International Network of Engineers and Scientists Against Proliferation

Measurements of Krypton-85 to Detect Clandestine Plutonium Production

Martin B. Kalinowski, Heiner Daerr, Markus Kohler

The biggest challenge for nuclear safeguards that aim to verify compliance with the nuclear Non-Proliferation Treaty (NPT) is the capability to detect unreported nuclear materials, activities, and facilities. The same task has to be solved for a verified agreement on the cutoff of fissile materials production. Novel technologies are being sought that are able to detect proliferation indicators at a distance. Environmental sampling is attractive because the indicating substances are transported through the atmosphere over large distances. However, there is still serious doubt about the efficient applicability of this approach and no particular concept has been convincingly proven yet. The case of krypton- 85 as an indicator for plutonium separation is discussed in this paper. The current understanding of the detectability of reprocessing plants based on atmospheric sampling of Kr-85 is presented, the need for further investigation is described, and proposals are made for possible improvements of concepts for this approach.

After the uncovering of clandestine nuclear operations in Iraq, the 93+2 program[1] was conducted by the International Atomic Energy Agency (IAEA) in order to strengthen the efficiency and effectiveness of nuclear safeguards. One of the most important goals was to identify new safeguards methods to detect undeclared reprocessing facilities. For the first time, environmental sampling was seriously considered as a method for nuclear safeguards.

It took a couple of years before the Additional Protocol[2] was concluded in 1997. It establishes not only the possibility to take environmental samples at declared sites and facilities. It also has two specific provisions based on complementary access that enable the IAEA to take environmental samples at a distance from an undeclared facility:

• location-specific environmental sampling,
• wide-area environmental sampling.

It is generally understood that location-specific environmental sampling allows for a visit at one location that might last for one or two days, during which the inspectors take a couple of samples at that location. This provision could be implemented without a new decision by the IAEA Board of Governors as soon as the technology is available and the related procedures are developed. This has not yet been achieved.

Wide-area environmental sampling is generally understood as taking samples at multiple locations over a limited period of time. Article 9 of the Additional Protocol defines that “the Agency shall not seek such access until the use of wide-area environmental sampling and the procedural arrangements therefor have been approved by the Board,” i.e. by the IAEA’s Board, of Governors. A proposal for procedural arrangements requires a detailed and scientifically sound analysis of the practical requirements that has not yet been achieved.

Even though the provisions for location-specific and wide-area environmental sampling were put into the Additional Protocol, the IAEA has no scientific and technological capabilities to implement them.

Krypton-85 as Ideal Tracer for Detection of Clandestine Plutonium Separation

Satellite imagery has proven to be very effective in locating and characterizing relevant facilities as well as monitoring construction progress and other relevant changes.[3] However, image analysis does not reveal the presence of nuclear materials or activities. The applicability of satellite imagery also depends on knowledge about the existence of a facility and its approximate location.

Environmental and material sampling are now playing an essential role in safeguards for detection of uranium enrichment and plutonium separation, though challenges remain in collection strategy, robustness of findings, and limited laboratory analysis capacities.[4] The Additional Protocol has driven the use of advanced analytical methods for detecting undeclared activities. However, besides a few exceptional applications, environmental sampling has been restricted to swipe samples taken inside declared facilities. These are not suited to detect unreported nuclear facilities. Environmental sampling outside inspected facilities is attractive because the indicating substances are transported through the atmosphere over large distances.

