CEN/TC
Date: 2012
TC WI
CEN/TC
Secretariat:
Determination of Size-Weighted Potential Respirable Fraction – SWeRF
and
Size-Weighted Potential Respirable Fraction of Crystalline Silica - SWeRFcs
Einführendes Element — Haupt-Element — Ergänzendes Element
Élément introductif — Élément central — Élément complémentaire
ICS:
Descriptors:
Document type: European Standard
Document stage: PQ
Document language: E
PQdraft_SWeRF (E)October 2012_ FINAL_25102012.docx
TC WI :2012 (E)
Contents
Page
Foreword ..............................................................................................................................................................4
Introduction .........................................................................................................................................................5
1
Scope ......................................................................................................................................................6
2
Normative references ............................................................................................................................6
3
Terms and definitions ...........................................................................................................................6
4
Methods for determination of SWeRF and SWeRFcs..........................................................................6
5
5.1
5.2
5.3
Determination of SWeRF and SWeRFCS by sedimentation ...............................................................7
Determination of sedimentation time ..................................................................................................7
Selection of sedimentation liquid ........................................................................................................8
Sample preparation, sedimentation and SWeRF determination .......................................................8
6
Determination of SWeRF and SWeRFcs by calculation. ...................................................................9
7
Bibliography ........................................................................................................................................ 10
Annex
A.1
A.2
A.3
A (informative) Guidance on the use of the calculation and sedimentation methods .................. 11
Recommendations on using the calculation and/or sedimentation method................................ 11
Recommendations for using the sedimentation method ............................................................... 11
Guidelines for the Determination of Crystalline Silica.................................................................... 11
Annex B (informative) Round Robin test to establish a SWeRF Reference Sample ................................. 13
B.1
Introduction ......................................................................................................................................... 13
B.2
Test material ........................................................................................................................................ 13
B.3
Evaluation and appraisal method ..................................................................................................... 14
B.4
Results ................................................................................................................................................. 14
Annex C (informative) Derivation for calculating the sedimentation parameters for determining
by sedimentation ................................................................................................................................ 16
Annex D (informative) The Determination and Isolation of the Size Weighted Potential Respirable
Fraction (SWeRF) of Kaolins and Kaolinitic Clays by Sedimentation........................................... 20
D.1
Introduction ......................................................................................................................................... 20
D.2
Scope ................................................................................................................................................... 20
D.3
Equipment / Consumables ................................................................................................................. 20
D.4
Method ................................................................................................................................................. 21
D.5
Illustrations .......................................................................................................................................... 23
Annex E (informative) Other Minerals which may be treated in a similar way to Kaolins / Kaolinitic
Clays for SWeRF and SWeRFcs Determination .............................................................................. 25
E.1
Introduction ......................................................................................................................................... 25
E.2
Andalusite ............................................................................................................................................ 25
E.3
Mica ...................................................................................................................................................... 26
E.4
Vermiculite ........................................................................................................................................... 26
E.5
Talc ....................................................................................................................................................... 26
Annex F (informative) The Determination of the Size Weighted Potential Respirable Fraction
(SWeRF) of Diatomaceous Earth (DE) by calculation and by sedimentation ............................... 28
F.1
Introduction ......................................................................................................................................... 28
F.2
Scope ................................................................................................................................................... 28
F.3
Equipment / Consumables ................................................................................................................. 28
F.4
Method ................................................................................................................................................. 28
F.5
Determination of SWeRF by sedimentation ..................................................................................... 30
F.6
Determination of SWeRF by calculation .......................................................................................... 31
F.7
Determination of SWeRFcs ................................................................................................................ 32
2
TC WI :2012 (E)
F.8
Example ................................................................................................................................................ 32
Annex G (informative) The Determination of the Size Weighted Potential Respirable Fraction
(SWeRF) of feldspar products by sedimentation ............................................................................. 34
G.1
Subject .................................................................................................................................................. 34
G.2
Scope .................................................................................................................................................... 34
G.3
Equipment ............................................................................................................................................ 34
G.4
Method .................................................................................................................................................. 34
Annex H (informative) The relation between density and SWeRF ............................................................... 37
H.1
Subject .................................................................................................................................................. 37
H.2
Determining SWeRF from size distribution ...................................................................................... 37
H.3
Determining SWeRF by sedimentation. ............................................................................................ 39
3
TC WI :2012 (E)
Foreword
This document has been prepared by the Industrial Minerals Association, IMA-Europe.
This document is currently submitted to the Primary Questionnaire procedure.
4
TC WI :2012 (E)
Introduction
This standard describes the determination of the Size Weighted Potential Respirable Fraction (SWeRF) in
bulk materials. The method can be used for comparing the potential health hazard of different bulk samples.
This method does not predict how a material will disperse in air, but it quantifies the respirable fraction in a
sample. Also, the more dangerous particles are weighted more. The advantage of this method is that it
provides an unambiguous characterization of the bulk material.
This standard uses the term Potential Respirable Fraction to indicate that it does not analyze airborne
particles, but it evaluates the presence of particles in a bulk material that, based on their particle size, have a
potential risk of being respirable when they would become airborne.
Evaluating bulk materials using SWeRF is complementary to determining the dustiness according to EN
15051 - Measurement of dustiness of bulk materials. The difference between the 2 standards is that SWeRF
quantifies the potential respirable fraction in a bulk material (it characterizes the bulk material as such) while
dustiness quantifies the respirable fraction that comes out of the bulk material after a standardized activity (it
characterizes the material with relation to the workplace atmosphere when working with the bulk material).
This standard describes a method using sedimentation and a calculation method based on particle size
distribution. The calculation method can only be used after experiments have shown that the results are
accurate and consistent with the results from sedimentation for that particular bulk material.
The SWeRF is, in any case, a method to provide valuable information to the legitimate questions of customers
about the content of respirable sized particles in products they purchase.
5
TC WI :2012 (E)
1
Scope
This standard specifies a method for evaluating the Size Weighted Potential Respirable Fraction (SWeRF) in
terms of weight percent in bulk materials. This method expresses the relative content of particles in bulk
materials that, based on particle size, should be considered respirable when they would become airborne.
The method can also be used to evaluate the content of potential respirable crystalline silica particles in bulk
materials.
2
Normative references
EN 481:1994 Workplace atmospheres - Size fraction definitions for measurement of airborne particles
ISO 13317-1 Determination of particle size distribution by gravitational liquid sedimentation methods — Part 1:
General principles and guidelines.
ISO 13317-2 Determination of particle size distribution by gravitational liquid sedimentation methods — Part 2:
Fixed pipette method
ISO 14887 Sample preparation — Dispersing procedures for powders in liquids
ISO 24095: 2009 Workplace air: Guidance for the measurement of respirable crystalline silica.
3
Terms and definitions
In this European standard, the terms and definitions given in EN 481 apply. In addition, for the purpose of this
European standard, the following terms and definitions apply.
3.1
Potential Respirable Fraction
The Potential Respirable Fraction is that fraction of a material which contains the particles that, because of
their physical properties (e.g. size, density), can be considered as respirable particles when they are airborne.
NOTE It should be clearly understood that they are not respirable as such but they are qualified to become
respirable.
3.2
Size Weighted Potential Respirable Fraction (SWeRF)
The Size Weighted Potential Respirable Fraction (SWeRF) is that fraction of a material weighted by its
probability to reach the alveoli as given in EN 481.
3.3
SWeRFcs
Size Weighted Potential Respirable Fraction of Crystalline Silica, calculated by multiplying the SWeRF of a
sample by the crystalline silica content of the sample.
4
Methods for determination of SWeRF and SWeRFcs
SWeRF and SWeRFcs can be evaluated in two possible ways; either by sedimentation (Clause 5 of this
standard) or by calculation (Clause 6 of this standard). The calculation method can only be used after
experiments have shown that the results are accurate and consistent with the results from sedimentation for
that particular bulk material.
The guidance given in appendix A1 explains the context in which each method can be used. In appendix B,
information is given on the validation and repeatability of the method based on a Round Robin test.
The methods described are based on the following assumptions:
6
TC WI :2012 (E)

For sedimentation method
1. All liquid that is extracted from the sedimentation vessel comes only from the space above the
extraction nozzle, and not from the space below the extraction nozzle.
2. The particle size distribution of the bulk material is constant over the interesting particle size
range.
3. Stokes law is valid for both liquid and gasses (at low Reynolds numbers). The velocity of a
particle settling in a medium is limited by the drag force and this depends on the Reynolds
number for that particle. Although the density of liquids are much higher, so is their viscosity so
that in the end the difference between the Reynolds number of a particle in air and in e.g. water is
only a factor of 5. And although the constants for calculating the drag coefficient of particles
depend on the Reynolds number, the variation with Reynolds number within this range is very
small and can be neglected. This means that the Stokes diameter is the same (or at least very
close) to the aerodynamic diameter. Therefore the dynamic form factor is assumed to be equal in
both air and liquid.
4. In the material of interest all particles have the same and known density.

