Helium Release from 238PuO2 Fuel Particles Mohamed S. El-Genk and Jean-Michel Tournier Institute for Space and Nuclear Power Studies and Chemical and Nuclear Engineering Dept. The University of New Mexico, Albuquerque, NM 87131, USA (505) 277–5442, email:
[email protected] Abstract. Coated plutonia fuel particles have recently been proposed for potential use in future space exploration missions that employ radioisotope power systems and/or radioisotope heater units (RHUs). The design of this fuel form calls for full retention of the helium generated by the natural radioactive decay of 238Pu, with the aid of a strong zirconium carbide coating. This paper reviews the potential release mechanisms of helium in small-grain (7-40 µm) plutonia pellets currently being used in the General Purpose Heat Source (GPHS) modules and RHUs, during both steady-state and transient heating conditions. The applicability of these mechanisms to large-grain and polycrystalline 238PuO2 fuel kernels is examined and estimates of helium release during a re-entry heating pulse up to 1723 K are presented. These estimates are based on the reported data for fission gas release from granular and monocrystal UO2 fuel particles irradiated at isothermal conditions up to 6.4 at.% burnup and 2030 K. It is concluded that the helium release fraction from large-grain (> 300 µm) plutonia fuel kernels heated up to 1723 K could be less than 7%, compared to ~ 80% from small-grain (7-40 µm) fuel. The helium release fraction from polycrystalline plutonia kernels fabricated using Sol-Gel techniques could be even lower. Sol-Gel fabrication processes are favored over powder metallurgy, because of their high precision and excellent reproducibility and the absence of a radioactive dust waste stream, significantly reducing the fabrication and post-fabrication clean-up costs.
INTRODUCTION The early technology of coated fuel particles came as a spin off from the ROVER and NERVA Nuclear Rocket Programs carried out in the sixties and early seventies at Los Alamos National Laboratory, Atomic International, and Westinghouse Electric Company. The coated plutonia particles design also draws on a vast and proven fabrication technology of UO2 and (U,Pu)O2 fuels for High-Temperature Gas-Cooled Reactors (HTGRs) (Minato et al., 1997). The attractive features of coated particles fuel in fission reactors have been their ability to operate at high temperatures (up to 1700 K in HTGRs and 3000 K in nuclear thermal propulsion reactors) and achieve high fuel burnup (> 30 at.%), with little concern about fuel swelling and fission products release. The primary coating of the fuel particles (SiC at < 1300 K or ZrC for operation at higher temperatures) serves as a pressure vessel for containing fission products and constraining fuel swelling during irradiation. The fuel kernel could be fabricated using Sol-Gel techniques (Haas et al., 1967; Huschka and Vygen, 1977; Förthmann and Blass, 1977; Lahr, 1977) or powder metallurgy processes (Burnett et al., 1963; Ford and Shennan, 1972; Allen et al., 1977). Unlike powder metallurgy, Sol-Gel techniques do not involve any milling and grinding and hence, do not generate a postfabrication radioactive solid waste stream. In addition, all used chemicals are recycled and reused. The literature on the fabrication, gas release, and material properties of coated particles fuel has recently been reviewed and documented (El-Genk and Tournier, 1999). Coated plutonia (238PuO2) fuel particles are being investigated for potential use in Radioisotope Heater Units (RHUs) and Radioisotope Power Systems (RPSs) (Scholtis et al., 1999; Tournier and El-Genk, 2000). The proposed fuel particles consist of a plutonia kernel (~ 300–1200 µm in diameter) coated with a thin layer (5 µm) of Pyrolytic Graphite (PyC) and a strong outer coating of zirconium carbide (see Figure 1). The thickness of the ZrC coating depends on the fuel kernel size and internal pressure of the released helium gas (Tournier and El-Genk, 2000). The amount and the rate of helium release depend on the fuel temperature and microstructure (granular or polycrystalline), and the helium inventory in the fuel matrix (or storage time).
The 238PuO2 coated particles fuel offers promise for enhanced safety and unique design flexibility, as the heat Pyrolytic sources can be fabricated in a variety of shapes, sizes, Carbon (PyC) 238PuO and power densities. The coated fuel particles could be 2 5 µm thick dispersed in a solid graphite matrix made into different 300-1200 µm shapes or used in the form of a heating tape or a paint, for a variety of applications. The coated particles fuel ZrC coating graphite-compact (CPFC) in a pellet form can potentially (variable be used to develop high specific power RHUs using the thickness > 10 µm) current Fine Weave Pierced Fabric (FWPF) aeroshell structure. When the fuel pellet, Pt-alloy cladding, and inner graphite sleeve in the Light Weight RHU FIGURE 1. Coated plutonia fuel particle. (LWRHU) were replaced with CPFC, the thermal power output increased from ~ 1.0 Wth to 1.54, 2.4 and 2.52 Wth, for 100%, 10%and 5% helium release from the fuel kernels, respectively (Tournier and El-Genk, 2000). Such higher thermal powers are possible at essentially the same LWRHU mass (~ 40 g), and conservatively assuming a pre-launch storage time of 10 years and a fuel temperature of 1723 K. This temperature is the predicted maximum during an accidental re-entry heating pulse (Schock, 1981; Tournier and El-Genk, 2000). The higher the release fraction of He, the thicker is the ZrC coating needed to accommodate the internal gas pressure in the coated fuel particles, reducing the loading of the plutonia fuel in the CPFC. Therefore, there is a need to investigate He gas release in coated plutonia fuel kernels as a function of temperature and storage time before launch in order to maximize the fuel loading in the CPCF. In addition to the storage time and temperature, the He release depends on the grain size and the fuel microstructure. The objectives of this paper are to: (a) review the release mechanisms of helium in small grain (7-40 µm) plutonia pellets currently being used in the General Purpose Heat Source (GPHS) modules and RHUs; (b) examine the applicability of these mechanisms to the helium release from large-grain (> 300 µm) and polycrystalline 238PuO2 fuel kernels; (c) provide best estimates of the He release fraction from coated plutonia particles as a function of temperature, up to 1723 K. These estimates are based on the reported fission gas release data for granular and monocrystal, spherical UO2 fuel particles irradiated at isothermal conditions up to 6.4 at. % burnup (Turnbull, 1974; Friskney et al., 1977; Turnbull et al., 1977; Friskney and Turnbull, 1979).
