BS ISO 11352:2012

BSI Standards Publication

Water quality — Estimation

of measurement uncertainty
based on validation and quality
control data
BS ISO 11352:2012 BRITISH STANDARD

National foreword
This British Standard is the UK implementation of ISO 11352:2012.
The UK participation in its preparation was entrusted to Technical
Committee EH/3/2, Physical chemical and biochemical methods.
A list of organizations represented on this committee can be
obtained on request to its secretary.
This publication does not purport to include all the necessary
provisions of a contract. Users are responsible for its correct
application.
© The British Standards Institution 2012. Published by BSI Standards
Limited 2012
ISBN 978 0 580 61295 4
ICS 13.060.45
Compliance with a British Standard cannot confer immunity from
legal obligations.
This British Standard was published under the authority of the
Standards Policy and Strategy Committee on 31 July 2012.
Amendments issued since publication
Date Text affected
 BS ISO 11352:2012
INTERNATIONAL ISO
STANDARD 11352

First edition
2012-07-01

Water quality — Estimation of

measurement uncertainty based on
validation and quality control data
Qualité de l’eau — Estimation de l’incertitude de mesure basée sur des
données de validation et de contrôle qualité

Reference number
ISO 11352:2012(E)

© ISO 2012
BS ISO 11352:2012
ISO 11352:2012(E)

COPYRIGHT PROTECTED DOCUMENT

© ISO 2012
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 [email protected]
Web www.iso.org
Published in Switzerland

ii  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

Contents Page

Foreword............................................................................................................................................................................. iv
Introduction......................................................................................................................................................................... v
1 Scope....................................................................................................................................................................... 1
2 Normative references.......................................................................................................................................... 1
3 Terms and definitions.......................................................................................................................................... 1
4 Symbols.................................................................................................................................................................. 4
5 Principle.................................................................................................................................................................. 5
6 Procedure............................................................................................................................................................... 5
7 Preparative considerations for the estimation of measurement uncertainty...................................... 6
7.1 Specification of the measurement................................................................................................................... 6
7.2 Specification of the parametric form in which the measurement uncertainty is reported............... 6
8 Evaluation of available precision and bias data.......................................................................................... 7
8.1 Approach and criteria.......................................................................................................................................... 7
8.2 Within-laboratory reproducibility..................................................................................................................... 8
8.3 Method and laboratory bias............................................................................................................................. 10
9 Calculation of the combined standard uncertainty................................................................................... 14
10 Calculation of the expanded uncertainty..................................................................................................... 15
11 Estimation of measurement uncertainty from reproducibility standard deviation.......................... 15
12 Report.................................................................................................................................................................... 15
Annex A (normative) Estimation of the standard uncertainty from range control charts............................. 16
Annex B (informative) Examples of the estimation of measurement uncertainty........................................... 17
Bibliography...................................................................................................................................................................... 26

© ISO 2012 – All rights reserved  iii

BS ISO 11352:2012
ISO 11352:2012(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 11352 was prepared by Technical Committee ISO/TC 147, Water quality, Subcommittee SC 2, Physical,
chemical and biochemical methods.

iv  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

Introduction

The basic principles of the estimation of measurement uncertainty are set out in ISO/IEC Guide 98‑3. There
are several ways of estimating measurement uncertainty depending on the purpose of the estimation and the
available data; Eurolab TR 1[9] gives an overview of the main approaches.

This International Standard specifies a set of procedures to enable laboratories to estimate the measurement
uncertainty of their results, using an approach based on quality control results and validation data. It is structured
in a way that is applicable to analysts that do not have a thorough understanding of metrology or statistics.

NEN 7779[8] and Nordtest TR 537[10] have been used as a basis for developing this International Standard.
The approach taken is “top-down”, contrary to the mainly “bottom-up” strategy adopted in ISO/IEC Guide 98-3.

It is statistically acceptable to combine a precision estimate and the uncertainty associated with the bias into
one uncertainty measure. The sources of data for this approach are method validation and analytical quality
control. The experimental approach specified in this International Standard enables a greater coverage of the
sources of variation observed during routine use of the analytical method.

© ISO 2012 – All rights reserved  v

BS ISO 11352:2012
 BS ISO 11352:2012
INTERNATIONAL STANDARD ISO 11352:2012(E)

Water quality — Estimation of measurement uncertainty based

on validation and quality control data

1 Scope
This International Standard specifies methods for the estimation of measurement uncertainty of chemical and
physicochemical methods in single laboratories based on validation data and analytical quality control results
obtained within the field of water analysis.

NOTE 1 The principles of the estimation of uncertainty specified in this International Standard are consistent with the
principles described in ISO/IEC Guide 98-3.

In this International Standard, the quantification of measurement uncertainty relies on performance

characteristics of a measurement procedure obtained from validation and the results of internal and external
quality control.

NOTE 2 The approaches specified in this International Standard are mainly based on QUAM [11], NEN 7779 [8], Nordtest
TR 537[10], and Eurolab TR 1[9].

NOTE 3 This International Standard only addresses the evaluation of measurement uncertainty for results obtained
from quantitative measurement procedures. The uncertainties associated with results obtained from qualitative procedures
are not considered.

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/IEC Guide 98-3:2008, Uncertainty of measurement — Part 3: Guide to the expression of uncertainty in
measurement (GUM:1995)

3 Terms and definitions

For the purposes of this document, the following terms and definitions apply.

NOTE 1 The terms and definitions listed are generally reproduced without the Notes which are associated with the
terms and definitions in the respective references.

NOTE 2 The terms concerning precision data from interlaboratory trials are taken from ISO 3534-2:2006[1] because the
definitions in ISO/IEC Guide 99:2007[7] are wider than those in ISO 3534-2:2006 as they include different measurement
procedures, which is not appropriate for this International Standard.

3.1
trueness
closeness of agreement between the average of an infinite number of replicate measured quantity values and
a reference quantity value

[ISO/IEC Guide 99:2007[7], 2.14]

3.2
precision
closeness of agreement between indications or measured quantity values obtained by replicate measurements
on the same or similar objects under specified conditions

[ISO/IEC Guide 99:2007[7,] 2.15]

© ISO 2012 – All rights reserved  1

BS ISO 11352:2012
ISO 11352:2012(E)

3.3
error
measurement error
measured quantity value minus a reference quantity value

[ISO/IEC Guide 99:2007[7], 2.16]

3.4
systematic error
systematic measurement error
component of measurement error that in replicate measurements remains constant or varies in a predictable manner

[ISO/IEC Guide 99:2007[7], 2.17]

3.5
bias
measurement bias
estimate of a systematic measurement error

[ISO/IEC Guide 99:2007[7], 2.18]

3.6
random error
random measurement error
component of measurement error that in replicate measurements varies in an unpredictable manner

[ISO/IEC Guide 99:2007[7], 2.19]

3.7
repeatability conditions
observation conditions where independent test/measurement results are obtained with the same method on
identical test/measurement items in the same test or measuring facility by the same operator using the same
equipment within short intervals of time

[ISO 3534‑2:2006[1], 3.3.6]

3.8
repeatability
precision under repeatability conditions

[ISO 3534‑2:2006[1], 3.3.5]

3.9
batch
series of measurements made under repeatability conditions

3.10
intermediate precision conditions
conditions where test results or measurement results are obtained with the same method, on identical
test/measurement items in the same test or measurement facility, under some different operating condition

NOTE There are four elements to the operating condition: time, calibration, operator and equipment.