For calculation method
1. The dynamic form factor is neglected in the determination of the particle size distribution with
methods other than methods based on sedimentation.
2. The precondition is that all particles have the same density or in a mixture the different materials
have the same particle size distribution. If this is not the case a separation using the
sedimentation should be used.
5
5.1
Determination of SWeRF and SWeRFcs by sedimentation
Determination of sedimentation time
The time of sedimentation can be calculated using the following equation (see formula 1), which is based on
Stokes’ law and the convention described in EN 481. The derivation of it is given in annex C.
(1)
t = time (s) at which the size of the fraction of fine particle fraction separated by sedimentation approaches the
size of the respirable fraction according to EN 481
h = height (m) of the column of the supernatant liquid that is extracted after time = t.
η = dynamic viscosity of the liquid (kg/m∙s)
2
g = acceleration due to gravity (m/s )
3
ρs = density of the solid particles (kg/m )
3
ρl = density of the liquid medium (kg/m )
ρ0 = unit density (1000 kg m-3 [= 1 g cm-3])
7
TC WI :2012 (E)
For determining the SWeRF of a sample, the density of this sample should be used for ρs. When SWeRFcs is
determined, the density of crystalline silica (quartz or cristobalite) should be used.
NOTE
The height h to be decanted is usually set at 0,1 m, but other heights may be used depending on the nature of
the sample. In any case, the height h should be less than half the total height H of the sedimentation liquid in the
measuring cylinder (see step 7 in clause 5.3).
5.2
Selection of sedimentation liquid
A suitable sedimentation liquid should be selected in order to meet the following requirements:

The particles in the sample must be completely deagglomerated.

The particles in the sample should not dissolve, swell or disintegrate.

