Search Documents

By Yongjun Xian

The release of fluid inclusions has a strong potential for the unintentional activation of pyrite during flotation. The present study aims to characterize fluid inclusions in a natural pure pyrite. The results indicate that a certain amount of fluid inclusions exist in the pyrite. The fluid inclusions assume a variety of shapes, including long strips, oval shapes and irregular shapes, with sizes that range from a few microns to dozens of microns. From the EDS results of hollow positions that resulted from the fracture of the fluid inclusions on the pyrite surface, certain amounts of Al, Cr, C, O, Ag and Cu were clearly detected; these elements were not detected in FeS2 pyrite itself. The contents of the inclusions were released into solution when the sample was subjected to a grinding process. Under experimental conditions of 2 g of pyrite cleaned in 40 mL of pure deionized water under an inert atmosphere, the concentrations of Cu, Fe, Cl- and SO4 2- released from the inclusions reached concentrations of 3.29 × 10-6, 32.52 × 10-6, 51.57 × 10-6 and 75.90 × 10-6 mol/L, respectively. These values are significantly greater than those from the experimental nonoxidative and oxidative dissolution of the pyrite. The fluid inclusions of pyrite were, therefore, concluded to represent the dominant sources of Cu, Fe, Cl- and SO4 2- in the aqueous solution. The present investigation provides a new understanding of the source of the unavoidable metal ions in the pulp and may benefit understanding of the flotation theory and environmental geochemistry.

By R. A. Franks, G. F. Friel, J. C. Edwards, L. Yuan

Five mine fire experiments were conducted in a 2.08m high and 2.90m wide, ventilated mine entry in the National Institute for Occupational Safety and Health (NIOSH)'s Safety Research Coal Mine (SRCM) to determine the effect of the dispersion of carbon monoxide (CO) on mine fire detection. CO measurements were made at distances from 7.6 m to 45.2 m downwind from the fire with diffusion mode CO sensors positioned near the roof. For small intensity fires, less than 30 kW heat release rate, generated by 14 kg coal in a 0.61 m square tray, it was determined that air flow and sensor spacing were significant for fire detection at the 10 ppm CO alarm level. Within 15.0 m downwind distance from the fire, 10 ppm CO alarm values occurred for volumetric air flows less than 11.5 m3/s. However, the 10 ppm CO alarm value did not occur 30.0 m downwind from the fire for air quantities greater than 6.2 m3/s due to dilutive mixing of the CO in the air stream. The criterion that the mine fire alarms occur within 15 min of the onset of flaming combustion could not be consistently met with the 10 ppm CO alarm. This suggests the use of lower CO alarm values, or reduced CO sensor spacings for mine fire protection. It is demonstrated how computational fluid dynamics (CFD) can be used to model the CO dispersion downwind from the fire in support of a plan to optimize sensor spacings.

By R. Larry Grayson

Barrett EA, Kowalski KM [1995]. Effective hazard recognition training using a latent-image, three-dimensional slide simulation exercise. Pittsburgh, PA: U.S. Department of the Interior, Bureau of Mines, RI 9527. NTIS stock number: PB95-170379. Beus MJ, Orr TJ [1995]. Computer modeling and analysis of the Greens Creek Mine, Admiralty Island, AK. Spokane, WA: U.S. Department of the Interior, Bureau of Mines, RI9530. NTIS stock number: PB95-176418. [ ]

By R. G. Gürtunca

Over the past century coal miner safety and health has seen tremendous improvements, the fatality and injury rates continue to decrease while productivity continues to increase. Many of the hazards that plagued miners in the past, such as coal bumps, methane and coal dust explosions, ground fall accidents and health issues have been significantly reduced. The contribution of NIOSH research includes products for prevention and survival of mine fires, methane control measures, design procedure for underground coal mines, methods for excavation surface controls, methods and procedures for blasting, laser usage in underground mines and prevention of electrocution from overhead power lines that have reduced accidents and injuries in underground coal mines. Health research has produced products such as the personal dust monitor, noise abating technologies and ergonomic solutions for equipment operators. Research priorities at NIOSH are set by considering surveillance statistics, stakeholder inputs and loss control principles. Future research in coal mining is directed towards respiratory diseases, noise induced hearing loss, repetitive musculoskeletal injuries, traumatic injuries, falls of ground and mine disasters. The recent spate of accidents in coal mines resulted in the Miner Act of 2006 which includes a specific role for NIOSH in future mine safety research and development. The mine safety achievements in the USA reflect the commitment of industry, labour, government and research organizations to improving the safety of the mine worker.

