Unlikelihood of Localized Corrosion of Nuclear Waste Packages Arising from Deliquescent Brine Formation, The

JOM, Jan 2005 by Apted, M, King, F, Langmuir, D, Arthur, R, Kessler, J

The Nuclear Waste Technical Review Board (NWTRB) recently postulated a scenario for the formation of a deliquescent, divalent-cation (Ca, Mg) chloride brine from wind-blown dust on the surface of an alloy 22 container designed to hold radioactive waste. The NWTRB suggested that such brines could lead to localized penetrations of waste packages at a proposed repository in Yucca Mountain, Nevada. In response, the Electric Power Research Institute (EPRI) sponsored an independent analysis, specifically examining and responding to a series of decision points in the NWTRB's scenario. Only if all of these questions could be answered affirmatively would NWTRB's scenario result in potential noncompliance with regulatory criteria. The EPRI analysis found that none of the questions has an affirmative answer.

INTRODUCTION

The U.S. Department of Energy (DOE) is developing aconceptual design and license application for the disposal of civilian spent nuclear fuel and defense high-level radioactive waste at a proposed geological repository within Yucca Mountain, Nevada. The nuclear waste is to be encapsulated into a robust waste package (Figure 1) composed of an inner vessel of 316NG stainless steel and an outer vessel of alloy 22. The loaded waste packages are to be emplaced horizontally in open tunnels within the mountain. The tunnels will be 5.5 m in diameter, located within the hydrologically unsaturated (i.e., 300 m above the water table) tuff formations approximately 500 m below the ground surface at Yucca Mountain. Within the emplacement tunnels, waste packages are to be overlain by a titanium drip shield. The repository system also includes gas phases in the tunnels, other engineered materials (concrete, steel, backfill, etc.), the surrounding wall rock (tuff), pore waters in tunnel walls, and any dust on waste package surfaces that contains soluble salts.

As part of the strategy for safe, longterm isolation of the waste, the heat from radioactive decay in waste packages initially raises the temperature within the emplacement tunnels above the boiling point of water. This keeps the relative humidity (RH) correspondingly low within the drift and prevents any water contact with the waste packages during this thermal "dry-out" period for approximately the first 1,000 years after permanent closure of the repository (Figure 2). Furthermore, the temperature gradient (the waste package surface remains slightly hotter than the tunnel walls) causes a reciprocal RH gradient that persists for many thousands of years. The lower humidity on the waste package surfaces acts to extend the containment time by the alloy 22 container, thereby substantially delaying any possible release of radionuclides via a groundwater pathway.1

The Nuclear Waste Technical Review Board (NWTRB) recently postulated2 a scenario for the formation of a deliquescent, divalent-cation (Ca, Mg) chloride brine from wind-blown dust on the surface of the alloy 22 container. The NWTRB further speculated that such brines could lead to localized penetrations of waste packages within the first several hundred years at Yucca Mountain, and thereby compromise compliance with regulatory criteria.

In response to the NWTRB speculative scenario, the Electric Power Research Institute (EPRI) sponsored an independent analysis,3 specifically examining and responding to a series of decision points posed as questions in the NWTRB's scenario. If, and only if, all of these questions could be answered affirmatively would NWTRB 's scenario result in potential non-compliance with regulatory criteria. The following is a brief summary of the analyses reported in EPRI.3 This analysis shows that none of the questions has an affirmative answer, and that significant localized corrosion of waste packages in the repository is highly unlikely.

Can Pure Calcium-Magnesium Chloride Brines Form from Wind-Blown Dust?

The most corrosive brines are those composed of calcium and magnesium chloride. During the thermal dry-out period, the only source of chloride salts on waste packages is dust. The NWTRB's concern was whether salts in this dust might absorb moisture from air (deliquesce) in the repository tunnels even at temperatures well above the boiling point of water (-150-160�C), resulting in a corrosive brine.

The composition of dusts found in the exploratory studies facility (ESF), which will become part of the repository, has been reported by Peterman et al.4 More important over the long term, it is anticipated that wind-blown dust will enter the repository and settle on the emplaced waste packages. Windblown dust has been studied by Reheis et al. of the U.S. Geological Survey.5 Key data for both dusts are their composition and weight percentage of specific soluble salts, and of minerals such as feldspars, clays, micas, and carbonates that will neutralize any acidity from the brines. The EPRI has analyzed the detailed mineralogy of these dusts. The predominant solids in both dusts (>90%) are aluminum-silicates (feldspar and clay), silicates, and carbonate minerals. Wind-blown dust has

If a Chloride Brine Forms, Will It be Stable and Persist?

If calcium-magnesium chloride brines do form, it is likely that they will also dissolve co-existing nitrate and sulfate salts in dust, creating a less corrosive, mixed-anion brine. Further, geochemical modeling3 indicates that any magnesium in such a brine will be removed by reaction with silicates in the dust.

Geochemical modeling3 shows that at 146�C, corrosive calcium chloride brines are thermodynamically stable only in a closed system, isolated from the tunnel atmosphere, a highly unlikely occurrence. The breakdown of calcium chloride and formation of calcium carbonate in the actual open system of the tunnel produces a gradient of decreasing HCl gas pressures from the waste packages toward the tunnel walls. Because the waste is always hotter than the tunnel walls, gaseous HCl will continue to migrate in the thermal and concentration gradients toward the cooler tunnel walls, which mighthave atemperature of 96�C, for example. This process is enhanced by the fact that the reactive surface area of the waste packages is perhaps less than 1 % of the reactive surface area of the tunnel walls. Modeling shows3 that at the tunnel walls the HCl will react with pore waters and tuff minerals and be neutralized, forming clays and sodium chloride, with the pH stabilizing indefinitely at near-neutral values. It is concluded, therefore, that even in the unlikely event of initial brine formation, the brine would not be stable and would not persist.