Static liquefaction and strength loss of tailings dams due to undrained failure has become a topic of interest in tailings management following the Fundão and Mount Polley tailings dam failures.

Static liquefaction is the sudden loss of strength when loose soil, typically granular material such as sand or silty sand, is loaded and cannot drain. Strength loss due to undrained failure is also associated with fine grained materials of low hydraulic conductivity, such as clays or plastic silts. The two phenomena are related; loading and deformation produces a tendency for the materials to contract and develop excessive pore pressure faster than drainage systems can relieve the pressure.

Static liquefaction is a rare event but happens very quickly and without warning, so it is an extremely dangerous phenomenon. Although it is called static liquefaction, a triggering event usually causes the rapid strength loss. There are many potential triggers, including:

• vibrations from construction equipment (not strictly static liquefaction);
• rise in water pressure in a slope;
• stress increase due to a dam raise;
• stress concentrations due to a higher dam;
• loss of horizontal confining stress due to lateral strains in the foundation or dam (extrusion-related failure).

Much of the risk depends on the in-situ stress regime, which is difficult to measure and monitor.

Static Liquefaction in Upstream Constructed Dams

A key point in understanding static liquefaction is that sands, including tailings sands, can become less dense relative to the critical state as the in-situ stress increases, for example as a dam is raised. A material consolidated by desiccation may appear to be dense in an exposed tailings beach, however it can behave like loose sand at higher stress levels. Pre-consolidation due to desiccation effects are usually overcome at vertical stresses of 200 kPa to 400 kPa, equal to a depth of about 15 m to 40 m, depending on the water pressure and tailings density.

The problem faced by many mine sites is that upstream constructed tailings dams are often higher than the 15 m to 40 m range (or are planned to go higher). Since the pre-consolidation benefits are being exceeded as the dams are raised, the previous “good” performance of the dam may not be an indicator of its future performance. This is especially true where there is a clay foundation. Such a foundation can induce lateral strains and decrease horizontal stress, which combined with increased vertical stress, promotes collapse and potential static liquefaction.

Undrained failure in foundation clays can cause deformations in the dam, which generates pore pressures induced by the failure process. This phenomenon was first recognised in tailings dams built over soft ground which failed at much flatter slopes than predicted because the design did not consider shear-generated pore pressure increases.

Clays also become more susceptible to this kind of undrained failure as the confining stress increases. The phenomenon is easier to understand for clays since the concept of pre-consolidation stress is well understood. The pre-consolidation pressure is the maximum past pressure that the clay has been subjected to. This can vary from modest stress imposed by desiccation, to very high pre-consolidation stress imposed by kilometer-thick ice sheets. Pre-consolidation stress can be reliably measured in the lab and confirmed by understanding the geologic history of the site.

The issue for tailings dams becomes acute when the loads imposed by the dam exceed the pre-consolidation stress, or the “apparent” pre-consolidation stress for soils desiccated or weathered in-situ. For in-situ weathered soils, such as tropical clays, it is not uncommon for the apparent pre-consolidation stress to be around 200 kPa to 400 kPa or 15 m to 40 m of fill height. Lightly consolidated lacustrine clays in pro-glacial environments often show similar pre-consolidation stress.

It is not possible to obtain samples of undisturbed sands and silty sands by drilling and it is very challenging to get undisturbed samples of fine tailings, therefore in-situ tests are needed to assess liquefaction. The potential for static or seismic liquefaction in sands can be evaluated with the help of seismic cone penetration tests (SCPTu). These tests can estimate the in-situ void ratio relative to the critical state void ratio at a given depth.

For clays, it is possible to get reasonably undisturbed samples. Laboratory tests can be used to get a reasonable estimate of the in-situ pre-consolidation pressure, and the static and post-seismic drained and undrained strength of the clays.

Conclusion

Tailings dam failures in recent years have shined a spotlight on static liquefaction and strength loss caused by undrained failure. KCB’s forensic investigation into both Mount Polley and Fundão have highlighted the role of undrained shear strength in the dam failures, where either clay foundation or tailings can become susceptible to failure as in-situ stress increases.

As dams increase in height, particularly upstream tailings dams, designers need to keep in mind the potential for undrained failure and that the future performance of a dam can change. Care must be taken to not exceed the pre-consolidation or the “apparent” pre-consolidation stress range. The potential strength reduction must also be accounted for in the design.

Upstream constructed dams should be discouraged and extra care is required if they are to be considered. For existing upstream dams, it is best to keep the fill height below 40 m.