Answer to Question #8446 Submitted to "Ask the Experts"
The following question was answered by an expert in the appropriate field:
In gamma spectroscopy, we dismiss most of the stuff in the "noise/background" region of the on-screen display. Unless a specific peak emerges, nothing in this region is of any interest. Since most nuclides decay by beta emission, how do we account for the initial beta particle? Since photoelectric and Compton scattering predominate, how can we be so dismissive of the junk on the low end of the spectrum?
Wouldn't it seem logical that the "junk" represents beta-initiated x rays and the continuum of photoelectric and Compton reactions?
The "proof" for me is that the "noise" region is always huge when compared against even the largest peak in a highly radioactive sample and that it tends to be proportional (peak-height to peak-height comparisons) to the sample under any condition.
The real question is, What is the average photon energy in a commercial nuclear plant? Unless we factor in the low-energy stuff, we can't suitably answer that question. If we dismiss the low-energy stuff as "noise," are we doing a disservice to the field of radiation protection?
A related issue is, What is the average photon energy in the plant since, once again, the predominant means of decay are beta particles and gammas predominantly interact along the photoelectric and Compton continuum, which interact with piping systems?
Your question is not a trivial one and, unfortunately, the answer is not simple. As you have inferred, the pulse height distribution (PHD) that might be associated with exposure of a gamma-ray detector, such as a germanium detector, used with a multichannel gamma analyzer in a reactor environment, is the result of a number of possible events. Primary photons, emitted by contaminants, that deposit their full energy in the detector produce pulses in the PHD that are associated with the typical photopeaks that are used for radionuclide identification and quantification. When primary photons interact with materials before reaching the detector, their energies are degraded and the Compton scattered photons that are incident on the detector produce pulses that fall below the photopeak energies and, in theory, may range from near zero energy up to the Compton edge energy. Additionally, the beta radiation that you note may interact within reactor components and produce bremsstrahlung radiation that, in turn, may intercept the detector and produce a continuum of pulses that become superimposed on the PHD and add to the pulses observed in the low-energy portion of the PHD. To complicate things more, when any photon enters the detector it may undergo various possible interactions. If the full energy gets deposited in the detector, photopeak counts arise (although there may not be enough of a given energy to make the photopeak visible), but if the photon undergoes Compton scatter in the detector, the Compton photon may deposit all or a portion of its energy in the detector, thus leading to more partial energy events and adding to the continuum of counts observed in the PHD. All of these secondary scattered and bremsstrahlung photons, produced outside or inside the detector, account for much of the distribution of pulses that are seen in the low-energy portion of the PHD. How much of the “junk” is real junk and how much is good stuff incidentally occupying the “junk” bins?
When multiple photopeaks associated with a number of primary photons are part of the PHD, there will typically be an entire continuum of Compton-induced pulses that extend from each photopeak into the low-energy portion of the PHD; these are superimposed one upon the other, all adding to the gross continuum of pulses. Add to these other events—such as pair production interactions of high-energy photons inside and outside the detector that lead to single and double escape peaks and annihilation photon peaks, sum peaks associated with cascade gamma rays from a specific gamma emitter and sometimes with simple coincidence counts when the field intensity is high, characteristic x rays produced in detector-associated components, and backscatter events leading to distorted peaks in the PHD (most notable if a shield is used around the detector)—and the PHD may be extremely complex.
In order to use this PHD to assess the true average energy of photons in the reactor environment, it would be necessary to weed out all of the non-full energy deposition pulses that originated from partial-energy deposition events that occurred in the detector—e.g., if a Compton-scattered photon entered the detector, did it undergo a partial-energy deposition in the detector? If so, the associated count should be eliminated. Similarly, any Compton-induced pulse produced by a primary photon interaction in the detector should be rejected. Similar restrictions would apply to pulses from bremsstrahlung radiation. The reality is that such analysis is virtually impossible using simply the information obtained from the PHD. The analysis is possible through applications of the Monte Carlo method. Such analysis requires extensive modeling of the detector and the radiation source(s). For in-plant measurements, the source modeling could be very complex. The Monte Carlo simulations are time consumptive and relatively costly but can be successfully applied to problems such as this. Naturally, if the source term can be properly modeled, it is also possible and easier to perform simulations to evaluate directly the photon energies that would contribute to dose at a given location, independent of any detector response.