By Ray Brainard

Charleston Water System?s Daniel Island Extension Tunnel, initially thought to be the simplest of the four replacement tunnels to design and build, became the most challenging after the discovery of an unexpected 30-ft thick lens of sand in the normally consistent Cooper Marl while sinking a shaft on Daniel Island. This paper will explore the challenges involved with determining the extent of the unforeseen lens of sand, the accelerated boring program, and the redesign of the vertical alignment?keeping the alignment between the sand layer and the bottom of the Cooper River. It will describe how the challenges were overcome, and how the Owner, Engineer, and Contractor were able to team together to achieve the desired results.

By B. Silliman

THE occurrence of diamonds of some size in the gold-fields of California is by no means uncommon, and was noticed by me in a communication, to the California Academy of Science in 1867, when specimens of this gem, from at least five different localities, were exhibited. I then suggested that a more attentive examination of the heavy sands left in the sluices of hydraulic washings would in all probability detect diamonds, mingled with other rare species not commonly believed to have occurred in these sands. Mr. George A. Treadwell, of San Francisco, has 1at.ely sent me a small package of these sands, collected by him from the sluices of the " Spring Valley Gravel Mining Claim," Cherokee, Butte County, California. A microscopic examination shows these sands to abound in beautiful colorless zircons (hyacinths), of the form well known in the hyacinths of Expailly (France), associated with crystals of topaz, quartz in fragments, rounded grains 'of chromic and titanic iron, and

By R. C. McFarlane, T. D. Mueller, J. E. Walstrom

In evaluating uncertainty, experiments are usually performed repeatedly and then conclusions are drawn from the distribution of results. With the advent of high-speed electronic computers, it is possible to perform experiments using mathematical models constructed to simulate complex experiments or operations. Statistical methods are then applied to the results of the simulated experiments. This procedure forms the busis of this paper. Demonstrated is the need for properly accounting for uncertainty in petroleum engineering problems. How uncertainty affects solutions is evaluated in three example illustrations. The method used to evaluate uncertainty in petroleum engineering studies is the Monte Carlo simulation procedure.'-" INTRODUCTION The solution to most technical problems may be derived from interrelationships among several quantities called variables or parameters. There may be only a few variables or several hundred. Interrelationships among parameters may be explicit or implicit, well established or only approximate. Some variables that fully or partially depend on the magnitude of others are called dependent variables. Input variables for most practical problems are not precisely known; there is usually an uncertainty in their value. The degree of uncertainty may vary from one variable to another. Variables that are known accurately are called determinates.' For instance, the gravity of crude obtained from a particular pool may be known precisely, and therefore is a determinate. The degree of precision with which a quantity can be determined increases as data describing the pool are accumulated during the development of the field and the producing life of the pool. The uncertainty of a parameter may result from difficulty in directly and accurately measuring the quantity. This is particularly true of the physical reservoir parameters which, at best, can only be sampled at various points, and which are subject to errors caused by presence of the borehole and borehole fluid or by changes that occur during the transfer of rock and its fluids to laboratory temperature and pressure conditions. Uncertainty may also result in attempting to predict future parameter values. This type of uncertainty is particularly evident in investment analyses involving future costs, prices, sales volumes and product demand. Uncertainty in the solution to investment problems is often called risk, and its study is called risk analysis.' Uncertainty also enters into biological and sociological analyses in which indeterminate factors are often important due to limited control of the experimental material. It is customary, in evaluating uncertainty, to perform repeated experiments and to draw conclusions from the distribution of the results of these experiments. With the advent of the high-speed electronic computer, it is possible to construct mathematical models which simulate complex experiments or operations and to perform the experiments repeatedly, utilizing the models. Statistical methods are then applied to the results of the simulated experiments This method forms the basis of the investigation reported here. PROBABILITY DISTRIBUTIONS FOR VARIABLES The uncertainty in the value of a variable may be indicated by a probabilistic description accomplished by expressing the quantity by a probability distribution. Many recognized probability distributions can be used to describe physical quantities. Recent studies used various types of distributions to describe core analysis data.',' However, for the examples in this paper, the uniform and triangular distributions are believed to reasonably approximate the data used (Fig. 1). The uniform distribution confines the variable between an upper and a lower limit. The variable may lie anywhere between the two limits. This distribution is used when no one range of values for a variable is more probable than any other, but information or intuitive reasoning indicates the variable will lie somewhere between the chosen limits. The triangular distribution is used for a variable when more data are available to indicate a central tendency of distribution. This allows postulating a "most likely" value to the distribution and upper and lower limits. In this case, as for the uniform distribution, the variable is not expected to assume a value less than the lower limit or greater than the upper limit. However, with improved quality of data it can be postulated that the variable will tend to assume a value close to the most likely value, and that there will be a decreasing probability for values away from the most likely value. The area under either of these probability distributions is equal to unity since it is assumed that there is a 100 percent probability that the variable will lie somewhere under the curve. An ordinate erected at any particular value of the variable divides the area under the curve into two parts: the area to the left of the ordinate represents the probability that the value of the variable will be equal to or less than the value of the variable at the position of the ordinate, and vice versa. The probability is zero that the variable will have any specific deterministic value. If two ordinates are drawn for any two values of the variable, the probability that the variables will have a value lying between these ordinates is equal to the area under the curve lying between the ordinates.

