Ind. Eng. Chem. Res. 2001, 40, 4649-4653
4649
Fire Protection with Bromoalkene/Nitrogen Gaseous Mixtures Yong Zou,† Nader Vahdat,*,† and Michelle Collins‡ Chemical Engineering Department, College of Engineering, Architecture, & Physical Sciences, Tuskegee University, Tuskegee, Alabama 36088, and NASA Kennedy Space Center, Kennedy Space Center, Florida 32899
The fire suppression efficiencies of 1-bromo-1-propene/nitrogen and 2-bromo-3,3,3 trifluoro-1propene/nitrogen gas mixtures were studied by a cup burner method. It was shown that addition of 1-bromo-1-propene and 2-bromo-3,3,3 trifluoro-1-propene to nitrogen can improve the suppression effectiveness of the inert gas by 34.5 and 64.3%, respectively. An arrangement for mixing organic compounds with inert gases before application to fire was proposed, and it was shown that droplets of bromoalkenes with diameters as large as 100 µm can be vaporized in less than 1 s. 1. Introduction
Table 1. Chemical and Physical Properties of Two Bromoalkenes
The international ban on the production of ozonedepleting halons has forced scientists to search for environmentally benign fire suppressants. A wide variety of chemical groups has been considered for possible use as halon replacements during the past decade. Metal compounds, sulfur compounds, bromoalkenes, and bromoethers are among the new classes of compounds that have been identified as having high fire suppression efficiencies and low environmental impacts.1-4 Halon 1301 is a highly effective total flooding agent because of its ability to inhibit flames at low concentrations. It is a nonconductive and clean agent with a low boiling point and low toxicity. Most of the new compounds considered for halon replacement have high boiling points and are liquids at room temperature,1-7 so that they cannot be used as total flooding agents. However, mixtures of these compounds in an inert gas such as nitrogen can produce a gaseous agent suitable as a halon 1301 replacement. The purpose of this project was to investigate the effectiveness of mixtures of bromoalkenes and nitrogen as total flooding fire suppressants. Inert gases extinguish fire by reducing the oxygen concentration to between 10 and 14%. They are nontoxic, electrically nonconductive, and clean fire suppressants. The main problem of inert gases is that large volumes of these agents are needed to extinguish a fire. Storage volumes required for inert gases are 11- 13 times larger than those for halon 1301.8,9 Addition of an effective fire suppressant to inert gases can reduce their extinguishing concentrations and, thus, reduce storage volume requirements. The concentration of a vapor (such as a bromoalkene) in a gas depends on its vapor pressure and the total pressure. The partial pressure of a vapor in a mixture is given by
pv ) Pyv
(1)
where pv is the partial pressure of vapor, yv is the mole fraction of vapor, and p is the total pressure. As long as * To whom correspondence should be addressed. Tel.: 1-334727-8978. Fax: 1-334-727-8090. E-mail address: vahdatn@ tusk.edu. † Tuskegee University. ‡ NASA Kennedy Space Center.
property CAS number purity molecular formula molecular weight appearance boiling point density solubility in water
1-bromo-1propene
2-bromo-3,3,3trifluoro-1-propene
590-14-7 98% CH3-CHdCHBr 120.98 colorless to yellow liquid 58-62 °C 1.415 g/mL immiscible
1514-82-5 97% CH2dCBrCF3 174.95 colorless liquid 33-33.5 °C 1.79 g/mL immiscible
the partial pressure of vapor in the mixture is less than its vapor pressure (at the system temperature), it will not condense, and the mixture can be used as a clean fire suppressant. Inert gases are usually stored at pressures over 2000 psia. At such high pressures, the maximum concentration of vapor in the gas is very small. Therefore, the organic compound should be stored in the liquid form in a gas cylinder or in a separate container. It should be vaporized and mixed with the inert gas before it is applied to the fire. 2. Experimental Section 2.1. Materials and Reagents. In this study, two bromoalkenes, namely, 2-bromo-3,3,3 trifluoropropene and 1-bromo-1-propene, from Lancaster Synthesis Ltd., were studied. Representative characteristics of the two bromoalkenes are reported in Table 1. The atmospheric lifetimes of many bromoalkenes have been measured or estimated to be as low as a few days.3 This means that, for all practical purposes, they rapidly degrade within the troposphere and never reach the stratosphere; therefore, they have an ozone depletion potential (ODP) close to zero. 2-Bromo-3,3,3 trifluoro-1-propene has been tested for toxicity.3 The results of acute inhalation toxicity and Ames mutagenicity tests have shown that the compound is nontoxic. No information on the toxicity of 1-bromo-1-propene is available in the literature. n-Heptane (HPLC grade) was from Fisher Scientific Inc. The air and nitrogen (99.998% purity) were both from Air Products and Chemicals Inc. 2.2. Apparatus and Procedure. 2.2.1. Apparatus. A schematic diagram of the experimental apparatus for measuring of the extinguishing concentration is shown in Figure 1. It has four main parts: a cup burner (I), a liquid fuel feeder with a fuel level control device (K), a
10.1021/ie010281z CCC: $20.00 © 2001 American Chemical Society Published on Web 09/18/2001
4650
Ind. Eng. Chem. Res., Vol. 40, No. 21, 2001 Table 2. Extinguishing Contcentrations of Pure Compounds compound
extinguishing concentration (vol %)
N2 CH3-CHdCHBr CH2dCBrCF3 CF3Br
33.6 3.0a 3.5a 3.8b
a From ref 4 after accounting for differences in the dispersion of a liquid and a gas. b From ref 10.
