SIST-TP CEN/TR 16176:2012

SLOVENSKI STANDARD
01-februar-2012
.DUDNWHUL]DFLMDRGSDGNRY5HãHWDOQHPHWRGH]DHOHPHQWQRVHVWDYR]UHQWJHQVNR
IOXRUHVFHQþQRVSHNWURPHWULMRQDNUDMXVDPHP
Characterization of waste - Screening methods for elemental composition by X-ray
fluorescence spectrometry for on-site verification
Charakterisierung von Abfällen - Anwendung von Screening-Verfahren bei der Vor-Ort-
Prüfung - Bestimmung der elementaren Zusammmensetzung mittels
Röntgenfluoreszenzspektrometrie
Caractérisation des déchets - Méthodes de dépistage pour la détermination de la
composition élémentaire par spectrométrie à fluorescence de rayons X pour les
vérifications in-situ
Ta slovenski standard je istoveten z: CEN/TR 16176:2011
ICS:
13.030.01 Odpadki na splošno Wastes in general
2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.

TECHNICAL REPORT
CEN/TR 16176
RAPPORT TECHNIQUE
TECHNISCHER BERICHT
November 2011
ICS 13.030.01
English Version
Characterization of waste - Screening methods for elemental
composition by X-ray fluorescence spectrometry for on-site
verification
Caractérisation des déchets - Méthodes de dépistage pour Charakterisierung von Abfällen - Anwendung von
la détermination de la composition élémentaire par Screening-Verfahren bei der Vor-Ort-Prüfung - Bestimmung
spectrométrie à fluorescence de rayons X pour les der elementaren Zusammmensetzung mittels
vérifications in situ Röntgenfluoreszenzspektrometrie

This Technical Report was approved by CEN on 3 October 2011. It has been drawn up by the Technical Committee CEN/TC 292.

CEN members are the national standards bodies of Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia,
Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland,
Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Switzerland and United Kingdom.

EUROPEAN COMMITTEE FOR STANDARDIZATION
COMITÉ EUROPÉEN DE NORMALISATION

EUROPÄISCHES KOMITEE FÜR NORMUNG

Management Centre: Avenue Marnix 17, B-1000 Brussels
© 2011 CEN All rights of exploitation in any form and by any means reserved Ref. No. CEN/TR 16176:2011: E
worldwide for CEN national Members.

Contents Page
Foreword .3
Introduction .4
1 Purpose .5
2 Description of the XRF technique .5
2.1 General .5
2.2 Principle of XRF .5
2.3 Interferences .6
2.4 Measurement .7
2.5 Calibration/evaluation .7
2.6 Validation .8
3 Overview XRF applications .9
4 Influence of the sample preparation on the result .9
5 Evaluation of the XRF screening technique . 13
6 Robustness study: description and results . 15
6.1 General . 15
6.2 Technical description of the instruments . 15
6.3 Description of the selected samples and their characterisation . 16
6.4 Results of the field trial . 17
6.4.1 Defining performance criteria . 17
6.4.2 Evaluation of the repeatability, reproducibility and accuracy . 18
6.4.3 Influence of the sample pretreatment . 19
6.4.4 Evaluation of false positive / false negative results. 22
6.4.5 Limit of detection . 22
6.4.6 General evaluation of the portable XRF instruments. 22
6.4.7 Conclusions of the robustness study . 22
7 Conclusions . 23
Annex A (informative) Pre-normative robustness study . 24
Annex B (informative) Summary of EPA report on XRF technologies for measuring trace elements
in soil and sediment . 38
Bibliography . 42