[ISO 3534‑2:2006[1], 3.3.16]

3.11
intermediate precision
precision under intermediate precision conditions

[ISO 3534‑2:2006[1], 3.3.15]

2  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

3.12
within-laboratory reproducibility
intermediate measurement precision where variations within one laboratory alone are included

3.13
reproducibility conditions
observation conditions where independent test/measurement results are obtained with the same method
on identical test/measurement items in different test or measurement facilities with different operators using
different equipment

[ISO 3534‑2:2006[1], 3.3.11]

3.14
reproducibility
precision under reproducibility conditions

[ISO 3534‑2:2006[1], 3.3.10]

3.15
uncertainty
measurement uncertainty
non-negative parameter characterizing the dispersion of the quantity values being attributed to a measurand,
based on the information used

[ISO/IEC Guide 99:2007[7], 2.26]

3.16
standard uncertainty
standard measurement uncertainty
measurement uncertainty expressed as a standard deviation

[ISO/IEC Guide 99:2007[7], 2.30]

3.17
combined standard uncertainty
combined standard measurement uncertainty
standard measurement uncertainty that is obtained using the individual standard measurement uncertainties
associated with the input quantities in a measurement model

[ISO/IEC Guide 99:2007[7], 2.31]

3.18
relative standard measurement uncertainty
standard measurement uncertainty divided by the absolute value of the measured quantity value

[ISO/IEC Guide 99:2007[7], 2.32]

3.19
target measurement uncertainty
measurement uncertainty specified as an upper limit and decided on the basis of the intended use of
measurement results

[ISO/IEC Guide 99:2007[7], 2.34]

3.20
expanded uncertainty
expanded measurement uncertainty
product of a combined standard measurement uncertainty and a factor larger than the number one

NOTE The term “factor” in this definition refers to a coverage factor.

[ISO/IEC Guide 99:2007[7], 2.35]

© ISO 2012 – All rights reserved  3

BS ISO 11352:2012
ISO 11352:2012(E)

3.21
coverage factor
number larger than one by which a combined standard measurement uncertainty is multiplied to obtain an
expanded measurement uncertainty

[ISO/IEC Guide 99:2007[7], 2.38]

4 Symbols
b bias estimated as the difference between mean measured value and an accepted reference
value

bi bias of the ith reference material respectively deviation from the complete recovery (100 %) of
the ith recovery experiment

brms root mean square of individual bias values respectively of the deviations from recovery
experiments

Di difference between the measurement result and the assigned value of the ith sample of the
interlaboratory comparison

Drms root mean square of the differences

d2 factor for the calculation of the standard deviation from the mean range R

i variable related to an observation of a series

j variable related to a source of uncertainty

J total number of sources of uncertainty

k coverage factor

nilc number of analysed interlaboratory comparison samples

nM number of measurements

np,i number of participating laboratories for sample i

nr number of reference materials

nη number of recovery experiments

R mean range

s standard deviation

sb standard deviation of the measured values of the reference material

sR,i reproducibility standard deviation from the interlaboratory comparison for sample i

s Rw standard deviation of the quality control results

U expanded uncertainty

Urel relative expanded uncertainty

uc combined standard uncertainty

uc,rel combined relative standard uncertainty

uj standard uncertainties from different sources j

4  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

uj,rel relative standard uncertainties from different sources j

uadd standard uncertainty in the concentration of the analyte added

ub standard uncertainty component associated with method and laboratory bias

uconc standard uncertainty of the concentration of the addition solution

u Cref mean standard uncertainty of the reference values or

mean standard uncertainty of the assigned values of the interlaboratory comparison samples

u Cref standard uncertainty of the reference value

u Cref ,i standard uncertainty of the assigned value of the interlaboratory sample i

u Rw standard uncertainty component for the within-laboratory reproducibility

ur,range standard uncertainty component from the range control chart (obtained under repeatability
conditions)

u R w ,bat standard uncertainty component resulting from variations between batches

u R w ,stand standard uncertainty component of the results from the standard solution which is used as
quality control sample

uV standard uncertainty component of the volume added

uV,b systematic standard uncertainty component of the volume added

uV,rep random standard uncertainty component of the volume added (obtained under repeatability
conditions)

eV,max maximum deviation of the volume from the specified value (producer information)

h recovery

5 Principle
A measurement result of a laboratory is an estimate of the value of the measurand. The quality of this estimate
depends on the inevitable uncertainty that is inherent to the measurement result. In principle, the measurement
uncertainty is a property of individual measurement results. The estimation of the measurement uncertainty
for each individual measurement result is usually not necessary, if the measurement result originates from a
controlled measurement process. In this International Standard, the measurement uncertainty is, therefore,
determined for a set of similar measurement results. Generally, it is assumed that the set of measurement
results obtained with a specific analytical method is obtained under controlled conditions. The estimation of the
measurement uncertainty applies to all of the measurement results within the set, independently of, for example,
sample matrix or analyst, provided that the measurement is carried out under a quality assurance programme.

This International Standard specifies procedures for the estimation of measurement uncertainty within the
scope of the analytical method, and generally, random and systematic errors need to be considered. The
estimation of measurement uncertainty is based on analytical quality control results and validation data which
represent the within-laboratory reproducibility, and the method and laboratory bias.

6 Procedure
The procedure for the estimation of measurement uncertainty consists of the steps shown schematically in
Figure 1. The figure gives references to appropriate clauses and sub-clauses within this International Standard.

© ISO 2012 – All rights reserved  5

BS ISO 11352:2012
ISO 11352:2012(E)

In general, the method and laboratory bias (systematic error) and the within-laboratory reproducibility (random error)
are determined independently using suitable data from method validation and analytical quality control results.

The combined measurement uncertainty, i.e. the root of the quadratic sum of the uncertainty component for the
within-laboratory reproducibility and the uncertainty component associated with method and laboratory bias,
is multiplied by a factor of 2 to obtain the expanded uncertainty at a confidence level of approximately 95 %.

If the measurement uncertainty varies significantly, depending on the matrix and/or concentration range, the
uncertainty estimation shall be made separately for each matrix and/or concentration range.

7 Preparative considerations for the estimation of measurement uncertainty

7.1 Specification of the measurement

Before starting the estimation of measurement uncertainty, it is necessary that the analyst specify the analytical
method under consideration, and the objectives and purposes of the measurement. The following list is a
minimal checklist for this specification.

The specification comprises:

— the measurand;

— the measurement procedure;

— the field of application (matrices, concentration range).

7.2 Specification of the parametric form in which the measurement uncertainty is reported
The expanded uncertainty, U, is reported either as an absolute uncertainty value or as a relative uncertainty
value. For results near the limit of quantification, the uncertainty is often found to be constant and can therefore
be expressed as an absolute value. When results are well above the limit of quantification, the uncertainty is
often proportional to the analyte concentration and can therefore be expressed as a relative value.