BASIC MICROSTRUCTURES OF FUEL MICROSPHERES There are two basic microstructures for the fuel microspheres or kernels (Figure 2): (a) the granular microstructure obtained by powder metallurgy processes; and (b) the polycrystalline microstructure obtained by the Sol-Gel processes. A granular fuel kernel consists of a number of polycrystalline grains of almost the same size (typically 7–40 µm). These grains are separated by common grain boundaries that develop during sintering at high pressure and temperature. This process also controls the as-fabricated porosity in the fuel grains and at the triple interface of the grains to accommodate fission gases during operation in nuclear reactors. Typical values of the as-fabricated fuel kernel porosity range between 5 and 15% (see Figure 2a). Granular fuel kernels are fabricated using the binderless agglomeration process (Burnett et al., 1963; Ford and Shennan, 1972; Allen et al., 1977). This process could be used to fabricate highly spherical and virtually monosized oxide fuel kernels in the range from 200 to 1000 µm in diameter, with a porosity between 5 and 20% (see Figure 2a). Green spheroids are initially prepared from ceramic-grade, plutonium dioxide powder, having a specific surface area of 2-4 m2/g. The fuel powder is mixed (using ball milling) with proprietary carbon blacks, having a specific surface area of 15-30 m2/g. The resulting powder is granulated to form seeds which are then grown to the desired kernel size in a vibrating pan fed with fresh fuel powder. The fuel green spheres, or “seeds” are formed by the simple action of gyrating the fine submicron fuel powder into a bowl. These seeds, which have a size distribution in the range between 150 and 200 µm in diameter, are spherical in shape. Fine fuel powder is added at a controlled rate to the charge of seeds while being continuously gyrated into a bowl. The PuO2/C “green” particles are then spheroidized in a rotary sieve or planetary mill. Porous PuO2 kernels are fabricated from fuel green microspheres using a two-stage heat treatment process to control the as-fabricated porosity and the stoichiometry of the oxide fuel. First, the green fuel microspheres are
Polycrystalline fuel grain with intragranular voids
As-fabricated intergranular voids
(a) Granular fuel kernel fabricated Using powder-metallurgy
As-fabricated voids
Homogeneous crystalline fuel matrix
(b) Polycrystalline fuel kernel fabricated using Sol-Gel techniques
FIGURE 2. Basic fuel microstructures for coated fuel particles.
consolidated by sintering in carbon monoxide at 1923 K to prevent carbide formation. The sintered particles are then decarbonized in a flowing CO/CO2 at 1573 K in a continuous furnace, where the removal of the carbon leaves the desired porosity within the fuel microspheres (Figure 2a) (Burnett et al., 1963; Ford and Shennan, 1972; Allen et al., 1977). Polycrystalline fuel kernels with diameters as large as 1200 µm can be fabricated using Sol-Gel techniques (Haas et al., 1967; Huschka and Vygen, 1977; Förthmann and Blass, 1977; Lahr, 1977). These techniques do not require milling or grinding (as during the fabrication of granular fuel by powder metallurgy), thus do not generate any radioactive dust or aerosols. These all wet chemical processes use solutions or sols (or suspensions) of fissile or fertile materials dispersed into uniform liquid (gel like) droplets. The spherical shape of these droplets is fixed by the gelation (by water or ammonia extraction) achieved by either precipitation or dehydration reactions. The gelled fuel microspheres with an almost perfect sphericity are then washed, dried, and fired at high temperature to remove water and volatile additives and sinter the spheres to the desired density and fuel stoichiometry. The sintered fuel kernels have a polycrystalline fuel structure with tiny, intragranular voids (< 1µm in diameter) (see Figure 2b). The amount of inter-granular and intra-granular porosity in the fuel kernels (5-20%) is controlled during the sintering process. The former tends to be very tiny, of sub-micron size, while the latter could be a few microns in diameter. Thus, for helium release, the polycrystalline fuel particles (Figure 2b) may be regarded as a single grain of the kernel diameter, with as-fabricated intragranular porosity made of tiny pores. The release mechanisms of fission gases in granular, mixed-oxide and plutonia fuel are summarized next. In addition, the similarities and differences pertaining to helium release in plutonia fuel pellets and microspheres are indicated and discussed.
MECHANISMS OF GAS RELEASE IN (U,Pu)O2 AND PuO2 FUELS There are strong similarities between the release mechanisms of fission gases and volatile fission products in granular oxide fuel and those of helium in granular 238PuO2. The major difference is the restructuring and cracking of oxide fuel occurring during irradiation in fission reactors. In fission reactors, the power density in the fuel pellets is orders of magnitude higher than that by natural radioactive decay of 238Pu in plutonia pellets. When combined with the inherently low thermal conductivity of oxide and mixed-oxide fuels, fuel pellets in fission reactors develop a steep, parabolic radial temperature distribution and significantly higher fuel temperatures, particularly at the pellet centerline. The combination of high temperature and radial temperature gradient causes significant restructuring of the oxide fuel pellets in fission reactors (Olander, 1976). However, owing to the similar ceramic nature of both oxide and plutonia fuels, for the same operating temperature and as-fabricated porosity and grain size, the gas release mechanisms in both fuel types are identical.
