Direct Contact

European Commission Joint Research Centre Institute for Transuranium Elements

Vincenzo Rondinella
Hot Cells
Herman-von-Helmholtz-Platz 1
76344 Eggenstein-Leopoldshafen
Tel.: +49 (0)7247-951-0 E-mail


        False colour tomography of bundles (IRSN Cadarache)

        Kühl Finger Anlage (KüFA)


TMI-2 Accident

On March 28, 1979, there was a nuclear accident that led to a partial meltdown of the reactor core of TMI-2 (Three Mile Island-unit 2) near Harrisburg, Pennsylvania in USA. The accident involved a relatively minor malfunction in the secondary cooling circuit which caused the temperature in the primary coolant to rise. This in turn caused the reactor to shut down automatically. Shut down took about one second. However, a relief valve failed to close, but instrumentation did not reveal the fact, and so much of the primary coolant drained away that the residual decay heat in the reactor core was not removed and the core suffered severe damage as a result.
TMI-2 Examination was then launched as a OECD-NEA led consortium under the initiative of the US Dept. of Energy (US-DOE). INEL Idaho extracted and examined the samples from TMI-2 and delivered samples to North American & European institutes:
• AECL Canada, Argonne National Lab, FZK-Karlsruhe, JRC-ITU Karlsruhe (see figure below), PSI Würenlingen, Studsvik, CEA Saclay, AEA Windscale, JAERI Japan

Positions of the ITU samples in the TMI Core

Principal tasks of the examination

what was corium (& other phases) composition
what temperatures were reached
what conditions (oxygen potentials) prevailed

The examination indicated the following temperatures during cool down:
Edge of reactor: T < 800°C (slight degradation) - transient rise in temp.
Agglomerated material around the melted core: T~1500°C (stainless st. mp) - more rapid & variable
fully molten core T= 2000-2500°C - with slow cooling (over 2-54 h)
some molten UO2 was seen T=2850°C

The phases formed were:

  1. Central Core: This was composed of UO2 fuel and Zry cladding melt (U, Zr)O2: the U, Zr-containing ceramic was made up of a U-rich phase, a Zr-rich phase and smaller amounts of Fe, Ni, Cr oxides & Ag nodules
  2. Outer Agglomerate: This was mixed metallic and ceramic phases from fuel/cladding /structure interactions: eg. (U,Zr)O2 phases, (Fe,Ni)-Zr-U oxides, Ni-Fe-Sn metal, Ni,Fe partially oxidised nodules, & occasional Ag metal nodules

Main results- The temperatures attained were not as high as originally thought.- Certain interactions between zircaloy -stainless steel (especially the Zr-N and Zr-Fe low melting eutectics) clearly initiate the degradation before the melting of zircaloy or UO2 occurs.



Phebus FP Project

The Phebus FP (fission project) project is an international collaboration led by the French Institut de radioprotection et de sûreté nucléaire (IRSN) at Cadarache in which severe nuclear accidents of various types are simulated. Six integral tests are planned with steam flow limitation and bundle overheating. There is on-line monitoring of bundle degradation and fission product release. The instrumentation of the simulated primary circuit and containment downstream permits the subsequent behaviour of the fission products after release to be monitored.

Phebus Reactor in Cadarache

The contributions of the ITU-Hot Cells are:

  • sectioning of degraded bundle into 14 x 2cm thick discs
  • microscopic examination and analysis of phases at selected points of the bundle to establish the principal interactions.
  • examination of  some simulated primary circuit samples positioned downstream of the bundle (PTA).
  • revaporisation tests of fission product deposits

Scheme of the Phebus facility

Phebus PF test matrix

  • FPT0: (base-line test) degradation of a non-irradiated, 20 x 1m fuel rod bundle under a steam atmosphere (Dec ‘93).
  • FPT1: degradation of 18 irradiated fuel rods (23 GWd/tU mean burn-up) - bundle was degraded under steam (July ‘96).
  • FPT4: examination of the late stages of an accident with the degradation of a debris bed of irradiated fuel pieces (33 GWd/tU) and oxidised cladding pieces under steam. Here the semi- and non-volatile fission products were particularly investigated (July ‘99).
  • FPT2: degradation of irradiated fuel bundle (35 GWd/tU) with a period of low steam flow, to see the effect of reducing conditions on bundle degradation interactions (Oct. 2000).
  • FPT3: degradation of irradiated fuel bundle with B4C absorber present in the reactor (compared to Ag-In-Cd for previous tests) - (~Oct 2004)


Phebus FP Project: Post irradiation examination (PIE)

Phebus shows how bundle degradation is a regular process, with the formation of a molten pool as the central (hottest) section of the bundle melts and collapses together onto cooler structures where a solidified base enables the corium pool to collect, below a central cavity. The remaining upper structure is still partially in place but is weakened, highly oxidised and brittle.
Sectioning of the bundle is performed at key positions of the bundle in  order to verify the tomographic sections seen below.  Further sampling of areas of special interest (molten pool, degraded pool, candling rods) is done to carry out a detailed examination and determine the mechanisms of fuel/cladding or structural material interaction (see Fig. below).



