Evaluation of uncertainty components associated with alpha-particle spectrometric measurements

Evaluation of uncertainty components associated with alpha-particle spectrometric measurements of uranium isotopes in water

Meryem Seferino?lu n,Abdullah Dirican,P?nar Esra Erden,Demet Er?in

Turkish Atomic Energy Authority,Sarayk?y Nuclear Research and Training Center,06983Kazan,Ankara,Turkey

H I G H L I G H T S

A radiochemical separation method for U using extraction chromatography.

A procedure for evaluation of the uncertainty of the measurement.

Combined relative standard uncertainty was approximately1.4Bq kgà1(7.9%)for238U

Accuracy and precision were checked by analysing of IAEA-PT-samples.

Results are in good agreement with recommended values.

a r t i c l e i n f o

Article history:

Received2June2014

Received in revised form

4September2014

Accepted4September2014

Available online16September2014

Keywords:

Measurement uncertainty

Alpha-particle spectrometry

Radiochemical separation

Uranium in water

a b s t r a c t

Quali?cations of uncertainties associated with the measurement of speci?c activity concentration of

uranium radioisotope(238U)in water samples by alpha-particle spectrometry are presented.Possible

sources of uncertainty are identi?ed and quanti?ed in the activity concentration measurements of238U

isotope;the major source being the statistical counting uncertainty as expected.The combined relative

standard uncertainty Uea238

Tof the measurement was calculated as 1.4Bq kgà1(7.9%)for the

investigated NPL sample.The accuracy and precision of recommended procedure were checked

analysing six spiked water samples supplied from IAEA-pro?ciency test exercises.The results were

evaluated in terms of relative bias,z-score,u-score,trueness and precision.These results show that the

activity values and their uncertainties are in good agreement with recommended values.

1.Introduction

Assessment of any release of radioactivity to the environment

is very important for the protection of public health,especially if

the released radioactivity might enter the food chain or result in the

radiation exposure to the population(IAEA,1989).Monitoring the

radiological content of water is necessary to ensure lower levels of

radioactivity(Council Directive2013/51/EUROATOM2013,Jobbágy

et al.,2009;W?tjen et al.,2010).The EU Commission Drinking Water

Directive sets standards on the most common parameters that need to

be checked in drinking water.Among many other parameters relevant

to microbiological and chemical properties of the water,uranium and

radium isotope content has to be regularly monitored in drinking water

(Vasile and Benedik,2008)since the alpha activity in water is mainly

due to the uranium and radium isotopes.The derived concentrations

for radioactivity of uranium isotopes in water intended for human

consumption in the Council Directive2013/51/EURATOM(2013)are

set to 3.0Bq Là1for238U and 2.8Bq Là1for234U based on its

radiological properties.Uranium is a naturally occurring radioactive

heavy metal and is widely distributed in the nature(Condon et al.,

2010).It has three long-lived,naturally occurring isotopes;238U(t1/2?

4.468?109a,99.27%abundance(LNHB recommended data2014,

Jaffey et al.,1971)),235U(t1/2?704?106a,0.72%abundance

(LNHB recommended data2014,Jaffey et al.,1971))and234U

(t1/2?2.455?105a(LNHB recommended data2014,Cheng et al.,

2000),0.0054%abundance(Jia et al.,2006)).

The WHO recommended values for uranium in the document

“Guidelines for drinking water quality”set the most stringent

limitation with2m g Là1(WHO,1998).The recommended value

was raised to15μg Là1based on its chemical toxicity for the

kidney in the3rd edition of this document published in2004

(WHO,2004)and later it was raised to30m g Là1in the4th edition

published in2011(WHO,2011).Consequently,the determination

of the activity concentration of uranium isotopes at trace and

ultra-trace levels with high accuracy and precision is of interest

due to the radiotoxicity of these isotopes.

The measurement of radioactivity via any radiometric techni-

ques should be performed with suf?cient accuracy and realistic

Contents lists available at ScienceDirect

journal homepage:https://www.360docs.net/doc/d512309982.html,/locate/apradiso

Applied Radiation and Isotopes

https://www.360docs.net/doc/d512309982.html,/10.1016/j.apradiso.2014.09.002

0969-8043/&2014Elsevier Ltd.All rights

reserved.

n Corresponding author.Tel.:t903128101551.

E-mail address:meryem.seferinoglu@https://www.360docs.net/doc/d512309982.html,.tr(M.Seferino?lu).

