Metrologia 2011 48 Tech. Suppl. 06015

Comparisons of the standards for air kerma of the GUM and the BIPM for 60Co and 137Cs gamma radiation

P. J. Allisy-Roberts, C. Kessler, D.T. Burns Bureau International des Poids et Mesures, F-92312 Sèvres Cedex,

M. Derlaciński, J. Kokociński

Główny Urząd Miar, Miar, Elektoralna 2, Warsaw, Poland

Abstract Direct comparisons of the standards for air kerma of the Główny Urząd Miar (GUM, Poland) and of the Bureau International des Poids et Mesures (BIPM) were carried out in the 60Co and 137Cs radiation beams of the BIPM in 2006. The results, expressed as ratios of the GUM and the BIPM standards for air kerma, are 1.0023 with a combined standard uncertainty of 2.4 × 10–3 in 60Co, and 0.9995 with a combined standard uncertainty of 2.9 × 10–3 in 137Cs. The result in 60Co agrees with the direct comparison carried out in 1996 when the new correction factors adopted by the GUM and the BIPM in 2007 and 2009 are applied for the present comparison.

1. Introduction Direct comparisons of the standards for air kerma in 60Co and 137Cs gamma radiation of the Główny Urząd Miar (GUM), Poland and the Bureau International des Poids et Mesures (BIPM) were carried out in April 2006 in the BIPM 60Co and 137Cs radiation beams. The last direct comparison of the air kerma standards for 60Co gamma radiation was in 1996 [1]. No previous comparison has been made for 137Cs air kerma although some measurements were made with the primary standard in the BIPM 137Cs beam in 1996. The standard for air kerma of the GUM is a cavity ionization chamber constructed at the Orszagos Mérésügyi Hivatal (now known as the Magyar Kereskedelmi Engedélyezési Hivatal – MKEH), Budapest, Hungary in 1983 (type ND 1005, serial number 8303). The main characteristics are given in Table 1. The standards of the BIPM are parallel-plate graphite-walled cavity ionization chambers described in [2, 3].

1/18

Metrologia 2011 48 Tech. Suppl. 06015

Table 1. Characteristics of the GUM standard of air kerma

Type: cylindrical graphite cavity standard ND1005 - 8303 Nominal values Chamber Outer height / mm

Outer diameter / mm 19 19

Inner height / mm Inner diameter / mm

11 11

Wall thickness / mm 4 Electrode Diameter / mm 2 Height / mm 10 Volume Air cavity / cm3

relative uncertainty / cm3 1.013

0.002 Wall Material ultrapure graphite Density / g cm–3 1.71 Impurity fraction < 1.5 × 10–4 Applied tension (both polarities) Voltage / V 250

2. Determination of the air kerma For the BIPM standards and the GUM standard, the air kerma rate is determined from

iaccaen ksgeWmIK Π)/()1)(/)(/( ,,1 ρμ−−=& (1)

where I is the ionization current measured for the mass m of air in the cavity, W is the average energy spent by an electron of charge e to produce an ion pair in dry

air, g is the fraction of electron energy given to radiative processes, ( ) ca,en / ρμ is the ratio of the mean mass-energy absorption coefficients of air and graphite, sc,a is the ratio of the mean stopping powers of graphite and air, ∏ki is the product of the correction factors to be applied to the standard. Physical data and correction factors The values of the physical data used in (1) are consistent with the CCEMRI(I) 1985 recommendations [4]. These values and those for the various corrections needed for 60Co and 137Cs radiation are also shown in Tables 2 and 3, respectively, for both the GUM and the BIPM standards, together with their associated uncertainties. Although the comparison was held in 2006, the results have been updated for the BIPM reference standards adopted in 2007 and 2009 for 60Co and 137Cs respectively [5, 6].

2/18

Metrologia 2011 48 Tech. Suppl. 06015

Table 2. Physical constants and correction factors with their estimated relative uncertainties of the BIPM and GUM standards for the 60Co radiation beams at the BIPM

BIPM standard GUM standard

values uncertainty(1) values uncertainty(1) 100 si

100 ui 100 si 100 ui

Physical Constants

0ρ dry air density(2) / kg m–3 1.2930 0.01 1.2930 0.01

ca,en )/( ρμ 0.9989 0.01 0.04 0.9985 0.05

acs , 1.0010 0.11(3) 1.0011 0.11(3)

