Introduction
In accelerator facilities, nuclear reactions are induced by collisions of high-energy accelerated particles with targets, beam losses due to collimators, and unexpected beam losses on the beamline, causing materials to become locally radioactive (radioactivation). Subsequently, among the particles produced by the above nuclear reactions, neutrons mainly activate the surrounding materials over a wide area. When an accelerator facility is decommissioned, the radioactivated materials must be decontaminated and treated as radioactive waste. Because concrete floors and walls cannot be easily detached, decontamination requires immense labor. In addition, the large volume of concrete has a significant impact on the cost as well as labor in decommissioning. Therefore, it is important to investigate the activation characteristics of concrete for each type of accelerator facility in preparation for the future decommissioning of accelerator facilities. Specifically, proton therapy facilities, which are rapidly becoming popular worldwide, are one of the most important accelerator facilities that should be investigated for their activation characteristics.
A proton therapy facility is a cancer treatment facility where high-energy proton beams are directly irradiated to malignant tumor sites in a patient. The proton beam has a physical property of energy deposition called the Bragg peak, which allows it to concentrate lethal damage on a malignant tumor. The depth position in the human body where damage is inflicted can be adjusted using proton energy. Although the treatment principles are the same, there are two main types of proton therapy facilities with different accelerators. One is the “cyclotron type,” which uses a cyclotron as the proton accelerator, and 61 facilities are in operation worldwide [1]. The other is the “synchrotron type,” which uses a synchrotron as the proton accelerator, and 32 facilities are in operation worldwide [1]. A significant difference between the two types lies in the energy control methods for the proton beam. In the cyclotron type, the acceleration energy is fixed, and the energy of the protons irradiating the patient is adjusted by a degrader. In contrast, in the synchrotron type, the accelerator can adjust the acceleration energy.
In our previous study [2], the activation characteristics of cyclotron-type proton therapy facilities, namely the National Cancer Center Hospital East located in Kashiwa, Chiba, Japan and Aizawa Hospital located in Matsumoto, Nagano, Japan, were investigated. A typical proton therapy facility consists of the following four rooms, starting from the upstream side of the proton beam: (1) the “accelerator room,” where the proton accelerator is located; (2) the “beam transport line room,” where the proton beam is transported; (3) the “gantry room,” where the gantry changes the direction of the proton beam; and (4) the “treatment room,” where the proton beam irradiates the patient. In the cyclotron-type proton therapy facilities, the beamlines around the degrader in the accelerator room are strongly radioactivated locally due to beam losses. The neutrons produced by the nuclear reactions in and around the degrader radioactivate the entire concrete in the accelerator room. The predicted specific activity levels of the concrete after 30 years of operation were close to the clearance level. However, in the beam transport line, gantry, and treatment rooms, the predicted specific activity levels of the concrete after 30 years of operation were negligible compared to the clearance level.
Herein, as a comparison study with cyclotron-type proton therapy facilities, we investigated specific activity levels of concrete at two synchrotron-type proton therapy facilities: the Medipolis Proton Therapy Research Center (Kagoshima, Japan), hereafter abbreviated as “MPTRC,” and Proton Medical Research Center, University of Tsukuba Hospital (Ibaraki, Japan), hereafter referred to as “PMRC.” The accelerator rooms in the synchrotron-type proton therapy facilities are larger than those in the cyclotron-type facilities. The larger concrete volume of the synchrotron-type proton therapy facilities has a greater potential impact on decontamination. The most important radionuclides in concrete during decommissioning are 152Eu (half-life: 13.542 years [3]) and 60Co (half-life: 5.2714 years [3]) [2], produced by thermal neutron capture reactions. In this study, the specific activities of 152Eu and 60Co were predicted using the following procedure at the synchrotron-type proton therapy facilities, as described in a previous study [2], for the cyclotron-type proton therapy facilities. First, the residual radioactivity on the beamline, which is an indicator of beam loss, was surveyed directly with γ-ray detectors. Next, the thermal neutron fluence rates on the concrete floor and walls were measured, primarily around the beam loss locations. Finally, from the obtained thermal neutron fluence rates, we predicted the specific activities of 152Eu and 60Co in the concrete after 30 years of operation and examined whether the concrete in the synchrotron-type proton therapy facility would be activated.