Applied Radiation and Isotopes94(2014)355–362

measurement uncertainty to obtain valid result with high accuracy and high precision.The measurement uncertainty is an essential part of the assessment of the measurement result(Jerome and Eccles,2007).Previously the uncertainty in the radioactivity measurements was de?ned by the standard deviation of repeated measurements or was not reported.The standard deviation in a radioactive measurement given by the square root of the number of counts re?ects only the statistical uncertainty of counting (Vre?ek and Benedik,2003;Kim et al.,2008).Recently,more attention is given to the evaluation of the measurement uncer-tainty and to the preparation of uncertainty budget.There is, however,no common procedure for the estimation of a realistic uncertainty.The ISO Guide“Guide to the Expression of Uncertainty in Measurement”(ISO,1995)sets up general rules for evaluating and expressing uncertainty in a physical measurement.Some recent European documents give further details about calculating uncertainty in calibration and in quantitative chemical analysis (EA European Co-Operation for Accreditation,2013;Eurochem, 1995).Nonetheless,the application of these concepts to environ-mental radioactivity measurements involving radiochemistry is not as straightforward as it should be,because the evaluation of uncertainty requires the analyst to take into account all stage of the method and all possible sources of uncertainty(Spasova et al., 2007;Kim et al.,2008).

The main purpose of this study is to demonstrate a procedure for evaluation of the measurement uncertainty in the analysis of uranium isotopes in water using a radiochemical separation method and alpha-particle spectrometry.The steps in the analytical proce-dure that contribute considerably to the combined measurement uncertainty are also identi?ed.Identi?cation of these steps might help lower the overall uncertainty of the measurements through decreasing the uncertainties in these steps.

2.Experimental

2.1.Instrumentation and calibration

Activity concentrations of uranium isotopes were determined by using alpha-particle spectrometry.Alpha sources were counted by a Passivated Implanted Planar Silicon(PIPS)detector with active surface area of450mm2(Canberra).The energy calibration was performed by using an electroplated mixed standard source containing238U,1.67Bq(75.2%);234U,1.63Bq(74.9%);239Pu, 1.91Bq(75.5%)and241Am,1.81Bq(75.5%)supplied from Eckert &Ziegler Isotope Products.

2.2.Reagents and tracer

All reagents used in this study were of analytical grade and supplied from Merck.The UTEVA(100–150μm particle size) extraction chromatography resin was supplied from Eichrom Technologies,Inc.(Darien,Illinois,USA).

The National Institute of Standards and Technology(NIST) traceable radionuclide,232U(SRM4324B)was used for determina-tion of recovery of radiochemical separation.The232U was purchased in the form of UO2Cl2dissolved in2M HCl with an activity concentration of38.22Bq gà1.The solution was diluted with2M HCl to reduce the activity concentration to0.111Bq gà1 (70.085%,k?1).

2.3.Sample description

A spiked water sample from IAEA-CU-2008-03,two spiked water samples from IAEA-CU-2010-03,two spiked water samples from IAEA-CU-2010-04Pro?ciency Test Exercises and an Alpha Low(AL)sample of NPL Environmental Radioactivity Pro?ciency Test Exercise2009were used.The IAEA pro?ciency test samples were used to validate the sample preparation and measurement procedures.The water sample taken from NPL2009pro?ciency test exercise was used for detailed measurement uncertainty analysis and determining the parameters contributing to the measurement uncertainty.The pH of water samples were checked and if necessary,the water samples were acidi?ed with concen-trated HNO3to pH2to ensure the trace elements and radio-nuclides are kept in the sample.

2.4.Blank description

Deionised water was used for reagent blank analysis.A number of replicate reagent blanks were analysed by the same procedure used for water sample analysis to determine the presence of tracer levels of uranium in the reagents.

2.5.Sample preparation

The Eichrom ACW02analytical procedure for separation of uranium in water samples(Eichrom,2001)was used.The AL sample of NPL Environmental Radioactivity Pro?ciency Test Exer-cise2009was diluted at the beginning of the analysis.10g of AL sample was weighed on an analytical balance calibrated with SI traceable weights and diluted to1.0L with deionised water.A sub-sample of50mL of this solution was taken by volumetric?ask. The sub-sample was weighed on an analytical balance and found as49.9864g.This sub-sample was used for uranium analysis.

A weighed aliquot of232U tracer was added to the samples at the beginning of the procedure to quantify the yield throughout the radiochemical separation procedure and to determine the radio-chemical recovery in alpha-particle spectrometry analysis.The schematic diagram of the sample preparation steps is presented in Fig.1A.0.5mL of1.25M Ca(NO3)2was added into the sample and heated with stirring until boiling.After addition of phenolphtha-lein indicator and 3.2M(NH4)2HPO4,the uranium was co-precipitated with Ca3(PO4)2by addition of concentrated NH4OH. The precipitate was allowed to settle overnight.The supernatant was discarded by decantation.The precipitate was then centri-fuged and washed twice using deionised water.The precipitate was dissolved with5mL of concentrated HNO3and evaporated to incipient dryness.The residue was dissolved in10mL of3M HNO3–1.0M Al(NO3)3.