EXAMPLE Determination of a heavy metal (limit of quantification: 5 µg/l).

for concentration range 5 µg/l to 20 µg/l U = 1 µg/l

for concentration >20 µg/l Urel = 5 %

Usually, the measurement uncertainty is determined for a certain matrix and concentration range. In some situations,
an interpolation function may be applied between different concentration ranges (see QUAM:2000[11], E.4).

6  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

NOTE It is recognized that, while the analysis of at least six samples in the bias estimation is appropriate for the vast
number of situations, there are occasions when this is not so. The greater the number of determinations, the greater the
confidence in the estimation.

Figure 1 — Schematic procedure for the estimation of measurement uncertainty (including

references to appropriate clauses and subclauses within this International Standard)

8 Evaluation of available precision and bias data

8.1 Approach and criteria

In this International Standard, the uncertainty component for the within-laboratory reproducibility, u R w , and the
uncertainty component from method and laboratory bias, ub, form the basis for the estimation of the measurement
uncertainty. Selected validation data and analytical quality control results shall be representative of the
measurement specification as described in Clause 7:

— measurement procedure: Are all steps of analysis considered (pretreatment, is the reference sample
comparable to test samples)?

— within-laboratory reproducibility: Are all conditions of execution of the measurement procedure considered
(e.g. different operators using different equipment)?

— measurement object: Are matrix variations (e.g. drinking water, surface water, sea water, waste water) and
all possible interferences considered?

Choose, as a basis for the estimation of the uncertainty, the source of experimental information that provides
the best coverage of uncertainty contributions in actual practice. Evaluate whether information exists on missing
uncertainty contributions that are not known to be negligible.

© ISO 2012 – All rights reserved  7

BS ISO 11352:2012
ISO 11352:2012(E)

If necessary, complete the uncertainty estimation, based on validation data and quality control results, with
information from additional experiments or with estimations on the basis of existing knowledge.

If overestimation of the measurement uncertainty is acceptable, i.e. if the target measurement uncertainty is
not exceeded, the result of this experimental approach including possible overestimation is still acceptable. If
overestimation is not acceptable, a more detailed estimation shall be performed, which can encompass part of
a modelling approach described, for example, in QUAM[11].

Figure 1 shows a schematic representation of the procedure.

Uncertainty calculations can be performed using either absolute or relative values (see 7.2).

Annex B contains worked examples of the estimation of measurement uncertainty according to this
International Standard.

NOTE Measurement uncertainty can be estimated in several ways. Ideally, the estimates should lead to a statistically
identical result. If not, this can be the consequence of not covering all sources of uncertainties. In this case, the uncertainty
is underestimated. It can also occur, however, that even applying the method correctly leads to an overestimation of the
uncertainty. This high value can be the consequence of uncertainty components not associated with the measurement
method under consideration, e.g. those associated with the reference value of a reference material or the uncertainty of
the concentration of a solution used in recovery experiments.

8.2 Within-laboratory reproducibility

8.2.1 Within-laboratory reproducibility conditions

The estimation of the random variations of measurement results shall be made under the same conditions as
used when routine analysis is carried out.

Thus, it is necessary that the measurements be made under “within-laboratory conditions” (i.e. on different
days and, depending on the conditions in the respective laboratory, with different equipment and different
operators). These conditions fall between repeatability and reproducibility conditions and are referred to as
within-laboratory reproducibility in this International Standard.

These conditions are usually employed when quality control samples are analysed that are similar to test samples.

For the estimation of the uncertainty component for the within-laboratory reproducibility, u R w , three approaches
are described, see 8.2.2, 8.2.3 and 8.2.4, respectively. Examples of each approach are given in Annex B.

8.2.2 Control samples covering the whole analytical process

If stable quality control (QC) samples that cover the whole analytical process, including all sample preparation
steps, are analysed regularly using the conditions described in 8.2.1, and if these QC samples are similar in
matrix and analyte concentration levels to test samples, then the uncertainty component for the within-
laboratory reproducibility, u R w , at this concentration and for this matrix, can be estimated from the standard
deviation of these QC results (e.g. as obtained from quality control charts).

u Rw = s Rw (1)

where s R w is the standard deviation of the QC results.

A minimum number of eight measurements is required for the estimation of this uncertainty component.

NOTE It is recognized that, while the analysis of at least eight measurements is appropriate for the vast number of
situations, there are occasions when this is not so. The greater the number of measurements, the greater the confidence
in the estimation.

If the analytical method covers a broad concentration range or range of matrices, and the uncertainty component
for the within-laboratory reproducibility, u R w , varies with the concentration or matrix, it is necessary to analyse
quality control samples comprising different matrices and concentration levels.

8  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

8.2.3 Using standard solutions as quality control samples

If quality control samples with an identical matrix to test samples are not available and synthetic standard
solutions (with a matrix which differs from that of routine samples) are used, the additional uncertainty
component due to possible increased inhomogeneity of the analyte in the matrix should also be considered.

The additional uncertainty due to inhomogeneity can be estimated, e.g. from range control charts using
samples of different matrices. For the calculation of the standard deviation from the mean range, see Annex A.

Since the uncertainty component from the range control chart, ur,range, covers only the repeatability component,
it shall be combined with the uncertainty of the results from the quality control sample analysed, i.e. the
standard solutions, u R w ,stand , to obtain a reliable estimate of the within-laboratory reproducibility.

u R w = u R2 + u r2,range (2)
w ,stand

where

u R w ,stand is the uncertainty component of the results from the standard solution which is used as
quality control sample;

ur,range is the uncertainty component from the range control chart.

A minimum number of eight measurements is required for the estimation of both uncertainty contributions.

NOTE It is recognized that, while a number of at least eight measurements is appropriate for the vast number of
situations, there are occasions when this is not so. The greater the number of measurements, the greater the confidence
in the estimation.

The uncertainty component from the repeatability is partly included twice in this estimation, repeatability from
standard solutions (part of u R w ,stand ) as well as repeatability from test samples. However, in general, the
repeatability part of the uncertainty component for the within-laboratory reproducibility, u R w , is small if
measurements are performed at a concentration level well above the limit of quantification.

8.2.4 Unstable control samples

Where stable quality control samples are not available, e.g. for the determination of oxygen in water, the
uncertainty component from the repeatability can be calculated from the mean of the ranges of replicate
analyses (see Annex A). A minimum number of eight samples for the determination of ranges is required.

NOTE It is recognized that, while the analysis of at least eight samples is appropriate for the vast number of situations, there
are occasions when this is not so. The greater the number of replicate analyses, the greater the confidence in the estimation.

For the uncertainty component resulting from variations between batches, u R w ,bat , other procedures for the
estimation are required. In many cases, this component relies on scientific judgement based on the analyst’s
experience (see ISO/IEC Guide 98-3:2008, 4.3.1).

u R w = u r2,range + u R2 (3)
w ,bat

where

ur,range is the uncertainty component from the range control chart;

uR is the uncertainty component resulting from variations between batches.

w ,bat

© ISO 2012 – All rights reserved  9

BS ISO 11352:2012
ISO 11352:2012(E)

8.3 Method and laboratory bias

8.3.1 General

If possible, sources of bias should always be eliminated. ISO/IEC Guide 98-3 states that if bias is significant
and can be reliably estimated, then a measurement result should always be corrected. This is usually specified
during development of a measurement procedure.

In many cases, the observed bias can vary depending on the matrix and the concentration of the analyte. This
may be taken into account by using several matrix reference materials.