Materials and Methods
1. Investigated Facilities
1) MPTRC
At MPTRC, proton therapy for patients was initiated in 2011. Specific information on the proton therapy facility at MPTRC is summarized in Table 1. Fig. 1A shows a horizontal sectional view of the proton therapy facility at MPTRC. The green arrows in Fig. 1 indicate the proton beam streams. The proton therapy facility at MPTRC is a typical synchrotron-type proton therapy facility consisting of four types of rooms in order from the proton beam upstream: the accelerator room, beam transport line room, gantry room, and treatment room. All the rooms are located on the same floor. A synchrotron that can accelerate up to 235 MeV of energy was installed as an accelerator in the accelerator room. The protons are accelerated to the desired energy (120, 150, or 210 MeV) by the synchrotron and delivered to a patient in the treatment room. The conventional broad-beam irradiation method (the wobbler and spread-out Bragg peak methods) combined with respiratory synchronization is applied to proton therapy.
The MPTRC has three treatment rooms, each equipped with a rotating gantry. During one treatment, the other rooms are ready for subsequent treatments, saving time and allowing more patients to be treated. Treatment rooms 1, 2, and 3 are used almost equally. Five treatment days per week are typically scheduled. According to the operating records from January 4, 2018 to December 29, 2019, during the period of regular treatment, an average of 50 patients were treated with 60 irradiations per day.
As shown in Fig. 1A, a radiation monitor for neutrons (nA-1; Fuji Electric Co., Ltd.) and a radiation monitor for γ rays (γA-1; Fuji Electric Co., Ltd.) are installed in the accelerator room. By analyzing these radiation levels, the operation time of the accelerator (the time from the start to the end of a day’s operation) and the fluctuation of the beam power can be indirectly determined. Our previous study [2] confirmed a correlation between the dose rate of the radiation monitor for neutrons and that of the radiation monitor for γ rays. As the dose rate values of the radiation monitor for neutrons showed a tail even after the accelerator was stopped as a detector characteristic, the dose rates of the radiation monitor for γ rays were used to analyze the accelerator operation history. Analysis of the dose rates of the radiation monitor for γ rays from December 1, 2019 to December 28, 2019 showed an average of 12.6 hours of operation time per day. The analysis results of the beam power fluctuations were used to correct the measured neutron fluence rates obtained from the radioactivity, which are described later because they are affected by the beam power fluctuations.
2) PMRC
Proton therapy was initiated at the PMRC in 2001. An overview of the proton therapy system at PMRC is shown the study by Umezawa et al. [4]. Specific information on the proton therapy facility at PMRC is summarized in Table 1. Fig. 1B shows a horizontal sectional view of the proton therapy facility at PMRC. The green arrows in Fig. 1 indicate the proton beam streams. As in the proton therapy facility at MPTRC, the proton therapy facility at PMRC also consists of four typical room types. All rooms are located on the same floor. A synchrotron capable of accelerating protons up to 250 MeV is located in the accelerator room. Protons are accelerated by the synchrotron to the desired energy (180, 200, or 250 MeV) and irradiate the patient in the treatment rooms. The conventional broad-beam irradiation method (the passive scattering method [5]) is applied to proton therapy.
The proton therapy facility at PMRC has two treatment rooms, each equipped with a rotating gantry. Treatment rooms 1 and 2 are used alternately for efficient treatment. Thus, the two rooms are used nearly equally. Typically, treatment is provided 5 days per week. Based on the operation records from July 27, 2018 to August 23, 2018, an average of 39 patients were treated with 95 irradiations per day.
As shown in Fig. 1B, a radiation monitor for neutrons (DAM-251; Hitachi Aloka Medical Ltd.) and a radiation monitor for γ rays (DAM-131; Hitachi Aloka Medical Ltd.) are installed in the accelerator room. As in the analysis at MPTRC, the dose rates of the radiation monitor for γ rays were used to analyze the accelerator operation history. From the analysis of the monitors from July 27, 2018 to August 23, 2018, the average operation time was 10.3 hours per day. The analysis results of the beam power fluctuations were used to correct the beam power fluctuations in the measured neutron fluence rates obtained from the radioactivity.
2. Survey for Residual Radioactivity in the Beamlines
The residual radioactivity in the beamline after accelerator operation indicates the magnitude of the proton beam loss [6]. Except for accelerator components that are intentionally bombarded with proton beams, most radioactivation occurs in stainless steel beam ducts. Therefore, the relative contact γ-ray dose rates on the beamlines from the residual radionuclides are a reference for the magnitude of the beam loss. As the beam losses in the beamlines originally produce the neutrons that radioactivate the concrete, it is important to determine the location of the beam loss throughout the accelerator. Therefore, we measured the contact γ-ray dose rates on the beamline with survey meters. In addition, γ-ray spectrometry was performed to identify the source radionuclides contributing to the contact γ-ray dose rates.
