The LWR-PROTEUS Phase II measurements 1. Samples

With the aims of Phase II mentioned previously, burnt fuel samples were first identified to constitute a relevant set for systematic analyses. Criteria used for the selection were: (i) to use modern fuel samples (e.g. high ²³⁵U enrichment and high discharge burnup values) from the same fuel vendor and extracted from the same nuclear power plant after different cycles, in order to reduce the uncertainties in their characterization (design, construction, irradiation history, etc.); (ii) to cover both UO2 and MOX fuels; (iii) to have a wide range of burnup values, from low to very high burnups; and (iv) to select the fuel samples from the axial region in the fuel rod where the burnup profile is flat and the burnup level is high, while also avoiding the effect of perturbations due to heterogeneities (e.g. the presence of spacers). Using these criteria, 6 UO2 and 3 MOX fuel rods were finally identified; their basic characteristics are shown in Table I. From each of these rods, a sample of 40 cm length was cut from an axial section between two spacer locations.

Table I. Selected Burnt Fuel Rods in Phase II

Fuel Type Initial Enrichment Discharge Burnup Discharged

UO2 4.1% ²³⁵U 36 GWd/ton June 1997

UO2 3.5% ²³⁵U 46 GWd/ton June 1995

UO2 3.5% ²³⁵U 64 GWd/ton June 1995

UO2 3.5% ²³⁵U 65 GWd/ton June 1995

UO2 3.5% ²³⁵U 82 GWd/ton June 1995

UO2 3.5% ²³⁵U 80 GWd/ton June 1995

MOX 5.5% Pufiss 21 GWd/ton June 1998

MOX 5.5% Pufiss 37 GWd/ton July 1999

MOX 5.5% Pufiss 53 GWd/ton July 2000

Each of the burnt fuel samples has been overcanned with a special zircaloy cladding and welded tight using a certified procedure to guarantee leak-tightness and absence of contamination. The over-canning is a safety measure usually employed for the return of tested, highly active fuel rods from PSI back to the nuclear power plants. It follows general “defense-in-depth” criteria to prevent contamination in case of an accidental leakage of activity from the primary cladding.

Special fresh UO2pellets doped with various additives have been fabricated by Westinghouse Atom, the different pellet types being shown in Table II. The undoped pellets serve as a reference for the experimental determination of the reactivity effect of the replacement of a fresh fuel rod by a burnt one. The enrichment of 4.3% is identical with that of the rods in the PWR test region, whereas 3.5% corresponds to the initial composition of the majority of the burnt UO2 samples. The fuel pellets doped with ¹⁰B are used for the calibration samples against which the relative reactivity worth of the burnt samples is measured. The reactivity worth of the 5 most important fission products (in terms of contribution to the reactivity loss) are investigated directly using pellets doped with the corresponding nuclides.

4.2. Post-irradiation examinations

The first set of experimental results has been obtained in relation to the reception of the full-length burnt fuel rods from KKG and the performance of post-irradiation examinations (PIE) at the PSI Hotlabor. These examinations include: (i) visual inspection, cladding inspection and determination of off-nominal conditions; (ii) measurement of fuel rod elongation and fuel rod diameter as a function of the rod axial length; (iii) axial gamma scans with a resolution in the order of millimetres for the determination of burnup and fission product axial profiles.

The fuel rods were then punctured and cut into a few fuel segments in order to perform extensive analyses including fission-gas release determinations and metallographic studies over the segment cross-sections to study outer and inner oxide thickness growth, pellet grain size and structure, as well as the uniformity of the cladding.

Table II. Special fresh UO2 samples

Enrichment Additive

4.3% ---

3.5% ---

2.1% ---

0.71% ---

3.5% ¹⁰B 2 concentrations

3.5% ¹⁴⁹Sm 3 concentrations

3.5% ¹⁵⁵Gd 2 concentrations

3.5% ¹⁰³Rh 2 concentrations

3.5% ¹⁴³Nd

3.5% ¹³³Cs 3 concentrations

3.5% 5 fission products

3.5% Gd enriched in ¹⁵⁵Gd and ¹⁵⁷Gd

4.3. Reactivity measurements

The introduction of the 40 cm long samples, one at a time, into the central PWR test region in PROTEUS allows the experimental determination of the reactivity loss due to burnup. Due to the complex three-dimensional geometry of the driven PROTEUS system, it would be very challenging to calculate the absolute reactivity worth. For the validation of lattice codes however, it is more appropriate to interpret the response of the samples of interest relative to those of well-characterized reference and calibration samples. The design calculations have shown that, particularly for small signals (a few cents), these ratios in the driven configuration are the same as in the corresponding (hypothetical) single-zone system.

