IRSN, Institut de radioprotection et de sûreté nucléaire

Search our site :


Contact us :

En Fr

All our expertise to protect you


Research program

OECD-STEM 2 projet

The STEM2 project (Follow-up of Source Term Evaluation and Mitigation) is a project initiated by IRSN under the aegis of the OECD/NEA, aiming to improve our understanding of the behavior of radioactive substances likely to be released to the environment in the event of a core melt accident at a nuclear facility. The project, which ran from the beginning of 2016 to the end of 2019, followed on from the initial STEM project developed between 2011 and 2015. 

The results obtained during that initial phase improved our knowledge of the behavior of fission products under severe accident conditions, and have been used to enhance and validate the models integrated in ASTEC, the code used to simulate severe accidents. It became apparent that the test program needed to be pursued, first, to explore test conditions that would be even more representative of actual core melt accident conditions; and second, to further our understanding of certain phenomena identified during Phase I (for example, the decomposition and reactivity of iodine oxides). As for Phase I, Phase II explored two aspects, one studying the behavior of iodine subject to radiation within the reactor containment, and the other on the chemistry of ruthenium in the reactor coolant system, focusing on Ru revaporization from deposits.


Background and aims

As in the initial phase, this involved eliminating some significant uncertainties affecting the evaluation of releases to the environment (the source term1), regarding some of the complex phenomena encountered during a core melt accident in a pressurized water reactor (PWR). Examples include the behavior of irradiated iodine subject to radiation in the reactor containment and that of ruthenium in the reactor coolant system. The latter phenomenon is a source of significant uncertainty regarding the evaluation of the source term in certain accident scenarios entailing air ingress into the reactor vessel. As in the initial phase, the STEM2 experimental project sought to reduce these uncertainties, focusing on two key objectives:

  • to advance our knowledge and thereby develop more accurate numerical simulation tools, with the ultimate aim of helping the teams in charge of managing a nuclear accident to make a more robust diagnosis and prognosis of the progress of the accident;
  • to identify additional measures that could be taken to further reduce (compared to present practice) the release of radioactive substances to the environment (prevention and mitigation, for example).

Project outline and research themes

With regard to the "iodine problem", the ability to evaluate releases of gaseous iodine (as molecular iodine, I2, and organic iodine, RI) and particulate iodine over a period of one week is extremely important for the purposes of accident management, since this can help determine the length of time during which emergency measures such as evacuating or sheltering the population must be maintained within a given area. Recent progress in R&D in the field has confirmed some of the assumptions made regarding possible new phenomena, such as the chemical instability of iodine aerosols, which could significantly modify this "medium-term source term".

As for the "Ruthenium problem", this is related to PWR safety issues under accident conditions (typically a reactor loss-of-coolant accident followed by ingress of a gaseous oxidizing mixture following vessel failure). However, it could also prove crucial for managing other types of accident situation, such as a spent fuel pool uncovery accident or accidents during fuel handling. These various cases involve different temperature ranges and the possible chemical hardening of gaseous ruthenium tetroxide, Ru04(g), to give a compound that then becomes metastable, which could result in the release of ruthenium as RuO4(g), bearing in mind that the radioactive isotopes of ruthenium have serious consequences for human health. While relatively comprehensive data on the behavior of RuO4 within the reactor containment is available, we do not have sufficient data regarding the behavior of RuO4 as it is transported through the reactor systems - from the fuel to the containment - to understand and model it. It therefore appeared necessary to perform tests as part of the STEM2 project to cover all the various accident situations that might occur.

The technical content of the STEM2 project was based on:

  1. the key conclusions of the initial STEM project, as presented at the international OECD-NEA/NUGENIA-SARNET meeting held in Marseille from March 30 to April 1, 2015. This workshop primarily focused on the behavior of iodine under accident conditions and on management during accident phases,
  2. technical discussions between the STEM project partners during meetings of the Program Review Group (PRG),
  3. recommendations made as a result of the OECD-NEA/NUGENIA-SARNET workshop held in 2015.