Figure 1. Schematic diagram of the experimental apparatus for measuring extinguishing concentration.
Tedlar bag (made of polyvinyl fluoride and purchased from Cole-Parmer, Vernon Hills, IL) that provides chemical/inert (such as bromoalkenes/nitrogen) gaseous mixtures (E), and the gaseous mixture/air mixture supply unit (A and G). The burner (made of quartz) is constructed using the specifications given in ISO 14520 (Determination of the Flame Extinguishing Concentration of Gaseous Extinguishants by the Cup Burner Method). The dimensions of the burner are 30 mm for the o.d. of the cup and 85 mm for the i.d. of the chimney. 2.2.2. Procedure. Experiments for measuring the extinguishing concentrations were carried out as follows: (1) A small amount of organic compound (1-5 mL) was injected into the Tedlar bag. (2) A specific volume of nitrogen (3-8 L) was also introduced into the Tedlar bag immediately, and then the Tedlar bag was closed and kept in the pressure vessel (F) until the liquid chemical was completely vaporized (8-12 h). (3) Vessel (F) was closed and connected to a compressed nitrogen gas cylinder to keep its pressure constant at 11 psig. (4) n-Heptane was introduced into the cup burner and the fuel level in the cup was adjusted to within 5-10 mm of the top of the cup. (5) Air with a flow rate of 10 L/min as determined by the mass flow controller (A) was passed through the burner. The fuel was then ignited and allowed to burn for a period of at least 2 min. (6) The flow of the flooding agent was then started at a constant pressure (∼11 pisg). The flow rate, controlled by the digital flow meter, was increased in increments until the flame was extinguished. (7) A waiting period of about 30 s was allowed between each successive increase in the flow rate of the agent so that the new proportion of air and agent in the manifold reached the cup position. The flow rate of the flooding agent at extinction was found from the digital flow meter. Each experimental result was repeated three times to verify data reproducibility. 3. Results and Discussions 3.1. Extinguishing Concentration of Flooding Agents. The concentration of suppressant in volume percent was calculated as follows
C)
Vsup × 100% Vair + Vsup
(2)
Figure 2. Effect of concentration of 1-bromo-1-propene in the flooding agent on the extinguishing concentration of the flooding agent.
Figure 3. Effect of concentration of 2-bromo-3,3,3-trifluoro-1propene in the flooding agent on the extinguishing concentration of the flooding agent.