The burnt samples are transferred to PROTEUS in a special transport cask. This cask consists of a steel container for shielding, similar to those routinely used for the transport of active samples within PSI, and includes a carousel system and drive mechanisms for the remote-controlled insertion of the selected samples into the reactor. The design of the cask allows the simultaneous transfer of up to 4 burnt fuel samples plus two unirradiated reference samples.

The transport cask is lifted on top of the PROTEUS reactor and located on a specially prepared support structure in the centre of the upper plate (see Figure 1). After the mechanical and electrical connections are made, the PROTEUS shielding doors are closed and the transport cask remains in place above the reactor throughout the reactivity measurements. The samples are then moved, one at a time, into and out of the guide tube in the centre of the reactor.

The measurements are being performed for two different moderation conditions, and hence different neutron spectra, in the PWR test region, viz. full-density light water and a mixture of

H2O and D2O in the proportion of about 2/3 to 1/3. The latter mixture simulates the water density in an operating PWR (at 300 ºC) which is about 0.7 g/cm³

The reactivity effects of the samples are determined by compensation with a calibrated fine control rod, which is moved automatically so as to maintain the reactor exactly critical. In addition, reactivities are also obtained from the observed evolution of the neutron flux (with the automatic control rod in a fixed position) by solving the kinetics equations. The expected reactivity worths are of the order of a few cents. The measurement of the burnt samples along with reference and calibration samples during continuous reactor operation helps to reduce systematic errors considerably; the repeated insertion and withdrawal of the sample (oscillation) allowing to eliminate the effects of possible drifts.

Photoneutrons are emitted in the regions containing heavy water, which influence the time-dependent behaviour of the reactor in an analogous way to the delayed neutrons, but with longer time constants. This effect is taken into account in the analysis of the experiments.

Since only parts of the fissions occur in the D2O-moderated region, a “photoneutron efficiency factor” is determined experimentally for each configuration. Burnt fuel samples, particularly high burnup and MOX fuel, emit neutrons by spontaneous fission and (α,n) reactions. The effect of this external neutron source is discriminated from the reactivity variation by performing measurements at different flux levels.

Experience from previous PROTEUS experiments has shown that reactivity effects of 10 cents and more can be measured to an accuracy of better than 0.1%, whereas effects of about 1 cent can be measured with an accuracy of about 1% using the techniques described above. Calculations made at PSI to estimate the reactivity effect of replacing a central fresh UO2 rod with different burnt fuel samples showed that rods differing by 2% in k_∞ would yield signals differing by about 0.08 cent per rod. Since signals of one cent can be measured with an accuracy of around 1%, i.e. 0.01 cent, this indicates that the methods proposed will be able to provide a reliable validation base, even for small effects.

4.4. Chemical assays

In addition to the 40 cm long rodlet for the reactivity measurements, a second (contiguous) sample of ~1 cm length (i.e. the size of a pellet) was extracted from each of the burnt fuel rods (see Table I). This is used to conduct destructive chemical assays at the PSI Hotlabor for the accurate quantification of individual isotopes.

The isotopic analysis of the burnt fuel samples is based on chemical assays in the PSI hotcells.

The choice of isotopes to be analysed corresponds to the ARIANE “basic programme” [4] and includes the most important actinides,

234U, ²³⁵U, ²³⁶U, ²³⁸U, ²³⁸Pu,²³⁹Pu,²⁴⁰Pu,²⁴¹Pu, ²⁴²Pu,²³⁷Np, ²⁴¹Am, ^242mAm, ²⁴³Am, ²⁴²Cm,

243Cm,²⁴⁴Cm,²⁴⁵Cm, and fission products,

142Nd, ¹⁴³Nd, ¹⁴⁴Nd, ¹⁴⁵Nd, ¹⁴⁶Nd, ¹⁴⁸Nd, ¹⁵⁰Nd, ¹³³Cs, ¹³⁵Cs, ¹³⁷Cs, ¹⁴⁷Sm, ¹⁴⁹Sm, ¹⁵⁰Sm,

151Sm,¹⁵²Sm,¹⁵³Eu, ¹⁵⁴Eu, ¹⁵⁵Eu, ⁹⁵Mo,⁹⁹Tc, ¹⁰⁹Ag, ¹⁵⁵Gd, ¹⁰³Rh,¹⁰¹Ru,¹⁴⁴Ce, ¹⁰⁶Ru,¹²⁵Sb,

129I, ¹⁴⁷Pm.

In addition to the fission products important from the neutronics point of view, i.e. for their contribution to the reactivity loss, this list also includes nuclides used as burnup indicators such as ¹³⁷Cs and the Nd isotopes.

The chemical assays are performed using high performance liquid chromatography/ion chromatography (HPLC-IC) and inductively coupled plasma mass spectrometry (ICP-MS).

These methods were applied successfully by PSI in the ARIANE project. The LWR-PROTEUS Phase II programme will thus benefit from available expertise and well-established experimental techniques for obtaining isotopic compositions of burnt nuclear fuel.

5. Calculational analysis