Date: 27 Oct 1998
Place: Dept. of Nuclear Engineering, Universitad Politčcnica de Catalonya

Materials

One randomly selected RadAl™-1/4, S/N 1014, equipped with an RS-232 cable and connected to a laptop computer equipped with program Radal.ht.
One Am²⁴¹ source, reported to be 12 µCi ± 10%
One Cs¹³⁷ source, certified to be 7.11 µCi on 30 Sep 1986, computed to be 5.39 µCi on present date.

Experimental

The instrument was turned on and calibrated on the local background, which resulted to be:
BGD: 2428 ± 35 (average of two readings) expressed as counts over 10 minutes (Standard Count Rate or scr)
The Am source was placed at a distance of 83 cm (est. ± 1 cm) along the vertical to the reference point marked on the instrument. Four readings gave:
12 µCi Am²⁴¹: 2939 ± 27 scr
The Cs source was placed at a distance of 83 cm (est. ± 1) as above. Five readings gave:
5.39 µCi Cs¹³⁷: 3145 ± 25 scr

Results and Calculations

Am²⁴¹ sensitivity, 12 µCi source: 511 ± 44 scr @ 83 cm = 352 ± 30 scr @ 1 m.
Sensitivity is 29.3 ± 5.4 scr/µCi @ 1 m (including calibration source uncertainty).
Cs¹³⁷ sensitivity, 5.39 µCi source: 717 ± 43 src @ 83 cm = 494 ± 30 @ 1m.
Sensitivity is 91.6 ± 5.6 src/µCi @ 1 m
The alarm threshold can be computed based on knowledge of the alarm algorithm. The alarm algorithm triggers the alarm when the measured signal is 6 standard deviations over background. In this case,
Threshold = 296 scr over BGD
Am²⁴¹ alarm threshold (50% level): 296/29.3 = 10.1 µCi @ 1 m, 2.51 µCi @ 50 cm, 0.63 @ 25 cm
Cs¹³⁷ alarm threshold (50% level): 296/91.6 = 3.23 µCi @ 1 m, 0.80 µCi @ 50 cm, 0.20 µCi @ 25 cm
An error of 20% should be placed on the computed thresholds, in consideration of source and statistical errors as well as possible Geiger-Müller tube inter-lot variability.

Conclusions and Comments

The background at location was very high compared to that found at most installation (typically 1300 to 3000 scr). Hence these results are conservative, and better sensitivity should be expected in locations with lower background, as well as by use of lead shielding.
The RadAl-1/4 instrument clearly complies with the FEMA requirements of being able to detect 1 µCi Cs¹³⁷ across a 1 m distance, if used in pairs placed at 1 m distance.
The claim of higher sensitivity to 60 keV vs. 661 keV is supported when converting to dose rate rather than source strength. In first approximation, taking into account the 661 keV to 60 keV gamma ray energy ratio, and a 36% gamma ray conversion efficiency for Am²⁴¹ vs. 85% for Cs¹³⁷, the ratio of signals at equal dose rate becomes (29.3/91.6)*(85/36)*(661/60) = 8.36
In other words, 1 µGi/h (for example) due to Am²⁴¹ will produce a signal 8.4 times higher than the same 1 µGi/h due to Cs¹³⁷. This behavior makes the RadAl-1 optimal for the detection of Special Nuclear Materials and other alpha emitters, which are the most dangerous of all radioactive materials.
Finally, inter-lot variability was not investigated, and further investigations are warranted to assess the importance of this source of uncertainty.