Spectrometric measurements of uranium radioisotopes by alpha-particle spectrometry require very pure,uniform and thin sources.Therefore radiochemical separation is necessary to sepa-rate uranium from all other interfering elements(radioactive or not)present in the water,since those elements eventually lead to lower spectral resolution and higher detection limit(Kiliari and Pashalidis,2010;Reyes and Marques2008).The radiochemical separation is based on the isolation of uranium isotopes from other radionuclides such as Pu,Am,Th and Np in the sample using UTEVA resin prior to measurement by alpha-particle spectrometry. UTEVA resin does not absorb mono-,di-and tri-valent ions.Thus, the resin effectively separates uranium isotopes from any inter-fering elements prior to electrodeposition(Amoli et al.,2006; Thakkar,2002).2mL of freshly prepared0.6M ferrous sulphamate solution must be added before radiochemical separation to reduce plutonium(IV)to plutonium(III)(Thakkar,2002).This step is necessary to avoid the plutonium isotopes retain on the UTEVA resin.

The schematic diagram of the sequential separation is shown in Fig.1B.The resin was pre-conditioned with5mL of3M HNO3,at a ?ow rate of1mL minà1.Sample that had been dissolved in3M HNO3–1.0M Al(NO3)3was then passed through the resin,followed

M.Seferino?lu et al./Applied Radiation and Isotopes94(2014)355–362 356

by a10mL wash of3M HNO3.The resin was converted to the chloride system with5mL of9M HCl.20mL of5M HCl–0.05M oxalic acid was added to remove any neptunium,plutonium and thorium from the column.Finally,uranium was eluted with15mL of1M HCl,and the puri?ed uranium fraction was evaporated to incipient dryness.The thin sources for alpha spectrometric mea-surements were prepared by electrodeposition using a sulphate system on stainless steel discs(Eichrom,2001).The sources were counted by alpha-particle spectrometry.

2.6.Determination of the activity concentration of238U and uncertainty sources

The measurand is the activity concentration of238U in the water samples supplied from NPL-2009pro?ciency test exercise. The activity concentration was calculated using the following equation(Kanisch,2004):

a A?A A

f1f2f3f4in Bq kgà1e1T

where a A is activity concentration of238U isotope,referred to the date of sampling in Bq kgà1,A A is the amount of activity of238U isotope on the electroplated source in Bq,m s is mass(or volume) of water sample(kg,or L),f1is correction for decay of238U isotope in the time interval from the end of sampling to the beginning of the measurement,f2is correction of decay of238U isotope during the counting interval,et GT,f3is correction for decay of tracer radionuclide in the time interval from its calibration date to the beginning of the measurement and f4is correction for decay of tracer radionuclide during the counting intervalet GT.All these f factors are usually very close to unity.

The decay correction factors(f1,f2,f3and f4)are determined as follows:

f1?expeλAet Sàt ETTe2T

f2?λA t G=e1àexpeàλA t GTTe3Tf3?expeàλTet Sàt CTTe4T

f4?λT t G=e1àexpeàλT t GTTe5TwhereλA is the decay constant of238U isotope in sà1,λT is the decay constant of tracer radionuclide in sà1,t S is the start time of the measurement,t E is the sampling date,t C is the calibration date of the tracer radionuclide solution and t G is counting period.The values ofλA andλT are calculated using the formulaλ?Lne2T=T1=2, in which T1/2is half-life of analyte(238U)or tracer radionuclide (232U).

The activity of238U isotope on the electroplated source at the sampling date(A A)is calculated with using the following equation(Kanisch,2004).

A A?C T m T

R GAàR BA

R GTàR BT

àq1

pαT

PαA

in Bqe6T

where C T is certi?ed activity concentration of the tracer solution, m T is mass of added tracer,R GA is gross counting rate of238U isotope in sà1,R BA is blank counting rate of238U isotope in sà1,R GT is gross counting rate of the tracer radionuclide in sà1,R BT is blank counting rate of the tracer radionuclide in sà1,q1is isotopic impurity ratio of uranium isotope in the tracer solution,PαT is the sum of alpha emission probabilities of those individual alpha lines of the tracer radionuclide and PαA is sum of alpha emission probabilities of those individual alpha lines of238U isotope.PαT and PαA are usually very close to unity.

