Document k9nb1BrwOxVGr427gnNRRwBVb

JOURNAL OF THE MINE VENTILATION SOCIETY OF SOUTH AFRICA Published monthly by the Mine Ventilation Society of South Africa VOL. 39. No. 11 NOVEMBER, 1986 PRICE R4,50 (Excl. tax) President -- DR. A.M. PATTERSON Vice Presidents -- L.J.C. PRETORIUS AND S.J. BLUHM Hon. Editor -- R. RAMSDEN Editorial Committee P. DEGLON, C.R. EAGAR, C FRITZ. R. HEMP, W. HOLDING, S.P. JANSEN VAN VUUREN, A. STOCKHUSEN Hon. Treasurer C.J. NISSEN Secretaries -- ASSOCIATED SCIENTIFIC AND TECHNICAL SOCIETIES OF SOUTH AFRICA. KELVIN HOUSE 2 HOLLARD STREET, JOHANNESBURG TELEPHONE 832-2177 P.O. BOX 61019 MARSHALLTOWN, TVL. The Society acknowledges, with gratitude, the financial assis tance received from the Department of National Education for the publication of the Journal. The opinions expressed by contributors do not necessarily represent the official view of the Society. Copyright 1986 by the Mine Ventilation Society of South Africa. All rights reserved. Suppression of Respirable Dust and Airborne Asbestos Fibres Using Sonic Atomizing Water Spray Nozzles by H.H.E. Schroder & F.M. Runggas.... CONTENTS Call for Papers...................................................................155 News from Western Branch.................................................................155 .145 25 Years Ago..................................................................... 156 SUPPRESSION OF RESPIRABLE DUST AND AIRBORNE ASBESTOS FIBRES USING SONIC ATOMIZING WATER SPRAY NOZZLES by H.H.E. Schroder and F.M. Runggas A sonic water atomizing spray nozzle, has been shown to suppress airborne respirable dust particles at efficiencies exceeding 80 per cent. The efficiency increases with an increase in intake relative humidity, an increase in water flow rate to the nozzle, the addition of a coarse counterspray to the sonic nozzle and the efficiency of a modified commercial lamella impact mist eliminator. It was found that airborne asbestos fibres at concentrations of up to 20 fibres/mf could be suppressed at an efficiency exceeding 70 per cent, using the above-mentioned configuration, and an even higher fibre suppression efficiency could be achieved when an ionic surfactant was employed. Journal of the Mine Ventilation Society ofSouth Africa, November, 1986 145 Suppression of Respirable Dust and Airborne Asbestos Fibres Using Sonic Atomizing Water Spray Nozzles INTRODUCTION The suppression of dust with water is an age-old tech nique. This may be very effective for larger airborne dust particles, but efficiency of collection of lung-damag ing particulates with aerodynamic diameters of less than 7pm by conventional water sprays may be low. Experiments and aerodynamic capture theory (1,2) have indicated that the collision of aerosol particles with water drops may be inefficient when the ratio p of the particle radius to the drop radius is small, because airf low currents created by the falling water drop may sweep the smaller dust particles out of the way of the drops. However, if the size of the water droplets is comparable with that of the dust particles, these particles will possess enough momentum to overcome the airflow streamlines, which will result in higher collision efficiencies. Suppres sion efficiencies are therefore highest for 0,2 < p < 0,8, the overall dust scavenging efficiency being increased with increase in droplet (or particle) flux, i.e. number of particles per unit volume(3). A high concentration of micron-size water droplets may be produced by the use of sonic atomizing waterspray nozzles(4). These are air driven acoustic oscillators in which a liquid, such as water, is atomized by being passed through a field of high frequency sound waves. The field is created by means of a small resonator cup situated opposite the orifice of the nozzle. A supply of compressed air expands through a convergent/divergent section into the resonator cup, from which it is reflected back to complement and amplify the primary shock Only sonic atomizing nozzles are capable of producing water droplets comparable in size with those of respirable dust particles waves. The result is an intense field of sonic energy focused between the nozzle body and the resonator cup. Water being pumped into the high energy acoustic field is sheared into micron-size droplets, and air bypassing the resonator carries the atomized droplets down-stream in a soft low velocity spray. A `well tuned' nozzle will produce a dense fine spray containing droplets less than 20pm in diameter, which is considered by meteorolog ists to constitute a dry, non-wetting fog(5). . Of a number of commercially available water-spray nozzles, tested experimentally, only sonic atomizing nozzles have been found to be capable of producing water droplets comparable in size with those of respira ble dust particles(6). All these require compressed air for the atomization of the water. They can therefore be employed underground in gold mines where compressed air is freely available. A characterization of sonically atomized water spray plumes has indicated that the air-supply pressure had a critical optimum value below and beyond which there was a significant reduction in droplet concentration as well as an increase in droplet size. This, together with a significant degree of evaporation which was linearly dependent on the ambient relative humidity, was consid ered to be