ISO 20904:2006

INTERNATIONAL ISO
STANDARD 20904
First edition
2006-10-01


Hard coal — Sampling of slurries
Houille — Échantillonnage des schlamms





Reference number
ISO 20904:2006(E)
©
ISO 2006

---------------------- Page: 1 ----------------------
ISO 20904:2006(E)
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but
shall not be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In
downloading this file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat
accepts no liability in this area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation
parameters were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In
the unlikely event that a problem relating to it is found, please inform the Central Secretariat at the address given below.


©  ISO 2006
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or
ISO's member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Published in Switzerland

ii © ISO 2006 – All rights reserved

---------------------- Page: 2 ----------------------
ISO 20904:2006(E)
Contents Page
Foreword. v
1 Scope. 1
2 Normative references. 1
3 Definitions . 1
4 Principles of sampling slurries . 2
4.1 General. 2
4.2 Sampling errors. 3
4.3 Sampling and overall variance . 6
5 Sampling schemes. 7
6 Minimization of bias and unbiased increment mass. 13
6.1 Minimizing bias . 13
6.2 Volume of increment for falling stream samplers to avoid bias. 14
6.3 Volume of increment for manual sampling to avoid bias. 14
7 Precision of sampling and determination of increment variance. 15
7.1 Overall precision. 15
7.2 Primary increment variance. 15
7.3 Preparation and testing variance . 16
8 Number of sub-lots and number of increments per sub-lot. 16
9 Minimum mass of solids in lot and sub-lot samples . 17
9.1 General. 17
9.2 Minimum mass of solids in lot samples. 17
9.3 Minimum mass of solids in sub-lot samples .17
9.4 Minimum mass of solids in lot and sub-lot samples after size reduction . 17
10 Time-basis sampling. 18
10.1 General. 18
10.2 Sampling interval. 18
10.3 Cutters. 18
10.4 Taking of increments. 18
10.5 Constitution of lot or sub-lot samples. 19
10.6 Division of increments and sub-lot samples . 19
10.7 Division of lot samples. 19
10.8 Number of cuts for division . 19
11 Stratified random sampling within fixed time intervals. 19
12 Mechanical sampling from moving streams. 20
12.1 General. 20
12.2 Design of the sampling system. 20
12.3 Slurry sample cutters. 22
12.4 Mass of solids in increments. 23
12.5 Number of primary increments . 23
12.6 Routine checking. 23
13 Manual sampling from moving streams. 23
13.1 General. 23
13.2 Choosing the sampling location . 23
13.3 Sampling implements. 24
13.4 Mass of solids in increments. 24
13.5 Number of primary increments . 24
© ISO 2006 – All rights reserved iii

---------------------- Page: 3 ----------------------
ISO 20904:2006(E)
13.6 Sampling procedures . 25
14 Sampling of stationary slurries . 25
15 Sample preparation procedures. 25
15.1 General. 25
15.2 Reduction mills . 25
15.3 Sample division. 25
15.4 Chemical analysis samples. 25
15.5 Physical test samples. 26
16 Packing and marking of samples . 26
Annex A (informative) Examples of correct slurry devices. 27
Annex B (informative) Examples of incorrect slurry sampling devices. 30
Annex C (normative) Manual sampling implements. 34
Bibliography . 35

iv © ISO 2006 – All rights reserved

---------------------- Page: 4 ----------------------
ISO 20904:2006(E)
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO 20904 was prepared by Technical Committee ISO/TC 27, Solid mineral fuels, Subcommittee SC 4,
Sampling.

© ISO 2006 – All rights reserved v

---------------------- Page: 5 ----------------------
INTERNATIONAL STANDARD ISO 20904:2006(E)

