Proficiency test : a quality assurance method for high-purity gamma spectrometry system

As part of the implementation of the quality system on ISO 17025 and accreditation, SALROM laboratory participated in the last year in intercomparison exercises, national and international, for determination of natural and man-made radionuclides. This paper describes results obtained in intercomparison exercise organized by IAEA in 2011. The reported values were mostly in good agreement with the resulting


Introduction
The activity of SALMROM laboratory, part of Life and Environmental Physics Department from IFIN-HH, consists in regular monitoring of the environmental radioactivity, on daily/monthly bases, in the area of the institute and nearby.Routinely measured isotopes in environmental samples are 40 K, 60 Co, 137 Cs, 241 Am and progenies of 238 U and 232 Th series.Some other activities, devoted to research programs, are related to the evaluation of radon concentration in salt mines, caves, construction sites or former industrial areas, to radionuclide inventory in water, soil or vegetation in the environment.
The laboratory has a quality system implemented and accredited to ISO 17025.In the frame of accreditation systems, the use of reference materials, both for quality control and proficiency testing, has therefore increased in recent years [1].In addition, proficiency testing that includes distribution of homogenous portions of the test material for analysis as an unknown is a method for assessing and documenting the reliability and accuracy of the analytical data produced [2].This paper presents the results obtained by the laboratory in an intercomparison exercise organized by IAEA in 2011, the first year when the laboratory has participated in such proficiency test.

Materials and methods
The equipment used for certified activities such as radiological monitoring of the environment is a germanium detector, with a diameter of 59.1 mm and a length of 54.3 mm, corresponding to a volume and mass of active germanium of 149 cm 3 and 0.8 kg, respectively.It is a coaxial p-type HPGe detector (ORTEC, model GEM 30P4) having a relative efficiency of 35% and energy resolution of 1.85 keV at 1332.5 keV, 60 Co.The detector is surrounded by a specially designed shield consisting of 10 cm lead and 2 mm copper.The detector is linked to an appropriate data acquisition system a Corresponding author: rileana@nipne.roDigiDART Ortec and to a spectrum analysing software tool MAESTRO-32 6.06.The metrological traceability of the equipment is assured by using radioactive standards provided by Radionuclide Metrology Laboratory (RML) IFIN-HH.In this paper, the quality assurance of the laboratory for gamma spectrometry measurements has been assessed by the annually organized proficiency test (PT) by IAEA.These PTs evaluate the validity and reliability of the analytical results.Through the provision of reference materials, control and performance assessment by the organization of proficiency tests and inter-comparison exercises, IAEA has been assisting laboratories in testing, improving and maintaining the reliability and quality of analyses of radioactive material [2].The data presented are the results submitted for the IAEA-TEL-2011-03 world-wide open proficiency test on spiked water and soil samples.The spiked samples were distributed to the participating laboratories and, using a rating system, the obtained analytical results are compared to the reference values assigned.For this PT four different samples were received and measured: three spiked water samples of 500 ml each and one soil sample of 200g.It is of major importance to obtain accurate analytical results of radionuclides, our laboratory producing acceptable results for most of the radionuclides reported: 60 Co, 133 Ba, 137 Cs, 152 Eu and natural radionuclides.For few radionuclides, 134 Cs, 241 Am and 226 Ra, the need for corrective actions in the analysis process is indicated.
To reach the final score "Acceptable", "Warning" or "Not acceptable" for the laboratory results various statistics such Z-scores and U-scores are used.The final evaluation includes both the total combined uncertainty associated with the target value of proficiency testing samples (unc IAEA ) and the total uncertainty reported by the participating laboratories (unc Laboratory ).According to this approach, the reported results are evaluated against the acceptance criteria for trueness and for precision.A result must pass both criteria to be assigned the final status of ''acceptable''.Based on the uncertainties assigned to the reported values the laboratory can establish its individual acceptance range and from the methodological point of view the credibility of uncertainty statement is checked.
In the final evaluation report for each laboratory one of the first parameter calculated is the bias of the reported result "Value Laboratory ".This is calculated relative to the assigned value "Value IAEA " according to Eq. 1: A high relative bias is the first indication that a result obtained by the laboratory is not in agreement with the target value.Further, the Z-score is calculated from the laboratory results, the assigned value and a standard deviation in accordance to Eq. 2 as follows: where the standard deviation (σ) is 0.10 × Value IAEA (for the "fitness for purpose") and the laboratory performance is evaluated as satisfactory if |Z Score |<2, questionable for 2<|Z Score |<3 and unsatisfactory for |Z Score |>3.As can be observed from Eqs.1-2 the uncertainty of the laboratory is yet not taken into account.This is considered in U-score, Eq. 3: The ''acceptable'' criteria for trueness of the laboratory result is based on A1≤A2 where: The limiting value for the U-test parameter is 2.58 for a level of probability at 99% to determine whether a result passes the test (U<2.58).
For the evaluation of precision, an estimator "P" is calculated.
DOI: 10.1051/ C Owned by the authors, published by EDP Sciences, 2014 ,