Optimization of process parameters for the sintering of MOX fuel

Optimization of process parameters for the sintering of MOX fuel

Journal of Nuclear Materials 178 ( 199 I ) 152- I57 North-Holland 152 Optimization of process parameters for the sintering of MOX fuel R. Giildner a...

2MB Sizes 4 Downloads 87 Views

Journal of Nuclear Materials 178 ( 199 I ) 152- I57 North-Holland


Optimization of process parameters for the sintering of MOX fuel R. Giildner and H. Schmidt SemensAG,


P.O. Box 110060, W-6450 Hanau II, Germany

The quality of MOX fuel is strongly influenced by the sintering process. For this reason experiments have been carried out in a batch-type laboratory scale sintering furnace to determine the optimal process parameters. It has been proved that good and stable quality in the fuel can be obtained with relatively short sintering times with an adequate temperature profile and gas humiditication adapted to the actual partial process. The most essential results are transfered to the routine fabrication of MOX fuel for Light-Water Reactors.

The requirements for characteristics in MOX fuel result from those of its in-pile behaviour and the solubility of the irradiated fuel during the head end process in the reprocessing plant. To meet the solubility requirements two different processes have been developed by Siemens, Hanau (formerly ALKEM), the OCOM process (eptimized aMilling) and the AU/PuC process (Ammonium-IJranyl/-Plutonyl-Carbonate) [ 11. -The flow diagrams of both processes are shown in fig. 1. In both processes the powders are pressed after the mixing operation without any additional powder preparation, e.g. granulation. Until now around 130 t of MOX fuel have been produced by the OCOM process and around 17 t by the AU/PuC process with a Pu-content of up to 7% (around 5O/alissile Pu). Sufficient solubility is reached with both processes. In addition to the two virgin powders mentioned above, reactivated sintered material (RSM) was added during mixing. The in-pile behaviour of the fuel is influenced by the characteristic properties such as density, microstructure and resintering activity. These characteristics are formed during the sintering of the green pellets. In addition sintering is usually the slowest process step during pellet fabrication and it is sensitive regarding quality aspects. Therefore an optimization of the sintering parameters may well increase the capability of the entire plant, if it allows a reduction in the sintering times.



1. Introduction PUOZ ex






ex AK













Fig. 1. Flow diagrams of the OCOM and AU/PuC process.

2. Properties of powders and green pellets As shown in fig. 1 in both processes three different powders are used for the fabrication of the mixture which

0022-3 1 I5/9 I /$03.50 0 199 1 - Elsevier Science Publishers B.V. (North-Holland)

R. Giildner, H. Schmidt /Process parameters for the sintering of MOXfuel


Table 1 properties of powders

OCOM U/Pu02 ex OCOM Pu content (%) Proportion in the mixture (%) O/M Particle size dSO(pm)

Pour density (g/cm’)

-30 lo-25 2.15-2.25

-0.8 -3.5



U/PuO, ex AU/P&




2-7 6-10 - 2.05

-2.0 -4.0

70-80 2.10-2.15 - 10.0


lo-25 2.20-2.30 - 10.0


6-10 -2.05

-2.0 -4.0


70-80 2.10-2.15 - 10.0


‘) RSM= reactivated sintered materials.

serves as press-feed. Additionally around 0.2 wt% zincstearate is added as a lubricant. It is obvious that the course of the sintering process is influenced by the properties of the powders and the composition of the mixture. The most important properties of the powders and the range of their proportion in the mixture are summarized in table 1. Table 1 shows that - especially during the OCOM process - powders with very different properties are mixed together. This means that the homogeneity of the fuel, which is important for the in-pile behaviour, is produced during the sintering of the fuel by diffusion processes. The press-feed is pressed to green pellets with a density of 5.8-5.9 g/cm3 and dimensions of around 11.2 mm (diameter) and around 14 mm (height). The O/M ratio varies between 2.13 and 2.2 1.

3. Course of the sintering process During sintering the reduction to an O/M ratio 52.0, the densification from - 5.8 to - 10.4 g/cm3, the formation of the microstructure (pore structure, grain size, homogeneity), the decomposition and evaporation of organic additives, and the evaporation of residual humidity and volatile impurities occur simultaneously or one after the other. As far as the method is concerned, two different sintering processes, exist. The commonly used sintering in a reducing atmosphere, where the fuel is reduced during the heating-up phase with densification and formation of the microstructure afterwards, and the so-called NIKUSI process, where first densification occurs in an oxidizing atmosphere followed by reduction in a reducing atmosphere [ 2,3]. Because of the gas separation required, realization of