Foreword
This document (CEN/TR 16176:2011) has been prepared by Technical Committee CEN/TC 292
“Characterization of Waste”, the secretariat of which is held by NEN.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. CEN [and/or CENELEC] shall not be held responsible for identifying any or all such patent rights.
Introduction
Although bench-top instruments generally provide much more conclusive results, hand-held XRF instruments
are becoming an interesting screening tool for a wide range of applications. Their portability and their ability to
identify, characterise and also analyse a wide range of elements rapidly, along with the fact that little technical
expertise is needed to operate them, make the hand-held XRF instruments very useful. The recent
developments in the XRF technology tends to create hand-held instruments with performance levels
approaching bench top equipment. Some years ago, hand-held instruments required the use of radioactive
materials to provide a source of X-rays, resulting in very stringent regulatory demands. The development of
miniaturised low-power X-ray tubes overcomes these problems and provides new opportunities for the hand-
held instruments. Recent advances in the improvement of the detector efficiency led to a significant decrease
in the detection limits for hand-held systems compared to the older ones. Due to the required compact
configuration for hand-held XRF systems only energy dispersive X-ray fluorescence (EDXRF) are on the
market. On the other hand wavelength dispersive XRFs (WDXRF) are generally more laborious.
The use of the XRF technique in field screening trials can provide a number of benefits compared to the
traditional laboratory techniques. On-site analyses ensure a fast turnaround between the measurement itself
and the availability of data results. Sample preparation is frequently unnecessary or will be limited. Screening
can gain a large sample data set on a short time frame, but that can be at the expense of the accuracy and
precision. When better accuracy is required confirmative analysis has to be performed. This approach will
surely result in a significant reduction of analysis time and costs.
This report focuses on hand-held XRF instruments, although portable bench-top instruments are also on the
market for this type of application. Whenever portable instruments are specifically addressed in this report,
both types of instruments can be considered.
1 Purpose
In the framework of the EU Directive 99/31/EC on the landfill of waste and the EU Directive 2000/76/EC on the
incineration of waste there is a growing need for fast, easy-to-handle screening tools. In this respect, low
costs, fast analyses, control of truck loads and yes/no-acceptance decisions are relevant criteria. The X-ray
fluorescence (XRF) technique meets these requirements as a screening tool for on-site verification on the
landfill and for entrance control on the incineration plants.
Recent developments of the XRF technology have made this technique a method of choice for on-site
analysis, namely miniaturisation of the XRF system (X-ray tube), the optimisation of the calibration
programmes and the improvement of the detectors. Therefore, a state-of-the-art document on the current
progress of the XRF technology and instruments available for on-site analysis shall support the key
arguments, dealing with the pro‟s and contra‟s, and the performance of these systems to be expected.
The XRF standard EN 15309, is validated for Na, Mg, Al, Si, P, S, Cl, K, Ca, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn,
As, Se, Br, Rb, Sr, Y, Zr, Nb, Mo, Ag, Cd, Sn, Sb, Te, I, Cs, Ba, Ta, W, Hg, Tl, Pb, Bi, Th and U, and
describes in the informative annex the procedures for hand-held XRF systems together with the
portable/transportable systems (placed in mobile labs). Although XRF can analyse a broad range of elements,
the main focus of this document is on the series of elements that is also being covered by EN 15309. Of that
series the following elements are related to the landfill directive: As, Ba, Cd, Cr, Cu, Hg, Mo, Ni, Pb, Sb, Se,
Zn.
The information in this document will be useful in all cases in which on-site determination of the elemental
compositions of waste is needed and hand-held instrumentation is therefore used. These cases may include,
beside landfills and incineration plant, also waste treatment plants, contaminations soil sites and controls of
transports of waste.
2 Description of the XRF technique
2.1 General
X-ray fluorescence spectrometry is a fast and reliable method for the analysis of the total content of certain
elements within different matrices. The quality of the results obtained depends very closely on the type of
instrument used, e.g. hand-held, bench top. When selecting a specific instrument several factors have to be
considered, such as the matrices to be analysed, elements to be determined, detection limits required and the
measuring time. The quality of the results depends on the element to be determined and on the surrounding
matrix, together with the applied sample preparation method, and the heterogeneity of the test sample.
2.2 Principle of XRF
An electron can be ejected from its atomic orbital by the absorption of a light wave (photon) of sufficient
energy [1]. The energy of the photon (h ) must be greater than the energy with which the electron is bound to
the nucleus of the atom. When an inner orbital electron is ejected from an atom, an electron from a higher
energy level orbital will be transferred to the lower energy level orbital. During this transition a photon maybe
emitted from the atom. This fluorescent light is called the characteristic X-ray of the element (Figure 1). The
energy of the emitted photon will be equal to the difference in energies between the two orbitals occupied by
the electron making the transition. Because the energy difference between two specific orbital shells, in a
given element, is always the same (i.e. characteristic of a particular element), the photon emitted when an
electron moves between these two levels, will always have the same energy. Therefore, by determining the
energy (wavelength) of the X-rays (photon) emitted by a particular element, it is possible to determine the
identity of that element.
For a particular energy (wavelength) of fluorescent light emitted by an element, the number of photons per unit
time (generally referred to as peak intensity or count rate) is related to the amount of that analyte in the
sample. The counting rates for all detectable elements within a sample are usually calculated by counting, for
a set amount of time, the number of photons that are detected for the various analytes‟ characteristic X-ray
energy lines. It is important to note that these fluorescent lines are actually observed as peaks with a
semi-Gaussian distribution depending on the resolution of modern detector technology. Therefore, by
determining the energy of the X-ray peaks in a sample‟s spectrum, and by calculating the count rate of the
various elemental peaks, it is possible to qualitatively establish the elemental composition of the samples and
to quantitatively measure the concentration of these elements.

