FUEL ACCIDENT PERFORMANCE TESTING FOR ... - Science Direct

4 downloads 0 Views 2MB Size Report
Limits of Coated Particle Fuel III: Fission. Product Migration in HTR Fuel, Dragon Project Report ... lung eines Matrixmaterials zur Herstellung gepresster. Turner ...
19

Journal of Nuclear Materials 171 (1990) 19-30 North-Holland

FUEL ACCIDENT W. SCHENK

PERFORMANCE

‘, G. PO’IT

’ Institui ftir Reakiorwerkstoffe Fed. Rep. Germany

TESTING

’ and H. NABIELEK and 2 HTA/HBK

FOR SMALL HTRs

*



Projekt, Kemforschungsanlage

Jiilich, Postfach 1913, D-51 70 Jiilich I,

Received 15 July 1989; accepted 10 November 1989

Irradiated spherical fuel elements containing 16400 coated UOz particles each were heated at temperatures between 1600 and 1800°C and the fission product release was measured. The demonstrated fission product retention at 1600 o C establishes the basis for the design of small modular HTRs which inherently limit the temperature to 1600°C by passive means. In addition to this demonstration, the test data show that modem TRISO fuels provide an ample performance margin: release normally sets in at 1800°C; this occurs at 1600°C only with fuels irradiated under conditions which significantly exceed current reactor design requirements.

1. Introduction Small high temperature reactors (HTRs) are designed such that maximum fuel temperatures in accidents remain below 1600°C without active control mechanisms [l]. This temperature limit is based on thermohydraulic calculations and the temperature limiting features have been demonstrated by experiments using the Jiilich AVR reactor [2]. At the same time, experiments have been started to determine the maximum temperature where modem HTR fuel retains all fission products of radiological significance. The data discussed below show that it is indeed possible to obtain fission product retention at the temperature maximum with available coated particles. The coated particle consists of a 0.5 mm diameter fuel kernel and several coating layers. First, a porous buffer layer to provide the free volume for fission gases and sacrificial material to stop fission fragments. This is followed - in the TRISO particle - by three highly dense layers of pyrocarbon (PyC), silicon carbide and pyrocarbon. Both F’yC and SIC retain fission gases completely and SIC provides the ultimate barrier to volatile fission products. These three dense layers constitute an effective pressure vessel due to the high strength of SIC and the prestressing effects of PyC during irradiation. Typical particle data are given in

* Dedicated to Prof. H. Nickel on the occasion of his 60th birthday. 0022-3115/90/$03.50

table 1. Older fuel particles of the BISO type (e.g. THTR fuel) have only one thick pyrocarbon coating layer. Spherical fuel elements of 60 mm diameter, as used in AVR and THTR, have been subjected to irradiation and heating tests in the demonstration programme. They have an outer fuel free zone of 5 mm thickness and a fuel zone with a low particle packing fraction [3].

Table 1 Characterization of coated UO, fuel particles Coating batch no.

EUO 2308

HT 232-245

EUO 2309

Irradiation tests

HFR-K3 FRJZ-Kl3 SL-Pl HFR-P4/1,3

AVR

HFR-P4/2

497

500

497

Kernel diameter (pm) Uranium enrichment

9.82%

9.828

9.82%

Coating thickness (pm) Buffer layer Inner PyC layer Sic layer Outer PyC layer

94 41 36 40

93 38 35 40

93 37 51 38

919

912

922

Particle diameter (pm)

0 1990 - Elsevier Science Publishers B.V. (North-Holland)

W. Schenk et al. / Fuel accident testing for HTRs

20

2.1. Heating tests

1

temperatuie

1, 0

I

100

,

200

I

300

I

400

500

Time (hours) Fig. 1. Temperature evolution during of a small HTR, and in the heating elements.

a loss-of-coolant accident tests with irradiated fuel

The 1600°C temperature limit in the MODUL reactor is achieved by a core design with a thermal power of 200 MW and a low power density, 3 MW/m3. In the tall and slim core, the heat is transported away in the case of a loss of coolant accident by passive means, i.e. by radiation and conduction through the walls. Also, this reactor design places very modest requirements on fuel and graphite during normal operations. Irradiation temperatures are 700-900°C the target bumup is 9% FIMA and the accumulated fluence of fast neutrons is 2.1 x 10z5 rnm2. The maximum fuel temperature during uncontrolled core heatup is 1150” C in a pressurized system. For a complete loss of coolant, a nominal maximum temperature of 1520 o C is attained in a small portion of the core [4]. When all calculational uncertainties are added up in a conservative manner, the peak core temperature may rise to just above 1600°C in 30 h (see fig. 1). Accident simulation tests reported here are conducted at constant 1600, 1700 and 1800°C for much longer times to determine the performance margins of the fuel system.