the main factor causing the poor dust suppres sion efficiencies obtained with sonically water atomizing systems. The subsequent removal of the small droplets and/or agglomerates from the airsteam in which they were entrained was also rather difficult, but this could be accomplished by means of a counterspray which pro duced larger water droplets, and a lamella impact mist eliminator. Presented in this paper are the results obtained from experiments on the suppression of respirable gold mine dust and asbestos fibres in a test duct set at the deter mined optimum conditions. The use of a modified mist eliminator with the counterspray resulted in a respirable dust suppression efficiency of up to 80 per cent. Air borne asbestos fibres could be suppressed with an effi ciency of up to 94 per cent (RH 98 per cent). This could be increased to as much as 98,6 per cent by the use of an ionic surfactant in the water. EXPERIMENTS Layout of test duct Suppression efficiency tests were conducted in a modular test duct, 760 mm in diameter, 18,5 m long, and consist ing of 18 interchangeable 1 m segments as previously described (6,7). The system included an upstream dustdispersion point, air heater/humidifier, sonic nozzle, counterspray, lamella impact mist eliminator, flow straightener and axial flow fan (2m3/s capacity), situated at the downstream end of the test duct. A number of modules were fitted with water collection ports (at dis tances of 3, 6, 8 and 11m from the sonic nozzle). In the tests the air velocity was fixed at 1,2 m/s. Dust dispersion Dust collected from underground gold mine filter bags was run into a vibrating, variable-speed, screened funnel from which it was dispersed by means of compressed air (30 /min) and then introduced into the test duct (1 m downstream of the radiator) by a centrifugal fan (0,4 m3/s, with a duct intake volume of 0,14 m3) at an angle of 45 degrees countercurrent to the airflow. This geometry ensured dispersion of an adequate quantity of dust particulates into the test duct Sonic atomizing nozzles A sonic atomizing nozzle was mounted on a holder at tached to a flange so as to be in the centre of the duct, with the orifice facing downstream. Compressed air, fil tered to remove oil, water and particulates, was passed through the nozzle at a maximum pressure of 1 400 kPa and 1 500 m/min. Water, also filtered (1 pm ele ment), was supplied at a maximum pressure of 1000 kPa. Water droplet size-distribution and concentration were determined using the gelatin film impingement method and the mean number (MND), mean surface (MSD), mean volume (MVD), number median (NMD) and mass median diameters (MMD), 50th percentile of cumulative oversize of the droplets were calculated as previously de scribed^). The coarse counterspray consisted of a cluster of three 146 Journal of the Mine Ventilation Society of South Africa, November, 1986 Suppression of Respirable'Dust and Airborne Asbestos Fibres Using Sonic Atomizing Water Spray Nozzles sonic nozzles, set to produce droplets larger than 25 pm in diameter. The plotting of data points, curve-fitting and the solu tion of polynomial and other regression functions were done by using a computer. Mist eliminator A commercially available lamella impact mist eliminator was modified to improve its water/dust agglomerate removal efficiency. Sixty fluted hard polypropylene lamella plates were packed vertically at a spacing of 10 mm into a 750 mm-square stainless steel frame coupled to the duct by means of a conversion plate. The fluted angle was about 100 degrees and the downstream edge was rolled-tubular (half-circle of radius 2 mm) so as to prevent collected droplets from being re-entrained. At an air velocity of 1,2 m/s the pressure drop across the mist eliminator was 153 Pa. Sampling of airborne particulates Airborne dust was sampled isokinetically (10 to 30 min through a 6 mm diameter probe, at 2,04 hnin) by means of a gravimetric dust sampler(8), onto 25 mm cel lulose nitrate filter membranes. Numbered membrane filter discs were weighed with an accuracy of 1 pg on a micro-balance and were loaded into numbered sampling heads in a dust-free laboratory. Used filters holding dust samples were dried in a heated desiccator at 50C for 20 min to 2 hours, after which they were left to stand in an air conditioned lab oratory overnight, together with correction filters. This allowed for mass changes due to variations in humidity when the filters were reweighed after samples had been taken. The dust concentration (mg/m3) was calculated from the mass increase and the sampled volume (sam pling time x 2,04 /min). For particle-size frequency analysis, used filter mem branes were rendered transparent by the acetone-triacetin mounting procedure(9). Airborne asbestos fibres were also sampled isokinetically (1,19 m/s through a 6 mm probe) onto filter membranes (5 to 10 min for the intake air and 25 to 30 min for the return air), or by using a konimeter without a size selector(10). The membranes were rendered transparent by the acetone/triacetin procedure and fibres were counted by the AIA method(9). Water-borne particulates in the collected fall-out water were counted by the deep cell method (10). Dust suppression efficiency (E%) was calculated as being equal to: 100 (1 - ) Where: C, = Dust concentration (mg/m3) at any point i downstream of the nozzle; and C,, = Dust