Small-scale containers provide a chance to study many more materials

Gas generation by pure and impure plutonium dioxide materials

Over the last several years, DOE has produced more than 4,500 containers of plutonium-bearing material that meet the requirements of DOE's 3013 Standard. These containers hold approximately 14.0 metric tons of metal, oxide, and impure oxide (30 weight percent or greater of plutonium plus uranium). The packaging configuration specified by the standard consists of robust, nested, welded containers with a minimum design pressure of 699 pounds per square inch gauge (psig). This is in contrast to interim storage containers in use at TA-55 designed by Roland Hagan that are fitted with filters to allow "breathing," or to crimp-sealed cans, which can withstand only minor internal pressures. Gases generated by the material over the fifty-year-allowed storage lifetime will remain within the container and could lead to pressurization and corrosion of the packaging.

During the development of the current 3013 Standard, there was limited information concerning gas generation from actual plutonium-bearing materials. It was assumed that the only mechanisms for gas generation were the production of helium from alpha decay (a minimal source of gas) and radiolysis of water. It was also assumed that all of the water in the container would radiolyze to hydrogen gas and the oxygen from the radiolysis of water would be taken up by the material.

The lack of oxygen in the gas-phase reduces the hazard of the system in two ways. First, the overall pressure from radiolysis of water is reduced by a third, one mole of hydrogen for every mole of water rather than one mole of hydrogen and a half mole of oxygen for every mole of water. Second, the possibility of a flammable mixture within the container is removed. These assumptions were based on a handful of experiments and extensive experience throughout the DOE complex with plutonium-bearing materials. However, the gas-generation behavior of materials that are actually being packaged had not been studied.

The DOE 3013 Standard specifies a set of robust, nested, welded containers. Shown are the British Nuclear Fuels, Ltd.-designed outer, inner, and convenience containers.

During the last two years, Los Alamos' Surveillance and Monitoring Program (formerly the 94-1 Program) has studied the gas-generation behavior of materials provided by Rocky Flats Environmental Technology Site (RFETS), Hanford, and Los Alamos, which are representative of the classes of materials that have been packaged by sites across the DOE complex. Large-scale instrumented containers that can hold five kilograms (kg) of material within a 2.38-liter internal volume were made by modifying a British Nuclear Fuels, Ltd. (BNFL) inner container to include a Raman spectroscopy chamber, thermocouples, pressure sensor, corrosion sensors, and a gas manifold to extract samples for gas chromatography or mass spectrometry analysis. These containers are meant to provide conditions as close as possible to the conditions experienced by materials packaged under the 3013 Standard.

Because of the size of the containers and the amount of material needed, only a limited number of studies are planned using the large-scale containers. An array of 45 Los Alamos-designed, small-scale containers, scaled to 1/500th the volume of the BNFL 3013 inner container, or 5 milliliters, provides a chance to study many more materials and to vary the moisture content. Results from these small-scale containers are compared to the large-scale container results to verify that they are applicable to packaged materials. The pressure of each small-scale container and the temperature of the heated block that holds it are monitored continuously. Each heated block contains nine small-scale containers and the temperature of each of the five blocks, which is held constant by a controller, can be varied independently. Each small-scale container has a 45-microliter sampling loop that provides the capability of gas composition analysis using a gas chromatograph.

Left: A small-scale container is shown beside a large-scale container. Right: Forty-five of the small-scale containers are assembled in an array.

Gas generation by pure oxides

A high-purity plutonium oxide material was studied in both the large- and small-scale containers. This material contained 87.8 percent plutonium with 0.15 percent identified impurities (stoichiometric PuO₂ of this purity and isotopic composition should be 88.1 percent), had a specific surface area of 1.08 square meters per gram (m²/g), and a pycnometer density of 11.5 grams per cubic centimeter (g/cm³). This material is a very fine brown powder.

Five kilograms of high-purity plutonium oxide were calcined to 950 degrees Celsius, and the freshly calcined material was sealed in a large-scale container with pure helium as the fill gas. The temperature, pressure, and gas composition were monitored for 469 days. No change in the gas composition was observed. Neither hydrogen nor oxygen was detected; the only gas observed within the container was helium. The pressure increased by only 0.3 kilopascal (kPa). The majority of the helium from alpha decay of the plutonium apparently remained trapped within the oxide particles. The amount of helium from alpha decay would have been sufficient to raise the pressure by 1.2 kPa had all of the helium entered the gas phase.

A high-purity plutonium oxide material was studied in the large- and small-scale containers.

Results of experiments on gas generation by high-purity plutonium dioxide exposed to humidified helium are shown at left.

