Martin Hertzberg

Flammability Of Methane, Propane, And Hydrogen Gases

This paper reports the results of flammability studies for methane, propane, hydrogen, and deuterium gases in air conducted by the Pittsburgh Research Laboratory. Knowledge of the explosion hazards of these gases is important to the coal mining industry and to other industries that produce or use flammable gases. The experimental research was conducted in 20 L and 120 L closed explosion chambers under both quiescent and turbulent conditions, using both electric spark and pyrotechnic ignition sources. The data reported here generally confirm the data of previous investigators, but they are more comprehensive than those reported previously. The results illustrate the complications associated with buoyancy, turbulence, selective diffusion, and ignitor strength versus chamber size. Although the lower flammable limits (LFLs) are well defined for methane (CH4) and propane (C3H8), the LFLs for hydrogen (H2) and its heavier isotope deuterium (D2) are much more dependent on the limit criterion chosen. A similar behavior is observed for the upper flammable limit of propane. The data presented include lower and upper flammable limits, maximum pressures, and maximum rates of pressure rise. The rates of pressure rise, even when normalized by the cube root of the chamber volume (V1/3), are shown to be sensitive to chamber size.

20‐l explosibility test chamber for dusts and gases

The Bureau of Mines has designed a 20‐l test chamber for the explosibility testing of dusts, gases, and their mixtures. It can be used to measure lean and rich limits of flammability, explosion pressures and rates of pressure rise, minimum ignition energies, minimum oxygen concentrations for flammability, and amounts of inhibitor necessary to prevent explosions. The 20‐l chamber can be used at initial pressures that are below, at, or above atmospheric as long as the maximum explosion pressure is less than 21 bar, which is the rated pressure of the chamber. The chamber instrumentation includes a pressure transducer, optical dust probes, an oxygen sensor, and multichannel infrared pyrometers. Ignition sources used include electric sparks and electrically activated chemical ignitors. Examples of the various types of data that can be obtained for dusts and gases are shown.

Selective diffusional demixing: Occurrence and size of cellular flames☆

The data on cellular flame structures, including those recently reported by several independent investigators, are reviewed and analyzed in terms of the established mechanisms and concepts of homogeneous flame propagation. A perturbation analysis is used to show how the balance between the processes of selective diffusional demixing and flame stretch readily accounts for the composition domains in which cellular flames occur, and also predicts their dimensions. The important, determining properties of the fuel-oxidant mixture are: Di − Dj, the difference in molecular diffusivity between the more diffusive component, i, and the less diffusive component, j; and, ∂Su/∂Ci, the slope of the burning velocity curve with respect to the concentration of the more diffusive component. Cellular structures are stable only for stoichiometries in which the more diffusive component, i, is also the deficient component, so that ∂Su/∂Ci > 0. The perturbation analysis also shows that, at the opposing stoichiometry, where the more diffusive component is present in excess, so that ∂Su/∂Ci < 0, selective diffusion tends to erase cellular structures and to stabilize the laminar or one-dimensional ones. A phenomenological theory based on a quantitative balance between selective demixing and flame stretch gives the following result for the cellular flame diameter, dcell: dcell(Di − Dj)(∂Su∂Ci ≈ α2(ρuρb) where α is the effective diffusivity (thermal and mass) across the flame front, and ϱu/ϱb is the expansion ratio. That equation adequately predicts the absolute magnitude of the cellular flame diameters; their dependence on the difference in the molecular diffusivities between fuel and oxidant; their dependence on pressure; their dependence on stoichiometry; their dependence on flow velocity when measurements are made with nonadiabatic burners, their dependence on molecular weight of added inertants; and the influence that selective diffusion exerts on the flammability limits. The same analysis also shows that when Di − Dj and/or ∂Su/∂Ci are very large, there is a natural lower limit on the raduis of curvature of the cellular flame. That limit is reached when the process generates a composition at the cell apex that corresponds to the most reactive mixture, and a composition at the cell edge that is at, or beyond, the flammability limit. The minimum value of dcell cannot then be smaller than the tubular quenching diameter for the most reactive mixture composition.

