Journal of Nuclear Materials North-Holland, Amsterdam
SHORT-LIVED PELLETS I.J.
115
135 (1985) 115-121
FISSION
PRODUCT
RELEASE
FROM
UO,: ANNULAR
VERSUS
SOLID
HASTINGS, C.E.L. HUNT, J.J. LIPSETT and R.D. DELANEY
Fuel Materials Received
Brunch. Chalk River Nuclear Laboratories.
20 July 1984: accepted
6 February
Chalk River. Onturio KOJ IJO, Canadu
1985
in a “dual sweep” test assembly, one element containing solid UOz pellets with four. 1 x 1 mm surface containing annular UO, pellets with a central hole two mm in diameter. Data are reported for a linear power range of 17-57 kW/m; thermocouples monitored operating temperatures. Hee28H z carrier gas swept short-lived fission products from both elements past independent spectrometers for identification and measurement. The releases from the solid and annular fuel were power-dependent from 17-57 kW/m, with the annular fuel release up to eight times greater than that for solid fuel. The annular fuel also showed a larger released iodine inventory in fuel and sheath areas with access to the carrier gas We have irradiated, slots and one element
1. Introduction The release of short-lived fission products from operating UO, pellets to the fuel-to-sheath gap is important in determining availability of the species for transport and release under accident conditions. Recent work has focussed on the “carrier-gas” technique to determine short-lived fission product behaviour under normal operating conditions to provide the fuel “source term” for accident studies. Results from isothermal samples have defined fundamental diffusion processes [I], and have been complemented by data from operating fuel elements [2-61. Release from fuel element tests involving solid UO, pellets indicates that R/B (atoms released/atoms born) is proportional to h-‘.’ [4,5], where X is the appropriate decay constant. AdditionaIIy. retease appeared only weakly dependent on linear power up to about 40.-50 kW/m [7]. An exception has been the CONTACT I tests [4] in which the fuel had a central hole. Release from CONTACT I at a linear power of 40 kW/m was generally about an order of magnitude higher than that from the Chalk River FIO-122 test with solid pellet fuel at 45 kW/m [S]. One inte~retation [5,7] assigned the difference to the fuel centre contributing most release in the CONTACT case, and the fuel periphery having the major contribution in FIO-122. The different temperatures of the fuel contributing to release thus give rise to different diffusion, and, consequently, release rates. The
measured central temperature for CONTACT I fuel was 1475’C 141; the measured peripheral temperature in fuel during experiments similar to FIO-122 was about 825°C [5]. An alternative interpretation of the data is that release from the solid fuel does originate from the high temperature central region, but since only a small part of the central region is intersected by cracks, its contribution to release through crack paths is limited. To assess directly the release differences, we have irradiated, in a “dual sweep” assembly, instrumented elements containing solid and annular UO, fuel pellets. Steady state data at a linear power of 57 kW/m are reported, as are stepped start-up data from 17--56 kW/m. Helium-2%H2 carrier gas swept short-lived fission products past independent on-line gamma spectrometers for identification and measurement. In addition to power (temperature) dependence of release, we determined the iodine inventory of the fuel and sheath internal surfaces to which the carrier gas had access. In solid fuel. this is primarily the fuel-to-sheath gap and fuel cracks with access to the gap; in annular fuel, the surface of the central hole.
2. E~~riment~l 2.1.
Fuel
characteristics and operating conditions
The experiment comprised a trefoil assembly, inserted in the X-4 loop of the NRX reactor at Chalk
Table I Fuel characteristics
/b
and
FUPI enrichment density grain size pellet diameter pellet length pellet configuration
fuel stack length
operating
conditions,
Sintered UO, pellets 5.9 wtR U-235 in U IO.64 M&m’ 10 pm 11.71 mm 14.0 mm (a) four. 1 mmX 1 mm surface slots (surface sweep) (h) one. 2 mm diameter central hole (central sweep) (c) filler element: half the pellets with slots, half with central holes 477 mm
Sheath
304 L stainless-steel
wall thickness
0.63 mm
(annealed)
CROSS SECTION OF 3rd PELLET
FUEL PERIPHERAL TMERMOCOUPLE
lJOt PELLETS CROSS SECTION
OF PELLETS
STST.