By J. J. Taber, I. S. K. Kamath, R. L. Reed

Alcohol floods of consolidated sandstone cores have shown the process to be strongly dependent on the phase behavior of the particular alcobol-oil-water system used. This means that in many cases the mechanism does not conform to the idea of a piston-like displacement. Instead, it is found that by changing the alcohol it is possible to change the relative velocities of the oil and water and, in fact, the entire mechanism of the process. The effects of rate, viscosity, initial saturation, distance travelled and hysteresis of relative permeability on the alcohol flooding mechanism are discussed. INTRODUCTION Reasons for interest in the use of alcohol to miscibly displace oil and water from a porous medium appear in the existing literature.l-7 The mechanism of the displacement has been considered and the apparent implications formulated into a theory which presumably would enable one to predict the essential features of the process.6 Unfortunately, most of the reported experiments have been performed with unconsolidated or artificially consolidated sands. With these systems some of the noteworthy facets of the process are obscured and resulting data appear uncertain. It is the purpose of this paper to show how the use of consolidated sandstones has led to revision of the mechanism and, hence, the theory of alcohol flooding. The practical result is increased pessimism toward the possibilities of commercial application of the simplest form of the alcohol-slug process. However, elucidation of the mechanism has made it possible to define the essential characteristics of a system of slugs which will behave in a nearly piston-like fashion and, thus, yield the best possible result. EQUILIBRIUM PHASE BEHAVIOR Fig. 1 is a diagram of the ternary system isopropyl alcohol (IPA)-Soltrol-calcium chloride brine. Brine was used to prevent plugging of the core and calcium chloride was used because sodium-chloride brine exhibits a solid phase with Soltrol and IPA. If alcohol is added in increments to the immiscible mixture of water and oil represented by Point A, the path followed by the successively equilibrated samples will be on the Line ABC and pass from the immiscible region to the miscible region by crossing the binodal curve at B. Consider the intersection D of this path with the tie Line EF. The quantity of oleic phase is proportional to the Segment ED and the quantity of the aqueous phase is proportional to DF. Compositions of the two phases are specified by Points E and F. It is clear that in the case shown the oleic phase is diminishing and entirely disappears when miscibility is achieved. Such will

By Andrew B. Cecala

While various types of equipment are used in bulk material processing plants to control respirable dust, the equipment often doesn't adequately protect workers. One cost-effective way to supplement your existing dust control equipment is to install a whole-plant ventilation system. This article outlines research by the National Institute of Occupational Safety and Health (NIOSH) on using whole-plant ventilation in two mineral processing operations to reduce respirable dust levels. Information also covers how to design a ventilation system for any bulk material processing plant.