Key
A excitation: X-rays from X-ray source
B fluorescence: Characteristic X-ray
Figure 1 — Principle of XRF
The basic configuration of an EDXRF system consists of an excitation source and a detector, coupled to data
processing unit, as shown in Figure 2.

Key
A X-ray source
B detector
Figure 2 — Basic configuration of an XRF system
2.3 Interferences
Interferences in X-ray fluorescence spectrometry are due to spectral line overlaps, matrix effects, spectral
artefacts and particle size or mineralogical effects.
Spectral line overlaps occur when an analytical line cannot be resolved from the line of a different element.
Corrections for these interferences are made using the algorithms provided with the software.
Matrix effects occur when the X-ray fluorescence radiation from the analyte element is absorbed or enhanced
by other elements in the sample before it reaches the detector. In the case of complex matrices these effects
generally have to be corrected.
Spectral artefacts e.g. escape peaks, sum peaks, pulse pile up lines, dead time, Bremsstrahlung correction,
are accounted for by the provided software. Spectral artefacts differ for energy dispersive and wavelength
dispersive XRF spectrometry.
Particle size effects can be reduced by milling the sample, and both particle size and mineralogical effects can
be eliminated by preparing bead samples. It is vital for quantitative analysis that the same sample preparation
procedure is applied to both the standards and the samples to be analysed.
2.4 Measurement
All X-ray spectrometers are supplied with a spectrometer software programme to operate the instrument. The
software packages are manufacturer depended and contain two major modules:
analytical measurement programme for data collection. This module controls the measurement of a
sample using a certain set of measurement parameters e.g. tube setting (kV, mA), targets and crystals,
detectors, measurement times. The analytical programme is always linked to a selected evaluation and
calibration programme. Actually, the same measurement conditions have to be applied for both the
standards of the calibration curve and the samples. Because in screening analysis the measurements will
be performed with the predefined analytical programmes, no further detailed descriptions will be given of
the analytical measurement parameters. Follow the manufacturer‟s instruction for further operation and
handling of the analytical software package;
evaluation programme for data processing. This module converts the measured intensities of the different
element lines to elemental concentrations taking all corrections into account. There are various types of
evaluation programmes available and each manufacturer has set up his own programmes for data
processing based on the XRF principles.
Sensitivity, instrumental detection limits and precision are instrument dependent and should therefore be
investigated and established for each individual analyte line on that particular instrument, and, if relevant, as a
function of matrix type and sample preparation procedure.
The XRF systems (hand-held as well as bench top) for screening purposes are all provided with precalibrated
measurement programmes, setup by the manufacturer, and therefore the user does not need to calibrate the
system itself. The user may be able to calibrate the XRF system itself, but it is not a requirement. It should be
stressed out that the user needs to verify the calibration programme using reference samples in order to gain
insights in the expected accuracy and precision of the XRF calibration programme. More information about the
validation will be described in 2.6.
2.5 Calibration/evaluation
For calibration purposes the measurement of analyte lines of samples of known composition is needed. The
basic equation implies a linear relationship between the intensity and the concentration.
c a a I (1)
i i,0 i,1 i
where
c is the concentration of the element of interest, expressed as mg/kg or percentage dry matter;
i
a is the offset of the calibration curve;
i,0
a is the slope of the calibration curve;
i,1
I is the net intensity of the element of interest, expressed as counts per second.
i
Matrix effects have to be taken into account in X-ray spectrometry according to the following equation:
c (a a I ) M (2)
i i,0 i,1 i
where
M is the correction term due to the matrix effects.

