J.W. Adams, P.R. Lageraaen, P.D. Kalb, and B.R. Patel
Abstract
Depleted uranium, in the form of uranium trioxide (UO3) powder, was encapsulated in
molten polyethylene forming a stable, dense composite henceforth known as polyethylene encapsulated DU (patent
pending). Materials were fed by calibrated volumetric feeders to a single screw extruder where
they were heated and mixed to form a homogeneous molten extrudate. Oxide loadings as high
as 90 wt% UO3 were successfully processed, yielding a maximum product density of 4.2 g/cm3.
Performance testing included compressive strength, water immersion and leach testing.
Compressive strengths of samples with 50-90 wt% UO3 were nearly constant, with a mean value
exceeding 15 MPa (2200 psi). Leach rates, which increased as a function of sample waste
loading, were less than 1.1% after 11 days at ambient temperature for samples containing 90 wt%
UO3. Ninety day water immersion tests showed sensitivity to "batch" processed UO3 for samples
containing >85 wt% of the oxide. Considering that UO3 should be insoluble in water, these
results indicate the probable presence of other, more soluble uranium compounds. Samples
containing UO3 produced by a "continuous" process showed no deterioration at up to 90 wt%
waste loadings.
Introduction
Department of Energy (DOE) facilities maintain large inventories of depleted uranium
(DU). Novel applications are currently being sought to convert these materials to stable, useful
secondary products. Uses that provide a positive benefit to society while allowing potential
recovery or extraction of the uranium are desirable, but techniques for stabilization of DU for
long-term storage or disposal are also being evaluated. Potential applications will likely exploit
the high density, shielding effectiveness and nuclear applicability of these materials. This study,
in particular, was initiated to investigate the feasibility of processing depleted uranium (e.g., UO3
powder) by polyethylene microencapsulation, to mitigate potential health effects and produce
useful radiation shielding and other products.
Natural uranium ore in the form of U3O8 contains about 0.7 weight percent of the
fissionable isotope 235U, with the remainder of uranium present as 238U. Reactor fuel is produced
from the ore by converting it to uranium hexafluoride gas and enriching the proportion of 235U to
around 3.5 percent, leaving the remaining portion depleted in 235U. This residual material, with
235U concentrations at around 0.25 percent, is known as depleted uranium. Approximately
560,000 metric tons of DU in the form of UF6, containing an equivalent mass of 379,000 metric
tons of uranium, are stored at the DOE Paducah, Portsmouth and Oak Ridge Gaseous Diffusion
plants. Some of the UF6 has been converted to uranium trioxide (UO3); about 20,000 metric tons
of DU are currently stored at the Savannah River Site. UO3 from Savannah River was used in this
preliminary investigation.
Alternatives for management of the DU inventory under consideration by the U.S. DOE
include 1) continue current management plan (no action); 2) revise current practices for long-term
storage as either UF6 or in an oxide form; 3) use of DU in shielding or high density applications;
4) disposal of DU.[(1)] Since uranium and uranium oxides are considered valuable resources, use
of the material (option 3) is most attractive. As polyethylene encapsulated DU will attenuate neutron as well as gamma
radiation, there is currently significant interest for its use in spent fuel storage and transport casks
and low level waste storage, transport and disposal packages. The high density and workability
of the material make it appealing for use as a ballast in nautical and aeronautical applications.
Treatment of DU materials by polyethylene encapsulation is a desirable option because of
the immediate availability of the technology and proven record to effectively and efficiently
process similar powder and granular materials. In addition, the process is very flexible.
Polyethylene products can be heated and reworked if future needs change. DU can potentially be
retrieved from the matrix by chemical and/or thermal processing if needed as a resource in the
future. Over the last twelve years, Brookhaven National Laboratory (BNL) has extensively
developed the polyethylene encapsulation extrusion process for low-level radioactive, hazardous,
and mixed wastes.[(2),(3),(4),(5),(6),(7),(8),(9)] During processing, filler materials (e.g., waste) are
mechanically mixed into the molten polyethylene binder, producing a workable homogeneous
product. The process is not susceptible to chemical interactions between the waste and binder,
enabling a wide range of acceptable waste types, high waste loadings, and technically simple
processing under heterogeneous waste conditions. The process has evolved from proof-of-principle, through bench-scale development and testing, to full-scale technology demonstration
and technology transfer.
