Volume 115, Number 1, January-February 2010 Journal of Research of the National Institute of Standards and Technology [J. Res. Natl. Inst. Stand. Technol. 115, 7-13 (2010)] Comparison Between the NIST and the KEBS for the Determination of Air Kerma Calibration Coefficients for Narrow X-Ray Spectra and 137 Cs Gamma-Ray Beams Volume 115 Number 1 January-February 2010 Michelle O’Brien, Ronaldo Air kerma calibration coefficients for a the said facility at the time of the Minniti reference class ionization chamber from measurements. The comparison of the narrow x-ray spectra and cesium 137 calibration coefficients is based on the National Institute of Standards gamma-ray beams were compared between average ratios of calibration coefficients. the National Institute of Standards and and Technology (NIST), Technology (NIST) and the Kenya Bureau Gaithersburg, MD 20899 of Standards (KEBS). ANISTreference- USA class transfer ionization chamber was calibrated by each laboratory in terms of Key words: air kerma; free-air ionization and the quantity air kerma in four x-ray chamber; gamma-ray, cesium 137; primary reference radiation beams of energies standard; reference radiation qualities; Stanslaus Alwyn Masinza between 80kVand 150kVand in a x-ray calibration. cesium 137gamma-ray beam. The Kenya Bureau of Standards reference radiation qualities used for this (KEBS), comparison are described in detail in the Nairobi, Kenya ISO4037 publication.[1] The comparison began in September Accepted: January 20, 2010 2008 and was completed in March 2009. The results reveal the degree to which the participating calibration facility can [email protected] demonstrate proficiency in transferring air [email protected] kerma calibrations under the conditions of Available online: http://www.nist.gov/jres 1. Introduction 2. Procedure The objective of this proposal was to compare the air 2.1 Object of Comparison kerma calibration coefficients of a reference class chamber determined in reference beams of x rays and The object was to compare the value of the air kerma gamma rays. The comparison was performed between calibration coefficient of the chamber determined at the the National Institute of Standards (NIST) and the KEBS with the value obtained at the NIST. The trans- Kenya Bureau of Standards (KEBS). ANISTreference- fer chamber used during the comparison was an class transfer ionization chamber was calibrated by Exradin1chamber model A5, SNXY051605. each laboratory in terms of the quantity air kerma in four ISO narrow x-ray spectra beams with energies between 80kV and 150kV and in a 137Csgamma-ray 1Certain commercial equipment, instruments, or materials are iden- tified in this paper to foster understanding. Such identification does beam. The reference radiation qualities used for this not imply recommendation or endorsement, nor does it imply that the comparison are described in detail in the ISO4037 equipment identified are necessarily the best available for the publication [1]. purpose. 7 Volume 115, Number 1, January-February 2010 Journal of Research of the National Institute of Standards and Technology 2.2 TransferChamber described in ISO4037 as the narrow spectra [1]. The NISTand the KEBS HVL’s are listed in Table 2. The chamber used is spherical and made of C552 air equivalent plastic and has a nominal volume of Table 2. Characteristics of reference radiation qualities 100cm3. The air kerma calibration coefficient N was K determined by evaluating the ratio NK=Kair/Icorr, NIST KEBS where K is the reference air kerma rate (expressed in Reference Generating tube potential air units of Gy/s) and I is the measured ionization Radiation (kV) corr Half-value layer (mm Cu) current (expressed in units of C/s) corrected for influ- ence quantities. Abias voltage of 400volts was applied NS80 80 0.59 0.56 to the chamber. This is the maximum voltage allowed NS100 100 1.14 1.07 by the electrometer owned by the KEBS. Table 1 NS120 120 1.76 1.67 NS150 150 2.41 2.34 summarizes the chamber specifications. Cs-137 Not Applicable 10.8 10.9 Table 1. Description of the chamber The gamma-ray measurements were made in the Type Serial Sensitive Outside Diameter Chamber NIST 137Cs horizontal beam gamma-ray calibra- Number volume Diameter of inner Voltage tion facility. The air kerma rates in this facility were (nominal (nominal) electrode obtained by applying a decay correction to the refer- ence air kerma rate. The reference air kerma rates were ExradinA5 XY051605 100 cm3 6 cm 6 mm 400V determined using a suite of six graphite wall cavity ionization chambers. The beam is collimated providing 2.3 Reference Radiation Qualities a circular beam size of 15cm radius at a source-to- detector distance of 195cm. The beam is uniform over The measurements made at the NIST were per- the area across the chamber used