cl CROSS SECTION
OF 2
FROM TME BOT 0
Operating conditions coolant pressure coolant flow coolant inlet temperature linear power range (stepped start-ups) linear power range (steady operation) burnup
GAS OUTLET
“dual sweep”
SHEATH
FUEL PERIPHERAL TNRh4OCOWLES (21
0
cl
8.5 MPa 1 .O kg/s 260°C
kW/m
15-56
57-63 kW/m 50 MW h/kg U
Fig. 2. Schematic diagram of a typtcal instrumented. svvep~ element with solid pellets. Annular arrangement is identical, except pellets have 2 mm diameter central hole and no surface SIOIS.
River Nuclear Laboratories. Table 1 gives fuel details and operating conditions; steady operating power range was 57-63 kW/m. Steady state data at 57 kW/m are reported in this paper. Additionally, data were collected at intermediate linear powers between about 17 and 56 kW/m, during stepped reactor startups. Fig. 1 shows the solid (“gap-swept”) and annular (“centre-swept”) peliet designs. In the solid element, four, 1 X 1 mm
cross-section surface slots permitted easy carrier gas flow through the fuel-to-sheath gap. The solid. annular and filler elements had similar UO, weights. to ensure similar operating powers and simplify calorimetry. Fig 2 is a schematic diagram of a typical instrumented element. The solid and annular fuel elements each had two peripheral chromel/alumel thermocouples, situated in the second pellet from the carrier gas inlet end. about 1 mm from the fuel surface. Fxh element also had one peripheral thermocouple in the third pellet from the carrier gas outlet end. The filler element had fuel peripheral thermocouples at the inlet and outlet ends, and also a central thermocouple. 2.2. Sweep gas system operation
-
--
Fig. 1. Solid (a) and annular (b) fuel designs.
Fig. 3 shows a schematic diagram of the dual sweep arrangement. During operation, bottled He-Z%H, gas was introduced via separate Iines into the fuel elements containing solid and annular pellets, respectively. at an operating pressure of about 3 MPa, after passing through an oxygen remover and moisture trap. In the solid fuel
f. J. Hasrings
ei al. /
Short-lived fission product release
117
:TOR
FILTER
Fig. 3. Schematic flow diagram of “dual-sweep” system.
elements the carrier gas swept the fuel surface slots; in the annular fuel the carrier gas swept the central hole. Gaseous fission products were swept out of the fuel elements past two gamma spectrometers each independently situated on the gas lines from the solid or annular fuel elements, respectively. The volume flow rate was about 6 ml/s. After passing the spectrometers, the carrier gas and fission products passed through iodine traps into delay tanks before final release. 2.3. Data analysis Continuous on-line activity monitoring for fission products was carried out using gamma ray spectrometer equipment with high-purity Ge detectors. We were able to identify and measure species with half-lives as short as Xe-139 and Kr-90. Collimators, detectors and spectrometers were changed from one gas line to the other during test operation to confirm data. Fission product data detected by the spectrometers were recorded automatically on tape and transmitted for archiving and analyses. A program package calculated concentrations and release values of the detected fission products as a function of time. Final data are reported as R versus time or R/B versus h; fuel-to-sheath gap iodine inventories were derived from the R/B values. Full details of all techniques used are given elsewhere [S]. 3. Results and discussion 3.1. Effect
of power
and temperature
During a number of reactor startups, reactor power was stepped in stages to determine release dependence
on power. With one hour holds at each of a range of powers, there was insufficient time for the longer-lived isotopes to reach equilibrium. We therefore concentrated on Xe-138 and Kr-89 behaviour as a reasonable compromise between appropriate half-lives (15 min and 3 min, respectively), and independence of precursor effects. Fig. 4 shows R/B plotted versus linear power in the range 17-56 kW/m, for (a) Xe-138 and (b) Kr-89; data from a stepped startup after about 25 MW . h/kg U. Corresponding temperatures at each step are also shown in fig. 4a. Peripheral temperatures are measured, central temperatures are calculated using the ELESIM fuel performance code [9.10]. In both cases, the strongest R/B dependence on power begins at about 45 kW/m. Also, in both cases, R/B for annular fuel is from about three to eight times greater than that for solid fuel, with the larger differentials generally at higher powers. By interpolating release data for Xe-138 in fig. 4a, we obtain R/B = 3 X 10-O at 1475°C for the annular fuel, and R/B = 4 X IO-’ at 825°C for solid fuel. These R/B values correspond closely with those for CONTACT I [4] and FIO-122 [5]. respectively. 3.2. Steady begs power operation We obtained R and R/B data for steady operation at a linear power of 57 kW/m after a burnup of about 35 MW. h/kg U. At this power the measured peripheral temperatures in the annular and solid fuel were 99O’C and 960°C respectively. The corresponding calculated [9,10] central temperatures were 1700°C and 1750°C respectively. Fig. 5 shows R plotted against time for Xe-138 from solid and annular elements over
118 I
I
I
I
I
I
lo”r----I_
A
A
.
