Annals of Occupational Hygiene Advance Access originally published online on October 16, 2008
Annals of Occupational Hygiene 2008 52(8):717-725; doi:10.1093/annhyg/men062
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Determining the Dustiness of Powders—A Comparison of three Measuring Devices
Department of Safety Engineering/Environmental Protection, University of Wuppertal, Rainer-Gruenter-Str., Building FF 42119, Wuppertal
* Author to whom correspondence should be addressed. Tel: +49-202-4393949; fax: +49-202-4393957; e-mail: sbach{at}uni-wuppertal.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSION AND CONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The dustiness of 12 test powders was determined using three different measuring methods. One of the methods, the continuous drop method, is a reference test method according to the EN 15051 ‘Workplace atmospheres—Measurement of the dustiness of bulk materials—Requirements and reference test methods’. A test of equivalence between the reference test method and the other two methods, the modified Heubach Dustmeter, a rotating drum method and the Palas Dustview, a single-drop method, has been carried out as provided in Annex D of the European standard. No equivalence was found between any of the test methods. An applied best-case scenario yielded a slightly better outcome, but the results lead to the conclusion that it is impossible to generate viable values using the test of equivalence provided in the standard. This outcome was expected and is due to the different handling procedures applied—which, however, relates to the reality of the variety of material-handling procedures in the workplace.
Keywords: continuous drop method • dust • dustiness classification • EN 15051 • rotating drum method • single-drop method
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSION AND CONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Plinke et al. (1992) defined dustiness as ‘the propensity of a material to generate airborne dust during its handling’. Dustiness can be tested by various techniques using mechanical energy in different ways to stress test material for a defined period of time. The material is dispersed, and the amount of dust released into the air is analysed either gravimetrically or optically, depending on the measuring principle. Even if it is not always possible to relate the test results to the actual worker exposure, dustiness measurement still is an adequate tool for the exposure risk management (Lidén, 2006). Lidén also states that the ranking of materials depends on the test method and may vary strongly. However, as the different methods represent different workplaces, we suggest that no one single test should be standardized.
In April 2006, EN 15051 ‘Workplace atmospheres—Measurement of the dustiness of bulk materials—Requirements and reference test methods’ was released by the European Committee for Standardization (CEN). It defines two methods with different ways of transferring mechanical energy to the material as ‘reference test methods’- one being a rotating drum method and one a continuous drop method in a back flow (CDD). Both are capable of dividing the dust-laden air into size fractions according to the EN 481 (1993). The rotating drum separates the dispersed material into the inhalable (I), thoracle (T) and respirable (R) fraction, the CDD into the inhalable and respirable fraction.
The objective of this project was to apply a test of equivalence described in EN 15051 (2006) between the alternative test method ‘Heubach Dustmeter’ (a rotating drum method) and the reference test method B of the standard. To do so, it was necessary to use a modified version of the Dustmeter as described in Methods, to be able to sample the different size fractions of particles. The European standard gives results for seven reference bulk materials. Unfortunately, these reference materials are not affordable in the minor amounts needed for research applications. Therefore, for this project, 12 similar test materials were chosen, able to cover a wide range of characteristics of bulk materials used in practice.
With the simultaneous application of a third measuring device—the Palas Dustview as a single-drop apparatus—it was possible to compare the results of three instruments, all using different ways of stressing the material and so representing different processes and handlings of materials applied in practice. The Palas DustView however, shows some major differences from the other two methods; it uses light attenuation rather than gravimetric analysis and it does not pre-separate the dust before the measurement. This of course makes the comparison difficult. On the other hand, the authors thought it would be of a scientific interest to see how the evaluation according to the EN 15051 works for this different method which has been widely accepted and used to characterize workplace conditions.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSION AND CONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Test methods
Continuous drop test apparatus CDD, system IGF.
This apparatus is described in the EN 15051 as reference test method B.
Figure 1 shows the constructive assembly of the continuous drop method used to characterize the powders. From the sample tank (1), the bulk material is being transported via the metering device (2) through the drop pipe (3) into the flow of the back flow pipe (6). The drop mass flow can be varied by changing the speed of the metering device and is to be set by weighing to a value between 6 and 10 g min–1 (EN 15051, 2006). The constant back flow has a velocity of 0.05 m s–1 and is drawn in via drill holes in the detachable collector tank (7). Partial flow pumps [(4) and (9), 
= 2 l min–1] separate the inhalable and respirable aerosol fractions simultaneously to adequate filters mounted in the sampling heads (5) and (8) (designed by the BGIA—Berufsgenossenschaftliches Institut für Arbeitsschutz). Each individual measurement required 20 min and according to EN 15051 (2006), five individual measurements per test powder have been carried out. The amount of sample material needed for a measurement is calculated from the drop mass flow and the measuring duration plus an allowance for adequate filling of the metering device. For the tests in this project, it has been set to 180 g, given a drop mass flow of 8 g min–1.
|
The pumps for the BGIA sampling heads have been calibrated to an exact flow rate of 2 l min–1 before each measurement.