Devolatilization wave structures and temperatures for the pyrolysis of polymethylmethacrylate, ammonium perchlorate, and coal at combustion level heat fluxes

Abstract Data are presented for the surface pyrolysis temperatures, T s , and for the develatilization mass loss rate per unit area, ṁ, for polymethylmethacrylate (PMMA) exposed to a laser beam at incident flux levels of 15–500 W/cm 2 . After a brief induction time during which the surface is being heated to T s , the measured ṁ values are thereafter steady state in time, and are linearly related to the net absorbed flux, I abs . The measured values for ṁ are in good agreement with that derived from the first law of thermodynamics: m = I abs ∫ T s T o C(T) dT+δ H v (T s ) −1 Kinetically, this equation represents a devolatilization wave front whose propagation rate and maximum temperature are both heat transport controlled, and determined by the magnitude of the driving flux, I abs . The available data for ammonium perchlorate and for coal are also shown to obey a similar relationship, and the same heat transport-controlled process appears to govern their pyrolysis and devolatilization kinetics. At high fluxes, the measured ṁ and T s values reported here for PMMA, and those reported by other investigators, can be interrelated to one another through an Arrhenius plot whose slope gives an “activation energy” that is essentially equal to the heat of vaporization, ΔH v ( T s ). That measured relationship between ṁ and T s is thus a reflection of the Clausius-Clapeyron equation for the equilibrium vapor flux emanating from the polymer melt at its surface decomposition temperature, T s . However, at low fluxes, the measured ṁ values are significantly lower than those predicted by the Clausius-Clapeyron line. The departure is attributed to the additional presence of mass transport limitations. Photographic data are presented for bubble structure and bubble location within the quenched PMMA samples that independently confirm the increasing significance of mass transport limitations at the lower fluxes.

Flammability limit measurements for dusts and gases: Ignition energy requirements and pressure dependences

Data are presented for the flammability limits of Pittsburgh Seam bituminous coal dust, polyethylene powder, and methane in air at pressures in the range 0.5 to 3 bar. Explosibility test chambers of 20 and 120 L were used, and ignitability limitations were overcome with efficient pyrotechnic ignitors with nominal energies of 500 to 5,000 J. The propagation criterion used was based on the maximum explosion pressure and the size-normalized maximum rate of pressure rise. The latter is a dynamic criterion that tends to minimize “overdriving” effects as the 20-L data are taken to their asymptotic limits at high ignition energies. The measured lean limits in air at atmospheric pressure are 90 g/m 3 for the coal dust, 35 g/m 3 for polyethylene, and 4.9 vol pct for methane. The rich limit for methane is 18–19 vol pct, whereas the dusts have no rich limits out to concentrations as high as 4,000 g/m 3 . A linear, lean-limit pressure dependence was measured for the dusts which was essentially the same as the pressure dependence measured for methane when all are expressed in comparable units: namely, mass concentration of fuel per unit volume of air. This observation further confirms a lean limit dust flame propagation mechanism that is controlled by the gas-phase reaction rate. The limit concentrations are then determined by the dust loading required to generate a lean limit concentration of pyrolysis products in the volatiles-air mixture, and the dust behaves as an equivalent “homogeneous,” premixed gas, regardless of the initial pressure.