The particles in the sample should not react.
NOTE 1
Water in most cases is a suitable sedimentation liquid.
NOTE 2
When necessary, a dispersant or deflocculant additive may be used in appropriate quantities (see guidance in
annex A).
5.3
Sample preparation, sedimentation and SWeRF determination
NOTE 1
In order that the sedimentation is performed correctly, it is advised to use the Andreasen pipette
method as described in BS 1377-2.
The following steps should be executed to determine the SWeRF or SWeRFcs of a sample:
1)
Take a sample of the material of approx. 5.0 g.
2)
Determine the weight M of the sample with a precision of 0.001 g.
3)
Disperse the sample in 50 ml of sedimentation liquid in a 100 ml pre-weighed, dry and clean
beaker. The weight of the beaker should also be determined with a precision of 0.001 g.
4)
Treat the sample in an ultrasonic bath or a shaker until completely deagglomerated.
NOTE 2
The deagglomeration time will depend on the type of material, see for further information Annex
A.
5)
If necessary, add a suitable dispersant or deflocculant to keep the particles from flocculating or
coagulating. See for further information Annex A.
6)
Pour the dispersed sample in a 250 ml measuring cylinder. Rinse out the sample jar using the
sedimentation liquid to ensure that no residue remains. Fill the cylinder up to 250 ml with
sedimentation liquid. Seal the open end of the cylinder and shake the contents thoroughly. Then
replenish the cylinder up till 250 ml and homogenise.
NOTE 3
The volume of solids should be maximum 1 % of the volume of the total liquid to ensure
unhindered sedimentation of the individual particles.
NOTE 4
For some minerals it might be necessary to use a plunger agitator. With the cylinder in place,
agitate. Then remove the plunger to replenish the cylinder up till 250 ml and homogenise.
7)
8
Place the cylinder in a location where it is at constant temperature and free from effects that could
cause currents in the liquid. The constant temperature can be achieved when using a water bath. Leave
to settle for the calculated time (t). Determine the depth of the liquid column H (m).
TC WI :2012 (E)
8)
From the cylinder, after calculated time (t), lower a pipette to an insertion depth of h below the surface,
draw the volume of the supernatant (Vt) above the tip of the pipette with a siphon or pipette and transfer
into the pre-weighed beaker.
Make sure that, whilst using the pipette, the tip of the pipette remains in the same place in the cylinder;
do not insert it deeper as the level descends.
NOTE 5
Placing a clip on the pipette and attaching it to the edge of the cylinder will make it easier to keep the
pipette in the right place.
9)
Place the beaker containing the supernatant on a hot plate to evaporate the liquid and heat gently until
dry. Then place the beaker in the oven at 103 °C for one hour and transfer into a desiccator, leaving it
to cool down to ambient temperature .
10)
Re-weigh the beaker to within 0.001 gram accuracy and note the post-weight. Determine the weight of
the residue m by subtracting the weight of the pre-weighed beaker (see step 3).
NOTE 6
The above procedure should initially be repeated twice in order to check the reproducibility of the
sedimentation method for SWeRF determination.
11)
Determine the SWeRF of the sample using equation 2.
SWeRF 
H m
 100%
h M
(2)
12) The SWeRFcs of the sample may be calculated by using equation 3.
SWeRFcs 
NOTE 7
6
H m
  cs  100%
h M
(3)
For a more detailed description of this method for the particular application on Kaolins and Kaolinitic
Clays, see informative Annex D. For SWeRF and SWeRFcs determination of some other minerals
(Andalusite, Mica, Vermiculite and Talc) which may be treated in a similar way to Kaolins / Kaolinitic
Clays see informative Annex E. For the particular application on diatomaceous earth (DE) see
informative Annex F, and for the particular application on feldspar products see informative Annex G.
Determination of SWeRF and SWeRFcs by calculation.
The SWeRF of a sample is calculated by first determining its particle size distribution.
A size weighting is then applied, based on the probability function given in EN 481, i.e. the probability function
for particles reaching the alveoli when inhaled. See formula 4.
D 
SWeRF 
 f ( psd )
D
 P( D)dD
(4)
D 0
f(psd)D = particle size distribution for aerodynamic diameter D.
P(D) = probability of reaching the alveoli for particles of aerodynamic diameter D, according to EN 481
D = aerodynamic diameter = d   (SG), where SG is the specific gravity.
NOTE 1
A standardized method should be used to determine the specific gravity, e.g. by using a He or
liquid pyknometer in accordance with ISO 1183-1.
9
TC WI :2012 (E)
A more detailed explanation of the derivation of formula (4) is given in Annex C.
The SWeRFcs is calculated by multiplying the SWeRF of the sample with the crystalline silica content of the
sample, see formula 5. It is assumed that the particle size distribution of the sample and of its crystalline silica
content are the same.
SWeRFcs  SWeRF   cs
(5)
cs = mass fraction of crystalline silica in the sample
NOTE 2
7
The content of crystalline silica of the sample can be determined using techniques such as Xray Powder Difractometry (XRD) or Infra-red Spectroscopy (IR) as described in ISO 24095. The
methods are derived from the standards used to determine crystalline silica in respirable
airborne dusts, e.g. MDHS 1010.
Bibliography
ISO 1183-1:2004 Plastics - Methods for determining the density of non-cellular plastics - Part 1: Immersion
method, liquid pyknometer method and titration method
ISO 10725:2001 Acceptance sampling plans and procedures for the inspection of bulk materials
ISO 11648-1:2003 Statistical aspects of sampling from bulk materials - Part 1: General principles
ISO 11648-2:2001 Statistical aspects of sampling from bulk materials - Part 2: Sampling of particulate
materials
ISO 24095:2009 Workplace air - Guidance for the measurement of respirable crystalline silica
EN 932-1:1996 Tests for general properties of aggregates – Part 1: Methods for sampling
EN 932-2:1999 Tests for general properties of aggregates – Part 2: Methods for reducing laboratory samples
EN 15051:2006 Workplace atmospheres - Measurement of the dustiness of bulk materials - Requirements
and reference test methods
pr-EN 15051-1:2011 Workplace exposure - Measurement of dustiness of bulk materials - Part 1:
Requirements and choice of test methods
BS 1377-2:1990 Methods of test for soils for civil engineering purposes. Classification tests
MDHS 101 Crystalline silica in respirable airborne dusts - Direct-on-filter analyses by infrared spectroscopy
and X-ray diffraction (February 2005)
NIOSH 7500 SILICA, CRYSTALLINE, by XRD (filter redeposition) - Issue 4 (15 March 2003)
Validation of the Analysis of Respirable Crystalline Silica (Quartz) in Foams Used with CIP 10-R Samplers, C.
Eypert-Blaison et al., Ann. Occup. Hyg., Vol 55, N°4, pp 357-368, 2011
10
TC WI :2012 (E)
Annex A
(informative)
Guidance on the use of the calculation and sedimentation methods
A.1 Recommendations on using the calculation and/or sedimentation method
There are two ways to determine the SWeRF and SWeRFCS:
1. the sedimentation method in a liquid.
2. the calculation method;
Experiments have been done on quartz sand and flours, feldspar, limestone, leucophyllite, clay and
diatomaceous earth and it has been shown that:
-
-
For quartz flours and sand, the calculation method can be applied directly and provides an accurate
result which is consistent with the sedimentation method.
For some minerals the calculation method may provide an overestimation of the SWeRFCS, in which
case the sedimentation method will give a more accurate figure. This is the case when the PSD
(particle size distribution) of the CS (crystalline silica) in the material is significantly coarser than the
PSD of the bulk material.
For some minerals, the choice of the suitable separation medium (liquid and/or dispersant agent)
requires investigations.
A.2 Recommendations for using the sedimentation method
In order to get reliable results from the sedimentation test, the particles in the sample must be completely
deagglomerated. If necessary, a suitable dispersant or deflocculant should be added to keep the particles
from flocculating or coagulating.
In this case, to adjust the weight of the samples when an additive has been added, a blank cylinder must be
prepared by adding only the dispersant to the liquid. Also, three blank beakers must be prepared. These
beakers should be treated the same as described in steps 3 and 4 of clause 4.2.3. Determine the weight of
the residual additive by taking the average of the residue weight in the 3 blank beakers. Deduct this average
weight of residual additive from the mass of the sample in step 10.
The three blank beakers should also be used as controls to check that the additive does not interfere with
subsequent IR or XRD measurement on the respirable fraction represented in the supernatant.
A.3 Guidelines for the Determination of Crystalline Silica
Currently, the two fundamental techniques used for routine estimation of the crystalline silica content of a
respirable dust sample are X-ray diffraction (XRD) and infrared (IR) absorption analysis, e.g. using ISO 24095.
Crystalline silica can be quantified in bulk samples, and in the supernatant of the sedimented samples:
-
If they are uncontaminated with minerals and/or materials yielding overlapping and/or interfering peaks
with the principal crystalline silica reflections.
If standard reference materials equivalent in peak half-width (crystallinity) and particle size distribution are
available for comparison with the mineral or material sample under study.
If suitable software allows the use of standardless Rietveld refinement to obtain quantitative phase
information from a (crystalline) multi-component mixture.
11
TC WI :2012 (E)
The choice of the analytical technique depends largely on other materials present in the sample which may
interfere in the analysis (overlapping, matrix effects, high mass absorption coefficient, etc.). The general
approach to this problem is to identify the other phases present prior to quantification of the crystalline silica
content.
-
-
-
Non-exhaustive examples of commonly occurring phases whose XRD peaks may overlap the most
intense peaks of quartz are: cristobalite, albite, anorthite, aragonite, biotite, graphite, kaolinite,
magehemite, mullite, muscovite, sillimanite, wollastonite, and zircon.
Non-exhaustive examples of commonly occurring phases whose diffraction peaks may overlap the most
intense peaks of cristobalite are: quartz, tridymite, albite, anorthite, orthoclase, calcite, cordierite, kaolinite,
muscovite, and talc.
Common IR interference minerals in crystalline silica determination are: kaolinite, mullite, muscovite,
pyrophyllite, montmorillonite, and amorphous silica.
All analytical methods for crystalline silica analysis (XRD and IR) are particle size dependent. It is well known
that as particle size increases the response of XRD increases while the response of IR decreases.
Crystalline silica methods require calibration adapt into reference standards of known purity, specific particle
size and distribution and sample-to-sample homogeneity. Ideally the particle size distribution of the calibration
adapt into reference standards would closely match that of the samples being analysed.
The following recommendations apply whether using XRD or IR analysis:
-
-
Periodic phase identification of the minerals or material under investigation (by a mineralogist or a
petrologist or other experts) from representative bulk samples.
If possible, remove the interfering phases:
- by dissolution (e.g. carbonate minerals)
- by magnetic separation (e.g. maghemite)
- by ashing (e.g. organic material).
Choose the more appropriate technique accordingly.
Standard reference samples should have ideally the crystallinity and the same particle size distribution as
the samples under consideration.
Standard reference samples and samples must be prepared in the same manner.
Use rotating sample holder to maximise count statistics and random orientation of particles.
Select a quartz (or cristobalite) line free from interference from other phases present. Note that sensitivity
is reduced if secondary peaks are used.
For internal standard methods, the crystalline silica content of the samples under study should be in the
same range as those used for the calibration curve.
Specific recommendations when using XRD technique:
-
-
12
Monitor the performance of the diffractometer and correct for the gradual decline in X-ray tube emission.
