1、NA-04-1 -3 Burning Velocity and Refrigerant Flammability Classification Tony Jabbour Student Member ASHRAE ABSTRACT Classifcation offlammable refrigerants has taken a new momentum due to the use offlammable refrigerants in some refrigerating applications such as domestic refrigeration. ASHRAE Standa
2、rd 34 has been continuously updated for ten years to integrate both improved test methods and the use of new refrigerants and blends. ASHRAE flammability class 2 includes a wide flammability range of moderately flammable substances, and additional criteria are required to create an orderly ranking o
3、fflammability within this class. The burning velocity (B v) is a fundamental property that can be used fvr this purpose. This paper presents the main results of tests that have permitted the measurement of the BV of six pure substances and three blends. Three classes of flammability appear when usin
4、g the maximum BVfor ranking. BV is not a substitute for the lower flammability limit (LFL) and the heat of combustion (HOC) but is a complementary and essential criterion, permitting a continuous scaling of$ammability$-om moderate to highly flummable substances. INTRO DU CTIO N Refrigerant choice ha
5、s been radically changed by the Montreal Protocol, which is leading to a progressive and complete phase-out of chlorinated substances. Moreover, the application of the Kyoto Protocol-especially in Europe- leads to uncertainties on the refrigerant choices for the long term. Interest is increasing on
6、the possible use of“moderate1y“ flammable refrigerants or refrigerant blends due to the lower GWP of molecules having lower fluorine content or higher hydrogen content. The substitution of fluorine by hydrogen increases the flammability of those substances. ANSUASHUE Standard Denis F. Clodic, Ph.D.
7、Associate Member ASHRAE 34-200 1 (ASHRAE 200 1) has defined three classes for flam- mability. This classification has been adopted de facto by the European standard EN 378 (EN 2000) and is the reference document for the work of the IS0 TC86 / SC8 / Working Group 5 in charge of the update of IS0 8 17
8、 standard, which is the IS0 standard equivalent to ASHRAE 34. Until now, this standard is the de facto international standard and, as such, deserves a significant interest. The three classes for flammability are defined in the latest public version of ASHRAE Standard 34 (2001) as follows: Class 1 in
9、dicates refrigerants that do not show flame propagation when tested in air at 101 kPa (14.7 psia) and 2 1 “C (70F). Class 2 signifies refngerants having a lower flammability limit (LFL) of more than 0.10 kg/m3 (0.00625 lb/ft3) at 21C and 101 kPa (70F and 14.7 psia) and a heat of com- bustion (HOC) o
10、f less than 19,000 kJkg (8,174 Btu/lb). Class 3 indicates refrigerants that are highly flammable, as defined by an LFL of less than or equal to O. 10 kg/m3 (0.00625 lb/ft3) at 21C and 101 Ea (70F and 14.7 psia) or a heat of combustion greater than or equal to 19,000 kJ/kg (8,174 Bhdlb). Class 1 defi
11、nes the nonflammable substances, class 2 defines the moderately flammable substances, and Class 3 defines the highly flammable ones. The LFL is measured by a test defined in ANSUASTM Standard E68 1-0 1 (ASTM 2001). This test determines the concentration of flammable substance in the air, which permi
12、ts sustainable flame propagation under an arc of 90“, as indicated in Figure 1, Tony Jabbour is a Ph.D. student and Denis F. Clodic is deputy manager at the Center for Energy Studies, Ecole Nationale Suprieure des Mines de Paris, Paris, France. 522 02004 ASHRAE. volume during the combustion-and the
13、BV, which character- izes the burning rate through the flame front. Figure 1 Flask used by ASTMStandard E 681-01. In summary, the classification is based on the calculation of the heat of combustion (HOC) of complete reaction of refrigerant and on the measurement of the LFL. As indicated by Kataoka
14、(2001), within Class 2 refrigerant, there is a wide flammability range of flammable fluids, with no criterion for flammability ranking within the class, This paper shows how burning velocity (BV) can be used as an additional criterion to reach sound theory-based ranking of flammable refrigerants, es
15、pecially for moderately flammable substances. BURNING VELOCITY AND HAZARDS Every flammable compound has some risk of ignition, and a combustion reaction may occur when specific condi- tions are met (minimum ignition energy, flammable range, etc.). The direct consequences of that ignition can vary fr
16、om a simple local combustion flame to a large deflagration-and even an explosion-which might result in spreading fire to the surroundings as well as damaging the building structure from the generated pressure. In general, the flammability hazard can be related to the mixture composition of the flamm
17、able component. The conse- quences of ignition will vary according to the mixture compo- sition, which varies from the LFL to the upper flammability limit (UFL). The combustion theory and the thermochemical fundamentals allow the interpretation of the combustion mechanism. This mechanism is governed
