by David Albright
August 1, 1994

The irradiated or spent "magnox" fuel rods discharged from North Korea's 5 megawatt (electric) reactor cannot be stored safely for long under the current conditions in the ponds near the reactor. The magnesium metal alloy jacket or "cladding" around the uranium metal fuel is corroding. If the corrosion creates holes in the cladding, radionuclides in the uranium fuel will escape into the environment. In addition, corrosion of the underlying uranium metal fuel will commence. Under certain conditions, the uranium metal can ignite spontaneously if exposed to air, possible causing a serious radiation accident.

North Korea has done little to slow the corrosion of the fuel cladding, or even to clean the water sufficiently to allow visual inspection of the fuel to check for cladding corrosion. If steps are not taken immediately to clean up the pool and extend the storage life of the fuel, North Korea might argue that this fuel must be reprocessing for safety reasons. Since the unloading of the spent fuel from the ponds must begin in time to remove the last of the fuel before these rods significantly corrode, the North might decide to begin the fuel transfer soon, possibly as early as August.

The movement of spent fuel to the Radiochemical Laboratory for plutonium separation would violate North Korea's June pledge to freeze its nuclear program. Thus, the rate of corrosion of the fuel could put the U.S./North Korean negotiations, set to resume August 5, under enormous time pressures unless the fuel's storage life is extended.

To allow more time for the negotiations to succeed, the United States and other Western governments have offered the North help in slowing the corrosion of the cladding. These techniques could extend the length of wet storage of this fuel by months or even a few years, but they must be instituted soon and require North Korean cooperation to succeed.

Under optimal water storage conditions, Western nations have found that the type of fuel used by North Korea can be stored in spent fuel ponds for up to two years. Since the North's fuel has already been in the ponds since May or June, such long storage times might not be possible with this fuel even if instituted immediately. But there is still time to extend fuel storage in the ponds up to a year. Storage up to five years is possible with the use of special canisters placed in the spent fuel ponds.

Besides the spent fuel ponds, North Korea has a small dry storage area for damaged fuel elements, but this area is also not maintained adequately to prevent continued corrosion and the possibility of a serious fire of the fuel. If North Korea built acceptable dry storage facilities, its spent fuel could be maintained for a decade or more without significant corrosion or risk of fire.

Although storage times can be greatly expanded, the fuel will likely need eventual reprocessing either in North Korea or abroad. Although the fuel currently in the ponds could be dried off in special facilities and placed in dry stores, this technique is complicated, unproven, and time-consuming to carry out. This approach might also be beyond North Korea's capabilities without extensive Western assistance.


In May and June, the North discharged about 8,000 spent fuel rods (or about 50 metric tons of fuel) from the 5 megawatt (electric) reactor. The fuel rod itself is a cylinder of natural uranium metal encased in a magnesium alloy cladding that contains about 0.55% zirconium. Each rod is about 50 centimeters in length and about 2.9 centimeters in diameter (including the cladding). The rod is surrounded by fins, bracers, and other structures necessary to support the rod in the reactor fuel channel.

Upon discharge, the reactor operators placed about forty rods in a basket. The basket was put in a concrete cask and transferred by trolley from the reactor hall through a 50-meter tunnel to the spent fuel storage building. This building contains an inspection bay, two connected spent fuel ponds, and a dry storage area.

In the inspection or decontamination bay, operators inspected the fuel rods in each basket for any damage. If the rods appeared intact, a crane placed the basket in one of two connected spent fuel ponds, about 7.5 meters deep. Because few rods had failed, nearly all the rods were placed in the spent fuel ponds in about 200 baskets stacked up to three high.

If a fuel rod in a basket appeared damaged, the entire basket was sent to the dry storage area. In total, about 300 rods were placed in dry storage during this procedure. There, they joined another 300 damaged rods unloaded from the reactor between 1989 and the time of the most recent fuel unloading.

Since visual inspection procedures are not believed capable of detecting all damaged fuel, some fuel sent to the spent fuel pond could be damaged, and vulnerable to accelerated corrosion.

Spent Fuel Ponds

The current conditions in the spent fuel ponds do not bode well for long term storage. The North has implemented little Western technology to slow corrosion of the cladding.

North Korea appears to operate with the assumption that it sends the fuel for reprocessing after waiting a few months for the radioactive decay of short-lived elements, such as iodine-131, but well before significant corrosion might occur. According to Selig Harrison at the Carnegie Endowment for International Peace, North Korean officials told him in early June, 1994 that depending on the condition of the rods, some rods would need to be moved after two months, or in August. He said that North Korean officials also told him that complete unloading of the ponds would take about two to three months. (This last statement implies that the North expects to take about this length of time to reprocess all this fuel.)

