Document K9qjQe5GYvjLN6DZ5j8jVomw

Monitoring of low, medium and high respirable dust concentrations using South African cyclones B.K. Belle, CSIR-Miningtek, Johannesburg, South Africa, H.R. Phillips Head, School of Mining Engineering, University of Witwatersrand, South Africa Introduction With the advent of newly available medical surveillance results on pneumoconiosis, worker exposure limits are being revised in many countries and new, more stringent dust stan dards are being recommended across the globe. It is, there fore, increasingly necessary to measure dust concentrations as accurately and in as practicable a way as possible to assess the dust exposure of workers. Trials involving underground dust measurements have been carried out using South African sam plers positioned side-by-side on the continuous miner opera tor's cabin, in machine face areas and at intake and return locations. The performance of the South African cyclones is similar to that of the European Higgins-Dewell type cyclones at a flow rate of 2,2 LImin and they respond according to the CEN/ISO/ACGIH respirable curve. During the trials dust sam ples were collected for full production shifts, excluding travel ling time. It was found that dust concentrations in the sample population ranged from 0,18 to 17,78 mg/m3. The ratio of larger to smaller dust concentration values for pairs of samples at each location was determined. The correlation coefficient between the pairs of samples for the intake, operator's cabin, face area and return positions were 0,98, 0,95, 0,88 and 0,95 respectively. The variations between the pair-wise samples were at their highest when sam pled closest to the face area. This could be due to the various patterns of dust cloud around the face area and turbulent air flow patterns around the front portion of the continuous miner. The results also indicated that the use of a single dust mea surement at high dust concentrations by itself sometimes underestimates the "true" dust concentration, depending on the position of the sampler. This paper recommends the cau tious use and interpretation of single engineering dust concen tration sample data collected for administrative dust control purposes. Introduction With the advent of new medical surveillance results on pneumoconiosis, worker exposure lim its are being revised in many countries and more stringent dust standards are being recommend ed across the globe. It is therefore, increasingly necessary to measure the dust concentrations as accurately and practicably as possible to assess the dust exposure of workers. Currently, in South African mines, workers' exposure to dust is assessed by a gravimetric sampling technique using samplers such as the Higgins-Dewell type cyclone, Casella 10 mm cyclone, Gillian cyclone, MSA cyclone, and CIP10. All these units are designed for monitoring dust and are approved by the South African Government's Department of Minerals and Energy Affairs (DME). Samplers are operated at a conventional flow rate of 1,9 L/min, except for the CIP-10, where the flow rate is 10,0 L/min in agreement with the BMRC res pirable convention (BMRC, 1952). However, Kenny, Baldwin and Maynand (1998) have sug gested that SA samplers operated at 2,2 LVmin would lead to better agreement with the new ISO/CEN/ACGIH respirable dust curve, with a 50 per cent cut-off point (d50) of 4 pm. The new flow rate confers an immediate advantage in sensitivi ty since, presently, cyclones in South Africa sam ple 16 % less air per minute. Mines are obliged to submit "engineering samples" to the DME, where gravimetric samples are collected at pre-determined continuous mining machine operator posi tions. Apart from this, the DME separately collects bi-annual "personal samples" from each section of the South African coal mines for assessing per sonal exposure. Previous research work The measurement uncertainties while estimat ing true worker exposure are not new to the min ing industry. The National Research Council in the USA (1980) concluded that uncertainties associ ated with spatial and temporal variation in dust estimates from machine mounted samplers pre cluded this method for estimating personal expo sures. During the past three decades, researchers worldwide have been presenting and publishing research work in the area of dust mea surement techniques and problems associated Journal of the Mine Ventilation Society of South Africa, April/June 2002 49 with its interpretation. Extensive work has been done to identify the factors associated with the variability of measured dust concentrations and the possible reasons identified for the variability. A Government Accounting Office report to the US Congress (1975) indicated that under certain con ditions the error associated with the respirable mine dust samples could be as great as 50%. The US National Bureau of Standards (1975) submit ted an investigation report to the senate commit tee studying respirable mine dust sampling and analysis and concluded that under tightly con trolled conditions