Given that many reactor facilities do not employ methods such as Monte Carlo simulations to evaluate photon energies expected at various locations, we can make some cursory but not complete judgments about whether the lack of information about the effects of secondary photons such as Compton-scattered photons and bremsstrahlung is important. As you have observed, most of the radioactive species, both fission products and activation products, produced in power reactors are beta emitters, and many of these emit characteristic gamma radiation during their decay. When a light-water reactor is operating, the dose rate in containment is typically dominated by the short-lived 16N produced by an (n,p) reaction on 16O. The 16N decays by beta emission with the associated emission of high-energy gamma rays (weighted average energy of 6.15 MeV). As you know, reactor entries during operation are highly restricted and most personnel exposures occur during various maintenance procedures when the reactor is shut down. Following shutdown and decay of 7.1 second 16N, the major radionuclides present are (1) activation products that have been produced by neutron activation of reactor components and ions dissolved in the reactor coolant (some of these get deposited throughout the reactor systems/components that have been in contact with reactor coolant water) and (2) fission products and possibly uranium and transuranic nuclides produced from reactor fuel during operation of the reactor. The prevalence of fission products and transuranic species outside of the reactor core depends on the history of cladding integrity of the fuel and on the extent of tramp uranium that might have contaminated fuel cladding. The distribution of gamma rays that might be observed in containment depends on the specific radioactive species that are present in accessible reactor system components. Because many of these species are beta emitters that are often contained within reactor components (e.g., inside reactor coolant pipes and valves) the beta radiation is often attenuated in such components, and this attenuation may result in the production of bremsstrahlung x rays, which appear as a continuum of photon energies with a maximum energy equal to the maximum beta-particle energy. Since bremsstrahlung yield is proportional to electron energy, high-energy beta emitters will yield relatively more (and more penetrating) bremsstrahlung radiation than will low-energy beta emitters. In all cases, however, the bremsstrahlung prevalence in a distribution produced by a single beta emitter may be expected to increase with decreasing energy.
In light-water reactors that have not had a history of leaky fuel, a common dominant activation product is 60Co, produced by activation of 59Co that comes from dissolution of small amounts of cobalt alloy metal used in some reactor coolant components. The maximum beta-particle energy from 60Co is a bit greater than 0.3 MeV. The bremsstrahlung radiation from the complete stopping of 60Co beta radiation will be rather low in yield and in effective energy (likely less than 0.1 MeV) and will be attenuated quite effectively by reactor piping, valve casings, and the like, with the net result that its contribution to personnel dose is probably negligible. In reactors that have leaky fuel problems, there is potential for the buildup of higher-energy beta emitters, such as 144Pr, the daughter of 144Ce, that can lead to greater yields and higher-energy bremsstrahlung radiation. However, even in these cases, the contribution of the bremsstrahlung to the total dose is generally very small relative to the direct gamma dose from the many gamma emitters that would also be present.
For many exposure situations, the major gamma contributors to personnel dose are relatively high-energy gamma emitters, in the 1+ MeV region, and these moderate- and higher-energy photons tend to scatter in the forward direction. Certainly, during reactor operation the dominance of 16N is clear, and scattered photons tend to be scattered in the forward direction and maintain a large fraction of their initial energy. This is not to say, however, that multiple scattering events do not occur, and some photons do get notably degraded in energy. When an individual is working close to a reactor component contaminated with fission products and/or activation products, many of the secondary photons contributing to dose will have undergone only a single scattering event with a rather small reduction in energy. Keep in mind, also, that the photon exposure and dose per unit photon fluence increase with increasing energy above about 60 keV, making the higher-energy photons often dosimetrically more important than the lower-energy photons.
In conclusion, deconvolution of the photon PHD obtained with a typical spectrometric system to determine the complete distribution of photons present in the measured radiation field is possible with specialized techniques (Monte Carlo simulations) but often not practical. For many exposure situations the determination of the frequencies and energies of the primary gamma radiation is sufficient to characterize the work environment for dosimetry purposes. It is always recommended that dose measurements be made in any radiation field that workers enter and in which they might receive a significant dose. Such (photon) measurements are often made with an acceptable ionization chamber that has a flat energy response over the range of energies of interest. Such a chamber measures dose (rate) or exposure (rate) from both the primary and secondary photons that might be present. A worthwhile exercise is to use the PHD to evaluate the fluence rates and respective energies of photons for which photopeaks are evident. (This requires knowledge of the detector efficiency, preferably as a function of incident photon energy and direction, along with the directional characteristics of the incident photons.) With such information one can then determine the expected dose rate or exposure rate at the detector location and compare this to the value obtained with an ionization chamber or other good photon dose measuring instrument. Acceptable agreement would validate an assumption that, from a dosimetry standpoint, the secondary photons are not contributing a significant amount to dose.
As a final note I should mention the possibility of using a tissue equivalent proportional counter (TEPC) to obtain more information about the energy characteristics of the photon field of interest. A number of reactor facilities have purchased TEPCs to obtain neutron energy information, but they are also useful for photon measurements. The TEPC is a specially designed low-pressure proportional chamber that can be purchased commercially. The PHD obtained from the TEPC is used to generate the LET distribution (linear energy transfer distribution) produced by photon interactions in the detector. Such a distribution can be used to interpret the average energy of the electrons set free by the photon interactions, the effective value of the quality factor and, with some assumptions, the average energy of the photons.
You raised a good issue, and there is more that could be discussed, but I have already gone on longer than I probably should have. Good luck.
George Chabot, PhD, CHP