eW / J/C 33.97 33.97 g bremsstrahlung loss 0.0031 0.02 0.0032 0.02

Correction factors: kg re-absorption 0.9996 – 0.01 – – – ks recombination losses 1.0022 0.01 0.02 1.0023 0.01 0.02 kh humidity 0.9970 – 0.03 0.9970 – 0.03 kst stem scattering 1.0000 0.01 – 0.9992 0.01 0.01 kwall wall attenuation and scattering 1.0011 – (4) 1.0211(5) 0.01 0.06 kan axial non-uniformity 1.0020 – (4) 1.0000 0.04 – krn radial non-uniformity 1.0015 – 0.02 1.0002 – 0.02 V chamber volume / cm3 6.7967 (6) – 0.08 1.013 – 0.20 I ionization current / pA 0.01 0.02 0.03 0.02 Relative standard uncertainty quadratic summation 0.02 0.15 0.06 0.25 combined uncertainty 0.15 0.26 Relative standard uncertainty neglecting contributions from physical constants and kh quadratic summation 0.02 0.09 0.06 0.21 0.09 0.22 combined uncertainty 0.24

(1) Expressed as one standard deviation si represents the type A relative standard uncertainty estimated by statistical methods, ui represents the type B relative standard uncertainty estimated by other means (2) At 101 325 Pa and 273.15 K (3) Combined uncertainty for the product of acs , and eW /(4) Uncertainties included in the uncertainty of the chamber effective volume (5) See text for Monte Carlo calculated value that has been adopted (6) For standard CH5-1, the measured volume 6.8028 cm3 reduced by the factor 1.0009.

3/18

Metrologia 2011 48 Tech. Suppl. 06015

Table 3. Physical constants and correction factors with their estimated relative uncertainties of the BIPM and GUM standards for the 137Cs gamma radiation beam at the BIPM

BIPM standard GUM standard

values uncertainty(1) values uncertainty(1) 100 si

100 ui 100 si 100 ui

Physical Constants

0ρ dry air density(2) /kg m-3 1.2930 – 0.01 1.2930 – 0.01

ca,en )/( ρμ 0.9990 – 0.05 0.9990 – 0.05

ac,s 1.0104 – 1.0101 –

eW / J/C 33.97 – 0.11(3)

33.97 – 0.11(3)

g bremsstrahlung loss 0.0012 – 0.02 0.0012 – 0.02 Correction factors: ks recombination losses 1.0018 0.01 0.02 1.0019 0.01 0.02 kh humidity 0.9970 – 0.03 0.9970 – 0.03 kst stem scattering 0.9998 0.01 – 0.9986 0.01 0.04 kwall wall attenuation and scattering 1.0278(6) 0.01 0.10 1.0002 0.01 (5)

kan axial non-uniformity 1.0018 – 0.04 1.0000 – 0.01 radial non-uniformity(4) krn 1.0011 0.01 0.10 1.0002 0.01 0.02

chamber volume /cm3 V 6.8283 – 0.08 1.013 – 0.20

I ionization current / pA 0.02 0.02 0.03 0.10 Relative standard uncertainty quadratic summation 0.03 0.19 0.04 0.28 combined uncertainty 0.19 0.28 Relative standard uncertainty neglecting contributions from physical constants and kh quadratic summation 0.03 0.14 0.04 0.25 0.14 0.25 combined uncertainty 0.29

(1) Expressed as one standard deviation si represents the relative standard uncertainty estimated by statistical methods, type A ui represents the relative standard uncertainty estimated by other means, type B (2) At 101 325 Pa and 273.15 K (3) Combined uncertainty for the product of acs , and eW /(4) For the 20 cm diameter 137Cs beam. (5) The non-statistical uncertainty for kwall is included in the determination of the effective volume of the

standard [6]. (6) See text for Monte Carlo calculated value that has been adopted Reference conditions Air kerma at the BIPM is determined under the conditions given in Tables 7 and 11 of [7]: - the distance from source to reference plane is 1 m, - the field size in air at the reference plane is 10 cm × 10 cm for the 60Co beam and 20 cm diameter for the 137Cs beam.

4/18

Metrologia 2011 48 Tech. Suppl. 06015

Reference values The value is the mean of four measurements made over a period of six months before and one after the comparison and is about 1.4 mGy s–1, 11 mGy s–1 and 17µGy s–1 for the Picker, CISBio and 137Cs beams, respectively. By convention they are given at the reference date of 2006-01-01 T 00:00:00 UTC using the half-life value of 1925.21 d, σ = 0.3 d for 60Co and 10976 d, σ = 30 d for 137Cs [8].

BIPMK&

Beam characteristics A comparison of the 60Co and 137Cs beams at the GUM and the BIPM is given in Table 4.