_.
-.___
_
ANNULAR
0
A 0
A
A
Cl 0
Xe-138
0
A ANNULAR 0
_.__L__._
SOLID
--_..__ _..
...- ..___~._._i. ____._.
IO
1?
MARCH I
I
I
I
20
10
30
40
1
I
50
60
Fig. 5. K versus time (date) for Xe-138 from annular and solid fuel during steady operation at a linear power of 57 kW/m. Data taken at a burnup of about 35 MW.h/kg L’.
LINEAR POWER (kW/m) Fig. 4. (a) R/B versus linear power for Xe-138, during stepped startup. ~orr~sp~ndin~ temperatures at each power step are given in table: Power
Temperature
(“C)
Annular
17 26 35 45 56
Centre
Periphery
Sofid ~I~ Centre
650 770 1020 1300 1630
625 680 790 850 950
670 790 1050 1360 1700
Peripheq 615 685 795 825 960
about 40 h, showing release enhancement for the annular fuel. Fig. 6 shows R/B plotted versus X ior the solid and annular elements. Also plotted are the best fit lines 15.71 to data from earlier tests. designated FIO-124 and FIG133, which operated at 55-60 kW;m with solid pellets. The agreement with data from the solid
__rII/rljl._
.-._.,“T ._...____-:._i-r ‘.‘7‘__‘,
A
A
It: iii ? cx
0 Kr-89 0 xl-5
A ANNULAR 0 SOtID
0
i
I
t
I
10
20
30
t0
LINEAR Fig. 4. (b) R/B
0
POWER
5
I
50
60
(kW/m)
versus linear powser for Kr-89.
j_ i_&___._
I
S5mK,
,
L.&.L.______.i.ii.i. 13Sx,
@K,
10-5 DECAY
_ ._.
135mxe 87K,‘3n
i___l_-il..
s9K, ‘YXt-
Xe
1O-L
10
-3
CONSTANT
is-’
t
L
$39 .7
30
Fig. 6. R/B versus X for solid and annuiar fuel during stead\ operation at a linear power of 57 kW/m. Best fits to data from similar tests [5.7] of solid fuel are also included.
1
119
J. Hastings et al. / Short -lived fission product release
fuel element in the current test is excellent. R/B for the annular fuel element is generally about eight times greater than that for the solid fuel element; in both cases there is an R/B dependence on X close to -0.5, confirming diffusion as the major mechanism of shortlived fission product release in both solid and annular fuel under steady power conditions. Note that our total test burnup is relatively low. about 50 MW. h/kg U. In a previous test [ll], we observed a burnup enhancement of release in solid pellet fuel of about a factor of two, at 220 MW h/kg U, related to fuel structural development. Significantly larger enhancements have been reported for solid fuel at higher burnups typical of light-water reactor (LWR) fuel [4].
Table 2 Iodine inventories Isotoue
of fuel areas with access to carrier
Test and inventory FIO-124 [5] (60 kW/m)
gas
in GBq (Ci)
FIO-141
(current)
(57 kW/m) Gap-swept
Centre-swept
140 (4) a 140 (4) a 28 (0.8)
680 (18.5) ’ 680 (18.5) ’ 136 (3.7)
I-131 I-133 I-1 35
100 (3) il 100 (3) 19 (0.5)
a Estimated kryptons.
from ratio assuming
behaviour
as for xenons
and
3.3. Iodine itwentory No iodines were directly measured at the spectrometers during any phase of the test, for solid or annular elements. This tendency of the iodine species to resist transport under normal operating conditions is as observed previously [5,7,12]. We have deduced the iodine behaviour in the fuel-to-sheath gap or central hole, by following the decay of iodine to xenon during continuous sweeping of both elements after reactor shutdown.