Metal dust combustion: Explosion limits, pressures, and temperatures

Measurements are reported for the explosibility behavior in air of a variety of metallic and other elemental dusts. The data are useful for evaluating the hazards involved in their manufacture, transport, storage, and use. This study was also designed to help resolve, a fundamental issue of whether the combustion reactions in those dust flames proceed homogeneously in the gas phase or heterogeneously at particle surfaces. Data are reported for 14 dusts: from the more volatile metals such as magnesium and zinc, to the refractory metals such as tantalum, tungsten and niobium. Also included are some intermediate metals: hafnium, titanium, aluminum, iron, lead and copper; as well as the nonmetallic elements: boron, silicon, and carbon. Explosion pressures, rates of pressure rise, and continuum radiation temperatures were measured as a function of the dispersed dust concentration, using the Bureau of Mines 20-L explosibility test chamber and strong pyrotechnic ignitors. The lean limit fuel equivalence ratios varied from as low as 0.13 for the more reactive and volatile dusts to well in excess of 1.0 for the less reactive and less volatile ones. The copper and lead dusts were nonexplosible. For some dusts, preliminary data on particle size dependencies were obtained. The data are analyzed in terms of calculated adiabatic flame temperatures and the equilibrium vapor pressures or nonequilibrium vapor fluxes at those temperatures. An inverse correlation is observed between the measured lean limit equivalence ratio and the calculated equilibrium flame zone volatility for the dust. The correlation is analogous to one previously established for the carbonaceous dusts. Estimated absolute vaporization fluxes for the more flammable dusts are adequate to support a homogenous mechanism. Some refractory dusts, however, exhibit marginal (but finite) explosibility for the finer particle sizes even though their estimated vaporization fluxes appear to be too low to support a homogeneous mechanism.

Effect of volatility on dust flammability limits for coals, gilsonite, and polyethylene

Lean limit concentrations and limestone rock dust inerting requirements were measured in a 20-L chamber for a range of carbonaceous dusts including various ranks of coal, gilsonite, and artificial mixtures consisting of polyethylene and graphite. Although there is some uncertainty in the true yield of volatiles for some of the fossil mineral dusts, especially at flame flux levels, the data show that the combustible volatile content of the carbonaceous dusts is the dominant factor governing their combustion behavior in the dust explosions. The char residues and graphite are essentially inert on the rapid time scale required for flame propagation processes in explosions. The lean flammability limits of the various dusts correspond to invariant combustible volatile concentrations of about 30 to 40 g/m 3 . Total interting of the dusts occurs when the inert content of the mixture exceeds about 90 to 95 wt pct or, equivalently, when the combustible volatile content is less than 5 to 10 wt pct. The chars role is equivalent to that of the other inert constituents, such as ash and rock dust. In addition to the 20-L explosibility experiments, volatile yields of the various coals were measured using a high power, carbon dioxide laser at flux levels of 110 to 115 W/cm 2 . These measurements provide new data on the volatilities of 110 μm particles of the various coals, but some uncertainties remain in the absolute volatilities of the various sized coals used for the 20-L tests. A relatively simple volatility model for the various fossil mineral dusts can be used to explain their explosion propagation behavior and their extinction limits. When the uncertainties in volatility were removed by using the artificial mixtures of a completely volatilizable fuel (polyethylene) and the nonvolatilizable graphite, the model was confirmed.

Domains of flammability and thermal ignitability for pulverized coals and other dusts: Particle size dependences and microscopic residue analyses