Table 4. Parameters of the 60Co and 137Cs beams at the GUM and the BIPM

Nominal source

activity at 01/01/06

Scatter contribution/

energy fluence

Approximate air kerma rate / (mGy s–1)

Source diameter and length

60Co beam Field size *

GUM 6 TBq 7 mm × 10.4 mm# 21 % 16 cm diameter 0.5

BIPM Picker 20 TBq 14 % 1.4 20 mm × 5.6 mm 10 cm × 10 cm

BIPM CISBio 130 TBq 21 % 10.8 20 mm × 14 mm 10 cm × 10 cm 137Cs beam

GUM 5 TBq 17.6 mm × 13 mm# 25 % 15.5 cm diameter 0.1

BIPM 1 TBq 8.3 mm × 13 mm# 30 % 20 cm diameter 0.02 * at 1 m # active dimensions 3. Correction factors for the GUM standard With the exception of kwall, krn and ks, the correction factors for the GUM standard were determined at the GUM. Attenuation and scattering in the chamber wall (kwall) and axial non-uniformity (kan)

The correction for kwall has previously been made at the GUM by measuring the ionization chamber response as a function of the wall thickness using build-up caps and then extrapolating linearly to zero wall thickness. The value obtained this way is then multiplied by a calculated value of kcep. This method of extrapolation is understood to underestimate the effect of the chamber walls on attenuation and scatter of the beam but the GUM is not yet in a position to calculate kwall with Monte Carlo (MC) methods. Such a calculation for their primary standard has been made by the BIPM and confirms other values for the wall correction identified in this way for a graphite density of 1.71 g cm–3 of 1.0211 (5) and 1.0278 (8) for 60Co and 137Cs beams, respectively [9]. Investigations of Büermann et al [10] strongly support the use of calculated wall and axial non-uniformity corrections and this has been endorsed by the CCRI [11]. Consequently, the GUM has declared that they have adopted the BIPM calculated values for the wall effects of their primary standards. However, there is a difference between the GUM measured value of the graphite density of the build-up caps (1.71 g cm–3), taken as being equivalent to the graphite of the chamber itself, and the value determined by the MKEH for the chamber that was constructed in 1983

5/18

Metrologia 2011 48 Tech. Suppl. 06015

(1.75 g cm–3). The GUM measured values were confirmed at the BIPM by measurements of the four build-up caps, which indicated they were from the same piece of graphite as the agreement was better than 0.2 %. These measurements imply that the graphite used for the construction of the standard was perhaps not the same as for the build-up caps, as believed at the time. It is important to note that the application of MC calculated correction factors rather than those obtained using the extrapolation method have lead to a relative increase in the air kerma response of the GUM cavity chambers by 0.99 × 10–2 for 60Co γ-rays and 1.15 × 10–2 for 137Cs γ-rays. These values assume a graphite density of 1.71 g cm–3; calculations for a density of 1.75 g cm–3 may increase these values by a further 5 × 10–4. Consequently, the mean calculated value is used by the GUM and the Type B uncertainty for the wall correction is increased to take the density uncertainty into account. Since the date of the comparison, the BIPM has also completed MC calculations for its own primary standards and the new reference values have been approved by the CCRI and adopted [5, 6] Radial non-uniformity of the beam (krn)

The correction factor krn, for the radial non-uniformity of the BIPM beams over the cross-section of the GUM standard, has been estimated from measurements carried out at the BIPM. The values are included in Tables 2 and 3 for the 60Co and 137Cs beams. Recombination loss (ks)

The air kerma rates at the GUM and the BIPM are significantly different, as shown in Table 4, so the corrections for losses due to recombination, ks, were also measured at the BIPM. The results are presented in Figure 1 and the corrections are consistent with the value of 1.0019 (3) measured at the GUM in their 137Cs beam. For the recombination measurements, the ratio of the ionization currents with applied voltages of 250 V and 80 V (using both polarities) was measured for four different air kerma rates (using both 60Co beams and a set of brass filters; this is permissible as recombination is insensitive to the spectrum). Applying the method of Niatel and the notation given in [12],

VnVV IVgmnVAnII )()1()1(1 222 −+−+= (2)

where, Iv is the uncorrected, measured current at the normally applied voltage, V n is any number, not necessarily an integer A is a constant dependent on the chamber type m2 is the volume recombination parameter for ionization chambers g is the geometrical factor dependent on the chamber shape The current, IV, is the current as measured by the chamber, not corrected for decay and not normalized for temperature and pressure. Figure 1 illustrates the measurements made for n = 250/80 = 3.125. The recombination correction ks can be expressed from Figure 1 as

VIkkk volinits 1 ++= (3)

6/18

Metrologia 2011 48 Tech. Suppl. 06015

and Table 5 gives the values and uncertainties for kinit and kvol. Consequently, a correction factor of 1.0023 (2) for ion recombination at 250 V was applied to the GUM standards in the BIPM reference CISBio beam. The appropriate value in the Picker beam is 1.0020 (2). This former value is used in Table 2. The equivalent value for the 137Cs beam used in Table 3 is 1.0019 (2).