3.Results and discussion

3.1.Evaluation of standard measurement uncertainty

The components contributing to the combined measurement uncertainty such as count rates of sample and tracer,chemical recovery,tracer activity and mass(or volume)of the sample and the tracer are discussed in detail in this study.The overall measurement uncertainty was evaluated based on Eq.(1)com-bined with Eq.(6),which describe the relationship of238U activity concentration with the variables considered in the accurate measurement.The source of uncertainties in the overall method can be grouped in two categories as Type A and Type B components.Type A components are evaluated from statistical distribution of the results of a series measurements.In Type B components,uncertainties arising from experimental results and from other information such as calibration certi?cates,manufac-turer's speci?cations or other directly related or incompletely speci?ed information are taken into account(Liu et al.,2006; Miller,1996).

Fig. 1.Schematic diagrams of steps involved in the analytical procedure for

separation of uranium in water.

M.Seferino?lu et al./Applied Radiation and Isotopes94(2014)355–362357

3.2.The evaluation of uncertainty associated with the sample mass The evaluation of uncertainty associated with dilution of NPL sample was estimated using uncertainties due to weighing and dilution of NPL sample.The uncertainty component associated with the mass of the NPL sample was estimated using the equation U (m s )?2.66?10à4g t3.05?10à6R taken from the calibration certi ?cate of the analytical balance,where R is the weight of the sample and was calculated as 0.0023%.

The uncertainty components associated with the dilution process was estimated using uncertainties due to ambient tem-perature dependence,repeatability and uncertainty data from the calibration certi ?cate (given by manufacturer of the ?asks).The uncertainty of the internal volume for 1000mL of volumetric ?ask was recommended as 0.2mL (k ?1)at 201C.Standard uncertainty was calculated as 0:2=???

6p ?0:082mL by assuming a triangular distribution.A standard uncertainty due to repeatability was estimated as 0.03mL using the standard deviation of repeatability measurements.According to the EURACHEM Guide (4),a variation of 731C of the room temperature results in an uncertainty component of V T ?3?2.1?10à4?0.630mL.A rectangular distri-bution was assumed for the uncertainty contribution of tempera-ture variation and this was converted to standard uncertainty by dividing square root of 3giving 0.364mL.These four uncertainty components are combined to give the standard uncertainty of the dilution volume of the sample and calculated as 0.037%for 1.0L.

The uncertainty component due to the weighing of the sub-sample was found as 0.0006%.This uncertainty component was also taken into account in the calculation of the ?nal standard uncertainty of the sample mass.This uncertainty was calculated as 0.037%.

3.3.The evaluation of uncertainty associated with chemical recovery The radiochemical separation is done to isolate uranium iso-topes from other radionuclides such as Pu,Am,and Th in the sample prior to measurement by alpha-particle spectrometry to prevent interference in the alpha energy spectrum.The chemical recovery of separation process is determined by the addition of a suitable tracer (232U)that is not present or expected in the sample.The uncertainty components associated with chemical recovery are the uncertainties due to tracer activity (supplied by the manufacturer of the standard tracer solution)and weighing measurements.

Approximately 1.00g of the tracer was added into the sample by using pycnometer.The uncertainty component associated with the mass of the tracer was estimated using the equation U (m t )?2.66?10-4g t3.05?10à6R taken from the calibration certi ?cate of the analytical balance,where R is the weight of pycnometer ?lled with standard 232U solution.The mass of 232U solution added into the water sample was calculated by subtracting the weight of the pycnometer before and after the addition of a certain amount of 232U standard solution into the sample.The uncertainty com-ponent due to the mass of the tracer U em T Twas calculated by the square root of the sum of the uncertainties of the weights of pycnometer before and after adding the 232U solution and found as 0.019%.

The uncertainty associated with activity concentration (C T )of the certi ?ed 232U solution was estimated using data from the calibration certi ?cate and U (C T )/C T was found as 0.0077.

3.4.Uncertainty of measured counting rates

The regions of interest (ROI)were set as 3.900–4.300MeV for 238

U and 5.000–5.420MeV for 232U isotopes.The spectrum of uranium isotopes by alpha-particle spectrometry is shown in Fig.2.The peak resolution was quite good upon visual inspection and no signi ?cant overlapping peaks were observed.This made peak area determination easier and no spectral deconvolution techniques were used.The uncertainty of the gross counting rate R GA of analyte (238U)peak and R GT of tracer radionuclide (232U)peak in the sample spectrum for acquisition live time t G is given by the following equation,respectively (Kanisch,2004):

U eR GA T?????????R GA t G s and U eR GT T?