1) MPTRC
At MPTRC, the contact γ-ray dose rates on the beamline were measured using a NaI(Tl) scintillation survey meter (TCS-172B; Hitachi Aloka Medical Ltd.) and an ionization chamber survey meter (ICS- ICS-331B; Hitachi Aloka Medical Ltd.). The dose rates in the accelerator room were measured within 3 hours after the operation on December 19, 2019, and those in the gantry and treatment rooms were measured within 3 hours after the operation on December 20, 2019.
To confirm the dominant γ-ray source radionuclides in the contact dose rates, a 38 mmϕ×38 mm LaBr3 detector (IPROL-1; Mirion Technologies Inc.) in combination with a handheld multi-channel analyzer (InSpector 1000; Mirion Technologies Inc.) was used for γ-ray spectrometry. The γ-ray spectrometry was performed within 1–5 hours after the end of the operation on December 18, 2019. Therefore, the γ-ray spectrometry was equivalent to the contact dose rate measurements with respect to the time interval between the end of the operation and measurement.
2) PMRC
At PMRC, the contact γ-ray dose rates on the beamline were measured using a NaI(Tl) scintillation survey meter (TCS-171; Hitachi Aloka Medical Ltd.). The dose rate measurements were performed from 9:00 AM to 1:00 PM on July 27, 2018, the day after the end of the treatment operation at 6:00 PM on July 26, 2018.
To confirm the dominant γ-ray source radionuclides in the contact dose rates, γ-ray spectrometry was also performed on the beamline using the same 38 mmϕ×38 mm LaBr3 detector used at MPTRC. However, the γ-ray spectrometry was initiated 1 hour after the end of the operation on June 28, 2018; therefore, the time between the end of the operation and the start of the measurements was shorter than that for the dose rate measurements.
3. Measurements of Thermal Neutron Fluence Rates by the in situ 24Na Measurement Method
In our previous study on cyclotron-type proton therapy facilities [2], the “in situ 24Na measurement method” was used for measuring the thermal neutron fluence rate in concrete with high detection sensitivity. This method was also applied at the electrostatic accelerator facility and the synchrotron radiation facilities [7, 8]. First, to determine the specific activity of 24Na (half-life: 14.9590 hours [3]) produced by the thermal neutron capture reaction of 23Na in the concrete, the 1,369 keV γ rays from 24Na were counted by putting a γ-ray spectrometer on the concrete floor in situ after the operation. Subsequently, the thermal neutron fluence rate in the concrete was estimated from the 24Na specific activity in the concrete, the typical Na concentration in concrete (2.1 wt% [7]), and the thermal neutron capture cross-section of 23Na (0.53 b [3]).
1) MPTRC
The γ-ray spectrometry was performed using a Ge detector (GR2018, relative efficiency for 1,333 keV γ ray of 60Co: 23.6%; Canberra Inc.) and a 38 mmϕ×38 mm CeBr3 detector (CEBRS-1.5×1.5 combined with Osprey-DTB; Mirion Technologies Inc.), which were shielded with 6.5 cm thick lead. The detection efficiencies of the detectors for 1,369 keV γ rays from 24Na in the concrete were calculated using In Situ Object Counting System (ISOCS) [9, 10]. Here, the γ-ray source was assumed to be uniformly distributed in the concrete. The measurement sites were selected mainly around the areas where high contact dose rates were observed on the beamline. Measurements in the treatment room were performed immediately after the operation on December 19, 2019 and in the other rooms immediately after the operation on December 20, 2019. The effect of beam power fluctuations on 24Na radioactivity production was corrected using the dose rate fluctuations observed by the radiation monitor for γ rays.
2) PMRC
The γ-ray spectrometry was performed using a Ge detector (GR2018; Canberra Inc.) shielded with 6.5 cm thick lead. The calculation method of detection efficiency is the same as that described for the MPTRC section. The measurement sites were selected mainly around the areas where high contact dose rates were observed on the beamline. Measurements were performed immediately after the operation on August 23, 2018. The effect of beam power fluctuations on 24Na radioactivity production was corrected using the dose rate fluctuations observed by the radiation monitor for γ rays.
4. Measurements of Thermal Neutron Fluence Rates Using Thermoluminescence Dosimeters and Au Foils
Thermoluminescence dosimeters (UD813PQ4; Panasonic Corp.), referred to as “TLD,” were used as detectors for thermal neutrons. TLDs can measure thermal neutrons using a pair of a 0.5 mm thick Cd-covered TLD and an uncovered TLD [11, 12]. The pairs of TLDs were installed on the concrete before and removed after the operation. The TLDs were then read using a reader (UD7900; Panasonic Corp.) to obtain the thermal neutron fluence during the operation.