The material was then exposed to a flow of humidified helium gas-60 percent relative humidity at 25 degrees Celsius (with a water-vapor pressure of about 1.9 kPa)-for 24 hours. The material picked up 1.25 grams of water-0.025 weight percent (wt %), which is equivalent to one monolayer of water. The fill gas was changed to helium with about 10 percent air to reproduce the conditions experienced by containers being packaged at RFETS. Under these starting conditions, we initially observed nitrogen, oxygen, and water in addition to helium as the gas composition of the fill gas. The water was observed by Raman spectroscopy-gas chromatography (GC) analysis using our instrumentation does not detect water. Oxygen was depleted to below detection limits within twenty-five days. The water in the gas phase, which is a small fraction of the water added to the container, decreased by about half and remained at a steady partial pressure for more than one hundred days. No hydrogen gas was observed.

Rates of gas generation in units of kilopascal per day for carbon dioxide (CO2), nitrous oxide (N₂O), and hydrogen (H₂).

The calcined high-purity plutonium oxide material was also used for studies in the small-scale containers that started about one year after the large-scale containers were loaded. The material was exposed to the dry glovebox atmosphere during that time. The material apparently adsorbed atmospheric gases during that time. We studied high-purity plutonium oxide with no water added, with 0.5 wt % added water, and with 2.0 wt % added water. The starting material would have picked up some water during its one-year exposure to the glovebox atmosphere. The amount of water in the starting material is not known but is expected to be less than 0.1 wt %.

Several gaseous species were produced under these conditions that were not seen in the large-scale studies. Gas chromatography analysis of the fill gas immediately after the reactors were sealed showed mainly helium with nitrogen and oxygen and barely detectable hydrogen. Within five days, gas chromatography analysis of the fill gas shows CO₂, N₂O, and H₂ present in quantities well above the sensitivity of the GC. Interestingly, during the first five days, the gas-generation rates for CO₂ are substantially larger than the gas-generation rate of H₂ regardless of the amount of water added to the container. Nitrous oxide is produced at the second fastest rate in the first five days. For the containers with water added, the rate of N₂O production is slightly less than the rate of CO₂ production, both of which are larger than the H₂ gas-generation rate by a factor of 4 to 5 times greater. When water is not added to the container, the rate of N₂O production is smaller by a factor of 10 as compared to the containers with water added yet is still larger than the hydrogen gas-generation rate when no water is added. Hydrogen is generated at the slowest rate during the first five days.

Pressure versus time of the containers for pure plutonium oxide samples with no water added (green curve), 0.5 weight percent water added (blue curves), and 2.0 weight percent water added (red curve).

During the next thirty-one days, the overall gas-generation rate decreases because the CO₂ and N₂O rates decrease substantially to about 10 to 20 percent of what they were in the first five days. The H₂ rate increases during this time for the containers with water added and decreases in the containers with no added water. Gas-generation rates of 0.05 kPa/day or less are difficult to measure, so the rates reported will have an error of approximately that magnitude. The next twelve days show very little change in the gas-generation rates from days six to thirty-seven. The H₂ gas-generation rate in the containers with added water seems to increase, but this is probably due to measurement errors.

The pressure versus time (PVT) of the containers for the sample with no water added (green curve), four samples with 0.5 wt % water added (blue curves), and one sample with 2.0 wt % water added (red curve) is shown at left. The curve for the container with no water added is easily distinguished from the curves of the containers with added water. The curve for the container with 2.0 wt % added water appears quite similar to the curves for the containers with 0.5 wt % water added. The curves were adjusted for clarity to be on the same scale and slightly offset by adding a constant value to the pressures.

The observation that H₂ gas-generation rates are very similar if not the same for pure oxides with 0.5 and 2.0 wt % added water is best explained by considering how much water the material can reasonably absorb. A pure plutonium oxide particle will adsorb water onto its surface until the surface is saturated, which occurs at three to five monolayers when the particle is exposed to a high relative humidity atmosphere. One monolayer of water on high-purity plutonium oxide with a specific surface area of 1.08 m2/g occurs at about 0.025 wt %. Saturation of the surface with water will start to occur at about 0.1Ð0.2 wt %. Additional water added to the container will seek the coldest part of the container, which in this case is not the plutonium oxide because it self heats. Thus, the containers with 0.5 and 2.0 wt % added water will behave similarly because the extra water is condensed away from the plutonium oxide and not exposed to significant alpha radiation. This is not the case for impure oxides containing salts.

Results of experiments on gas generation by impure oxides.

