Introduction

The Comprehensive Nuclear Test-Ban Treaty (CTBT) was established to stop the testing of nuclear weapons around the globe [1]. Within the CTBT, the International Monitoring System (IMS) was established to create a world-wide detector network for treaty verification. The IMS contains seismic, hydroacoustic, infrasound, and radionuclide monitoring stations. The radionuclide monitoring stations include capabilities to monitor for both aerosols and radioxenon. For radioxenon monitoring, the isotopes of interest are 131mXe (T 1/2 = 11.93 days), 133mXe (T 1/2 = 2.19 days), 133Xe (T 1/2 = 5.25 days), and 135Xe (T 1/2 = 9.14 h). The radioxenon monitoring is the focus of the work detailed herein.

Radioxenon monitoring is an important technology for monitoring covert nuclear activities. Since xenon is a noble gas, it is hard to contain and has a high probability for leakage out of most systems. A prime example of the utility of radioxenon is the Democratic People’s Republic of Korea (DPRK; North Korea) first nuclear weapon test on October 9, 2006. Radioxenon was the only radionuclide detected from the release of this nuclear test [2].

The IMS routinely detects radioxenon from anthropogenic sources other than nuclear explosions [3]. These sources include both commercial nuclear power plants and radiopharmaceutical facilities. Out of these two sources, the radiopharmaceutical facilities dominate the radioxenon releases to the atmosphere [3]. While the nuclear forensic signatures of commercial nuclear power plants are readily distinguished from nuclear explosions [4], Saey et al. [5] has shown that the source signatures of radiopharmaceutical facilities is not easily distinguishable from that of nuclear weapon explosions.

The goal of this work is to quantify the factors that control the radioxenon source signature from radiopharmaceutical production facilities. Many factors are investigated including neutron flux, neutron energy, uranium target enrichment, irradiation time, decay time, and the use of holding tanks.

Modeling of source signatures

The radioxenon source signatures were calculated using ORIGEN 2.2 [6]. This code was utilized rather than recent versions of ORIGEN due to the presence of libraries for fast fission and the availability of the PRO card. Fast fission libraries were necessary to estimate the radioxenon production in nuclear explosions. The PRO card was originally designed to model nuclear fuel cycles with reprocessing. For this work, the PRO card was utilized to separate radioxenon from its parent radionuclides. This is necessary for both modeling fractionation within a nuclear weapon explosion and for modeling the separation of radioxenon from the radioiodines in a medical isotope production process. The ORIGEN 2.2 calculations were benchmarked against both ORIGEN-ARP and Excel calculations.

Radioxenon source signatures

Nuclear explosions and commercial nuclear reactors

The plotting methodology utilized in this work is similar to that shown in Kalinowski et al. [4]. for comparison purposes. Figure 1 illustrates the radioxenon source signature regions for both nuclear explosions and commercial nuclear power plants. The data show that there is minimal difference between the source signature from a 235U weapon and a 239Pu weapon. The nuclear weapon signature at the time of detonation is the point at the top of the curves. If radioxenon is immediately fractionated from the parents (e.g. radioiodine), the source signature follows the right most line as it decays. If no radioxenon is separated from its parents, then the source signature follows the left most line as it decays. These scenarios give the boundary values possible nuclear explosion radioxenon source signatures. So, the lines shown in this plot should be viewed as the window of possibilities for nuclear weapon radioxenon source signatures.

Fig. 1
figure 1

Comparison of radioxenon source signatures. Note that the source signature region for commercial nuclear reactors is clearly to the left of the source signature window set for nuclear weapon explosions

Full size image

The source signature region for commercial nuclear reactors is represented by the circle to the left of the nuclear explosion region shown in Fig. 1. This region is representative of the data shown in Kalinowski et al. [4]. As a result of these data, it is predicted that the radioxenon source signature from commercial nuclear reactors is discernable from the signature of nuclear weapon explosions for many scenarios.