Different procedures for correcting matrix effects may be used according to the analytical accuracy required:
the scattered radiation method is based on the principle that the intensities of the analyte line and of the
Compton line are affected in the same proportion due to the overall mass absorption coefficient of the
sample. This linear relationship holds when all analytes are at low concentrations (trace elements) and
their absorption coefficients are not affected by an adjacent absorption edge. In this case an internal
Compton correction can be used. Beside that, a correction method using the Compton intensity with mass
absorption coefficients (MAC) is also applicable. In this method, the intensities of the major elements are
measured to apply a jump edge correction for the analysed trace elements;
correction using the fundamental parameter approach;
correction using theoretical correction coefficients (alphas) taking basic physical principles, instrumental
geometry etc. into account;
correction using empirical correction coefficients (alphas) based on regression analysis of standards with
known elemental concentrations.
For more information about the various calibration procedures is referred to EN 15309.
2.6 Validation
Prior to analysis of a sample, the available pre-calibrated analytical method has to be validated. The validation
can be performed on different levels. First of all, and this is also a requirement when performing screening
analysis, a verification of the calibration programme using a reference material with known composition has to
be carried out on a regular base in order to follow-up the stability and drift of the system. Secondly, control
analyses can be performed by using reference samples with a similar composition as the samples under
investigation to gain insights in the expected measurement uncertainty. If no reference materials with a
comparable matrix are available, only a qualitative analysis with indicative concentration values can be
performed.
The reference sample can consist of:
in-house or commercially available reference materials - if possible certified - with matrices similar to that
of the sample;
synthetic samples, made by weighing the appropriate amount of each pure reagent;
site specific or batch specific samples, similar to the matrix of the sample;
standard addition method or spiked samples may also be used to create standards for which appropriate
reference materials are not available for an element of interest. The matrix material needs to match that
of the sample.
The element concentrations of the reference samples have to be known by certification or by determination
with a different analytical technique.
The reference samples have to be analysed under the same analytical conditions as the sample, meaning the
same sample preparation (pellet, powder etc.), the same analytical measurement method etc.
3 Overview XRF applications
An overview of „portable/benchtop‟ and „hand-held‟ instruments that are at present commercially available, is
given in Annex A. The „hand-held‟ instruments are generally less than 2 kg in weight, whereas the
„portable/benchtop‟ instruments are not below 9 kg.
Waste samples can consist of different matrix types, each of them may require an appropriate sample
preparation. See Annex B in EN 15309, where five different types are described: 1. soil, sediment, fly ash and
sludge samples, 2. samples consisting of carbon matrices, 3. liquid samples, 4. paste-like materials, 5. scrap
samples. The procedures of the different sample preparation techniques are described in ISO 11464,
EN 15002.
Application of XRF in „on-site‟ verification at the entrance of landfill or a waste incinerator, will usually consist
of the check of the presence of one or more components in solid material, with relatively low carbon content.
For large landfills en plants, limited laboratory facilities will be present for sample preparation and
„portable/benchtop‟ instruments available.
Application of XRF during transport in traffic, will require a hand-held instruments at different sizes and forms
of load must be approached.
Applications of XRF ‟in the field‟, e.g. looking for hot spots in suspected areas; hand-held instruments are
preferred because of their mobility.
4 Influence of the sample preparation on the result
Although the XRF technique is a non-destructive technique and it is often stated that no or only a limited
sample preparation is required, the applied sample preparation method plays an important role in the final
quality of the obtained results. The moisture content and the particle size of a solid sample have a significant
influence on the obtained quality of the results. The highest quality can be obtained when the samples are
dried and finely ground. However, on the field it is not always possible to perform an extended sample
pretreatment.
XRF analysis can be performed on samples with different levels of pretreatment resulting in different quality of
result:
In situ analysis: The XRF is placed directly onto the sample and the analysis is performed without any
sample preparation (Figure 3). A limited handling of the sample (flattening, make it more compact) can
improve the quality of the obtained results. Nevertheless, the obtained results will only be qualitative and
can only be considered as indicative.