with a "well-trained" technician, the average standard deviation associated with the process was 0,39 mg/m3 (at the 2,0 mg/m3 standard). NIOSH (1976) found that in high-risk mine sections, which failed to comply with the 2,0 mg/m3 standard, the coefficient of variation in dust measurements was 91,6 %. A study by Page and Jankowski (1984) attributed the sampling mea surement variations to the differences in the aerosol cloud being sampled, air flow velocity at the face and cyclone orientation, in 1986, a study by Kissell, Ruggier and Jankowski reviewed sev eral factors contributing to the measured dust concentration variability. They concluded that sampler position, geological variation in the com position of coal (for variability in measured free sil ica), production factors such as deep or continu ous cutting and failure to control known sources, such as shuttle car loading, play a crucial role in dust concentration sample results. Another study on dust concentration levels and their variability indicated that in coalmines, sampler location made an important contribution to the variability (Hall, Corn and Zeger, 1997). and DME-approved 10 mm cyclone (GME-G05) was used and is shown (operator's position) in Figure 1. The cyclone body is fabricated from plastic. The dust cup (grit pot) is also fabricated from plastic and is fitted to the cyclone body. Filtration is accomplished with a 37-mm filter hold er, which is pressed on over an "O" ring seal. Measurement of the size-selection characteristics of this cyclone, as tested in the UK (Kenny, Baldwin and Maynand, 1998), confirmed that they are similar to those of the Higgins-Dewell (HD) designs commonly used in the UK and elsewhere in Europe. The dust-monitoring set-up contained two HD type samplers, positioned side by side. The HD type samplers consist of an air pump, which draws 2,2 IVmin of air through a cyclone, which in turn selectively collects the fraction of airborne respirable dust with particles less than 10 pm on a pre-weighed filter disc. Filters from the samplers were weighed on an analytical electronic balance capable of an accuracy of 0,0001 mg. The proce dure for determining the particulate mass was fol lowed as per the DME guidelines (DME, 1997). Well-maintained pumps were used to avoid the effect of pump pulsations and fluctuations in the flow rate. Vy DATA COLLECTION Dust Measurement South African studies on systematic compari son of cyclones are limited to mechanized under ground coal mining conditions. The dust samples during the study were collected replicating condi tions encountered during the actual production section under various dynamic ventilation condi tions. Dust samples were collected in a bord-andpillar continuous mining machine (continuous miner or road header) sections. The samplers were positioned in the section intake (40m outbye of the face), in the operator's cabin position, face area position (on the machine) and in the section return airway (40m outbye of the face). The objec tives of the study were to determine if the variance of samples collected at the identified sampling locations, are comparable with different sampling environments and also to investigate the effect of concentrations levels and dynamic airflow pat terns on the variance of samples collected. Test samplers For all test purposes, the locally manufactured Figure 1. Dust sampling train at the CM operator posi tion. Data analysis Paired samples were defined for this study as two HD type cyclones collecting the dust samples at the specified location for the same sampling period and positioned approximately 250mm apart. For the analysis, the ratios (larger to small er) of airborne respirable dust (ARD) concentra tions were used. A rejection protocol was applied to the sample data to maintain consistency. The criteria used to reject sample data pairs was fail ure of an individual pump in the sample pair dur ing the sampling period. Rejection due to uncon trolled activity was a subjective call invoked on occasions such as when stone dusting took place in the section. 50 Journal of the Mine Ventilation Society of South Africa, April/June 2002 Results and Discussions Summary of paired sample results The statistical evaluation was carried out on the sample data collected from HD type sampler pairs at four positions. Descriptive statistics were calculated for the sampler concentration ratios including the mean, minimum, maximum, vari ance, median, standard deviation and standard error of the mean. Table 1 shows a summary statistic of respirable dust concentration values obtained by the HD samplers at different posi tions. The correlation coefficient (r) between all the HD type samplers is 0,95. The correlation coefficients (r) at the intake, operator's cabin, face area and return position are 0,98, 0,95, 0,88, and 0,95 respectively. Statistic Mean Variance Median Std. Dev. Std. Error Minimum Maximum Size HDin 1,116 0,016 1,081 0,124 0,025 1,000 1,451 25 Hdop 1,151 0,011 1,133 0,106 0,019 1,002 1,408 30 Hdface 1,253 0,065 