For decades, krypton-85 has been considered as a remote indicator for plutonium production.[5] It has been used to determine the plutonium stockpile of the Soviet Union.[6] Krypton- 85 has a combination of unique features which makes it an ideal tracer for plutonium separation activities anywhere in the world. It is always generated along with plutonium and 99.9% remains within the fuel cladding. Due to its half-life of 10.76 years, significant amounts of krypton- 85 still remain in the spent fuel even after long cooling times. Since krypton is a noble gas, it is chemically inert and has a low solubility in water. Therefore, krypton is not removed from the atmosphere by any processes like chemical reactions or wash-out. Furthermore, there are no other relevant sources of krypton-85 other than reprocessing. In comparison with other candidate tracer nuclides (e.g. T-3, I-129, Sr-90, Ru-106, Cs-137 etc.), krypton-85 has the best sensitivity next to Pu-239.[7]

Possible evasion scenarios might speak against krypton-85 as a reliable indicator for plutonium separation. Though retention of noble gases is technically possible, it is expensive but not very reliable and, therefore, was never applied on a large scale. A more realistic but also elaborate scenario in addition to retention would be to funnel the off-gases through a delay bed. This could hold up the plume in time from hours to days and result in a reduction of peak release concentrations by one order of magnitude. Another evasion method would simply be to wait for favorable meteorological conditions, when the winds are blowing the plume away from the air sampling site. This scenario requires the implementation of sampling schemes (e.g. by setting up multiple stations or by taking unannounced surprise samples) that would minimize the chances for the proliferator to escape detection.

The IAEA detection goal is to detect the production of one significant quantity of weapons-useable material, i.e. 8 kg of separated plutonium.[8] The timeliness for detection of unreported facilities has not been defined, but all relevant studies assume a one year period for the production of one significant quantity of plutonium. Since the most likely scenario is that the proliferating country wants to produce weapon-grade plutonium, it would reprocess fuel with a very low burn-up. Since reprocessing is done in a batch mode, there is no continuous release of krypton-85 but there are pulses lasting as long as it takes the fuel to dissolve in the acid. There are various scenarios of running the production campaign. Studies on krypton-85 as proliferation indicator often select the case of processing one batch per week as reference scenario.

The major difficulty of using krypton-85 as indicator is the high atmospheric background level of the isotope and its variability, which limits the distance at which a signal can be clearly distinguished from the background noise. The atmospheric background is principally caused by accumulation from reprocessing activities in the past.

Measurement Technologies and Sampling Schemes for Krypton-85

he standard procedure for measuring the atmospheric krypton-85 concentration is by detection of its beta activity. Before the krypton-85 can be measured, the other gas components in the sample (nitrogen, oxygen, carbon dioxide, water, radon, and other trace elements) have to be completely eliminated.

In case of wide-area environmental sampling, complete analysis would have to be carried out in the field in an automated fashion. This kind of technology has been developed for the monitoring system of the Comprehensive Test Ban Treaty (CTBT) organization to analyze daily air samples for radioactive xenon isotopes.[9] This new technology is still being tested under field conditions.

For location-specific environmental sampling, samples taken in the field could be sent to laboratories. The size of air samples required for analysis has implications on the complexity of sampling instrumentation and may limit the number of samples. Sampling of 10 up to 250 liter of air in the field is currently necessary and requires a method for reducing the sample volume. A small volume of air may be taken as grab sample and compressed to fit into a small bottle. Larger volumes are pumped into a cryo-absorption device that requires electric power and liquid nitrogen for cooling.

The novel technology of atom trap trace analysis (ATTA) has been demonstrated in the laboratory by the physics group at Argonne National Laboratory.[10] This is an ultra-sensitive analysis technique able to detect single atoms.

A direct comparison of three trace analysis methods for krypton-85 (low-level counting, accelerator mass spectrometry, and ATTA) shows that ATTA is the superior method.[11] It has not only the highest sensitivity and selectivity, it can also operate on the smallest samples. The current stateof- the-art ATTA system requires a fairly large volume of air (ca. 25 liters), but produces results much faster than beta counting: ATTA takes 30 minutes, while it takes about eight hours to achieve the same accuracy with beta counting. ATTA technology has still considerable potential for efficiency improvement. It seems promising to investigate sampling schemes and inspection procedures based on ATTA analysis in the laboratory rather than in the field. The laboratory equipment has a significantly higher sensitivity, therefore smaller air samples could be used.