the NIKUSI process is difficult under the conditions stipulated for handling plutonium. Therefore the experiments have been restricted to the reductive sintering. The quality of the fuel is influenced by the evolution of the different partial processes during sintering particularly the formation of the microstructure. With regard to this, one of the most important effects is the diffusion of cations within the crystal lattice of the fuel. The diffusion rate depends on the temperature and the stoichiometry of the crystal. It increases with increasing temperature but decreases significantly if the crystal becomes hypostoichiometric, this means if oxygen interstitials disappear, and reaches a minimum at an O/M ratio of 1.98 [ 4,5]. In (U,Pu)-mixed oxides hypostoichiometry can be reached if the valence of plutonium decreases to under 4.0. According to the model of Breitung [6] the reduction of the plutonium can be shifted to higher temperatures and the stoichiometry of the material can be kept well above the minimum of the diffusion process during the entire sintering process by means of an adjustment of the oxygen potential of the sintering gas. The oxygen potential of a moistened Ar/H, sintering gas depends on temperature and the H20/H2 ratio, which means that at a given temperature and with a constant H2 concentration the oxygen potential can be adjusted by the Hz0 partial pressure. Some results concerning the fabrication of FBR-fuel have been published by Elbel et al. [ 7 1.

4. Experiments

In a batch-type laboratory scale sintering furnace about 120 experiments were carried out to optimize the temperature profile and the gas composition for the sintering of


R. Giildner, H. Schmidt /Process parameters.for the sintering of MOXfuel

MOX fuel. The main aim of the experiments was to determine favourable conditions for the cation diffusion to obtain a homogeneous and well developed microstructure. Ar/H, (4% HZ) dry or with variable moisture was used as sintering gas. Due to the design of the furnace the influence of the reaction products of the sintering process particularly water - on the atmosphere is limited to the time of their evaporation. The heating rate was 300 and 600”C/h respectively and the temperature maximum was I 700’ C. The influence of isotherms was studied at different temperatures. To obtain further information on intermediate stages, some experiments were stopped at different temperatures and the pellets were analyzed. Additionally the influence of the characteristics of the mixture, especially Pu and RSM content and O/M ratio, was examined. Green pellets from around 30 mixtures, from both OCOM and AU/PuC process, were obtained from routine pellet fabrication with properties in the range shown in section 2. The Pu content varies from 3.5-6.7% and the RSM content from 6-g%.

5. Results Essential criteria for the assessment of the experimental results are: density, O/M ratio, grain size and solubility of the sintered pellets. The last two characteristics can be regarded as a measure of the diffusion process. With sufficient moisture (log pH2JpH2 = - 1.5 to - 1.O) up to the maximum temperature it is certain that the stoichiometry can be kept above 1.99 during the entire sintering process. Compared with this, a stoichiometry of around 1.98 was reached with dry sintering gas, which significantly reduces the diffusion velocity. Fig. 2 shows a comparison of the grain structure after a sintering test with dry and constantly moistened Ar/H* respectively. With dry gas the grain size was 2-3 pm (linear intercept) at a density of 10.45 g/cm3 and an O/M ratio of 1.982. The corresponding values for the moistened experiment are 8 pm, 10.35 g/cm3 and I .996. With a variable moisture of log p,&pH2 = - 1.5 to - I .O in the temperature range from 600 to 1700°C the hypostoichiometry is reached between 1000 and 1200°C depending on the heating rate and the Pu content. Table 2 shows some values from experiments which were stopped at different temperatures. The O/M ratio of the mixture utilized was 2.17. The humidification of the sintering gas not only influences the reduction process but also the densitication. Fig. 3 shows the dependence of the density from the temperature for dry and moistened gas respectively. In these experiments several mixtures were examined and the heat-

Fig. 2. Comparison of grain structure of MOX fuel sintered with dry and moistened Ar/H* (4% H,), respectively. (a) Dry sintering gas, (b) moistened sintering gas.

Table 2 Dependence of density the heating rate Temperature

and O/M


ratio on the temperature






O/M: 2.002 Density: 8.6 g/cm3

O/M: 2.007 Density: 7.75 g/cm3


O/M: 1.997 Density: 9.55 g/cm3

O/M: 2.002 Density: 9.03 g/cm3

R. Giildner, H. Schmidt /Process 10.5

Ar/H2 Ar/H2

Q dry Clmoist.


Hzl Hz1

parameters for the sintering of MOXfuel















temperature (“C) Fig. 3. Densification of LWR-MOX fuel with dry and moistened sintering gas (heating rate 300”C/h).

was 300”C/h. The figure shows that the densification starts significantly earlier with moistened gas and is considerably advanced at a temperature of 1400°C. The diffusion process is noticeable above 14OO”C,but the diffusion velocity is rather high under experimental conditions with moistened sintering gas. Fig. 4 shows SEM photographs of the grain structure after sintering tests with variable moistened gas and a heating rate of 300”C/h. The first test (fig. 4a) was stopped at 14OO”C,the second (fig. 4b) at 1600°C. The maximum temperature in the third test was 1670°C with an hold time of 1 h (fig. 4~). The increase of the grain size is significant and the velocity of grain growth becomes obvious in considering that the total sintering time is only around 2 h longer in the third test compared to the first one. The different properties of the mixture require different moisture profiles. At temperatures from 600 to 1400°C the moisture is essential to control uranium reduction and to stimulate densitication (see fig. 3). In this range the humidification must be increased with increasing O/M ratio of the green pellet to reach the required density of the product. In the higher temperature range the moisture serves to ing rate

Fig:. 4. Development of the grain structure of MOX fuel (heatin rat1e 300”C/h). (a) T,,,,,=14OO”C, (b) T,,,.,=l6OO”C, (C T,,,,,= 1670°C (isotherm 1 h).