Figure 1. Recombination measurements made at the BIPM

for the GUM standard ND1005-8303

I250 / pA

0 100 200 300 400 500

I 250

/ I 80

1.003

1.004

1.005

1.006

1.007

1.008

1.009

ND1005-8303

Table 5. Results of ion recombination measurements made at the BIPM

for the GUM standard

Standard uncertainty GUM Standard ND1005-8303 Correction

19.1 × 10–4 3 × 10–5 Initial recombination and diffusion, kinit

Volume recombination coefficient, kvol, / pA–1 9.5 × 10–7 2 × 10–7

ks in the BIPM Picker beam, BIPM value 1.0020 2 × 10–4

ks in the BIPM CISBio beam, BIPM value 1.0023 2 × 10–4

ks in the BIPM 137Cs beam, BIPM value 1.0019 2 × 10–4

7/18

Metrologia 2011 48 Tech. Suppl. 06015

Polarity effect (kpol)

The collecting voltage applied to the GUM standard was 250 V using both polarities. The chamber was left for 30 min after each voltage change to allow it to stabilize before each measurement. The polarity effect, determined as the ratio of positive and negative currents, was measured as 1.0034 (1) in the 60Co beam (kpol = 1.0017). This value agrees with the polarity effect measured at the BIPM in 1996. At the GUM, only the positive polarity is used normally and the polarity correction factor applied at the GUM implies a polarity effect of 1.0032. No polarity corrections were foreseen in the present comparison at the BIPM as both polarities were used on each occasion. Leakage correction

The raw ionization current measured with the GUM standard was corrected for the leakage current. It is important to note that the GUM standard had a leakage current that was related to both the air kerma rate and its recent radiation history when positive polarity was applied to the high-voltage electrode of the chamber. In 1996, no significant leakage was detected in the first series of measurements, which were performed in the low air kerma rate 137Cs beam, but a similar relative radiation-induced leakage was measured immediately after irradiation in the 60Co beam in which the air kerma rate was about 100 times greater. In the present comparison, the measurements were performed alternately in the three beams, having started in the old Picker 60Co reference beam where the relative correction was about 7 × 10–4. For the subsequent measurements in the lower air kerma rate 137Cs beam, this correction was initially 3 × 10–2 increasing to 6 × 10–2 during the first 3 series; however, as an additional effect was noted in this beam, two further sets of measurements were made after the chamber had been measured in the other beams. This investigation, related to the polarity effect, is described in a later paragraph. 4. Comparison of the air kerma standards for 60Co and 137Cs radiation The values of the ionization current measured by the GUM standard and used to determine the air kerma rate in the three BIPM beams are given in Table 6. These values are for both polarities, corrected for leakage and for decay from the measurement date to the reference date in each case of 2006-01-01, 0 h UTC. The currents are also normalized to the reference conditions of air temperature 273.15 K and pressure 101.325 kPa. Three independent measurements were made with the GUM standard in each beam.

Table 6. Ionization currents measured with the GUM standard at the BIPM

100 s† GUM ND1005 8303 standard; current / pA Mean values

BIPM Picker 60Co beam 53.775 53.781 53.768 53.775 < 0.01 BIPM CISBio 60Co beam 410.345 410.423 410.365 410.377 < 0.01 BIPM 137Cs beam‡ 0.6429 0.6424 0.6435 0.6429 < 0.06

† relative statistical standard uncertainty of the measurements ‡ mean of the measured values for each polarity.

8/18

Metrologia 2011 48 Tech. Suppl. 06015

In Table 6, the first three measurement series made in the 137Cs beam are shown. The polarity effect of these three measurements was significantly different for each set of measurements and was not the same as in the 60Co beams. Consequently, two further sets of measurements were made. It became apparent that the current measured when positive polarity was applied is a function of a differential leakage effect. The maximum leakage measured in the 137Cs beam was only 57 fA but this is a significant proportion (about 10 %) of the measured ionization current for this volume of ionization chamber in this beam. Figure 2 shows the leakage-corrected ionization current (with positive polarity) as a function of the leakage current.