Figure 3 — In situ analysis
Analysis of a bagged sample: A simple way of improving the quality of the XRF results can be achieved
by homogenisation of the sample. When putting the sample in a plastic bag, the sample can be
homogenised roughly and crushed into smaller parts by hand. The XRF measurement can be performed
directly on the bagged sample (Figure 4).
Figure 4 — Analysis of a bagged sample
Analysis of a wet sieved sample: Although a wet sample is not easy to sieve, it can significantly improve
the XRF data quality. The sieved sample can be put in a bag and analysed as such or it can be placed in
a sample cup and analysed with the XRF in stand mode (Figure 5). Using this procedure the XRF results
can be considered as semi-quantitative.

Figure 5 — Analysis of a wet sieved sample
Analysis of a dried and ground sample: If a mobile lab is present at the site under investigation, it can be
equipped with a drying oven (e.g. microwave) and grinding tools. Analysing a dried and finely ground
sample will significantly improve the accuracy and precision of the XRF results. The prepared sample can
be analysed with the XRF in stand mode (Figure 6). Using this procedure, one have to keep in mind that
longer testing times are required and that the sample throughput will be lower compared to the direct XRF
analysis.
Figure 6 — Analysis of a dried and ground sample
In a publication of Vanhoof et al [2] several types of XRF instruments were compared, together with the
influence of the applied preparation method on the acquired data using a field portable instrument. The
influence of the pretreatment on the acquired data is shown in Figure 7. Performing in situ analysis with the
field portable Spectrace 9000 resulted in low regression coefficients R of 0,7 for both elements. Also the
slopes were very low, therefore, the XRF values with a low concentration level were overestimated while the
higher levels were underestimated. A limited homogenisation (sieving over 2 mm) of the soil sample
significantly improved the regression parameters resulting in more comparable results.
Figure 7 — Method comparison for the elements Zn and Pb, measured with the field portable EDXRF,
as a function of the pretreatment
These results confirm that in situ analyses are strongly influenced by the heterogeneity of the samples. A
limitation of the in situ analysis, without any homogenisation, is that only the surface top layer is analysed on
just a small soil area (± 5 cm ) corresponding to the XRF measuring window. Furthermore, the
homogenisation process also assure that the same sample will be analyzed by the XRF system and the
reference ICP-AES technique, while in situ analysis can be regarded as a one point surface analysis.
Field analysis are particularly suitable to define „hot spots‟ and to differentiate contaminated areas giving a
measure of the degree of contamination. For these purposes the sample strategy is of major concern to obtain
useful results. Further pretreatment i.e. drying and fine grounding of the soil sample, still improve the data
quality but less significant. However, the largest progress in achieving comparable data between the XRF and
the ICP-AES system is obtained when a minimal homogenisation is performed.
In commission of the RWS Waterdienst (The Netherlands) a similar study was performed by N. Walraven [3]
to evaluate the possibilities of XRF in order to measure Pb, Cu, Zn, As, Cd , Hg, Ni, Cr, Mo, Ba, Sn, V and Se
in sediments. The performance characteristics of the portable Niton XL3t 600 XRF analyser were determined
for the analysis of sediments, and special attention was spent to the evaluation of field moisture samples.
Using certified reference materials the accuracy of the XRF system was determined. In the project two
different hand-held Niton XRF systems were used (same type). Based on the results shown in Table 1, it can
be stated that different performance characteristics were obtained with the same type of XRF instruments.
Nevertheless it is possible to improve the calibration programme to obtain more accurate results. These data
confirm that it is a necessity to validate always the XRF instrument in use.
Table 1 — Yield determined on certified reference materials
Yield Zn Pb Cu As Cr Ba Ni Sn Hg Cd V Mo
% % % % % % % % % % % %
84 to 89 to 73 to 101 to 57 to 90 to 96 to 79 to 104 to 84 to 74 to
XRF 1 ND
89 100 107 112 86 132 137 86 155 108 79
78 to 79 to 69 to 54 to 47 to 77 to 80 to 111 to 85 to 144 to 76 to 71 to
XRF 2
86 92 97 89 85 129 129 121 113 177 77 76
ND: not determined
For this study sediment samples (51) were collected and homogenised manually. The sediment samples were
analysed as such with the XRF system and also after drying. The results indicate that a moisture content of
20 % to 30 % in the samples has an effect on the quality of the measured values. Correction of the obtained
XRF data for the moisture content improve the final results, as shown in Figure 8.