To evaluate the uncertainty associated with method and laboratory bias, ub, two components shall be estimated:

a) the bias itself (as difference from the nominal or certified reference value);

b) the uncertainty of the nominal or certified reference value.

NOTE The bias uncertainty component can be neglected if it is < u R / 3.

w

In 8.3.2, 8.3.3 and 8.3.4, respectively, three approaches to estimating the uncertainty associated with method
and laboratory bias are outlined:

— analysis of suitable reference materials;

— participation in interlaboratory comparisons;

— recovery experiments with suitable samples.

8.3.2 Analysis of suitable reference materials

Results from regular analysis of suitable reference materials can be used to estimate the measurement
uncertainty component associated with method and laboratory bias, ub. Therefore each reference material
should have been analysed in at least six batches of analyses.

NOTE 1 It is recognized that, while the analysis of the reference materials in at least six batches of analyses is
appropriate for the vast number of situations, there are occasions when this is not so. The greater the number of results,
the greater the confidence in the estimation.

The uncertainty in the reference value of the certified reference material can be obtained from the producer’s
certificate. It may be necessary to convert the uncertainty given on the certificate into a standard uncertainty,
e.g. if the uncertainty is expressed as an expanded uncertainty or as a confidence interval. For other reference
materials, the uncertainty in the reference value is taken from suitable statistical data, e.g. for material from
interlaboratory trials from the reproducibility standard deviation (see 8.3.3).

To obtain a reliable estimate of the measurement uncertainty component associated with method and
laboratory bias, ub, it is advisable to use several reference materials covering the scope of the analytical
method (different matrices and concentration levels). If only one reference material is used, the uncertainty
may be underestimated.

If several reference materials are used, different values for bias are obtained which are used to calculate
the uncertainty component brms. The uncertainty component associated with method and laboratory bias,
ub, is given by:

u b = u C2 2
+ brms (4)
ref

where

u Cref is the mean uncertainty of the reference values;

brms is the root mean square of the individual bias values, given by

10  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

∑ (bi )
2
brms = (5)
nr

in which

bi is the difference between the mean measured value and the accepted reference value of the ith
reference material;

nr is the number of reference materials.

If the individual bias values and the uncertainties of the reference values vary significantly, it can be necessary
to separately estimate uncertainties for the different cases.

If only one reference material is available, the results of analyses of this reference material are treated as the best
available estimate for the measurement uncertainty component associated with method and laboratory bias, ub.

When only one reference material is used, the uncertainty component associated with method and laboratory
bias is given by:

2
 s 
2  + u C2
ub = b +  b (6)
 n  ref
 M 
where

b is the difference between mean measured value and an accepted reference value;

sb is the standard deviation of the measured values of the reference material;

nM is the number of bias measurements on the reference material;

u Cref is the uncertainty of the reference value.

NOTE 2 If only one reference material is used, the uncertainty of the bias estimation s b / nM can also make a
significant contribution to ub and is therefore included in the equation (see Reference [12]).

8.3.3 Participation in interlaboratory comparisons

Results from interlaboratory comparisons may be used in the same way as results from analyses of reference
materials, if it is assumed that the assigned value in the interlaboratory comparison is a sufficiently good
estimate of the true value.

NOTE 1 For each certified reference material, an estimate of bias, with a mean value based on several measurements
performed on different days, can be obtained. In proficiency testing schemes, often, only single measurements are
performed on a single day. Therefore, the difference between a laboratory result and the assigned value is calculated for
different interlaboratory samples. This difference then includes contributions from both the uncertainty component
associated with method and laboratory bias, ub, and the uncertainty component for the within-laboratory reproducibility,
u R . The contributions from both components can lead to an overestimate of measurement uncertainty.
w

To determine an estimate of the bias from interlaboratory comparison results, a laboratory should have analysed
at least six different samples within one or more rounds of interlaboratory comparisons.

NOTE 2 It is recognized that, while the analysis of at least six different samples is appropriate for the vast number of situations,
there are occasions when this is not so. The greater the number of samples, the greater the confidence in the estimation.

© ISO 2012 – All rights reserved  11

BS ISO 11352:2012
ISO 11352:2012(E)

The differences, Di, between the measurement results and the assigned values for the different samples
can be both positive and negative. All difference values are used to estimate the root mean square of the
differences, Drms:

Drms =
∑ Di2 (7)
nilc

where

Di is the difference between the measurement result and the assigned value of the ith sample of the
interlaboratory comparison;

nilc is the number of interlaboratory comparison samples analysed.

If the individual differences and the uncertainties of the assigned values vary significantly, it may be necessary
to separately estimate uncertainties for the different cases.

NOTE 3 Usually the uncertainty of an assigned value of an interlaboratory comparison sample is larger than the
uncertainty of the certified reference value of a certified reference material. Thus, usually the estimated uncertainty
component is larger. In some cases, the uncertainty of the assigned value of the interlaboratory comparison sample is so
large that it cannot be used for the estimation of the uncertainty component associated with method and laboratory bias.

The mean uncertainty of the assigned values of the interlaboratory comparison samples, which have been
calculated as robust or arithmetic mean from the results of the participating laboratories (i.e. consensus value),
u Cref , is calculated as:

u Cref =
∑ u Cref ,i (8)
nilc

where

s R,i
u Cref ,i = 1, 25 ×
np,i

if the median or robust mean is used as consensus value or

s R,i
u Cref ,i =
np,i

if the arithmetic mean is used as consensus value, where

u Cref ,i is the uncertainty of the assigned value of the interlaboratory sample i;

nilc is the number of analysed interlaboratory comparison samples;

sR,i is the reproducibility standard deviation from the interlaboratory comparison for sample i;

np,i is the number of participating laboratories for sample i.

NOTE 4 As specified in ISO 13528,[5] it is necessary to introduce a factor of 1,25 if the median or a robust mean is used
for the calculation of the consensus value.

If the assigned value is not established as mentioned above, the uncertainty shall be obtained directly from the
organizer of the interlaboratory comparison scheme.

12  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

Finally, the standard uncertainty component associated with method and laboratory bias, ub, is calculated as:

2
u b = Drms + u C2 (9)
ref

where

Drms is the root mean square of the differences;

u Cref is the mean uncertainty of the assigned values of the interlaboratory comparison samples.

8.3.4 Recovery experiments

Recovery experiments, which estimate the recovery of a known amount of analyte added to a previously
analysed sample, can also be used to evaluate bias.

In this case, the uncertainty associated with method and laboratory bias, ub, consists of two components: the
difference between observed and complete recovery of the analyte, and the uncertainty in the concentration
of the analyte added.

The recovery experiments should be performed with at least six different samples of the relevant matrix.

NOTE It is recognized that, while the analysis of at least six different samples of the relevant matrix is appropriate
for the vast number of situations, there are occasions when this is not appropriate. The greater the number of recovery
experiments, the greater the confidence in the estimation.

The standard uncertainty associated with method and laboratory bias, ub, estimated from recovery experiments is:

2 2
u b = brms + u add (10)

where

brms is the root mean square of the deviations from the recovery experiments;

uadd is the uncertainty in the concentration of the analyte added.