The processes by which gas bubbles nucleate within the fuel grains, grow, diffuse, and eventually coalesce at the grain boundaries are extremely complex (Olander, 1976). The fuel temperature, however, appears to provide a useful index for determining the release mechanisms and rates of fission gases in oxide and mixed-oxide fuels in fission reactors and of helium in plutonia fuel in RHUs. In summary, the prevailing release mechanisms of the gases and volatile products generated in ceramic fuel by either fission in nuclear reactors or radioactive decay of 238PuO2 (Lyons et al., 1972; Mueller et al., 1974; Olander, 1976; Peterson et al., 1984; Scaffidi-Argentina et al., 1997) can be divided into four sequential mechanisms: (a) Atomic diffusion near the surface of the fuel pellet or particles when the fuel temperature is < 900 K. The release fraction in this case is typically very low since it is limited by the available area for release at the fuel surface and the relatively low diffusion coefficient in the fuel matrix. At such low fuel temperatures, most of the generated gases are accumulated within the fuel grains. (b) Accumulation in and diffusion to newly forming intergranular gas bubbles, due to the enhanced mobility of tiny intragranular gas bubbles and the growth of fuel grains at temperatures ~ 1150 K or higher. At such temperatures, some open porosity may form at the grain boundaries, resulting in a higher gas release. The gas release fraction, however, remains well below 10%, due to the limited surface area for release. (c) Separation and formation of open porosity at grain boundaries, which occur at fuel temperatures of 1150-1500 K. At these high temperatures, most accumulated gases in the grain boundary bubbles are released through open porosity at the grain boundaries. (d) Atomic and volume diffusion in fuel grains, to the open porosity at the grain boundaries. This mechanism, which is dominant at fuel temperatures in excess of 1600-1700 K, solely depends on the fuel temperature, since gas release is not limited by release surface area. At these temperatures, the open porosity at the grain boundaries are fully accessible for gas release from the fuel grains. These release mechanisms have been confirmed by experimental data on helium gas release from GPHS and RHU plutonia fuel pellets as a function of temperature, storage time, and mode of heating (steady-state and transient reentry heating) (Angelini et al., 1970; Mulford and Mueller, 1973; Mueller et al., 1974; Land, 1980; Peterson and Starzynski, 1982; Peterson et al., 1984). Results described later in this paper also indicate the strong dependence of the release fraction of He on the as-fabricated grain size of the fuel. Results of the helium gas release experiments from granular plutonia pellets (7-40 µm grain size) have indicated that transient heating increases the helium release rate, but the release fraction of helium depends on the temperature reached during the transient (Peterson and Starzynski, 1982; Peterson et al., 1984). Similar release fractions were reported at the same temperatures in steady-state heating experiments of granular plutonia pellets and smaller samples (Mulford and Mueller, 1973; Mueller et al., 1974). Therefore, it is evident that increasing the fuel grain size and/or the elimination of grain boundaries could significantly reduce the release fraction of He gas in 238PuO2 fuel. The grain boundaries are absent in polycrystalline fuel particles fabricated using Sol-Gel techniques (Figure 2b). Therefore, in such polycrystalline plutonia fuel particles, helium gas release is expected to be much lower than in granular fuel particles of the same diameter. The only contributing mechanisms for He gas release in polycrystalline plutonia particles will be atomic and volume diffusion, which are limited by the effective mass diffusion coefficient of the gas in the fuel matrix and the availability of surface area for gas release. The absence of grain boundaries in polycrystalline fuel particles limits the gas release to the geometrical surface area of the fuel kernel, even at temperatures in excess of 1600 K. The next section presents a summary of the equivalent-sphere diffusion model for the release of fission gases and volatile fission products and discusses the effects of the grain size and the decay constant (or half-life) on the release fraction of various radioactive species. The applicability of the results to predicting the release fraction of noble, non-radioactive gases, such as helium in plutonia fuel, is also discussed.
EQUIVALENT-SPHERE DIFFUSION MODEL FOR FISSION GAS RELEASE The equivalent-sphere model, originally proposed by Booth and Rymer (Olander, 1976), has been used to describe fission gas and helium release from granular, oxide fuel in fission reactors and granular plutonia pellets in GPHS and RHU units (Angelini et al., 1970; Mulford and Mueller, 1973; Mueller et al., 1974; Land, 1980; Peterson and Starzynski, 1982; Peterson et al., 1984). The model assumes that gases diffuse to the free surface of so-called
equivalent spheres, from which it is released. The equivalent-sphere model has successfully been used to correlate gas-release experiments. The model treats granular fuel as a collection of equivalent spheres of uniform size, which have the same surface-to-volume ratio as the average release unit in the fuel matrix. Thus, the diameter of the equivalent sphere for gas release is d = 2a = 6V/SR, where a in the sphere radius and V and SR are the equivalent sphere’s volume and surface area, respectively. The total fuel surface area for the gas release, however, can be measured using gas-adsorption techniques, thus the effective value of d (or a) is experimentally determined. When the average grain size of as-fabricated fuel is known, it may be substituted as an approximate value for d. The actual value of d, however, could be different, particularly as intragranular fuel cracking occurs, typically at temperatures < 1400 K. Once the effective geometry of the fuel specimen for gas release has been characterized, the results of gas-release experiments can be used to determine the effective mass diffusion coefficient of the gases in the fuel matrix, D. In order to account for the effect of radioactive decay, the following diffusion equation: ∂C & D ∂ 2 ∂C = B+ 2 r − λC ∂t ∂t r ∂r
(1)
is used to determine the release fraction, F, assuming an initial gas concentration C(a,0) = 0. In this equation, B& is the production or birth rate of the gaseous isotope of interest and λ is its radioactive decay constant. The approximate solution of Equation (1) for stable gases (i.e. when the third term on the right hand side of Equation (1) is zero) gives the release fraction, F, as (Olander, 1976): F ≈6 τ π ,
(2)
t
where, τ = ∫ D ′dt is a dimensionless time, and o
D ′(T ) = D(T ) a 2 (T ) ∝ (S R V )2 × D .
