.05 0.84 MESGmix = 1.014 Using TABLE III, the gas mixture is thus classified group D, (IIA). It is interesting to note that the MESG value obtained for the mixture using Le Chatelier’s principle has a value greater than all MESGs for each of the individual material components. This is due to the effect nitrogen gas has on calculating the MESG of the mixture. Inert gases will influence the MESG of a mixture by reducing the oxidant quality of the mixture. Le Chatelier’s principle will tend to overemphasize the effect of the inert gas on the MESG of an inert-flammable gas mixture and will produce odd results when used with mixtures that contain a large percentage of inert gas. The same is true of mixtures that contain oxygen in excess of a nominal 21% concentration. The higher concentration of oxygen increases the volatility of the ignition producing higher pressures which in turn render the MESG test values for the individual components invalid. Materials such as carbon monoxide or carbon disulfide that do not incorporate hydrogen bonds can also result in erroneous results. Laboratory test results has shown that Le Chatelier’s principle is reasonably accurate provided the following situations are avoided: [16][17] 1. 3. 4. Flammable mixtures where a significant portion of the gas is an inert. Mixtures that incorporate oxygen as one of the components. Mixtures that contain greater than 5% carbon monoxide. Mixtures that contain acetylene. D. Estimating LFL and UFL of Mixtures 2. Estimating the lower flammable limit of a mixture is required for point source calculation methods and for determining the level of adequate ventilation in an enclosed area using the fugitive emissions calculation procedure described in appendix B of API RP 500 and 505. Upper flammable limits are seldom used as a basis for area classification. The LFL of a mixture may be estimated by applying a derivation of Le Chatelier’s principle. The derivation uses only the flammable components of the mixture with their mole fractions adjusted to their relative percentage of the flammable mix as follows: 1 %LFLmix= ∑ %LFLi i where (3) xiflam %LFLmix = Estimated LEL of the mixture %LFLi = %LFL of the individual constituent xiflam = Corresponding mole fraction on a flammable bases Consider the flammable mixture example previously used for estimating MESG. TABLE VIII EXAMPLE: ESTIMATING %LFL Material Diethyl Ether Methanol Nitrogen Isopropyl Ether Methyl Ether xi 50% 20% 15% 10% 5% 100% xiflam 58.8% 23.5% 11.8% 5.9% 100% LFL 1.7% 6.0% Le Chatelier’s principle provides flammability limits that are close to the experimental values of simple hydrocarbon mixtures with sufficient accuracy for most area classification analysis. The approximation is valid with mixtures in air at ambient pressures [18]. E. Flammability of Mixtures There are often situations where gas or fluid mixtures contain flammables in low concentrations. The question arises whether the mixture as a whole is flammable and therefore must be considered as part of an area classification. To determine the flammability of a mixture, an assessment must be made to determine if the mixture contains sufficient flammables to exceed the lower flammable limit of the mixture as a whole. The calculation procedure is relatively easy for mixtures in a gaseous state. It is more complex for fluid mixtures where the flammable components may flash to atmosphere based on ambient, operating or storage temperatures. 1) Flammability of a Gas Stream: The flammability of a process gas stream may be determined by first estimating the %LFL of the mixture using Le Chatelier’s principle and then comparing the %LFL to the total %flammables contained in the mixture. If the %flammables exceed the %LFL, the mixture is considered flammable [19]. Consider the example in TABLE IX where a relatively small percentage of flammables make up a gas mixture: TABLE IX EXAMPLE: DETERMINING THE FLAMMABILITY OF A GAS STREAM Material Methane Butane Ethane Air % Vol 3 1.5 1 94.5 100% xiflam 54.5% 27.3% 18.2% Estimating the %LFL for the mixture using Le Chatelier’s principle: 1 X 100 = 3.2% %LFLmix = 1.4% 3.4% LFL 5.0% 1.9% 3.0% 0.545 0.273 + 0.05 0.019 + 0.182 0.03 The flammable percent volume of the mixture is: 1 X 100 %LFLmix = 0.588 0.235 + 0.017 0.06 + 0.118 0.014 + 0.059 0.034 %LFLmix = 2% The UFL of a mixture can also estimated using Le Chatelier’s principle in a similar way. %Flammix = 3% + 1.5% + 1% = 5.5% Since the %Flammix exceeds the %LFLmix the mixture is considered flammable and must be considered within the context of an area classification. 