The root mean square of the deviations from the recovery experiments, brms, is obtained from:

brms =
∑ bi2 (11)
nη

where

bi is the deviation from the complete recovery (100 %) of the ith recovery experiment or from the mean
recovery, if the results are corrected with this mean recovery;

nη is the number of recovery experiments.

If the individual bias values vary significantly, it may be necessary to separately estimate uncertainties for the
different cases.

Bias from recovery experiments consists of contributions from both uncertainty components, ub and u R w .
These contributions can lead to measurement uncertainty being overestimated. Therefore each recovery which
is used to calculate the deviation, bi, should ideally be a mean recovery based on six determinations in order
to reduce the contribution of the u R w component.

The uncertainty in the concentration of the analyte added, uadd, consists of two components: the uncertainty of
the volume added, uV, and the uncertainty in the concentration of the solution added, uconc.

© ISO 2012 – All rights reserved  13

BS ISO 11352:2012
ISO 11352:2012(E)

The uncertainty of the volume added, uV, can often be estimated from information provided by the manufacturers
of volumetric labware. The systematic and the random (repeatability) errors shall be taken into account. The
systematic error is often referred to as “maximum deviation”. Where insufficient information is available, it is
necessary that a rectangular distribution be assumed and that the systematic standard uncertainty component
of the volume, uV,b, be added, calculated as:

ε V ,max
u V ,b = (12)
3

where eV,max is the maximum deviation of the volume from the specified value (producer information).

If the temperature deviates from that specified by the manufacturer of the volumetric labware, this shall also be
taken into account (for further information, see QUAM[11]).

The random uncertainty component of the volume added, uV,rep, provided by the manufacturer is often given
as a standard deviation.

The uncertainty component of the volume added, uV, is given by:

uV = uV2,b + uV2,rep (13)

where

uV,b is the systematic uncertainty component of the volume added;

uV,rep is the random uncertainty component of the volume added (repeatability conditions).

If the solution used in recovery experiments is a certified reference material, the uncertainty of the concentration
can be obtained from the certificate. If the solution is prepared by the laboratory, the uncertainty in the
concentration shall be estimated in a suitable way (e.g. see B.3).

The uncertainty in the concentration of the analyte added, uadd, is calculated as:

u add = uV2 + u conc

2
(14)

where

uV is the uncertainty component of the volume added;

uconc is the uncertainty of the concentration of the addition solution.

Finally, the standard uncertainty component associated with method and laboratory bias, ub, is calculated
according to Equation (10).

The recovery experiment might not cover all possible causes of uncertainty. Relevant (>uc/3) contributions, e.g.
those caused by interfering substances, shall be considered separately.

9 Calculation of the combined standard uncertainty

Calculate the combined standard uncertainty, uc, from the standard uncertainties, uj, or the combined relative
standard uncertainty, uc,rel, from the relative standard uncertainties, uj,rel, respectively, of all J sources of uncertainty:

J
uc = ∑ u 2j (15)
j =1

14  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

J
u c,rel = ∑ u 2j,rel (16)
j =1

If there are no further uncertainty components, other than u R w and ub, the combined standard uncertainty is
calculated according to Equation (17).

u c = u R2 + u b2 (17)
w

10 Calculation of the expanded uncertainty

Calculate the expanded uncertainty, U, or the relative expanded uncertainty, Urel, using a coverage factor of
k = 2. This approximately corresponds to a symmetrical confidence interval of 95 %.

U = 2uc (18)

Urel = 2uc,rel (19)

11 Estimation of measurement uncertainty from reproducibility standard deviation

Measurement uncertainty can also be approximately estimated, if data from an interlaboratory comparison for
method validation as specified in ISO 5725-2[2] are available, using the reproducibility standard deviation, sR,
from the interlaboratory comparison according to ISO 21748.[6] For this purpose, sR is multiplied by a factor of
2 and the result is an estimate for the expanded uncertainty, U (for more details see ISO 21748[6]). If data from
an interlaboratory comparison are used, it is necessary that the laboratory ensure that the within-laboratory
standard deviation from replicate analyses is comparable to the repeatability standard deviation obtained in
the interlaboratory comparison. It is also necessary that the laboratory check for gross errors regarding bias.

Reproducibility standard deviations for a method from proficiency tests where no repeatability standard
deviation is stated can be used in a similar way.

12 Report
The complete report of measurement uncertainty consists of the uncertainty itself, the chosen confidence
level, and the method used for the estimation of the measurement uncertainty.

EXAMPLE Mass concentration of SO 2− [4]

4 in waste water (ISO 10304-1 ): (100 ± 8) mg/l

With the following footnote:

The measurement uncertainty was derived from results of interlaboratory trials. It represents the
expanded uncertainty and was obtained with a coverage factor of k = 2. This corresponds to a confidence
level of approximately 95 %.

© ISO 2012 – All rights reserved  15

BS ISO 11352:2012
ISO 11352:2012(E)

Annex A
(normative)

Estimation of the standard uncertainty from range control charts

The standard uncertainty can be calculated from the mean range of a range control chart using Equation (A.1):

R
u r,range = (A.1)
d2

where

R is the mean range;

d2 is taken from Table A.1 and is dependent on the number of values from which the range is
calculated.

Table A.1 — Factors for the calculation of the standard deviation from the mean range
(Source: ISO 8258[3])

No. of values from which the range Factor

is calculated d2
2 1,128
3 1,693
4 2,059
5 2,326

16  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

Annex B
(informative)

Examples of the estimation of measurement uncertainty

B.1 Example 1 — Estimation of measurement uncertainty using reference material

NOTE See 8.2.2 and 8.3.2.

B.1.1 Origin of quality control data

For quality control for the determination of orthophosphate in sea water samples, a certified reference material
is analysed in 30 batches over a period of about three months.

B.1.2 Calculation of the uncertainty component for the within-laboratory reproducibility

As the analysis of the quality control sample covers the whole analytical process except sampling, the standard
deviation of the quality control results, s R w , is equivalent to the uncertainty component for the within-laboratory
reproducibility, u R w , at this concentration level. As the results are well above the limit of quantification, the
uncertainty can be expressed as a relative value. See Table B.1.

Table B.1 — Results of the analyses of the quality control sample

Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

3−
PO 4 -P , µmol/l 2,16 2,40 2,31 2,33 2,36 2,27 2,37 2,27 2,27 2,10 2,26 2,58 2,23 2,47 2,37

Number 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

3−
PO 4 -P , µmol/l 2,39 2,30 2,26 2,42 2,67 2,36 2,37 2,36 2,30 2,50 2,17 2,43 2,35 2,16 2,30

From these quality control data the following values are calculated:

Mean value: c = 2,336 µmol/l

Standard deviation: s R w = 0,122 µmol/l

From this the relative uncertainty component for the within-laboratory reproducibility, u R w ,rel , can be calculated as:

s Rw 0,122
u R w ,rel = = = 0, 052 1 = 5, 21 %
x 2, 336

B.1.3 Calculation of the uncertainty component associated with method and laboratory bias
Since only one reference material is available, it is necessary that three components be considered for the
estimation of the uncertainty component associated with method and laboratory bias, ub:

a) the difference between the mean measured value and an accepted reference value, bias;

b) the standard deviation of the measured values of the reference material, sb;

c) the uncertainty of the reference value, u Cref .