The particle size dependence for the lean limits of flammability in air of Pocahontas coal dust (16% as received volatility, by ASTM proximate analysis), Pittsburgh coal dust (35% volatility), and polyethylene powder (100% volatility) were measured using narrow size distributions with average diameters ranging from 2 μm to over 400 μm. In all cases, the lean limits measured in an 8-liter chamber became insensitive to particle size below some characteristic diameter. Above those characteristic diameters, the lean limit concentrations increased significantly with increasing particle size until a critical size was reached, above which the dust was nonflammable for any concentration at ambient temperature and pressure. Both the characteristic diameter and the coarse size limit of flammability (critical diameter) were observed to increase monotonically with increasing dust volatility and increasing oxygen content of the dispersing gas. The same narrow size distributions were also used to measure the particle size dependenceof the thermal ignitability of those same dusts dispersed into preheated air in a 1-liter furnace. The minimum autoignition temperatures (AITs) display behavior similar to that of the lean limit concentrations, except that heating of the mixture to the elevated temperature required for such thermal ignitability studies markedly increases its flammability range. At those elevated initial temperatures, the characteristic sizes are shifted to larger diameters and no critical sizes were observed for some of the AIT measurements even at the coarsest sizes studied. Also presented are scanning electron microscope (SEM) data that reveal the structuralchanges in coal particles that resulted from their participation in dust explosions. Char residues that are pockmarked with blow holes are the typical remnants from all dust explosions of bituminous coals. At high dust concentrations, those char residues are fused into large, agglomerated masses. Preignition and postignition residues sampled from the thermal ignitability furnace show the same pockmarked appearance for their char residues. In another experiment, using a high power, pulsed CO 2 laser beam, it was possible to carefully control the pyrolysis exposure time and to obtain SEM photomicrographs during the early stages of the pyrolysis-devolatilization process. A theory is presented which isolates the important physical processes in their simplestform. The processes are: (1) heating and devolatilization of the dust particles, (2) mixing of those volatiles with air, and (3) gas-phase combustion of the volatiles-air mixture. There are two loss processes that compete with that propagation sequence and can therefore quench propagation: (a) natural convection or buoyancy, operating through the mechanism of flame stretch, and (b) conductive-convective heat losses to dust particles, which, in addition to being sources of needed fuel, are also heat sinks. The measured insensitivity of the lean limit concentrations to particle size for the finer dusts reflects propagation under rate control by process 3 and quenching by loss process a. A particle size dependence appears at the characteristic diameter because rate control of propagation is shifting from process 3 to process 1, while quenching is still by loss process a. The dust mixture is nonflammable above some critical diameter when the propagation process is entirely rate controlled by process 1 and quenching occurs by a combination of loss processes a and b.

Volatility model for coal dust flame propagation and extinguishment

A theoretical analysis is presented for the propagation and extinguishment of coal dust flames and of dust and gas flames containing inhibitor powders. The analysis is based on the established mechanisms for homogeneous flame propagation and the well known concept of a constant limit flame temperature for a given class of homogeneous fuels. The analysis is expanded to phase-heterogeneous systems such as coal dust by means of a volatility model. The analysis includes the singly heterogeneous system of a solid fuel dust in air; the singly heterogeneous solid inhibitor dust in a homogeneous fuel-air flame: and the doubly heterogeneous system consisting of a solid fuel and inhibitor dust mixture in air. The data for measured explosion pressures, flammability limits, and extinguishant requirements for heterogeneous systems are shown to be consistent with the established mechanisms and processes for homogeneous flame propagation provided that one adds an additional process: the heating and devolatilization of the solid fuel or inhibitor. The limitations on the rates of devolatilization of the solid particles become rate controlling at high burning velocities, at high dust loadings, and for large particle sizes. Devolatization rates are controlled by the intrinsic devolatilization rate constant for the solid fuel or inhibitor and the effective heating flux of the approaching flame front. The effective vield of volatiles is a function of those factors, the decomposition chemistry, and the time available for devolatilization. The fraction of the total volatiles that can be generated in the time available is the β-factor, and it determines the effective yield of fuel or inhibitor that participates in the flame propagation process. The data for explosion pressure, Flammability limits, and extinguishant requirements are readily understood in terms of those β-factors.

Infrared temperatures of coal dust explosions

Abstract A comprehensive set of temperature measurements is presented for coal dust explosions as a function of dust concentration. Data include the dependence of explosion temperature on particle size, volatility, oxygen content, and added rock dust. The majority of the data are for constant volume explosions in a laboratory-scale vessel, but data from a coal dust burner flame and full-scale mine dust explosions are also included. In all of the measurements, the gas temperature was significantly higher than the dust particle temperature. The measurements were made with a three-wavelength near-ir pyrometer and a six-wavelength ir pyrometer (in the 1–5-μm-wavelength region). Some thermocouple measurements are shown for comparison.