All this led to the definition of the following priorities for research on iodine:


  • the impact of paint ageing on the production of organic iodides (RI), the "Paint Ageing" series of tests
  • the decomposition of iodine oxides (oxides are made up of fine particles formed by the oxidation of volatile iodine), the "IOx" series of tests
  •  the chemical reactivity of iodine oxides with gases such as CO and H2, large quantities of which may be present within the reactor containment, the "Gas-IOx" series of tests
  • the radiochemical stability of multi-component iodine-bearing metallic aerosols transported through the reactor coolant system and entering the reactor containment, the "MC-AER" (Multi-Component AERosols) series of tests

The research priorities below were defined for ruthenium, focusing on the effects on its behavior in the reactor coolant system:


  • the characteristics of the surface of the transport tube (quartz versus pre-oxidized stainless steel)
  • different compositions of air-/steam-rich carrier gas
  • the oxidizing conditions (presence of nitrogen oxides as NOx)
  • the speciation of ruthenium entering the reactor coolant system (RuO2 or RuO4)


Tests were performed using IRSN's CHROMIA experimental radiochemistry platform. To be more precise with regard to the iodine tests, these were prepared at the LEAR "hot laboratory" and the "severe accident" phase, under pressure-temperature-irradiation conditions in the reactor containment, was reproduced using the EPICUR irradiator, the only facility of its kind in the world, to obtain on-line measurements of volatile iodine. This provided data on the kinetics specific to the reaction studied. Twelve EPICUR tests were carried out using iodine-131.


EPICUR facility used for testing iodine © IRSN on the CHROMIA platform

The parametric experiments for the "ruthenium" part of the STEM2 project were carried out using the START facility (Study of the TrAnsport of RuThenium in the primary circuit). Seventeen tests were carried out using the START test bench.


START © IRSN, part of the CHROMIA platform

Key results

The key results from the "iodine" part of STEM2 were:


  • For the tests on the impact of paint ageing on the production of organic iodides (Paint Ageing): Three irradiation tests covering the "medium-term" phase (30 to 60 hours) were performed on steel test specimens painted with EPOXY paint with molecular iodine (I2) deposits, to assess the kinetics of gaseous iodine (I2 and CH3I) release from these paints. Unlike the tests performed under the initial STEM project ("LD" test series), in which the paint was subject only to thermal ageing to simulate ageing (the Arrhenius equation) of paint present within the reactor containment, these test specimens were subjected to a pre-accident phase with thermal ageing due to steam and radiolytic ageing (as an effect of radiation on the paint structure). Last, the results from STEM and STEM2 show that the rate of organic iodide (RI) formation due to interactions between I2 and the paint is relatively slow regardless of the type of ageing process applied to the paint. A production model has been developed based on these data (Bosland and Colombani, J. Radioanal Nucl Chem 314, 1121–1140, 2017) and has been integrated in the ASTEC code. However, the simulations conducted on the PHEBUS-FP tests tend to show that the levels of organic iodine measured by experiment are one order of magnitude higher than the simulated levels. One plausible explanation for this is that there may be other sources of production. This is to be investigated as part of the OECD's future "ESTER" project (Experiments on Source TErm for delayed Releases).

  • For the tests on the decomposition of iodine oxides (IOx) and on the chemical reactivity of iodine oxides with gases such as CO and H2 (Gas-IOx): six irradiation tests covering the "medium-term" phase (30 to 60 hours) were performed on test specimens made of quartz coated with iodine oxides (generated by reactions between ozone and molecular iodine) were carried out to evaluate the kinetics of gaseous iodine release (mostly of I2) under various conditions. The thermal decomposition of iodine oxides produces significant releases of I2, especially at temperatures above 100°C. This instability is consistent with the chemical composition of the oxides tested, namely I2O4. At temperatures higher than 100°C, radiolytic decomposition predominates. In terms of the impact that the gaseous composition has on stability, the most significant factor is water vapor which promotes decomposition.

