Problems continue to plague the high active radioactive waste vitrification facility at Japan Nuclear Fuel Ltd's (JNFL) Rokkasho Reprocessing Plant. Sixty glass canisters were produced in just under two months after vitrification tests began in November 2007, but in the latter half of these tests, JNFL experienced difficulties controlling the temperature of the glass melting furnace and the tests had to be stopped. Tests resumed on July 2 this year, after being suspended for six months, but within half a day, without a single canister being produced, equipment problems arose and the facility had to be shut down once again. Stable production of glass canisters is a condition of government acceptance that active testing has been successfully completed. If JNFL is unable to demonstrate stable production of glass canisters, the whole reprocessing enterprise, including acceptance of the active tests and approval of full commercial operation of the plant, will have to be reconsidered.
The designs for the main processes used in the Rokkasho Reprocessing Plant were imported from the reprocessing plant in La Hague, France. However, the design of the vitrification equipment was developed at Japan's own Tokai Reprocessing Facility. The basic concept for making glass canisters is the same. It involves heating glass until it melts, mixing it with high active liquid waste, then cooling it until it vitrifies.
The method used in France, known as "Atelier vitrification de Marcoule" (AVM), involves a two-stage process, using a furnace and a melter. After water and nitric acid in the liquid waste have been evaporated off in the furnace, the waste and the glass are heated together and melted in a melter made of Inconel metal alloy. In the AVM method the life of the melter is around 6 months to a year, meaning that it has to be replaced frequently. Also, due to the fact that it uses induction heating, there is a limit to the extent to which it can be scaled-up.
The method used at Tokai is called "Liquid Fed Ceramic Melter" (LFCM). The melting furnace developed at the Tokai Vitrification Facility (TVF) is made of a fire-proof ceramic material within a metal casing (see diagram). An alternating current, which passes between electrodes fitted into the body of the melting furnace, Joule heats the glass and maintains the temperature at about 1,200oC. In the LFCM method, high active liquid waste and glass ingredients are put directly into the melting furnace, where evaporation of water and nitric acid and melting of the glass occur at the same time.
The reasons given for adopting the Tokai LFCM method at Rokkasho were that it was easy to scale-up, the life of the melting furnace was longer than in the AVM method (about 5 years) and the quantity of radioactive material produced was small. The melting furnace at Rokkasho is 3m x 3m x 3m, five times larger than TVF. However, there were lots of problems with TVF. Difficulties controlling the temperature meant that platinum group elements (palladium, ruthenium, etc.) in the highly active liquid waste accumulated in the melting furnace, blocking the outlet so that the molten glass could not flow freely down into the stainless steel containers where the glass canisters were formed. These structural problems were inherited by the Rokkasho vitrification facility.
The reason why the vitrification tests at Rokkasho were stopped was that the temperature of the melting furnace could not be maintained at 1,200oC. Platinum group elements accumulated at the bottom, causing it to take three times longer than normal for the molten glass to flow down. Two months after the tests were suspended, JNFL submitted a report about the causes of the problem and the measures it had taken in response. JNFL wanted to restart the tests, but the Nuclear, Industrial and Safety Agency ruled that the report was inadequate. It said, "no details are provided about the method of stable operation". JNFL finally provided the requested details in June, six months after the tests were suspended.
The fact is that LFCM's problems with platinum group elements have still not been solved. In the first place, contrary to the claims of JNFL and the government, it is impossible to mix all the high active liquid waste into the glass matrix. Evidently it is difficult to mix platinum group elements into the glass matrix. At both Tokai and Rokkasho they tried to prevent platinum group elements from sedimenting out by reducing the temperature at the bottom of the melting furnace to below the temperature at the top, thus making the glass more viscous. However, they were unable to control the temperature and the outlet kept getting blocked.
Measures responding to sedimentation of platinum group elements and conditions for stable operation announced by JNFL are as follows:
1) adjust the concentration and composition of waste liquid by mixing it with other waste liquid;
2) when signs of sedimentation appear, stop adding radioactive waste liquid and just add glass ingredients instead ("cleaning operation");
3) if that doesn't work, stir with a mixing rod;
4) when all else fails, remove all the waste liquid from the melting furnace.
JNFL's "stable operation" method is not so much a method to overcome the problem of platinum group elements, as an ad hoc survival response.
Having obtained the government's approval for this "stable operation" method, at 12:00 midday on July 2 JNFL recommenced testing of the vitrification facility. When the melting furnace reached 1,200oC, JNFL poured in high active liquid waste. At 21:11 it checked the flow-through of the glass-waste mixture, but the flow stopped immediately. Thinking that the temperature was probably not high enough, JNFL temporarily stopped trying to make the mixture flow through. It raised the setting of the electric heater and restarted the operation, but the molten glass mixture would not flow through. At 23:11 an alarm went off activating an interlock and the test ended in failure.
According to JNFL, temperature control was not the problem, but it was possible that the temperature of the flow-through nozzle that connects the melting furnace to the steel container did not rise, causing the glass to harden. A camera was inserted to check the nozzle and glass-like material was found around the nozzle. JNFL says that in order to find the cause of the problem it will take off the connecting device (about 40cm wide by 1m long) and carry out investigations. However, apparently this flow-through nozzle has not caused problems in the past, so this is a new problem.
There are still eleven batches of molten glass mixture containing one batch of high active liquid waste in the melting furnace. The temperature has been reduced by a few hundred degrees from 1,200oC and is being maintained at this lower temperature, but it is unclear how long this situation will continue. The issue of safety assurance is a big problem. Apparently, just a few hundred grams of molten glass found its way into the glass canister container.
It is certain that at the end of July an announcement will be made that the date for completion of construction and testing of the plant has been postponed for the thirteenth time. The truth is that no one knows when the plant will be completed.
By Masako Sawai (CNIC)