in this work. Various formed in the x-ray and gamma-ray beam calibration types of reference class ionization chambers are used facilities. The air kerma rates at both of these facilities routinely to track the reference air kerma rate value. are well established and are determined using the NIST primary standard instruments. Details on the determi- 2.4.1 Reference Conditions at the NIST nation of the air kerma rates using the primary standard instruments and the NIST calibration service can be At the NIST each x-ray calibration was made by found in Refs. [2-5]. alternating between the transfer chambers and the For the x-ray measurements, the air kerma rates were standard free-air chamber, through the translation of the determined directly using the Wyckoff-Attix free-air chambers to one beam center line. Beam alignment to chamber. This primary standard for x rays is located in the axis is accepted to be 0.1mm and reproducible to the NIST300kVcalibration facility. The x-ray source better than 0.01mm, as observed by an alignment used at the time of the comparison was a 320kVx-ray telescope. The reference plane for all measurements at generator with a metal ceramic x-ray tube, both sup- the NIST was positioned at 1000mm from the x-ray plied by Seifert-Pantak.1 The x-ray generator is a radiation source and 1950mm from the 137Cs high-frequency, highly stabilized voltage source. The source. The beam diameter in the reference plane was tungsten anode x-ray tube, model MXR-321, has a at least 90mm for the x-ray measurements and at least beryllium window of thickness 3mm and a focal spot 150mm for the gamma-ray measurements. The exact 8.0mm in diameter. The materials used for the filtra- beam geometry was not prescribed in the protocol so tion and for the measurement of Half Value Layer that the KEBS could use the appropriate geometry (HVL) were at least 99.99% pure with thicknesses depending on the field size and field uniformities known with an uncertainty of 0.01mm. The reference available and the conditions that best represents routine radiation qualities used for the comparison are calibrations at their facility. The influence of field size was not investigated for this comparison. The geometry 1Certain commercial equipment, instruments, or materials are iden- and conditions of the air kerma measurement were kept tified in this paper to foster understanding. Such identification does consistent at NIST for all measurements. Table 3 not imply recommendation or endorsement, nor does it imply that the describes the radiation reference field sizes used for the equipment identified are necessarily the best available for the comparison. purpose. 8 Volume 115, Number 1, January-February 2010 Journal of Research of the National Institute of Standards and Technology standard free-air chamber, through the translation of the Table 3. Beam geometry used at both facilities chambers to the central beam axis indicated by the laser beam. The details of the KEBS standard can be found Geometry NIST KEBS NIST KEBS X-ray spectra 137Cs in Ref.[6]. The x-ray apparatus at KEBS consists of several elements placed between the x-ray tube and the Calibration distance (mm) 1000 2000 1950 1950 detection distance. Starting at the x-ray source these Beam diameter (mm) 90 853 150 682.5 elements include a diaphragm/aperture wheel holder at 200mm from the source, a filter wheel holder at At the NIST the chamber background current was 220mm, a permanent tube inherent filter holder at measured before and after the chamber was exposed to 230mm, a monitor chamber at 240mm, and a HVL the source of radiation. The average of the background filter holder. The HVL diaphragm holder is only used currents was determined and subtracted from the for HVLmeasurements and is placed at 630mm from measured ionization current produced in the chamber the tube. The diaphragm holder is remotely adjustable when exposed to the source of radiation. Typical back- from the control unit. The filters can be selected from ground currents were of the order of 0.01% of the the control unit. The comparison ionization chamber ionization current. The integration times used at the was placed at the measuring distance of 2000mm from NIST were 120seconds for the x-ray measurements the source. Details of the KEBS laboratory setup can be and 60seconds for the gamma-ray measurements. The found in Ref.