Equipment and Procedures
Representative samples of depleted UO3 powder were obtained from Westinghouse
Savannah River Company (WSRC). Savannah River alone maintains over 22 million kg (50
million pounds) of depleted UO3 stored in 55 gallon drums. This inventory consists of material
of two distinct lots corresponding to two different processes used to prepare the oxide, batch and
continuous. Over 99% of this inventory was produced with the batch process. The remaining 1%
of the inventory, produced by a newer continuous process, is chemically identical but is
characterized by a slightly larger particle size. Extrusion process runs described below were
carried out using the batch process material.
The UO3 materials shipped from WSRC were characterized in a recent report by Carolina
Metals, Inc.[(10)] The drummed material was described as a 200 mesh (74 m average particle
size), 96.5% uranium trioxide with trace impurities of aluminum, iron, phosphorous, sodium,
silicon, chromium and nickel. The material has a bulk density range of about 2.5 g/cm3
(uncompacted) to about 3.6 g/cm3 (compacted). The 235U content was assayed at approximately
0.2% and the plutonium content at 3 ppb. Gross gamma was 53,100 dpm per gram of uranium.
A 32 mm (1.25 in diameter) single-screw, non-vented, Killion extruder was used for
processibility testing. The extruder is equipped with a basic metering screw, three heating/cooling
barrel zones and an individually heated die. The polyethylene and the DU powder were metered
to the extruder through AccuRate, 300 Series, volumetric feeders. These feeders are designed to
provide a constant volume output at a given operating setting (varies as a percentage from zero
to 100% output). Due to differing densities of feed materials, calibration of the feeders was
required with each material. The resulting data provided feeder output in g/min versus feeder
speed setting.
For a given process run, a number of different samples were taken in replicate (typically
ten for statistical assurance) to monitor the process and characterize the extruder output. Rate
samples were one minute collections of the extruder output to determine process consistency over
time. Low variation between replicate rate samples indicated the output was continuous and that
the material was successfully processed at a given DU loading. Grab samples were small
(approximately 3-10 g) specimens of the output product taken for pycnometer density
measurement. Low variation between replicate grab samples indicated that the DU material fed
well and was consistently well mixed with the polyethylene as it was processed in the extruder.
2x4 samples, nominal 5.1 cm (2.0 in) diameter by 10.2 cm (4.0 in) tall right cylindrical
specimens, were fabricated for compressive strength[(11)] and immersion testing. Specimens were
formed in pre-heated brass molds, compressed with up to 0.2 MPa (25 psi) pressure. ALT
samples, for use in Accelerated Leach Testing[(12)], were nominal 2.5 cm (1.0 in) diameter by 2.5
cm (1.0 in) tall right cylinders fabricated in individual Teflon molds. These samples were also
compression molded with up to 1.7 MPa (250 psi) pressure. Disk samples, for use in future
attenuation studies, were formed in glass petri dishes with <0.1 MPa (10 psi) pressure. Disks
were nominally 11.7 cm (4.5 in) diameter and were fabricated at varying thicknesses, up to
approximately 2.5 cm (1.0 in).
Process Results
Processibility testing with UO3 was initially conducted at a loading of 50 weight percent
(wt%). This loading was selected based on previous microencapsulation experience. Process
results at 50 wt% were not immediately successful because the extrudate contained trapped gases,
producing voids in samples during cooling. DU powders were oven-dried overnight at the
maximum process temperature (160șC) to ensure removal of any excess moisture prior to
extrusion.
Processing with dried DU produced excellent results. DU was successfully processed at
loadings of 50, 60, 70, 75, 80, 85, and 90 wt%. As DU loading increased, the extrudate became
more viscous and dry, with a rough surface texture due to the decreasing amounts of polyethylene
available to encapsulate and lubricate the DU particles. However, even at 90 wt% the DU was
readily processible and samples could be successfully cast with the aid of compression
molding.
At 95 wt% DU, the material plugged causing output to cease and die pressures to rise
above their alarm set point (25 MPa). At this loading there was insufficient polyethylene to mix,
wet and convey the DU through the extruder barrel. DU flow was stopped immediately after
noting the plugged condition. The clog was voided within several minutes of stopping the DU
flow, introducing pure polyethylene to the screw. Motor load, as measured by current draw, rose
slightly during this episode, but remained within acceptable limits.