Hard coal — Sampling of slurries
1 Scope
This International Standard sets out the basic methods for sampling fine coal, coal rejects or tailings of
nominal top size < 4 mm that is mixed with water to form a slurry. At very high ratios of fine solids to water
when the material assumes a soft plastic form, the mixture is correctly termed a paste. Sampling of pastes is
not covered in this International Standard.
The procedures described in this International Standard primarily apply to sampling of coal that is transported
in moving streams as a slurry. These streams can fall freely or be confined in pipes, launders, chutes, spirals
or similar channels. Sampling of slurries in stationary situations, such as a settled or even a well-stirred slurry
in a tank, holding vessel or dam, is not recommended and is not covered in this International Standard.
This International Standard describes procedures that are designed to provide samples representative of the
slurry solids and particle size distribution of the slurry under examination. After draining the slurry sample of
fluid and measuring the fluid volume, damp samples of the contained solids in the slurry are available for
drying (if required) and measurement of one or more characteristics in an unbiased manner and with a known
degree of precision. The characteristics are measured by chemical analysis or physical testing or both.
The sampling methods described are applicable to slurries that require inspection to verify compliance with
product specifications, determination of the value of a characteristic as a basis for settlement between trading
partners or estimation of a set of average characteristics and variances that describes a system or procedure.
Provided flow rates are not too high, the reference method against which other sampling procedures are
compared is one where the entire stream is diverted into a vessel for a specified time or volume interval. This
method corresponds to the stopped-belt method described in ISO 13909-2.
2 Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments) applies.
ISO 1213-1, Solid mineral fuels — Vocabulary — Part 1: Terms relating to coal preparation
ISO 1213-2, Solid mineral fuels — Vocabulary — Part 2: Terms relating to sampling, testing and analysis
ISO 13909-1, Hard coal and coke — Mechanical sampling — Part 1: General introduction
ISO 13909-4, Hard coal and coke — Mechanical sampling — Part 4: Coal — Preparation of test samples
ISO 13909-8, Hard coal and coke — Mechanical sampling — Part 8: Methods of testing for bias
3 Definitions
For the purpose of this document, the definitions given in ISO 13909-1, ISO 1213-1 and ISO 1213-2 apply.
© ISO 2006 – All rights reserved 1

---------------------- Page: 6 ----------------------
ISO 20904:2006(E)
4 Principles of sampling slurries
4.1 General
For the purposes of this International Standard, a slurry is defined as fine coal, coal rejects or tailings of
nominal top size < 4 mm that is mixed with water, which is frequently used as a convenient form to transport
coal, rejects or tailings though plant circuits by means of pumps and pipelines and under gravity in launders or
chutes or through long distances in slurry pipelines. Tailings from wet plants are also discharged as a slurry
through pipelines to the tailings dam. In many of these operations, collection of increments at selected sample
points is required for evaluation of the coal or rejects in the slurry.
A lot or sub-lot sample is constituted from a set of unbiased primary increments from a lot or sub-lot. The
sample container is weighed immediately after collection and combination of increments to avoid water loss by
evaporation or spillage. Weighing is necessary to determine the mass percentage of solids in the lot or sub-lot
sample. The lot or sub-lot sample can then be filtered, dried and weighed. Alternatively, the lot or sub-lot
sample may be sealed in plastic bags after filtering for transport and drying at a later stage.
Except for samples for which their characteristics are determined directly on the slurry, test samples are
prepared from lot or sub-lot samples after filtering and drying. Test portions may then be taken from the test
sample and analysed using an appropriate and properly calibrated analytical method or test procedure under
prescribed conditions.
The objective of the measurement chain is to determine the characteristic of interest in an unbiased manner
with an acceptable and affordable degree of precision. The general sampling theory, which is based on the
additive property of variances, can be used to determine how the variances of sampling, sample preparation
and chemical analysis or physical testing propagate and hence determine the total variance for the
measurement chain. This sampling theory can also be used to optimize mechanical sampling systems and
manual sampling methods.
If a sampling scheme is to provide representative samples, it is necessary that all parts of the slurry in the lot
have an equal opportunity of being selected and appearing in the lot sample for testing. Any deviation from
this basic requirement can result in an unacceptable loss of accuracy. A sampling scheme having incorrect
selection techniques, i.e. with non-uniform selection probabilities, cannot be relied upon to provide
representative samples.
Sampling of slurries should preferably be carried out by systematic sampling on a time basis (see Clause 10).
If the slurry flow rate and the coal-solids concentration vary with time, the slurry volume and the dry solids
mass for each increment will vary accordingly. It is necessary to show that no systematic error (bias) is
introduced by periodic variation in quality or quantity where the proposed sampling interval is approximately
equal to a multiple of the period of variation in quantity or quality. Otherwise, stratified random sampling
should be used (see Clause 11).
Best practice for sampling slurries is to mechanically cut freely falling streams (see Clause 12), with a
complete cross-section of the stream being taken during the traverse of the cutter. Access to freely falling
streams can sometimes be engineered at the end of pipes or by incorporating steps or weirs in launders and
chutes. If samples are not collected in this manner, non-uniform concentration of coal solids in the slurry due
to segregation and stratification of the solids can lead to bias in the sample that is collected. Slurry flow in
pipes can be homogenous with very fine particles dispersed uniformly in turbulent suspension along the length
and across the diameter of the pipe. However, more commonly, the slurry in a pipe has significant particle-
concentration gradients across the pipe and there can be concentration fluctuations along the length of the
pipe. These common conditions are called heterogeneous flow. Examples of such flow are full-pipe flow of a
heterogeneous suspension or partial-pipe flow of a fine suspension above a slower moving or even stationary
bed of coarser particles in the slurry.
For heterogeneous flow, bias is likely to occur where a tapping is made into the slurry pipe to locate either a
flush-fitting sample take-off pipe or a sample tube projecting into the slurry stream for extraction of samples.
The bias is caused by non-uniform concentration profiles in the pipe and the different trajectories followed by
particles of different masses due to their inertia, resulting in larger or denser particles being preferentially
rejected from or included in the sample.
2 © ISO 2006 – All rights reserved