If samples with a high mass absorption coefficient for the radiation used are encountered (e.g.,
maghemite with CuK radiation), use a different anode or remove the interfering phase (e.g., by magnetic
separation).
Use preferably the peak area instead of intensity peak height.
If using standardless Rietveld refinement, check and correct for the presence of amorphous substances.
TC WI :2012 (E)
Annex B
(informative)
Round Robin test to establish a SWeRF Reference Sample
B.1 Introduction
The IMA-Europe Metrology Working Group has arranged a Round Robin test in order to make a SWeRF
reference sample available. A full report on this work is available from the IMA Metrology WG. In this
appendix, the most relevant results are summarized.
B.2 Test material
Typical quartz flour from Quarzwerke was used as the test material. It was produced from processed silica
sand of Frechen deposit by iron-free grinding with subsequent air separation.
The particle size was selected so that the expected SWeRF was not close to the cut-off values of either 1 %
or 10 %.
Essentially the quartz content is 100 %, therefore in this case SWeRF is equal to SWeRFcs.
NOTE
Content is given in mass% throughout this appendix.
A 75 kg sample was homogenized and divided by coning and quartering to give a number of representative
portions, in accordance with ISO 14488/2007. Reference samples are still available upon request.
For further material data of the flour sample, see full report. Sample homogeneity was checked by laser
diffraction at the Quarzwerke central laboratory. A random sample from each of the 50 subsamples was
measured and evaluated (Fraunhofer approximation).
For the homogeneity check, the between-samples standard deviation (ss) was compared with the standard
deviation obtained from the proficiency assessment ( ) for laser diffraction [ISO 13320].The standard
recommends that for samples smaller than 10 µm, the coefficient of variation for D50 should be less than 6 %
and for D10 and D90, less than 10 %.
According to ISO 13528 samples may be considered as homogenous if:
ss  0.3 x
The subsamples comply with this requirement (cf. Table ).
13
TC WI :2012 (E)
Passing
Passing
Passing
D10
D50
D90
Mean
[µm]
0.83
3.18
9.70
Between-samples standard deviation (ss)
[µm]
0.01
0.02
0.04
Between-samples standard deviation (ss)
[%]
1.2
0.6
0.4
[%]
10
6
10
[%]
3
1.8
3
Standard deviation for proficiency assessment (
Requirement: ss  0,3 x
)
Table 1 Homogeneity check by laser diffraction (Fraunhofer) with 50 subsamples of the SWeRF reference
sample
B.3 Evaluation and appraisal method
The main focus of the round robin test is the estimation of SWeRF by laser diffraction and SediGraph
analysis. For the SediGraph analyses, automated X-Ray sedimentation was used according to ISO 133173:2001.
For laser diffraction, results based on the Fraunhofer theory was requested in a first run. The Mie theory
provides that particles "are homogeneous, isotropic and their optical properties are known" (cf. ISO
13320:2009 appendix A5). Because these data, especially effects depending on shape and surface of the
used quartz flour were not known, the Fraunhofer theory was in principle preferred. Nevertheless, some
laboratories sent in results using the Mie theory.
The initial laser diffraction data using the Mie theory showed large differences between the data of the
different laboratories. When the results were evaluated it became clear that the participants used different
optical data. Therefore, it was decided that analyses with harmonized Mie parameters from literature (e.g. ISO
13320: 2009) should be carried out in a second run. The following parameters were defined:
Refractive index of quartz:
1.54
Imaginary index of quartz:
0.1
Absorption coefficient of liquid (water): 1.33
The mean values of each participating laboratory were used as the basis of the evaluation. The results were
first checked for outliers after Dixon (cf. DIN 53804 T1) with significance of α = 0.05.
The statistical model is based on the means and standard deviations obtained from all laboratories.
The accepted tolerance was set using two different methods:
-
Relative reproducibility standard deviation (R). Acceptable tolerance within 10 %.
-
Calculation of z-score according to DIN 38402-45 with z <│2│delivered an acceptable result.
B.4 Results
The SWeRF results were spread out over a wide range. Only in the case of automated X-Ray sedimentation
(ISO 13317-3 : 2001) analysis was a statistical comparability (i.e. R > 10 %) obtained. Nevertheless the
resulting SWeRF mean lies on a similar level (range: 36.3 % to 39.0 %). Figure 1 gives an overview of the
results
14
TC WI :2012 (E)
Figure 1: Analysis results of the Round Robin test (green dotted line: mean of all values without outliers, blue
dotted lines: statistical tolerance limits of each method with outliers included, red lines: tolerance around mean
of all methods with Rrel ± 10 % of mean)
Gravitational Liquid
Sedimentation
Laser Diffraction
Fraunhofer
Mean of SWeRF
σRabs
σRrel
Statistical Tolerance
Mie (1)
X-ray
Fixed
Technique
Pipette
All methods
Mie (2)
[%]
[%]
[%]
[%]
[%]
[%]
38.3
36.3
38.1
39.0
39.8
38.0
3.62
14.6
3.77
2.46
2.45
7.32
9.5
40.4
9.9
6.3
6.1
19.2
31.0 - 45.5
7.0 – 65.6
30.5 – 45.6
34.1 – 43.9
35.0 – 44.7
23.4 – 52.7
Number of Participants
Outliers
z-score: tolerable level
21
11
8
11
10
61
2
0
2
0
1
5
18
11
6
11
9
z-score: warning signal
1
0
1
0
0
z-score: action signal
2
0
1
0
1
not
determined
Figure 2: Analysis results of the Round Robin test for different methods with statistical characteristics. Mie (1)
are results without harmonised parameters and Mie (2) are with harmonized parameters (refractive index
quartz np = 1.54 and kp = 0.1, refractive index water nm = 1.33). The statistical tolerance is the mean +/- 2
σRabs.
15
TC WI :2012 (E)
Annex C
(informative)
Derivation for calculating the sedimentation parameters
for determining by sedimentation
The probability that an inhaled particle reaches the alveoli of the lungs is given in EN 481 by the respirable
convention including the inhalability factor. This separation of the respirable particles is a function of the
aerodynamic diameter with decreasing probability with increasing diameter.
In order to separate the SWERF a sample is dispersed in a liquid and left to settle. After a calculated time the
top volume is extracted to a certain depth. Since larger particles settle faster than smaller once, this will
have resulted in a separation between coarse and fine particles. By choosing the right time and depth the
separation by sedimentation approaches the respirable convention. See Figure 1.
Separation efficiency
100
Respirable convention; R(D)
80
Sedimentation; S(D)
%
60
40
20
0
0,1
1
Dmax
10
µm
100
Figure 1: Graph of the separation efficiencies of the respirable convention and the sedimentation.
Both the respirable convention as well as the separation by sedimentation are probability functions. R(D)
describes the probability for a particle to enter the alveoli (EN 481) and S(D) describes the probability that a
particle remains in suspension. The SWERF is found by choosing the parameters for sedimentation in such a
way that the sum of all probabilities for all diameters is equal for both the respirable convention as well as
the sedimentation. This is done by equating the integrals of both functions.
(1)
The integral of R(D)
According to EN 481, sampling of the respirable fraction of particles with diameter (D) is the percentage ER of
the inhalable fraction EI.
EI is given as: 50(1+e(-0.06D)).
ER is given as the cumulative log-normal distribution with mean diameter of 4.25 µm and a geometric
standard deviation of 1.5 µm.
16
TC WI :2012 (E)
EN 481 gives the result in percentages but since the fraction is needed instead, this has to be divided by
100%. Function R(D) then becomes:
(2)
The integral of R(D) cannot be solved analytically but can be determined numerically and this integral (for
aerodynamic particles with unit density (1 g cm-3) ) is:
(3)
Using Stokes law the integral for particles with density ρp becomes:
(4)
where ρp is the density of the particles (kg m-3) and ρ0 the unit density (1000 kg m-3 [= 1 g cm-3]).
The integral of S(D)
According to Stokes law the sedimentation separation function for a fixed time is quadratic and can be
represented as:
(5)
where C is dependent on several factors but is constant during sedimentation.
(6)
(7)
(8)
For the special case of
(see fig.1) the probability is zero:
(9)
(10)
Combining (8) and (10):
(11)
(12)
(13)
Equating the integrals:
(14)
17
TC WI :2012 (E)
(15)
(16)
For the separation of the SWERF by sedimentation a sample is dispersed in a liquid and left to settle for a
certain time (t) after which the top volume is extracted to depth h. The largest particle in this volume has a
aerodynamic diameter of
. This means:
(17)
for which
is the settling velocity of a particle with diameter Dmax .
According to Stokes law:
(18)
where ρl is the density of the liquid (kg m-3).
Using (17) and (18):
(19)
With this equation the depth (h) can be calculated to which the volume has to be extracted after a certain
time of sedimentation. Or when a fixed depth is chosen, the time of sedimentation can be calculated as:
(20)
The sedimentation time is determined at the point where the integral of function R(D) is equal to that of the
sedimentation (see equation 1). However the function R(D) is different from S(D) so that the collection
efficiency for the sedimentation is only exactly the same as what may be expected according to EN 481 in
the case of a constant size distribution. In practice the differences are small and acceptable.
Calculation of the SWERF after sedimentation
The SWERF of a sample is calculated as:
(21)
For which:
 H is the height of the total column of fluid that is
h
used for sedimentation.
 h is the height to which the supernatant is extracted
at the calculated time.
 M is the total mass that was dispersed.
 m is mass of the residue in the extracted
supernatant.
18
s
H
Particle settles with
velocity (v) and covers
distance (s) in time (t).
TC WI :2012 (E)
Mixtures with not equal densities
The SWERF can only be accurately determined for particles with a well defined density. It is however still
possible to determine the SWERF of a constituent in a mixture with particles of different density. First the
sedimentation is performed using the time and height of extraction as calculated using the density of the
constituent (not the density of the bulk even if the total fraction of the constituent of interest is small). Then,
the amount of the constituent needs to be analysed in the residue using methods like XRD, IR, chemical
analyse, etc., resulting in:
(22)
For which:
 C is the constituent
 SWERF(C) is the SWERF of constituent C.
 f(C) is the fraction of constituent C in the residue in the supernatant.
In the same way it is possible to determine the SWERF of multiple constituents. In that case the
sedimentation needs to be performed for each density separately after which the fraction of the constituent
of interest is determined by analysis.
E.g. the SWERF of crystalline silica (cs) in a ceramic powder can be determined by performing the
sedimentation procedure twice; one time for the density of quartz and one time for the density of
cristobalite. The quartz fraction has to be determined in the residue of the supernatant of the ‘quartz
sedimentation’(Q) and the cristobalite fraction in that of the ‘cristobalite sedimentation’(crist.).
(23)
19
TC WI :2012 (E)
Annex D
(informative)
The Determination and Isolation of the Size Weighted Potential
Respirable Fraction (SWeRF) of Kaolins and Kaolinitic Clays by
Sedimentation
D.1 Introduction
This procedure describes the method which has been used at Imerys Minerals Ltd. for the determination and
isolation of the Size-Weighted Potential Respirable fraction (SWeRF) of kaolins and kaolinitic clays. The
procedure is based on the general instructions for the determination of SWeRF by sedimentation as described
in clause 4.3 of this standard. The method described in this annex is not a replacement for this procedure, but
it is supplementary in the sense that it gives a more detailed specification for a specific application of the
method. It essentially ‘scales-up’ the general procedure to allow sufficient respirable fraction to be obtained for
subsequent quantitative mineralogical analysis by X-ray powder diffraction.
D.2 Scope
The method applies to most kaolins and kaolinitic clays and may be adapted for other coarse to fine-grained
materials for which water is generally a good medium for separation by sedimentation.
Water is a suitable separation medium provided the following conditions are met:



The particles in the sample must be completely de-agglomerated
The particles in the sample should not dissolve, swell or disintegrate
The particles in the sample should not react
The method is therefore not particularly suitable to bentonite type minerals or kaolins / kaolinitic clays that
contain significant bentonite or other swelling minerals.
D.3 Equipment / Consumables
This method requires standard laboratory equipment and some additional items as described below.
Beakers (preferably glass) or similar containers to hold ~ 3 litres by volume
Buchner vacuum flask (preferably glass) of >2 litre capacity with implosion protection
Water bath with temperature control (+/- 1ºC) to accommodate the 3 litre glass beakers (a bath that holds 3-4
beakers gives a good compromise between taking up too much bench space and adversely affecting sample
throughput).
Siphon tube as described and illustrated below (alternatively, a curved glass-siphon can be used as well)
Demineralised (deionised) water
Analytical balance with an accuracy of 0.001g
Ultrasonic bath or shaker
Dessicator
20
TC WI :2012 (E)
Oven with temperature control to maintain at temperature of 105ºC +/- 5ºC.
Buchner Filter assembly (flask + funnel for filtering respirable fraction)
Whatman No. 50 filter papers
Millipore filter assembly
0.8 µm Cellulose filter membranes e.g. Whatman ME27
Evaporating basins
Dispersant – e.g. sodium hexametaphosphate (Calgon), sodium polyacrylate etc.
Dilute (~10%) Sodium carbonate solution for pH adjustment
Dilute (~10%) Sulphuric acid for pH adjustment
D.4 Method
D.4.1
Weigh out on a clean and dry dish or weighing boat approximately 40g of the sample material to be analysed
with a precision of 0.001g. Record the weight for use in the calculation.
Note:
Ideally the sample under test should be dried / free of moisture to allow accurate weighing. However, if it is not
possible / desirable to dry the sample beforehand then an equivalent ‘dry-weight’ may be determined if the moisture
content is known.
D.4.2
Add approximately 300ml of deionised water (ideally - that has been pre-heated to 25 ºC in a beaker in the
water bath) to a 600ml clean glass beaker and then add dispersant if desired at an appropriate dosage for the
type / particle size range of the mineral to be sedimented. The optimum dose of dispersant will depend on the
particle size distribution (surface area) of the material under test. The finer (higher surface area) materials will
require a higher dosage of dispersant. Typically, for kaolins, a dispersant dose of between 0.1 wt.% and
0.3wt.% on the equivalent dry-weight should suffice. E.g. 6g (mls) of 1 wt.% dispersant on 40g dry-weight of
kaolin would be equivalent to 0.15wt.% active dispersant.
D.4.3
Carefully add the pre-weighed sample material into the water/dispersant and gently mix until fully wetted-out.
Ensure that the entire sample is transferred to the mixture by either brushing or washing out with deionised
water. Full dispersion of the sample may then be achieved by using an ultrasonic probe or placing the beaker
of slurry into an ultrasonic bath for a specified time. The pH of the slurry should also be checked and adjusted
as necessary by the addition of a few drops of sodium carbonate solution (to raise pH) or dilute sulphuric acid
(to lower pH). The optimum pH will depend on the dispersant type used, e.g. for sodium hexametaphosphate
(Calgon) a pH of around 7 is optimum. If a polyacrylate dispersant is used then a higher pH around 9 – 9.5 is
preferred.
D.4.4
Transfer the dispersed slurry into a clean 3 litre glass beaker and dilute the slurry with further pre-heated deionised water ensuring thorough mixing. The level (height) of liquid in the beaker should then be made up to
the required height above the take-off level (see illustrations in 5.1). A height of 12cm (0.12m) has been used
at Imerys as this gives a sedimentation time of little over 2 hours. The use of 40g of sample in approximately 2
litres of water gives a slurry which is approximately 2 wt.% solids. This is sufficiently dilute to allow unhindered
settling but allows enough respirable fraction (‘fines’) to be collected for subsequent analysis.
21
TC WI :2012 (E)
D.4.5
Place the 3 litre beaker containing the sample slurry into the water bath and check that the temperature is
correct (typically 25ºC). If not, then the beaker and its contents must be allowed to adjust to the correct
temperature before starting the timed sedimentation. If the temperature is OK then give the slurry a final stir to
thoroughly mix and then leave to stand in the water bath, undisturbed, for the calculated time (see illustration
in 5.2). For quartz, this would be 2 hrs and 6 mins for a height of 12cm at 25 ºC in water. The excel file
provided by IMA may be used to calculate alternative times if alternative heights, temperatures etc. are used.
D.4.6
After the calculated time the siphon device (see illustrations in 5.3) is gently lowered into the slurry and the
supernatant transferred into a buchner flask using vacuum suction. The siphon device should be gradually
lowered into the supernatant until it reaches (touches) the base of the beaker. As soon as no further
supernatant can be removed the siphon device should be placed into a clean beaker containing approximately
50ml of de-ionised water. This clean water should be allowed to pass through the siphon device and be
collected along with the sample supernatant. This step ensures that all of the supernatant is transferred and
none remains within the siphon device.
D.4.7
The residue remaining in the 3 litre beaker (coarse fraction) can now be transferred into a clean, pre-weighed
evaporating dish using deionised water to wash-out the beaker into the dish. The dish with the residue should
then be placed in an oven (105 ºC) and dried to effectively zero moisture content. Once dry the evaporating
basin should be transferred into a dessicator and reweighed once cooled to ambient temperature. The dryweight of sample may be obtained by subtracting the weight of the pre-weighed evaporating basin from the
total weight (i.e. evaporating basin + sample). The weight of residue may then be used to calculate the wt.%
of residue and fines (SWeRF). e.g. If the starting sample weight was 40g and the dry-weight of sample
residue is weighed as 15.65g then the fine fraction is 40.0 – 15.65 = 24.35g. This value may then be
substituted into the SWeRF calculation Excel file to obtain the wt. percentage SWeRF – see example below.
D.4.8
The supernatant which was collected in the buchner vacuum flask may now be transferred initially to a clean
beaker (~2 litres) and the pH adjusted to ~4.5 with dilute sulphuric acid. This should flocculate the slurry and
allow it to be filtered on a vacuum filter using a Whatman No. 50 filter paper (or similar). If the filtrate is not
completely clear it will be necessary to re-filter it through a millipore type filter – 0.8µm Whatman type ME27 or
22
TC WI :2012 (E)
equivalent. If this second filtering process is necessary then the fines collected on both the paper and
cellulose filter must be combined. The total fines may now be dried in an oven at 105 ºC and the weight
recorded. This can be used as a cross-check to the weight of residue recorded previously. In theory the two
portions should, of course, add up to the original weight. However, in practice it is likely that a small amount of
material will be lost from the fines. This is why it is better to use the weight of residue to calculate both coarse
and fine (SWeRF) fractions. The dried respirable fraction may now be analysed using X-ray powder diffraction
in order to quantify the crystalline silica content.
D.5 Illustrations
D.5.1
General image of the siphon device (left), the height between the ‘take-off’ level and the top of the liquid
(centre) and the base of the siphon device showing the inlet slit in the tube to allow the supernatant to be