2. Experimental procedures Accident simulation tests have been performed since the mid-seventies both in this laboratory and elsewhere [5] in which KFA has concentrated on heating complete spherical fuel elements. An early experimental programme consisted of heatup ramp tests with (Th,U)O, BISO fuel up to 2500°C [6]. This programme was followed by work with fuel elements containing (Th,U)O, TRISO and UO, TRISO particles [7,8] which is still going on.

The equipment consists of a heating furnace located in a gastight box in the hot cell. The furnace is connected to a helium loop working at a slight overpressure with blowers and filters outside the shielded area (HTRs have a helium pressure between 1 and 7 MPa during normal operations; the atmospheric pressure in the heating tests is representative of the loss of coolant accident). To simulate the slow heatup characteristic of HTRs and for experimental considerations, the following 30 h procedure (fig. 2) was established prior to each test: (i) measurement of *‘Kr release at room temperature as a qualitative indication of the state of the fuel; (ii) heating at 300°C for 5 h to clean the helium circuit and to remove moisture from graphite components; (iii) simulation of operating temperatures at 1050/125O”C for equilibration of fuel and fission products; (iv) heatup to target temperature at a rate of 47”C/h. Outside the hot cell box are noble gas traps which contain charcoal cooled by liquid nitrogen. In these, the activity of “Kr is mo m‘tored with NaI(T1) detectors which generate the release curves during the heating test. To enable measurements of metallic fission products, a water-cooled finger intrudes into the heating furnace, which can be removed to exchange the condensation plate without interruption of the heating test (fig. 3). The activity of the plate is measured in a different laboratory with low level background [S]. In the case of large releases (> 1% of the sphere inventory), the 13’Cs and 134Cs loss can also be determined by comparing gammaspectrometric measurements of the sphere before and after the heating test. 2.2. Fuel samples Eleven spherical fuel elements and 8 fuel compacts containing modem TRISO-coated fuel have been heated. Coated particles and fuel bodies were manufactured at 2000 ,-

Fig. 2. Details

of temperature

profile tests.

18oooc

in standardized

heating

21

W. Schenk et al. / Fuel accident testing for HTRs

Inset-l new condensation plate

Furnace’with

spherical

fuel ele;ent

Fig. 3. Schematic diagram for replacement of cold finger in heating

furnace cold finger plate.

HOBEG in 1981 with less than 5 X 10e5 Sic layer defect fraction and less than 1 x 10F6 heavy metal contamination fraction. Pertinent data for the low-enriched UOz particles are given in table 1. Four spheres each were irradiated in the High Flux Reactor Petten in test HFR-K3 and four spheres in the Jiilich Dido reactor in FRJ2-K13. Beginning in July 1982, 24000 low-en-

6

i

4

6

8

IO

12

14

Bumup (% FlYA)

Fig. 4. Fast neutron fluence versus heavy spherical fuel elements (FE) and compacts 1800°C.

metal burnup of heated at 1600-

and subsequent

measurement

of deposited

activity

on

riched TRISO spheres went into the AVR core [9]. The cylindrical compacts were manufactured from spherical fuel elements with a fuel zone of only 20 mm diameter to be inserted into three HFR-P4 experiments and one Silo& experiment in Grenoble. The irradiation experiment HFR-P4/2 used particles with a 51 pm thick silicon carbide layer. During irradiation in the Material Test Reactors, the release of short-lived fission gases was measured continuously. The measurement of long-lived fission products in the irradiation capsule components outside the sphere, when compared to the total inventory, yields the fractional release from the fuel element. The conclusion from both the in-reactor fission gas release and the release of long-lived metallic fission products is that no irradiation induced particle failure occurred in all of the irradiation tests with modern TRISO fuel. As part of the postirradiation examinations, spheres and compacts were selected for the accident condition heating programme. The combination of fast neutron fluence/heavy metal bumup values of the selected samples shows that the irradiation conditions of the MODUL are covered well by the heating tests performed so far (fig. 4).