2) Flammability of a Process Fluid Stream: Determining the flammability of a process fluid stream requires estimating the flashpoint of the mixture. A common industry practice for estimating the flashpoint of mixtures is to use the lowest published flashpoint for any of the pure components. If the flashpoint temperature for the single lowest pure component in the mixture is exceeded at ambient, process or storage handling conditions, the mixture is considered flammable. This assumption is thought to provide a conservative value for the flashpoint of a mixture however; further research has shown that this assumption is not always valid. Mixtures can exhibit lower flashpoints than any of the pure component values depending on if the mixture is classed as ideal or nonideal. An ideal mixture is defined as a mixture where all molecular interactions are the same as if the individual components were in their pure state at the same pressure and temperature as the mixture [20]. The mixture will follow the properties of Raoult’s law which allows the thermodynamic properties of the mixture to be predicted based on the properties of the individual components. Using this principle, the vapor pressure and flashpoint of a mixture can be estimated diagrammatically as illustrated in Fig 6. The vapor pressure or the flashpoint of a mixture is directly influenced by the mole fraction of the individual constituents of the mixture. PsatB PsatA Vapour Pressure Tf,B Flashpoint xi xi (mole fraction) (mole fraction) Tf,A Fig. 6 Vapor Pressure and Flashpoint Behavior of Ideal Mixtures Mixtures that exhibit ideal behavior include water-alcohol and heptane-hexane hydrocarbon mixtures. Non-ideal mixtures deviate from Raoult’s law in a positive or negative manner depending on the molecular attraction of the mixture components. Fig. 7 illustrates the behavior of a mixture that deviates in positive manner from Raoult’s law. Note that under certain conditions, the flashpoint of the mixture may be less than the flashpoint of either component in the mixture. Most mixtures exhibit non-ideal behavior to a degree. Methanol hydrocarbon mixtures are highly non-ideal and have been shown to exhibit flashpoint temperature lower than either component. This complicates the development of a simple rule of thumb for determining the flammability of a fluid mixture. Unless a mixture can be positively confirmed to exhibit ideal behavior in accordance with Raoult’s law, the mixture must be assumed to be flammable with the flashpoint lower than what is published for any of the flammable components. The only reliable way of determining the flashpoint and hence the flammability of a fluid mixture is by test procedure. Actual Vapour Pressure PsatA PsatB Tf,B Ideal Ideal Actual Flashpoint xi xi (mole fraction) (mole fraction) Tf,A Fig. 7 Vapor Pressure and Flashpoint Behavior of Non-Ideal Mixtures F. Autoignition Temperature Estimating an AIT for a mixture is a challenging problem due to the complex nature of the autoignition-combustion process. There are no established mathematical relationships to help in this regard. The only way to determine the AIT of a mixture is by laboratory test. If AIT data for a mixture is not available and the situation precludes testing, often the lowest AIT of any pure component in the flammable mixture is used as the AIT. This typically results in a very conservative AIT value and may not reflect the true nature of the hazard. A rule of thumb often used to estimating the AIT of hydrocarbon mixtures is to take the lowest AIT of any component exceeding 5% (by volume) of the total mixture. This is based on the assumption that the autoignition temperature of a mixture will be influenced more by the properties of the major components than by the minor components in the mixture. This assumption has not been verified and additional research in this area is required. Rules in the NEC and the CEC require the maximum surface operating temperature of electrical equipment to be less than the ignition temperature of the specific gas or vapor in a hazardous area. While the rules do not specify the use of AIT’s (as opposed to hot surface ignition temperature), the use of published AIT’s has been standard practice. In some situations it may make sense to use a temperature other than the published AIT