The goal is to work on one liter samples with five minutes analysis time. This sample volume reduction would be a significant step, since one liter can be taken as a grab sample by sucking it directly into pre-evacuated bottles at atmospheric pressure. This, in turn, would allow sampling procedures that are as simple and effective as swipe samples, since neither liquid nitrogen nor electric power is required in the field. As a consequence, more air samples can be taken while keeping the costs relatively low.

Sampling Schemes

The distance from the source is the most critical parameter for detecting unreported plutonium production, because the signal gets weaker during the transport through the atmosphere. A plume with a krypton-85 load dilutes by mixing with ambient air and, at some point, the maximum krypton- 85 concentration in the plume is too low to be detected against the variation of the atmospheric background concentration. Consequently, efficient sampling approaches are distinguished by their distance from the source. A global network, e.g., is not an option because it would require a maximum distance between stations of ca. 100 km. This would be prohibitively expensive. Instead of a large network of stations, it would be more cost efficient to use a relatively small number of strategically placed sampling stations. This is how the term wide-area environmental sampling should be understood.

Basically, there is a short-range and a long-range option as described below.

Scheme I: Short Range

The short-range option – within about one kilometer from a known site – could be used for the detection of unreported plutonium production at declared sites. Under the NPT Additional Protocol, the inspectors could take air samples within the facilities in a similar way as they already take swipe samples under the provision of location specific environmental sampling. The samples are sent to a laboratory for evaluation. This type of air samples are of particular interest for complex equipment or facilities inside industrial districts that are not completely under IAEA nuclear safeguards. Under a Fissile Material Cutoff Treaty (FMCT), the goal could be to verify the status of stand-by or shut-down facilities and to detect undeclared operations verification of military production facilities. This would likely demand non-intrusive inspection procedures which are able to ensure a high confidence in compliance without compromising too much the perceived national security by releasing sensitive information. This is achieved best by remote detection methods, in particular by air sampling, because air samples can be taken just beyond the fence and do not even require the inspector to enter a facility.

Scheme II: Long Range

The long-range option – at 1-100 km distance – could be used in the case of an unreported but suspected plutonium production facility, where the location is not precisely known and signals from large areas would need to be detected. The goal would be to detect relevant krypton-85 signals in air samples and to trace them back to their source. Under the NPT Additional Protocol, two different approaches can be taken and these could well be adopted by an FMCT as well: ■ Wide-area environmental sampling: With this approach a certain region would be monitored with a network of stations for a certain period of time so as to maximize the detection probability. The drawbacks are the high cost and the fact that the proliferator can decide not to reprocess as long as this would be detectable. Accordingly, the benefit of this approach would mainly result from its deterrence effect. Depending on the quality of the signals, one of the following two analysis methods can be applied:

• A signal that is clearly above the background indicates the detection of a plume. Methods for determining the possible source location can then be applied.
• In the absence of clear signals in single samples, the average signal level over a period of time can be compared to meteorological simulations. An increase may be observed in the presence of an undeclared source. This assumes that all reprocessing plants would report their krypton-85 emissions, which are then used for meteorological modeling.

■ Location-specific environmental sampling: The inspectors would use complementary access for a surprise visit at a location where according to the meteorology data the maximum plume from a suspected reprocessing facility can be expected. The disadvantage is the high risk of not detecting anything, mainly because reprocessing campaigns and therefore the related radioactivity releases are infrequent and their time can only be guessed.

In any case, the sampling scheme has to be designed such that locating the source is facilitated. Very efficient methods for determining possible source regions have been developed and implemented in the context of the CTBT.[12] The method for source location uses a correlation of concentration-weighted source-receptor-sensitivities.

Case Studies

The reprocessing plant Karlsruhe (Wiederaufarbeitungsanlage Karlsruhe, WAK) has been used for a case study. During a period of two and a half years in 1985-1988, weekly air samples were taken at four distances (0.29 km, 0.68 km, 5.1 km, and 39 km) downwind in the main wind direction of the WAK.[13] A significant advantage of this study is that the time and amount of emissions is well known. Besides the krypton-85 emission data (precise to the hour), the amounts of plutonium separated at those times is publicly available.