By Thomas M. Barczak, Dennis R. Dolinar, Gabriel S. Esterhuizen

Longwall mines typically use some form of standing support for secondary roof support in longwall tailgate entries. Although there have been several new support products developed for this application, there remains no universal design criteria to optimize the application of these support technologies. The requirement for optimization and proper support selection is to understand the degree of control that the support has on the ground behavior. The ground reaction curve and numerical modeling was used to evaluate the impact of standing support on ground behavior. LaModel was used to evaluate the impact of standing support on main roof and floor behavior and pillar yielding. The conclusion drawn from this study was that standing supports do not have sufficient capacity to control main roof or floor loading or prevent the resulting convergence of the tailgate entry. However, it is imperative that this “uncontrollable convergence” be considered in the support design to prevent premature failure of the support. A FLAC model was used to evaluate the near-seam roof and floor behavior in conjunction with the global vertical and horizontal stresses. The model suggests that standing roof supports can have some impact on the ground behavior as the elastic response of the rock is exceeded and rock structure deteriorates from the stress concentrations that develop around the tailgate opening. During this phase, the capacity and stiffness of the standing support can be critical to the stability of the opening, as eventually the rock mass will be transformed into a partially detached structure whose weight must be supported by the standing support. This work is part of a ground control program of research at the National Institute for Occupational Safety and Health (NIOSH) aimed at improving mine safety by reducing roof fall injuries and fatalities.

By T. R. Waters, F. Li, R. L. Huston

This paper describes the development of a multi-body biomechanical model that can be used to assess the risk of low back disorders due to occupational exposure to jarring and jolting from operation of heavy mobile equipment (e.g., trucks, haulers, graders, tractors, etc.) The four-element dynamic spinal model is capable of providing estimates of the force and acceleration responses of the head/neck, upper, lower, and middle torso resulting from vibration input collected in field studies. The paper also presents a method for using the model output to assess an individual's risk of low back disorder due to a specific jarring and jolting exposure.

By Gerald J. Joy, Paul J. Middendorf

This study examines the patterns and trends in noise exposure documented in data collected by Mine Safety and Health Administration inspectors at U.S. coal mines from 1987 through 2004. During this period, MSHA issued a new regulation on occupational noise exposure that changed the regulatory requirements and enforcement policies. The data were examined to identify potential impacts from these changes. The overall annual median noise dose declined 67% for surface coal mining and 24% for underground coal mining, and the reduction in each group accelerated after promulgation of the new noise rule. However, not all mining occupations experienced a decrease. The exposure reduction was accompanied by an increase of shift length as represented by dosimeter sample duration. For coal miners exposed above the permissible exposure level, use of hearing protection devices increased from 61% to 89% during this period. Participation of miners exposed at or above the action level in hearing conservation programs rapidly reached 86% following the effective date of the noise rule. Based on inspection data, the occupational noise regulation appears to be having a strong positive impact on hearing conservation by reducing exposures and increasing the use of hearing protection devices and medical surveillance. However the increase in shift duration and resulting reduction in recovery time may mitigate the gains somewhat.

By Brian G. White, Theodore J. Williams, Jeffrey K. Whyatt

Detailed mapping and examination of the site of a large rock burst at the Lucky Friday Mine, Mullan, ID, demonstrated conclusively that damage was caused by a splitting and buckling failure mechanism initiated by separation of rock into layers along steeply dipping metamorphic shear foliation subparallel to the affected crosscut. The thin tablular layers then buckled into the crosscut from both sides. Rigorous control of the burst by these layers suggests practical measures for avoiding bursts in crosscuts in the future by repositioning and reorienting openings so that splitting and buckling are reduced or precluded.

By E. Gaujot

IN connection with the question of coal waste and economy in mining, we would call the attention of those interested to an apparatus invented by M. Dernencourt, Superintendent of the Anzin Division of the Anzin Coal Company, North of France. This apparatus is known in France and Belgium under the name of prop screw-jack (vis botte), and has been used with good results in the above-named countries for some years past. M. Ponson gives, in his " Trait6 de I'exploitation des Mines," a description of it (page 533), but as the description given is old, the work not widely known in this country, and as some valuable improvements have been made since, I thought it might be useful to give a description of it, having seen it in 'operation in 1867, while in Europe to attend the great Exposition, and on a scientific tour to the mining districts of France, Germany, Belgium, and Italy. The prop screw-jack is composed of the following pieces : I. The body or column B, Fig. 4 [Plate (I)], of oak or yellow pine, sawed square, the bark taken off, and any length, proportionate to the vein. The smallest are 10 inches diameter at the base, and 9 inches at the top, the height being from 12 to 15 inches less than the size of the vein. In the axis of this column, and at the top end, is a cylindrical opening or hole, C, 22 inches diameter, 10 inches deep, for the screw, S. II. Two wrought-iron rings, R R, to strengthen the two extremities of the column ; the rings must be put on hot. III. A cast,-iron washer, E, with hole in centre a little larger than the screw, to be put on the top of the column, and on which the nut is to rest; the top side of this washer must be faced on the lathe, also the bottom of the nut, to prevent friction. IV. The screw, S, having a total length of 12 inches and a diameter of 24 inches (for very high veins this diameter may be increased to 3 inches), and has on its top a wrought-iron plate, T,