withdrawn without disturbing the sedimented material (right). The siphon tube is made by a piece of stainless
steel tubing with a slit cut into it near the base (as shown in the image) and a small cup or 'inverted umbrella'
attached to the base of the tube which also serves to block the end of the tube (below), so that the
supernatant fines can only enter the tube via the slit.
23
TC WI :2012 (E)
D.5.2
Image showing a general view of the water-bath with three samples in 3 litre glass beakers part-way through
the sedimentation.
D.5.3
Images showing the siphon device about to be lowered into the sedimented slurry (left) and the supernatant
collection flask connected to the siphon device and vacuum suction (right).
24
TC WI :2012 (E)
Annex E
(informative)
Other Minerals which may be treated in a similar way to Kaolins /
Kaolinitic Clays for SWeRF and SWeRFcs Determination
E.1 Introduction
The general SWeRF and SWeRFcs methodology developed for kaolins and kaolinitic clays may also be used
for the following minerals: Andalusite, Mica, Vermiculite and Talc.
E.2 Andalusite
Andalusite is a blocky, coarse alumina-silicate mineral but some products may contain fine particles.
Generally, Andalusite products are very coarse (> 1mm) and the procedure described below should be
followed.
E.2.1 Sedimentation procedure
1. Take approximately 100g of dry product and accurately weigh to 2 decimal places (weight A). Waterwash the 100g of product through a 1mm screen and collect the < 1mm fraction (the portion passing through
the screen) and the > 1mm fraction (the portion remaining on the screen) separately.
2.
Dry, to constant weight, the >1mm fraction in an oven at ~ 100°C (weight B).
3.
Calculate the weight percentage of > 1mm fraction (weight% C).
Weight % of >1mm fraction C = weight B/weight A x 100%
4. If the weight % of > 1mm is greater than 99.00 then no further action is required. (If weight % >1mm is >
99.00 then wt.% < 1mm will be <1wt.%).
5. If the weight % of > 1mm is less than 99.00 then screen sufficient product to carry out a sedimentation
analysis as detailed for kaolins and kaolinitic Clays.
6. Obtain the SWeRF value from the sedimentation of the < 1mm fraction obtained as above and multiply
this value by (100 – weight% C) to give the true SWeRF value for the whole fraction of the product.
7.
If the SWeRF value for the whole fraction of the product is < 1wt.% then no further action is required.
8. If the SWeRF value for the whole fraction of the product is > 1wt.% then carry out a mineralogical
analysis of the fine fraction obtained by sedimentation to obtain the crystalline silica content (quartz and / or
cristobalite).
9. The SWeRFcs value may then be obtained by multiplying the percentage of quartz or cristobalite
(whichever is the higher) in the fine fraction by the SWeRF value for the whole product.
E.2.2 Example
A sample of an Andalusite product (99.86g = weight A) is taken and screened at 1mm.
91.23 g (weight B) of the material passes through the 1 mm sieve.
Therefore 91.23 (weight B) / 99.86 (weight A) x 100 = 91.35wt.% (weight% C)
25
TC WI :2012 (E)
As the weight % (weight% C) of the >1mm fraction is less than 99.00 then sedimentation is necessary.
20.23g of <1mm screened product was subjected to the sedimentation procedure,
4.35g of material was found to be in the fine fraction,
i.e. SWeRF (of <1mm fraction) = 4.35/20.23 x 100 = 21.50wt.%
Therefore SWeRF in whole fraction = 21.50/100 x (100-91.35)/100 = 1.83wt.%
As the SWeRF for the whole sample is greater than 1 wt.% then mineralogical analysis is required to
determine the crystalline silica content of the fine fraction.
Using X-ray powder diffraction, no cristobalite was detected but 2.3 wt.% quartz was detected.
Therefore the SWeRFcs for this sample was calculated as 1.83 x 2.3/100 = 0.04 wt.%
As the wt.% crystalline silica was less than 1 wt.% then this example product is not classified as hazardous.
E.3 Mica
Mica is a platy mineral with similar particle shape to kaolins but generally a coarser particle size. For very
coarse mica products an initial pre-screening should be carried out to determine whether the full
sedimentation procedure is required (see E.2 Andalusite).
E.4 Vermiculite
Vermiculite, in its unexpanded form, is another platy mineral similar to a mica and kaolin but with an even
coarser particle size. Because of the coarse nature of Vermiculite, especially when expanded, it should be
given the same pre-screening process as for coarse micas and Andalusite (see E2 Andalusite).
E.5 Talc
Talc is a platy mineral, it exhibits some degree of hydrophobicity which means that a suitable surfactant
should be added to the talc / water suspension in order to thoroughly wet-out the mineral particles and give a
fully dispersed and deflocculated suspension.
E.5.1 Surfactant/Dispersant preparation
Add 2ml of Triton X100 to 500 mg of Calgon (Sodium metaphosphate) in a 2000 ml beaker, complete to 2000
ml with demineralised water and mix it during 45 min using a magnetic stirrer.
E.5.2 Sedimentation procedure
a)
Take a sample of the material to be analysed of approx. 5 g
b)
Determine the weight M of the sample to be analysed with a precision of 0.001 g.
c)
Disperse the sample in 80 ml of dispersant in a 175 ml dry and clean beaker.
NOTE: To avoid the presence of foam, introduce the demineralised water and the dispersant in several steps
in order to obtain a “pastry” and then a “liquid mixture”.
d)
26
Place the sample in an ultrasonic bath during 4 minutes.
TC WI :2012 (E)
e) Pour the dispersed sample in a 250 ml measuring cylinder. Rinse out the sample jar using the
demineralised water to ensure that no residue remains. Fill the cylinder up to 250 ml with demineralised water.
Seal the open end of the cylinder and shake 4 times the contents thoroughly.
NOTE: The volume of solids should be maximum 1 % of the volume of the total liquid to ensure unhindered
sedimentation of the separate particles.
f) Place the cylinder in a location where it is at constant temperature and free from effects that could cause
currents in the liquid and leave to settle for the calculated time (t = 1 hour). Determine the depth of the liquid
column H (m).
g)
Determine the weight M1 of an aluminium cup with an accuracy of 0.001g.
h) From the cylinder, after calculated time (t), lower a pipette to an insertion depth of H ( 5cm) below the
surface, draw the volume of the supernatant (Vt) above the tip of the pipette with a siphon or pro-pipette and
transfer into the pre-weighed aluminium cup.
i) Make sure that, whilst using the pipette, the tip of the pipette remains in the same place in the cylinder; do
not insert it deeper as the level descends.
NOTE: Placing a clip on the pipette and attaching it to the edge of the cylinder will make it easier to keep the
pipette in the right place.
j) Re-weigh the aluminium cup to within 0.001 gram accuracy and note the post-weight (M2). Determine the
weight of the residue (m) by subtracting the weight of the pre-weighed aluminium cup (m= M2-M1).
27
TC WI :2012 (E)
Annex F
(informative)
The Determination of the Size Weighted Potential Respirable Fraction
(SWeRF) of Diatomaceous Earth (DE) by calculation and by
sedimentation
F.1 Introduction
This procedure describes the method for the determination of the size-weighted potential respirable fraction
(SWeRF) of diatomaceous earth (DE). The general instructions for the determination of SWeRF by
sedimentation or by calculation described in the standard were adapted by taking into account a feature of the
DE ie, its internal porosity. The porous structure of DE varies as a function of the size of the particles and
influences their own density. Consequently, the determination of the SWeRF requires the preliminary
measurement of the effective density (ρeff) of the fine fraction of the products, which reflects the one of the
respirable fraction.
F.2 Scope
The DE are classified in three categories of product : natural, calcined and flux-calcined.