(3) -1
In this Equation, D', the effective diffusion coefficient (s ), depends on the release surface-to-volume ratio of the fuel and the mass diffusion coefficient of the gaseous species, D (m2/s), which are both functions of the fuel temperature, T. The release fraction, F, can be expressed in terms of the effective radius of the average release unit, a, fuel temperature, T, and release time, t, as: S F ∝ R × Dt V
∝
1 Tγ t . a(T )
(4)
As this Equation indicates, F is inversely proportional to the effective radius of the average release unit in the fuel, and increases with the square root of time. Note that the fractional gas release, F, in Equation (3) is not the releaseto-birth rate ratio, R& / B& . For radioactive fission gas species that have attained a steady-state concentration in the fuel, the fractional release, F* equals in this case the release-to-birth rate ratio, and is given by (Olander, 1976): S F * ≡ R& / B& ≈ R × V
D
λ
∝
D ′(T )
λ
∝
1 T γ T1 / 2 . a (T )
(5)
This equation shows F* to be inversely proportional to the average fuel grain size, a, and to increase with the square root of the half-life (or inversely proportional to the square root of the radioactive decay constant). As the half-life increases, F* approaches the release fraction of non-radioactive species, given by Equation (2). Thus, for the same fuel material, temperature, and grain size, F* increases exponentially with the half-life of the gas species, approaching an asymptote. This conclusion is critical to the applicability of the release data of fission gases to the prediction of helium release in plutonia fuel particles, as detailed later. The effective diffusion coefficient, D', increases exponentially with temperature and accounts for the increases in both the mass diffusion coefficient, D, and the effective release area, SR, (including fuel cracks and open grain boundaries) with temperature. Neglecting the as-fabricated open porosity, the surface-to-volume ratio of plutonia
The normalized surface area for helium gas release from plutonia fuel particles can thus be written as: SR Sp
d = (1 − α ) + α p , dg
(8)
where Sp is the geometrical surface area of the asfabricated fuel particle.
Normalized Surface Area for Release (SR/Sp)
fuel microspheres for gas release can be expressed as: 6 6 SR +α , (6) = (1 − α ) dp dg V where, T < 900 K, α = 0, for 0 < α < 1, for 900 K < T < 1450 K, (7) α = 1, for T > 1450 K.
100
granular fuel: grain size = 10 µm c (~ 1450 K) d
Particle size: dp = 1000 µm
10
incipient separation of grain boundary (~ 1150 K)
granular fuel: grain size = 200 µm c
incipient formation of open pores (~ 900 K)
b
1
b
polycrystalline fuel E
D
volume atomic diffusion diffusion
a
C
B
separation of grain boundary
increase in open porosity
A diffusion from fuel matrix
The effect of fuel temperature on the effective area for gas release in granular and polycrystalline fuels, SR, is 5 6 7 8 9 10 11 12 illustrated in Figure 3. At temperatures below ~ 900 K, 4 -1 10 / T (K ) gas release occurs by atomic diffusion from the fuel FIGURE 3. Effective release area for gases in granular matrix, and SR is equal to the geometrical surface area of and polycrystalline fuels. the as-fabricated fuel sample (i.e. α = 0). Between 900 and 1150 K, 0 < α < 1 and SR increases with temperature in granular fuel due to the formation and the coalescence of grain boundary bubbles. Above ~1150 K, SR increases rapidly with temperature as α approaches unity, due to the formation of open porosity caused by the separation of the grain boundaries. At ~1450 K, SR reaches its maximum value and α = 1.0, as the separation at the grain boundaries is complete. Above this temperature, the release fraction in granular fuel is no longer limited by the surface area available for release, but rather by the atomic and volume diffusion of the gases in the fuel grains. It should be noted that the separation of the grain boundaries does not necessarily cause powdering of the fuel or breakup of its structure, but rather a full release of the gas at the grain boundaries through the formation of open porosity or tunnels. Figure 3 illustrates the effect of the fuel grain size on SR and potentially on the helium gas release in plutonia pellets or particles. For the same fuel temperature and particle geometry, increasing the grain size from a typical 10 µm to 200 µm could reduce the release fraction at high fuel temperatures (>1450 K) by more than an order of magnitude. As indicated earlier, a polycrystalline fuel kernel could be regarded as a single grain with a size equal to the kernel diameter. Thus, the effective surface area for gas release in polycrystalline particles would remain essentially unchanged and equal to the geometrical surface area of the particle. Therefore, the helium gas release in polycrystalline PuO2 fuel would be expected to be significantly lower than in granular fuel pellets or microspheres. To illustrate this point, consider a plutonia fuel microsphere, 1-mm in diameter, at high temperatures > 1450 K. A granular fuel made of 200 µm grains has a surface area for release that is 5 times that of a polycrystalline particle, while a 10 µ-grain fuel particle would have 100 times the surface area for helium release of a polycrystalline particle of the same diameter (Figure 3). These estimates are based on the experimental data showing complete release of the gas at the grain boundaries at fuel temperatures > ~ 1450 K (Angelini et al., 1970; Mulford and Mueller, 1973; Mueller et al., 1974; Land, 1980; Peterson and Starzynski, 1982; Peterson et al., 1984). Most gas release data reported in the literature for coated fuel particles were obtained for oxide fuels irradiated in fission reactors. In order to apply the reported data to the helium gas release in plutonia fuel kernels, fission gas release measurements must be made at isothermal conditions, which is the case in plutonia fuel particles. Fortunately, such measurements were obtained at the Berkeley Nuclear Laboratories (U.K.). The release fractions of radioactive fission gases (Xe and Kr isotopes) and volatile fission products (Cs, I and Te isotopes) were measured for small-grain, large-grain and monocrystal UO2 particles. The applicability of the reported data to the He release in spherical plutonia particles operating at the same temperatures and having same microstructure is discussed next.
