Washability

Analysis

Anand Anupam
Assistant Professor
Department of Fuel, Minerals and
Metallurgical Engineering
IIT (ISM), Dhanbad

March 2020

for academic purposes only

INTRODUCTION

Washability analysis is the laboratory technique of determining the ease with which a coal or an
ore may be cleaned using density separation. Washability analysis also delivers the information about the
expected yields, characteristic ash (or grade), Near Gravity Material (NGM) content, separation density,
etc. it is a crucial characterisation technique for density separator selection and performance audit.
Washability analysis comprises of experimentation and data analysis. These have been discussed in the
following sections.

EXPERIMENTATION: FLOAT-SINK TESTS

The experimental part of the washability analysis

consists of float-sink test, also known as, sink-float test
or heavy liquid analysis. The float-sink test is a simple Preparation
mineral engineering laboratory test, which obtains its of media
footing from the Archimedes’ principle. The objective of
the test is to separate a given sample into pre-determined
Float-sink
density fractions. This is achieved by preparing fluid tests
media of different densities and submerging the test
sample into those media successively, either in increasing
or in decreasing order. The density fractions thus Ash/grade
analysi
obtained are weighed, after cleaning off the media,
ash/grade analysed and the data is tabulated and
appraised. This process is briefly illustrated in Figure 1. Data
analysis
The float-sink tests can be of two types – simple
float-sink tests and comprehensive or complete float-sink
tests; the latter being the default connotation of the term Figure 1 Washability analysis work-flow
‘float-sink test’. The difference between the two types being in the number of the different density media
used. In simple float-sink test only one medium of a particular density is used. It is commonly used by plant
operators for a quick assessment of their operating density separation process. On the other hand, the general
(comprehensive) float-sink test is carried out over the entire range of density of particles present in the
sample.

The float-sink tests are common for coal coarser than around 0.5 mm, as only the coarser coal is
cleaned by density separation methods. Finer coal is generally cleaned through froth flotation, for which a
different prerequisite test, known as release analysis, is used.

Coal Preparation: Lecture Notes 1

PREPARATION OF MEDIA

The range of densities of the media that need to be prepared for the float-sink tests needs to cover
the entire range of particle densities present in the test sample (coal or ore). Therefore, the common range
of medium relative densities for coal samples is 1.3 to 1.8; for Indian coal samples, the higher limit should
be at least 2.0 and can be as high as 2.4, depending upon the coal. The lower limit of the range of relative
densities for ores is generally 2.55, or the relative density of the anticipated lightest particle (usually gangue).
The higher limit of the range ideally depends upon the density of the highest density particle (valuable
mineral), however, in most cases it is impractical to prepare media of such high density. Therefore, the
higher relative density limit is most commonly determined based on the available media. From the
preceding, it is clear that the density of the media is always greater than 1 g cm -3, the density of water.
Therefore, the media used in float-sink tests are also known as heavy liquids and the test work is also known
as heavy liquid analysis.

There are two types of media that are used for float-sink tests – aqueous solutions and organic
liquid mixtures. The aqueous solutions are prepared by mixing such inorganic salts in water which lead to
appreciable increase in the density of the solution thus formed. Table 1 lists the commonly used salts which
are used for float-sink tests. The solution density is controlled by adjusting the concentration of the salts in
the solution. The maximum solution density that can be attained depends upon the solubility of the salt used
which in turn depends upon the purity of the salts and the solution temperature. Heavy media can also be
prepared by proportionate mixing of the heavy organic liquids, like the ones listed in Table 2. That is, for
preparing a medium of any desired density an organic liquid of lower than the desired density can be mixed
in proportionate amount with an organic liquid of the higher than the desired density. The following
equation may be used for calculating the volumes of the different liquids to be used for making the heavy
medium.
𝜌𝑀 − 𝜌𝐿
𝑉𝐻 = 𝑉𝑀 ( )
𝜌𝐻 − 𝜌𝐿
where, 𝑉 and 𝜌 represent volume and density respectively, and the subscripts 𝐻 , 𝑀 and 𝐿
correspond to the higher density liquid, the medium and the lower density liquid respectively. The organic
liquid mixtures offer the advantages of low medium viscosity, high medium density without any adverse
effect on the medium rheology and fast and complete removal from the test sample after the test. However,
such mixtures are toxic, volatile, have unstable density (over long periods of time) and are generally quite
expensive; more so for higher densities. It is noteworthy that the densities indicated in Table 2 depend upon
the ambient temperature and the purity of the liquids. Recently, heavy liquids for float-sink tests have also
been prepared using magneto-hydrostatics for float-sink tests of nonmagnetic samples coarser than 50 µm.