Heubach Dustmeter.
The test equipment consists of a drum with a diameter of 140 mm, a pre-separator, the sampling section and an integrated pump. By rotating, the drum continuously moves the sample material. A variably adjustable horizontal airflow conducts the dust released from the dropping material through the pre-separator to the size-selective sampling section. Possible caking inside the drum is avoided by three small hammers (tappers) clapping alternately on the outside wall while the drum rotates.
At the end of the pre-separator, sampling devices for the respirable and inhalable fraction like in the continuous drop test apparatus—but designed for a flow rate of 10 l min–1 instead of 2 l min–1—were mounted (see Figs 2 and 3). Modifying the Heubach Dustmeter in this manner makes it possible to analyse the dustiness size selectively and compare the results directly with those from the continuous drop apparatus.
|
|
The sampling devices GSP 10 and FSP 10 (also designed by the BGIA) are designed for a flow rate of 10 l min–1. All other parameters were chosen following the DIN 55 992 (2006): A weighted sample of 50 g was given into the drum which rotated with 30 revolutions per minute. The flow rate of 10 l min–1 was applied for 300 s. Apart from the sample weight, these parameters were set at the internal computer of the Dustmeter. Five individual measurements were carried out per sampling material.
The sampling devices of both the Dustmeter and the continuous drop apparatus were equipped with 37-mm glass fibre round filters with a pore width of 8 µm. Concerning the Heubach Dustmeter, it was necessary to use 100 mm filters for some of the sampling materials, to make sure the designated flow rate could be maintained. Those filters were fixed externally to the sampling devices, replacing the 37 mm filters. The hose connections were as short as possible, minimizing the dust loss. For both type of filters, the collection efficiency per mass is assumed to be 100%. The output weight of the filters after each measurement minus the initial filter weight gives the dust mass resulting from the measuring process.
Palas Dustview.
For this single-drop method, a sample of 30 g is given into a feed funnel. At the beginning of the measurement, the funnel valve opens automatically and the sample falls into the dust reservoir. Measurement is effected by extinction of a laser beam: the dust formation causes a reduction of the laser beam intensity. During the measuring time, this extinction is observed, detected on a scale from 0 to 100% and saved as a series of dust values. With this measuring principle, the refractive index of the particles has some influence on the extinction and therefore on the outcome of the measurements. If this only affects the absolute values or also the ranking of different materials has not been examined yet but should be subject of further research.
From the start value of the first 0.5 s and the extinction value at 30 s, the dustiness number is calculated. For the present project, those two values were also used as reference values for the inhalable (0.5 s) and respirable (30 s) dust fraction. This delineation is an assumption and based on research experience, it cannot be scientifically justified until now. The start value is the maximum value displayed by the Palas Dustview software and thus seems to represent the total suspended and inhalable dust fraction in the measurement chamber. After 30 s most of the material has sedimented to the ground. This point of time was chosen to represent the respirable aerosol fraction. For a particle with an aerodynamic diameter of 4 µm, which corresponds to the respirable aerosol fraction according to the EN 481 (1993), the settling velocity in the Stokes regime is 5x10–4 ms–1, which again means 16 mm in 30 s. Assuming a relatively consistent fall of all particle sizes through the tube, and taking in account that the optical system of the apparatus is installed in the upper third of the 22-cm high dust chamber, the 30 s seem to be a suitable point of time to characterize this smaller size fraction of particles being airborne because of turbulences and rebounds. Five individual measurements were carried out with each tested material. The measured dustiness with a single-drop unit is usually below these assessed by other methods (Hamelmann and Schmidt, 2003). Figure 4 shows a schematic diagram of the device.
|
The measurements of each test material with the three apparatus took place at the same time under identical circumstances to eliminate errors ascribed to different sampling conditions. The relative humidity in the laboratory averaged 46% in a range of 40 and 48%. Temperature averaged 25.5°C (23.2 to 27). The moisture content of all bulk materials was determined and documented according to the procedure given in Annex B of the EN 15051, the bulk density according to Annex C.
Test powders.
Table 1 shows a list of the 12 test materials chosen for the measurements including the moisture contents, bulk densities and distinctive features observed during the measurements.
|
Dustiness mass fractions and dustiness characteristics
For the continuous drop apparatus, the inhalable and respirable dustiness mass fractions
and 
are calculated with formulas (1) and (2).
 | (1) |
 | (2) |
and are the inhalable and respirable dustiness mass fractions in milligrams per kilograms; 
and 
are the masses of the dust collected by the sampling device for the inhalable and respirable dust, in milligrams; 
is the drop mass in the collector tank, in kilograms; 
is the total flow rate, in litres per minute; 
and 
are the flow rates of the sampling device for the inhalable and respirable dust, in litres per minute.