Key
X-axis: concentration Zn in mg/kg, results of accredited laboratory
Y-axis (left figure): results of mobile XRF (without moisture correction), Zn in mg/kg
Y-axis (right figure): results of mobile XRF (after moisture correction), Zn in mg/kg
Figure 8 — Influence of moisture content on the accuracy of the XRF results
The U.S. Environmental Protection Agency's (EPA) Superfund Innovative Technology Evaluation (SITE)
Programme encourages the development and implementation of innovative treatment technologies for
hazardous waste site remediation and monitoring and measurement technologies (www.epa.gov/ORD/SITE)
[4].
In the SITE Demonstration Programme, the technology is field-tested on hazardous waste materials.
Engineering and cost data are gathered on the innovative technology so that potential users can assess the
technology's applicability to a particular site. Data collected during the field demonstration are used to assess
the performance of the technology, the potential need for pre- and post-processing of the waste, applicable
types of wastes and waste matrices, potential operating problems, and approximate capital and operating
costs.
A filed demonstration of eight XRF instruments (hand-helds and bench top) to measure trace elements in soil
and sediments was conducted in 2005. Separate verification reports have been prepared for all the XRF
instruments that were evaluated as part of the demonstration. The objectives of the evaluation included
determining each XRF instrument‟s accuracy, precision, sample throughput, and tendency for matrix effects.
To fulfil these objectives, the field demonstration incorporated the analysis of 326 prepared samples of soil
and sediment that contained 13 target elements. The following reports are available on the website:
EPA/540/R-06/001 XRF Technologies for Measuring Soil and Sediment (Rigaku ZSX Mini II XRF)
http://www.epa.gov/ORD/SITE/reports/540r06001/540r06001.pdf
EPA/540/R-06/002 XRF Technologies for Measuring Soil and Sediment (Innov-X XT400 Series XRF)
http://www.epa.gov/ORD/SITE/reports/540r06002/540r06002.pdf
EPA/540/R-06/003 XRF Technologies for Measuring Soil and Sediment (Niton XLi 700 Series XRF)
http://www.epa.gov/ORD/SITE/reports/540r06003/540r06003.pdf
EPA/540/R-06/004 XRF Technologies for Measuring Soil and Sediment (Niton XLt 700 Series XRF)
http://www.epa.gov/ORD/SITE/reports/540r06004/540r06004.pdf
EPA/540/R-06/005 XRF Technologies for Measuring Soil and Sediment (Rontec PicTAX XRF)
http://www.epa.gov/ORD/SITE/reports/540r06005/540r06005.pdf
EPA/540/R-06/006 XRF Technologies for Measuring Soil and Sediment (Xcalibur ElvaX XRF)
http://www.epa.gov/ORD/SITE/reports/540r06006/540r06006.pdf
EPA/540/R-06/007 XRF Technologies for Measuring Soil and Sediment (Oxford ED2000 XRF)
http://www.epa.gov/ORD/SITE/reports/540r06007/540r06007.pdf
EPA/540/R-06/008 XRF Technologies for Measuring Soil and Sediment (Oxford X-Met XRF)
http://www.epa.gov/ORD/SITE/reports/540r06008/540r06008.pdf
In Annex B a summary is presented of the performed demonstration programme.
5 Evaluation of the XRF screening technique
The XRF technique can be applied for screening purposes on waste and when doing this specific
requirements are needed typical for these applications.
In a very short time-frame an XRF analysis can generate a fingerprint analysis of a waste sample which
arrives at the entrance of the landfill or the incineration plant, and this directly on the truck load or on selected
samples taken from the loading. Although this screening analysis will not be performed under „ideal‟ laboratory
conditions (no extended sample preparation, homogenisation), and therefore results will be only qualitative,
the obtained information can be very useful to verify if the waste matches the expected characteristics for that
type of waste. Because the chemical and physical properties of the arrived waste is known from the basic
characterisation step, the appropriate fingerprint or spot check parameters can be chosen easily, since the
purpose of the screening analysis is to verify that the waste arrived at the entrance is the actual waste
expected.
To verify the waste directly on the truck only hand-held instruments can be applied, performing
non-destructive element analyses. Qualitative results will be obtained within a few minutes and can be used
as a yes/no indicator for acceptance of the waste. As these analysis are performed under „non-ideal‟
laboratory conditions and using short measurement times, only hot spot parameters with concentration levels
above 500 mg/kg to 1 000 mg/kg (element and matrix dependent) can be screened. In comparison, under
optimal conditions the limits of detection are situated in the low mg/kg region.
If sufficient time is available samples can be collected from the truck and analysed by a bench top XRF
system placed in a mobile lab. Using this procedure a limited sample homogenisation and preparation step
(e.g. homogenisation in a plastic bag) can be included resulting in more reliable results.
It has to be clear that all the obtained information acquired by the XRF analyses can only be considered as a
guide value. In case there is any discussion on the validity of the results, representative samples have to be
taken and send to an accredited laboratory for confirmative analysis.
The evaluation of the XRF screening technique has to be performed on different levels.
First of all, the performance of the XRF equipment itself has to be evaluated. By analysing samples under
ideal conditions the performance of the XRF equipment can be checked taking into account: the accuracy, the
precision (repeatability and reproducibility) and the detection limits. Therefore, XRF analysis has to be