????????R GT

t G

s e7T

It is necessary to subtract the background from the total count

of analyte and tracer radionuclides in the region of interest (ROI)to get accurate peak areas (Vre ?ek and Benedik,2003).The standard uncertainty of the corrected counting rates,U eR A Tof analyte and tracer,U eR T Twas calculated by Eq.8,where R U-238and R U-232,are background count rates of 238U and 232U radionuclides,respectively,and t B is the acquisition live time of background.The standard uncertainties of the count rates were found as 8.5?10à5s -1for 238U and 2.9?10à4s à1for 232U.U eR A T????????????????????????????R GA t G tR U à238t B s and U eR T T?

???????????????????????????R GT t G tR U à232

t B

s e8T

In this study,a series of blank analyses were performed with

the analysis procedure used for the sample to determine different blank sources.The blank counting rates of the analyte and the tracer peak and their uncertainties in a series of blank sample spectrum were calculated by Eq.9.Blank count rates were found as 2.0?10à5s à1for 238U and 1.1?10à5s à1for 232U and their uncertainties were 1.3?10à5s à1and 8.2?10-6s à1,respectively.

R BA ?∑n i ?1R BA ;i

;U eR BA T?

?????????????????????????????∑n i ?1eR BA ;i

àR BA T2

n à1

q and

R BT ?

∑n i ?1R BT ;i

;U eR BT T?

??????????????????????????????????????

∑n i ?1eR BT ;i àR BT T2

n à1

s e9T

The uncertainty due to acquisition live time correction in the counting rate calculation was not taken into account since this correction is negligibly

small.

Fig.2.Spectrum of water sample taken from NPL 2009pro ?ciency test exercise obtained from alpha-particle spectrometry measurement (net peak area corrected with background of 238U and 232U are 190.5and 2149.6counts,respectively,counting period is 162,613s and half-life of 232U is 70.6(11)a).

M.Seferino ?lu et al./Applied Radiation and Isotopes 94(2014)355–362

358

3.5.Uncertainty of decay correction factors,sums of emission probabilities and other sources

The correction factors for radioactive decay are approximately equal to 1in the case of 238U and 232U.The uncertainty associated with the decay correction factors of the analyte and the tracer were calculated using the following equations:U ef 1T?f 1et s àt E TU eλA Te10T

U ef 2T?f 2e1àf 2exp eàλA t G TTU eλA T

λA

e11TU ef 3T?f 3et s àt c TU eλT Te12T

U ef 4T?f 4e1àf 4exp eàλT t G TT

U eλT TλT

e13T

where λA and U(λA )are the decay constant of 238U and its uncertainty in s à1,respectively,λT and U (λT )are the decay constant of 232U and its uncertainty in s à1,respectively (U (λA )and U (λT )values are calculated by using the formula λ?Ln e2T=U eT 1=2Tin which U (T 1/2)is the uncertainty of 238U or 232

U half-life),the t E is the sampling time,t s is beginning date of the acquisition,t c is reference date of the 232U standard solution,t G is the acquisition live time in s .The uncertainty contributions of the decay correction factors f 1,f 2for 238U and f 3,f 4for 232U were calculated as 3.0?10–14, 5.4?10à8,and 7.8?10à4, 3.9?10à7,respectively.

Uncertainties of the sum of emission probabilities P αA of 238U,(or U(P αA )/,P αA )and P αT of 232U (or U(P αT )/P αT )were found as 0.0071and 0.0085.

The uncertainty due to isotopic impurities of the analyte in the tracer was not taken into account since the radioimpurities in the 232

U standard solution are not detected (taken from calibration certi ?cate).

3.6.Calculation of the combined standard uncertainty

3.6.1.Calculation with mathematical formula

The activity concentration of 238U radionuclide in the water sample taken from NPL-2009pro ?ciency test exercise was found as 18.3Bq kg à1.The combined standard uncertainty,U ea 238U Tof the measurement was calculated as 1.443Bq kg à1(k ?1)(7.9%)according to the following formulas (Eqs.14–16):

U ea 238U T?a 238U

U eA 238U TA 238U

2tU em s T

m s 2tU ef 1Tf

U ef 2T

tU ef 3T3 2t

U ef 4T4

2!1=2e14T

with:

U eA 238U T?A 238U

U eC T TT

U em T TT

U ey T

U eP α;A Tα;A

2tU eP α;T T

α;T

1=2

and

y ?R GA

àR BA

R GT àR BT

àq 1;e15T

The uncertainty in the measurement result,y related to the counting rates of radionuclide can be calculated by following equation.