Figure 2. Ionization current measurements made at the BIPM as a function of leakage for the GUM standard ND1005-8303

leakage current / pA

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07

I +250

V /

pA

0.640

0.641

0.642

0.643

0.644

0.645

0.646

0.647

0.648

0.649

0.650

Extrapolating to zero leakage current indicates an ionization current of 641 fA. It should be noted that the current measured when negative polarity is applied remained constant to within 5 × 10–4, while the current with positive polarity applied had a relative standard deviation of 3.3 × 10–3, if no extrapolation is made. As no cause could be identified for this behaviour of the chamber, and the GUM normally measures just a few fA leakage current with a higher activity 137Cs source, they agreed to investigate the leakage at the GUM laboratory, following the comparison. The results of this investigation showed no significant leakage at the GUM. At the BIPM, the decision was made, in agreement with the GUM, to extrapolate the current (positive polarity) to zero leakage as making a leakage correction actually increased the measurement uncertainty. The current proposed for the comparison was the mean of the negative and the extrapolated positive polarity measurements. An uncertainty for the extrapolated value was deduced from the slope of the fit to the differential leakage measurements. However, in view of this unexplained behaviour, it was decided to use the negative polarity measurements and apply the kpol correction measured in the BIPM 60Co beam. An uncertainty of 1 × 10−3 was included to account for the unexpected behaviour with positive polarity.

9/18

Metrologia 2011 48 Tech. Suppl. 06015

5. Comparison results and discussion The comparison result is given by,

RK. = K

.GUM / K

.BIPM , (4)

where K.

is the value of the air kerma rate at the BIPM as measured by the GUM and BIPM standards, respectively. The results are given in Table 7 together with their uncertainties.

Table 7. Results of the comparisons of the GUM standard for air kerma

BIPMK&GUMK&100 uc RK Beam

/mGy s–1 /mGy s–1 60Co CISBio 10.884 10.859 1.0023 0.24

137Cs * 17.24 × 10–3 17.25 × 10–3 0.9995 0.29

Result of the check in the 60Co Picker beam

60Co Picker 1.4259 1.4221 1.0027 0.30

* result using negative polarity measurements only, with a polarity correction, see text. As some constants (such as air density, W/e, ρμ en , g, sc,a and kh) are derived from the same basic data in both laboratories, the uncertainty in RK is due only to the uncertainties in the correction factors, the volumes of the standards, the measured ionization currents and the distance to the source, the values of which are given in the final rows of Tables 2 and 3. The relative standard uncertainty arising from the positioning of each chamber at the BIPM is less than 10–4. Each air kerma rate value measured using the GUM standard in Table 7 is derived from the mean of each measurement series in Table 6 using the volume in Table 1 and the physical constants and correction factors given in Tables 2 and 3. The value is taken as the mean of the four measurements made around the period of the comparison for each beam. Air kerma rates were verified immediately before the comparison measurements. The values refer to an evacuated path length between source and standard and are given at the reference date of 2006-01-01, 0 h UTC, as are those measured using the GUM standard.

&KBIPM

&KBIPM

Since 2003, each NMI has been encouraged by the Consultative Committee for Ionizing Radiation (CCRI) [13] to verify its correction factors and to publish any changes to its national standards that it feels are appropriate so that the results may be included in the BIPM key comparison database (KCDB). All the previous results of air kerma comparisons in 60Co at the BIPM have been re-evaluated [14], taking into account the effect of changes being made in national standards following the recommendations of the CCRI and of changes to the BIPM standard itself [5]. In May 2007, the CCRI(I) approved an overall change in the BIPM

10/18

Metrologia 2011 48 Tech. Suppl. 06015

air-kerma in the CISBio reference beam by a factor of 1.0054. The combined relative standard uncertainty on the air-kerma determination is now evaluated as 1.5 × 10–3. As indicated in [14], the reference beam for air kerma comparisons at the BIPM is the CISBio 60Co beam since the characterization was completed and the values adopted as the reference in 2007.

The measurements in the Picker beam were used as a check. The ratio of the air kerma rates determined previously in 1996 by the GUM and BIPM standards in the BIPM Picker 60Co beam was 0.9987 (28). When the updates to each of the standards are applied retrospectively to these earlier results, the agreement between 1996 and the latest Picker result in Table 7 is within 4 × 10–4. In May 2009, the BIPM presented to the CCRI(I) the re-evaluation of the air kerma standard for the 137Cs beam, which results in an increase of 3 parts in 103 of the reference air kerma value; the change was approved and was adopted in September 2009. This change has been included in the comparison result of 0.9995 (29) given in Table 7. In 1996, although no 137Cs comparison result was reported for the GUM standard, possibly due to lack of agreement at that time over the stopping power ratio that is appropriate for the GUM standard, measurements had been made in the 137Cs beam at the BIPM. A comparison result of 0.9918 (33) can be deduced from the measurements made in 1996 which becomes 1.0002 using the present correction factors. This agrees within the uncertainties with the 2006 comparison value in Table 7. Several other national laboratories have also made 137Cs comparisons with the BIPM. It is of note that the air kerma determinations in a 137Cs beam made by the national metrology institutes have recently undergone re-evaluation to take account of changes to the correction factors in particular. Once all the results have been re-evaluated, they will be the subject of a summary report and the results will be placed in the KCDB under the comparison identifier BIPM.RI(I)-K5 [15]. 5. Degrees of equivalence