18、 by fundamental properties of combustion-mainly the heat released per The BV is the velocity at which the flame propagates in a normal direction relative to the unburned gas ahead of it. The heat released during this exothermal reaction and the BV are related, and the higher the heat release, the hi
19、gher the temperature rise and the faster the burning rate. However, two different flammable substances having the same HOC will not necessarily have the same BV because one may react more or less rapidly than the other. The reaction rate introduces the time during which the reaction occurs and is ex
20、pressed by the BV, so the BV is an essential parameter, measuring how fast a given substance will react when it is ignited. The flamma- bility hazard can be related to the BV, especially when a rela- tive scale is established. Moreover, the BV permits one to estimate the minimum ignition energy as w
21、ell as the pressure generation, which are essential properties to evaluate the flam- mabi iity hazards, BURNING VELOCITIES OF SIX PURE SUBSTANCES AND THREE BLENDS The results presented are the BVs and flammability limits in air measured for propylene (R- 1270), propane (R-290), 1,l- difluoroethane (
22、R- 152a), ammonia (R-7 17), 1 ,l, 1 -trifluoroet- hane (R-l43a), and difluoromethane (R-32), as well as blends of R- 152a with CO, (R-744) and blends of R-32 with R- 125 and R- 134a, respectively. Table 1 summarizes some of the main characteristics of these rehgerants. BVs have been measured using t
23、he tube method at 23C and atmospheric pressure and for concentrations ranging from the LFL to the UFL. Many experimental methods have been developed for measuring the BV that can be classified into two categories as follows: Nonstationaryflame methods, in which the flame propa- gates through the ini
24、tial standstill unburned mixture. They include the tube, the spherical bomb (constant vol- ume), and the soap bubble (constant pressure) methods. Stationary flame methods, in which a stream of pre- mixed gas flows into a stationary flame with a velocity equal to the BY They include the burner and th
25、e flat flame methods. The choice of the method depends on many parameters, among which are the precision, the reliability, the repeatabil- ity, and the simplicity of implementation. Technically, there is no general agreement on a standardized method for measuring the BV, though each method presents
26、its own particularity, advantages, and disadvantages. The tube method is seen as one of the best methods for measuring BVs below 100 cm/s (Andrews and Bradley 1972; Linnett 1953). Measuring the BV in a vertical tube is known as a “simple” method (Gerstein ASHRAE Transactions: Symposia 523 R-290 R-12
27、70 I R-717 I ammonia I NH? I 17.03 I 45.9 I 22.5 I 21.9 I propane C H 3 C H 2 C H 3 44.10 -103.8 50.3 4.0 propylene CH?CH=CH? 42.08 + 20.0 48.9 4.5 R-744 R-125 R-134a Data of standard molar heat of formation are given at 298.15K (CRC 2002) - 44.01 -393.5 Non flammable - carbon dioxide CO2 pentafiuor
28、oethane CHF2CF, 120.02 -1 100.4 Non flammable - 1 , 1 ,I ,2-tetrafluoroet- CH,FCF, 102.03 - - Non flammable hane et al. 1951; Henderson and Hill 1956; Wheatley 1950). The constant flame velocity, shape, and minimal stretch rate favor an accurate measurement of the BV (Bradley 1999). The diffi- culty
29、 remains in the flame instabilities encountered with rich concentrations, induced turbulence, and acoustic wave inter- actions occurring with high-velocity propagating flames, and for which the recommended conditions of uniform flame speed and shape, essential for an accurate measurement of the BV,
30、are no longer valid. Even though these effects can be mini- mized to a certain limit by adjusting the exit orifice diameter of the combustion products at the lower end of the tube, these difficulties remain of minor importance for the presented tests because the studied substances have relatively lo
31、w BVs except for hydrocarbons. Measurement of Burning Velocity with a Vertical Tube The flammable refrigerant-air mixture is ignited at the lower end of a vertical tube and propagates up to the upper end. The flame front propagation is at a constant speed; the shape and the size of the flame are als
32、o constant. As the temperature and pressure of the unburned mixture ahead of the flame are constant, the volume of burned gas can be calculated. The expression for the BV Su is obtained from the equa- tion of mass conservation for the unburned gas. The volume of burned gas per second and per unit ar
33、ea, or the SV, is obtained by dividing the mixture volume consumed per second (at the test temperature and pressure) by the flame surface area, A. The volume consumption of the mixture per second is the volume swept by a cross-sectional area of the flame base, a, with a speed equal to the propagatio