U y eT?R GA àR BA GT àR BT

2"U 2eR GA TtU 2eR BA T

GA BA

U 2eR GT TtU 2eR BT TeR GT àR BT T2

!#!

tU 2

eq 1T!1=2e16T 3.6.2.Calculation with the spreadsheet approach

The combined standard uncertainty was also calculated using the spreadsheet approach (Holmes,2004)to see the percentage contribution of each parameter in the measurement of 238U activity concentration with alpha-particle spectrometry.Although the measurement equation is somewhat complicated,these com-plications can be overcome fairly easily using uncertainty spread-sheet approach since it can be used to account for the correlations between the input variables in the measurement equation and the uncertainty estimates.The uncertainty spreadsheet approach requires only a measurement equation that shows how all the quantities of parameters are combined to obtain the desirable result and the individual numerical values and their uncertainties of the parameters in the measurement equation.The method is detailed elsewhere by Holmes (2004).

The values of the activity concentration of 238U isotopes,a A are related to the quantities of parameters given by Eq.(1)which is further combined with Eq.(6).These equations were used in the spreadsheet approach to calculate the combined standard uncer-tainty of the measurement of the activity of 238U with alpha-particle spectrometry.Each parameter value along with its uncer-tainty and the combined standard measurement uncertainty calculation in accordance with spreadsheet approach are given in Table 1.The results indicate that the combined standard uncertainty (1.441Bq kg à1)found with spreadsheet method is about the same as the uncertainty calculated using the formula (1.443Bq kg à1).It is straightforward to come to the conclusion that the spreadsheet approach gives a good estimation on the overall uncertainty.

This method also gives the percentage contributions of each parameter to the overall uncertainty on the result.The analyte (238U)peak area uncertainty is the highest contributor (approxi-mately 87.7%)to the total uncertainty value.The uncertainty of tracer (232U)peak area has also considerable effect on the total uncertainty with a value of 7.2%.The other uncertainties have small effect on the combined uncertainty,especially the uncer-tainties associated with decay correction factors and volume and mass determination are negligible.

As a conclusion,the uncertainties due to counting rates are major sources in the result of uranium analysis by alpha-particle spectrometry.Reduction of these uncertainties is very dif ?cult.The only way to reduce these uncertainties are to increase the measurement time and/or increase the sample quantity (Vre ?ek and Benedik,2003)while maintaining good source quality (thin and homogeneous).

3.7.Validation of recommended procedure for uranium analysis The accuracy and precision of the recommended procedure for

the determination of the activity concentration of uranium iso-topes (234U and 238U)were checked by analysing six spiked water samples supplied from IAEA-pro ?ciency test exercises.The results were evaluated in terms of relative bias,z -score,u -score,trueness and precision as shown in Table 2.The uncertainty of the participant ’s measurement result is not taken into account in the z -score value for the evaluation of performance.u -score however includes uncertainties of the participant measurements and the uncertainty of the assigned value.The limiting value for the u -test parameter is assigned as 2.58for a level of probability at 99%to determine if a reported result passes the test (u o 2.58)(IAEA,2011).The u -score values in this study are found to change from 0.00to 1.26for 234U and from 0.11to 1.34for 238U.These results indicate that the overall measurement uncertainties of the results obtained by alpha-particle spectrometry are within the reference interval.

M.Seferino ?lu et al./Applied Radiation and Isotopes 94(2014)355–362359

The pro?ciency test scoring system applied by IAEA Terrestrial Environmental Laboratory takes into consideration the trueness and the precision of the reported results.According to IAEA scoring system,the pro?ciency test results are evaluated against the acceptance criteria for trueness and precision and then different statuses“Acceptable”,“Warning”or“Not Acceptable”are assigned accordingly.The reported result is assigned“Accep-table”status if it passes both trueness and precision criteria.The criterion for trueness is passed if A1r A2where equations for A1 and A2are given in the footnote of Table2.The claimed measurement result uncertainty is assessed according to the precision evaluation expressed by an estimator P as given in Table2.P directly depends on the measurement uncertainty of the activity values reported.The Limit of Acceptable Precision (LAP)is de?ned by the IAEA in advance for each analyte including any adjustment due to the consideration activity level of the analyte concerned and the complexity of the analytical problem. The reported results passes precision criterion if P r LAP.The LAP values are15%for234U and238U radionuclides in the water samples interested in this study.A result must obtained“Accep-table”score in both criteria to be assigned the?nal status of “Acceptable”.The results presented in Table2show that trueness and precision criteria are satisfactorily ful?lled.In other words,the activity values and its uncertainty measured in this study are in good agreement with recommended values.