Comparison of a given NMI with the key comparison reference value

Following a decision of the CCRI, the BIPM determination of the dosimetric quantity, here KBIPM

(KB), is taken as the key comparison reference value (KCRV), for each of the CCRI radiation qualities [16]. It follows that for each NMI i having a BIPM comparison result R,i (denoted xi in the KCDB) with combined standard uncertainty ui, the degree of equivalence with respect to the reference value is given by a pair of terms:

the relative difference Di = (Ki – KBi)/ KBi = RKi – 1 (5)

and the expanded uncertainty (k = 2) of this difference, Ui = 2 ui. (6)

The results for Di and Ui, are expressed in mGy/Gy. Table 8 gives the values for Di and Ui for each NMI, i taken from the BIPM.RI(I)-K1 values published in the KCDB and this report, using (5) and (6), and forms the basis of the entries in

11/18

Metrologia 2011 48 Tech. Suppl. 06015

the KCDB of the CIPM MRA. These data are presented graphically in Figure 3 where the black square indicates a result that dates prior to 1998 although the ARPANSA has made a comparison quite recently. The results of three published regional metrology organization (RMO) comparisons are also included [17, 18, 19]. It is of interest to note that the previous result of the GUM participation in the EUROMET comparison was prior to the update of their primary standard, using Monte Carlo corrections for the wall effects in particular. Although that result agreed with the KCRV within the expanded uncertainties at the time, the agreement using the updated standard has improved, and the uncertainties have also decreased.

Comparison of any two NMIs with each other

The degree of equivalence between any pair of national measurement standards is expressed in terms of the difference between the two comparison results and the expanded uncertainty of this difference; consequently, it is independent of the choice of key comparison reference value.

The degree of equivalence, Dij, between any pair of NMIs, i and j, is thus expressed as the difference

(7) jijiij RRDDD −=−=

and the expanded uncertainty (k = 2) of this difference, Uij = 2 uij, where

( ) ( )∑∑ −−+=k

jkkk

ikkjcicij ufufuuu 2corr,

2corr,

2,

2,

2 (8)

and the final two terms are used to take into account correlation between the primary standards, notably that arising from the physical constants and correction factors for similar types of standard. For example a number of national primary standards have a similar shape and size to the GUM standard [14] for which the wall correction factors are strongly correlated. As yet, no correlation has been assumed for the volume estimations of identically shaped standards

The results for Di and Ui are given in Table 8 in the format in which they appear in the key comparison database. The values are also given for Dij and Uij although following a decision of the CCRI(I) in 2011, these values no longer appear in the KCDB. Note that the data presented in the table, while correct at the time of publication of the present report, become out-of-date as NMIs make new comparisons. The formal results under the CIPM MRA [20] are those available in the key comparison database.

12/18

Metrologia 2011 48 Tech. Suppl. 06015

6. Conclusion The GUM standard for air kerma in 60Co gamma radiation compared with the present BIPM air kerma standard gives a comparison result of 1.0023 (0.0024). This compares favourably with other primary standards for which the wall correction factor has been calculated using Monte Carlo methods. All the comparison results of the national metrology institutes (NMIs) and designated laboratories are used as the basis of the entries in the KCDB set up under the CIPM MRA to which the comparison result of the GUM has now been added. The 137Cs air kerma standards of the GUM and the BIPM have been compared for the first time. The result for this comparison, RK (137Cs) = 0.9995 (29), which agrees within the uncertainties with (unpublished) measurements made at the BIPM in 1996. The result will be included in the KCDB once all the other NMI results have been updated similarly for the Monte Carlo calculations of the wall effects.

13/18

Metrologia 2011 48 Tech. Suppl. 06015

Table 8. Degrees of equivalence with the BIPM.RI(I)-K1 and RMO.RI(I)-K1 comparisons for the GUM participation

• Key comparison BIPM.RI(I)-K1

MEASURAND : Air kerma

The key comparison reference value is the BIPM evaluation of air kerma. The degree of equivalence of each laboratory i with respect to the reference value is given by a pair of terms both expressed in mGy/Gy:

Di and Ui, its expanded uncertainty (k = 2), with Ui = 2ui.

(See Final Report for the computation of Di)

The degree of equivalence between two laboratories is given by a pair of terms both expressed in mGy/Gy:

Dij = Di - Dj and Uij, its expanded uncertainty (k = 2).

The approximation for Uij is explained in the Final Report.