34、n speed of the flame, S,. This relation can be written as follows: S,=S,xalA At a given temperature and pressure, the BV is only a function of the flammable substance, its concentration with the oxidant, and, to a limited extent, of the experimental appa- ratus. 1. Equation 1 implies the knowledge o
35、f three parameters: The flame propagation velocity, S, deduced from the measurement of the flame front displacement at the time given by the camera acquisition frequency. Special atten- tion should be paid to the assumption of a constant S, made when deriving Equation 1. This condition is true only
36、with a uniform movement of flame. The flame-base cross section, deduced from its diameter measured at the base of the flame. The flame front area, A, which needs an accurate method for its integration. Even though, in many cases, the flame front shape is symmetrical, this shape cannot be generated b
37、y the revolution of a parabola nor by approximation by an ellipsoid segment. To obtain better accuracy in its measure- ment, the flame front profile is marked with fitting points, then divided into two or more sections, and for each section a fitting equation of appropriate order is elaborated in or
38、der to reach the best-fitting curve of the selected section. The area of each section is then calculated separately by divid- ing it into a large number of small elementary sections (more than a thousand). The area of each elementary section is calculated based on the assumption of a revolu- tion sh
39、ape. 2. 3. EXPERIMENTAL TEST BENCH Measurement of the BV in a tube consists of propagating a flame in a vertical transparent tube, open at the lower ignition end and closed at the upper end and filled with the flammable mixture, 524 ASHRAE Transactions: Symposia measuring the flame speed, and record
40、ing by a camera the flame front area. Figure 2 presents a global view of the test bench as devel- oped by the laboratory. The different elements are: the tube, a mirror, a camera, a vacuum pump, a mixing vessel, a room temperature control system, and a gas extraction system. Tests are performed at r
41、oom temperature, set at 23“C* 1 K, using a special coolingheating controlled system, and at atmospheric pressure. Preparation of Mixtures The mixtures of flammable refrigerant and air are prepared in a 13 L internal volume vessel. The mixing vessel and all leading connections are evacuated beforehan
42、d to 10 Pa abs. (and less). A high-precision pressure transducer is used to measure the partiai pressures (-+0.025% of reading for values ranging from 1% to 100% full scale and -t0.00025% full scale for values below l%.). A mixing fan, 2.5 x 18 cm, magneti- cally driven from the outside, is placed i
43、nside the vessel and ensures the full mixing of the mixture components. Dry recon- Temperaiun, _ measurement Temperaiure - measurement Test tube -Tube inlet Quenching and smoothing suwn / Elecbodes - Purginggasline To vacuum PressUre Pump measurement To inlet PVC pipe 7- Magktic SUmK Quenching scree
44、n / Figure 2 Global vim of the test bench. ASHRAE Transactions: Symposia 525 stituted air (O2 2 1?40 - N2 79%) is used as an oxidant. The flam- mable gas has a 99.5% per weight purity or higher. Samples from the mixtures are taken and analyzed by gas chromatog- raphy in order to verify that the mixt
45、ures are homogeneously prepared and that the compositions are within the precision range. The Test Tube The test tube is made of special glass and is 1.5 m long. A 40 mm internal diameter has been chosen as a compromise between narrower tubes that will increase the quenching effect, and larger tubes
46、 that will increase instabilities (Coward and Hartwell 1932; Lewis and von Elbe 1961). To reduce the possible deformations of the flame front and to ensure a more symmetrical one, the tube is placed in a vertical position and purged with a continuous flow from the mixing vessel, with an equivalent v
47、olume of at least 12 times the internal tube volume. The gas mixture enters the upper end of the tube and exits from its lower end. The lower end can be closed after purging to avoid any possible concentration variation by dilu- tion in the neighborhood of the electrodes. The lower end of the tube i
48、s then opened to the atmosphere, whereas its upper end is connected to the mixing vessel from which the mixture flows into the tube and is closed after the end of purging until the end of the flame propagation. To prevent any hazard to the surroundings, both ends of the tube are mounted with quench-
49、 ing screens of a mesh size 0.3 ds for saturated and unsaturated HCs and highly flammable substances. As can be observed in Figures 4 and 5, the BV criterion can achieve a clear distinction between these categories. It will be interesting to add other refrigerants, permitting one to draw a more complete table of these three classes. The only pure substances are R-161 for the ethane series and R-41 for the methane series. Flammability Limits of Simple Compounds LFL and UFL define the flammable range of concentra- tions of a flammable compound in the mixture with the ox