SHUTDOWN Xe-135m ANNULAR M)WWc x,
x/ x
I 23 FEE
Fig. 7. Release of Xe-135m as a function of time from sweeping during a reactor shutdown, for annular fuel. Time from shutdown (arrowed) to completion of data collection was eight hours.
The approach [5] gives the steady-state inventory of iodine on fuel or sheath surfaces with access to the carrier gas. Fig. 7 shows the release of Xe-135m as a function of time for the annular fuel during a shutdown. and is typical of results obtained. We also followed the behaviour of the solid fuel during the same shutdown. For both solid and annular fuel, the behaviour is characterized by a prompt decrease in release corresponding with the shutdown time, due to cessation of xenon release, followed by a decay period characteristic of the I-135. The measured I-135 inventories of areas with access to the carrier gas for the solid and annular cases during operation at 57 kW/m are 28 GBq (0.8 Ci) and 136 GBq (3.7 Ci), respectively. Steady-state R/B values for the gap- and centre-swept I-135 fit the xenon and krypton data in fig. 6 well. reflecting the similar behaviour of iodine, xenon, and kryptons under normal operating conditions reported earlier [5,7,12]. Table 2 shows estimates of the I-131 and I-133, as well as measured values for I-135 inventories for the accessible areas in the solid and annular fuel in the current tests, and compares them with those from a previous test [5] with a solid element under similar conditions. The data from the current test confirm the relative immobility of iodine under normal operating conditions, once it has deposited on internal fuel or sheath surfaces, probably as CsI. We measured negligible amounts of iodine on the piping downstream of both elements. The only time we have observed iodine moving in the system downstream of the fuel section was under defect conditions in a previous test [12] when coolant had access to the fuel. Data are confirming that the iodines give less cause for concern under accident conditions, as recently reviewed [13].
120
1.J. Hustrngs
4. Practical
er ul. / Short -hued fiwon produci relea.~r
implications
Considerable data are available to show that release of stable fission gases from annular UO, fuel with a large starting grain size is less than that for solid fuel. up to about 30 kW/m [14]. At higher powers. the differences are less marked. From the current test, release of short-lived fission products from annular UO, fuel is greater than that for solid pellet fuel over a wide range of linear powers encompassing both LWR and CANDU reactor conditions. However. we have not fully resolved the origin of release for the solid pellets. R/B for the annuhar fuel is about eight times that for solid fuel during steady operation at 57 kW/m. The difference is not explained by release resulting from diffusion characteristics of the central hole and solid pellet surface temperatures; this implies a release difference of several orders of magnitude. The key lies in assigning correct values for S/V (surface-to-volume) in the Booth formulation [l,lS]: R = -ST .__ B
_
v v’ A’
(11
in which D and h are appropriate diffusion coefficients and decay constants for the species studied. Qualitatively, the effective free surface area of the central hole in the annular fuel is greater than that for the cracks in solid fuel. We will confirm this by post-irradiation examination. The most plausible interpretation from our test and other data [15] showing dependence of release from solid fuel on central temperature, is that release from the solid fuel does in fact originate from the high temperature central region, but is limited by crack access to the fuel-to-sheath gap. We note the equilibrium inventory of iodine with access to the carrier gas is significantly higher for the annular fuel. ff one is producing a fuel source term based on data from solid pellets, the possible enhancement due to use of annular fuel must be taken into account, even assuming the beneficial effects of large grains. However, preliminary data from the current test suggest that transient release from annular fuel under reactor start-up and shutrdown conditions may only bc about one-tenth of that for solid fuel, offsetting the enhanced steady state release. These data will be detailed in another paper 1171.
5. Conclusions (1)
Release
ments
of short-lived fission products from containing solid and annular pellets
elewas
power dependent from 17--57 kW/m. with !hc at: nular fuel releasing up to eight times that for k&d fuel. (2) Release showed a -0.5 dependence of X. conhrming diffusion as the major mechanism of release in both solid and annular fuel. invento(3) Preliminary data suggest that equilibrium ries of iodine in areas accessible to the carrier gab are about a factor of five higher for annular fuel. compared with those for solid pellet fuel. (4) As previously observed. iodine diffused in the fuel similarly to the xenons and kryptons under normal operating conditions. but did not exit from thz element.