  • For the tests on the radiochemical stability of multi-component iodine-bearing metallic aerosols (MC-AER): three tests were performed to complete the series carried out previously under the STEM project, examining the stability of cesium and cadmium iodide aerosols, which may be present during the gaseous phase in the form of deposits within the reactor containment, under irradiation for lengths of time designed to cover the "medium term" phase. The aim of these three additional tests was to examine the impact of a multi-component aerosol (CsI), including molybdene (Mo) for example, and to study the behavior of a representative non-soluble aerosol such as AgI. The results obtained confirm the formation of gaseous molecular iodine (I2) for soluble CsI aerosols, while, in the case of insoluble Agl aerosols, there is some decomposition but it is very slow and occurs on a very small scale. This phenomenon, which had been suspected but had never been demonstrated before, may have a significant influence on concentrations of gaseous iodine in the medium term within the reactor containment and ultimately lead to delayed releases.

Based on all the results obtained, the next step will be to improve models of iodine's behavior and then integrate these in the ASTEC simulation code. A study will then be conducted to assess the impact of these models on potential releases (source term) under accident conditions.

The key results from the "ruthenium" part of STEM2 were:

  • the gaseous ruthenium transported through the thermal gradient tube results in the direct transport and revaporization of ruthenium deposits at high temperature (T > 900°C).

  • even in the presence of excessive quantities of water vapor compared to air in the carrier gas, the presence of gaseous Ru was observed. This gaseous Ru fraction increases as air content increases.

  • at the end of the different phases of vaporization and revaporization of deposits in the system, release containing a few percent of gaseous ruthenium was observed (under test conditions) at the system outlet.

  • during the revaporization tests, almost all the ruthenium transported was in gaseous form.


  • the type of tube (quartz versus pre-oxidized stainless steel) has a significant impact on the quantity of Ru transported, which may be explained by possible interactions between the oxides at the surface and the ruthenium.  


  • nitrogen oxides (NO2 or N2O at 50 ppm/vol.) do not increase the fraction of Ru transported under test conditions (T > 400°C).

All the experimental results improve our understanding of the complex phenomena (condensation, chemical kinetics, thermodynamics, etc.) that govern Ru transport through the reactor coolant system and can thus be used to model the physical and chemical behavior of ruthenium oxides. The ultimate goal is to re-evaluate the ruthenium source term integrating these new models and data relative to the behavior of RuO4(g) and factoring in the containment venting and filtration systems.


In the case of iodine, the Paint Ageing tests have meant that it is possible to close the inquiry into the production of organic iodides caused by interactions between I2 and epoxy paint since they confirm the validity of the model developed using the data obtained during the initial STEM project (Phase I: LD test series). Nonetheless, this model does not obtain the same levels of organic iodides as measured in the PHEBUS-FP tests, which points to the need to extend R&D work to factor in other sources of formation (the ESTER project).

Regarding the thermal and radiolytic stability of iodine oxides, it has been demonstrated that these fine particles are not very stable and decompose to form gaseous molecular iodine. The kinetics at work are highly dependent on temperature and humidity, both of which promote decomposition into molecular iodine. This new data will be useful for developing and finessing modeling of severe accident conditions. Regarding metallic iodide aerosols arriving from the RCS into the reactor containment, the small amount deposited on the containment walls and internal structures is likely to be oxidized as gaseous molecular iodine. Regarding soluble iodine aerosols (such as CsI), they are totally and rapidly converted into gaseous iodine, which is not the case for insoluble aerosols (AgI, for example) which are converted very slowly and on a very small scale.

In the case of ruthenium, IRSN studied the behavior of this chemical element in the reactor coolant system, the main source of uncertainty regarding evaluation of the ruthenium source term for scenarios entailing the ingress of air into the reactor coolant system. A few percent of RuO4(g) may reach the reactor containment even if the oxygen content of the carrier gas is very low. To conclude, the experimental data obtained using the START test facility improves our understanding of the transport and the reactivity of ruthenium in the reactor coolant system under oxidation conditions. These data also provide guidelines that we can use to develop and validate the models used in severe accident simulation tools. Last, the ruthenium source terms for air ingress scenarios need to be re-evaluated based on the results obtained.



Following on from STEM2, and under the aegis of the OECD's Nuclear Energy Agency (NEA), IRSN has proposed a new project, called ESTER (Experiments on Source TErm for delayed Releases).