By J. M. Adams

It is my purpose to give some of the results obtained by an experience of nearly seven years in working ores by the method frequently called the Washoe Process, and in several mills of which I have had charge, but principally in the Owyhee Mill, at Silver City, Idaho, which had twenty 650-pound stamps, and sixteen pans. I shall discuss here merely the mechanical details for working ores generally, subdividing the subject as follows : 1. Preparation of the ore for the stamps; 2. The crushing in the battery; 3. The settling of sand or pulp in vats or tanks; 4. The treatment in the pans; 5. The results obtained in settlers, agitators, and concentrators; 6. The straining of quicksilver, cleaning of amalgam, and retorting; 7. The saving of slimes and their subsequent treatment; 8. The loss of quicksilver. For descriptions of the various kinds of orebreakers, stamps, tanks, pans, settlers, etc., I must refer to the various works on these subjects, prominent among which are the reports of the United States Commissioner of Mining Statistics,and the volume on "Mining Industry," the third volume of the United States Geological Survey on the Line of the Fortieth Parallel through the gold and silver bearing regions of the great West, undertaken under the able guidance of Clarence King, assisted by Mr. James D. Hague and others. In this discussion it will be assumed that the general arrangement of the quartz mill is understood; and the question will be treated how to secure, from such a mill, the greatest economy in working, combined with the largest results. This place seems fittest for a single preliminary suggestion—namely, that there should be double "floors" throughout the mill, so that nothing can sift through and be lost.

By T. A. Gray, L. M. Crayne, P. Glogowski, Trevits. M. A.

Of the 19 major mine fire events (including thermal events) that have occurred in the last 6 years (2000-2005), it is estimated that remotely constructed mine seals could have been used at 65% of the events to control or suppress the fire. Underground observations of seals that have been remotely installed during mine fire events show that they often do not fully achieve mine roof-to-floor and rib-to-rib closure. Unfortunately, the inability to reliably close the mine void has limited or precluded the regular use of this technology. The National Institute for Occupational Safety and Health is conducting full-scale underground experiments at the Lake Lynn Experimental Mine to identify and remedy existing remote sealing technology shortcomings, to develop novel technologies, and to transfer the new or improved technologies to the mining industry. This paper will discuss the remote mine seal testing program and will provide the results of the in-mine experiments. Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the National Institute for Occupational Safety and Health.