W. Schenk et al. / Fuel occident testing for HTRs

22 3. Results and discussion

The objective of the present series of accident simulation tests is to obtain fission product release data during heating tests at constant temperature. All results for modern UO* TRISO fuels in the temperature range 1600-1800°C are summarized in tables 2 and 3. A second series of tests will be performed later where the simulation of normal operating and accident conditions will follow more closely the scenarios projected for the MODUL reactor. 3.1. Fission product release and distribution 1800°C

at 1600 to

The gas release data from spheres during heating tests are shown in fig. 5. The measured isotope is *‘Kr which has been shown previously to give the same release as ‘33Xe and I31I [8]. As expected, release increases with heating temperature and duration. All 1600°C release results remain below the level of one particle failure (6 X 10P5 fraction for 16400 particles) as long as the burnup is below - 10% FIMA. A case of one particle failing with a complete breach of the coating can be observed in the 1700°C release curve shown in fig. 5a. With a 50% kernel retention of krypton at 17OO”C, the fractional release from the sphere is 3 x 10m5. The 2100°C release curves in fig. 5a have been obtained with the graphite furnace which does not have a cold finger installed [7]. Fuel compacts irradiated to burnup/fluence combinations significantly beyond the maximum MODUL conditions show higher gas release during the post

irradiation heating (fig. 5b), indicating failure or deterioration of the coating. With 1600 particles per compact, the level of one particle failure is 6 X 10K4. Fig. 5c demonstrates that gas release remains negligibly small for the first 50 h up to 1800 o C, but increases later when heating temperatures are 2 1700°C. The shape of the release curves can be explained by two phenomena: - Deterioration of the SIC layer results in permeability to fission products, but the remaining intact outer pyrocarbon layer delays the release of krypton. - On rare occasions a burst of gas release can be observed which is due to pressure induced complete coating failure. Several fission product isotopes are shown in figs. 6 and 9 indicating the release sequence Ag/Cs/Kr which is observed in all heating tests, even those extending to 2500°C [7]. ‘34Cs and 13’Cs exhibit an identical release behaviour, as expected. “Sr is always retained in spheres to a higher degree than “‘Cs. ‘romAg is an activated fission product isotope present in very low quantities due to the low fission yield. Therefore, it plays no role in safety considerations. It is, however, the only radionuclide which diffuses through intact silicon carbide during irradiation at 1200°C [lo]. There, silver release from the sphere can reach 1% of the sphere inventory. During heating at 1600 to 18OO”C, “OrnAg release values between 2 and 100% have been observed. Caesium is the main indicator of SIC deterioration or failure. The sphere release is further delayed by retention in matrix graphite. Strontium follows a similar pattern, but remains in the fuel sphere to a much higher

Table 2 Results of accident simulation tests with irradiated spherical fuel elements Fuel element

Irradiation bumup (W FIMA)

AVR 71/22 HFR-K3/1 FRJ-K13/2 AVR 82/20 AVR 82/9 AVR 74/11 FRJ-K13/4

3.5 7.5 8.0 8.6 8.9 6.2 7.6

HFR-K3/3 AVR 76/18 AVR 74,‘lO AVR 70/33

10.6 7.1 5.5 1.6

Heating test

Fractional release

Temperature

Duration

(“C)

(h)

1600 1600 1600 1600 1600 1700 1600 1800 1800 1800 1800 1800

500 500 160 100 500 185 138 100 100 200 90 175

*sKr

“‘cs

4x10-’ 2x1o-6 6x10-’ 2x10-’ 5x10-’ 3x10-s 3x10-’ 7x10-s 7x1o-4 1x1o-4 2x10v3 2x1o-3

2x1o-5 1 x 1o-4 4x10-5 6~10~~ 8~10-~ 8x10-s 3x1o-6 1x10-2 6~10-~ 5x10-2 1x10-’ 2x10-*

W. Schenk et al. / Fuel accident resting for HTRs

100

23

(al

10-I

Compact Fuel element --

(b)

10-Z

JI 10-Z i .