Au foils were also installed alongside a pair of TLDs as another detector for the thermal neutrons. The thermal neutron fluence rate measurements using Au foils were applied in other studies [13, 14]. The Au foils (purity: 99.99%, diameter: 6 mm, thickness: 0.02 mm, shape: circular) can measure thermal neutrons using a pair of a 1 mm thick Cd-covered Au foil and an uncovered Au foil. After the operation, the 198Au (half-life: 2.69517 days [3]) radioactivity in one of the Au foils was determined using a Ge detector (GC2518, relative efficiency for 1,333 keV γ ray of 60Co: 25%; Canberra Inc.) as a reference. Here, the counting efficiency was calculated using ISOCS [9, 10]. All Au foils, including the reference, were exposed on an imaging plate (MS2025; Fujifilm Corp.). After exposure, the imaging plate was analyzed using a reader (Typhoon FLA7000; GE). The radioactivities of 198Au for all the Au foils were determined from the values read from the imaging plate relative to the reference Au foil.
1) MPTRC
The TLDs and Au foils were installed before the operation on December 20, 2019 and collected after the operation on the same day. The operation time for this day was 10.3 hours. The beam power obtained from the γ radiation monitor on this day was slightly weaker than the average beam power from December 1 to December 28, 2019, with a correction of 1.02.
2) PMRC
The Au foils were placed on the concrete during operations from July 27 to August 23, 2018. The operation time for this period was 206 hours. The effect of beam power fluctuations on the 198Au radioactivity production was corrected using the dose rate fluctuations observed by the radiation monitor for γ rays.
Results and Discussion
1. MPTRC
1) Beam loss locations
Fig. 2 shows the contact γ-ray dose rates on the beamline measured at MPTRC. The measurements were taken within 3 hours after the end of the operation. The units for the values in Fig. 2 are μSv/hr, and only values above 5 μSv/hr are shown in the figure. In particular, the locations above 10 μSv/hr are marked with yellow stars. Two locations in the beam transport line room exceeded 100 μSv/hr. These two locations had aluminum beam stoppers, and the observed radionuclides were 22Na (half-life: 2.6019 years [3]) and 24Na, which were different from the other locations. At the other locations, similar radionuclides were detected; they were 44mSc (58.6 hours), 46Sc (83.8 days), 48V (16.0 days), 51Cr (27.7 days), 52Mn (5.6 days), 54Mn (312.1 days), 56Co (77.3 days), 57Co (271.8 days), 58Co (70.8 days), and 57Ni (35.6 hours) (values in parentheses indicate half-lives [3] rounded to one decimal place). The radioactivity of 56Co was particularly high, and the γ-ray contribution of 56Co was the largest in the contact dose rate. The contact dose rate on the beamline can be used to indicate the proton beam loss. Considering the half-life of 56Co, the measured contact γ-ray dose rates reflect the results of beam losses in the last couple of months. It was found that the proton beam loss generally occurred only at the beam bend locations at MPTRC.
2) Thermal neutron fluence rates on the concrete
Fig. 3 shows the thermal neutron fluence rates on concrete measured at MPTRC. As previously mentioned, three different methods were used to measure the thermal neutron fluence rates. In the figure, each value is labeled with the corresponding measurement method. The measurements of the thermal neutron fluence rates were performed primary in the vicinity of locations where the high contact dose rates were observed, that is, where beam losses occurred. The thermal neutron fluence rates measured using the three different methods were in close agreement, although there were some differences due to the different detection principles.
In the accelerator room, thermal neutron fluence rates of 102 neutrons/cm2/s were observed. In addition, thermal neutron fluence rates of 102 neutrons/cm2/s were observed around the two beam stoppers on the beam transport line. In all other locations, the thermal neutron fluence rates were very low, at 101 neutrons/cm2/s.
3) Specific activities in concrete after 30 years of operation
The specific activities of 152Eu and 60Co produced in concrete after 30 years of operation were estimated using the measured thermal neutron fluence rates. Here, the target stable Eu and Co concentrations in the concrete were assumed to be 0.7 ppm [15] and 8 ppm [15], respectively. These concentrations are the averages of the values determined for the shield concrete of the various accelerators. In addition, 5,900 b (152Eu ground state production) [3] for 151Eu and 37.18 b [3] for 59Co were used as the cross sections for the thermal neutron capture reactions. Using these values, we predicted the sum of the specific activities of 152Eu and 60Co in the concrete, assuming that the typical operation at the time of the investigation would continue for 30 years. The validity of this prediction method has been confirmed in detail in the previous study [2]. The sum of the predicted specific activities of 152Eu and 60Co in concrete is shown in Fig. 4, and these values are color-coded by level.