The detection and false alarm rates were determined from the weekly samples.[14] For sampling sites closer than 1 km to the source and within the main wind direction, a detection rate of more than 80-90% is achieved; for a distance of 5 km it is still 70%; at 40 km it is reduced to 40%. However, not a single false alarm was sounded at the 40 km site. Only one or two false alarms occurred at the three sampling sites located closer to the source resulting in a false alarm rate of 2-3%. Because of the specific geographic conditions of the Rhine Valley, at an additional sampling site at Freiburg, located 130 km from the WAK and in a direction opposite to the prevailing winds, a 14% detection rate was achieved with only one false alarm. As a result of this study it became clear that a single air sampler positioned in the main wind direction at a distance of several hundred meters from the emission point can achieve a high detection probability, if the samples are evaluated on a weekly basis. Cheating may be easy, because operators can wait for the wind blowing in the opposite direction. This can be avoided by placing several stations around the site to cover all wind directions. Even with a single station, a proliferator can never be sure that weather conditions stay steady enough to not carry air masses with a high krypton-85 content back to the sampler.

Another proof of detectability was found in 2001.[15] The shutdown of the Japanese reprocessing plant at Tokai Mura between April 1997 and July 2000 was detectable with this single detector at a distance of 80 km.

Previous studies were able to find evidence for pulsed emissions of krypton-85 from large sources which were up to several hundreds of kilometers away. Global atmospheric modeling was able to predict the qualitative variances of monthly krypton-85 concentrations at various sampling points quite well.

Conclusion

Within a range of one hundred kilometers from small-scale reprocessing facilities, pulsed emissions of krypton- 85 can clearly be identified with a high probability. Large sources can be detected at even larger distances. Improvements in implementing a sam- Measurements of Krypton-85 to Detect Clandestine Plutonium Production INESAP Information Bulletin No. 27, December 2006 12 pling scheme based on the novel technology of atom trap trace analysis (ATTA), high-quality atmospheric transport simulations, mobile air samplers, and shorter sampling periods could raise the usefulness and quality of krypton-85 sampling in comparison to the monitoring scheme that were previously studied. However, it still remains unclear to what extend and under what conditions remote sampling in combination with transport modeling can detect clandestine plutonium separation of significant quantities with sufficiently high detection and low false alarm probability. In order to evaluate this more precisely, further studies are under way.