concentration (mg/m3) upstream of the sonic nozzle. When applied to the efficiency of the mist eliminator C0 and Cj were the number of droplets per unit of sam pling area found before and after the demister. Water loss due to evaporation was calculated by means of Barenbrug's psychrometric charts(ll), the apparent specific volume (m3/kg) being used to calculate the apparent specific humidity of the water introduced by the sonic nozzle. The surfactant dose rate was such as to obtain the stated concentration in the spray water entering the sonic atom izing spray nozzle. RESULTS AND DISCUSSION Dust suppression When the sonic atomizing nozzle was set at the optimum air pressure (470 kPa)(6), an overall dust suppression efficiency of about 74 per cent was obtained for `respira ble' dust particles (Table 1). The efficiency increased FIGURE 1 Linear regression curves for overall dust suppression efficiency (counterspray plus demister) versus ambient relative humidity, for different water flow rates (470 kPA air pressure) Relative humidity (%RH) Journal of the Mine Ventilation Society of South Africa, November, 1986 147 Suppression of Respirable Dust and Airborne Asbestos Fibres Using Sonic Atomizing Water Spray Nozzles with increase in water flow rate, being 66,7 per cent for a flow rate of 300 mf/min and reaching a maximum of 87 per cent for a flow rate of 600 m/min. The effi ciency was even higher for higher ambient relative humidities (89,1 per cent for 98 per cent RH), which was consistent with previous findings that droplet concentra tion increased with water flow rate and relative humidity for any given sonic nozzle set at optimum air press ure^). The sonic nozzle alone was not very effective in scav enging dust particles. Its average efficiency was only 35,2 per cent. Again the efficiency improved with higher water flow rates and relative humidities, but the highest flow rate (700 mf/min) did not produce the highest effi ciency. It thus appears that this flow rate was so high as to disrupt the nozzle's field and cause unsheared larger droplets to be formed, with a concomitant loss in effi ciency. Dust suppression efficiency with the nozzle alone increased with relative humidity at lower water flow rates of 300 and 400 mUmin but this was not apparent for the higher flow rates of 500 and 600 mUrn i n for which a negative, though low, slope as obtained for the linear regression lines. However, correlation coefficients for these were insignificant. This might have been be cause the higher flow rates negated the effect of the rela tive humidity. When the counterspray was operating, the dust sup pression efficiency increased to an average of 53,5 per cent. There was a tendency for higher water flow rates to give higher efficiencies but, again, there was no positive correlation with initial ambient relative humidity. The combined effect of sonic spray plus counterspray and demister resulted in an overall suppression efficiency of up to 89 per cent, with an average of 73,9 per cent (Figure 1). TABLE 1 Dust concentration (mg/m3) and dust suppression efficiency (% E) at points along the test duct for different water flow rates (mUmin) and relative humidities (% RH) obtained with a sonic water atomizing spray nozzle (at 470 kPa air pressure) WATER FLOW RATE (rnf/min) 300 Average RH (%) 34 70 90 98 400 Average 20 36 70 90 98 500 Average 25 39 90 98 600 Average 28 50 90 98 700 Average 31 90 Overall average FEED 8,58 2,51 3,84 2,97 5,34 15,22 5,03 2,41 8,91 2,02 7,43 6,12 18,80 3,15 1,95 6,09 6,32 2,73 6,57 2,94 4,39 3,67 3,86 3,75 5,68 DUST CONC. (mg/m3) AND EFFICIENCY (%) 3 m AFTER NOZZLE 3 m AFTER COUNTERSPRAY 3 m AFTER DEMISTER (mg/m3) (%E) (mg/m3) (%E) (mg/m3) (%E) 6,47 24,6 4,23 50,7 2,95 65,6 1,73 31,1 1,52 39,4 0,94 62,6 2,70 29,7 2,28 40,6 1,11 71,1 1,99 33,0 1,37 53,9 0,87 70,7 3,90 27,0 2,73 48,9 1,78 66,7 10,08 33,8 7,46 51,0 4,68 69,3 3,62 28,0 3,38 32,8 1,52 69,8 1,85 23,2 1,36 43,6 0,84 65,2 4,94 44,6 4,41 50,5 2,38 73,3 1,20 40,6 0,82 59,4 0,48 76,2 4,90 34,1 3,78 49,2 2,20 70,4 3,60 41,2 2,61 57,5 1,64 73,2 10,71 43,0 8,09 57,0 5,17 72,5 2,05 31,8 1,85 41,3 0,79 74,9 1,11 43,1 0,85 56,4 0,43 78,0 3,61 40,7 2,74 55,0 1,54 74,7 3,49 44,8 1,94 69,3 0,95 85,0 1,74 36,3 1,33 51,3 0,36 86,8 3,78 42,5 2,71 58,8 0,89 86,5 1,77 39,8 1,34 54,4 0,32 89,1 2,65 39,6 1,63 62,9 0,57 87,0 2,38 35,2 1,42 61,3 0,81 77,9 2,35 39,1 1,38 64,3 0,70 81,8 2,37 36,8 1,40 62,7 0,82 78,1 3,68 35,2 2,64 53,5 1,48 73,9 148 Journal ofthe Mine Ventilation Society of South Africa, November, 1986 Suppression of Respirable Dust and Airborne Asbestos Fibres Using Sonic Atomizing Water Spray Nozzles Coefficients of correlation were significant for the linear regression between relative humidity and dust supression efficiency. A significant correlation (r=0,765) between mean dust suppression efficiency (different per cent RH) and water flow rate was also obtained (slope: 4,17 %E/100 m/min). Dust particle size distribution A tendency for larger dust particles to be scavenged in preference was suggested by the mean and median diameters calculated for the airborne dust particles sam pled at various points along the test duct. These diameters were smallest downstream of the demister, dif fering on average by 27 per cent from those obtained for the feed dust (Table 2). A small difference (average 10 per cent) between the latter and the original