Radiopharmaceutical facilities

Production of medical isotopes through high enriched uranium (HEU) target activation was the method modeled for this work because it is one of the most common production methods. A standard production methodology was chosen to include an irradiation time of 5 days, a decay time of 12 h, and subsequent separation [7]. The radioxenon signature for the standard medical isotope irradiation is shown and compared to that of a 235U nuclear explosion in Fig. 2. The data show that the two source signatures are very close and likely not discernable in an environmental monitoring scenario.

Fig. 2
figure 2

Radioxenon source signature from standard medical isotope facility run in comparison to those from a 235U nuclear weapon explosion

Full size image

Figure 3 shows the radioxenon signature measured from the emissions of the Ezeiza Atomic Center in Argentina [8]. These data show that the signature is very similar to that predicted by the modeling for a radiopharmaceutical production facility. The data point is also within the bounds of what is expected from a nuclear explosion.

Fig. 3
figure 3

Comparison of radioxenon emission from Argentina with nuclear explosion signature

Full size image

Sensitivity study

The above work has shown the basic radioxenon signature range for medical isotope production facilities. The goal of this work is to show how this signature changes as a function of different operational parameters and experiment methodologies. Parameters studied include neutron flux, neutron cross-section, fuel enrichment, irradiation time, decay time, and the use of accumulation tanks.

Neutron flux

Each reactor utilizes different neutron flux for the irradiation of HEU targets for radiopharmaceutical isotope production. The neutron flux is largely dependent on nuclear reactor design and the position of the irradiation facility within the reactor core. A range of neutron fluxes were considered: 1011–1015 n cm−2 s−1. This is expected to cover the possible neutron fluxes that are most likely to be utilized.

For this sensitivity study, an irradiation time of 5 days was assumed. Samples were calculated to decay for 12 h before separation. The only parameter that was altered during this modeling was neutron flux. Figure 4 illustrates how the radioxenon signature changes as a function of neutron flux. These results show that higher neutron fluxes suppress the production of 135Xe. This is expected due to increased loss of 135Xe through neutron absorption because of the very large thermal cross-section of this isotope. Even though the radioxenon signature does appear to change as a function of neutron flux, it does not move to a region where the signature is readily distinguished from a nuclear explosion.

Fig. 4
figure 4

Radioxenon signature variation as a function of neutron flux. Note that the ratios during irradiation, decay, and after separation are all included in the sample line for this figure

Full size image

Neutron cross-section

ORIGEN 2.2 utilizes one-group neutron cross-sections for the calculations shown in this work. These are flux-weighted averaged cross-sections that represent specific reactor designs. The neutron cross-section library was the parameter altered for this study. The weighted cross-sections correspond to the neutron energy distributions within the reactor designs.

Figure 5 shows the radioxenon signatures for boiling water reactor (BWR), pressurized water reactor (PWR), and pure thermal cross-section libraries. Two BWR cross-section libraries were used to calculate radioxenon signatures from irradiations utilizing fresh fuel (0 burnup) and self-generation fuel. Three PWR libraries were utilized including fuel with 33,000 MDW/MTU burnup, 50,000 MWD/MTU burnup, and self-generation fuel. While slight differences are calculated, they are not discernable when plotted on a log scale as shown in Fig. 5. There are also no significant changes resulting from fuel burnup.

Fig. 5
figure 5

Sensitivity of neutron cross-section library utilized. This relates to differences in neutron energy distributions within the different reactor systems

Full size image

One difference is observed when pure thermal cross-section values are used. The calculations with pure thermal cross-sections result in lower 135Xe values. This is again due to the extremely large thermal absorption cross-section for 135Xe.

These results have two conclusions. The first is that changes in reactor design and fuel burnup do not significantly affect the radioxenon signal from radiopharmaceutical isotope production. As a result, all thermal reactor designs will produce similar radioxenon signatures given the same irradiation and decay times. The second conclusion is that one must not utilize pure thermal cross-sections when modeling radioxenon signatures produced from a reactor. If just thermal neutron cross-sections are utilized, the calculated 135Xe activities will be underestimated.