performed on homogenised and finely ground waste samples. Additionally, facts as the sample throughput,
the required skills of the operator, analysis time, should be considered.
Secondly, XRF analysis has to be performed on the field to gain insight in the contribution of the error when
performing on-site analysis. This evaluation will provide information whether the XRF system and software
(calibration programme) are capable of handling analyses under non-ideal conditions i.e. wet samples, non-
flattened samples. The influence of the applied sample pretreatment should be evaluated.
At the moment there are several suppliers of XRF instruments on the European market. The XRF instruments
can be divided into 2 groups i.e. the laboratory systems for confirmatory analysis and the portable ones. The
latter group consists of hand-held systems and portable bench top systems (maximum weight around 10 kg).
For the application of the first group of XRF instruments the EN 15309 is developed. For field application
another European guideline is required. As pre-normative investigation a robustness study, initiated by the
German states working group on waste (LAGA), was carried out [5].
In this robustness study the following items have been surveyed:
Method detection limit;
Accuracy;
Precision;
Inventarisation/evaluation of false negative results;
Effect of interferences;
Effects of sample characteristics (sample type, moisture content, particle size, heterogeneity);
Skills and training of operators;
Health and safety aspects;
Portability (weight, size).
6 Robustness study: description and results
6.1 General
An XRF robustness study was initiated by LAGA [5] to prove the capability of hand-held XRF systems for
screening waste loads on hazardous substances as incoming inspection at waste handling plants in order to
identify the waste material or to classify it by its key variables (critical parameters). A summary of this
robustness study is presented below, referring to the final report of this study for a more detailed description
and extended presentation of the results.
In the study a market survey of the available hand-held XRF-instruments on the European market was
performed, together with a field trial to demonstrate the capability of hand-held XRF instruments. All relevant
manufacturers as evaluated in the market survey were invited to a workshop. Selected waste samples were
arranged in 6 different rooms (one room for each type of sample) in the same way in order to guarantee
similar test conditions. The participants were asked to analyse all samples in 5-fold. All instruments with the
exception of one portable bench-top instrument were pistol-like hand-held XRF instruments. One
manufacturer measured directly on the sample while the others used the hand-held instrument in a bench top
stand.
The measurements were focused on the following elements: As, Ba, Cd, Co, Cu, Cr, Hg, Mn, Mo, Ni, Pb, Sb,
Tl and V, while optionally the elements Ag, Fe, Sn, Zn, S and halogens were considered.
Confirmatory analysis were performed by the reference methods to determine the reference value.
6.2 Technical description of the instruments
All together 5 hand-held XRF instrument manufacturers participated to the workshop. Since some of the
manufacturers used several hand-held XRF instruments, in total 8 different portable XRF instruments were
available for the workshop. In Annex A an overview is given of the technical specifications of the instruments
used during the field trial. In general the following findings can be formulated:
All instruments are equipped with a miniaturised X-ray tube using different anodes (Ag, Ta, Rh, W). The
use of radio-isotopes as X-ray source is out of the market;
The measurement window size is instrument dependent and is in the range between 3 mm and 10 mm;
The use of a Si-PIN detector (resolution < 190 eV) is common practice, while the use of an SDD detector
(silicon drift detector, resolution between < 145 eV and < 175 eV) is in progress. Six of the eight
instruments in the field trial were provided with an SDD detector;
Most of the instruments are equipped with primary beam filters to eliminate interferences;
Most of the instruments use different calibration programmes depending on the type of sample to be
analysed. Selection of the suitable programme is based on a visual evaluation of the sample or on a short
pre-screening of the sample;
All instruments are provided with the necessary protection tools regarding X-rays e.g. cover shield, IR
sensor and backscatter control, radiation proximity switch;
All instruments can operate using rechargeable Li-ion batteries or AC-power;
For all instruments data storage is possible on the instrument itself (with or without PDA), together with
data transfer to PC using e.g. USB, bluetooth;
The pistol-like hand-held instruments can be used for in situ measurements directly on the sample, but
beside that they can also be placed in a bench top stand for use in a mobile lab;
The weight of all instruments is below 10 kg, the pistol-like hand-held XRF systems even below 3 kg, and
the size of all instruments are below 500 mm wide x 500 mm deep x 200 mm high;
Optional some instruments have the capability of measuring the low atomic mass elements by using e.g.
He purge, vacuum system;
Optional GPS technology is available on some instruments.
6.3 Description of the selected samples and their characterisation
For the field trial 6 different waste materials have been prepared: construction waste, shredder material,
contaminated soil, waste wood, Pb granulate, and slag from a municipal incineration waste (Figure 9). One
part of these samples has been examined as raw material, whereas another part has been homogenised and
ground to a particle size of less than 250 µm (Figure 10).