---------------------- Page: 7 ----------------------
ISO 20904:2006(E)
In slurry channels such as launders, heterogeneous flow is almost always present, and this non-uniformity in
particle concentration is usually preserved in the discharge over a weir or step. However, sampling at a weir or
step allows complete access to the full width and breadth of the stream, thereby enabling all parts of the slurry
stream to be collected with equal probability.
Sampling of slurries in stationary situations, such as a settled or even a well-stirred slurry in a tank, holding
vessel or dam is not recommended, because it is virtually impossible to ensure that all parts of the slurry in the
lot have an equal opportunity of being selected and appearing in the lot sample for testing. Instead, sampling
should be carried out from moving streams as the tank, vessel or dam is filled or emptied.
4.2 Sampling errors
4.2.1 General
The processes of sampling, sample preparation and measurement are experimental procedures, and each
procedure has its own uncertainty appearing as variations in the final results. When the average of these
variations is close to zero, they are called random errors. More serious variations contributing to the
uncertainty of results are systematic errors, which have averages biased away from zero. There are also
human errors that introduce variations due to departures from prescribed procedures for which statistical
analysis procedures are not applicable.
The characteristics of the solids component of a slurry can be determined by extracting samples from the
slurry stream, preparing test samples and measuring the required quality characteristics. The total sampling
[5] [6]
error, E , can be expressed as the sum of a number of independent components (Gy, 1982 ; Pitard, 1993 ).
T
Such a simple additive combination is not possible if the components are correlated. The total sampling error,
E , expressed as a sum of its components, is given by Equation (1):
T
E=+EE+E+E+E+E+E (1)
TQ1 Q2 Q3 W D E P
where
E is short-range quality fluctuation error associated with short-range variations in quality of the solids
Q1
component of the slurry;
E is long-range quality fluctuation error associated with long-range variations in quality of the solids
Q2
component of the slurry;
E is periodic quality fluctuation error associated with periodic variations in quality of the solids
Q3
component of the slurry;
E is weighting error associated with variations in slurry flow rate;
W
E is increment delimitation error introduced by incorrect increment delimitation;
D
E is increment extraction error introduced by incorrect increment extraction from the slurry;
E
E is the preparation error introduced by departures (usually unintentional) from correct practices, e.g.
P
during constitution of the lot sample, draining and filtering away the water, and transportation and
drying of the sample.
© ISO 2006 – All rights reserved 3

