Re-examine the design?

The US Nuclear Waste Technical Review Board (NWTRB) expressed its reservations about the corrosion resistance of waste packages under repository conditions in a letter to DoE in October 2003 and raised the issue again in June 2004 in its review of the previous year’s work. In its executive summary the NWTRB observed that:

• Localised corrosion processes are insidious because initiation is difficult to predict and propagation rates can be very rapid. Data emerging from DoE’s Yucca Mountain project and from the Nuclear Regulatory Commission’s Center for Nuclear Waste Regulatory Analyses (CNWRA) suggests to the board that crevice corrosion of Alloy 22 is likely to begin during the thermal pulse (the first thousand years after repository closure).

• Project data shows that initiation of crevice corrosion during the thermal pulse is likely in concentrated brines (with or without nitrates) formed through deliquescence processes at temperatures well below the peak temperatures on the waste package surface expected in DoE’s proposed repository design.

• Crevice corrosion initiated during the thermal pulse is likely to propagate during the remainder of the thermal pulse (and afterwards).

• Welds and thermal treatment (ageing) increase susceptibility to crevice corrosion. The DoE’s modified waste package design has both welded areas (closure welds) and many features that offer opportunities for crevice formation.

• Most generalised corrosion data reported to date is for surface temperatures on the waste package of 95ºC or lower. This data may constitute an adequate technical basis if the surface temperatures of the waste packages in the repository never exceed 95ºC. Little data exists, however, for the higher temperatures of the thermal pulse.

Because of the seriousness of these corrosion concerns, the NWTRB strongly urges DoE to re-examine the current repository design and proposed operation. The board believes that the high temperatures of the current design and operation will result in perforation of the waste packages, with possible release of radionuclides. The data currently available to the NWTRB indicates that perforation is unlikely if waste package surface temperatures are kept below 95ºC. Finally, the NWTRB believes that total system performance assessment should not be used to dismiss these corrosion concerns.

FURTHER DISCUSSIONS

The question of corrosion was the main topic under discussion at the NWTRB’s most recent meeting in May.

Various speakers discussed the issue and detailed further research being carried out on behalf of DoE.

It was noted, for example, that several hundred open circuit potential tests and cyclic polarisation tests, had been completed, most since the NWTRB last reported. The tests were providing data on the likely sodium chloride, sodium nitrate, calcium nitrate type conditions and examined where nitrate and sulfate effects were beneficial.

Some data had been collected at higher temperatures – up to 220°C. The high temperature data was still being collected.

Some speakers at the meeting agreed there was a potential for a chemistry that could lead to low nitrate/chloride ratios during one period after the closure.

How much water would be around? The mean rates for seepage in a monsoonal or glacial transition climates would be several litres, but this climate was associated with time periods beyond 600 years. Before then just a tenth of a tenth of that water ingress was possible and that water had to concentrate and evaporate. That would require tens or hundreds of millilitres of water per year per package to produce the high concentrations.

Depending on the temperature, the conditions to initiate localised corrosion could occur if the salt solutions contact the waste package and they concentrate. But “you have to have all those ifs lined up,” one speaker said.

If corrosion was initiated, would it propagate? Speakers said there was ‘some uncertainty’. It was also not clear whether there was a strict temperature boundary that would affect the potential for corrosion. Additional research was required.