Washability Analysis 2
Such heavy liquids are prepared by subjecting a ferrofluid or a paramagnetic salt solution to external
magnetic field. Through this technique, separations with relative densities of up to 18 have been realised
for float-sink tests.
Table 1 Properties of selected aqueous solutions used as media for float-sink tests

Maximum solution
Solute Remarks
density (g cm-3)

Quite commonly used; low-cost;

Zinc chloride (ZnCl2) 2.0
corrosive; substantial viscosity

Sodium polytungstate [Na6(H2W12O40)] 3.1 Non-toxic, low viscosity, costly

Equal parts of thallium formate The aqueous solution is commonly

[Tl(HCO2)] and thallium malonate 5.0 known as Clerici solution; very costly;
[Tl(C3H3O4)] highly toxic

For float-sink tests, selection of the appropriate Table 2 Some organic liquid densities
medium densities is critical for subsequent data analysis. The
range of medium densities chosen should be such that it Organic Liquid Density (g cm-3)

virtually covers the entire range of particle densities present Naphtha 0.70
in the sample. Within this range, it is recommended to use
Petroleum spirit 0.77
intervals of at least 0.05 g cm-3 between the medium densities
for preparing media with densities in the range 1.25 g cm-3 White spirit 0.77
to 1.7 g cm-3. Intervals of 0.1 g cm-3, or less, should be kept
Acetone 0.78
for preparing media with densities higher than 1.7 g cm-3.
While preparing the media It should also be ensured that the Toluene 0.86

mass present in each of the end density ranges is not more Tetracholoroethylene 1.60
than 5% of the total mass. Besides, for reliable data analysis
Bromoform 2.89
the mass in none of the density ranges should exceed 15% of
the entire sample mass. If need be, even smaller density Tetrabromoethane 2.96
intervals like 0.02 g cm-3 should be used for increasing the
Diiodomethane 3.32
reliability of the results. The float-sink tests that are done for
determining the separation efficiency of a density separator should use density intervals of 0.02 g cm -3 for
0.1 g cm-3 around the anticipated separation density. The maximum interval between the medium densities
should normally not be more than 0.1 g cm-3.

The mass of the sample being used for a float-sink test is also of grave significance. The mass of
the test sample should be enough so that there are at least 2000 particles in the test sample, failing which

Coal Preparation: Lecture Notes 3

the reliability and applicability of the results will be questionable. The minimum mass that should be used
for float-sink test depends upon the top size of the sample. For Indian coals, the minimum mass that should
be used in a float-sink test can be obtained as shown in Table 3.
Table 3 Minimum mass of sample for Indian coal to be float-sink tests

100 to 50 to 25 to 13 to 6 to 3 to 1.5 to 0.5 to Less than

Top size (mm)
200 100 50 25 13 6 3 1.5 0.5

Minimum mass
1,500 1,000 500 250 50 15 5 2 1
(kg)

FLOAT-SINK TESTS

After the preparation of the different density media, they are taken in suitable containers, which
are placed successively in the order of the densities of the media that they contain. The media stored in the
containers are also referred to as baths. The size of the container depends upon the top size of the particles
being treated and the amount of the increment desired to be tested; for a 75 mm top size sample, the
container size could be 100 litres, while for fine particles having sizes around 0.5 mm, glass beakers of 1-
2 litre capacity would be sufficient. Then, based on the experience of float-sink tests carried with a similar
sample, it is decided, to either proceed the test starting with the lowest density medium or the highest density
or with an intermediate density medium; the results being independent of this choice, as long as the correct
test procedures are followed. The test sample is successively immersed in each of these media, in increments,
and sufficient amount of time is allowed for the particles to either float or sink, completely. The time taken
for the complete floating or sink varies between half an hour for particles as fine as 0.5 mm to about a
minute for coarse particles, for coal. For still finer particles of coal, it is a general practice to provide more
than 12 hours for the particle movement within the media. To shorten this time, centrifugal force can also
be used, although with caution, as this may lead
to size degeneration of the particles – thus
affecting the accuracy of the results.
Afterwards, the floated and sunk fractions in a
medium are taken out separately, cleaned from
any adhering medium and weights noted down.
The removal of floated fraction is simple and
simply involved careful skimming off of the
floating layer using a strainer or similar
equipment. The sunk fraction is removed after Figure 2 Float-sink test process

Washability Analysis 4
removal of the floats from the container. The sinks are collected either by first decanting the medium and
then collecting the sinks over a different strainer, or by using a wire mesh bucket inserted within the medium
container before any increment of the test sample was introduced. Figure 2 shows an example of the flow
of a float-sink test of coal. with the test being conducted in increasing order of the densities of the media
used.