ISOTHERMAL RELEASE IN GRANULAR AND MONOCRYSTAL FUEL PARTICLES The only detailed studies of the isothermal release of fission gases and volatile fission products from granular and monocrystal fuel particles (Figure 4) were those conducted at Harwell and at Berkeley Nuclear Laboratories in the U.K. (Turnbull, 1974; Friskney et al., 1977; Turnbull et al., 1977; Friskney and Turnbull, 1979). These studies investigated the effects of grain size, radioactive decay, fuel burnup (up to 6.4 at.%) and temperature (up to 2023 K) on the release-to-birth rate ratio of the various gaseous and volatile isotopes. The temperature of the fuel particles during irradiation was kept nearly uniform, as would be expected during actual operation in fission reactors at low power density and in plutonia fuel particles in RHUs. UO2 fuel microspheres consisting of small grains (average grain size of 10 µm) (Figure 4a) and large grains (an effective grain size between 300 and 600 µm) (Figure 4b) were irradiated up to a 6.4 at.% burnup. Monocrystal right cylinders of natural (0.72 wt.% 235U) stoichiometric UO2 were also irradiated (Figure 4c). In addition, granular fuel specimens of 2.0% enriched UO2 of near theoretical density, in the form of small cylinders, 10 mm long and 3 mm in diameter, having 7-µm and 40-µm grain sizes were irradiated (Turnbull, 1974).
10-µm grains r = 1.46% 235U B = 0.4-6.4 at%
~ 300-600 µm grains r = 1.46% 235U B = 4.0-6.4 at%
Monocrystal r = 0.72% 235U B = 1.6 at%
1.2 mm
1.2 mm
2.5 mm
(a) Small-grain spherical particle
(b) Large-grain spherical particle
(c) Monocrystal right cylindrical particle
2.5 mm
During irradiation, the individual fuel particles were wrapped in tungsten mesh to prevent them from touching each other or the walls of the molybdenum container. The particles were irradiated isothermally in an electrically heated rig in the UKAEA reactor DIDO. Fission heating (thermal neutron flux of ~ 2.6 x 1017 m2/s) produced a temperature drop between centerline and outer surface of the specimens of less than 100 K. The fuel particles were continuously swept with a He–2%H2 gas mixture to carry the released fission gases and volatile fission products. The volatile fission products were deposited using a cold finger while fission gases were collected using charcoal traps cooled with liquid nitrogen. The amounts of the various gaseous species released were measured by γspectroscopy, after correcting for their radioactive decay. The release rates of fission gases and volatile fission products were calculated based on these measurements, while their birth rates were calculated using computer codes (Friskney and Turnbull, 1979).
FIGURE 4. UO2 fuel samples used in isothermal fission gas release studies at Harwell and Berkeley Nuclear Laboratories (Turnbull, 1974; Friskney et al., 1977; Turnbull et al., 1977; Friskney and Turnbull, 1979).
Measured Release Data Figure 5 presents the measured release-to-birth rate ratios of 133Xe, which has a half-life of 5.2 days, for the different fuel particles, as a function of irradiation temperature. The small-grain (10 µm) UO2 microspheres released about 20 times more gas that the large-grain (~300–600 µm) ones. This ratio is comparable to that of the grain size ratio, consistent with Equation (5), which predicts the release-to-birth rate ratio to be inversely proportional to the grain size. As expected, the monocrystal, natural uranium particles released less gas than the large-grain microspheres, due to the absence of grain boundaries in the former. However, the difference was not as large as might be expected, because the rate of fission was about half that in the large-grain fuel particles, due to the lower 235U enrichment. Figures 5 and 6 show a change in the fission gas release-to-birth rate ratio in granular fuel particles at ~ 1350–1450 K, indicating a change in the release mechanism. Below ~ 1100 K, fission gases and volatile fission products
235
r = U enrichment B = fuel burnup
10
10 µm grains (r = 1.46%, B = 6.4 at%)
1 300-600 µm grains
0.1
0.01 1000
133
Xe
2500-µm monocrystal (r = 0.72%, B = 1.6 at%)
1200
1400
1600
Temperature, T (K)
FIGURE 5. Effect of grain size and temperature on isothermal release of 133Xe (T1/2 = 5.24 days).
1800
Release-to-Birth Rate Ratio (%)
Release-to-Birth Rate Ratio (%)
100
10
133
Xe (5.24 days) Kr (4.48 hrs) Kr (2.83 hrs) 138 Xe (14.1 mins) 85m 88
1
300-600 µm grains (B = 6.4 at%)
0.1
0.01
0.001 1000
1200
1400
1600
1800
Temperature, T (K)
FIGURE 6. Effect of half-life and temperature on isothermal release of noble fission gases from large-grain (~ 300-600 µm) UO2 spheres irradiated to 6.4 at.%.
diffuse to the grain boundaries, where they are trapped in the grain boundary bubbles or released by diffusion along the grain boundaries. Above ~ 1100 K, the growth and coalescence of intergranular bubbles cause separation of the grains, forming open porosity and effectively increasing the surface area available for release. The increase in fission gas release due to grain boundary separation does not occur in the monocrystal fuel particle due to the absence of grain boundaries. This explains why release-to-birth rate ratio data for the monocrystal particles did not exhibit a change in slope with increasing fuel temperature below 1600 K (Figure 6). The increase in the release-tobirth rate ratios for granular and monocrystal particles at temperatures > 1600 K is caused by the higher mobility of gas atoms and tiny intragranular bubbles in the fuel grains. The annealing of fission defects in the fuel matrix at such high temperatures could also have contributed to the increase in the mobility and the diffusion coefficient of fission gases, hence, increasing the gas release rate.