dust used (bulk sample), sug gested that some agglomeration of dust particles had occurred. The difference was, however, statistically in significant (p<0,01, f-test for equality of variances and t-test). Likewise, the differences between the values be fore and after the sonic nozzle were insignificant. This suggested that the scavenging achieved with the sonic nozzle alone (35,2 per cent) occurred with equal effi ciency for all particle sizes under investigation, i.e. for respirable dust particles. Spent water collected The particle-size frequency distribution of dust in the water collected along the duct also did not differ signifi cantly from that of the airborne dust in the test duct (Table 3). Some agglomeration of dust particles was, however, evident as was revealed by the microscope. When geometric diameters of the observed agglomerates were included, mass median diameters of up to 275 pm were obtained. TABLE 2 Particle-size frequency distribution on the dust fed into the test duct and at various points along the duct SAMPLED AT Percentage particles in size range (pm) > 94 94 - 38 38 - 19 19 - 9,4 9,4 - 3,8 3,8 - 1,9 1,9 - 0,9 0,9 - 0,4 < 0,4 DUST USED (BULK SAMPLE) DUST FEED (2 m BEFORE SONIC NOZZLE) 3 m AFTER SONIC NOZZLE 0 0 0 0,40 5,14 10,28 23,72 60,46 0 0 0 0 0 8,37 14,81 28,34 48,48 0 0 0 0 0 8,27 16,53 28,36 46,84 0 3 m AFTER COUNTER SPRAY 0 0 0 0 5,75 15,76 23,15 55,34 0 2 m AFTER DEMISTER 0 0 0 0 2,66 8,97 27,74 60,63 0 MND (pm) 1,34 1.60 1.63 1,44 1,15 MSD (pm) 2,02 2,23 2,25 2,00 1,56 MVD (pm) 2,87 2,85 2,86 2,59 2,07 NMD (pm) 0,68 0,93 0,96 0,81 0,71 MMD (pm) 7,5 7,5 7,1 5,9 5,4 TABLE 3 Mean and median diameters (pm) of dust particles found in the water collected at various points along the test duct WATER COLLECTED AT MND (pm) MSD (pm) MVD (pm) NMD (pm) MMD (pm) 3 m AFTER SONIC NOZZLE 1,52 2,42 3,44 0,72 9,3 1,5 m AFTER COUNTER SPRAY (6 m) 1,47 2,25 3,52 0,82 13,1 3 m AFTER COUNTER SPRAY (8 m) 1,51 2,18 2,89 0,80 8,2 OFF DEMISTER (10 m) 2,21 3,10 3,92 1,23 11,4 Journal of the Mine Ventilation Society of South Africa, November, 1986 149 Suppression of Respirable Dust and Airborne Asbestos Fibres Using Sonic Atomizing Water Spray Nozzles The actual amount of water collected was smallest at the point 3 m downstream of the sonic nozzle being, on average, only 4,3 per cent of the total amount collected. As was to be expected, there was an increase in the per centage collected with an increase in water flow rate (r = 0,992, slope: l,25%/100m/min). The largest amount of water (53 per cent of the total) was collected at the point 1,5 m downstream of the coun terspray. A relatively small amount of water was collected off the demister (11,1 per cent of the total) and there was no significant correlation with water flow rate at the demis ter. The total amount of water collected, as a percentage of the inflow water (sonic nozzle plus counterspray) was, however, dependent on water flow rate to the sonic nozzle (r = 0,972, at constant counterspray flow rate). Thus 84,8 per cent was collected when the water flow rate was 300 m/min. This increased to 90,3 per cent when the flow rate was 700 mf/min. Further downstream (8 m from the nozzle, 3 m after the counterspray) an average of 31,4 per cent of the total inflow was collected. A plot of the percentage of water collected against dis tances at various points along the test duct, produced quadratic regression curves having significant correlation X cc FIGURE 2 >. Graphic presentation of -4-* the rapid increase in D relative humidity along E -3C the test duct, induced by the sonic atomized water spray, with concommitant > decrease in air frCtS temperature. cc FIGURE 3 Air moisture loss (or gain) within the test duct at various ambient relative humidity and temperature levels (negative figures imply condensation) 150 Journal of the Mine Ventilation Society of South Africa, November, 1986 Suppression of Respirable Dust and Airborne Asbestos Fibres Using Sonic Atomizing Water Spray Nozzles coefficients (r = 0,829 to 0,986). The optimum of these, i.e. at the point at which the theoretically largest amount of water was collected, correlated strongly (r>0,936) with the water flow rate into the sonic nozzle. Higher flow rates displaced the point of optimum collection towards the sonic nozzle. The magnitude of this displacement was, however, small, the point of maximum collection being 5,47 m from the nozzle for a water flow rate of 700 mUmin, which shifted to 5,79 m from the nozzle for a flow rate of 300 mf:/min. The magnitude of displace ment could not be correlated with the relative humidity. The amount of water collected at the point of maximum collection was about 47 per cent, of the theoretical water intake flow. The point of maximum water collection would obviously depend on the relative position of the nozzle, counterspray and demister. The above observations sug gest, however, that such a system could be housed in an underground airway only a few metres long. Deep cell dust counts, performed on the fall-out water, were found to vary with distance along the test duct, probably because of dilution effects. Thus counts were invariably high in the water collected downstream of the sonic nozzle (6,05 Mp/m) because little water was precipitated there, whereas the water introduced by way of the counterspray diluted any scavenged dust. On average the