After successive float-sink tests has been completed for all the media and the sample particles have
been cleansed off the media, the different density fractions of the sample, so obtained are carefully labelled
for the density ranges that they represent and their respective weights are noted. Each of these density
fractions are then subjected to ash/grade analysis. The mass and ash/grade data obtained are then analysed.

DATA ANALYSIS

A sample washability data obtained from float- Table 4 Washability data for coal
sink tests followed by ash analysis for coal has been
shown in Table 4. For carrying out the data analysis on RD* Ranges Weight (kg) Ash (%)

the raw washability data, as shown in Table 4, several < 1.30 7.23 4.96
calculations are performed to obtain the complete 1.30 - 1.40 27.33 10.89
washability table, as shown in Table 5. The first step is
1.40 - 1.45 42.38 15.41
to convert the raw weight values (obtained in the units of
1.45 - 1.50 43.61 21.39
mass) to weight (%) values. Please note that, the first row
1.50 - 1.55 36.79 27.63
in Table 5 indicates the column number. The values in
Table 5 are calculated row-by-row, starting from the top, 1.55 - 1.60 20.65 32.11
except for Column 8. The operations that are performed 1.60 - 1.70 31.09 37.59
for obtaining each of the columns in Table 5 have been 1.70 - 1.80 15.88 44.17
listed in The CFM values (Column 9) of Table 5 have
1.80 - 1.90 10.90 49.54
been obtained assuming a linear increment in the RD of
1.90 - 2.00 9.53 56.37
particles within the RD ranges. These values also assume
that the particle RDs are uniformly distributed within the 2.00 - 2.10 7.97 63.57
RD ranges. These assumptions only hold good, if there 2.10 - 2.20 6.58 67.66
are sufficient number of particles present in the test > 2.20 38.07 79.32
sample. The NGM refers to the group of all those

*
Relative Density

Coal Preparation: Lecture Notes 5

particles which have RDs near about the separation RD. Classically, RDs which are in the ±0.1 range of
the separation RD are considered as ‘near about’ the separation RD.

Table 6. Apart from the operations listed in The CFM values (Column 9) of Table 5 have been
obtained assuming a linear increment in the RD of particles within the RD ranges. These values also assume
that the particle RDs are uniformly distributed within the RD ranges. These assumptions only hold good, if
there are sufficient number of particles present in the test sample. The NGM refers to the group of all those
particles which have RDs near about the separation RD. Classically, RDs which are in the ±0.1 range of
the separation RD are considered as ‘near about’ the separation RD.

Table 6, a few other methods are also available in the literature; all of which provide the same
results within the limits of approximation error.

In Table 5, the differential (also known as individual for fractional) values (Columns 2 and 3) are
with respect to the corresponding RD ranges (Column 1), whereas the cumulative values (Columns 5
through 8) are not with respect to the RD ranges. The cumulative values (Columns 5 through 8) are
calculated and interpreted with respect to the corresponding NRD (Column 4).
Table 5 Washability table for coal

1 2 3 4 5 6 7 8 9 10

Differential Cumulative Float Cumulative Sink

RD NRD CFM NGM
Weight Ash † Weight Ash Weight Ash ‡ §
Ranges
(%) (%) (%) (%) (%) (%)

< 1.30 2.43 4.96 1.30 2.43 4.96 97.57 36.05 1.21 -

1.30 - 1.40 9.17 10.89 1.40 11.60 9.65 88.40 38.66 7.01 38.03

1.40 - 1.45 14.22 15.41 1.45 25.82 12.82 74.18 43.11 18.71 45.80

1.45 - 1.50 14.64 21.39 1.50 40.45 15.92 59.55 48.45 33.14 48.13

1.50 - 1.55 12.35 27.63 1.55 52.80 18.66 47.20 53.90 46.63 39.18

1.55 - 1.60 6.93 32.11 1.60 59.73 20.22 40.27 57.65 56.26 29.71

1.60 - 1.70 10.43 37.59 1.70 70.16 22.80 29.84 64.66 64.94 15.76

†
Nominal Relative Density
‡
Cumulative float weight at mid RD in the RD range (%)
§ Near Gravity Material (±0.1) weight (%)

Washability Analysis 6
 1.70 - 1.80 5.33 44.17 1.80 75.49 24.31 24.51 69.11 72.82 8.99