The overall success encountered during processing was evidenced from rate and grab
sample statistics. The low deviation and percent error between ten replicate samples taken at each
DU loading, shown in Table 1, indicate that the UO3 powders processed continuously and
consistently. These runs were conducted at screw speeds of either 60 or 65 rpm and at combined
(DU + binder) feed rates of 100 to 120 g/min. Feed rates were increased proportional to DU
loading, as the actual volume of feed material decreased with increased DU loadings.
Table 1. Process Rate Samples (g/min)
|
DU Loading |
| 50 wt% |
60 wt% |
70 wt% |
75 wt% |
80 wt% |
85 wt% |
| Mean |
114.23 |
109.23 |
111.69 |
117.78 |
125.63 |
124.13 |
| Std.Dev. |
3.45 |
2.71 |
3.37 |
1.48 |
2.27 |
2.87 |
| 2 sigma |
2.47 |
1.94 |
2.41 |
1.06 |
1.62 |
2.05 |
| % Error |
2.16 |
1.76 |
2.16 |
0.90 |
1.29 |
1.65 |
Product Characterization
The polyethylene encapsulated DU product was characterized by density measurement, compressive strength
testing, and leach and immersion testing in deionized water. Physical characteristics and
performance of the product varied significantly as a function of DU loading. Most
obvious of these was product density. polyethylene encapsulated DU densities ranged from 1.38 to 3.93 g/cm3 for
uncompressed samples (disk, 2x4, and uncompressed ALT forms) for the range of 50 to 90 wt%
DU. A density increase of approximately 10-15% was observed using compression molding, with
mean values ranging from 1.62 to 4.25 g/cm3 for compressed ALT forms at 50 to 90 wt% DU.
DU density as a function of wt% DU loading is depicted in Figure 1 for both compressed and
uncompressed samples. DU densities were calculated by multiplying the mean polyethylene encapsulated DU density
times the weight percent of DU in the sample, for a given DU loading.
Initial process runs were conducted using batch process UO3. Process runs using
continuous process UO3 produced nearly identical values for compressed forms, whereas
uncompressed sample densities were slightly higher than corresponding batch process samples.
This artifact was probably attributable to improved molding technique with subsequent runs,
allowing fewer voids to be trapped in the product while filling the molds. For both batch and
continuous process, DU densities for 90 wt% samples were higher than the reported
density of a vibration compacted sample of the dry powder (3.5 g/cm3). Uncompacted UO3
powder, which has a density of about 2.5 g/cm3, is surpassed at about 80 wt% for
compressed samples and about 85 wt% for uncompressed samples. In other words, at these DU
loadings, the polyethylene encapsulated DU process represents a volume reduction compared with disposal of untreated
UO3. Such high product densities are achieved because of an increased volume packing efficiency
for the DU particles during processing. This effect may be attributed to reduced particle
agglomeration due to drying of the particles during thermal treatment, comminution of the
particles due to mechanical abrasion during processing, or compressive forces exerted during
forming.
Polyethylene encapsulated DU 2x4 samples were compression tested in accordance with ASTM D-695, "Standard
Method of Test for Compressive Properties of Rigid Plastics".[11] Testing was done using a
Soiltest hydraulic compression tester at an unloaded crosshead deflection rate of 1.3 ± 0.3 mm
(0.05 ± 0.01 in.)/min. Crosshead speed and total deflection were monitored using a dial gauge
and lab timer. Load and deformation were recorded at 60 second intervals. Compressive yield
strength is plotted against DU loading in Figure 2. At least 8 samples were tested at each DU
loading.
With batch and continuous process polyethylene encapsulated DU data averaged together (filled squares),
maximum yield strength is relatively constant between 50 and 85 wt% DU considering the range
of measurement error. At 90 wt%, a statistically significant increase was noted, probably due to
particle-to-particle contact of the DU in the matrix, with barely enough polyethylene present to
fill void spaces. This fact is reflected in the percent deformation at yield, reduced from
approximately 26% for 50 wt% samples to only 7% for 90 wt% samples.