[7]. measured ionization currents were normalized to The gamma-ray measurements were made in the 295.15K and 101325Pa. The humidity was monitored KEBS 137Cs gamma-ray calibration laboratory. The air and recorded and was typically in a range of 30%, kerma rates in this facility are obtained by applying a however the currents were not corrected for humidity. decay correction to the reference air kerma rates that Only the negative polarity measurements were used for were established on July 2007, at the time the facility the comparison results, resulting in a negative was commissioned. These daily air kerma rate values charge.collected on the electrometers. No correction are tracked using a laboratory developed spreadsheet was applied to the calibration coefficient to account for and confirmed through routine measurements using polarity effect since at both facilities the same polarity ionization chambers models LS01 and LS10 as part of was used. The NIST performed measurements at both the quality control of the facility. It is important to men- polarities to verify the influence of the polarity effect. tion that the air kerma standards at KEBS are traceable The value of the calibration coefficients obtained with indirectly to the Bureau International des Poids et both polarities agreed within less than 0.10% for all Measures (BIPM) standards through the International reference beams used. This implies that the polarity Atomic Energy Agency (IAEA). The reference air effect is negligible. No corrections were applied to the kerma rates at KEBS were determined using the KEBS calibration coefficients to account for electronic recom- secondary standard chamber model W-32002, SN 247 bination effects. However an estimate of the recombi- calibrated at the International Atomic Energy Agency nation correction was obtained using the two-voltage (IAEA). The value of the most recent 137Cs calibration method. For this, the ratio of the ionization currents coefficient of the KEBS standard chamber obtained measured at the full voltage of 400 Vand half voltage from IAEA is 2.52×104Gy/C. This value includes a of 200Vwas obtained. correction introduced by the BIPM in the year 2009 The KEBS was requested to provide the calibration that applies to all calibration facilities that are directly coefficients for the transfer chambers in terms of air and indirectly traceable to the BIPM, including KEBS. kerma per charge in units of Gy/C and refer to standard Furthermore, all results presented in this work and conditions of air temperature at 295.15K, pressure at obtained at KEBS, include all corrections introduced 1013.25hPa and 50% relative humidity. The KEBS by the BIPM up to the date of this publication. was requested to use the same polarity for the voltage The KEBS gamma beam geometry uses a collimated bias of the Exradin chamber. circular beam with a radius of approximately 340mm at a source-to-detector distance of 1950mm. The beam is uniform over the area across the chamber. At a 2.4.2 Reference Conditions at the KEBS distance of 1000mm, the variation of the beam intensi- X-ray calibrations for each beam quality were made ty across a circular area of 2850mm does not exceed by alternating between the transfer chambers and the 5% when the largest collimator is used. 9 Volume 115, Number 1, January-February 2010 Journal of Research of the National Institute of Standards and Technology The NIST chamber was connected to the PTW Table 4. Uncertainties associated with the NIST standard for the x-ray spectra electrometer. Prior to performing the measurements, a voltage was applied to the chamber overnight to allow NISTstandard sufficient time for the chamber to stabilize. Relative standard uncertainty u u iA iB Background currents were measured before and after Ionization current 0.0008 0.001 the chamber was exposed to the source of radiation. Volume 0.0004 0.0001 The average of the background currents was deter- Positioning 0.0001 mined and subtracted from the measured ionization Correction factors (excl. kh) 0.0012 0.0024 Humidity k 0.0003 current produced in the chamber when exposed to the h Physical constants 0.0015 source of radiation. Typical background currents were ⋅ of the order of 0.02% of the ionization current. The K 0.0015 0.003 Standard integration time used at the KEBS was 60seconds for 0.003 both the x-ray and the gamma-ray measurements. The ⋅ measured ionization currents were normalized to uncertainty for the air kerma rate, K , from the Standard 295.15K and 101325Pa. The humidity was monitored 137Cs gamma ray beam is 0.0029 (as listed in Table 7 and recorded and was typically in a range of 40%, ahead). This results in relative combined uncertainties however the currents were not corrected for humidity. of the air kerma rates of 0.30% and 0.29% for x rays The calibration coefficients for the transfer chamber and gamma rays respectively when expressed as a per- were given in terms of air kerma per charge in units of centage. Table 5 and Table 6 show the uncertainty Gy/C and corrected to standard conditions. analysis associated with the calibration of the transfer chamber at the NISTand the KEBS using x rays. The 2.5 Course of Comparison uncertainty of the calibration is given by the uncertain- ty