By P. J. A. Beukes, C. C. La Grange

INTRODUCTION Some years ago the authors proposed a modified procedure1 for carrying out micum index determinations2. The modifications consisted of using a drum of 50 cm internal length, i.e. half the length of the standard micum drum (a modification also adopted by the B.S.J.3), and the use of a + 25 mm coke instead of + 60 mm coke for the test. The proposed modified test and formulae for converting results obtained from it to standard micum test results and vice versa were based on the results obtained during the investigation of about 200 test cokes made in ovens of the South African Steel Industrial Corporation Limited (Iscor), the Fuel Research Institute co-operating with the experiments. Extensive testing of cokes has since been done at the Fuel Research Institute using both the standard and the modified procedures. The numerous test results obtained during this trial period fully support and strengthen the confidence originally expressed by the authors in the modified testing procedure. The scope of the modified micum index determination has in the meantime been extended by applying the testing procedure also to test cokes made in the Institute's experimental coke ovens, and also in this case the modified procedure has proved its usefulness. A difficulty is that when comparing the characteristics of cokes made in the Institute's two experimental coke ovens with each other and with those of cokes made from similar blends in commercial coke ovens, for example those of Iscor, the index values obtained invariably do not check exactly due mainly to inherent differences between the different types of oven4. For instance, resistance to abrasion of a coke made in Iscor's ovens is always appreciably higher than that of a coke made from the same blend (and having a similar moisture content) in the experimental ovens. This makes direct comparisons and the prediction of quality to be expected from commercial products, based on experimental coke oven results, difficult. It may also be added that workers in Great Britain some years ago reported a similar experience5. In the authors' opinion the main reason for the discrepancy is the relatively low height of the experimental ovens resulting in a much lower static pressure on the charge during coking than that obtaining in a commercial oven. The difficulty was pointed out and briefly discussed in a publication4 which appeared at a time when the micum test for coke evaluation had only just been introduced at the Institute, so that only B.S. shatter and abrasion test results were available for discussion in the publication. Needless to say, the introduction of the micum test by no means solved nor even alleviated the problem. On numerous occasions during the past few years the Institute has had to conduct contract investigations in its experimental coke ovens on behalf of companies which invariably desired an indication of the quality of coke likely to be obtained from certain coals or blends when coked commercially, more particularly in Iscor's type of coke ovens. It was, therefore, essential to accumulate statistical data which would enabe the prediction to be readily made. Fortunately the compilation of comparative coke-quality index values for the three types of oven under consideration has become possible as a result of coking investigations conducted in co-operation with Iscor over a number of years. During this period Iscor, on numerous occasions, kindly made available to the Institute portions of coals and blends coked in the Corporation's coke ovens, thus enabling parallel coking tests to be carried out in the Institute's experimental ovens. In this way enough statistical data became available to establish relationships between coking results obtained in Iscor's ovens and results obtained on similar coal charges in the Institute's ovens. The most practical and acceptable method of overcoming the problems of finding a suitable and simple basis enabling direct comparison between the cokes of the different origins mentioned to be made, is the assignment of descriptive arbitrary quality ratings, as indicated below, to the cokes obtained, it being argued that the quality rating assigned to coke made from a given coal or blend in the different coke ovens under standardized conditions should be the same, irrespective of differences in the index values obtained when subjecting the cokes to the usual coke evaluation tests. The system, which should enable the sponsors of coking investigations in the Institute's experimental ovens to make their own interpretations of the results obtained, is explained below. ARBITRARY QUALITY CLASSIFICATION OF COKES The Ml0m index1 (percentage material smaller than 10 mm after the test) of a blast furnace coke is a measure of its tendency towards breeze formation during handling -a low value indicating a low tendency. This is generally of greater importance under ruling South African conditions, than the M'40 index1 (percentage material larger than 40 mm after the test), which can be regarded as a measure of resistance to shatter. In fact, if the M10m value of a coke as made in Iscor's ovens is higher than about 11, the tendency of the coke to form breeze during handling and in the blast furnace is relatively so high that hardly any notice need be taken of its M'40 value, even if this is comparatively high. For this reason, therefore, much more weight should normally be assigned to Ml0m indices in the quality evaluation of cokes made from South African coking coals and blends, as the coals

By R. H. Grau, S. B. Robertson, R. B. Krog, T. P. Mucho

Stone mines represent the highest percentage of operating underground nonmetal mines in the United States. With the introduction of stricter Diesel Particular Matter (DPM) regulations, large-opening underground room-and-pillar stone mine operators are faced with implementing one or more alternative methods to reduce concentration levels including: install and maintain an effective ventilation system; install lower DPM emission diesel engines; the adding of DPM filters. The National Institute for Occupational Safety and Health (NIOSH) is actively researching various aspects of fan and stopping designs in an effort to improve the ventilation of these mines. NIOSH has identified and further developed three alternative ventilation methods for use in large-opening underground mines: split-mine, perimeter, and unit-ventilation. This paper describes the implementation of the three methods with respect to their use in mines of various levels of development.

By R. A. Franks, J. C. Edwards, L. Yuan

To determine the critical air velocity for preventing smoke rollback, diesel-fuel fire experiments were conducted in the National Institute for Occupational Safety and Health’s (NIOSH) Pittsburgh Research Laboratory’s Safety Research Coal Mine. Such information is necessary for preplanning and implementation of ventilation changes during mine fire fighting and rescue operations. The fire intensity varied from 50 to 300 kW depending on the fuel tray area. Airflow in the 2-m- (6.6-ft-) high, 2.9-m- (9.5-ft-) wide coal mine entry was regulated during the course of each experiment. The airflow was measured upwind from the fire as an average over the entry cross-section with an ultrasonic airflow sensor and was recorded dynamically with a mine monitoring system. The extent of smoke reversal was monitored with light-obscuration monitors, ionization smoke sensors and visual observations. Experimental results for the critical air velocity for smoke reversal as a function of fire intensity compared very well with model predictions based upon a computational fluid dynamics fire dynamics simulator.