free web site hit counter
Search CNIC
Rokkasho Reprocessing Plant

Problems with the Production of Glass Canisters

The Rokkasho Reprocessing Plant, owned and operated by Japan Nuclear Fuel Ltd. (JNFL), is in the final stage of active testing, but problems with vitrification of high-level radioactive waste to form glass canisters have become an obstacle to completing the tests. The problems are fundamental. They will take a long time to solve and I predict that it will be necessary for equipment to be replaced.

As a by-product of reprocessing spent nuclear fuel to extract uranium and plutonium, large quantities of radioactive waste containing many types of radioactive materials is produced. The most dangerous waste is referred to as "high active waste", or "high level waste" (HLW).

It is hard to imagine any way of disposing of this waste other than vitrifying it to form glass canisters and storing it deep underground ("geological storage").

There are two methods of producing glass canisters. One method is to evaporate off the nitric acid by heating the HLW nitric acid solution, mixing HLW oxide powder produced at high temperature with glass, melting this mixture and sealing it in stainless steel canisters. This is the method used in France. The other method, used in Japan, is to mix HLW in liquid form with glass and then heat the mixture to melt the glass.

Borosilicate glass (Na2O-B2O3-SiO2) is used. Some people might think that there should not be any problems, because glass has been manufactured in Japan from ancient times. However, it is by no means a simple process when all sorts of heat producing radioactive isotopes are incorporated into the glass.

The majority of the radioactive substances in the glass canisters are elements produced as a result of nuclear fission reactions ("fissile products"). Besides these, there are also radioactive elements produced by neutron irradiation ("neutron-induced radionuclides"). Table 1 shows the elements contained in 1 ton of spent light water reactor (LWR) fuel (equivalent to the quantity in one glass canister). Elements which are not sent to the vitrification facility are not included in the list.

Table 1: Weight and radioactivity of elements contained in 1 ton of spent LWR fuel
(Burn-up 33,000 MWd/t, cooled for 10 years after removal from LWR)

Weight (kg)
Radioactivity (Bq x 1012)
Strontium (38Sr)
Yttrium (39Y)
Zirconium (40Zr)
Molybdenum (42Mo)
2 x 10-5
Technetium (43Tc)
Ruthenium (44Ru)
Rhodium (45Rh)
Palladium (46Pd)
Tellurium (52Te)
Cesium (55Cs)
Barium (56Ba)
Lanthanum* (57La)
Cerium (58Ce)
Praseodymium (59Pr)
Neodymium* (60Nd)
Promethium (61Pm)
Samarium (62Sm)
Europium (63Eu)
Neptunium (93Np)
Americium (95Am)
(* These elements are not radioactive after one year cooling.)

Chemists will find something rather troubling in this table. That is that not all these elements can be incorporated into glass. One would not expect that platinum group elements with high melting points (ruthenium, rhodium, palladium) could be incorporated into glass. Also, molybdenum readily forms the molybdate ion, which behaves as an anion and is therefore difficult to incorporate into glass. This has become a real problem.

Some of the glass produced in the vitrification facility of the Rokkasho Reprocessing Plant did not flow into the stainless steel canisters. It is believed that this was due to the fact that platinum group elements had blocked the outlet. JNFL tried but failed to make a homogeneous molten glass mixture by stirring it with a metal stirring rod1. This is an extremely serious problem. The facility cannot be restarted until this problem is solved.

Also a substance referred to as "yellow phase", which does not dissolve in glass, has been discovered. This might not turn out to be a major problem, but it is certainly not a good thing. It destroys the fundamental assumption that the glass is homogeneous.

Why then were these problems overlooked until now? First, the knowhow that has been accumulated in Japan was not put to good use. The former Power Reactor and Nuclear Fuel Development Corporation (PNC) (now Japan Atomic Energy Agency (JAEA)) constructed and operated the Tokai Reprocessing Facility in Tokai Village, Ibaraki Prefecture. The vitrification technology used at Rokkasho is based on technology developed for the Tokai facility. Many problems arose during the course of the operation of the Tokai facility and it appears that the details of these problems were not adequately communicated. Second, the organizations that designed and constructed the equipment, JNFL and IHI Corporation, bear much of the responsibility. It is said that these organizations did not pay sufficient heed to the advice of PNC. Finally, insufficient tests were carried out. Had tests been repeatedly carried out using model spent fuel during the uranium tests (real spent fuel not used), which preceded the active tests (using real spent fuel), the current problems might have been avoided.

The current situation is very serious. Due to the paucity of information, there are aspects about which it is impossible to judge. However, replacing equipment that has already been contaminated with radioactive material will be no easy matter. Therefore, it can be expected that fixing the problems will take a long time.

By Emeritus Professor Michiaki Furukawa (nuclear chemist and member of CNIC Board of Directors)

Post Script re Stirring Rod
Vitrification tests have been attempted on three occasions: November to December 2007, for 12 hours on July 2, 2008, and October 10~18. Each time they ended in failure. The earlier tests only involved fission-product-containing high-level liquid waste and alkaline liquid waste. In October, for the first time waste liquid containing undissolved scrap was poured into the vitrification kiln. The vitrification process had not worked without the undissolved scrap, so not surprisingly, it did not work when this material was added. A rod was inserted to stir the platinum group elements that had accumulated at the bottom of the kiln, but then the rod could not be removed. A camera was inserted to see what was wrong and it was discovered that the rod was bent 90o (see diagram). Of course, there is no chance that active tests will be completed in February 2009 as scheduled.

By Masako Sawai (CNIC)

Return to NIT 128 contents


Citizens' Nuclear Information Center
Akebonobashi Co-op 2F-B, 8-5 Sumiyoshi-cho,
Shinjuku-ku, Tokyo, 162-0065, Japan