Acknowledgements
We acknowledge the support of L. Larson and ,4. English in the data analysis. There were significant contributions from P. Anderson, L.R. Bourque, P.J. Fehrenbach, R. Lavoie and D.H. Rose in the design. operation and fabrication of the test. We also acknowledge the excellent co-operation of NRX Reactor Branch.
References J.A. Turnhull. CA. Friskney, J.R. Findlay, F.A. Johnson and A.J. Walter, J. Nucl. Mater. 107 (1982) 168. C.J. Greatly and R. Hargreaves, 3. Nucl. Mater. 79 (1979) 235. A.D. Appelhans and J.A. Turnbull, Measured release of radioactive xenon, krypton and iodine from UO, during nuclear operation and a comparison with release models. Eighth Water Reactor Safety Research Information Meeting. Gaithersburg, MD, 1980. M. Bruet, J. Doddier, P. Melin and M.-L. Point& CONTACT 1 and 2 experiments: behaviour of PWR fuel rod up to 15000 MWd/TU, IAEA Specialists’ Meeting on Water Reactor Fuel Element Performance Computer Modelling, Blackpool, 1980. I.J. Hastings, C.E.L. Hunt, J.J. Lipsett and R.D. MacDonald, Behaviour of short-lived fission products within operating UO, fuel elements, IAEA Specialists’ Meeting on Water Reactor Fuel Element Performance Computer modelling, Preston, 1982; also, Res Mechanica 6 (1983) 167. M. Charles, J.-J. Abassin. D. Baron, M. Bruet and I’. Mehn, Utilization of CONTACT experiments to improve the fission gas release knowledge in PWR fuel rods, IAEA Specialists’ Meeting on Water Reactor Fuel Element Performance Computer Modelling. Preston. 1982.
I.J. Hastings et ul. / Short -lived fission product release [7] 1.J. Hastings, C.E.L. Hunt, J.J. Lipsett and R.G. Gray, Transient fission product release during dryout in UO, fuel, Atomic Energy of Canada Limited. report AECL-7832 (1982). [8] I.J. Hastings, C.E.L. Hunt, J.J. Lipsett and R.D. MacDonald, Tests to determine the release of short-lived fission products from UO, fuel operating at linear powers of 45 and 60 kW/m: methods and results, Atomic Energy of Canada Limited, report AECL-7920 (1985). [9] M.J.F. Notley. Nucl. Technol. 44 (1979) 445. [lo] M.J.F. Notley and I.J. Hastings, Nucl. Engrg. Des. 56 (1980) 163. [ll] C.E.L. Hunt and F. Iglesias, Chalk River Nuclear Laboratories, unpublished results. (121 J.J. Lipsett. C.E.L. Hunt and I.J. Hastings, Behaviour of short-lived iodines in operating UO, fuel elements. Atomic Energy of Canada Limited, report AECL-7721 (1985). [13] S. Rippon. Accident source terms - cause for less concern, Atom 327/2 (1984).
121
[14] L.F.A. Raven, Gas retentive annular fuel pellets - palliative or panacea for the LWR?, Annual Meeting Am. Cer. Sot., Cincinnati, OH, 1979, BNFL (UK) report BNFL 852 (S) (1979). [15] A.H. Booth, A method of calculating fission gas diffusion from UO, fuel, Atomic Energy of Canada Limited, report AECL-496 (1957). [16] I.J. Hastings, C.E.L. Hunt and J.J. Lipsett, Release of short-lived fission products from UO, fuel: effects of operating conditions, Proc. Internat. Symp. on Thermodynamics of Nuclear Materials, Hamilton, Ontario. 1984, J. Nucl. Mater. 130 (1985) 407. [17] I.J. Hastings, J.J. Lipsett. C.E.L. Hunt and R.D. Delaney, Transient release of short-lived fission products: solid and annular UO, fuel, Proc. BNES Meeting on Fuel Performance, Stratford-upon-Avon, 1985. 1 (1985) 313.