4.Conclusions

The measurement uncertainty associated with the determina-tion of the238U activity concentration in water samples by alpha-particle spectrometry was evaluated.The combined standard uncertainty was calculated by using mathematical formula and the spreadsheet method.Easier implementation,the possibility to observe correlations between parameters and percentage contri-butions of different uncertainty components all in one table makes the spreadsheet method favourable over cumbersome mathema-tical calculations.The main sources of uncertainty were identi?ed as uncertainties associated with the counting rates which are very dif?cult to reduce.

The method accuracy and precision were also checked analys-ing spiked water samples supplied through IAEA-pro?ciency test exercises.The results show that the reported values for activity concentration of uranium isotopes and their uncertainties are in good agreement with recommended values.

Acknowledgements

This work was supported by Turkish Atomic Energy Authority (Project no.A2.H4.P1.03).

References

Amoli,H.S.,Barker,J.,Flowers,A.,2006.Electrodeposition and determination of nano-scale uranium and plutonium using alpha spectroscopy.J.Radioanal.

Nucl.Chem.268,497–501.

Cheng,H.,Edwards,R.L.,Hoff,J.,Gallup,C.D.,Richards,D.A.,Asmerom,Y.,2000.The half-lives of uranium-234and thorium-230.Chem.Geol.162,17–33. Condon, D.J.,McLean,N.,Noble,S.R.,Bowring,S.A.,2010.Isotopic composition (238U/235U)of some commonly used uranium reference materials.Geochim.

Cosmochim.Acta74,7127–7143.

Council Directive2013/51/EURATOM2013,Council Directive2013/51/EURATOM of 22October2013laying down requirements for the protection of the health of the general public with regard to radioactive substances in water intended for human consumption(OJ L296,7.11.2013,p.12).

EA European Co-Operation for Accreditation,2013.Evaluation of the uncertainty of measurement in calibration.Publication reference:EA-4/02M:2013. Eurochem,Quantifying Uncertainty in Analytical https://www.360docs.net/doc/d512309982.html,boratory of the Government Chemist Information Services,Teddington(Middlesex)/London(ISBN 0-948926-08-02).

Eichrom,2001.Uranium in water,ACW02,Rev.1.3,Analytical Procedures,Eichrom Technologies,Inc.

Holmes,L.,2004.Alpha-spectrometric analysis of environmental samples-spreadsheet approaches(IAEA-TECDOC-1401).Quantifying Uncertainty in Nuclear Analytical Measurements.International Atomic Energy Agency,Vienna, pp.141–149.

IAEA,1989.Measurement of Radionuclides in Food and Environment:A Guidebook (Technical Reports Series No.295).International Atomic Energy Agency, Austria.

ISO,1995.ISO Guide to the expression of uncertainty in measurement.Internatio-nalOrganization for Standardization,Geneva,ISBN:92-67r-r10188-9(Available onlinefrom https://www.360docs.net/doc/d512309982.html,/en/publications/guides/gum.html).

Jaffey,A.H.,Flynn,K.F.,Glendenin,L.E.,Bentley,W.C.,Essling,A.M.,1971.Precision measurement of half-lives and speci?c of235U and238U.Phys.Rev C4, 1889–1906.

Jerome,S.M.,Eccles,J.L.,2007.General concepts for traceability in environmental radioactivity monitoring.Metrologia44,S133–S139.

Table2

Experimental results for234Uand238U radionuclides compared to recommended values of activity concentration in the pro?ciency test samples organised by IAEA.

Pro?ciency test governing institution Sample name Analyte Measured

(Bqkgà1)

Relative

unc.(%)

Reference value

(Bq kgà1)

Relative

bias(%)

z-score b u-test c A1d A2e P f

(%)

Final

Score

IAEA CU-2010-04 ALMERA Sample2Spiked

tap water

U-234 1.30(11)a8.5 1.30(3)a0.000.000.000.000.298.77A

U-2380.71(5)7.00.70(2) 1.430.030.190.010.147.60A Sample3Spiked

tap water

U-2340.43(3)7.00.47(1)-8.51-0.17-1.260.040.087.29A

U-2380.28(2)7.40.31(1)-9.68-0.19-1.340.030.067.84A

IAEA CU-2010-03Sample2Spiked

tap water U-234 1.23(10)8.1 1.30(3)-5.38-0.11-0.670.070.278.45A U-2380.64(5)7.80.70(2)-8.57-0.17-1.110.060.148.32A

Sample3Spiked tap water U-2340.45(3) 6.70.47(1)-4.26-0.09-0.630.020.087.00A U-2380.30(2) 6.70.31(1)-3.23-0.06-0.450.010.067.41A

IAEA CU-2008-03Sample1U-2340.590(28) 4.80.56(2) 5.360.540.870.030.09 5.94A

U-2380.362(16) 4.40.36(1)0.560.060.110.000.05 5.22A Sample2U-234 1.202(69) 5.7 1.20(4)0.170.020.030.000.21 6.64A

U-238 1.188(69) 5.8 1.25(4)-4.96-0.50-0.780.060.21 6.63A a Combined uncertainty at k?1.

b z score?X expàX ref

?100%,where X ref is the assigned value,X exp is the experimental value and the target standard deviation(σ)is equal to0.10X ref.

c u

score?