based on documented engineering judgment. A real life example involves the installation of MI heat tracing on high temperature process piping in heavy oil upgrader facilities. In these situations, it is common for process piping to operate at elevated temperatures in the range of 325°C in areas where diluents (predominantly hexane - AIT 215°C), are present. Historically, the heat trace design would have been limited to a maximum surface operating temperature of 215°C on a pipe that normally operates at 325°C. This results in multiple passes of a lower wattage electric heat trace to limit the maximum surface operating temperature of the tracer to less than 215°C. Recent upgrader facility designs have made documented engineering decisions to allow the maximum design temperature of the heat trace to be as high as the operating temperature of the hot piping. This allows more flexibility in the design of the heat trace system and typically reduces the number of tracer circuits required to meet the process piping heating requirements. VI. APPLICATION GUIDELINES The hazardous area classification for a facility is usually documented using layout drawings and details that define the degree and extent of the classified areas. The drawings also provide pertinent information on the group classification and autoignition temperature defined for the facility. For complex facilities that incorporate a variety of flammable mixture streams, the drawings are usually supplemented by an engineering document that summarizes the basis for the area classification. The hazardous area classification study report should include: 1) A description of the process and the flammable materials handled. 2) A list of the codes, standards, recommended practices and material references used in the analysis. 3) A list of assumptions used to classify the facility including the basis for normal and abnormal operations. This would also include a section on flammable mixtures and how their material properties were estimated. 4) A hazardous materials worksheet that documents the properties of the flammable materials handled within the facility. The flashpoint, group classification, autoignition temperature, LFL/UFL and hazard classification is summarized for each process stream. 5) A source of release worksheet that identifies the most likely release points within a facility. 6) The basis for the group classification and autoignition temperature for the overall facility. 7) A summary of any fugitive emission and ventilation calculations. 8) Recommendations for gas detector placement including information on what flammable gases are to be detected and the calibration procedure required for gas detector maintenance. 9) A list of recommendations for operation and maintenance activities performed in classified areas. 10) The area classification layout drawings should incorporate a note that references the study report as the bases for the area classification design. The hazardous area classification study report provides essential information for understanding the basis for the original hazardous area classification and for planning any future modifications to the area classification design for the facility. the molar weight of air at ambient temperatures and pressures. Determining the group classification, estimating the LFL, UFL, and predicting the flammability of a gas mixture can be done using Le Chatelier’s principle. The complex nature of the autoignition process precludes the use of a simple mathematical model or industry rule of thumb that can be used to estimate the AIT of a flammable mixture. Additional research is required to develop a basis for an industry accepted practice for estimating the autoignition temperature of flammable mixtures. X. REFERENCES [1] NFPA 30, Flammable and Combustible Liquids Code, National Fire Protection Association, Quincy, MA., 2003. [2] Institute of Petroleum, Model Code of Safe Practice – Part 15, Area Classification Code for Installations Handling Flammable Fluids, 3rd Edition, Energy Institute (2003), Portland Press. [3] ANSI/API RP 500, Recommended Practice for Classification of Locations for Electrical Installations at Petroleum Facilities Classified as Class I, Division 1 and Division 2, American Petroleum Institute, Washington, DC, 1998. [4] ANSI/API RP 505, Recommended Practice for Classification of Locations for Electrical Installations at Petroleum Facilities Classified as Class I, Zone 