Answer to Question #5410 Submitted to "Ask the Experts"

Many new GM-based dose-rate instruments are available in addition to the standard ion-chamber instruments for gamma personnel dose-rate surveys in an operating power reactor. What instrument issues must be addressed in order to justify the use of GM instruments in a predominately ⁶⁰Co/¹⁶N environment for personnel dose-rate surveys? Most GM instrument manuals do not address ¹⁶N energy response. Instruments would be calibrated to ¹³⁷Cs. Why have ion-chamber instruments traditionally been recommended for personnel dose rates? No one here seems to know the origin of the requirement.

In response to the first part of your question, I would judge that, commonly, the major issues to be considered in justifying the use of Geiger Mueller (GM) detectors in a field dominated by gamma radiations from ⁶⁰Co and/or ¹⁶N are energy dependence of dose response and dose-rate range capability. GM detectors, as you are aware, are event detectors, and the rate at which individual events take place represents a count rate, which is converted to exposure rate or dose rate through the calibration process. There is no inherent property of the detector that establishes a valid correlation between exposure rate or soft-tissue dose rate and output of the detector.

When a detector is calibrated with ¹³⁷Cs at a photon energy of 662 keV, the major interaction mode in the detector is Compton scatter, pretty much regardless of the materials of which the detector is fabricated. Most GM detectors are fabricated of materials that have an atomic number, Z, higher than that of air or soft tissue. The walls are often metallic, and the most common filling gas is argon (e.g., steel with Z=26; Z for argon is 18). The effective atomic numbers for air and water are each close to 7. While the Compton interaction in the GM detector is dominant at 662 keV, this is not necessarily so at lower and higher energies. Thus, as energy decreases the reduced energy along with the elevated Z-values start to favor increased interactions by the photoelectric process, which shows a strong inverse dependence on energy, the cross section decreasing at about the inverse cube of photon energy, and a very strong Z dependence, increasing at about the 4^th to 5^th power of Z. Thus as energies decrease below a few hundred keV, the GM response (count rate per unit exposure rate or per unit soft-tissue dose rate) increases, yielding an inherent excess reading, relative to the actual exposure or dose rate.

On the high-energy end of the spectrum, another differential response is seen that can be attributed to the dominance of the pair production event at high photon energies and to its strong dependence also on atomic number, the pair production cross section increasing with the square of Z. This effect accounts for a common over-response of many GM instruments used in ¹⁶N photon fields, weighted heavily by 6 MeV gamma rays.

Many manufacturers reduce the lower-energy over-response by adding additional metal around the detector to attenuate some of the low-energy photons. This process of energy compensation is usually reasonably effective and may produce a detector that yields exposure rate readings within 20 percent of the true values from roughly 200 keV through energies somewhat greater than the approximate 1.25 MeV of ⁶⁰Co. Responses of such compensated detectors will usually drop off rapidly as energies go below about 100 keV. The detectors will still over-respond at the high energies such as those from ¹⁶N, with excess response of 50 percent to 100 percent being rather common.

Because of the very large gas discharge that accompanies a GM pulse, the time required to clear the detector of sufficient charge to allow the voltage to recover such that a subsequent event may be recorded is quite long. This results in innate limitations on how rapidly the detector may record events and places restrictions on how high a dose rate may be handled by a standard G-M detector. The maximum reliable dose rate varies depending on the design characteristics of the detector and the associated electronics. Some of the older GM detectors had the unfortunate characteristic of reading zero when placed in a saturating field in which the detector could not recover. Newer detectors generally show full-scale readings in such instances. Also, with technology available over the past decade or so, manufacturers have succeeded in using electronic means to provide deadtime compensation so that the ranges of some GM detectors can be extended considerably—e.g., a detector previously capable of a maximum reading of 50 mR h^-1 may be able to read as high as a couple of R h^-1.

Regarding your question about the often-expressed preference for ionization chambers for dose-related measurements in photon fields, this preference stems from the fact that a properly designed ionization chamber exhibits a response (e.g., amperes per unit exposure rate or per unit dose rate) that is directly related to the true exposure rate or dose rate—that is, the rate of collection of charge is equal to the rate of production of primary ionization in the detector volume. This is quite unlike the GM detector in which every output pulse is the same regardless of how much primary ionization was produced in the detector volume.

The proper design of the ionization chamber for exposure or dose measurements can be achieved to an acceptable degree by abiding by requirements of the Bragg-Gray principle, which I shall not discuss in detail here (you can find discussions of this in many references in health physics and instrumentation). One requirement, however, is that the detector wall thickness be at least as great as the range of the most energetic charged particle produced within it. Manufacturers generally try to select a kind of compromise thickness, sufficient to meet the needs for relatively high-energy photons but not so thick as to produce excessive attenuation at lower energies.

Ionization chamber wall materials used for exposure or dose measurements are often of low atomic number, and air is often the fill gas, so that the kinds of photon interactions that occur in the detector are very similar to those that occur in air or soft tissue at the same energies over a wide range of energies. Detector walls of commercial ionization chambers typically have density thicknesses on the order of a few hundred mg cm^-2, sufficient to provide a near equilibrium distribution of secondary electrons over a range of photon energies from less than 100 keV through ⁶⁰Co but never high enough to ensure equilibrium for the high-energy electrons from ¹⁶N photons. As it often turns out, however, in the power reactor environments in which ¹⁶N is a concern, the photons emitted by ¹⁶N in the coolant or steam lines frequently have had the opportunity to interact in the coolant, pipe walls, ambient air, and other materials, in the process producing a distribution of secondary electrons that reasonably simulate an equilibrium distribution so that an air ionization chamber placed in the field will often provide readings that are within about 20 percent of the true exposure-rate or dose-rate values.

I apologize for the lengthiness of this response, but you have asked some good questions that I felt required due consideration. I hope this satisfies your needs.

George Chabot, PhD, CHP