The actual rate of corrosion of the fuel in the ponds is unknown. The IAEA has reported that, based on its inspectors' observations, the water is opaque and dirty, and the ponds do not have adequate filtering or purification systems. A video of North Korean nuclear facilities made public by the IAEA in 1992 shows the spent fuel ponds green with algae.

More important, the IAEA has said that the North does not conduct detailed water chemistry analyses of the water in its ponds. It also does not control the level of certain impurities that can greatly accelerate cladding corrosion, particularly chlorides and sulfates. If the water contains more than certain concentrations of these impurities, the cladding might corrode significantly within a few months. To reduce the level of these impurities, Western magnox spent fuel ponds use demineralized water.

Dry Storage Area

The nearby dry storage area is also inadequate, especially since it holds damaged fuel elements. An IAEA official described it as a "moist dry storage area," because of its proximity to the spent fuel ponds. The moisture in the air can corrode both the cladding and the uranium metal fuel itself. Sometimes, the corrosion of the uranium metal can lead to the spontaneous combustion of the fuel at ambient temperatures, particularly if anyone disturbs the fuel.


Both the magnesium and uranium metal are vulnerable to corrosion when stored in water, moist air, or inert gases if moisture is present.1 The primary reaction for either the cladding or the uranium is oxidation, involving water and the subsequent release of hydrogen. Excess hydrogen can react with bare uranium metal, forming uranium hydride, a brownish-black or brownish-gray powder that burns spontaneously in air at room temperature.

The first risk from corrosion in the spent fuel ponds is that the cladding will fail and large quantities of radioactive materials will leak into the pools and eventually get into the air or leak into nearby groundwater, posing a risk to workers and the surrounding population. Once the cladding fails, the uranium metal can corrode, possibly resulting in the formation of uranium hydride. If the uranium hydride combusts it can ignite the uranium metal, leading to a serious radiation accident.

Cladding Corrosion

There are two basic types of corrosion of the magnesium alloy cladding--general superficial corrosion and localized or pitting corrosion. In the case of North Korea, localized corrosion is the more serious concern, because the possible presence of impurities in the water could significantly accelerate the rate of this type of corrosion compared to general corrosion rates.

General corrosion depends on the temperature and the pH levels of the water in the pool. Ideally, the water should be chilled to about 15 degrees centigrade and kept alkaline with a pH level of 11. 5 to 12. In contrast, the North's pond water reportedly is at about 30 degrees centigrade and has a pH of about 11. Once corrosion begins, the magnox corrosion product sludge can also markedly increase cladding corrosion.

Maintaining a high pH level requires the continual addition of hydroxide and the periodic removal of carbonate ions. These ions lower the pH and result from the dissolving of carbon dioxide from the atmosphere in the water.

Pitting or localized corrosion begins in cracks, crevices, or scratches in the cladding. Poor handling of the spent fuel during discharge can produce many sites where pitting can occur. Pitting corrosion is accelerated in the presence of sulfate and chlorine ions, and carbon depositions on the fuel.
Once corrosion creates holes in the cladding, soluble radioactive elements will leach out and diffuse throughout the spent fuel pond and migrate to the surface. The most immediate concern is the cesium 137 that emits penetrating gamma rays that pose a risk to workers near the ponds and any IAEA officials inspecting the pond. In addition, corroded fuel requires more careful handling to prevent the release of more radioactive material.

Uranium Corrosion and Hydride Formation

Once exposed, uranium metal will react with water to form uranium oxide and hydrogen. Although a uranium oxide layer could be expected to seal the metal and prevent further corrosion, the main type of oxide that forms on uranium metal, namely uranium dioxide, does not act as a seal. As the oxide layer builds up, it tends to crack exposing bare metal to further corrosion.

Hydrogen from the oxidation process or elsewhere will react with bare metal to form uranium hydride. Often, the hydride powder is coated with an oxide layer that can prevent its spontaneous combustion. But if the oxide layer is disturbed, such as can occur during fuel handling, the hydride will glow and spark, and possibly ignite the surrounding uranium metal, leading to the release of the radioactivity in the spent fuel element.