False colour tomography of bundles (IRSN Cadarache)

False colour tomography of a) FPT0 and b) FPT1 bundles showing the formation of dense molten pools (red) at 1/4 height with central cavities (purple & black) above them.

Cladding of outer fuel rod undergoing
degradation by corium attack

False colour tomography of bundles (IRSN Cadarache)




The main objective of the CIT (Corium Interaction Thermochemistry) and COLOSS (COre LOSS of Geometry) projects was the examination of high temperature material properties and their modelling relevant to reactor accidents. ITU carried out single effect tests to investigate the corium (melted reactor materials) formation and properties.

CIT and Coloss Main Tasks

  • natural UO2 and Zircaloy interaction
  • irradiated UO2 and Zircaloy interaction

Other interactions e.g. FeO-natural UO2, Ag-Zry-Fe-UO2 were investigated; thermodynamic data measurements and corium viscosity measurements were also carried out. On the theoretical side, phase diagram calculations and modelling of the interactions and reactor behaviour were performed.

Testing Conditions of UO2/Zircaloy interaction at ITU


  • Fuel  (5 mm segments)- irr. UO2 53 GWd/tU- non-irr. UO2 and depleted UO2
  • with Zry-cladding (outer oxide thickness 5 - 10 µm)
  • Direct electrical heating in graphite crucible
  • He - Atmosphere (550 ml.min-1)



  • Fuel (5 mm segments) with Zr-cladding - irr. UO2  ~90 GWd/tU- non-irradiated UO2- irr. MOX  ~45 GWd/tU- non-irradiated MOX
  • Indirect electrical heating in Y2O3crucible
  • He - Atmosphere (500 ml.min-1




  • optical microscopy, image analysis, microprobe analysis




In the CIT project the first investigations on fuel-cladding interaction were done. The pictures on the right show horizontal cuts of the samples after heating in the furnace. After the first high temperature tests the facility broke down and the further investigations of this single effect test were carried over to the COLOSS project.

2000°C-190s; non-irradiated fuel (left); irr. fuel (3 GWd/tU)


In the COLOSS experiments two samples, irradiated and non-irradiated, were tested together in order that both samples had identical conditions. The examples here are  irradiated MOX & natural UO2 heated to 2000°C for 10 mins. in He atmosphere

MOX (left); nat. UO2 (right)

The top view shows a quite uniform pattern for the nat. UO2 fuel. The surface of the fuel is still smooth and the cladding shows a uniform structure. The MOX sample underwent the same temperature history, but already it shows a more extensive attack. The fuel surface is much rougher and the cladding has no structure remaining.
After the experiment the samples were examined under the optical microscope to investigate, the extent of the attack and the phase distribution.
The microscopy confirms the first observations. The attack of the nat. UO2 by the liquid cladding material is quite uniform. The fuel is dissolved continuously from the outside. The frozen melt shows a regular two-phase dendritic structure of a bright, metallic-rich phase and and a darker ceramic (UO2)-rich phase, as the dissolved UO2 fuel precipitates out on cooling.
The cross section of the MOX sample shows a much more advanced attack compared to the non-irradiated sample. Large frozen fission gas bubbles that lifted the sample up can be observed. The melt shows a more heterogeneous 2-phase structure and with almost no dendrite structure. The irradiated MOX has porosity and cracks from the irradiation which enables the liquid cladding to penetrate more easily into the fuel and thus accelerate the dissolution process.
Considerable work remains to create a reliable database for severe accident modelling, however the burn-up of the fuel certainly is an accelerating factor due to its fission gas release and rapid fracturing. This behaviour will also depend on the type of fuel.



HT Reactor research & the KüFA-facility

In the 1950s, Dr Rudolf Schulten had the idea to compact silicon carbide-coated uranium granules into hard billiard-ball-like graphite spheres to be used as fuel for a new high-temperature, helium-cooled type of reactor. A 15 MW (megawatt) demonstration pebble bed reactor (AVR), was built and operated successfully for 21 years in Germany.
The AVR Research Reactor, a 40 MW (thermal) 15 MW (electrical) HTR based on the pebble bed fuel concept was then built to both test and demonstrate fuel design, fuel loading configurations and HTR safety characteristics. Over its 21-year lifespan, it reached a utilization factor of 67%.
The interest in the HTR technology has revived considerably in the last 5 years especially in Japan, China and in at Koeberg (South Africa) where demonstration reactors have been, or are being, built. The Generation IV International Forum (GEN IV forum) has also ranked it as one of the most promising design for sustainable development in the next generation of nuclear reactors.



Kühl Finger Anlage (KüFA)

ITU is collaborating with Forschungszentrum Jülich GmbH and others in the HTR Shared Cost Action (SCA) of EC’s 5th Framework Program in which ITU will carry out fission product (gas & volatile) release measurements on high burn-up irradiated fuel elements (pebbles) from the HTR, Jülich reactor.
The objective is to test their ability to retain the accumulated fission gas even at extended burn-ups and after long periods of storage. During operation they have already shown excellent fission gas retention in the kernels within the pebble and in the pebbles themselves at temperatures upto 1800°C: this represents a loss-of-coolant-accident for the reactor.



View of the cool finger and oven on the base
plate of the new caisson 103 behind the Hot Cells