© ISO 2012 – All rights reserved  17

BS ISO 11352:2012
ISO 11352:2012(E)

The certificate accompanying the reference material provides the following information:

— certified value, Cref, and confidence interval of orthophosphate-P: (2,43 ± 0,41) µmol/l;

— the confidence interval (±0,41) represents three standard deviations derived from the interlaboratory
testing for the generation of the certified reference value.

From this it follows that the uncertainty of the reference value, u Cref is:

0, 41
u Cref = µmol/l = 0,137 µmol/l
3

The relative uncertainty of the reference value, u Cref ,rel , is given by:

u Cref 0,137
u Cref ,rel = = = 0, 056
C ref 2, 43

The bias, b, is calculated using the mean of the measured values of the control sample x (see B.1.2):

b = x − C ref = 2, 336 − 2, 43 = −0, 094

The relative bias, brel, is given by:

x − C ref 2,336 − 2,43

brel = = = − 0,0387
C ref 2,43

The coefficient of variation of the measured values of the reference material, CV,b, is equal to the relative
uncertainty component for the within-laboratory reproducibility, u R w ,rel , (see B.1.2):

CV ,b = u R w ,rel = 0, 052 1

and

nM = 30

Using these data, the relative uncertainty component associated with method and laboratory bias, ub,rel is
calculated as:

2 2
 CV ,b   0,052 1 
u b,rel = 2
brel +
 n
 + u C2 ,rel =

( −0,038 7 ) 2 +  2
 + 0,056 2 = 0,068 9 = 6,89 %
 M 
ref
 0
3 

B.1.4 Calculation of the relative combined uncertainty, uc,rel

The relative combined uncertainty is calculated as:

u c,rel = u R2 + u b2,rel = 0,052 12 + 0,068 9 2 = 0,086 4 = 8,64 %

w ,rel

B.1.5 Calculation of the relative expanded uncertainty, Urel

The relative expanded uncertainty is calculated using a coverage factor k = 2:

Urel = kuc,rel

Urel = 2 × 0,086 4 = 0,172 8 ≈ 17,3 %

18  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

B.2 Example 2 — Estimation of measurement uncertainty based on data from pro-

ficiency tests
NOTE See 8.2.2 and 8.3.3.

B.2.1 Origin of quality control data

The total phosphorus content of sea water samples is routinely determined. For interlaboratory quality control
the laboratory regularly participates in proficiency tests. In the period from 2003-07 to 2004-04, the results
listed in Table B.2 were obtained.

Table B.2 — Results from proficiency tests

Proficiency u Cref ,i,rel

cass c Di Di,rel sR,i,rel np,i
test
µmol/l µmol/l µmol/l % % %
1 14,080 14,253 0,173 1,23 3,1 28 0,73
2 6,250 6,752 0,502 8,03 4,8 28 1,13
3 2,820 2,582 -0,238 -8,44 7,6 28 1,80
4 5,243 5,414 0,171 3,26 5,3 35 1,12
5 3,600 3,780 0,180 5,00 6,9 35 1,46
6 1,838 1,913 0,075 4,08 8,4 35 1,77
cass assigned value of the interlaboratory comparison sample

c mean value of the laboratory’s test results

Di difference of the mean value c and the assigned value cass

c − c ass
Di,rel relative difference as a percentage: D i,rel = × 100
c ass

sR,i,rel reproducibility standard deviation from the interlaboratory comparison for sample i
np,i number of participating laboratories

uC relative uncertainty of the assigned value of the interlaboratory sample

ref ,i,rel

Since the robust mean was used as the consensus mean, the single uncertainties are calculated as

s R,i,rel
u Cref ,i,rel = 1, 25 ×
np,i

B.2.2 Calculation of the uncertainty component for the within-laboratory reproducibility

As the results are well above the limit of quantification, the uncertainty can be expressed as a relative value.
The relative uncertainty component for the within-laboratory reproducibility, is calculated as the coefficient of
variation from the analyses of a stable control sample [r(P) = 8 mg/l] used in a mean control chart:

mean value from 20 batches ρ = 8, 03 mg/l

standard deviation s = 0,352 mg/l

coefficient of variation CV = u R w ,rel = 0, 043 8 = 4, 38 %

© ISO 2012 – All rights reserved  19

BS ISO 11352:2012
ISO 11352:2012(E)

B.2.3 Calculation of the uncertainty component associated with method and laboratory bias
For the determination of the uncertainty component associated with method and laboratory bias, ub, two
components shall be determined: root mean square of the differences, Drms, and the mean uncertainty of the
assigned values u Cref .

The relative root mean square of the differences Drms,rel is calculated as

Drms,rel =
∑ Di2,rel
nilc

0,012 3 2 + 0,080 3 2 + 0,084 4 2 + 0,032 6 2 +0,050 0 2 + 0,040 8 2

Drms,rel = = 0,056 2 = 5,62 %
6

The mean uncertainty of the assigned values is calculated as

u Cref =
∑ u Cref ,i
nilc

For calculation of the relative mean uncertainty of the assigned values the u Cref ,i,rel values listed in Table B.2
are divided by a factor of 100 (as tabulated values are expressed as percentages)

0,007 3 + 0,011 3 + 0,018 0 + 0,011 2 + 0,014 6 + 0,017 7

u Cref ,rel = = 0,013 4 = 1,34 %
6

Finally the relative uncertainty component associated with method and laboratory bias, ub,rel, is calculated as:

2
u b,rel = Drms,rel + u C2 = 0,056 2 2 + 0,013 4 2 = 0,057 8 = 5,78 %
ref ,rel

B.2.4 Calculation of the relative combined uncertainty, uc,rel

The relative combined uncertainty is calculated as:

u c,rel = u R2 + u b2,rel = 0,043 8 2 + 0,057 8 2 = 0,072 5 = 7,25 %

w ,rel

B.2.5 Calculation of the relative expanded uncertainty, Urel

The relative expanded uncertainty is calculated using a coverage factor k = 2:

Urel = kuc,rel

U rel = 2 × 0,072 5 = 0,145 = 14,5 %

B.3 Example 3 — Estimation of measurement uncertainty using a standard solution

as quality control sample and recovery experiments
NOTE See 8.2.3 and (8.3.4.

B.3.1 Origin of quality control data

The herbicide triflusulfuron-methyl is routinely determined in water samples. A quality control sample covering
the whole analytical process is not available. Therefore, a standard solution is used as quality control sample
for operation of a control chart. For quality control, recoveries are determined in each batch of analyses.

20  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

For this purpose, spiked water samples (triflusulfuron-methyl concentration, r = 0,1 µg/l) are analysed. The
triflusulfuron-methyl results of the real samples are corrected using averaged recoveries.

B.3.2 Calculation of the uncertainty component for the within-laboratory reproducibility

For calculation of the uncertainty component for the within-laboratory reproducibility, u R w , the results from the
following quality control measurements are required:

a) measurement of a stable quality control standard in organic solvent in each batch of analyses (triflusulfuron-
methyl concentration, r = 0,50 µg/ml).

b) repeated determination of spiked water samples (triflusulfuron-methyl concentration, r = 0,1 µg/l) for range
control chart.

As the results are well above the limit of quantification the uncertainty can be expressed as a relative value.