10”

i 8 2 10-S

10-6

10-7

10-a I/ 0

I 100

I

I I I 200 300 Heating time (h)

400

500

10-8 /; 0

100 200 Heating time (h) at 16CWC

300

10-3

8d t

10-l

B z

10-S

:

104

10-7 0

Fig. 5. Accumulated fractional release of “Kr fuel elements at 1600-1800 OC. (b) 1600°C

200 100 Heating time (h)

300

as a function of heating time at constant temperature. (a) Heating tests with spherical heating tests with compacts of 8-14% FIMA. (c) 1600-1800° C heating tests with compacts of 10-12X FIMA.

24

W. Schenk et al. / Fuel accident testing for HTRs

Table 3 Results of accident simulation tests at 1600-1800°C Fuel compact

HFR-P4/3/7 HFR-P4/1/8 HFR-P4/2/8 HFR-P4/1/12 SL-P1/6 SL-Pl/lO SL-P1/9 HFR-P4/3/12

with irradiated fuel compacts

Irradiation conditions

Heating test

Fractional release

Bumup (I%FIMA)

Fast fluence (lo*’ m-*)

Irr. temp

Temperature

Duration

(“C)

(“C)

(h)

13.9 13.8 13.8 11.1 10.7 10.3 10.7 12.0

7.5 7.2 7.2 5.5 6.7 6.0 6.3 5.5

1075 940 945 940 800 800 800 1075

1600 1600 1600 1600 1600 1700 1700 1800

304 304 304 304 304 304 304 279

degree than caesium. Krypton is always released later than caesium because of the additional holdup in the pyrocarbon layer. In particles without an outer pyrocarbon layer, krypton is released simultaneously with caesium [7]. Sphere 1 of the irradiation experiment HFR-K3 was irradiated hotter than most other fuels. Therefore, particles had already released some “OrnAg into the matrix graphite during irradiation. In the subsequent heating test at 16OO”C, silver in the matrix graphite is released

13’cs

1x1o-3 5x10-s 8~10-~ 5x10-7 7x10-7 9x1o-5 4x1o-5 1x1o-3

4x10-3 2x10-3 1x10-s 3x10-4 4x10-4 6x10-* 1x10-t 5x10-’

from the sphere indicating an in-reactor release level of 2% (fig. 6). Only during the later portion of the 500 h heating test can further increases in release be observed. Caesium, krypton and strontium releases remain extremely low during the first 200 h. This demonstrates that this fuel element had no initial coating defects, a low manufacturing contamination level, no in-pile coating failure and no particle failure during the initial phases of the heating test. The deconsolidation of the sphere, however, shows that by the end of the 500 h

HFR-K3/1

100

8sKr

7.5% fima 3.9. ld5 m-* (E> 0.1MeV) 1020-1200%

10-l AgllOm ,p-n-aa-a-a-a-_---~-~A-~~ 10-2

.?! t z

lo-:* Heating time at 16OOT

30

Fuel element radius (mm)

Fig. 6. Fission product release and distribution in sphere HFR-K3/1 after irradiation at lOOO-1200°C for 359 days and 1600°C heating.

W Schenk et al. / Fuel accident testing for HTRs

0

200

loo

400

300

500

Heating time (h) at 1600%

e

‘i

$

IO-4

I+-

-30

-20

I

after SO0 h at 16OOoC --t-I

-10

0

10

20

25

old contaminated fuel) is diffusing into the sphere matrix from the outside, and is then released during the heating test. Typically, this appears as a fractional release of 2 x 10m5 for low burnup spheres and 6 x 10v5 for high burnup spheres with their longer irradiation times (1250 full power days in the case of AVR 82/20). HFR-K3/1 is a much cleaner sphere with low initial release. After 500 h, caesium release and the profile through the sphere is higher owing to the burnup and fluence making the Sic permeable. Of particular interest is the fact that the average particle release is 1.2 x lop3 (table 4) while the sphere release is 1.1 x lop4 indicating an extremely high retention in the matrix. Figure 8 shows the 1600-1800°C caesium release curves from the fuel compacts irradiated to conditions beyond the design values in comparison to sphere results at 1600 o C, which remain below the level of release of one particle inventory. Release from compacts with modest irradiation conditions also stays at this low level.

