|
Title:
|
|
Open-Bio - Opening bio-based markets via standards, labelling and procurement: Deliverable 3.1: Performance characteristics for horizontal bio-based carbon content standard – round robin assessment results
|
|
|
|
Author(s):
|
|
|
|
|
|
Published by:
|
Publication date:
|
|
ECN
Environment & Energy Engineering
|
2-6-2015
|
|
|
|
ECN report number:
|
Document type:
|
|
ECN-E--15-028
|
ECN publication
|
|
|
|
Number of pages:
|
Full text:
|
|
52
|
Download PDF
|
Abstract:
This report presents the results of the round robin assessment that was organised with the aim to investigate the performance characteristics of the method that is described in CEN TS 16640 for the bio-based carbon content determination, in order to convert the available tech-nical specification into the European standard. The round robin assessment was initiated in the framework of the European Open-Bio project (www.biobasedeconomy.eu). The assess-ment involved 11 independent laboratories to whom in total 132 samples were delivered (11 equivalent sets of samples, 12 samples each set).
Next samples were involved in the round robin testing:
Sample 1. White water soluble matt paint, volatile components about 34% are present; pos-sible difficulties with ignition, combustion in an elemental analyzer is recommended. Non-hazardous.
Sample 2. White emulsion; non-volatile; non-hazardous; used as one of components of a sun lotion.
Sample 3. White emulsion; non-volatile; non-hazardous; used as one of components of a sun lotion (different from Sample 2)
Sample 4. A wheat straw panel, 10cm x 10cm; non-hazardous; can be used for different construction and building purposes
Sample 5. Highly flammable liquid (biodiesel); used as a fuel.
Sample 6. A container filled with bio-gas, pressurized to 2.5bar, H2S content 25ppm. The biogas contain approximately 60% of CH4 and 40% CO2.
Sample 7. White surfactant granules that are used in cosmetics; non-hazardous.
Sample 8. Cosmetic emulsion with high water content; non-hazardous.
Sample 9. Multilayer packaging film; presents no hazard.
Sample 10. Silk paint; non-hazardous.
Sample 11. Bio-based binder used in paints; non-hazardous.
Sample 12. Wooden particle board ground to 0.5mm; presents no hazard.
None of these samples demanded a special storage conditions.
These samples, together with the latest available version of CEN TS 16640, were sent to each participating laboratory.
Below the list of participating laboratories is presented:
Agroisolab GmbH, Germany
Beta Analytic, USA
Centre de Datation par le RadioCarbone/Institute of Analytical Sciences, France
Energy research Center of the Netherlands (ECN), the Netherlands
SGS, France
SKZ, Germany
Silesian University of Technology, Institute of Physics, Radiocarbon Laboratory, Poland
Scion/GNS Science, National Isotope Centre, Rafter Radiocarbon, New Zealand
University of Wageningen, Food and Bio-based Research, the Netherlands
University of Groningen, Center for Isotope Research (CIO), the Netherlands
University of York, Green Chemistry Centre of Excellence, United Kingdom
Due to the confidentiality agreements, the results obtained by each laboratory are presented in anonymous way. Every laboratory was prescribed a name known only to the organiser of the assessment and to that specific laboratory. In the final report, the results are presented using these names (Lab 1, Lab 2, … Lab 11) so that every laboratory can have an overview of all results, but is able to recognise only its own results. Laboratories were free to choose their own method when preparing a (sub)sample that would be homogeneous and repre-sentative of the received sample. For pre-treatment, CEN TS 16640 was advised to follow.
The round robin test was carried out to determine the influence of parameters which may vary between individual laboratories. Subsequently, the reproducibility standard deviations were calculated based on the results reported by each laboratory. Statistical evaluation of the results was done when analysing the results from all participating laboratories on each indi-vidual sample. Extremely biased results were investigated for possible errors. In the current study, the Grubbs test was used for statistical evaluation of the results that were reported for each sample by each laboratory. Outliers and stragglers that were defined based on the re-sults of Grubbs analysis, were excluded from the calculations of measured average numbers and the reproducibility standard deviations among all laboratories.
The results of performed assessment showed a good consistency. The maximum number of outliers/stragglers (1 outlier, 1 straggler) when analysing the reported results on the 14C con-tent, was observed for Sample 1 that was a paint with low carbon content and with a high volatile fraction (approximately 35%). This can be related to the combustion difficulties and possible loss of carbon that could be present in the volatile fraction. The maximum value for the variation of the coefficient of the reproducibility (17.7%) for the biogenic carbon content was observed for the same Sample 1 (10.2 ± 1.8 % of 14C as fraction of total carbon, see Table 5), that was one of the most challenging samples. Analysing the calculated perfor-mance characteristics for the total carbon content, one can observe that the highest value for the variation of the reproducibility standard deviation for total carbon content was 8.8% (15.9 ± 1.47) for Sample 8 that was cosmetic emulsion with high water content (see Table 3). Rela-tively high variation in the coefficient of the reproducibility for Sample 1 and Sample 8 can be caused by combustion difficulties of these two samples: some laboratories used combustion enhancers and some did not. This can explain somewhat high values for the reproducibility variations and has to be taken into account when converting paint-like or water-containing samples into carbon dioxide.
For the C14 analysis, the known LSC (Liquid Scintillation Counting)or the AMS (Accelerated Mass Spectrometry) techniques were used in this round robin assessment. 3 of 11 laborato-ries did the 14C analysis using the LSC method (no direct LSC was performed on any sam-ples). By 8 laboratories the AMS analysis was used in order to determine the 14C amount in the delivered samples. The results of the round robin assessments indicated that these two techniques give the equivalent results as no inconsistencies were observed for the results of the measurements when using AMS (Accelerated Mass Spectrometry) and LSC (Liquid Scin-tillation Counting) techniques.
Back to List