The extinguishing concentrations of the pure agents are given in Table 2. The suppressant effectiveness of the two bromoalkenes is very close to that of halon 1301. However, the main difference between these compounds and halon 1301 is their boiling points. The bromoalkenes are liquids at room temperature and cannot be used as total flooding agents. The results of cup burner experiments for the mixtures of the two alkenes in nitrogen are given in Figures 2 and 3. As indicated in the Introduction section, there is a limit on how much organic vapor can be added to the nitrogen (saturation concentration). For 1-bromo-1-propene, mixtures with concentrations up to 20% in nitrogen were prepared and tested. 2-Bromo-3,3,3-trifluoro-1-propene has a higher vapor pressure, and mixtures with concentrations of up to 30% chemical in nitrogen were prepared and tested. The extinguishing concentrations for both bromoalkene/
Ind. Eng. Chem. Res., Vol. 40, No. 21, 2001 4651
nitrogen mixtures decreased with increasing concentration of organic compound. The extinguishing performance of mixtures of two agents with different organic compound concentrations can be described by the following equations
Y ) 18.573 46 + 14.964 93e-X/3.702 74 (1-bromo-1-propene) or
Y ) -0.975 11 + 34.542 94e-X/33.277 09 (2-bromo-3,3,3-trifluoro-1-propene) which shows that the mixtures of the two bromoalkenes with nitrogen have synergism. As can be seen from Figure 2, addition of small quantities of 1-bromo-1-propene to nitrogen improves its extinguishing capabilities considerably. A 4.5% (by volume) mixture of the bromoalkene in nitrogen has an extinguishing concentration of 22%, compared to 33.6% for pure nitrogen, which represents a reduction of 34.5%. Further addition of the organic compound improves the fire suppressant effectiveness of nitrogen only slightly. For example, a 15.9% 1-bromo-1-propene mixture has an extinguishing concentration of about 18%. The extinguishing concentrations of 2-bromo-3,3,3trifluoro-1-propene/nitrogen mixtures are shown in Figure 3. The suppressant effectiveness of the mixture improves steadily as the concentration of bromoalkene is increased. For a 32% mixture of organic compound in nitrogen, the extinguishing concentration is about 12%. This is a 64.3% improvement in the effectiveness of nitrogen. One of the advantages of using inert gases for fire suppression is that they do not produce toxic byproducts. When mixtures of bromoalkenes and inert gases are used as fire extinguishers, they can produce acidic byproducts during suppression. 1-Bromo-1-propene would produce HBr, and 2-bromo-3,3,3-trifluoro-1-propene would produce HBr and HF. 3.2. Technology of Mixing Organic Compounds with Inert Gases. Inert gases are usually stored in cylinders at pressures over 2000 psia. As a result of condensation, high concentrations of organic compound vapors in inert gases cannot be stored at elevated pressures. The organic compound, therefore, should be stored in liquid form and converted to vapor and mixed with the inert gas when needed. Two arrangements are possible: (1) the organic compound can be stored as a liquid in the inert gas cylinder at high pressure or (2) the liquid can be stored in a separate container. In the first arrangement, the organic compound (such as a bromoalkene) is stored at high pressure in the gas cylinder. During application, as the gas is released, the pressure in the cylinder drops, and the liquid vaporizes, thus providing a continuous flow of the gaseous mixture. Because of the high pressure in the cylinder, however, the concentration of the vapor will not reach high values until a good portion of the inert gas has been released. This means that the fire will essentially be extinguished with only the inert gas initially. Another problem with this arrangement is that there might not be enough time for the liquid to vaporize. A better arrangement is to store the liquid in a separate container. During application, the liquid is sprayed into the inert gas as it leaves the high-pressure cylinder. The gas carries the droplets to the fire. If the droplets are vaporized before the
stream reaches the fire, a clean agent with suppressant effectiveness far better than that of the inert gas has been produced. It should be noted that a gaseous extinguishing agent permeates space and suppress the fire. The purpose of vaporizing the organic droplets is to distribute the organic compound to the entire space during fire suppression. The time required to completely vaporize the liquid droplets is calculated using the theoretical method described below. The rate of decrease in particle size during evaporation is given by11
(
[
)
d(dp) 4DM pv p* ) dt RFpdp T T0
2λ + dp
()
]
λ2 dp + 5.33 + 3.42λ dp
(3)
For droplets larger than 2 µm, the last bracket in eq 3 can be neglected.
(
)
d(dp) 4DM pv p* ) dt RFpdp T T0
(4)
Integration of eq 4 from the initial size of the droplet to zero gives droplet lifetimes
t)
RFpdp2 8DM
(
)
p* pv T0 T
(5)
If the temperature of the gas (T) is assumed to be equal to the temperature of the droplets, eq 5 is reduced to
t)
RTFpdp2 8DM(p* - pv)
(6)
The partial pressure of the vapor in the gas can be written in terms of the relative humidity (R) as
R)
pv p*
(7)
Substitution of eq 7 into eq 6 yields
t)
RTFpdp2 8DMp*(1 - R)
(8)
To use eq 8 to calculate the drying time for droplets of the organic compounds, the vapor pressure (p*) and diffusion coefficient of organic vapor in the inert gas (D) are needed. These properties are not available for the two bromoalkenes studied here. They can be estimated from correlations available in the literature. Myrdal et al.12 developed an equation for the vapor pressure of complex organic molecules. This equation, which requires knowledge of only the boiling point and molecular structure, has been shown to be fairly accurate for a wide variety of compounds and over a wide range of temperatures. The results of calculations for 1-bromo1-propene and 2-bromo-3,3,3 trifluoro-1-propene are given in Table 3. The diffusion coefficient of a binary mixture can be estimated from the Wilke-Lee modification13 of the Hirschfelder-Bird-Spotz method.14 This correlation requires knowledge of the boiling point and density of the liquid. The results of calculations for diffusion coefficient of the two bromoalkenes in nitrogen
4652
Ind. Eng. Chem. Res., Vol. 40, No. 21, 2001
4. Conclusions The fire extinguishing concentrations of 1-bromo-1propene/nitrogen and 2-bromo-3,3,3 trifluoro-1-propene/ nitrogen gas mixtures at various concentrations of 1-bromo-1-propene and 2-bromo-3,3,3 trifluoro-1-propene were determined by a cup burner method. The extinguishing concentrations for both bromoalkene/ nitrogen mixtures decreased with increasing concentration of organic compound, the extinguishing behavior of the two bromoalkene/nitrogen mixtures can be represented by Figure 4. Drying time for 1-bromo-1-propene droplets at 25 °C.