---

1.   The 93+2 program owes its name to the fact that it started in 1993 and was planned for a duration of 2 years so that it would produce sufficient results by the time of the NPT Review and Extension Conference in 1995, in order to provide satisfying answers to the critical issue of clandestine activities.
2.   Model Protocol Additional to the Agreement(s) between State(s) and the International Atomic Energy Agency for the Application of Safeguards, INFCIRC/540 (corrected), approved by the IAEA Board of Governors on 15 May 1997.
3.   B. Jasani and G. Stein, (eds.), Commercial satellite imagery: a tactic in nuclear weapon deterrence, Berlin: Springer, 2002.
4.   M.B. Kalinowski, J. Feichter, M. Nikkinen, and C. Schlosser, Environmental Sample Analysis, in: R. Avenhaus, N. Kyriakopoulos, M. Richard, G. Stein (eds.), Verifying Treaty Compliance. Berlin et al: Springer, 2006, pp. 367-387.
5.   See e.g. A. Sittkus and H. Stockburger, Krypton-85 als Indikator des Kernbrennstoffverbrauchs, Naturwissenschaften 63, 1976, pp. 266-272.
6.   F. v.Hippel and B.G. Levi; Controlling Nuclear Weapons at the Source: Verification of a Cutoff in the Production of Plutonium and Highly Enriched Uranium for Nuclear Weapons, in: K. Tsipis, D. Hafemeister, P. Janeway (eds.), Arms Control Verification — The Technologies that Make it Possible, Pergamon Brassey’s, Washington, 1986, pp. 338-388.
7.   P.W. Krey and H.L. Beck, Atmospheric monitoring for a nuclear fuel reprocessing plant, Environmental Measurements Laboratory, New York, 1993. P.W. Krey and K. W. Nicholson, Atmospheric sampling and analysis for the detection of nuclear proliferation, Journal of Radioanalytical and Nuclear Chemistry 248, No. 3, 2001, pp. 605-620.
8.   The IAEA defines a significant quantity of nuclear material (8 kg of plutonium, 25 kg of uranium-235 in highly enriched uranium, 75 kg of uranium-235 in natural or low enriched uranium) as “The approximate quantity of nuclear material in respect of which, taking into account any conversion process involved, the possibility of manufacturing a nuclear explosive device cannot be excluded.” www.iaea.org/Publications/Booklets/Safeguards/pia3810.html.
9.   M. Auer, A. Axelsson, X. Blanchard, T.W. Bowyer, G. Brachet, I. Bulowski, Y. Dubasov, K. Elmgren, J.P. Fontaine, W. Harms, J.C. Hayes, T.R. Heimbigner, J.I. McIntyre, M.E. Panisko, Y. Popov, A. Ringbom, H. Sartorius, S. Schmid, J. Schulze, C. Schlosser, T. Taffary, W. Weiss, and B. Wernsperger, Intercomparison experiments of systems for the measurement of xenon radionuclides in the atmosphere, Applied Radiation and Isotopes 60, 2004, pp. 863–877.
10.   C.Y. Chen; Y.M. Li; K. Bailey; T.P. O’Connor; L. Young; Z.-T. Lu, Ultrasensitive Isotope Trace Analyses with a Magneto-Optical Trap, Science 286, 5 November 1999, pp. 1139-1141. The device developed at Argonne is used for ground water and ice core dating studies and had not been previously considered for safeguards applications.
11.   P. Collon; W. Kutschera; Z.-T. Lu, Tracing Noble Gas Radionuclides in the Environment, Annual Review of Nuclear and Particle Science 54, 2004, pp. 39-67.
12.   M.B. Kalinowski, Atmospheric transport modelling related to radionuclide monitoring in support of Comprehensive Nuclear-Test-Ban Treaty verification; Kerntechnik 66/3, 2001, pp. 129-133. Gerhard Wotawa, Philippe Denier, Lars-Erik DeGeer, Martin B. Kalinowski, Harri Toivonen, Real D’Amours, Franco Desiato, Jean-Pierre Issartel, Matthias Langer, Petra Seibert, Andreas Frank, Craig Sloan & Hiromi Yamazawa, Atmospheric transport modelling in support of CTBT verification — Overview and basic concepts, Atmospheric Environment 37 (18), 2003, p. 2529-37.
13.   I, Levin, K.O. Muennich; H. Sartorius; H. Stockburger; W. Weiss; D. Papadopoulos; L.A. Koenig, and H. Huebner, Experimentelle Bestimmung der Langzeitausbreitungsfaktoren durch simultane C-14- und Kr-85-Messungen in der Umgebung der Wiederaufarbeitungsanlage Karlsruhe (WAK), Final report, March 1992.
14.   M.B. Kalinowski, H. Sartorius, H., S. Uhl, and W. Weiss, Conclusions on Plutonium Separation from Atmospheric Krypton-85 Measured at Various Distances from the Karlsruhe Reprocessing Plant, Journal of Environmental Radioactivity 73/2, 2004, pp. 203-222.
15.   A. Igarashi, M. Aoyama, K. Nemoto, K. Hirose, T. Miyao, K. Fushimi, M. Suzuki, S. Yasui, Y. Asai, I. Aoki, K. Fujii, S.Yamamoto, H. Sartorius, and W. Weiss, 85Kr measurement system for continuous monitoring at the Meteorological Research Institute, Japan, Journal Environmental Monitoring 3, 2001, pp. 688-696.