0, Zone 1, and Zone 2, American Petroleum Institute, Washington, DC, 1998. [5] Propst, John, E., Volatility and Mists—Electrical Area Classification's Important Variables, IEEE Transactions on Industry Applications, Vol. 36, No. 2, March/April 2000. [6] ASTM E 681-04, Standard Test Method for Concentration Limits of Flammability of Chemicals, American Society for Testing and Materials, Conshohocken, PA, 2004. [7] IEC 60079-4, Electrical Apparatus for Explosive Gas Atmospheres Part 4: Method of Test for Ignition Temperature, International Electrotechnical Commission, Geneva, Switzerland 1975. [8] ASTM E659-78, Standard Test Method for Autoignition Temperature of Liquid Chemicals, American Society for Testing and Materials, Conshohocken, PA, 2005. [9] NFPA 325, Guide to Fire Hazard Properties of Flammable Liquids, Gases and Volatile Solids, National Fire Protection Association Quincy, MA., 1994. VII. CONCLUSION Performing a hazardous area classification risk analysis requires an understanding of how flammable mixtures behave. In an ideal world, the properties of a flammable mixture would be determined by test procedure; however, this is impractical in most cases. Methods are required to estimate the properties of flammable mixtures to support a hazardous area classification analysis. Predicting the relative vapor density of a mixture is a simple calculation that compares the molar weight of the mixture to [10] API RP 2216, Ignition Risk of Hydrocarbon Liquids and Vapors by Hot Surfaces in Open Air, American Petroleum Institute, Washington, DC, 2003. [11] Colwell, J.D., Reza, A., Hot Surface Ignition of Automotive and Aviation Fluids, Fire Technology, 41, 105-123, 2005. [19] Crowl, Daniel, A., Understanding Explosions, American Institute of Chemical Engineers, New York, New York, 2003 [12] NFPA 497, Recommended Practice for the Classification of Flammable Liquids, Gases, or Vapors and of Hazardous (Classified) Locations for Electrical Installations in Chemical Process Areas, National Fire Protection Association Quincy, MA., 2004. [20] Perry’s, Chilton, Kirkpatrick, Perry’s Chemical Engineers Handbook, McGraw-Hill Inc. New York, New York, 1963. [13] Fire Protection Guide to Hazardous Materials, National Fire Protection Association, Quincy, MA., 2002. [14] IEC 60079-10, Electrical Apparatus for Explosive Gas Atmospheres – Part 10 Classification of Hazardous Areas, International Electrotechnical Commission, Geneva, Switzerland, 2002. [15] IEC 60079-20, Electrical Apparatus for Explosive Gas Atmospheres – Part 20 Data for Flammable Gases and Vapours, related to the use of Electrical Apparatus, International Electrotechnical Commission, Geneva, Switzerland, 2000. [16] Briesch, Edward, NEC Group Classification of Mixtures, American Institute of Chemical Engineers, 34th Loss Prevention Symposium, Atlanta, Georgia, March 8, 2000. [17] Report on Proposals NFPA 497, Report of the Committee on Electrical Equipment in Chemical Atmospheres, F2007. [18] Hristova, M., Tchaoushev, S., Calculation of Flash Points and Flammability Limits of Substances and Mixtures, Journal of the University of Chemical Technology and Metallurgy, 41, 3, Pg 291-296, 2006. H. VITA Allan Bozek, P.Eng., MBA, is a Principal with EngWorks Inc. providing consulting engineering services to the oil and gas industry. He is a registered professional engineer in the provinces of Alberta and British Columbia, Canada and has been a member of the IEEE since 1989. Allan’s areas of expertise include hazardous area classification, power systems design, protective relaying and grounding for large scale industrial and petrochemical facilities. Mr. Bozek graduated from the University of Waterloo in 1986 with BASc in Systems Design Engineering and a MBA from the University of Calgary in 1999. Allan may be contacted at ABozek@EngWorks.ca Vince Rowe retired from Shell Canada in 1994 and continues to be active in the Oil and Gas industry as a consultant. He provides training on electrical installations and maintenance in classified locations. He is a partner with Marex Canada Limited and a member of the steering committee for the Canadian Electrical Code. He is Chairman of the Hazardous Location and Electric Heating sections and one of the initiators of the Objective Based Industrial Electrical code for Canadian Industry. Mr. Rowe graduated from the University of Manitoba in 1960 with a Bsc in Electrical Engineering. Vince may be contacted at Vince.Rowe@shaw.ca