of the calibration coefficient N . Similarly Table 7 K The NIST shipped the chamber to the KEBS in and 8 show the uncertainty analysis associated with the September 2008 with a tentative scheduled return to the calibration of the chamber at both facilities in the NISTof November 2008. However, due to unexpected 137Cs beam. Note that in Tables 7 and 8, the relative delays at the KEBS resulting from equipment failure uncertainties of the chamber current corrected for influ- the measurements were completed in March 2009. ence quantities and the air kerma rate value provided After completion of the calibrations at the KEBS facil- by the standard, are combined, resulting in the relative ity and the return of the chamber to the NIST, constan- standard uncertainty of the calibration coefficient cy checks were performed in late March and early April N . Finally, Table 9 lists the relative combined uncer- K of 2009. The values of the calibration coefficients tainty of the comparison ratio, KEBS/ NIST, obtained obtained at the NIST before and after the shipment of both with x-rays and gamma-rays. The uncertainty of the chamber to the KEBS agreed as expected, indicat- the calibration coefficients N for the x-ray portion K ing that there were no alterations or damage to the of the comparison listed in Table 9 are grouped in chamber during the full period of the comparison. Type A and Type B evaluations, whereas the uncer- tainties for the gamma ray calibration coefficients are expressed directly as the relative combined 3. Uncertainties uncertainties. The uncertainties associated with the NIST primary x-ray standard are listed in Table 4. Following standard Table 5. Uncertainties associated with the NISTcalibration of the A5 Exradin transfer chamber in the x-ray beam guidelines for the expression of uncertainity published elsewhere [8,9], the uncertainties listed in the tables in NISTstandard this work are classified in Type Aand Type B accord- Relative standard uncertainty u u ing to the method used to evaluate them. Type Auncer- iA iB ⋅ tainties, denoted as u , are determined from statistical iA KStandard 0.0015 0.0030 analysis while Type B uncertainties, denoted as u , are Ionization current 0.00083 0.0010 iB determined by other means such as scientific judgment. Positioning 0.0001 As shown in the table, the relative fractional combined Humidity 0.0003 ⋅ 0.0015 0.0032 uncertainty of the air kerma rate, K , from the Standard N 0.00354 x-ray beam is 0.0030. Similarly, the relative combined K,NIST 10 Volume 115, Number 1, January-February 2010 Journal of Research of the National Institute of Standards and Technology Table 6. Uncertainties associated with the KEBS calibration of the A5 Exradin transfer chamber in the x-ray beam KEBS Relative standard uncertainty u u iA iB KEBS Reference chamber N 0.00200 K,reference Long term stability 0.00052 Positioning 0.00290 Ionization current 0.00210 Temperature and pressure 0.00003 Exradin chamber at KEBS Positioning 0.00290 Temperature and pressure 0.00070 Ionization current 0.00667 N 0.00732 0.00410 K,KEBS 0.00839 Table 7. Uncertainties associated with the NISTcalibration of the A5 Exradin transfer chamber in the 137Cs gamma-ray beam NIST Relative standard uncertainty u u iA iB Charge 0.0010 0.0010 Time 0.0005 Air density correction (temperature and pressure) 0.0003 Distance 0.0002 k , loss of ionization due to recombination 0.0001 0.0005 sat Probe orientation 0.0001 Humidity 0.0006 0.0010 0.0014 Chamber current corrected for all influence quantities 0.0017 ⋅ K 0.0029 Standard N 0.0034 K,NIST Table 8. Uncertainties associated with the KEBS calibration of the A5 Exradin transfer chamber in the 137Cs gamma-ray beam KEBS Relative standard uncertainty u u iA iB Stability of reference value K at 2m 0.0024 0.0000 air Temperature and pressure 0.0003 0.0015 Leakage current 0.0000 0.0100 0.0024 0.0.0101 Chamber current corrected for all influence quantities 0.0104 ⋅ K 0.0078 Standard N 0.0130 K,NIST 11 Volume 115, Number 1, January-February 2010 Journal of Research of the National Institute of Standards and Technology Table 9. Uncertainties associated with the comparison results X-rays Cs-137 Relative standard uncertainty u u u iA iB i N 0.0073 0.0041 0.0130 K,KEBS N 0.0015 0.0032 0.0034 K,NIST N /N 0.0075 0.0052 0.0134 K,KEBS K,NIST ⋅ ⋅ K /K 0.0091 0.0134 KEBS NIST 4. Results and Discussion The mean value of the calibration coefficients NK Table 10. Mean values of the calibration coefficients measured at measured at the NIST and the KEBS are shown in the NISTand the KEBS using the Exradin A5 Table10. The average air kerma rates used at both facilities are listed in Table11. The comparison results Calibration coefficients are listed in Table12 as the ratio of the mean calibra- 105GyC–1 tion coefficients, N /N for each reference