Natural DE are not concerned by the SWeRF method because they do not contain (or less than 1%)
crystalline silica.

Calcined DE : use the effective density of the finest calcined grade of the product range.

Flux-calcined DE : use the effective density of the finest flux-calcined grade of the product range.
F.3 Equipment / Consumables
Particle size analyser Mastersizer 2000 (Malvern) or equivalent
Mercury porosimeter Autopore IV 9500 (MICROMERITICS)
Mercury of high purity, hexadistilled.
Silica alumina standard from MICROMERITICS of about 0,5 ml/g.
Analytical balance with an accuracy of 0.001g
Gas Pycnometer Accupyc 1330, Micromeritics (LPX544)
Helium gas (quality: Classe 2-1°A)
F.4 Method
F.4.1 Determination of the absolute density
The absolute density (or skeleton density) is defined as the mass of a unit volume of the DE skeleton,
inaccessible to Helium. A known mass of DE, dried at 150°C over night, is introduced in the sample cell, and
its volume is measured by gas displacement.
3
3
The absolute density (ρs) is expressed as kg/m with an accuracy of 10 kg/m .
28
TC WI :2012 (E)
(Note that Mercury porosimetry allows also to determine the absolute density of powders, but Helium
pycnometry is more appropriate because the gaz can penetrate inside the smallest pores).
F.4.2 Determination of the intra-particular pore volume and intra-particle porosity
Note: the intra-particle porosity of a powder is composed with open and closed pores. Closed pores are completely
isolated from the external surface, not allowing the access of external fluids liquid or gaz. For the final calculation of the
effective density of the particles of DE, the volume of closed pores is considered as negligeable .
Pore volume and pore volume distribution of DE is measured with Autopore IV 9500 (MICROMERITICS) by
Mercury Intrusion with pressures of Mercury varying from 3580 Pa to 207 MPa. A representative mass of DE,
3
with a volume below 3 cm , is weighed and introduced in the penetrometer. This mass is chosen so that the
level of Mercury in the calibrated rod of the penetrometer varies between the beginning of the analysis and the
end (at maximal pressure) of a value between 20 and 90 % of the possible total variation. The method of
measurement used is a capacitive method.
The intrusion curves are generated using the adapted software (v1.05). A typical intrusion curve representing
the cumulative intrusion volume of mercury per gram of DE as a function of the diameter of the pores is
divided in several parts :
 The first part of the curve, with an apparent continuous intrusion, is due to the de-agglomeration of the
particles: they are completely separated and packed.
 The steep part of the curve corresponds to the mercury penetration in the inter-particle voids (space
between packed particles). This phase is materialized by an abrupt change in the filling rate on the
intrusion curve. The total volume of the inter-particle voids is considered to be the volume of mercury
intruded at the inflection point of the curve, for a size of pores named “critical diameter”.
 In the last part of the curve, internal pores are filled by mercury, up to the maximum selected pressure.
Note: the volume of mercury intruded are expressed in ml/g of sample. For the calculation, the values are converted in
m3/kg (SI units).
The intra-particle pore volume is calculated with the formula: Vi = Vtot – Vcrit
3
Vi = intra-particle pore volume for a given mass of sample (m /kg)
3
Vtot = total cumulative pore volume for a given mass of sample (m /kg)
3
Vcrit = cumulative pore volume at critical diameter for a given mass of sample (m /kg)
The particle porosity is defined as the ratio of the volume of open pores to the volume of the particle. The
volume of the particle is the combination of the volume of the open pores plus the volume of the skeleton. The
volum of a given mass of skeleton is calculated among the absolute (skeletal) density: Volum of skeleton =
1/ρs
Vi
P=
Vi + 1/ ρs
Vi
x 100
or
P=
Vi + Vs
x 100
P = Porosity (%)
3
ρs = absolute density (kg/m )
3
Vi = intra-particle pore volume for a given mass of DE (m /kg)
3
Vs = Volum of skeleton for a given mass of DE (m /kg)
29
TC WI :2012 (E)
F.4.3 Calculation of the effective density
DE is porous material. It behaves in sedimentation as having the density of the combination of the amount of
fluid (air, water or other media) contained in the open pores and the amount of skeletal structure. This density
called effective density (ρeff) is calculated as follow:
ρeff =
P x ρf + (100 – P) x ρs
100
ρeff = effective density (kg/m3)
P = Porosity (%)
ρs = absolute density (kg/m3)
ρf = fluid density (kg/m3)
F.4.4 Particle size analysis
Particle size distribution of the sample is determined by laser diffraction, in liquid (demineralised water), with a
presonication during 20s and the use of Mie model with the following parameters:
Refractive index: 1.45
Imaginary index: 0.001
Absorption coefficient of liquid (water): 1.33
F.5 Determination of SWeRF by sedimentation
F.5.1 Preparation of the sample
Boil 5g of a sample of DE in 50ml of demineralised water. Cool to 20°C. Transfer in a 250ml measuring
cylinder, adjust the volume to 250ml and let to settle for a calculated time (t), after which the supernatant
containing the respirable fraction is collected with a pipette, dried and weighted according to the standard
procedure.
F.5.2 Determination of sedimentation time
The time of sedimentation is calculated using the following equation which is based on Stokes’ law and the
convention described in EN 481.
t  h
30
18  η
4
SGs
 
2
x


(ρ s  ρ l )  g 9 

  EN 481dx 


 x 0

TC WI :2012 (E)
With:
SGs
8 η
4
 
2
x 

ρ l)  g 9 
  EN 481dx 


 x 0
 4,281 (µm).
=
t = time (s) at which separation of the particles by sedimentation in the liquid equals separation according EN
481
h= height (m) of the column of the supernatant liquid that is extracted after time = t.
η= dynamic viscosity of liquid (kg/ms)
g= acceleration due to gravity (m/s2)
ρs= density of the solid particles (kg/m3)
ρl= density of the liquid medium (kg/m3)
SGs = specific gravity of the solid particles (water = 1)
x = aerodynamic diameter (µm)
Conditions defined for DE analysis :
Height = 0.01 m
Sedimentation liquid = demineralised water
Density of water at 20°C: ρwater = 998.205 kg/m3
P x ρwater + (100 – P) x ρs
SG = ρeff in water =
100
F.6 Determination of SWeRF by calculation
The calculation method, being based on the probability for particles reaching the alveoli when inhaled,
concerns the moving of the particles in air. In this condition, the specific gravity of the particles corresponds to
their effective density in air.
P x ρair + (100 – P) x ρs
SG = ρeff in air =
100
3
ρair = 1.204 kg/m at 20°C, atmospheric pressure (101325 Pa)

Calcined DE : measure the porosity of the finest calcined grade and calculate the effective density ρeff
calcined in air. This value will be used for all the range of calcined grade (fine, medium, coarse) from the
same origin (mining, process plant).
31
TC WI :2012 (E)