count was 4,45 Mp/mf at the point 3 m downstream of the counterspray. A slight rise in the counts for water collected off the demister (4,77 Mp/mf) suggested a higher scavenging efficiency of the demister. The dust counts also invariably decreased with increase in water flow rate. Water evaporation For any given steady state air intake the sonic wateratomizing nozzle induced a very rapid increase in rela tive humidity. Near saturation was reached within a few metres downstream of the nozzle, i.e. within a few seconds (Figure 2). A sharp reduction in air temperature occurred simultaneously. An increase in initial ambient relative humidity had little effect on the temperature at any point along the duct. A marginal drop was obtained (0,2 deg C/10% RH, r=0,563 to 0,708). The air temperature at any given point was, however, directly proportional to that of the intake air. (Figure 3). The moisture content of the air in the duct down stream of the sonic nozzle depended largely on the intake temperature, and on the relative humidity. At low relative humidity levels droplets evaporated very rapidly, resulting in an increase in moisture content of the air. At TABLE 4 Droplet removal efficiency (%E) of the modified lamella impact mist eliminator for increasing water flow rates calculated from droplet concentration upstream (before) and downstream (after) of the mist eliminator (without the counterspray in operation) WATER POSI FLOWRATE TION DROPLETS (No. per mm GELATIN SLIDE AREA) FOUND IN SIZE RANGE (pm) TOTAL (%E) (m/min) >45 45-18 18- 9 9-4,5 4,5-1,8 1,8-0,9 <0,9 200 BEFORE AFTER %E 300 BEFORE AFTER %E 400 BEFORE AFTER %E 500 BEFORE AFTER %E 600 BEFORE AFTER %E OVERALL BEFORE AVERAGES AFTER %E 0 0 -- 0 0 -- 0 0 -- 4 0 100 0 0 -- 1 0 100 3 1 66,6 11 2 81,8 6 1 83,3 17 0 100 10 2 80,0 9 1 88,9 89 38 57,3 174 42 75,9 102 42 58,8 80 14 82,5 104 18 82,7 89 31 65,2 292 92 68,5 527 169 67,9 217 91 58,1 240 47 80,4 294 50 83,0 314 90 71,3 960 155 83,9 839 220 73,8 703 112 84,1 785 92 88,3 807 94 88,4 819 135 83,5 299 66 77,9 442 101 77,1 584 64 89,0 311 45 85,5 254 38 85,0 290 63 78,3 6 1 649 0 352 100 78,6 0 1 993 0 534 -- 73,2 0 1 612 0 310 -- 80,8 0 1 437 0 198 -- 86,2 0 1 469 0 202 -- 86,2 1 1 523 0 320 100 79,0 Journal of the Mine Ventilation Society ofSouth Africa, November, 1986 151 Suppression of Respirable Dust and Airborne Asbestos Fibres Using Sonic Atomizing Water Spray Nozzles high relative humidity, however, moisture was lost from the air. This loss was greater at higher intake tempera tures (Figure 4) which suggested that condensation had occurred under these conditions, probably under the in fluence of the relatively cold water. It might be envi saged that the condensation nuclei thus formed had greatly assisted in scavenging airborne dust particles. It could therefore be expected that sonic water atomizing nozzles should have excellent dust suppression efficien cies underground in gold mines where high relative humidities occur. There was a high correlation (r=0,939 to 0,997) between the relative humidity of the intake air and the moisture loss or gain, calculated on the basis of a second-order polynomial. The resulting equations could be used to calculate the relative humidity at which either no or total evaporation of the sonically atomized water droplets would occur. The following equations for moisture loss or gain (m,, g/kg) were obtained for steady-state tem peratures of 25, 30, 35 and 40 deg C (intake): m25 = - 0,000 685 6 RH2- 0,1603 RH + 15,2 m30 = 0,002 511 RH2 - 0,5912 RH + 24,8 m35 = 0,000 206 1 RH2 - 0,359 RH + 16,5 m40 = -0,000 388 4 RH2 - 0,452 RH + 16,9 Thus the solution of the equation for m, = 0 gave the relative humidity at which no water evaporated; e.g. at 35 deg C this was 47,2 per cent. Below this RH, water droplets would therefore evaporate, whereas condensa tion of water vapour would occur at higher RH values. Alternatively, the percentage loss of water could be calculated from the quadratic equations obtained for the percentage loss, or gain versus RH, for which high coef ficients of correlation (r=0,941 to 0,997) were also obtained. The results for no or total evaporation were identical. These results indicate that in order for a sonic nozzle to be effective in suppressing dust, a certain minimum amount of spray water is required at any given humidity and temperature condition, to overcome the effects of these variables on the droplet concentration and there fore on the dust suppression efficiency. Thus more water would be required at lower relative humidities, whereas increasingly larger amounts of the spray water could be converted to droplets at RH values above the critical, and so be more effective. It could therefore be expected that under highly humid conditions, as occurs under ground in gold mines, there would be very efficient con version of spray water to droplets. This might, however, result in `fogging-up' and pose problems regarding the removal of droplets and droplet/dust agglomerates entrained in the air. Efficiency of mist elimination A study of the droplet concentration upstream (before) and downstream (after) of the modified lamella impact mist eliminator indicated that an average droplet removal efficiency, based on droplet concentration, of 79 per cent could be achieved. The efficiency increased with an increase in water flow rate (2,82 %E/100m/min; r=0,812; Table 