B.3.2.1 Data for calculation of the uncertainty component for the within-laboratory reproducibility

See Table B.3.

Table B.3 — Results of stable quality control standard, assigned value: r = 0,50 µg/ml

r
Number of measurement
µg/ml
1 0,49
2 0,50
3 0,52
4 0,48
5 0,49
6 0,51
7 0,51
8 0,54
9 0,48
10 0,49
Mean value ρ 0,501

Standard deviation s 0,019 1

Coefficient of variation, CV = u R w ,stand,rel 0,038 2

For the operation of a range control chart, spiked water samples are analysed in duplicate with each batch of
samples. The relative range, R j,rel , which is used as quality control value for the range control chart, is
calculated as:

x i,max − x i,min
R j,rel =
xj

where j is the jth batch of i replicates.

See Table B.4 for results from the range control chart (d2 = 1,128 for duplicate measurements).

© ISO 2012 – All rights reserved  21

BS ISO 11352:2012
ISO 11352:2012(E)

Table B.4 — Results from the range control chart (d2 = 1,128)

Number of batch R j,rel

1 0,154 4
2 0,168 9
3 0,104 6
4 0,044 5
5 0,116 2
6 0,030 3
7 0,113 5
8 0,025 0
9 0,050 1
10 0,025 5

Mean value Rrel 0,083 3

Coefficient of variation,
R 0,073 8
CV = u r,range,rel = rel
d2

B.3.2.2 Calculation of the uncertainty component for the within-laboratory reproducibility

For calculation of the relative within-laboratory reproducibility, u R w ,rel , the relative uncertainty from the mean control
chart, u R w ,stand,rel , is combined with the relative uncertainty component from the range control chart, u r,range,rel .

u R w ,rel = u R2 + u r2,range,rel = 0, 038 2 2 + 0, 073 8 2 = 0, 083 1 = 8, 31%

w ,stand,rel

B.3.3 Calculation of the uncertainty component associated with method and laboratory bias
For the determination of recoveries, spiked water samples (triflusulfuron-methyl concentration, r = 0,1 µg/l) are
analysed in each batch of analyses. The uncertainty component associated with method and laboratory bias,
ub, is estimated from these recovery experiments. For the estimation of ub, it is necessary that two components
be determined:

a) the root mean square of the deviations from the recovery experiments, brms;

b) the uncertainty in the concentration of the analyte added, uadd.

B.3.3.1 Determination of root mean square of the deviations from the recovery experiments, brms

See Table B.5.

22  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

Table B.5 — Results from 10 recovery experiments and calculation of bias from mean recovery, η

hi η
bi,rel
% %
95,1 0,066 1
84,5 -0,052 7
98,3 0,102
86,3 -0,032 5
85,4 -0,042 6
89,2
92,8 0,040 4
88,0 -0,013 5
83,0 -0,069 5
83,8 -0,060 5
95,0 0,065 0

The relative bias from the mean recovery is calculated as:

ηi −η
bi,rel =
η

where hi is the recovery of the single measurement.

NOTE Here bi,rel refers to the mean recovery and not to the target value 100 %, because for calculation of results an
equation with a correction factor derived from averaged recoveries is used.

The relative root mean square of the deviations from the recovery experiments, brms,rel is calculated as:

brms,rel =
∑ (bi,rel ) 2
nη

brms,rel =

0, 066 12 + ( − 0,052 7) 2 + 0,102 2 + ( − 0,032 5) 2 + ( − 0,042 6) 2 + 0,040 4 2 + ( − 0,013 5) 2 + ( − 0,069 5) 2 + ( − 0,060 5) 2 + 0, 065 0 2
10
brms,rel = 0,059 1 = 5,91 %

B.3.3.2 Determination of the uncertainty in the concentration of the analyte added, uadd

The analyte-added uncertainty, uadd, consists of two components:

a) the uncertainty of the concentration of the addition solution, uconc;

b) the uncertainty component of the volume added, uV.

B.3.3.2.1 Determination of the uncertainty of the concentration of the addition solution, uconc

The spiking solutions are prepared in the laboratory. The uncertainty due to weighing is considered to be
negligible. For the calculation of the uncertainty of the concentration of the addition solution, uconc, the
uncertainties of the volumetric glassware are used.

For the preparation of the spiking solutions, 3 (20 ml) volumetric flasks and 2 (1 ml) pipettes are used. The
manufacturer gives details about the maximum deviation of the volumes (uV,b,i), which are reported to be 0,2 %
for the volumetric flasks and 0,7 % for the pipettes.

© ISO 2012 – All rights reserved  23

BS ISO 11352:2012
ISO 11352:2012(E)

For intra-laboratory quality control analyses, the volumetric instruments are checked by weighing and
the random uncertainty component (uV,rep) is calculated as a standard deviation. The random uncertainty
component for the volumetric flasks is 0,045 % and for the pipettes is 0,14 %.

Since there is no other information available regarding the deviation of the volumes, a rectangular distribution
is assumed and the systematic uncertainty component of the volume added, uV,b is calculated as:

ε V ,max
u V ,b =
3

The relative uncertainty of the concentration of the addition solution, uconc,rel is calculated by combining the
systematic and the random uncertainty components:

u conc,rel = ∑ uV2,b,i,rel + ∑ uV2,rep,i,rel

2 2
 0, 002   0, 007  2 2
u conc,rel =   ×3 +   × 2 + ( 0, 000 45 ) × 3 + ( 0, 001 4 ) × 2 = 0, 006 4 = 0, 64 %
 3   3 

B.3.3.2.2 Determination of the uncertainty of the volume added, uV

The uncertainty of the volume added, uV, consists of two components,

a) the systematic uncertainty component of the volume added, uV,b;

b) the random uncertainty component of the volume added, uV,rep.

B.3.3.2.2.1 Determination of the systematic uncertainty component of the volume added, uV,b

The spiking solution is added using a microlitre syringe. The maximum variation reported by the manufacturer
for the volume is 1 %. Since there is no other information available, a rectangular distribution is assumed and
the relative systematic uncertainty component of the volume added, uV,b,rel, is calculated as:

ε V ,max 0,01
uV ,b,rel = = = 0,005 8 = 0,58 %
3 3

B.3.3.2.2.2 Determination of the random uncertainty component of the volume added, uV,rep

The component uV,rep is determined by repeated weighing (see Table B.6) of water, which is pipetted using the
same or similar microlitre syringe.