30

Fuel element radius (mm) Fig. 7. duration: heating during caesium

Comparison of two the top diagram where AVR 71/22 irradiation in AVR; concentration in the

16OO’C heating tests of 500 h shows the release curve during is releasing caesium picked up the bottom diagram shows the fuel element outside the particles.

heating, certain amounts of Cs and Sr had been released from the particles, but had been strongly retained in the sphere. The ceramographic sections through a number of particles make a small degradation of the SIC layer visible (bottom photographs in fig. 11). The concentration of fission products in the matrix graphite (i.e. outside particles) along a cylinder through the centre of sphere HFR-K3/1 is shown to the right of fig. 6. With ?Sr being immobile in graphite, its radial profile represents the release from particles. ‘tomAg and t3’Cs show radial diffusion profiles through the sphere consistent with their release curves. The contrast in caesium behaviour between an MTR-irradiated and an AVR-irradiated sphere is shown in fig. 7. During irradiation in AVR for ca 480 full power days, caesium in the AVR primary circuit (from

0

100

200

300

Heating time (h) Fig. 8. Caesium release during heating of spherical ments (1600°C) and compacts (1600-1800°C).

fuel ele-

W. Schenk et al. / Fuel accident testing for HTRs

26 Table 4 Distribution

of caesium

in the components

of the fuel element

Heating temperature

Fuel element

CG3’content

after heating

tests

(%) in:

kernel

SIC

coating (PyC+ SIC)

particle

(“C) AVR71/22 HFR-K3/1 FJ2-K13/2

1600 1600 1600

81 29 58

_ a) 0.03 0.08

19 71 42

100 100 100

AVR 74/l 1 FRJ2-K13/4 AVR 76/18 HFR-K3/3

1700 1800 1800 1800

58 18 42 1.6

_ ?I) _ a) _ a) 0.13

42 57 44 54

100 75 86 56

sphere matrix 0.002 0.12 _ a) 0.03 23 14 8

‘) Not measured.

Sphere 3 of HFR-K3 was heated at 18OO”C, first for 25 h and later for 75 h. Fission product release curves and profiles through the sphere are shown in fig. 9. With 7 x 10e4 for krypton and 6 x 10e2 for caesium, the fractional release remains significantly below expected levels under these conditions. At 18OO”C, the influence of irradiation conditions is less pronounced than at 16OO“C. This is shown by the caesium release results listed in table 2 and caesium profiles shown in the right-hand diagram of fig. 9. All caesium release curves from the spheres are combined in fig. 10 as a function of heating times up to 500 h. The early part of the 1600°C curves is dominated by

AVR spheres which release caesium picked up in the AVR core. In fuel element AVR 74/11, the caesium inventory from 5 to 10 particles has been released into the matrix graphite during the 1700°C heating test to yield a final sphere release fraction of 8 X 10m5. Sphere FRJ2-K13/4 had first been heated for 138 h at 1600°C and subsequently for 100 h at 1800°C. Caesium release results are shown as if these were two independent heating tests. Earlier work [11,12] has shown that TRISO particles are more effective than BISO particles in retaining caesium at 1600°C. Due to experimental limitations at that time, the demonstrated improvement was of the

100

10-l

10-z

I E .E 10-a

I I z 10-3 ii s e 10"

E s 1o-4 ‘s

E IL 10-S

i

-

76ll8

200h

-- FRJ2-K13/4 100 h

'o-5- *..* HFR-K3/3 ___.__~

1OOh

1800°C 10-e

10-G 10-7

I

I

0

20

40

60

80

Heating time (h) at 1800°C Fig. 9. Fission

product

release and distribution

100

1o-7-*

30

Fud element radius (mm)

‘O-‘-M

30

Fuel element radius (mm)

in sphere HFR-K3/3 irradiated at 7OC-9OO’C and heated caesium profiles with other 1800’ C tests.

at 1800°C.