The calculation of the dustiness mass fractions gained by the measurements with the Heubach Dustmeter is shown in formulas (3) and (4).
 | (3) |
 | (4) |
and 
are the inhalable and respirable dustiness mass fractions in milligrams per kilograms; and are the masses of the dust collected by the sampling device for the inhalable and respirable dust, in milligrams; 
is the mass of the test sample given into the drum, in kilograms. The Palas Dustview displays the dustiness characteristic of a test material. To compare these results with the inhalable and respirable dustiness mass fractions of the continuous drop apparatus and the rotating drum method, the 0.5-s value (for the inhalable fraction) and the value after 30 s (for the respirable fraction) have been used as reference values. Table 2 summarizes the results of the measurements of all three measuring devices.
|
Correction function and test of equivalence
Evaluation procedure according to the EN 15051.
To carry out the test of equivalence for two measuring systems, the results for the dustiness mass fractions (as listed in Table 2) are used to estimate a correction function that determines the functional relationship between the outcomes of the two methods. To do so, the following parameters have to be defined:
, mean value of the dustiness mass fractions and dustiness characteristics of the inhalable or respirable dust fractions of the ith test material as listed in Table 2 (X = alternative test method, Y = reference test method).
, natural logarithms of the mean values.
The pairs ln Xi and ln Yi have to be plotted graphically. The functional relationship between the inhalable or respirable dustiness mass fractions determined with the reference test method and those determined with the alternative test method is estimated from the logarithmic data using a recognized statistical method. For this project, the curvilinear regression by the least-squares method has been applied.
, by means of the correction function transformed natural logarithm of the dustiness mass fraction and dustiness characteristic of the alternative test method (ith test material).
, transformed value of the dustiness mass fraction and dustiness characteristic of the alternative test method (ith test material).
, ratio of the transformed value and the dustiness mass fraction of the reference test method (ith test material).
The n calculated Ri then have to be evaluated statistically.
natural logarithm of the geometric mean of the n (number of evaluated test materials) ratios Ri.
, natural logarithm of the geometric standard deviation of the n ratios Ri.
, actual value of the geometric standard deviation.
Concerning EN 15051 (2006), the alternative test method can be considered as equivalent to the reference test method for the evaluated types of dust and over the range of measured Yi values, if GSDR 
1.5. According to the standards committee, this value has been arbitrarily chosen (Dirk Dahmann, personal communication).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSION AND CONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The test of equivalence has been carried out between the continuous drop apparatus CDD, system IGF and the Heubach rotating drum method (Test I, Figs 5 and 6) and the Palas single-drop apparatus (Test II, Figs 7 and 8). In these figures, the natural logarithms of the mean values of the five measurements for each tested material of the reference test method are plotted against those of the alternative test method. On the right, the correction functions for both the inhalable and the respirable particle fraction are given including the geometric standard deviation of the Ri, GSDR, which is the value to be compared according to EN 15051 (2006). None of the tests carried out with all test materials lead to the assumption of an equivalence between the measurement techniques.
|
|
|
|
It was possible to get closer to the value of GSDR
1.5 by excluding some test powders, although showing a relatively high deviation from the estimated correction function curve. Therefore, because EN 15051 (2006) provides only seven reference materials, both tests for this project have also been carried out with seven values, estimating a new correction function. For this ‘best-case scenario’, the results obtained for the GSDR are as expected consistently lower. For both the Heubach Dustmeter and the Palas DustView, different materials have been chosen to be excluded, namely those showing the highest deviation from the calculated correction function in each case. For the purpose of comparison, a test of equivalence has also been carried out between the two alternative test methods (Test III, Fig. 9) and the two reference test methods according to the EN 15051 (Test IV, Fig. 10), even though conclusions made from the results of Test III cannot be regarded as significant in terms of the European standard. The variation of the measurement values of each individual experiment (five measurements per test material and device) did not exceed 10%, and error bars were not plotted.
|
|
Dustiness classification.
To evaluate the dustiness, the measured values have to be compared to referring dustiness classifications. The dustiness ranges from very low over low and moderate to high. For the reference test methods A and B according to EN 15051 (2006), dustiness classifications are given. For the alternative test methods, these class limits have to be transformed by means of the evaluated correction functions. Table 3 shows these class limits for all methods at hand. Calculating the class limits for the single-drop method resulted in values higher than 100, which are not measurable with the Palas apparatus. Therefore, the classification ‘high’ is not available for the inhalable dust fraction. For the sake of completeness, the classification for reference test method A is indicated as well. Tables 4 and 5 give the evaluations of the dustiness for the test materials used in this project, Table 4 for the inhalable dust fraction and Table 5 for the respirable. Also declared is the compliance of the classifications of each alternative test method with those of the reference test method on hand, dependent on the number of test materials used to estimate the correction function.