Fuel enrichment

As mentioned previously, the majority of radiopharmaceutical production facilities currently use HEU targets [7]; however, non-proliferation and waste issues are pushing the industry towards using low enriched uranium (LEU) targets [7]. Figure 6 shows that the radioxenon signature models for both HEU and LEU targets lie on top on one another, and hence the results indicate that uranium enrichment does not affect the radioxenon signature.

Fig. 6
figure 6

Sensitivity of low enriched (19.9%) to high enriched (95%) uranium targets

Full size image

Irradiation time

The target irradiation time was varied in the ORIGEN 2.2 model from 5 to 21 days. Decay time was maintained at 12 h. Figure 7 shows that the radioxenon signature model moves to the left with increased irradiation time. After 2 weeks of irradiation, the radioxenon signature from the radiopharmaceutical production becomes notably separate from that of a nuclear explosion signature.

Fig. 7
figure 7

Sensitivity of irradiation time on radioxenon signature

Full size image

This is an important finding showing significant distinction between radiopharmaceutical production and nuclear explosion radioxenon signals may be achieved through longer radiopharmaceutical target irradiation time. The change largely results from proportionally increased levels of 131mXe produced. Because this is the longest lived of the four radioxenons of interest, it takes a longer time for 131mXe asymptotic activities to be achieved.

Decay time

For this work, decay time is defined as the time between the end of target irradiation and the time when a chemical separation is performed allowing xenon to be separated from other fission products. Calculations included a range of decay times from 6 h to 7 days. Figure 8 shows that the radioxenon signature does not significantly change as a result of decay time. These radioxenon signatures are also not significantly different from those of a nuclear weapon explosion.

Fig. 8
figure 8

Radioxenon signature sensitivity as a function of decay time (the time from the end of target irradiation to a chemical separation of radioxenon from the other fission products)

Full size image

Accumulation tanks

The use of accumulation tanks was also explored. Radioxenon emissions from a facility could be held in one or more tanks prior to release. Such tanks would accumulate the radioxenon releases from multiple runs. Figure 9 shows the results of emission accumulation for 30, 60, and 90 days. While there is not much differences between the different accumulation tank signatures, there is an appreciable difference between the accumulation tank radioxenon signatures and those signature from a nuclear weapon explosion.

Fig. 9
figure 9

Radioxenon signature when an effluent accumulation tank is utilized

Full size image

Conclusions

The results from this study are summarized in Table 1. It is shown that two main parameters alter the radioxenon signal released from a radiopharmaceutical production process: U target irradiation time and the use of accumulation tanks. Target irradiation times of 3 weeks may not be economically advantageous because the radionuclides of interest (e.g., 99Mo and 131I) start to asymptotically level off after just 1 week. Facilities produce less 99Mo and 131I per day of operation with irradiations longer than 1 week. As a result, facilities will likely not find it practical to significantly increase their irradiation times to alter their radioxenon emissions for CTBT monitoring purposes.

Table 1 Summary of sensitivity study
Full size table

In contrast, the use of accumulation tanks appears to hold many advantages. With respect to this study, accumulation tanks change the radioxenon source signature to a point where it is distinguishable from a nuclear weapon explosion. In addition, accumulation tanks allow for radioactive gas decay and consequently reduce the activity of emissions. This holds advantages for the radiopharmaceutical production facilities in maintaining releases under regulatory limits. In cases where these regulations limit isotope production, a radiopharmaceutical production facility may actually be able to increase production (and profit) with the use of accumulation tanks.

One last note of interest resulting from this study is the importance of using representative nuclear reactor based cross-sections to model fission product production in such facilities. If just thermal cross-sections are utilized, the calculated 135Xe activities will be low.