The sum of the predicted specific activities of 152Eu and 60Co in concrete slightly exceeded 1.0×10−3 Bq/g at the highest levels near the synchrotron and beam stoppers of the beam transport line. At the site showing the highest value of 1.6×10−3 Bq/g, the 30-year neutron fluence is 1.6×1011 neutrons/cm2. In all other locations, no concentration exceeded 1.0×10−3 Bq/g. Consequently, it was found that the specific activity concentration of concrete throughout the facility was negligibly lower than the clearance level of 0.1 Bq/g.
2. PMRC
1) Beam loss locations
Fig. 5 shows the contact γ-ray dose rates on the beamline measured at MPTRC. The contact γ-ray dose rates were measured on the day after the operation. The units for the values in Fig. 5 are μSv/hr, and only values above 5 μSv/hr are shown in the figure. In particular, the locations above 10 μSv/hr are marked with yellow stars. At the beamline, radionuclides of 46Sc, 48V, 51Cr, 52Mn, 54Mn, 56Co, 57Co, 58Co, 60Co, and 57Ni, which were produced in stainless steel beam ducts, were detected. As in MPTRC, the radioactivity of 56Co was particularly high, and the contribution of 56Co was dominant in the contact dose rates. Using the contact γ-ray dose rate as an indicator of beam loss, we found that at PMRC, the large proton beam losses that occurred were limited to the bent beam locations, similar to that at MPTRC.
2) Thermal neutron fluence rates on concrete
Fig. 6 shows the thermal neutron fluence rates on concrete during the operation measured at PMRC. The thermal neutron fluence rates on the concrete were measured using the two different methods described above. In the figure, each value is labeled with the corresponding measurement method. The thermal neutron fluence rates were measured mainly in the vicinity of sites where high contact dose rates were observed.
Thermal neutron fluence rates of 102 neutrons/cm2/s were observed in a limited area near the synchrotron in the accelerator room. In other places, the thermal neutron fluence rates on concrete were found to be very low, at 101 neutrons/cm2/s.
3) Specific activities in concrete after 30 years of operation
The sums of the specific activities of 152Eu and 60Co in the concrete after 30 years of operation were also predicted for PMRC, using the same method as for MPTRC. Fig. 7 shows the sums of the predicted specific activities of 152Eu and 60Co in the concrete after 30 years of operation. The sums of the specific activities of 152Eu and 60Co did not exceed 1.0×10−3 Bq/g anywhere. At the site showing the highest value of 7.5×10−4 Bq/g, the 30-year neutron fluence is 7.3×1010 neutrons/cm2. Consequently, the specific activities in the concrete throughout the facility were negligibly lower than the clearance level. The characteristics of the concrete activation at PMRC were similar to those at MPTRC. Therefore, the characteristics of the concrete activation obtained in this study are applicable to all typical synchrotron-type proton therapy facilities.
3. Comparison with Cyclotron-Type Proton Therapy Facilities
In our previous study [2], we measured the thermal neutron fluence rates in concrete during operations at two cyclotron-type proton therapy facilities and predicted the sums of the specific activities of 152Eu and 60Co in concrete after 30 years of operation. In the cyclotron-type proton therapy facilities, the thermal neutron fluence rates in concrete during operation exceeded 104 neutrons/cm2/s in the accelerator rooms. These thermal neutron fluence rates were two orders of magnitude higher than those in the synchrotron-type proton therapy facilities. As a result, the sum of the predicted specific activities of 152Eu and 60Co in concrete was closer to the clearance level at the cyclotron-type proton therapy facilities. The specific activities were particularly high near the degrader at the cyclotron-type proton therapy facilities. Proton deceleration in the degrader for energy control causes neutron production. Because there is no degrader at the synchrotron-type proton therapy facilities, there is no special neutron production. However, the predicted levels of specific activity in the concrete in other rooms, such as the beam transport line, gantry, and treatment rooms, at the synchrotron-type proton therapy facilities are similar to those at the cyclotron-type proton therapy facilities.
Conclusion
For the two typical synchrotron-type proton therapy facilities, the specific activities in concrete after 30 years of operation were predicted from the measured thermal neutron fluence rates during operation. The specific activities in concrete everywhere in typical synchrotron proton therapy facilities are negligible compared to the clearance level. Therefore, concrete does not need to be treated as radioactive waste when decommissioning the synchrotron-type proton therapy facilities. Notably, the specific activity level in the accelerator room concrete in the synchrotron-type proton therapy facilities where an accelerator controls the proton energy is much lower than that in the cyclotron-type proton therapy facilities, where a degrader controls the proton energy.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