Effect of Half-Life on Release-to-Birth Rate Ratio The database showing the effect of half-life on the release of noble gases and volatile fission products from the large-grain (~ 300–600 µm) UO2 microspheres is presented in Figures 7-8 (Friskney and Turnbull, 1974; Turnbull, 1974; Turnbull et al., 1977). Since gaseous radioisotopes decay as they diffuse through the fuel grains and are being trapped at the grain boundaries for a period of time, those having longer half-lives would exhibit higher release-tobirth rate ratios. According to Equation (5), the release-to-birth rate ratio is proportional to the square root of the half-life, for the relatively short-lived isotopes that reach equilibrium early in time. Such dependence of the releaseto-birth rate ratio on half-life is evident in Figures 7a and 8 for the species having half-lives < ~ 10 days. At 1723 K and 4.0 at.% burnup, the small-grain fuel particles (7-40 µm) released essentially all gases and volatile fission products (~ 80% release), whereas only ~ 7% was released from the large-grain (~ 300-600 µm) fuel particles (Figures 7 and 8). It is worth noting that the release fraction from the small-grain UO2 particles is almost the same as that reported for granular plutonia pellets, of the same average grain size (7-12 µm), during re-entry heating tests to 1723 K (Peterson and Starzynski, 1984). When the data in Figure 7a was plotted versus the square root of half-life, it exhibited an exponential increase with half-life, reaching asymptotic values at large half-lives (Figure 7b). These asymptotic values are representative of the release fractions of stable gases in granular UO2 fuel during fission and of helium in plutonia fuel particles of the same grain size. Based on these experimental data, helium gas release from small-grain (7-40 µm) plutonia fuel at 1723 K would be ~ 80%, which is in agreement with the experimental data generated at LANL for GPHS and LWRHU granular plutonia pellets (Angelini et al., 1970; Mulford and Mueller, 1973; Mueller et al., 1974; Land, 1980; Peterson and Starzynski, 1982; Peterson et al., 1984). For large-grain (> 300 µm) and polycrystalline fuel microspheres, when heated up to 1723 K, the helium gas release could be more than an order of magnitude lower (~ 7%). The data presented in Figure 7b is in excellent agreement with the theory (Equation (5)), showing the strong effect of the grain size on R& / B& of fission noble gases and volatile fission products, particularly evident for T1/2 > 1 year.
T = 1723 K
132
Te
85
Kr
10
133
135
Xe
Cs
Xe
136
Cs
137
134
Cs
129m
Te
85m 131
I
1
Release-to-Birth Rate Ratio (%)
Release-to-Birth Rate Ratio (%)
100
133
r = 1.46 %
Kr
I
235
U
87
Kr 138
0.1
0.01 -5
Xe
7 µm (0.4 at%, 2023 K) 10 µm (4.0 at%, 1723 K) 40 µm (0.4 at%, 2023 K) 300-600 µm (4.0 at%, 1723 K) 2500 µm (1.6 at%, 1723 K)
88
Kr r = 0.72 %
-4
-3
-2
-1
0
1
235
U
2
3
100 72% 10 6%
T = 1723 K 1 7 µm (0.4 at%, 2023 K) 10 µm (4.0 at%, 1723 K) 40 µm (0.4 at%, 2023 K) 300-600 µm (4.0 at%, 1723 K) 2500 µm (1.6 at%, 1723 K)
0.1
0.01
0
1
2
3
4
√T1/2 (years
Half-Life, T1/2 (days)
(a) Versus the half-life
1/2
5
6
)
(b) Versus the square root of half-life
FIGURE 7. Effects of half-life and fuel microstructure on isothermal release of fission gases and volatile fission products from granular and monocrystal UO2 fuel samples irradiated at high temperature.
The data presented in Figure 8 illustrates the effect of fuel temperature on the isothermal release-to-birth rate ratios of fission gases and volatile fission products from the large-grain (~ 300-600 µm) UO2 particles (Turnbull, 1974; Turnbull et al., 1977; Friskney and Turnbull, 1979). At 1042 K, only the data for the short-lived noble gases and volatile fission products was reported. This data was extrapolated to higher half-lives using a factor of three, which is the same as that for the hightemperature data between 10 and 105 days half-life (Figure 8). This extrapolation is appropriate since the difference between the release-to-birth rate ratios of the radioactive and the stable gases equals the decay rate of the former, which is independent of temperature.
Release-to-Birth Rate Ratio (%)
Effect of Fuel Temperature 10 136
Cs
137
Cs
134
Cs
129m
Te
132
300-600 µm grains 235 (r = 1.46 % U)
I
I
1
Te 133
131
85m
Kr
87
Kr
extrapolation 138
0.1
Xe
1723 K (B = 4.0 at%) 1042 K (B = 6.4 at%)
0.01
0.001 -5
Xe
133
-4
-3
-2
-1
135
Xe
0
88
Kr
1
2
3
Half-Life, T1/2 (days)
FIGURE 8. Effects of half-life and temperature on isothermal release of fission gases and volatiles from large-grain (~ 300600 µm) UO2 spheres enriched to 1.46%.
The data delineated in Figure 8 indicate that at 1042 K, less than 1% of the fission gases and volatile fission products were released. Similar release fraction would be expected for helium gas in plutonia fuel particles having 300 to 600 µm grains. Also, since the nominal operating temperature in LWRHUs (~ 800 K) is several hundred degrees lower than 1042 K, the helium gas release at the operating temperature in coated particles of large grain size (> 300 µm) plutonia fuel would be practically nil. In summary, the isothermal release data of the fission gases and volatile fission products presented in this section clearly demonstrated the strong effects of the fuel grain size and the half-life on the steady-state release-to-birth rate ratios of the species. The data is consistent with the theory, Equations (4)-(8), particularly the dependence of R& / B& on the half-life for the radioactive species that have attained equilibrium (T1/2 < 10 days). For the small-grain (7-10 µm) UO2 fuel, the release-to-birth rate ratio of the long-lived gases and volatile fission products at 1723 K was nearly 80%. For the large-grain (~ 300-600 µm) fuel particles, the measured release-to-birth rate ratios at same burnup of 4.0 at % and 1723 K were about an order of magnitude lower, ~ 7%, decreasing to < 1% at a fuel temperature of 1042 K and higher burnup of 6.4 at.%. The release-to-birth rate ratios for both small- and large-grain fuel particles increased exponentially with the half-life of the released species, as indicated by the theory (Figure 7b).