Polyethylene encapsulated DU forms containing 50, 70 and 90 wt% batch process UO3 were leach tested in
accordance with ASTM C1308.[12] This Accelerated Leach Test (ALT), developed at BNL, was
devised to enable prediction of a sample leach rate providing data fits a diffusion-controlled
model. In theory, the effective diffusion coefficient (De) is temperature dependent according to
the Arrhenius equation: De= Aexp(-Ea/RT), where Ea is the activation energy. Due to limited
scope, all leach testing was done at ambient temperature. Leachates were analyzed by inductively
coupled plasma (ICP) spectroscopy for their total uranium metal concentration.
Leach testing of batch process polyethylene encapsulated DU forms produced cumulative uranium releases of
approximately 1.1% for 90 wt% DU and approximately 0.07% for both 50 and 70 wt% DU
samples, after 11 days (Figure 3). These results are typical for high loadings of soluble salts
microencapsulated in polyethylene. However, considering the insolubility of uranium trioxide in
water[(13)], these data indicate the probable presence of other, more soluble uranium compounds.
While the UO3 is reportedly 96.5% pure (82.25-78.47% total U), it is likely that other
hygroscopic uranium compounds are present and unaccounted for in the DU. Although
identification of these salts was beyond the scope of this effort, all uranium halides are very
soluble, as are uranium and uranyl sulfates and nitrates. The high solubility of the as-received
batch DU was further evidenced in that a source term leach sample (50 g batch process DU in
3000 ml water) saturated within the first (2 hr.) leach interval. Continuous process polyethylene encapsulated DU
samples were not tested.
Water immersion testing was performed using one 2x4 and one ALT form of each DU type
and waste loading. Samples were immersed in distilled water to determine possible deleterious
effects of a saturated environment. Three or four similar samples were grouped together in a
single polyethylene container, with a water/sample ratio of 1000 ml per sample for 2x4 samples
and 200 ml per sample for ALT's. The test, done at ambient temperature, was a 90 day static
immersion after which time the sample weights and volumes were re-measured. Samples
remaining intact on completion of the test were compression tested to determine potential for non-visible degradation.
After 90 days, visible degradation was only evident on samples containing 85 and 90 wt%
batch process UO3. Samples containing 80 wt% or less batch process UO3 were unchanged, as
were all samples containing up to 90 wt% continuous process UO3. The 90 wt% batch process
samples began showing signs of cracking around the top and bottom perimeter within the first
week of immersion. Cracks in the 85 wt% batch process samples were not noticed until the third
month of the test. After 90 days, 90 wt% batch process samples were severely deteriorated, with
cracks running lengthwise penetrating nearly to the center of both samples. 85 wt% batch process
samples contained only three or four minor cracks (<1 cm) along the sample sides. Since the
only polyethylene encapsulated DU samples that degraded under water immersion testing contained high concentrations
of batch processed DU, this phenomena is thought to be related to the presence of soluble
impurities, as discussed above.
Post-immersion compressive strengths of 50, 60, 70, 75, 80 and 85 wt% batch process
UO3 samples were 16.9, 17.0, 9.6, 16.5, 13.6, and 9.2 MPa (2450, 2460, 1390, 2390, 1980, and
1340 psi), respectively. Post-immersion compressive strengths of 70, 80 and 90 wt% continuous
process UO3 samples were 18.5, 16.8, and 18.2 MPa (2680, 2440, and 2640 psi), respectively.
Except for the 70 and 85 wt% batch process UO3 samples (28 and 41% decrease in strength,
respectively), post immersion compressive strengths were statistically equivalent to non-immersed
samples.
Conclusions
Depleted UO3 powders, representative of the depleted uranium inventory at the Savannah
River Site, were successfully stabilized by polyethylene encapsulation using BNL's single-screw
extrusion process. Material corresponding to two different processes used to prepare the oxide,
batch and continuous, were processed. Loadings as high as 90 wt% DU were successfully
achieved for both lots. A maximum product density of 4.2 g/cm3 was obtained. Additional
process improvements are under consideration which are expected to lead to product
densities as high as 7.2 g/cm3.