Reference radiation NIST KEBS K,KEBS K,NIST radiation with an associated comparison uncertainty Mean NK Mean NK listed in Table9. The mean ratio of the comparison NS100 2.969 2.886 ratios for the x rays is 0.984. The corresponding NS80 2.929 2.898 standard deviation of the four comparison ratios is NS100 2.969 2.886 0.0083. Acceptable agreement is also achieved for the NS120 2.999 2.961 NS150 2.991 2.955 gamma-ray comparison with a resulting ratio of 1.006 Cs-137 3.047 3.065 with a standard uncertainty of the ratio of 0.0134. The differences observed in the values of the calibra- tion coefficients, resulting in agreement of approxi- mately 1.6% for xrays and 0.6% for gammarays, can Table 11. Average air kerma rates used at both facilities be due to the variations in the field sizes and source to detector distances used at both facilities. Although the Average air kerma rate beam uniformity and scatter characteristics at NIST Reference radiation 10–6Gys–1 meet the conditions as described in the ISO4037[1] , Beam NIST KEBS measurements at the NIST, using the ExradinA5 cham- ber at various distances and field sizes, result in differ- NS80 6.61 3.74 ences of the calibration coefficient of up to 1%. Since NS100 4.08 2.87 NS120 4.50 1.59 this comparison was not designed to compare beam NS150 3.53 12.9 geometry, the uncertainty due to differences in field Cs-137 26.0 3.31 sizes and scatter conditions was not directly addressed. The polarity effect and the electronic recombination effect were excluded as possible influences to the comparison results. The differences due to the polarity Table 12. Results of comparison of the comparison chamber bias voltage were measured to have a negligible effect for the specific chamber Reference radiation Mean volume and the rates used for the comparison. The air N /N kerma rates used at both laboratories were similar, K,KEBS K,NIST resulting in probable similar electronic recombination NS80 0.990 effects. The estimate of the recombination effect NS100 0.972 obtained using the two-voltage method resulted in NS120 0.987 NS150 0.988 negligible influences. The uncertainty analysis did not Cs-137 1.006 include these negligible components. 12 Volume 115, Number 1, January-February 2010 Journal of Research of the National Institute of Standards and Technology 5. References [1] ISO/IS4037-1: 1996 X and gamma reference radiations for calibrating dosimeters and dose rate meters and for determining their responses as a function of photon energy—Part1.: Radiation characteristics and production methods. [2] P. Lamperti and M. O’Brien, Calibration of X-Ray and Gamma-Ray Measuring Instruments, NISTSpecial Publication 250-58 (2001). [3] C. M. O’Brien, NIST Quality Manual, Ionizing Radiation Division QM-II, Procedure 3.Calibration of x-ray radiation detectors. Gaithersburg, MD, NIST(2004). [4] R. Minniti, NISTQuality Manual, Ionizing Radiation Division QM-II, Procedure 4.Calibration of gamma-ray radiation detec- tors. Gaithersburg, MD, NIST(2004). [5] R. Minniti, H. Chen-Mayer, S. M. Seltzer, S. M. Huq, L. Bryson, T. Slowery, J. A. Micka, L. A. DeWerd, N. Wells, W. F. Hanson, and G. S. Ibbott, Med Phys 33,1074 (2006). [6] IAEA, Austria, Calibration of Radiation Protection Monitoring Instruments, STI/PUB/1074 (January 2000). [7] A. S. Masinza, KEBS Laboratory Procedure, MET-LP-19/17, X-Calibration Procedures Quality Control in X-Ray, (2008). [8] B. N. Taylor and C. E. Kuyatt, Guidelines for evaluating and expressing the uncertainty of NISTmeasurement results, NIST Technical Note 1297, Washington, DC: U.S. Government Printing Office (1994). [9] International Organization for Standardization (ISO) Guide to the Expression of Uncertainty in Measurement, Geneva, Switzerland:ISO (1993). About the authors: Michelle O’Brien and Ronaldo Minniti are physicists in the Ionizing Radiation Division, Radiation Interactions and Dosimetry Group of the NIST Physics Laboratory. Mrs. O’Brien, an expert in x-ray dose traceability, maintains the NIST primary x-ray standards, while conducting compar- isons between other national measurement institutions and facilitating x-ray dose measurement quality stan- dards domestically. Dr Minniti, an expert in gamma- ray dose traceability, maintains the NIST gamma-ray measurement capabilities, while conducting compar- isons between other national measurement institutions and facilitating dose measurement quality standards domestically. Alwyn Stanslaus Masinza is a member of the Radiation Dosimetry Laboratory at KEBS where the national standards in the field of Ionizing Radiation are maintained. It is a Secondary Standard Dosimetry Laboratory (SSDL) and was established in 2007 as a result of the collaboration of International Atomic Energy Agency (IAEA) and the Government of the Republic of Kenya through the project KEN/6/016. 13