Comparison of the pressure versus time curve for two impure oxide samples and one high-purity sample.

Gas generation by impure oxides

Typical Los Alamos impure oxide materials in the packaged 3013 containers contain plutonium oxide with salt mixtures of sodium chloride (NaCl), potassium chloride (KCl), and magnesium chloride (MgCl₂). We have studied salt-containing materials in both the large- and small-scale containers. The salt-containing material in the large-scale container was produced by combining burned anode heals and electrorefining salt residues for a blend that would have sodium and potassium chloride salts with a trace of magnesium chloride salts. This material was calcined at 800 degrees Celsius for two hours.

Water was added by flowing humidified helium (with a water-vapor pressure of approximately 2.3 kPa) through the container from the bottom to the top. The material gained 8 grams of water (0.19 wt %) after 24 hours. Hydrogen was generated and oxygen was depleted. Water was seen in the gas phase, indicating that the salts were in equilibrium with the water vapor. The initial hydrogen gas-generation rate was 0.19 kPa/day.

Researchers in Los Alamos' Materials Identification and Surveillance (MIS) Program received from Hanford a low-purity plutonium-bearing salt with 70.0 percent plutonium, 5.5 percent chlorine, 1.9 percent potassium, 1.5 percent sodium, and 0.5 percent magnesium. Seven samples were studied in small-scale containers: one with no added water, five with 0.5 wt % added water, and one with 2.0 wt % added water. In contrast to the pure oxide material, there were large differences in the gas-generation rates of samples with different amounts of added water. In addition, at the higher water loading, oxygen was generated as well as hydrogen.

A comparison of the pressure versus time curve for the impure PuO₂ salt (Hanford) samples and one of the pure oxide samples shows that the salt materials generate gas more rapidly than the pure material. Analysis of the gas composition reveals that the pressure rise for the impure PuO₂ salt sample with 0.5 wt % water added is composed of hydrogen gas generation and oxygen gas depletion. The pressure rise for the impure PuO₂ salt sample with 2.0 wt % water added was caused by both hydrogen and oxygen gas generation. The hydrogen gas-generation rates obtained from gas analysis for these two samples are 5.1 and 5.7 kPa/day when normalized to 1 wt % added water.

Analysis of the gas constituents for the small-scale containers with impure PuO₂ salt at two weeks shows that oxygen is depleted in the containers with 0.5 wt % added water or less but that oxygen is generated in the container with 2.0 wt % added water.

The ratio of hydrogen to oxygen decreases from about 8 to about 4 at longer times. At these ratios the gases are flammable.

Gas-generation rates in terms of pressure over time are of interest to regulators and managers of storage facilities because they can readily understand how and if containers pressurize. In radiation studies on hydrogen containing materials the usual unit for providing a rough estimate of gas generation is the G value. The rate of radiolysis (e.g., the G value,) is thus defined as the number of H₂ molecules produced with every 100 electronvolts (eV) of absorbed energy.

G values can be derived from our data using the volume of the containers, the density of the material, the amount of material, and the increase in pressure. There are a number of methods of determining the G value and can include elaborate geometrical models of how the water is distributed on the plutonium dioxide and associated materials. Regardless of the simplifying assumptions used, the values determined in our studies can be compared to G values reported recently in the literature: for pure water GH₂=1, for magnesium chloride hexahydrate exposed to gamma radiation G_H2=0.1, and for pure plutonium dioxide powder exposed to a high-humidity atmosphere G_H2= 1.1 to 8.6.

These empirical observations clearly indicate the high variability in gas generation from pure and impure plutonium-bearing materials. Furthermore, pragmatic implications of these studies also provide guidance on which stored material types need more extensive surveillance. The later aspect is an explicit requirement stated in the DOE 3013-2000 Standard.

This article was contributed by D. Kirk Veirs, Nuclear Materials Technology Division. Thanks go to Dennis Padilla, Ed Wilson, Rebecca Maez, and Beverly Bender for preparing the material; and Alex Carrillo, Max Martinez, Adam Montoya, Dennis Padilla, and Dallas Hill for maintaining and operating the capability in PF-4. Special thanks go to John Berg for providing expertise in Raman spectroscopy; to David Harradine and Rhonda McInroy for providing expertise in data-acquisition program development and gas chromatography analysis; to Max Martinez, Alex Carrillo, and Adam Montoya for keeping us safe; and to Laura Worl for filling the project- and line-management roles.

NMT | LANL | DOE
Phone Book | Search | Help/Info

L O S  A L A M O S  N A T I O N A L   L A B O R A T O R Y
Operated by the University of California for the US Department of Energy