1,143 0,255 0,051 1,006 2,181 25 HDret 1,121 0,019 1,058 0,139 0,026 1,004 1,647 29 Figure 4. ARD concentration data at the face area. The summary statistics of dust concentration ratios are briefly summarized as follows: 3.1.1 Section intake The ARD ratios (larger to smaller) values exceeded 1,1 in 40 % of the paired samples and 1,2 in 20 % of the pairs. The intake sample data had a standard devia tion of 12,5 % with a maximum ARD ratio of 1,45. Table 1. Summary statistics of dust concentration ratios. Figures 2-5 display the plots of Airborne Respirable Dust (ARD) concentrations between pairs of dust samplers at various positions. Concentration measured by HD sampler in mg/m3 Figure 2. ARD concentration data at the section intake. Concentration measured by HD sampler in mg/m3 Figure 5. ARD concentration data at the section return. CM operator position The ARD ratios (larger to smaller) values exceeded 1,13 in half of the paired samples and 1,2 in 25 % of the pairs. The standard deviation at the operator sample was 10,5 % with maximum ARD ratio of 1,407. In addition, Figure 6 shows the cumulative fre quency distribution of ARD ratios (larger to small er) at various positions. Figure 3. Dust sampling train at the CM operator's cabin. Journal of the Mine Ventilation Society of South Africa, April/June 2002 --------------- u------------ ,-------------1-------------1 0 0.5 1 1.5 2 2.5 Ratio of paired samples (large/small) Figure 6. Cumulative frequency distribution ofARD ratios. 51 Face area The ARD ratios (larger to smaller) values exceeded 1,10 in 60 % of the paired samples and 1,4 in 20 % of the pairs. The variability was highest of all the sample positions (25 %) with a maximum ARD ratio of 2,181. Section return The ARD ratios (larger to smaller) values exceeded 1,06 in 50 % of the paired samples and 1,2 in 19 % of the pairs. The standard deviation in the return air samples was 13,9 % with a maximum ARD ratio of 1,647. From Table 2, it is observed that, at all sam pling positions, the measured mean dust con centration level from each pair of HD samplers are not significantly different and the null hypoth esis is accepted. A paired t-test was performed on the combined data of all four sampling positions (Figure 7) to determine if there was a statistical difference in the results obtained between two samplers at all loca tions. Statistical Analyses Preliminary data analysis indicated that loge -transformed data was an improved fit of the nor mal distribution. Therefore, for the statistical anal ysis, Loge(Ha) and Loge(Hb) at each sampling position were compared (paired t-test). The sub scripts, Ha and Hb are the sample concentration values in the sample pair (random) at various positions. Hypothesis tests were carried out at each of the sampling locations as the sampling environment varies in terms of air flow pattern, degree of turbulence, dust concentration profiles and fixed-point (intake position and return posi tion) or moving (operator cabin and face area) type of sampling. The null and alternative hypoth esis for the tested sample pairs were: H0: jumdiff = 0 Ha:jumdiff *0 In the paired t-test, hypothesis H0 states that the mean difference in concentration values (transformed values) between side-by-side sam ple pairs is equal to zero. On the other hand, the alternative hypothesis states that the two sam plers positioned side-by-side in fact measured dif ferent mean concentration levels or the difference was not equal to zero. For this research work, a standard 95% confidence level was chosen. Results of the paired t-test statistical analyses are given in Table 2. Statistic Sample pair concentration difference Hdin HDop Hdface HDret Hda]| 95 % LCL 95 % UCL Size f-Statistic p-value Hypothesis (accept/reject) -0,09 0,03 25 -1,01 0.32 accept -0,06 -0,17 0,06 0,05 30 25 -0,07 -1,15 0.94 0.26 Accept accept -0,07 0,05 29 -0,22 0.83 accept -0,06 0,01 109 -1,30 0,19 accept Table 2. Results of paired t - test (on transformed values). Concentration measured by HD sampler in mg/m3 Figure 7. Combined ARD concentration data. The result of the paired t-test was a test statis tic with 107 degrees of freedom, p = 0,197, indi cating no significant difference between mea sured mean concentration levels using two sideby-side HD samplers. Therefore, hypothesis H0 was accepted, in other words, dust readings measured by the two samplers side-by-side are not affected at the 95 % level of confidence. The data was further separated in order to determine if there was any difference between the side-by-side sampler pairs at measured concen trations greater than 5 mg/m3 for the sampling period. The correlation coefficient (r) between all the HD type sampler pairs where one of the sam ple pair concentrations was greater than 5 mg/m3 is 0,79. The comparison of the paired sample data (greater than 5 mg/m3) was a test statistic with 30 degrees of freedom, p=0,Q5. This indi cates that at higher dust concentrations (above the DME directive maximum), there appears to be a minor difference (due to