Comparison

of

W. Schenk et al. / Fuel accident testing for HTRs

0

100

200

H-W*:

Fig. 10. Caesium release from all heated spheres as a function of heating times up to 500 h.

order of one magnitude. The results presented here, when compared to the data from BISO particles, show an improvement by five orders of magnitude. Even at 1800” C, caesium retention is as good or better than previously assumed for 1600 o C. 3.2. Coated particle performance The HTR coated particles have been designed to withstand high bumups by providing the free volume for 100% internal gas release in the sacrificial buffer layer [13]. Therefore, the state of the fuel kernel is of minor concern. On the other hand, it is extremely important that the coating layer maintains its function as a pressure vessel, as an effective fission product barrier and as a protective layer around the fuel. One qualitative method to inspect the functionality of the particle coating is the visual examination in ceramographic sections. At the end of irradiation, changes in appearance are only observed in the fuel kernel, in the densification of the inner part of the buffer layer due to

27

fission fragments and in the optical activity of the PyC layer due to neutron bombardment. In particular, the SIC layer retains a dense, uniform appearance. No further changes can be observed after 160 h of heating at 1600°C (top of fig. 11). After the 500 h heating tests of AVR 71/22 and HFR-K3/1 some small SIC damage becomes apparent (middle and bottom photographs of fig. 11). It is much more pronounced in HFR-K3/1 with its higher irradiation conditions, but it also varies significantly from particle to particle as is visible from the three enlarged SIC sections. It is hypothesized that SIC degradation proceeds by a fission product enhanced change of SIC structure which affects particles on a statistical basis. At temperatures beyond 2000 o C, Sic thermal decomposition is observed to occur in unirradiated particles [7]. The rate of decomposition is accelerated by the effects of irradiation. Particles from sphere HFR-K3/3, heated at 1800 o C, have been mounted for microprobe examinations (fig. 12a). The observed 13’Cs loss in these particles ranges from 10 to 81%, underlining the stochastic nature of the release process [14]. Microprobe scans of 5 pm width through the coating layer at opposing ends of selected particles have been performed to show the distribution of fission product elements Ba, Ru, Ag, Pd, Cs and I. For one particle from HFR-K3/3 with 78% caesium loss, these radial profiles are shown in figs. 12b and 12d. Similar scans from a 1600°C heated particle from HFR-K3/1 with no observed caesium loss are shown in fig. 12~. Generally, no high fission product concentrations are observed in the Sic layer (microprobe measurements are not sensitive to very small concentrations as is shown for Cs in Sic in table 4). Palladium - suspected to be the essential agent for SIC corrosion [l&16] _ is piling up towards the PyC/SiC interface. This is consistent with the ceramographic observations which fail to show a systematic removal of SIC from the inside due to corrosion. Caesium and iodine are concentrated in the porous buffer layer. Since their radial concentration profiles are similar, this is interpreted as the formation of caesium iodide. A complete description of the fission product release mechanisms cannot be derived entirely from the microprobe observations. However, the available data taken as a whole imply fission product enhanced local structural changes of Sic enabling a grain boundary diffusion mechanism for caesium release. With SIC crystallites being nearly the same size as the layer thickness, random variations in release can be expected from

28

W. Schenk et al. / Fuel accident testing for HTRs

Fig. 11. Ceramographic sections through particles heated at 1600°C (complete particle followed by enlarged views from 3 different particles).

particle to particle. These variations are strongly apparent in the measurements of individual coated partitle release.

4. Conclusions The maximum accident temperatures in small HTRs are limited to 1600°C. Under these conditions, the heating data confirm that licensing applications can be

based on zero observed coated particle failures (out of 114 800 measured particles) and no additional release of safety-relevant fission products beyond normal operation. More heating tests are, however, required to extend the statistical data base for the prediction of the performance of approximately lo9 particles in the reactor core. During 500 h heating at 1600 o C, and 100 h tests at 1800 o C, progressive changes in the silicon carbide lead

29

W. Schenk et al. / Fuel accident testing for HTRs

10-l

]SiC

1 PyC Kernel

/ SIC 1

Pyc

10-Z E .P $ ,0-3 5 ? a” 10-4

(b

10-5

I

I

1

I

I

I

10 20 30 40 50 60 70 Measure points radial through coating 10-l

SiC Y

1

Pyc LTI

Kernel

~ Buffer layer

\

/ SIC FyC

FyC

Suffer layer

I Ln

LTI II

:) 0

10

20

30

40

50

Measure points radial through

60

70

80

coating

10-55

(d 0

10

20

30

40

50

Measure points radial through

60

70

1

coating

Fig. 12. Microprobe profiles of fission product elements through coatings of particles from HFR-K3. (a) Arrangement of sectioned particles (HFR-K3/3) for microprobe measurements.The numbers show the percentage of caesium loss from every single particle after heating at 1800 o C. (b) Ba, Ru, Ag profile in a particle with 78% Cs loss after 1800 o C test. (c) Cs, I, Pd profile in a particle from HFR-K3/1 after 1600 o C test. (d) Cs, I, Pd profile in a particle from HFR-K3/3.