4) to a maximum of 86,2 per cent. As was to be expected, large droplets were removed at a higher efficiency, those with a geometric diameter of more than 45 pm being totally removed. However, a small increase in mean diameters and also a higher scavenging efficiency for smaller size ranges suggested that the removal of small droplets (<4,5 pm) was more efficient than that of medium-size droplets (4,5 -- 18 pm). This may have been either because of agglomera tion of smaller droplets or simply because the concentra tion of smaller droplets was much higher, so that there was a greater chance of their colliding with other droplets(12). With the counterspray in operation still higher effi ciencies were obtained (maximum of 90,1 per cent). There was also an increase in efficiency with an increase 20 30 40 50 60 70 80 90 100 Relative humidity of intake air (%RH) FIGURE 4 Linear regression indicating an increase in fibre suppression efficiency with increase in relative humidity of intake air for a sonic water atomizing nozzle (No. 2) set at an optimum air pressure of 470 kPa and water flow rate of 600 m/min at a position 3 m downstream of the counterspray (No. 3 nozzle). 152 Journal of the Mine Ventilation Society of South Africa, November, 1986 Suppression of Respirable Dust and Airborne Asbestos Fibres Using Sonic Atomizing Water Spray Nozzles in water flow rate, (2,74 %E/100m/min; r=0,682). The higher efficiency obtained with the counterspray in oper ation might have been due to the presence of larger droplets (>45 pm), produced by the counterspray, which had not precipitated by the time the air reached the mist eliminator, but which were effectively removed by the mist eliminator. Furthermore, even though the counter spray was set to produce droplets larger than 20 pm, the concentration of smaller droplets was doubled, which also increased the frequency of collision. The advantage of using the counterspray together with the mist elimina tor was, however, demonstrated by the higher dust sup pression efficiency obtained with such a combination. In the underground mining environment the countersprfy could act also as a spray cooler, many of which are at present being employed in the gold mining industry. Suppression of airborne asbestos fibres When a sonic water atomizing jet, set at optimum con ditions, was used, airborne asbestos fibres could be sup pressed with an efficiency exceeding 70 per cent. The ef ficiency was higher at higher ambient relative humidity TABLE 5 (Figure 4) as was found with quartz dust (7). At a rela tive humidity of 98 per cent the efficiency was as high as 94,5 per cent. Fibre suppression efficiency correlated strongly (r=0,9898) with ambient relative humidity, the linear regression equation being: Efficiency (%) = % RH x 0,295 + 65,1 The maximum efficiency attainable at 100 per cent RH, calculated with this equation, is 95 per cent. Like wise the efficiency should always exceed 65 per cent, since this is the value obtained for zero RH. This, of course, applies only to the set optimum steady-state con dition used in the present investigation. Nevertheless this regression equation could be employed to predict the efficiency for any given situation. Thus a 71 per cent effi ciency could be expected for dry winter conditions on the highveld (20 per cent RH). The mean number fibre length did not correlate well (r=0,232) with the relative humidity, nor with the sup pression efficiency (r=0,342). This could have been because only a limited number (five) of size, intervals had been used for the size-frequency distribution, which ren dered the figure statistically inaccurate. Airborne asbestos fibre suppression efficiency (%) for different water flow rates (m/min) at various relative humidites (%RH) for a sonic water atomizing nozzle set at an optimum air pressure of 470 kPa at a position 3m downstream of the counterspray The nozzle water flow rate affected the fibre suppres sion efficiency significantly (Table 5) but an optimum was found (average 815 m/min) beyond which the effi ciency declined. This was previously found for quartz dust and again may have been due to the disruption of the high energy sonic force field by excessively high water flow rates. RELATIVE WATER FIBRE HUMIDITY FLOW RATE CONCENTRATION (f/mtj (%) (m/min) INTAKE AFTER SPRAYS SUPPRESSION EFFICIENCY (%) 25 300 20 500 20 600 20 700 20 8,80 6,10 5,45 4,65 56,0 69,5 72,7 76,8 The optimum water flow rate (highest attainable fibresuppression efficiency) could be calculated from the first derivative of a quadratic expression at which the flow rate correlated highly with efficiency (e.g. for RH of 25 per cent r=0,998) and: efficiency = 26,82 + 0,1175x(flow) - 6,6364x10 s (flow)2 40 300 20 500 20 600 20 700 20 6,60 4,65 4,25 4,70 67,0 76,7 78,7 76,5 50 300 15 4,98 66,8 500 15 4,08 72,8 600 18,4 4,00 78,3 700 18,4 4,05 78,0 75 300 15 4,56 69,6 500 15 4,02 73,2 600 18,4 3,20 82,6 700 18,4 3,72 79,8 90 300 15 2,76 81,6 500 15 2,04 86,4 600 18,4 1,78 90,3 700 18,4 2,24 87,8 98 300 15 1,78 88,1 500 15 1,20 92,0 600 18,4 1,02 94,5 700 18,4 1,22 93,4 An even higher fibre suppression efficiency (up to 100 per cent) could be obtained by incorporating an ionic surfactant into the spray water (Table 6). Increasing the surfactant dosing rate increased the efficiency, but a point of diminishing returns was reached, at 0,73 g/ in