Y ) 18.573 46 + 14.964 93e-X/3.702 74 (1-bromo-1-propene) and
Y ) -0.975 11 + 34.542 94e-X/33.277 09 (2-bromo-3,3,3-trifluoro-1-propene)
Figure 5. Drying time for 2-bromo-3,3,3 trifluoro-1-propene droplets at 25 °C. Table 3. Relations for Vapor Pressure of Bromoalkenes compound
vapor pressure equationa
CH3-CHdCHBr CH2dCBrCF3
log p* ) 36.27 - 3051/T - 4.667 ln T log p* ) 35.88 - 2804/T - 4.667 ln T
a
The mixtures of both bromoalkenes with nitrogen exhibited synergism. A mechanism for mixing organic compounds with inert gases prior to application to fire was proposed. In this arrangement, the organic compound is stored in a separate container and is sprayed into the inert gas as it leaves the high-pressure cylinder. A theoretical study showed that droplets of the two bromoalkenes with diameters as large as 100 µm will be vaporized in less than 1 s and that a completely dry gas will be produced. Acknowledgment This work was supported by NASA Kennedy Space Center under Grant NAG10-0253. Nomenclature
T is in K, p* is in atm.
Table 4. Diffusion Coefficients of Bromoalkenes in Nitrogen at 25 °C compound
diffusion coefficient (m2/s)
CH3-CHdCHBr CH2dCBrCF3
9.704 × 10-6 8.987 × 10-6
at 25 °C are given in Table 4. Details of the procedure for calculation of the vapor pressure and diffusion coefficient of the two bromoalkenes are listed in Appendixes A and B. Equation 8 gives the time needed to completely vaporize a droplet of size dp in terms of temperature and the amount of vapor in the gas (relative humidity). The results of calculations for drying times of droplets of 1-bromo-1-propene and 2-bromo- 3,3,3-trifluoro-1propene in nitrogen at 25 °C are given in Figures 4 and 5, respectively. The drying time for 1-bromo-1-propene is about three times larger than that for 2-bromo-3,3,3trifluoro-1-propene. This is because the vapor pressure of 1-bromo-1-propene (0.27721 atm at 25 °C) is much lower than that of 2-bromo-3,3,3-trifluoro-1-propene (0.75478 atm at 25 °C). It takes about 7 × 10-3 s for a 10-µm droplet of 1-bromo-1-propene (at 80% relative humidity) to completely vaporize. For a 100-µm droplet, the drying time is about 0.7 s. The results clearly show that the drying time is very short and that all of the droplets of the organic compound will be vaporized before the mixture reaches a fire.