The program has been defined on the basis of the experimental needs expressed at the end of the OECD/NEA workshop on source terms under accident conditions held in Paris on January 21–22, 2019. The program was set out in more detail at a preparatory meeting held at NEA headquarters on September 20, 2019.

The ESTER project will cover two lines of research, related to the source term, for which a need for further data has been identified.

The first relates to delayed releases of fission products (FP) re-emitted from deposits in the RCS; depending on the scale, such releases may have a significant influence on evaluation of the source term. This issue was brought to light during the accident at Fukushima-Daiichi where significant and unexpected delayed releases were observed several days after the beginning of the accident. Iodine, cesium and tellurium are among the volatile or semi-volatile fission products that cause the highest levels of exposure for people and the environment.

The second line of research relates to the chemistry of iodine within the reactor containment and, more specifically, the formation of organic iodides. The BIP (Behavior of Iodine Project) and STEM projects carried out under the aegis of the OECD have advanced our knowledge regarding the formation of organic iodides from the paint present in the reactor containment. However, one key conclusion has been that the production model developed on the basis of the experimental data does not enable an accurate simulation of the formation of the organic iodides identified in integral tests such as the PHEBUS-FP tests. The fraction of organic iodide calculated was underestimated. Given that organic iodine-131 is the major contributor to short-term radiological consequences, due to it being highly volatile and the fact that Filtered Containment Venting Systems (FCVS) are less efficient at trapping it than is the case for I2, a significant amount of organic iodine was probably released during the Fukushima Daiichi accident, and as previously suspected during the accident at TMI-2 (Three Mile Island).

Send Print


​Dates : 2016-2019

Partners : NRC (United-States of America), CNL (Canada), VTT (Finland), SSM (Sweden), KINS et KAERI (South Korea), GRS (Germany), JAEA et NRA (Japan), EDF (France) et NNL (England).

Involved IRSN laboratory



List of STEM and STEM 2 publications

    1. Bosland L., Colombani J., Study of the radiolytic decomposition of CsI and CdI2 aerosols deposited on stainless steel, quartz and Epoxy painted surfaces, Annals of Nuclear Energy 141(15), 2020, Article number 107241.
    2. Ohnet M.N., Leroy O., Mamede A.S., Ruthenium behavior in the reactor cooling system in case of a PWR severe accident, Journal of Radioanalytical and Nuclear Chemistry 316(1), 2018, Pages 161-177.
    3. Cantrel L., Albiol T., Bosland L., Colombani J., Cousin F., Grégoire A.C.,  Leroy O.,  Morin S., Research works on iodine and ruthenium behavior in severe accident conditions, Journal of Nuclear Engineering and Radiation Science 4(2), 2018, Article number 020903.
    4. Ohnet M.N.,, Leroy O., Boucualt K., Gomez C., Ruthenium Transport in the RCS in case of a PWR Severe Accident: a Parametric Study. The 27th International Conference NuclearEnergy for New Europe (NENE), Portoroz, Slovenia, 10-13 September 2018.
    5. Bosland L., Colombani J., Study of iodine releases from epoxy and polyurethane paints under irradiation and development of a new model of iodine-Epoxy paint interactions for PHEBUS and PWR severe accident applications, Journal of Radioanalytical and Nuclear Chemistry 314(2), 2017, Pages 1121-1140.
    6. Miradji F., Cousin F., Souvi S., Vallet V., Denis J., Tanchoux V., Cantrel L., Modelling of Ru behaviour in oxidative accident conditions and first source term assessments. In: The 7th European Review Meeting on Severe Accident Research (ERMSAR), Marseille, France, 24-26 March 2015.
    7. Bosland L., Dickinson, S., Glowa G.A., Herranz L.E., Kim, H.C., Powers, D.A., Salay, M., Tietze S.., Iodine-paint interactions during nuclear reactor severe accidents, Annals of Nuclear Energy 74(C), 2014, Pages 184-199


Send to a friend

The information you provide in this page are single use only and will not be saved.
* Required fields

Recipient's email:*  

Sign with your name:* 

Type your email address:*   

Add a message :

Do you want to receive a copy of this email?