to fission product release for fuel irradiated to bumup and fast neutron fluence levels beyond MODUL target design. While “OrnAgis the first radionuclide released during postirradiation heating of TRISO-coated fuel, it plays no role in safety considerations. It is released by a

diffusive transport mechanism through intact SIC. Silver release is followed by the simultaneous release of 13’Cs and ‘34Cs. Caesium release is initiated upon failure of the silicon carbide, likely by fission product enhanced local changes of SIC structure. After Sic failure, caesium release is controlled by a SIC grain boundary diffusion

30

W. Schenk et al. / Fuel accident testing

mechanism. Finally, “Kr and ?Sr are released; krypton being retained by its slow transport through the remaining intact pyrocarbon, and strontium being strongly bound in the oxide kernel and in the matrix graphite.

Acknowledgements Important in this work has been the active support of Prof. Nickel and of many colleagues in the KFA institute of reactor materials. Essential contributions have come from General Atomics, Oak Ridge National Laboratory, the Austrian Research Centre Seibersdorf and from the Harwell Laboratory. This work was carried out in the HBK project of the Entwicklungsgemeinschaft HTR supported by federal and state funds.

References [l] H. Reutler and G.H. Lohnert, Nucl. Eng. Des. 78 (1984) 129. [2] K. Krtiger, Experimentelle Simulation eines Ktihlmittelverlust-Stdrfalles mit dem AVR Reaktor, Report Jiil-2297 (1989). [3] M. Hrovat, H. Nickel and K. Koizlik, Uber die Entwicklung eines Matrixmaterials zur Herstellung gepresster Brennelemente fur Hochtemperaturreaktoren, Report Jill969 (1973). [4] W. Jahn, W. Rehm and K. Verfondern, Spezielle Analysen

for HTRc

zum Temperaturund Spaltproduktverhalten von HTRMODUL-Anlagen, Report Jiil-Spez-235 (1983). [51 D.T. Goodin, J. Am. Ceram. Sot. 65 (1982) 238. an bestrahlten KugelbrennPI W. Schenk, Storfallsimulation elementen bei Temperaturen von 1400 bis 2500° C, Report Jill-1883 (1983). W. Schenk, W. Heit, A.-W. Mehner and [71 H. Nabielek, D.T. Goodin, Nucl. Technol. 84 (1988) 62. PI W. Schenk, D. Pitzer and H. Nabielek, Fission Product Release Profiles from Spherical Fuel Elements at Accident Temperatures, Report Jtil-2234 (1988). G. Kaiser, H. Huschka, H. Ragoss, M. 191 H. Nabielek, Wimmers and W. Theymann, Nucl. Eng. Des. 78 (1984) 155. and E.H. Voice, DOI H. Nabielek, H. Hick, M. Wagner-LBffler Performance Limits of Coated Particle Fuel III: Fission Product Migration in HTR Fuel, Dragon Project Report 828, Part III (1974). - speziell Cs - in [111 H.-J. Allelein, Spaltproduktverhalten HTR TRISO Brennstoffteilchen, Report Jill-1695 (1980). and H. Nickel, J. Nucl. Mater. v21 W. Schenk, A. Naoumidis 124 (1984) 25. 1131 H. Nickel, Long-Term Testing of HTR Fuel Elements in the FRG, Report Jiil-Spez-383 (1986). Kugelbrennelemente mit 1141 W. Schenk and H. Nabielek, TRISO-Partikeln bei Storfalltemperaturen, Report JiilSpez-487 (1989). [15] O.M. Stansfield, F.J. Homan, W.A. Simon and R.F. Turner, Interaction of Fission Products and SIC in TRISO Fuel Particles: A Limiting HTGR Design Parameter, Report GA-Al7183 (1983). [16] R.J. Lauf, T.B. Lindemer and R.L. Pearson, J. Nucl. Mater. 120 (1984) 6.