Application to Helium Gas Release in Plutonia Fuel Kernels The reported data for the isothermal release of fission gases and volatile fission products from the granular fuel particles have provided a solid foundation for predicting the helium gas release in 238PuO2 kernels, having similar grain size and as-fabricated porosity. The application of the reported data to helium release in plutonia particles includes a certain degree of conservatism. For example, the weakening of the grain boundaries by the bombardment of fission products, which increases fission gas release, does not occur in the α-emitter plutonia fuel. In addition, the constraint imposed by the ZrC coating could decrease the release of helium gas from the 238PuO2 fuel kernels. In addition, due to the absence of grain boundaries, the helium release in polycrystalline 238PuO2 fuel kernels, fabricated using Sol-Gel techniques, is expected to be significantly lower than in large-grain fuel kernels. Based on the reported data for the isothermal release of fission gases in UO2 fuel particles (Turnbull, 1974; Friskney et al., 1977; Turnbull et al., 1977; Friskney and Turnbull, 1979), the following conclusions relative to the helium release in plutonia fuel can be drawn: (a) The helium release from small-grain (7-40 µm) plutonia fuel at 1723 K would be ~ 80%, decreasing to less than 10% at 1042 K. This conclusion is in excellent agreement with the actual helium release data obtained at Los Alamos National Laboratory for GPHS and LWRHU plutonia pellets. (b) The release from large-grain (> 300 µm) plutonia fuel at 1723 K could be ~ 7%, decreasing to ~ 0.8% at 1042 K. (c) In polycrystalline plutonia fuel kernels fabricated using Sol-Gel processes, the helium release fraction could be even lower than that reported for the large-grain 238PuO2 kernels (i.e. < 7% at 1723 K and < 0.8% at 1042 K).
SUMMARY AND CONCLUSIONS The coated plutonia fuel particles have recently been proposed for potential use in future space exploration missions that employ radioisotope power systems and/or RHUs. The particles vary in size from 300 to 1000 µm and consist of a 238PuO2 kernel with a thin (5 µm) PyC inner coating and a strong ZrC outer coating. The thickness of the ZrC is selected to ensure full retention of the helium gas generated by the radioactive decay of 238Pu. The thickness of the ZrC coating, therefore, depends on the actual release fraction of the gas during a simulated re-entry heat pulse, following a storage time of as much as 10 years. During such transient heating, the fuel temperature could reach 1723 K, compared to only ~ 800 K during nominal operation. Reducing the thickness of the ZrC coating increases the fuel loading of the coated particles, and hence the specific thermal power of the coated particles fuel compacts. This paper reviewed the potential release mechanisms of helium in small-grain (7-40 µm) plutonia pellets, currently being used in GPHS modules and LWRHUs, and in large-grain (> 300 µm) and polycrystalline plutonia kernels of the coated particles. The helium release mechanisms are similar to those of fission gases and volatile fission products in oxide and mixed-oxide fuels in fission reactors. The applicability of these mechanisms to small-grain, large-grain and polycrystalline 238PuO2 fuel particles is examined and estimates of the helium release during a reentry heating pulse up to 1723 K are presented. These estimates, based on the reported data of fission gas release from granular and monocrystal UO2 fuel particles irradiated at isothermal conditions up to 6.4 at.% burnup and 2030 K, performed at Harwell and Berkeley Nuclear Laboratories (U.K.), are in good agreement with the helium release tests performed at LANL for small-grain (7-40 µm) plutonia GPHS and LWRHU pellets. It is concluded that the helium release fraction from large-grain (> 300 µm) plutonia fuel kernels heated up to 1723 K could be less than 7%, compared to ~ 80% in small-grain (7-40 µm) fuel. The release fraction of helium in largegrain plutonia fuel kernels could be less than 1% at 1000 K and nil at the nominal operation temperature in RHUs of ~ 800 K. Due to the absence of grain boundaries, the fraction of helium gas released in polycrystalline plutonia kernels fabricated using Sol-Gel techniques could be significantly lower than those in large-grain fuel kernels, for the same storage time and operation temperature. Helium gas release experiments involving large-grain (> 300 µm) and polycrystalline, coated and un-coated plutonia fuel kernels are recommended. In these experiments, the release rate and release fraction of helium gas would be measured as functions of storage time and fuel temperature up to 1800 K, during both steady-state and transient heating conditions. The fuel kernels fabricated using both the Sol-Gel and the powder-metallurgy agglomeration techniques could be used. The PyC undercoating, the ZrC coating, and the PyC overcoating of the kernels would be applied using the state-of-the-art Chemical-Vapor-Deposition techniques. The proposed tests, which may run continuously for up to two years, could be performed at an approved and qualified DOE facility.
ACKNOWLEDGMENTS This research was funded by Sandia National Laboratories (SNL), Kirtland Air Force Base, Albuquerque, NM, under Contract No. BE-2543, to the University of New Mexico’s Institute for Space and Nuclear Power Studies. The opinions expressed in this paper are solely those of the authors. We are grateful to Dr. Ronald J. Lipinski, SNL, and Mr. Joseph A. Sholtis, Jr., Sholtis Engineering & Safety Consulting, for their continuous technical support.
