Gross counting is used to determine the presence of absence of radioactivity in an object. Once it has been determined that radiation is present above the ambient background levels, it is necessary to determine the specific radionuclides that are present. Specific gamma emitting radionuclides are determined by gamma spectroscopy, for example. This allows the person monitoring or searching for radiation to determine if it is natural, medical or industrial radionuclides or a radioactive source that should be detained and further investigated.
Gross counting measurements are used to screen samples for relative levels of radioactivity. Gross counting is generally performed first to determine the presence or absence of either gamma-ray or neutron radiation. Both portable and stationary systems first measure the background radiation as a count rate. Background may be measured for a preset number of seconds or a continuous moving average of the background may be calculated. If a gross count rate measurement is determined to be above the previously measured background level, an alarm is issued as a visible, audible and/or vibration indication.
Data Acquisition Rates – Over-sampling
Systems used for search or continuous monitoring for homeland security applications must be able to find radioactivity in very short times compared to laboratory counting measurements. Typically, measurements are made in less than a few seconds an in the case of moving objects, one second or less.
Over-sampling the counting interval provides fewer missed alarms. In the picture below, two data reporting frequencies are shown, 1 second and 100 milliseconds. If a vehicle, person or other object is moving in front of the detector, it may only be present for a second or less as shown in the previous table and graph of speed versus measurement time. If the system is sampling at 1 second intervals and the interval exactly corresponds to time and the object is in front of the detection system, then an alarm would occur. If, on the other hand, the 1 second sampling interval occurs such that one ends and another begins just as the object is centered on the detector, then 50% of the counts will be in the first interval and 50% in the second interval. This situation may result in the net counts not exceeding the alarm threshold setting. To overcome this possibility, the sampling interval can be reduced to a 100 milliseconds or less. We can then integrate the counts in each sampling interval such that the sum of 10 intervals equals a second of data. 100 milliseconds later we sample again and subtract the data from the oldest interval and add the new counts to the total. This moving average ensures that no more than 10% of the counts will be missed and in general, less than this amount.
Over-sampling, by rapid polling of the counting data, thereby reduces missed alarms while allowing the sensitivity and precision of the measurements to be retained.
When a source is moving, the length of time it is in front of a radiation monitor and its detectors. The table here shows the distance traveled in feet per second (fps) related to the speed in miles per hour (mph). As 88 fps corresponds to 60 mph, a 10cm, 30cm and 50 cm wide sensor would have a source in front of them only 4, 11 and 19 thousandths of a second or milliseconds (ms) respectively. At more reasonable speeds such as those required in the ANSI N42.35 standard of 5mph, a vehicle would be moving 7.3 fps. In this case the time in front of a detector would be 45, 134 and 224 ms respectively. Although some radiation would hit the detector before and after the source the was directly in front of the detectors, these times are minimum to collect data. For moving sources, the over-sampling data acquisition of a system is essential. Nucsafe was the first commercial company to implement this method during the IAEA ITRAP assessments.
|3He 10cm = 3000cm2 5000cm2|
|2″ tube with 1″ moderator on each side|
The graph displays the data in the table and indicates that there is a large change in the available measurement time especially at low speeds. Ideally systems measuring moving objects should sample at no more than 100ms.