• Linking an RMO.RI(I)-K1 to BIPM.RI(I)-K1

MEASURAND : Air kerma

The value xi is the comparison result for laboratory i participant in RMO.RI(I)-K1 having been normalized to the value of the linking laboratories (see RMO.RI(I) Final reports in the KCDB). The degree of equivalence of each laboratory i participant in the relevant RMO.RI(I)-K1 with respect to the reference value is given by a pair of terms both expressed in mGy/Gy:

Di and Ui, its expanded uncertainty (k = 2).

See the appropriate RMO.RI(I)-K1 Final Report for the computation of Di and the approximation used for Ui in the Matrix of equivalence. The degree of equivalence between two laboratories i and j, one participant in BIPM.RI(I)-K1 and one in RMO.RI(I)-K1, or both participant in an RMO.RI(I)-K1, is given by a pair of terms both expressed in mGy/Gy:

Dij = Di - Dj and Uij, its expanded uncertainty (k = 2). The approximation for Uij is given in the relevant RMO.RI(I)-K1 Final Report [17, 18, 19].

14/18

Metrologia 2011 48 Tech. Suppl. 06015

15/18

Table 8 continued. Degrees of equivalence with the BIPM.RI(I)-K1 and SIM.RI(I)-K1, COOMET.RI(I) and EUROMET.RI(I) comparisons for the GUM participation

Di Ui Dij Uij Di Ui Dij Uij Lab i

/ (mGy/Gy) / (mGy/Gy)

Lab i

/ (mGy/Gy) / (mGy/Gy)

ARPANSA -2.6 6.4 -4.9 7.3 CIEMAT -1.5 3.9 -3.8 6.0

DMDM 2.5 3.6 0.2 4.8 CMI -5.8 14.1 -8.1 16.0

NMIJ 1.8 4.8 -0.5 5.7 SSM 1.0 7.5 -1.3 8.7

NIM -0.8 3.6 -3.1 5.1 STUK -2.3 7.3 -4.6 8.6

LNE-LNHB -1.9 5.4 -4.2 5.5 NRPA 5.1 7.1 2.8 8.4

ENEA -0.3 5.2 -2.6 6.0 SMU 5.2 6.5 2.9 7.6

VSL -1.5 4.4 -3.8 5.4 IAEA 0.0 7.5 -2.3 8.7

MKEH 5.5 4.4 3.2 5.5 HIRCL 4.2 11.9 1.9 12.7

GUM 2.3 4.8 0.0 0.0 BIM -4.5 13.0 -6.8 13.7

NPL 1.1 7.6 -1.2 8.3 ITN -0.4 6.0 -14.8 17.9

NRC 3.2 5.6 0.9 6.6 PTB 8.4 3.4 -2.7 7.3

BEV 3.4 4.2 1.1 5.3 METAS -1.3 4.6 6.1 5.2

VNIIM 0.8 3.6 -1.5 4.9 NIST 3.4 6.4 -3.6 6.5

KRISS -0.5 3.2 -2.8 4.7 LNMRI 2.4 13.7 1.1 7.5

CNEA 1.8 10.0 0.1 14.4

ININ 3.2 16.4 0.9 16.9

BELGIM 12.5 21.8 10.2 22.2

CPHR 1.1 9.7 -1.2 10.4

RMTC -3.6 9.7 -5.9 10.4

BIPM.RI(I)-K1

SIM.RI(I)-K1 (2002)

COOMET.RI(I)-K1 (2006)

EUROMET.RI(I)-K1 (2005 to 2008)

All national metrology institute acronyms are available in the KCDB.

Metrologia 2011 48 Tech. Suppl. 06015

16/18

BIPM.RI(I)-K1, SIM.RI(I)-K1 (2002), COOMET.RI(I)-K1 (2006) and EUROMET.RI(I)-K1 (2005 to 2008)Degrees of equivalence with the KCRV for air kerma in 60Co

-35

-25

-15

-5

5

15

25

35AR

PAN

SA

DM

DM

NM

IJ

NIM

LNE-

LNH

B

ENEA

VSL

MKE

H

GU

M

NPL

NR

C

BEV

VNIIM

KRIS

S

ININ

BelG

IM

CPH

R

RM

TC

CIE

MAT CM

I

SSM

STU

K

NR

PA

SMU

IAEA

HIR

CL

BIM

ITN

PTB

MET

AS

NIS

T

LNM

RI

CN

EA

Di

/ (m

Gy

/ Gy)

Figure 3. Graph of degrees of equivalence with the KCRV

BIPM.RI(I)-K1

SIM.RI(I)-K1 COOMET.RI(I)-K1 EUROMET.RI(I)-K1

Metrologia 2011 48 Tech. Suppl. 06015

References [1] Allisy-Roberts P. J., Boutillon M., Referowski Z., Paz N., 1997, Comparison of the

standards of air kerma of the GUM and the BIPM for 60Co γ rays, 1997, Rapport BIPM-97/2.

