0% found this document useful (0 votes)
159 views30 pagesBas 3
Uploaded by
renata jazicCopyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
159 views30 pagesBas 3
Uploaded by
renata jazicCopyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
/ 30
ISBI raéun br: 04/22-528 - 2022-11-15 - Gross d.o.0. Gradika PJ Srebrenica Rudnik olovai cinka Sase -Bro|icenci: 1
BAS ISO 12743:2022
ISO 12743:2021(E)
Annex A
(normative)
Sampling stage method of estimating sampling and total variance
A.1 Components of sampling error and sampling variance
‘As chown by Gy/[3I the total sampling error TSE can be broken up into a number of componente
corresponding to each sampling stage 1, 2, mui, .., and hence is given by Formula (A.1):
TSE = TSE, + we TSE, + wae # TSE ay
where
TSE, is the sampling error for stage 1;
‘TSE, isthe sampling error for stage i:
TSE,, is the sampling error for stage u, the last stage.
‘The above break-up is possible because each component of sampling error is independent. The errors
may be random or systematic (i.e. bias).
Each sampling stage consists of two operations. These are selection (or sampling) and preparation. In
this context, preparation is a non-selective operation involving operations such as crushing or drying
Consequently, the total sampling error is given by Formula (4.2):
TSE = SE+PE (a2)
where
SE isthe selection error;
PE_ is the preparation error.
‘Typical preparation errors include sample contamination, sample loss, alteration of the chemical or
physical composition of the sample and operator mistakes.
‘The selection error can be further broken up into the integration error CE and the materialization error
ME and hence is given by Formula (A.3):
SE=CE+ME @3)
‘The integration error arises from the manner in which the sampling points are selected on the time or
mass axes. The materialization error arises from the physical manner in which increments are taken
and can be eliminated by correct cutter design and operation.
‘The integration error also consists of two components caused by variations in quality and flow rate and
is given by Formula (4.4):
cE=Qe+ WE (4)
where
(© 180 2021 - All rights reserved 51 br 04/22-528 - 2022-11-15 - Grossd.o.0. Gradiska Pd Srebrenica Rudnik olovai cinka Sase - Brojlicenci: 1
sal
BAS ISO 12743:2022
ISO 12743:2021(E)
QE_ isthe quality fluctuation error;
WE is the weighting error.
‘The quality fluctuation errors are of three types, namely short-range, long-range and periodic. Hence
the quality fluctuation error is given by Formula (A.5):
QE = QE; + QE, + QE, (A5)
where
QE, isthe short-range quality fluctuation error;
QE, is the long-range quality fluctuation error;
QE, isthe periodic quality fluctuation error.
The shurtrange fluctuations result from two properties related to the particulate nature of the
concentrate. These are the composition of the particles (fundamental error) and the manner in which
the particles are grouped (segregation or grouping error). Thus, the short-range quality fluctuation
error is given by.
QE; = FE+GE (a6)
where
PE is the fundamental error;
GE. is the segregation and grouping error.
‘The materialization error can be further broken up into the delimitation error DE and the extraction
eivur EE, as given in Eurmula (A.7).
E+ EE (a7)
The delimitation error is eliminated ifall parts of the concentrate stream are intercepted by the sample
cutter for the same length of time, The extraction error is eliminated if the increment is completely
extracted from the stream without any concentrate rebounding from the cutter.
Combining Formulae (4.2) to (A.7) gives Formula (A.6) for the sampling error at each stage:
TSE =FE + GE + QE, + QE, + WE + DE+EE + PE (Ag)
The last three error components in Formula (4.8), ‘e. DE, EE and PE, are systematic errors, which
introduce bias. They arise from not respecting the correct principles of sampling from the mechanical
standpoint. They can be eliminated by using correct sampling practices, which are described in
Clauses 8, 9, 10 and 11. Practical experience with concentrates also shows that the weighting error WE,
is negligible compared with QE,, even when there are significant variations in flow rate. Likewise, the
periodic quality fluctuation error QE; is also negligible, except in exceptional cases where production
52 @ 180 2021 ~Allrights reserved ISBIH racun br: 04/22-528 - 2022-11-15 - Gross d.0.0. Gradiska PJ Srebrenica Rudnik olove i cinka Sase -Broj icenci: 1
BAS ISO 12743:2022
ISO 12743:2021(E)
or stockpiling procedures introduce some periodicity. Hence, the formula for the total sampling reduces
to Formula (A.9) as follows:
TSE=FE+GE + QE, @s)
or, alternatively, Formula (4.10):
TSE = OF; +9, (a0)
‘These error components are random errors,
From Formulae (A) and (4.10), it is clear that the variance of the total sampling error is given by
Formula {A.1)):
Sash +5, +53 (aan)
s@ isthe total sampling variance;
s?, isthe fundamental variance;
sfy__ is the segregation and grouping variance;
SGp, is the long-range quality fluctuation variance;
or, alternatively, Formula (A.12):
B=, +e, =e (A.12)
where
Gz, _ is the short-range quality fluctuation variance:
se _ 1s the quality tluctuation variance.
In Formula (A.12), the long-range quality fluctuation variance is often referred to as the distribution
variance.
A.2_ Estimation of fundamental variance
Gylal has shown that the variance of the fundamental error, s2,., is given by Formula (4.13):
daz
(a3)
C isthe sampling constant for a given concentrate of given particle size and critical constituent;
d__ isthe nominal top size of the concentrate, in centimetres;
img is the sample mass at a given sampling stage, in grains;
a _ isthe fractional concentration of the constituent under consideration.
(© 180 2021 -Allrights reserved 53 ISBIH raGun br: 04/22-528 - 2022-11-15 - Gross d.o.0. Gradi8ka PJ Srebrenica Rudrik olove i cinka Sase - Broj Kcenci: 1
BAS ISO 12743:2022
ISO 1274:
021(E)
‘The sampling constant Cis given by Formula (4.14):
Cael fg (a4)
where
° is the mineralogical composition factor calculated using Formula (A.15);
I is the liberation factor;
F is the particle shape factor, which can usually be taken to be 0,5;
9 is the size range factor, ucually betweon 0,25 and 1,03,
[q/a) Whenliberation is incomplete, d, being the nominal top size at which complete liberation
occurs; != 1 when liberation is complete.
If d, is unknown, a conservative assumption is to set: ¢, = d.
‘The mineralogical composition factor is given by Eormula (A.15):
(1-a)[(1-a) p, +299) (Aas)
where
’P; is the density of the particles of the critical component, in grams per cubic centimetre;
P, is the density of gangue particles, in grams per cubic centimetre.
‘The size-range factor g can be estimated from the ratio d/d’of the nominal top size d to the lower size
limit d’ (about 5 % undersize) as follows:
Large size range (d/d' > 4) 9
Medium size range (2