X expàX ref

????????????????

ref

tu2exp

p,where U

ref

is the combined standard uncertainty of the reference value,U exp is the combined standard uncertainty of the experimental value.

d Trueness:A1?j X

ref

àX exp j.

e A2?2:58

???????????????????????U2reftU2exp q

f Precision:P?

??????????????????????????????????

U ref

ref

tU exp

exp

?100%:

M.Seferino?lu et al./Applied Radiation and Isotopes94(2014)355–362361

Jia,G.,Belli,M.,Sansone,U.,Rosamilia,S.,Gaudino,S.,2006.Concentration and characteristics of depleted uranium in biological and water samples collected in Bosnia and Herzegovina.J.Environ.Radioact.89,172–187.

Jobbágy,V.,Chmielewska,I.,Kovács,T.,Chalupnik,S.,2009.Uranium determination in water samples with elevated salinity from Southern Poland by micro coprecipitation using alpha spectrometry.Microchem.J.93,200–205. Kanisch,G.,2004.Alpha-spectrometric analysis of environmental samples(IAEA-TECDOC-1401).Quantifying Uncertainty in Nuclear Analytical Measurements.

International Atomic Energy Agency,Vienna,pp.127–139.

Kiliari,T.,Pashalidis,I.,2010.Simpli?ed alpha-spectroscopic analysis of uranium in natural waters after its separation by cation-exchange.Radiat.Meas.45, 966–968.

Kim,C.K.,Martin,P.,Fajgelj,A.,2008.Quanti?cation of measurement uncertainty in the sequential determination of210Pb and210Po by liquid scintillation counting and alpha particle spectrometry.Accredit.Qual.Assur.13,1691–1702.

Liu,H.K.,William, F.G.,Soto,J.,2006.Open-source excels tools for statistical metrology.In:Proceedings of the XVIII IMEKO World Congress,Metrology for

a Sustainable Development,September17–22,Rio de Janeiro,Brazil.

LNHB,2014.Recommended data:(?https://www.360docs.net/doc/d512309982.html,/DDEP_WG/DDEPdata.

htm?).

Miller,A.,1996.Uncertainty of dose measurement in radiation processing.Radiat.

Phys.Chem.47,479–482.Reyes,E.,Marques,S.L.,2008.Uranium series disequilibria in ground waters from a fractured bedrock aquifer(Morungaba Granitoids-Southern Brazil):Implica-tions to the hydrochemical behavior of dissolved U and Ra.Appl.Radiat.Isot.

66,1531–1542.

Spasova,Y.,Pommé,S.,Benedik,L.,W?tjen,U.,2007.Uncertainty budget for226Ra activity concentration in water by alpha spectrometry.Acta Chim.Slov.54, 857–858.

Thakkar,A.H.,2002.A rapid sequential separation of actinides using Eichrom's extraction chromatographic material.J.Radioanal.Nucl.Chem.252,215–218. Vasile,M.,Benedik,L.,2008.On the determination of Ra Po U and U in mineral waters(JRC Scienti?c and Technical Reports).

Vre?ek,P.,Benedik,L.,2003.Evaluation of measurement uncertainty in the determination of210Pb and210Po using beta counting and alpha spectrometry.

Accredit.Qual.Assur.8,134–137.

W?tjen,U.,Benedik,L.,Spasova,Y.,Altzitoglou,T.,Beyermann,M.,2010.EC comparison on the determination of226Ra,228Ra,234U and238U in water among European monitoring laboratories,Appl.Radiat.Isot68,1200–1206.

2nd ed.WHO Guidelines for Drinking Water Quality:Recommendations,vol.1.

WHO,Geneva.

3rd ed.WHO Guidelines for Drinking Water Quality:Recommendations,vol.1.

WHO,Geneva.

4th ed.WHO Guidelines for Drinking Water Quality:Recommendations,vol.1.

WHO,Geneva.

M.Seferino?lu et al./Applied Radiation and Isotopes94(2014)355–362 362