Conditions that favor high uranium hydride concentrations are high relative humidity, lack of free oxygen, and high hydrogen concentrations. These conditions can exist inside a failed fuel rod in a spent fuel pond, where water may penetrate through a small hole in the cladding and the hydrogen resulting from the oxidation of the uranium cannot diffuse out of the cladding effectively.


Given the condition of the North Korean storage facility, the priority is ensuring that the spent fuel is safely stored in the spent fuel ponds and the dry storage area for many months and perhaps years. Methods to accomplish this goal have been proven in France and Britain, who have extensive experience storing their own magnesium clad fuel.

Spent Fuel Ponds

Controlling the corrosion of the cladding requires rigorous and continual control of the pond water conditions. Even relatively small variations from the recommended conditions can significantly accelerate corrosion. Fortunately, the implementation of the appropriate controls can be accomplished quickly:

Improving the Water Chemistry. The fuel storage life in the ponds can be extended for several months by purifying the water, raising the pH level, and lowering the concentrations of particular ions, particularly chlorides and sulfates in the water. Since much of the fuel has already been stored for over two months, these steps need to be carried out as soon as possible if they are to be effective.

The following steps can extend the fuel storage life many months, perhaps for a year:

o Use only demineralized water in the pond to improve the visibility of the water and reduce impurity concentrations in the water. Chlorine ion concentrations should be below 0.5 grams per cubic meter, sulfate concentrations below 0.2 grams per cubic meter, and carbonate ions below 3 kilograms per cubic meter. The demineralized water should continuously be replaced (at least 10% per day) to maintain water purity and keep the deleterious ions below specified concentrations.

o Raise the pH levels to at least 11.5 and maintain them at this level by adding pure sodium hydroxide ions (NaOH);

o Regularly measure and record the pH level, chlorine and sulfate ion concentrations, and other impurities in the water;

o Reduce the water temperature to 15 degrees centigrade by portable water chillers to improve water clarity, retard algae growth, and to reduce roughly fourfold the rate of magnesium corrosion occurring at current pond temperatures.
o If possible, remove magnox corrosion product sludge, which settles to the bottom of the pool, that may have already formed.

Special Magnox Fuel Canisters. The corrosion rate can be reduced further, permitting wet storage up to five years without significant further cladding corrosion, by raising the pH level around the fuel to 13, while maintaining total chlorine and sulphate levels below 0.5 parts per million. This high of a pH cannot be achieved in the ponds because the pH level is lowered too much by dissolved carbon dioxide from the atmosphere.

However, Britain has successfully developed the Magnox Fuel Canister, which when loaded into a spent fuel pond, that can maintain a pH of 13 around the fuel in the canister. The canister works by isolating the water in the canister from the pond water by a gas space under the canister's loose fitting lid. The lid must be loose-fitting to ensure that hydrogen produced during any residual corrosion of the cladding or uranium does not build up to unacceptable levels.

The gas space above the water in the canister is called "ullage." Initially, nitrogen is injected into the space, which is replaced over time by hydrogen from the small continued corrosion of the fuel cladding. Oxygen from the water also gradually diffuses into the gas space, some of which combines with the hydrogen. The British have found that the gas space can be maintained for up to five years before it must be replenished.

Each British canister has a mass of about one metric ton, and can hold several hundred fuel rods; the exact number depends on the rod's outer structures. These canisters are stacked up to three high in a spent fuel pond. Drawbacks are that they have an elaborate design and require special handling equipment to establish the ullage and place the container in the spent fuel ponds.

Radioactive Clean-Up. Whatever the options developed, the West should supply North Korea portable equipment to clean up radioactive contamination, particularly radioactive cesium, from the pond water. Such a step would reduce exposures to workers and inspectors, lower the chance of environmental contamination, and prevent large-scale contamination of the spent fuel pond and associated equipment.

Dry Storage Area

Although this area contains a relatively small number of fuel elements, it poses a significant fire hazard from corroded or damaged fuel. Some alternatives include drying the air or flooding the area and treating it as another spent fuel pond with appropriately treated water.

Meanwhile, the mechanical handling of the fuel in this area should be kept to a minimum and conducted with utmost care.


The main long option for disposing the spent fuel is reprocessing the spent fuel in North Korea or abroad. The latter is the most acceptable to international interests.

Another option is to transfer the intact fuel to dry storage after drying off the fuel stored in water. This latter option, however, has never been proven.