---------------------- Page: 8 ----------------------
ISO 20904:2006(E)
The short-range quality fluctuation error consists of two components, as shown by Equation (2):
E=+EE (2)
Ql F G
where
E is the fundamental error due to variation in quality between particles;
F
E is the segregation and grouping error.
G
The fundamental error results from the composition heterogeneity of the lot, i.e. the heterogeneity that is
inherent to the composition of each particle making up the solids component of the lot. The greater the
differences in the compositions of particles, the greater the composition heterogeneity and the higher the
fundamental error variance. The fundamental error can never be completely eliminated. It is an inherent error
resulting from the variation in composition of the particles in the slurry being sampled.
The segregation and grouping error results from the distribution heterogeneity of the sampled material (Pitard,
[6]
1993 ). The distribution heterogeneity of a lot is the heterogeneity arising from the manner in which particles
are distributed in the slurry. It can be reduced by taking more increments, but it can never be completely
eliminated.
A number of the components of the total sampling error, namely E , E and E , can be minimized or reduced
D E P
to an acceptable level by correct design of the sampling procedure.
4.2.2 Preparation error
In this context, the preparation error, E , includes errors associated with non-selective sample preparation
P
operations that should not change mass, such as sample transfer, flocculation, draining and filtering, drying,
crushing, grinding or mixing. It does not include errors associated with sample division. Preparation errors
include sample contamination, loss of sample material, alteration of the chemical or physical composition of
the sample, operator mistakes, fraud or sabotage. These errors can be made negligible by correct design of
the sample plant and by staff training. For example, cross-stream slurry cutters should have caps to prevent
entry of splashes when the cutter is in the parked position and it is necessary to take care during filtering to
avoid loss of fines that are still suspended in the water to be discarded.
4.2.3 Delimitation and extraction errors
Delimitation and extraction errors arise from incorrect sample cutter design and operation. The increment
delimitation error, E , results from an incorrect geometry of the volume delimiting the slurry increment (see
D
Figure 1), and this can be due to both design and operation faults. Because of the incorrect shape of the slurry
increment volume, sampling with non-uniform selection probabilities results. The average of E is often non-
D
zero, which makes it a source of sampling bias. The delimitation error can be made negligible if all parts of the
stream cross-section are diverted by the sample cutter for the same length of time.
Sampling from moving slurry streams usually involves methods that fall into three broad operational
categories as follows:
a) taking the whole stream part of the time with a cross-stream cutter as shown in Figure 1 a) (after Pitard,
[6]
1993 ), usually when the slurry falls from a pipe or over a weir or step. Cuts 1 and 2 show correct
sampling with the cutter diverting all parts of the stream for the same length of time. Cuts 3, 4 and 5 show
incorrect sampling where the cutter diverts different parts of the stream for different lengths of time;
[6]
b) taking part of the stream all of the time as shown in Figure 1 b) (after Pitard, 1993 ) with an in-stream
point sampler or probe within a pipe or channel, which is always incorrect;
[6]
c) taking part of the stream part of the time as shown in Figure 1 c) (after Pitard, 1993 ), also with an in-
stream point sampler or probe within a pipe or channel, which is always incorrect.
4 © ISO 2006 – All rights reserved

---------------------- Page: 9 ----------------------
ISO 20904:2006(E)

a)  Taking all of the stream part of the time

b)  Taking part of the stream all of the time (always incorrect)


c)  Taking part of the stream part of the time (always incorrect)
a
Correct.
b
Incorrect.
Figure 1 — Plan view of slurry volumes diverted by sample cutters
The increment extraction error, E , results from incorrect extraction of the slurry increment. The extraction is
E
said to be correct if, and only if, all particles in the slurry that have their centre of gravity inside the boundaries
of the correctly delimited increment are extracted. The average of E is often non-zero, which makes it a
E
source of sampling bias. The extraction error can be made negligible by ensuring that the slurry increment is
completely extracted from the stream without any particulate material being lost from the cutter due to
splashes. It is necessary that the depth and capacity of the cutter be sufficient to avoid slurry reflux from the
cutter aperture, resulting in loss of part of the extracted slurry increment.
4.2.4 Weighting error, E
W
The weighting error is an error component arising from the selection model underlying Equation (1). In the
model, the time-dependent flow rate of the solids in the slurry stream is a weighting function applied to the
corresponding time-dependent quality characteristic over time, which gives the weighted-average quality
characteristic of the solids component of the lot. The weighting error results from the application of incorrect
© ISO 2006 – All rights reserved 5

---------------------- Page: 10 ----------------------
ISO 20904:2006(E)
weights to the quality characteristics. The best solution to reducing the weighting error is to stabilize the flow
rate. As a general rule, the weighting error is negligible for variations in flow rate of up to 10 % relative and
acceptable for variations in flow rate up to 20 % relative.
4.2.5 Periodic quality fluctuation error, E
Q3
Periodic quality fluctuation errors result from periodic variations in quality generated by some equipment used
for slurry processing and transportation, e.g. grinding and screening circuits, splitters and pumps. The
presence of periodic variations can be detected by determining the variogram (see ISO 13909-7). While in
most cases variogram values can be fitted with a simple linear or quadratic function, if periodic behaviour
(characterized by regularly spaced maxima and minima) is observed, the fitting function can include a sine-
[5]
wave term with a period and amplitude to be determined as parameters of the fit (Gy, 1982 ). In such cases,
stratified random sampling should be carried out as discussed in Clause 11. The alternative is to significantly
reduce the source of periodic variations in quality, which can require plant redesign.
4.3 Sampling and overall variance
4.3.1 Sampling variance
Assume that the weighting (E ), increment delimitation (E ), increment extraction (E ) and preparation errors
W
...