R. Giildner, H. Schmidt /Process parameters for the sintering of MOXfuel

control the reduction of the plutonium increased with increasing Pu content

and so it must be and also with in-

creasing RSM content to obtain a homogeneous microstructure. No significant difference was found between OCOM and AU/PuC mixtures with similar stoichiometry and Pu content. Concerning the characteristics of the sintered pellets the best results were obtained under the following conditions: - Heating rate 300”C/h; - Short hold ( 1 hour ) at 600’ C; - Variable moisture (10gpH20/PH2= - 1.5 to - 1.O) in the temperature range from 600 to 17OO”C, depending on the powder properties. To avoid crack formation, as described for FBR fuel in ref. [ 71, and reoxidation, the humidification must be cut off in the cooling down phase. Under these conditions a maximum temperature of 1700” C with a hold time of around 2 h, which means a total sintering time of 9 h excluding the cooling down phase, is sufficient to produce a MOX fuel which meets all the specified requirements. The microstructure of such a fuel is shown in fig. 5 (fig. 5a: grain structure; fig. 5b: pore structure; fig. SC: Ly-autoradiography). The mean grain size is 10 pm (linear intercept) and the solubility of the unirradiated fuel is in this case 99.99%, which means in practice it is completely soluble. Generally a solubility of > 99.95% was reached. This applies to both OCOM and AU/PuC fuel. Thus the experiments have shown that the characteristics of the sintered fuel depend on the properties and the composition of the initial powder mixtures. The sinter tests described are to be supported by thermogravimetry and differential thermoanalysis measurements.

6. Transfer of the experimental results to routine MOX fuel fabric&&

Fig. 5. Microstructure of MOX fuel sintered under Ar/H, (4% Hz) with variable humidification (T,,,., 1700°C; hold time 2 h; heating rate 300”C/h). (a) Grain structure, (b) pore structure, (c) cx-autoradiography.

Transfering the experimental results to routine fuel fabrication the different concepts of the furnaces must be considered. In the production lines pusher type furnaces are installed with a gas flow against the product flow, which means that the temperature profile is not freely variable and the sintering atmosphere is strongly influenced by the reaction products of the sintering process, especially water. Nevertheless some of the results could be transfered successfully to the MOX fuel production. The humidification of the high temperature zone of the production furnace was raised to a value of log pHz0/pH2 = - 1.5 to - 1.O depending on the Pu and RSM content of the powder mixture and the supply of dry gas in the presintering

R. Giildner, H. Schmidt /Process parameters for the sintering of MOXfuel

zone was reduced. Additionally the charge intervals were reduced to 60 minutes, corresponding to an hold time of around 3 h at a maximum temperature of 1700 to 1720’ C. With these measures the sintering process could be stabilized with a higher product quality. This becomes evident especially for the diffusion dependent characteristics such as grain size (S-10 pm) and solubility ( > 99.9%) as well as the mean variation of the pellet density. Not quite solved is the prediction of necessary variations in the sintering atmosphere depending on the different powder properties because of the influence of the reaction products.

7. Conclusions


ported in [ 8 ] that optimum sintering conditions are more effective to ensure sufficient diffusion than long sintering times. The results of laboratory scale sintering tests could be successfully transfered to actual fuel fabrication.


[ 1] D. Hanus, J. Krellmann and R. Liib, Jahrestagung Kemtechnik (1982) p. 377. [ 21 H. Assmann, W. Doerr and M. Peehs, J. Am. Ceram. Sot. 67 (1984) 631.

[ 3 ] H. Assmann, W. Doerr and M. Peehs, in: Proc. 13th Int. Conf. Temperature and especially moisture profile are essential for the sintering process used in production and therefore for the quality of the MOX fuel. In addition to the moisture, formed by the reduction process a variable humiditication of the sintering gas is necessary to ensure a sufficient diffusion. This applies to both the batch type and the pusher type furnace. With adjustment of the sintering parameters to the different powder properties like stoichiometry and Pu content a well characterized MOX fuel can be produced, even with reduced sintering times. This shows in good coincidence with some results re-

on Science of Ceramics ( 1985). [4] D. Glasser-Leme and Hj. Matzke, J. Nucl. Mater. 106 ( 1982) 211. [5] D. Glasser-Ixme and Hj. Matzke, Solid State Ionics 12 (1984) 217. [6] W. Breitung, KfK 2363 (1976). [7] H. Elbel, J. Klews and R. L6b, J. Nucl. Mater. 153 ( 1988) 160. [8] B.T. Bell, J. Edwards and H.M. Macleod, in: Proc. IAEA Technical Committee Meeting on Recycling of Plutonium and Uranium in Water Reactor Fuel, Cadarache, France, November 13-16,1989 (IAEA, Vienna, 1990) p. 124.