1.80 - 1.90 3.66 49.54 1.90 79.15 25.48 20.85 72.54 77.32 6.86

1.90 - 2.00 3.20 56.37 2.00 82.34 26.68 17.66 75.47 80.74 5.87

2.00 - 2.10 2.67 63.57 2.10 85.02 27.84 14.98 77.60 83.68 4.88

2.10 - 2.20 2.21 67.66 2.20 87.23 28.85 12.77 79.32 86.12 -

> 2.20 12.77 79.32

The CFM values (Column 9) of Table 5 have been obtained assuming a linear increment in the RD
of particles within the RD ranges. These values also assume that the particle RDs are uniformly distributed
within the RD ranges. These assumptions only hold good, if there are sufficient number of particles present
in the test sample. The NGM refers to the group of all those particles which have RDs near about the
separation RD. Classically, RDs which are in the ±0.1 range of the separation RD are considered as ‘near
about’ the separation RD.
Table 6 Methods for completing the washability table shown in Table 5

Column number
Method for obtaining values**
in Table 5
1
2 From the float-sink test work
3
4 Taking the higher limit of Column 1
Summing up of all Differential Weight (%) values (Column 2) from the
5
top row to the current row, including the value in the current row
Taking the weighted average of the values of Differential Ash (%)
(Column 3) by using the corresponding values of Differential Weight
6
(%) (Column 2) as weighting factor from the top row to the current row,
including the values in the current row
Subtracting the value of Cumulative Float Weight (%) (Column 5) from
7
100, in the current row
Taking the weighted average of the values of Differential Ash (%)
8
(Column 3) by using the corresponding values of Differential Weight

**
Column numbers refer to the columns in Table 5

Coal Preparation: Lecture Notes 7

 (%) (Column 2) as weighting factor from the bottom row to the current
row, excluding the values in the current row
Adding half of the Differential Weight (%) (Column 2) value of the
9 current row to the Cumulative Float Weight (%) (Column 5) value of the
immediately above row, if available
Taking the Weight (%) in the range ±0.1 of the Nominal Relative
10
Density (Column 4) value given in the current row
Based on the washability table (Table 5), the family of washability curves is plotted, as shown in
Figure 3. The description of each of the washability curves has been summarised in Table 7.

Figure 3 Washability curves

Washability Analysis 8
 Table 7 Description of the washability curves††

Washability curve Horizontal axis Vertical axis Utility

Cumulative float Cumulative To predict the yield of clean coal‡‡

Float Curve ash (%) (Column float weight corresponding to given clean coal ash content,
6) (%) (Column 5) and vice versa

Cumulative sink Cumulative To predict the yield of the reject

Sink Curve ash (%) (Column sink weight (%) corresponding to given reject ash content, and
8) (Column 7) vice versa

To predict the levels of characteristic ash

Characteristic Differential ash CFM (Column
corresponding to clean coal yield, and vice
Curve (%) (Column 3) 9)
versa

Cumulative
Yield-Gravity To predict the separation RD for a particular
NRD (Column 4) float weight
Curve clean coal yield, and vice versa
(%) (Column 5)

NGM (Column To predict the expected NGM weight (%) for

NGM Curve NRD (Column 4)
10) separation at a particular RD, and vice versa

††
Column numbers refer to the columns in Table 5
‡‡
Simply put, clean coal yield is the commercial equivalent of cumulative float weight (%)

Coal Preparation: Lecture Notes 9

BIBLIOGRAPHY

Bureau of Indian Standards, 1993. IS : 13810 - 1993. Code of practice for float and sink test of coal
(Reaffirmed 2000).

Leonard, J.W. (Ed.), 1991. Coal Preparation, 4th ed. American Institute of Mining, Metallurgical, and
Petroleum Engineers, New York, USA.

Osborne, D.G., 1988. Coal Preparation Technology. Graham & Trotman Limited, London.

Wills, B.A., Napier-Munn, T., 2005. Wills’ Mineral Processing Technology. Elsevier.
https://doi.org/10.1016/B978-0-7506-4450-1.X5000-0

IMAGE SOURCES

Image Source URL

https://www.science-sparks.com/wp-content/uploads/2018/09/shutterstock_1146846485-2-
Cover
768x620.jpg

Figure 2 https://www.uky.edu/KGS/coal/coal-analyses-float-sink.php

(Bureau of Indian Standards, 1993; Leonard, 1991; Osborne, 1988; Wills and Napier-Munn, 2005)

Washability Analysis 10