the sonic spray water beyond which there was little fur ther increase in efficiency. The fibre suppression effi ciency at this point was 98,6 per cent. Further dilution by the counterspray resulted in a calculated surfactant con centration of 0,275 g/ which thus appeared to be the `optimal' concentration. The 100 per cent efficiency which was recorded meant simply that no fibres could be collected within the sam pling period of 25-30 minutes. An extension of the sam pling period might well have resulted in finite quantities of fibres being collected. An extrapolation to zero surfactant dose rate indicated that water alone provided a fibre suppression efficiency of about 96 per cent under the experimental conditions. The use of a surfactant in combination with a sonic water atomizing jet appears to be beneficial only where it is imperative that very high efficiencies are required for the adequate safeguarding of the environment. Journal of the Mine Ventilation Society ofSouth Africa, November, 1986 153 Suppression of Respirable Dust and Airborne Asbestos Fibres Using Sonic Atomizing Water Spray Nozzles TABLE 6 Airborne asbestos fibre suppression efficiency (%) obtained for various dosing rates (100 m/min of 1 to 50 g/) of an ionic surfactant (RC 11a) fed into the water introduced into a sonic water atomizing nozzle set at an optimum air pressure of 470 kPa and a water flow rate of 600 mC/min, at a position 3m downstream of the counterspray, for a relative humidity of 98 %RH SURFACTANT DOSING RATE m FIBRE CONCENTRATION INTAKE (f/m) AFTER SPRAY (f/m) SUPPRESSION EFFICIENCY (%) 0,17 16,7 1,3 92,2 0,33 49 1,8 96,3 0,50 49 1,4 97,1 0,67 49 1,1 97,7 0,83 58 0,8 98,6 1,67 58 0,8 98,6 2,50 58 0,6 98,9 3,33 58 0,6 98,9 4,16 58 0,5 99,1 5,0 58 0,4 99,3 5,83 58 0,3 99,5 6,66 58 0,1 99,8 8,33 58 0 100 CONCLUSIONS The results indicate that: (i) Airborne respirable dust can be suppressed with an efficiency exceeding 80 per cent by passing the air through a sonic atomizing water spray nozzle set at optimum air pressure (470 kPa), and then through a counterspray and mist elimi nator. (ii) This combination of components was found to be necessary since the sonic nozzle alone gave an efficiency of only 35,2 per cent, whereas the addition of the counterspray increased it to 53,5 per cent. (iii) The dust suppression efficiency increased with increase in ambient relative humidity, which sug gested that sonic atomizing nozzles should be highly effective underground in South African gold mines where relative humidities are high. (iv) For excessive nozzle flow rates (>815 m/min) suppression efficiencies were lower, probably because these rates upset the high-energy sonic field. (v) Larger dust particles were scavenged at higher efficiency, as indicated by a decrease in the mean number, and the surface and volume diameters, as well as the number and mass median diameters. The differences, however, were statistically insignificant, probably because of the limited number of observations made. (vi) There was some dust agglomeration in the spray water collected along the test duct, as was evi dent from microscopic observation. (vii) A total of up to 90 per cent of the spray water was collected along the duct, the rest probably (viii) being blown out of the duct. Of the water col lected only 4,3 per cent precipitated downstream of the sonic nozzle, while the largest amount precipitated after the counterspray (53 per cent at 1,5 m downstream plus 31,4 per cent at 3 m downstream of the counterspray). Only 11,1 per cent of the total collected on the demister. The point at which the maximum collection occurred shifted marginally towards the nozzle when there was an increase in water flow rate, but remained at less than 6 m downstream of the sonic nozzle. Water droplets created by the sonic nozzle were found to evaporate almost instantly at low rela tive humidities, even at moderately low ambient air temeperatures. Second-order equations derived from the results (r => 0,94) could be used to calculate the RH for a given air tem perature at which either no or complete spraywater evaporation occurred. It was concluded that a certain minimum amount of spray water was required at any given temperature and humidity, which had first to saturate the air with water vapour, before further water could effec tively be converted to droplets, and hence to ob tain a satisfactory dust suppression efficiency. However, at higher ambient relative humidity the relatively cool spray water would aid in the condensation of water vapour to cause the for mation of condensation nuclei which would en- A certain minimum amount of spray water was required at any given temperature and humidity, which had first to saturate the air with water vapour, before further water could be converted to droplets hance dust scavenging. (ix) The modified lamella impact mist eliminator was highly effective in removing airborne micron size water droplets (>80 per cent efficiency). An effi ciency of up to 98 per cent could be calculated on the basis of relative masses (mass median diameter X particle count). (x) Airborne asbestos fibres (concentration of up to 20 fibres/mC) could be suppressed with an effi ciency exceeding 70 per cent. The efficiency was also higher at higher ambient relative humidities (94,4 per cent at RH of 98 per cent). Suppres sion efficiency increased with water