[2] Boutillon M., Niatel M.-TA., 1973, Study of a graphite cavity chamber for absolute measurements of 60Co gamma rays, Metrologia, 9, 139-146.

[3] Boutillon M., Allisy-Roberts P.J., 1996, Measurement of air kerma and ambient dose equivalent in a 137Cs beam, Rapport BIPM-96/7, 12 pp.

[4] BIPM, Constantes physiques pour les étalons de mesure de rayonnement, BIPM Com.Cons. Etalons Mes. Ray. Ionizants, Section (I), 1985, 11, p. R45 (Paris: Offilib).

[5] Burns D.T., Allisy P.J., Kessler C., 2007, Re-evaluation of the BIPM international standard for air kerma in 60Co gamma radiation, Metrologia, 2007, 44(6), L53-L56.

[6] Kessler C., Burns D.T., Allisy-Roberts P.J., Re-evaluation of the BIPM standard for air kerma in 137Cs gamma radiation, Metrologia, 2009, 46(5), L24-L25.

[7] Allisy-Roberts P.J., Burns D.T., Kessler C., 2009, Measuring conditions used for the calibration of National Ionometric standards at the BIPM, Rapport BIPM-2009/04, 20 pp.

[8] Bé M.-M., Chisté V, Dulieu C., Browne E., Baglin C., Chechev V., Kuzmenco N., Helmer R., Kondev F., MacMahon D., Lee K.B., 2006, Table of Radionuclides (Vol. 3 – A = 3 to 244) Monographie BIPM-5.

[9] Rogers, D.W.O., Treurniet, J., Monte Carlo calculated wall and axial non-uniformity corrections for primary standards of air kerma, 1999, NRCC Report PIRS-663, 31 pp.

[10] Büermann L., Kramer H.-M. And Csete I., Results supporting calculated wall correction factors for cavity chambers, 2003, Phys. Med. Biol., 48, 3581-3594.

[11] Comité Consultatif des Rayonnements Ionisants, 1999, The estimation of katt, ksc kCEP and their uncertainties, CCRI, 16, 145-146.

[12] Boutillon M., 1998, Volume recombination parameter in ionization chambers, Phys. Med. Biol., 1998, 43, 2061-2072.

[13] Comité Consultatif des Rayonnements Ionisants, 2004, Corrections to air kerma standards, 18th meeting (2003), 21-23.

[14] Allisy-Roberts P.J., Burns D.T., Kessler C., 2007, Summary of the BIPM.RI(I)-K1 comparison for air kerma in 60Co gamma radiation, Metrologia, 2007, 44, Tech. Suppl., 06006

[15] Allisy P.J., Burns D.F., Kessler, C., Summary report of the BIPM.RI(I)-K5 137Cs dosimetry comparisons, Metrologia, 2009, 46, Tech. Suppl., 06XXX (in preparation)

[16] Allisy P.J., Burns D.R., Andreo P., 2009, International framework of traceability for radiation dosimetry quantities, Metrologia, 2009, 46(2), S1-S8

[17] Ross C.K., Shortt K., Saravi M., Meghzifene A., Tovar V.M., Barbosa R.A., da Silva C.N., Carrizales L., Seltzer S.M., 2008, Final report of the SIM 60Co air-kerma comparison SIM.RI(I)-K1, Metrologia, 2008, 45, Tech. Suppl., 06010

17/18

Metrologia 2011 48 Tech. Suppl. 06015

[18] Büermann L, Oborin A V, Dobrovosky J, Milevsky V S, Walwyn Salas G, Lapenas A, 2009, COOMET.RI(I)-K1 comparison of national measurement standards of air kerma for 60Co γ radiation, Metrologia 2009, 46, Tech. Suppl. 06015

[19] Csete I, Leiton A G, Sochor V, Lapenas A, Grindborg J-E, Jokelainen I, Bjerke H, Dobrovodsky J, Megzifene A, Hourdakis C J, Ivanov R Vekic B, Kokocinski J, Cardoso J, Buermann L, Tiefenboeck W, Stucki G, van Dijk E, Toni M P, Minniti R, McCaffrey J P, Silva C N M, Kharitonov I, Webb D, Saravi M, Delaunay F, 2010, Report on EUROMET.RI(I)-K1 and EUROMET.RI(I)-K4 (EUROMET project no. 813): Comparison of air kerma and absorbed dose to water measurements of 60Co radiation beams for radiotherapy, Metrologia, 2010, 47, Tech. Suppl., 06012

[20] CIPM MRA: Mutual recognition of national measurement standards and of calibration and measurement certificates issued by national metrology institutes, International Committee for Weights and Measures, 1999, 45 pp. http://www.bipm.org/pdf/mra.pdf.

18/18