C ) extinguishing concentration, % (v/v) D ) diffusion coefficient dp ) particle diameter HBN ) hydrogen bond number M ) molecular weight, g/gmol p ) total pressure p* ) vapor pressure of liquid pv ) partial pressure of vapor in the gas R ) gas constant r ) molecular radius, nm rAB ) molecular separation at collision, nm T ) temperature of gas T0 ) temperature of droplets Tb ) normal boiling point Tm ) melting point t ) time Vair ) volumetric flow rate of air, L/min Vsup ) volumetric flow rate of suppressant, L/min v ) molar volume of liquid at the normal boiling point Greek Symbols AB ) energy of molecular attraction λ ) mean free path Fp ) particle density σ ) external rotational symmetry number τ ) empirical measure of the effective number of torsional bonds
Appendix A: Calculation of Vapor Pressure Myrdal et al.12 developed the following equation for estimating pure-component vapor pressures of complex
Ind. Eng. Chem. Res., Vol. 40, No. 21, 2001 4653
organic molecules
[56.5 - 19.2 log(σ) + 9.2τ](Tm - T) 19.1T [86.0 + 0.4τ + 1421HBN](Tb - T) + 19.1T Tb (-90 - 2.1τ) Tb - T (A.1) - ln 19.1 T T
log p ) -
(
)
For liquids, the first term of eq A.1 is eliminated. Therefore, the equation is reduced to
log p ) -
[86.0 + 0.4τ + 1421HBN](Tb - T) + 19.1T Tb (-90 - 2.1τ) Tb - T (A.2) - ln 19.1 T T
(
)
The results of vapor pressure calculations for the two bromoalkenes are given in Table 3. Appendix B: Calculation of Diffusion Coefficient The diffusion coefficient of gaseous mixtures can be estimated from the following equation13
DAB ) 10-4(1.084 - 0.249x1/MA + 1/MB)T3/2x1/MA + 1/MB pt(rAB)2f(κT/AB)
(B.1)
where
rAB )
rA + rB 2
(B.2)
r ) 1.18v1/3
(B.3)
) 1.21Tb K
(B.4)
f(κT/AB) is the collision function and is given by Figure 2.5 of ref 15.
The results of calculations for the diffusion coefficient of the two bromoalkenes in nitrogen are given in Table 4. Literature Cited (1) Tapscott, R. E.; Mather, J. D. Development of a Tropodegradable Total-Flooding Agent. Phase 2: Initial Screening; Report NMERI 96/22/30930; University of New Mexico, Albuquerque, NM, 1996. (2) Mather, J. D.; Tapscott, R. E. Tropodegradable Halocarbons and Main Group Element Compounds. In Proceedings of Halon Options Technical Working Conference; The University of New Mexico: Albuquerque, NM, 1999; Paper PP132-141. (3) Mather, J. D.; Tapscott, R. E. Tropodegradable Bromocarbon ExtinguishantssProgress Overview. In Proceedings of Halon Options Technical Working Conference; The University of New Mexico: Albuquerque, NM, 2000; Paper PP154-163. (4) Nyden, M. R.; Yang, J. C.; Mather, J. D. Screening of candidate fire suppressants. In Proceedings of Halon Options Technical Working Conference; The University of New Mexico: Albuquerque, NM, 2000; Paper PP104-111. (5) Twarowski, A. Combust. Flame 1993, 94, 91. (6) Fisher, E. M.; MacDonald, M. A.; Jayaweera, T. M.; Gouldin, F. C. Combust. Flame 1998, 116 (1/2), 166. (7) Ibiricu, M. M.; Gaydon, A. G. Combust. Flame 1962, 8, 51. (8) Kim, A. K.; Su, J. Z.; Crampton, G. P. In Proceedings of Halon Options Technical Working Conference; The University of New Mexico: Albuquerque, NM, 1999; Paper PP331-335. (9) Su, J. Z.; Kim, A. K.; Liu, Z.; Cramption G. P. In Proceedings of Halon Options Technical Working Conference; The University of New Mexico: Albuquerque, NM, 1999; Paper PP336-342. (10) Hamins, A.; Gmurczyk, G. W.; Grosshandler, W.; Rehwoldt, R. G.; Vazquez, I.; Cleary, T.; Presser, C.; Seshadri, K. Flame Suppression Effectiveness. Evaluation of Alternative In-Flight Fire Suppressants for Full-Scale Testing in Simulated Aircraft Engine Nacelles and Dry Bays; Grosshandler, W. L, Pitts, W. M., Gann, R. G., Eds.; NIST Special Publication 861; NIST: Gaithersburg, MD, 1994. (11) Hinds, W. C. Physical and Chemical Changes in the Particulate Phase. In Aerosol Measurement; Willeke, K., Baron, P. A., Eds.; Van Nostrand Reinhold: New York, 1993. (12) Myrdal, P. B.; Yalkowsky, S. H. Estimating Pure-Component Vapor Pressures of Complex Organic Molecules. Ind. Eng. Chem. Res. 1997, 36, 2494. (13) Wilke, C. R.; Lee C. Y. Ind. Eng. Chem. 1955, 47, 1253. (14) Hirschfelder, J. O.; Bird, R. B.; Spotz, E. L. Trans. ASME 1949, 71, 921. (15) Treybal, R. E. Mass Transfer Operations, 3rd ed.; McGrawHill Book Company: New York, 1980.
Received for review March 28, 2001 Revised manuscript received June 18, 2001 Accepted July 7, 2001 IE010281Z