flow rate but an optimum was found (815 m/min) beyond which it decreased again, probably because these high flow rates disrupted the high energy sonic field. (xi) An even higher fibre suppression efficiency (up to 100 per cent) could be obtained when an ionic 154 Journal of the Mine Ventilation Society of South Africa, November, 1986 Suppression of Respirable Dust and Airborne Asbestos Fibres Using Sonic Atomizing Water Spray Nozzles surfactant was employed. Increasing the surfac tant dosing rate increased the efficiency, but little additional benefit was obtained at dosing rates exceeding 0,73 g/ of nozzle spray water, at which the calculated efficiency was 98,6 per cent. 3RD U.S. MINE VENTILATION SYMPOSIUM, CALL FOR PAPERS The 3rd U.S. Mine Ventilation Symposium will be held at The Pennsylvania State University, University Park, Pennsylvania on October 12-14, 1987. The symposium is sponsored by the Underground Ventilation Committee of the Coal/Mining and Exploration Divisions of the Society of Mining Engineers of AIME. The advisory committee for the symposium invites authors to propose papers in the following areas: REFERENCES 1. WARRINGTON, G.B. A new technique for dust abatement in a crushing and screening plant. Paper presented at the 22nd Annual Meeting, Aggregate Producers Association of Ontario, Ottawa, March 2, 1979. 2. SCHOWENGERDT, F.D. & BROWN, J.T. Colo rado School of Mines tackles control of respirable coal dust. Coal Age, 81, April 1976, pp. 129-31. 3. LEONG, K.H., BEARD, K.V., STUKEL, J.J. & HOPKE, P.K. Factors affecting the collision of aerosol particles with small water drops. J. Aerosol Sci. Technol., 2, 1983, pp. 341-9. 4. EDITORIAL. Dust controlled at source by a sonic suppression system. Pit & Quarry, 70, October 1977, pp. 135-7. 5. HASSLER, H.E.B. A new method of dust separa tion using autogenous electrically charged fog. J. Powder Bulk Solids Technol., October 1977, pp. 10-4. 6. SCHRODER, H.H.E., RUNGGAS, F.M. & KRUSS, J.A.L. Characterization of sonically atom ized water spray plumes. Paper presented at the Third International Mine Ventilation Congress, Har rogate, 13-19 June 1984. Proceedings, Ed. Howes, M.J. & Jones, M.J., pp. 219-28. 7. RUNGGAS, F.M. & SCHRODER, H.H.E. Dust Suppression by sonically atomized water sprays. Sixth International Conference on Air Pollution, 23-25 October 1984, Pretoria, South Africa. 8. APPELMAN, G. Dust sampler. Chamber of Mines Services (Pty) Ltd., S.A. Patent No. 79/5060, filed 25th September, 1979, granted 29th October, 1980. 9. REFERENCE METHOD for the determination of airborne asbestos fibre concentrations at work places by light microscopy (membrane filter method). AIA Health and Safety Publication RTM1, 7 September 1979. 10. CHAMBER OF MINES OF SOUTH AFRICA. Measurements in Mine Environmental Control, Johannesburg, 1982, pp. 27-31. 11. BARENBRUG, A.W.T. Psychrometry and Psychrometric Charts, 3rd Edition, Chamber of Mines of South Africa, 1974, 59 p. & 31 charts. 12. SCHRODER, H.H.E. The properties and effects of dust. In: Environmental Engineering in South Afri can Mines. Ed. Burrows, J., Hemp, R., Holding, W. & Stroh, R.M. The Mine Ventilation Society of South Africa, Johannesburg, 1982, p. 322. Fans and Shafts Face Ventilation Booster Fans and Methane Control Recirculation Dust Generation, Transport Dust Characteristics and Control Mine Health Studies Diesel Ventilation Monitoring and Control Heating/Air Conditioning Ventilation System Design Network Analysis and Coal/Non-coal Case Studies Simulation Miscellaneous Studies Abstracts of 200 to 300 w irds should be submitted by December 30, 1986 to: Jan M. Mutmansky 120 Mineral Sciences Building The Pennsylvania State University University Park, PA 16802 (814) 863-1642 The symposium advisory committee will review the abstracts and authors will be informed concerning ac ceptance by February 15, 1987. An author's guide will be mailed to each successful author at that time. Completed manuscripts will be required in photo-ready form by June 8, 1987. Symposium proceedings will be available and are planned for distribution to all registrants at the symposium. WESTERN BRANCH BROSEAL VISIT An educational visit to Broseal was arranged on Friday 5 September 1986. Thirty nine delegates met at the Randfontein factory. After registration, the delegates were welcomed by Mr De Wet Mulder, the Broseal Managing Director. A brief resume of the history of Broseal was also given. Mr de Laney, the Technical Manager conducted the delegates on a tour of the factory, explaining the various processes that go into the manufacture of the galvanized iron vent ducts of which 200 of 406 mm x 36 mm can be manufactured daily. After this tour, delegates left for the Krugersdorp factory where they arrived at 10h45, and were invited to enjoy refreshments. Mr Kevan Carew conducted the delegates on a tour of the Krugersdorp factory, where special application appli ances are manufactured. These include: Special ducting, silencers, non-return flaps and special clamps. A demon stration on duct pressure testing as well as the effect of a silencer on a noisy fan was given. At 12h30, the delegates departed to the Krugersdorp Golf Club where they were treated to a very enjoyable lunch. During lunch Mr Ray Choveaux thanked Broseal on behalf of the delegates. Journal of the Mine Ventilation Society of South Africa, November, 1986 155