|
|
|
|
DISCUSSION AND CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSION AND CONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
What is noticeable about most tests made is the fact that there is a higher equivalence between the results for the respirable particle fraction than for the inhalable. Also interesting is the fact that there seems to be no coherence between better or worse handling of the powders (e.g. flow characteristics, clogging effects or the formation of dust inside the sample tank of the IGF apparatus, which especially had to be supervised carefully and sometimes adjusted manually to assure a constant mass flow) and the deviation from the correction function, as revealed by the direct comparison the locations of the materials with same characteristics in the diagrams in Figures 5 and 7. If for example the clogging of material inside the sample tank and the metering device of the IGF CDD would have had a direct effect on the results, we would have been able to see that effect for all clogging materials. The fact that no such effect can be noticed shows that the applied test methods seem to be able to deal with a wide range of particle characteristics without major consequences for the results.
Nevertheless, none of the tests lead to a sufficiently low value of the GSDR, except for the test where the Palas Dustview is compared to the IGF continuous drop apparatus with only seven powders (Fig 8). However, given the fact that the bulk materials to be tested for their dustiness can be exposed to all sorts of stresses in practice, no one single test principle/apparatus can represent the broad range of handling scenarios at workplaces.
The tests of equivalence carried out lead to the assumption that neither of the techniques are comparable, not even those two showing the greatest analogies in handling the test material. Only by excluding four outlying materials does the comparison lead to comparable results, 1.5 for the respirable and 1.3 for the inhalable particle fraction. Of course, such exclusion of data is scientifically disputable, even if no dustiness classification has been omitted and if the same range of dust characteristics has been plotted. In any case, it was possible to show that the European standard generally does not allow for testing devices other than the two reference test methods to be used to characterize the dustiness of powders and that even under the best circumstances it is very difficult to declare two test methods as equivalent.
At first sight now it seems to be suitable to ease the restrictions and increase the allowable GSDR of an equivalence test. For example, instead of 1.5 another arbitrarily chosen value of 2 could be used. Thereby, all tests showing a high correlation coefficient R would also result in an acceptable GSDR. But this enforcement of an equivalence of course would not erase the actual problem as mentioned above, that EN 15051 (2006) does not allow for the various sorts of handling procedures, powders are exposed to at working places and therefore for a reasonable application in practice. Another approach which would lead to a more practicable interpretation of EN 15051 (2006) would thus be to either include more measuring methods as references or to allow a variety of techniques and only propose standards for measuring circumstances and the evaluation of the results. It would be helpful to facilitate the acquisition of the test materials used in the standard, particularly because for this kind of test of equivalence always a small amount of material will be needed.
Figure 10 shows the test of equivalence between the two reference test methods in the EN 15051 (2006). Table 6 presents the corresponding compliances of the classifications. These results - no equivalence and a very high GSDR, especially for the inhalable fraction - should be taken as background information when viewing the experiments with the alternative test methods even if they were obtained with different test materials, namely those provided in the European standard. It shows again the importance of the possibility of choosing between the various test methods so as to represent the process at hand in the best way.
|
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSION AND CONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The authors would like to thank the Institut für Gefahrstoff-Forschung in Bochum for providing the continuous drop apparatus according to the EN 15051 (2006).
Received May 14, 2008; in final form September 4, 2008
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSION AND CONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
DIN 55 992. Bestimmung einer Maßzahl für die Staubentwicklung von Pigmenten und Füllstoffen, Part 1 (2006) Rotationsverfahren. (1996, new version 2006). Berlin: Beauth Verlag.
EN 481. Workplace atmospheres—size fraction definitions for measurement of airborne particles (1993) Berlin: Beauth Verlag.
EN 15051. Workplace atmospheres—measurement of the dustiness of bulk materials—requirements and reference test methods (2006) Berlin: Beauth Verlag.
Hamelmann F, Schmidt E. Methods of dustiness estimation of industrial powders - a review. KONA Powder and Particle (2003) 21:7–17.
Hamelmann F, Schmidt E. Vorhersage der arbeitsschutzrelevanten Staubfraktionen von Pulvern durch Adaptierung bewährter Probennahmesysteme an ein Rotationsmessgerät. Gefahrstoffe—Reinhaltung der Luft (2004) 64:35–40.
Institut für Gefahrstoff-Forschung (IGF). "Staubungsapparatur CDD. Bochum: System IGF Betriebs–und Wartungsanleitung", manual (2006).
Lidén G. Dustiness testing of materials handled at workplaces. Ann Occup Hyg (2006) 50:437–439.
Plinke MAE, Maus R, Leith D. Experimental examination of factors that affect dust generation by using Heubach and MRI testers. Am Ind Hyg Assoc J (1992) 53:325–30.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||