Waste form performance testing of microencapsulated UO3 included compressive strength,
water immersion and leach testing. Compression test results were in keeping with measurements
made with other waste materials encapsulated in polyethylene, typically around 14 MPa (2000
psi). Leach rates, which varied according to UO3 loading, were similarly in line with
polyethylene waste forms bearing high salt waste loadings: approximately 1% after 11 days for
samples containing 90 wt% UO3. Considering the insolubility of uranium trioxide, however, these
leach data indicate the probable presence of other, more soluble uranium compounds present in
the batch process UO3 powder. The continuous process polyethylene encapsulated DU material was not tested. Ninety
day water immersion tests concluded that water absorption was inconsequential except for batch
process samples at very high (>85 wt%) waste loadings. Sample degradation in batch
process samples corresponds to the increased leach rate observed during ALT testing, probably
resulting from hygroscopic impurities. In contrast, continuous process DU showed no evidence
of swelling/cracking during 90 day immersion testing even at the highest DU loading of 90 wt%.
Therefore, continuous process DU provides a significantly more stable and durable product.
References
1. U.S. DOE, "Alternative Strategies for the Long-Term Management and Use of
Depleted Uranium Hexafluoride, Notice of Intent," 61 Fed Reg 2239, January 25, 1996.
2. Kalb, P.D. and P. Colombo, "Polyethylene Solidification of Low-Level Wastes, Topical
Report," BNL-51867, Brookhaven National Laboratory, Upton, NY, October 1984.
3. Kalb, P.D., J.H. Heiser, III, and P. Colombo, " Polyethylene Encapsulation of Nitrate
Salt Wastes: Waste Form Stability, Process Scale-up, and Economics," BNL- 52293,
Brookhaven National Laboratory, Upton NY, July 1991.
4. Kalb, P.D., and M. Fuhrmann, "Polyethylene Encapsulation of Single-Shell Tank
Wastes," BNL-52365, Brookhaven National Laboratory, Upton, NY, September 1992.
5. Kalb, P.D., J.H. Heiser, and P. Colombo, "Long-Term Durability of Polyethylene for
Encapsulation of Low-Level Radioactive, Hazardous, and Mixed Wastes," Emerging
Technologies in Hazardous Waste Management III, D.W. Tedder and R.G. Pohland,
ed., American Chemical Society Symposium Series 518, Washington, D.C., 1993.
6. Kalb, P.D., J.W. Adams, M. Meyer, and H. Holmes Burns, "Thermoplastic
Encapsulation Treatability Study for a Mixed Waste Incinerator Off-Gas Scrubbing
Solution," Third International Symposium on Stabilization/Solidification of Hazardous,
Radioactive, and Mixed Wastes, November 1993.
7. Adams, J.W., and P.D. Kalb, "Thermoplastic Stabilization Treatability Study for a
Chloride, Sulfate and Nitrate Salts Mixed Waste Surrogate," Emerging Technologies
in Hazardous Waste Management VI, American Chemical Society, D.W. Tedder and
R.G. Pohland, ed., in press.
8. Kalb, P.D., and P.R. Lageraaen, "Full-Scale Technology Demonstration of a Polyethylene
Encapsulation Process for Radioactive, Hazardous, and Mixed Wastes," Journal of
Environmental Science and Health, in press.
9. Lageraaen,
P.R., B.R. Patel, P.D. Kalb, and J.W. Adams, "Treatability Studies for
Polyethylene Encapsulation of INEL Low-Level Mixed Wastes," BNL-62620,
Brookhaven National Laboratory, Upton, NY, October 1995.
10. Brady, S.D., D.A. Krough, and R.P. Laporte, "Depleted Uranium Trioxide (UO3)
Characterization and Storage," 94AB62387P-00F, Carolina Metals, Inc., January 1995.
11. ASTM, "Standard Method of Test for Compressive Properties of Rigid Plastics," D695,
American Society for Testing and Materials, Philadelphia, PA, 1969.
12. ASTM, "Accelerated Leach Test for Diffusive Releases from Solidified Waste and a
Computer Program to Model Diffusive, Fractional Leaching from Cylindrical Waste
Forms," C1308, American Society for Testing and Materials, Philadelphia, PA, 1995.
13. CRC Handbook of Chemistry and Physics, ed. R.C. Weast and M.J. Astle, CRC Press,
Inc, Boca Raton, FL, 1981.