Transportation to an Overseas Reprocessor

U.S. and British officials have said that the fuel could be removed from North Korea in six to nine months. Transportation casks for this type of fuel are plentiful and proven. Many are also located nearby. Japan, for example, has routinely transported a similar type of fuel from its Tokai gas-graphite reactor to a British reprocessing plant for many years.

The British cask weighs 60 metric tons, and contains water with a pH greater than 11.5. The equivalent French cask weighs about 50 metric tons. These casks can hold up to about 5 metric tons of undamaged fuel, although they might not be able to hold this amount of North Korean fuel. The list of potential reprocessors includes Britain, France, Russia, China, and the United States.

Reportedly, the crane over the storage pools at the Yongbyon spent fuel storage ponds can hold only up to 35 metric tons. Either a new crane would have to be supplied, or alternate, lighter casks secured.

Prior to shipment, the fuel would need to be inspected to determine if it was damaged, since a cask should not contain too many damaged fuel elements that could overly contaminate the cask water. In any case, special precautions need to be taken to ensure that any damaged fuel did not come into contact with air.

Going from Wet to Dry Storage

Another option is to take the wet spent fuel, dry it off, and place it in a dry air storage facility. The British have proven that storage of magnesium clad fuel in dry air is safe and delays the need for reprocessing for many years, and perhaps decades.2 However, this approach has been applied only to fuel that has never been wet.

Because of the potential attractiveness of dry storage, the British have studied the process of drying fuel stored in wet spent fuel ponds and placing it in dry stores since 1978. However, a British nuclear industry study of this route for British magnox fuel concluded that going from wet-to-dry storage is feasible, although expensive and difficult.3 This study recommended this procedure as an alternative to reprocessing only if absolutely necessary.

This study envisioned few difficulties in drying off fuel with intact cladding and transferring it to dry storage. But fuel with defective cladding is harder and riskier to dry, because water leaks through small holes in the cladding and uranium hydride forms.

Because of the likelihood of missing some damaged fuel rods during fuel inspections, this study concluded that each element would need to be treated in the same way.4 The study envisioned that each element would be individually dried at elevated temperatures in a tightly fitting drying tube in a specially designed facility. The heated air would also oxidize uranium hydride on the fuel, a necessary step for extended safe storage. Air might not reach all the uranium hydride underneath the cladding since hydrogen can migrate a significant distance from where it was formed before forming uranium hydride. As a result, the fuel would require careful handling to avoid breaking the fuel and exposing any hydride to air.

If uranium hydride inadvertently ignited the fuel during the drying process, the fire would occur in an isolated and contained environment where the amount of air could be restricted until the fire was extinguished.


Corrosion of the North's spent fuel is probably already occurring. Although corrosion needs to be avoided or slowed, it cannot be stopped. Western utilities and governments have extensive experience handling corroded fuel. This expertise can reduce the chance of extensive contamination or a serious radiation accident in North Korea. But the North Korean government must cooperate. If it does so, the fuel can remain stored for many months or a few years without the need to move it.


1. For a review of magnesium cladding corrosion problems see Appendix 2: The Storage of Magnox Fuel--CEGB Experience and Research, in The CEGB/SSEB Response to Recommendation 17 in the Environment Committee's Report on Radioactive Waste, Volume 2: Appendices, November 1986. See also chapter 7 on Uranium in C. R. Tipton (eds) Reactor Handbook, Volume I: Materials (New York: Interscience Publishers, Inc., 1960).

2. A. H. Speller, E. O. Maxwell, and R. J. Pearce, "The Long-Term Dry Storage of Irradiated Magnox Fuel," in Proceedings of BNES Conference on Gas-Cooled Reactors Today, held in Bristol, Britain, September 1982, volume 4, pp. 25-30. Air is preferred instead of inert gases because corrosion rates of uranium and magnesium in air are acceptably low, and moist air is a more benign medium in regard to uranium hydride suppression than moist nitrogen or argon, two common inert gases. The oxygen in the air inhibits the formation of uranium hydrides.

3. See for example, The CEGB/SSEB Response to Recommendation 17 in the Environment Committee's Report on Radioactive Waste, Volume 1, November 1986. The Environment Committee of the British Parliament asked the nuclear electricity utilities to study the feasibility of using only dry stores for Magnox fuel prior to geological disposal with no reprocessing. This assessment included a study of the process of drying off Magnox fuel once it had been stored in a cooling pond.

4. Ibid, p. 17; and National Nuclear Corporation, Dry Storage of Magnox Fuel: A Design Concept Document, C6996/DCD/001 Issue B, September 1986.