low p-value of 0,05) between measured mean concentration levels using two side-by-side HD samplers. Conclusions Trials involving underground dust measure ments were carried out using South African sam plers positioned side-by-side at low, medium and high concentration positions. During the trials dust samples were collected for full production shifts, excluding traveling time. It was found that dust concentrations in the sample population ranged from 0,18 to 17,78 mg/m3. The ratio of larger to smaller dust concentration values for a pair of samples at each location was determined. The variation in the sample dust concentration 52 Journal of the Mine Ventilation Society of South Africa, April!June 2002 between the pair wise samplers is at its highest when sampled closest to the face area. This could be due to various patterns of sample cloud around the face area and operator's cabin and turbulent airflow patterns around the machine. In general, test data have indicated that there are a variety of factors that bring into question the valid ity of samples collected underground for expo sure assessment. The perception of significant difference between pair of side-by-side samplers at the operator's position (engineering sample) seems to be persistent. The bias between the sample pairs was determined using the concentration ratio (larger to smaller) data. The analysis indicate that the sampler biases the measurement by 11,6 %, 15,1 %, 25,3 % and 12 % for the dust samples collected at section intake, operator's cabin, face area position and section return respectively. The bias is at its highest for the samplers located at the face area position. The results also indicate that the use of a single dust sampler may not by itself either underestimate or overestimate the actual dust concentration for engineering control purposes. The results also indicated that the use of a single dust measurement value at very high dust concentrations (face area) sometimes underestimates the "true" dust concentration for near-accurate exposure assessment. This paper therefore, recommends the cautious use and interpretation of single engineering dust concen tration sample data collected daily for administra tive dust control purposes. However, the question of greater bias for dust sample concentration val ues greater than 5 mg/m3 is not an immediate matter of concern as it fails to meet the DME directive. Therefore, mine management should continue to channel efforts into greater dust con trol measures in the face area. Finally, at the oper ator's cabin position, use of a single dust sampler would suffice for near-accurate exposure assess ment of the face condition under correct sampling and sample handling procedures. Acknowledgements The authors would like to acknowledge the financial support of the Safety In Mines Research Advisory Committee. Gratitude is also expressed to Dr. Brian Williams, formerly of CSIR-Miningtek for the advice in data handling during initial stages of the research work. The support of CSIRMiningtek and many of its employees is also appreciated. References American Conference of Governmental Industrial Hygienists (ACGIH). 1998. Threshold Limit Values for chemical substances and phys ical agents and biological exposure indices. USA. British Medical Research Council. 1952. Recommendations of the BMRC panels relat ing to selective sampling. 1952. From the min utes of a joint meeting of panels 1,2 and 3 held on 4th March. Comite European de Normalization. 1991. Size fraction definitions for measurement of airborne particles in the workplace. Pr EN 481. DME. 1997. Guidelines for the gravimetric sam pling of airborne particulate for risk assessment in terms of the occupational diseases in mines and works act # 78 of 1978, Parent Doc. 3rd Ed. SA. Hall, T. A., Corn, M., Zeger, S., 1997: Respirable dust and free Silica variation in mine environ ments, Strategies for mine dust measurement. Kissell, F. N. Ruggier, S. K. and Jankowski, R.A. 1986. How to improve the accuracy of coal mine dust sampling. Am. Ind. Hyg. Assoc. J. 47 (10): 602-606. International Standards Organization: Air Quality, 1993. Size fraction definitions for mea surement of airborne particles. CD 7708. Kenny, L. Baldwin, RE.J. and Maynard, A. D. 1998. Respirable dust sampling at very high concentrations, HSL, UK. National Resource Council. 1980.Measurement and control of respirable dust in mines, Washington D.C. USA. National Bureau of Standards. 1975. An evalua tion of the accuracy of the coal mine dust sam pling program administrated by the Department of Interior. Final report to the senate committee on labor and public welfare, Washington, D.C. USA. NIOSH, CDC. 1976. Statistical analysis of (Mining Enforcement Safety Administration) respirable dust data, Pittsburgh. USA. Page, S. & Jankowski, R. 1984. Correlations between measurements with RAM-1 and gravi metric samplers on longwall shearer faces. Am. Ind. Hyg. Assoc. J. 45(9): 610-616. Journal of the Mine Ventilation Society of South Africa, April!June 2002 53