S1: Pb granulate (0 mm to 8 mm); S2: Construction waste (0 mm to 10 mm); S3: Waste wood (0 mm to
2 mm); S7: Shredder (0 mm to 2 mm); S8: Contaminated soil; S11: Municipal incineration waste, extract
production (0 mm to 8 mm)
Figure 9 — Overview of the selected samples

Figure 10 — Example of raw and fine sample S2 (construction waste)
For the evaluation of this study the results of the hand-held XRF instruments have been compared with the
reference values of laboratory analyses (EDXRF-analyses according to EN 15309 and ICP-OES analyses
after HF:HNO :HCl digestion according to EN 13656).
6.4 Results of the field trial
6.4.1 Defining performance criteria
A visual inspection of the means of the different instruments was performed sample wise and evaluated by
colouring according to an acceptance value. Additionally, a comparison with the reference values was given.
As example, the data of sample S1 (Pb granulate) are shown in Table 2.
The table represents per element the mean of the 5 repetitive measurements (different sampling) for each
instrument, ordered according the list of instruments of the manufacturers. The column indicated by „Mean of
the means‟ gives the grand mean i.e. the mean of all instrumental means. The right column shows the
reference value.
Each cell (i.e. mean value) is coloured according to an acceptance value indicating the empirical tolerance of
deviation of the reference value:
For contents below 500 mg/kg the acceptance level is 5 x and 1/5 x the reference value;
For contents between 500 mg/kg and 5 000 mg/kg the acceptance level is 2 x and 1/2 x the reference
value;
For contents above 5 000 mg/kg the acceptance level is 1,3 x and 1/1,3 x the reference value.
These acceptance levels are arbitrary chosen, but based on expert judgement. The mean values within these
acceptance levels are marked green, those mean values above this acceptance level (overestimation) are
marked red and the mean values below these acceptance levels (underestimation) are marked blue. Elements
with too low concentrations (in the range of the detection limit) are not considered.
Table 2 – Sample S1 (Pb granulate) raw – coloured according to an accepta
...