REFERENCES Allen, P. L., Ford, L. H., and Shennan, J. V., “Nuclear Fuel Coated Particle Development in the Reactor Fuel Element Laboratories of the U. K. Atomic Energy Authority,” Nuclear Technology 35, 246–253 (1977). Angelini, P., McHenry, R. E., Scott, J. L., Ernst, W. S., Jr., and Prados, J. W., Helium Release from 238PuO2 Microspheres, Report No. ORNL-4507, Oak Ridge National Laboratory, Oak Ridge, Tennessee, March 1970. Burnett, R. C., Bisdorff, L., and Gough, J. R. C., Development of Coated Particle Fuel – Part I: the Fissile Fertile Particle, Dragon Project Report No. 151, also in Proceedings of the Dragon Project Fuel Element Symposium, held January 28-29, 1963, in Bournemouth, England, CONF-630103-8 (1963). El-Genk, M. S., and Tournier, J.-M., Study of a Coated-Particle Fuel Form for Advanced Radioisotope Heat Sources – Part I: A Review of Coated-Particle Fuel Performance, Fabrication Techniques and Materials Properties, Report UNM–ISNPS–5– 1999, Sandia National Laboratories, Contract BE–2543, (1999). Ford, L. H., and Shennan, J. V., “the Mechanism of Binderless Granulation and Growth of Ceramic Spheres,” Journal of Nuclear Materials 43, 143 (1972). Förthmann, R., and Blass, G., “Fabrication of Uranium-Plutonium Oxide Microspheres by the Hydrolysis Process,” Journal of Nuclear Materials 64, 275–280 (1977). Friskney, C. A., Turnbull, J. A., Johnson, F. A., Walter, A. J. and Findlay, J. R., “The Characteristics of Fission Gas Release from Monocrystalline Uranium Dioxide during Irradiation,” Journal of Nuclear Materials 68, 186–192 (1977). Friskney, C. A., and Turnbull, J. A., “The Characteristics of Fission Gas Release from Uranium Dioxide during Irradiation,” Journal of Nuclear Materials 79, 184–198 (1979). Haas, P. A., Kitts, F. G., and Beutler, H., “Preparation of Reactor fuels by Sol-Gel Processes,” Chemical Engineering Progress Symposium Series 63 (80), 16–27 (1967). Huschka, H., and Vygen, P., “Coated fuel Particles: Status of Fabrication Technology,” Nuclear Technology 35, 238–245 (1977). Land, C. C., Microstructural Damage Produced by Helium in Aged 238PuO2 Fuels, Report No. LA-8083, Los Alamos Scientific Laboratory, Los Alamos, New Mexico, January 1980. Lahr, H., “Fabrication and Properties of Particle Fuels,” Kerntechnik 19, 159 (1977). Lyons, M. F., Boyle, R. F., Davies, J. H., Hazel, V. E., and Rowland, T. C., “UO2 Properties Affecting Performance,” Nuclear Engineering and Design 21, 167–199 (1972). Minato, K., Ogawa, T., Fukuda, K., Sekino, H., Kitagawa, I. and Mita, N., “Fission Product Release from ZrC-Coated Fuel Particles during Post-Irradiation Heating at 1800 and 2000 oC,” Journal of Nuclear Materials 249, 142–149 (1997). Mulford, R. N. R., and Mueller, B. A., Measurements of Helium Release from Materials Containing 238PuO2, Report No. LA5215, Los Alamos Scientific Laboratory, Los Alamos, New Mexico, July 1973. Mueller, B. A., Rohr, D. D., and Mulford, R. N. R., Helium Release and Microstructural Changes in 238PuO2, Report No. LA5524, Los Alamos Scientific Laboratory, Los Alamos, New Mexico, April 1974. Olander, D. R., Fundamental Aspects of Nuclear Reactor Fuel Elements, U.S. ERDA Report No. TID-26711-P1, Technical Information Center, Energy Research and Development Administration, Oak Ridge, Tennessee, 1976, Chapters 13–15. Peterson, D. E., and Starzynski, J. S., Re-entry Thermal Testing of Light-Weight Radioisotope Heater Units, Report No. LA-9226, Los Alamos National Laboratory, Los Alamos, New Mexico, March 1982. Peterson, D. E., Early, J. W., Starzynski, J. S., and Land, C. C., Helium Release from Radioisotopic Heat Sources, Report No. LA-10023, Los Alamos National Laboratory, Los Alamos, New Mexico, 1984. Scaffidi-Argentina, F., Donne, M. D., Ronchi, C., and Ferrero, C., “ANFIBE: A Comprehensive Model for Swelling and Tritium Release from Neutron-Irradiated Beryllium – I: Theory and Model Capabilities,” Fusion Technology 32, 179–195 (1997). Schock, A., “Light-Weight Radioisotope Heater Unit,” in Proceedings of the 16th Intersociety Energy Conversion Engineering Conference, held in Atlanta, GA, 9–14 August 1981, Paper No. 819175, AIAA, 1981, pp. 343–354. Sholtis, J. A., Jr., Lipinski, R. J., and El-Genk, M. S., “Coated Particle Fuel for Radioisotope Power Systems (RPSs) and Radioisotope Heater Units (RHUs),” in Proceedings of the Space Technology and Applications International Forum – 1999, M. S. El-Genk, Ed., American Institute of Physics Conf. Proceedings No. 458, Woodbury, New York, 1999, 2, 1378–1384. Tournier, J.-M., and El-Genk, M. S., “Performance Analysis of Coated 238PuO2 Fuel Particles Compact for Radioisotope Heater Units,” in these Proceedings, M. S. El-Genk, Ed., American Institute of Physics, Woodbury, New York, 2000. Turnbull, J. A., “The Effect of Grain Size on the Swelling and Gas Release Properties of UO2 during Irradiation,” Journal of Nuclear Materials 50, 62–68 (1974). Turnbull, J. A., Friskney, C. A., Johnson, F. A., Walter, A. J., and Findlay, J. R., “The Release of Radioactive Gases from Uranium Dioxide During Irradiation,” Journal of Nuclear Materials 67, 301–306 (1977).