Flux-calcined DE : measure the porosity of the finest flux-calcined grade (filler) and calculate the effective
density ρeff flux-calcined in air. This value will be used for all the range of flux-calcined grade (fine,
medium, coarse) from the same origin (mining, process plant).
The SWeRF is calculated by entering the particle size distribution (determined by laser diffraction) of the
whole sample in the prepared calculation sheet, and the specific gravity of the corresponding product
(ρeff calcined in air or ρeff flux-calcined in air) as the density of the sample.
F.7 Determination of SWeRFcs
In DE, it is assumed that the particle size distribution of the sample and of its crystalline silica content are
related. So the SWeRFcs is calculated by multiplying the SWeRF of the sample with the crystalline silica
content of the sample.
F.8 Example
The following data are given as an example for the calculation of the effective density and their use in the
method of calculation or sedimentation.
Grade:
Fine Flux-calcined EAF 7053
Skeletal density:
measured
Pycnometry
by
Helium ρs = 2300 kg/m3
Vs = 4.35 x 10-4 m3/kg
Volum of skeleton Vs = 1/ρs
for 1 kg of DE:
Intraparticle
volume:
determined
by
mercury Vi = 0.95ml/g ie, in SI units
intrusion, per gram of sample
Vi = 0.95 x 10-3 m3/kg
Porosity:
0.95 x 10
Vi
P=
Vi + Vs
x 100
P=
-3
-3
-4
0.95 x 10 + 4.35 x 10
x 100
P = 68.6%
Effective density in
air
ρeff =
in air
P x ρair + (100 – P) x ρs
100
68.6 x 1.204 + (100 – 68.6) x 2300
ρeff =
100
in air
ρeff in air = 723 kg/m3
Use this value in the prepared calculation sheet for
all the range of flux-calcined grade (fine, medium,
coarse) from the same origin (mining, process plant),
and enter the Particle Size Distribution of each grade
of flux-calcined DE.
Effective density in
water
ρeff =
in water
P x ρwater + (100 – P) x ρs
68.6 x 998.205 + (100 – 68.6) x 2300
ρeff =
100
in water
100
ρeff in water = 1406 kg/ m3
Use this value in the prepared calculation sheet to
determine the time of sedimentation that must be
32
TC WI :2012 (E)
used for all the range of flux-calcined grade (fine,
medium, coarse) from the same origin (mining,
process plant).
33
TC WI :2012 (E)
Annex G
(informative)
The Determination of the Size Weighted Potential Respirable Fraction
(SWeRF) of feldspar products by sedimentation
G.1 Subject
This instruction is derived from NEN 5753-ISO/DIS 11277. The instruction describes a method for determining
the content of respirable particles in a sample of a substance containing crystalline silica.
G.2 Scope
This instruction applies to coarse to fine-grained material samples (maximum 100 µm), for which water is
generally a good medium for separation by sedimentation.
Water is suitable as a separation medium provided the following conditions are met:
a. the particles in the sample must be completely deagglomerated
b. the particles in the sample should not dissolve, swell or disintegrate
c. the particles in the sample should not react
e.g. clays or cement should not be analysed in water
G.3 Equipment
Standard laboratory glassware and the following supplies:
1)
2)
3)
4)
5)
6)
7)
8)
Measuring cylinders 250 ml high-model or equivalent cylinders
Demineralised (deionised) water
Sodium hewametaphosphate (NaPO3)6
Analytical balance with an accuracy of 0.001 grams
Ultrasonic bath or shaker.
Desiccator
o
Oven with an automatic temperature control set to 103  5 C
Sedimentation Excel sheet for calculations of sedimentation time
G.4 Method
4.1 Weigh in a sample jar approximately 5 g of the sample material to be analysed with a precision of 0.001
gram. Note the weight (M) and enter in the Sedimentation Excel sheet.
Prepare a small glass beaker of approximately 50 ml capacity by heating in the oven at 103 ± 5 oC for 1 hour.
Let the glass beaker cool down for 15 minutes in a desiccator and weigh it to within 0.001 gram accuracy.
4.2 Add approximately 20 ml demineralised water to the sample jar to prepare a slurry. Add 2.5 ml of sodium
hexametaphosphate 50g/l (0.05%). Place the sample jar in an ultrasonic bath for about 5 minutes, or for 1
hour in a shaker.
4.3 From the sample jar, transfer the slurry into a cylinder. It is good practice to rinse out the sample jar using
demineralised water to ensure that no residue is left behind. Fill the cylinder up to 250 ml with demineralised
water. Cover with a piece of parafilm and shake the contents thoroughly.
Note that for some materials it will be inadequate to shake the contents and therefore the use of a plunger
aggitator with the cylinder in place is recommended.
34
TC WI :2012 (E)
Then leave the cylinder to stand, vibration free, for the calculated time (t) using formula (3) of paragraph 5.1 of
this standard (approx. 2 hours). Keep the suspension of the material at a constant temperature by using e.g. a
thermal bath. Determine the depth of the liquid in the cylinder (H).
4.4 From the cylinder, after calculated time (t), lower a pipette to an insertion depth of (h) below the surface,
draw the entire supernatant (Vt) from the cylinder with a siphon or pipette and transfer into the pre-weighed
beaker. See the figure (Fig1) below for a siphon setting.
Make sure that, whilst using the pipette, the tip of the pipette remains in the same place in the cylinder; do not
insert it deeper as the level descends. Placing a clip on the pipette and attaching it to the edge of the cylinder
will make it easier to keep the pipette in the right place.
Figure 1. Example of a siphon setting for transferring the supernatant.
4.5 Place the beaker on a hot plate to evaporate the liquid and heat gently until dry.
Place the beaker in the oven at 103 ° C for one hour and transfer into a desiccator, leaving it to cool down for
a quarter of an hour.
Re-weigh the beaker to within 0.001 gram accuracy and note the post-weight (mpo).
4.6 The above procedure should initially be repeated twice in order to check the reproducibility of the
sedimentation method for SWeRF determination.
4.7 The following additional steps should be taken only if a dispersant or deflocculant additive has been used!
To adjust the weight of the samples when an additive has been added, prepare a blank cylinder by adding
only the dispersant to the liquid. Prepare three blank beakers and treat these beakers the same as described
above. Using the equation below, determine the weight of the residual additive mra. Deduct the mass of
residual additive from the mass of the sample.
The three blank beakers should now also be used as controls to check that the additive does not interfere with
subsequent FT-IR or XRD measurement.
mra 
1
[mb1,2  mb1,1   (mb2,2  mb2,1 )  (mb3,2  mb3,1 )]
3
Where:
mb1,2 = re-weight of beaker of blank 1 in g
mb2,2 = re-weight of beaker of blank 2 in g
35
TC WI :2012 (E)
mb3,2 = re-weight of beaker of blank 3 in g
mb1,1 = pre-weight of beaker 1 in g
mb2,2 = pre-weight of beaker 2 in g
mb3,1 = pre-weight of beaker 3 in g
The SWeRF (as a percentage) is calculated by using the following formula:
SWeRF 
H  m  mra 
 100%
hM
Weighing table for beakers
Beaker
Reference
Blank 1
Blank 2
Blank 3
Sample 001
Sample 002
Sample 003
Sample 004
Sample 005
Mean of blanks
to be deducted
from the result
36
Pre-weight
(g)
Post-weight
(g)
Difference
(g)
TC WI :2012 (E)
Annex H
(informative)
The relation between density and SWeRF
H.1 Subject
In determining the SWeRF of a material it is of paramount importance to consider that the actual aim is to
predict how the particles would move in air. There are two ways to determine the SWeRF of a material. One is
to first determine the size distribution of the material and then calculate the SWeRF from this PSD using the
probability function described in EN 481. The other way is to disperse the material in a liquid in which the
SWeRF is separated by removing the larger particles by sedimentation.
H.2 Determining SWeRF from size distribution
To calculate the SWeRF the aerodynamic diameter is needed. Aerodynamic diameter is an expression of a
particle's settling behaviour in air.
a perfect sphere with the same density as water reach the same
terminal velocity in air, then the aerodynamic diameter of the particle is equal to that of the sphere.
v
(=1)
v
In general this means that for most materials having a density greater than 1 the aerodynamic diameter is
larger than the geometric diameter. However particles with high aspect ratios (needles and platelets) or
porous particles will settle slower and the aerodynamic diameter can become even smaller than its geometric
diameter.
H.2.1 Aerodynamic diameter and laser diffraction
This is straight forward when dealing with solid spheres. When the particles are all perfect spheres, the
diameter can be converted to aerodynamic diameter. Using Stokes law this relation is given by:
ρ
As long as the particles have a reasonably rounded character the laser diffraction will report an
equivalent spherical diameter which is generally not far from the truth. When this is the case the particle
size distribution can be converted to an aerodynamic PSD using the equation above. For particles with
high aspect ratios (needles and platelets) laser diffraction will tend to overestimate their diameter and
additionally because of their shape these particles behave as smaller particles in air. As a consequence
of this laser diffraction should not be used for these kinds of particles because it will result in a SWeRF
that is too low.
H.2.2 Porous and hollow particles
When dealing with porous and hollow particles one should consider how these particles would behave
in air.
37
TC WI :2012 (E)
When the PSD is determined by laser diffraction the reported diameter will be close to the
equivalent spherical diameter (assuming the particles have a reasonably rounded aspect). To
convert to aerodynamic diameter the effective density of the whole particle including all voids
should be used. This is the weight of the particle divided by its volume. This effective density
can also be calculated using the open and closed porosity of the particles:
ρ
In which:
ρ
ρ
ρ
When the PSD is determined by sedimentation techniques one needs to consider that the
settling rate is directly proportional to the result of the gravitational force and the buoyancy of
the particle. This means that both for the analysis and to convert to aerodynamic diameter a
density should be used that compensates for air filled closed pores and open pores filled with
liquid. In this case the density of the particle should be calculated as:
ρ
ρ
ρ
In which:
ρ
ρ
This is to correct for the fact that the open pores are filled with liquid (and gain weight).
H.2.3 Aerodynamic diameter and sedimentation techniques
The size distribution determined by sedimentation has the advantage that the shape of the
particles affects the behaviour of these particles in the same way in a liquid as it does in air.
This is because with sedimentation based particle sizing techniques (such as Sedigraph) the
diameters are reported as equivalent Stokes diameters. The equivalent Stokes diameter of a
particle is the diameter of a perfect sphere with the same density as that of the particle and
which settles at the same rate. 1
Stokes law is valid for both liquid and gasses (at low Reynolds numbers). The velocity of a
particle settling in a medium is limited by the drag force and this depends on the Reynolds
number for that particle. Although the density of liquids are much higher than air their viscosity
is also higher . This results in only a small difference between the Reynolds number of a
particle in air and in, for example water where the difference is only a factor of 5. Although the
constants for calculating the drag coefficient of particles depend on the Reynolds number, the
variation with Reynolds number within this range is very small and can be neglected. This
means that the Stokes diameter (in liquid) is essentially the same to the aerodynamic diameter.
Since the shape of the particle is already taken into account, the aerodynamic diameter can be
calculated from the equivalent Stokes diameter (ESD):
ρ
1 This is only valid for low Reynolds numbers but because of the small size when dealing with/calculating SWeRF, this is
always the case.
38
TC WI :2012 (E)
H.3 Determining SWeRF by sedimentation.
When the SWeRF is determined by sedimentation the shape of the particles will have an effect on the result.
Particles with a high aspect ratio will result in a higher SWeRF when compared to rounded particles of the
same mass. However, as explained above, the particles will settle in a liquid according to their equivalent
Stokes diameter which can be converted to their corresponding aerodynamic diameter by multiplying the ESD
by the square root of the density.
For porous and hollow particles the effect of porosity on the density to be used should compensate differently
for open and closed pores. Particles with only closed pores can be seen as particles with the same outside
shape, only with a lower density. If a particle has an open porosity, liquid will fill these pores and add weight to
the particle increasing its density. Therefore, the density of these particles should be calculated as:
ρ
ρ
ρ
H.3.1 Determining SWeRF of mixtures
When a material consists of a mixture of multiple components, the separation by sedimentation should
be performed based on the density of the materials for which the value of the SWeRF is required. The
next step is to determine the concentration of the required component in the supernatant (E.g. by X-ray
powder diffraction, Infra-red spectroscopy or other suitable techniques depending on the component).
If the SWeRF value for more than one component is required, then the separation needs to be
performed multiple times. E.g. when a material contains both quartz and cristobalite and the SWeRF of
both are required, the sedimentation needs to be performed once using the density of quartz and a
second time using the density of cristobalite. The next step is to determine the concentrations of quartz
and cristobalite in their separated fractions.
39