Table 4a:

Uncertainty budgets for the SIRTI measurement of the ¹⁸F ampoules

Uncertainty contributions due to	Comments	Evaluation method	Relative standard uncert. × 104
18F to 94Nb meas. ratio including live-time, background, decay corrections and threshold setting	Standard uncertainty of the weighted mean of 33 measurements, taking into account the correlation due to the 18F half-life	A	2.4
Long-term stability of the SIRTI	Weighted standard deviation of 85 series, each series consisting of 10 measurements	A	0.3
Nb-l bias from the mean since 2007	Relative difference from the mean	B	1.8
Effect of decay on the live-time correction	Maximum measurement duration evaluated from [12]	B	< l
SIRTI drift at high count rate	Mean possible drift over all 18F measurements at the NIST.	B	2
Ampoule dimensions	From the IRMM report [13] and sensitivity coefficients from Monte-Carlo simulations	B	2*
Ampoule filling height	Solution volume is 3.6(l) cm3; sensitivity coefficients from Monte-Carlo simulations	B	2*
Solution density	Between l g/cm3 and l.0l g/cm3 as requested in the protocol; sensitivity coefficients from Monte-Carlo simulations	B	0.7
Measurement time	l s uncertainty	B	l
Relative combined standard uncertainty			3.9

^*Included in the type A uncertainty of the measurements of two ampoules of ¹⁸F