B.3.3.2.2.3 Calculation of the uncertainty of the volume added, uV

For calculation of the relative uncertainty of the volume added, uV,rel, the systematic and the random uncertainty
components of the volume added, uV,b,rel and uV,rep,rel are combined as follows:

uV ,rel = uV2,b,rel + uV2,rep,rel = ( 0,005 8 ) 2 + ( 0,002 7 ) 2 = 0,006 4 = 0,64%

B.3.3.2.3 Determination of the uncertainty in the concentration of the analyte added, uadd

For calculation of the relative uncertainty in the concentration of the analyte added, uadd,rel, the uncertainties
uV,rel (see B.3.3.2.2.3) and uconc,rel (see B.3.3.2.1) are combined as follows:

2
u add,rel = uV2,rel + u conc,rel
2
= (0, 006 4) 2 + ( 0, 006 4 ) = 0, 009 1 = 0, 91 %

24  © ISO 2012 – All rights reserved

 BS ISO 11352:2012
ISO 11352:2012(E)

B.3.3.3 Calculation of the uncertainty component associated with method and laboratory bias, ub

For calculation of the relative uncertainty component associated with method and laboratory bias, ub,rel, the
relative root mean square of the deviations from the recovery experiments, brms,rel (see B.3.3.1), and the
relative uncertainty in the concentration of the analyte added, uadd,rel, are combined as follows:

2
u b,rel = brms,rel 2
+ u add,rel = ( 0,059 1) 2 + (0,009 1) 2 = 0,059 8 = 5,98 %

Table B.6 — Standard deviation of the pipetted volume

Mass
Number of measurement
g
1 0,249 6
2 0,249 0
3 0,250 2
4 0,251 0
5 0,249 8
6 0,250 2
7 0,248 9
8 0,249 4
9 0,250 4
Mean value m 0,249 8
Standard deviation s 0,000 686
Coefficient of variation, CV = uV,rep,rel 0,002 7 = 0,27 %

B.3.4 Calculation of the relative combined uncertainty, uc,rel

The relative combined uncertainty is calculated as:

u c,rel = u R2
w ,rel
+ u b2,rel = ( 0,083 1) 2 + ( 0,059 8 ) 2 = 0,102 4 = 10,24 %

B.3.5 Calculation of the relative expanded uncertainty, Urel

The relative expanded uncertainty is calculated using a coverage factor k = 2:

Urel = kuc,rel

Urel = 2 × 10,24 % ≈ 20 %

© ISO 2012 – All rights reserved  25

BS ISO 11352:2012
ISO 11352:2012(E)

Bibliography

[1] ISO 3534-2:2006, Statistics — Vocabulary and symbols — Part 2: Applied statistics

[2] ISO 5725-2, Accuracy (trueness and precision) of measurement methods and results — Part 2: Basic
method for the determination of repeatability and reproducibility of a standard measurement method

[3] ISO 8258, Shewhart control charts

[4] ISO 10304-1, Water quality — Determination of dissolved anions by liquid chromatography of ions —
Part 1: Determination of bromide, chloride, fluoride, nitrate, nitrite, phosphate and sulfate

[5] ISO 13528, Statistical methods for use in proficiency testing by interlaboratory comparisons

[6] ISO 21748, Guidance for the use of repeatability, reproducibility and trueness estimates in measurement
uncertainty estimation

[7] ISO/IEC Guide 99:2007, International vocabulary of metrology — Basic and general concepts and
associated terms (VIM)

[8] NEN 7779, Milieu — Meetonzekerheid [Environment — Measurement uncertainty]

[9] Eurolab TR 1, Measurement uncertainty revisited: Alternative approaches to uncertainty evaluation.

Brussels: Eurolab, 2007

[10] Nordtest TR 537, Handbook for calculation of measurement uncertainty in environmental laboratories,
2nd edition, Magnusson, B., Näykki, T., Hovind, H., Krysell, M. Espoo: Nordtest, 2004. Available
(viewed 2011-12-22) at: http://www.nordtest.info/images/documents/nt-technical-reports/NT TR
537_Handbook for calculation of measurement uncertainty in environmental laboratories - edition 2_
Nordtest Technical Report.pdf

[11] QUAM, EURACHEM/CITAC Guide: Quantifying uncertainty in analytical measurement, second edition,
Ellison, S.L.R., Rosslein, M., A Williams, A., editors. 2000. Available (viewed 2011-12-22) at: http://
www.measurementuncertainty.org/pdf/QUAM2000-1.pdf

[12] Magnusson, B., Ellison, S.L.R. Treatment of uncorrected measurement bias in uncertainty estimation
for chemical measurements. Anal. Bioanal. Chem. 2008, 390, pp. 201-213

26  © ISO 2012 – All rights reserved

BS ISO 11352:2012
BS ISO 11352:2012
ISO 11352:2012(E)

ICS 13.060.45
Price based on 26 pages

© ISO 2012 – All rights reserved

This page deliberately left blank
NO COPYING WITHOUT BSI PERMISSION EXCEPT AS PERMITTED BY COPYRIGHT LAW

British Standards Institution (BSI)

BSI is the national body responsible for preparing British Standards and other
standards-related publications, information and services.
BSI is incorporated by Royal Charter. British Standards and other standardization
products are published by BSI Standards Limited.

About us Revisions
We bring together business, industry, government, consumers, innovators Our British Standards and other publications are updated by amendment or revision.
and others to shape their combined experience and expertise into standards We continually improve the quality of our products and services to benefit your
-based solutions. business. If you find an inaccuracy or ambiguity within a British Standard or other
The knowledge embodied in our standards has been carefully assembled in BSI publication please inform the Knowledge Centre.
a dependable format and refined through our open consultation process.
Organizations of all sizes and across all sectors choose standards to help Copyright
them achieve their goals. All the data, software and documentation set out in all British Standards and
other BSI publications are the property of and copyrighted by BSI, or some person
Information on standards or entity that owns copyright in the information used (such as the international
We can provide you with the knowledge that your organization needs standardization bodies) and has formally licensed such information to BSI for
to succeed. Find out more about British Standards by visiting our website at commercial publication and use. Except as permitted under the Copyright, Designs
bsigroup.com/standards or contacting our Customer Services team or and Patents Act 1988 no extract may be reproduced, stored in a retrieval system
Knowledge Centre. or transmitted in any form or by any means – electronic, photocopying, recording
or otherwise – without prior written permission from BSI. Details and advice can
Buying standards be obtained from the Copyright & Licensing Department.
You can buy and download PDF versions of BSI publications, including British
and adopted European and international standards, through our website at Useful Contacts:
bsigroup.com/shop, where hard copies can also be purchased. Customer Services
If you need international and foreign standards from other Standards Development Tel: +44 845 086 9001
Organizations, hard copies can be ordered from our Customer Services team. Email (orders): [email protected]
Email (enquiries): [email protected]
Subscriptions
Subscriptions
Our range of subscription services are designed to make using standards
Tel: +44 845 086 9001
easier for you. For further information on our subscription products go to
Email: [email protected]
bsigroup.com/subscriptions.
With British Standards Online (BSOL) you’ll have instant access to over 55,000 Knowledge Centre
British and adopted European and international standards from your desktop. Tel: +44 20 8996 7004
It’s available 24/7 and is refreshed daily so you’ll always be up to date. Email: [email protected]
You can keep in touch with standards developments and receive substantial
Copyright & Licensing
discounts on the purchase price of standards, both in single copy and subscription
format, by becoming a BSI Subscribing Member. Tel: +44 20 8996 7070
Email: [email protected]
PLUS is an updating service exclusive to BSI Subscribing Members. You will
automatically receive the latest hard copy of your standards when they’re
revised or replaced.
To find out more about becoming a BSI Subscribing Member and the benefits
of membership, please visit bsigroup.com/shop.
With a Multi-User Network Licence (MUNL) you are able to host standards
publications on your intranet. Licences can cover as few or as many users as you
wish. With updates supplied as soon as they’re available, you can be sure your
documentation is current. For further information, email [email protected].

BSI Group Headquarters

389 Chiswick High Road London W4 4AL UK