In Part A, the geotechnical background is presented to a stability problem regarding the North Spur dam wall at Muskrat Falls in Churchill River in Newfoundland, Canada. This land was formed in the regression of the sea during and after the last ice age with deposits of multiple layers of silty sands and silty sandy clays that formed the valleys and plains that are now above sea level. Some of these layers, deposited thousands of years ago in post-glacial times, are vulnerable to liquefaction when they are disturbed. These conditions have in the past repeatedly caused slides along the banks of the Churchill river. In the report, a specific type of possible progressive failure –the most dangerous one in respect of the safety of the North Spur – is discussed. This type of landslide development may be caused by the rising water pressure, when - or after - the dam is impounded. As will be explained, such a slide could force part of the North Spur ridge to slide along a failure surface sloping East-wards into the deep river whirlpool downstream of Muskrat Falls. In the following, a brief overview is provided of the geotechnical background behind our concerns, also discussing methods of mitigating the risk of the kind of slopefailure in question. Hence, we propose measures such as compacting the soil by piling or by methods of grouting and drainage. We also suggest the need for an expert Advisory Panel to look further into the long-term safety of the North Spur.

In Part B, the paper in Part A is discussed and the risk for a failure is denied.

In Part C, finally, the authors of Part A, give their answer to the criticism and uphold their view: There may be a serieois risk for a dam breach caused by weak soil layers in the dam. This risk should be properly investigated and mitigated. 

The paper presents the geotechnical background to one of the stability problems regarding the North Spur dam wall: This land was formed in the regression of the sea during and after the last ice age with deposits of multiple layers of silty sands and silty sandy clays that formed the valleys and plains that are now above sea level. Some of these layers, deposited thousands of years ago in post-glacial times, are vulnerable to liquefaction when they are disturbed. These conditions have in the past repeatedly caused slides along the banks of the Churchill river.

In the current paper, a specific type of possible progressive failure – the most dangerous one in respect of the safety of the North Spur – is discussed. This type of landslide development may be caused by the rising water pressure, when - or after - the dam is impounded. As will be explained, such a slide could force part of the North Spur ridge to slide along a failure surface sloping East-wards into the deep river whirlpool downstream of Muskrat Falls. 

In the following, we provide a brief overview of the geotechnical background behind our concerns, also discussing methods of mitigating the risk of the kind of slope failure in question. Hence, we propose measures such as compacting the soil by piling or by methods of grouting and drainage. We also suggest the need for an expert Advisory Panel to look further into the long-term safety of the North Spur.

The concerns regarding the stability of the North Spur can be summarized in three points:

(1) None of the most critical inclined failure surfaces have been studied by Muskrat Falls Corporation. These failure surfaces may be initiated on the upstream side of the dam containment. Here the effects of the deformations, caused by the pressure of the rising water level, have to be resisted by the metastable soil layers in the North Spur. A local failure may occur progressing downwards towards the downstream side of the Spur. A catastrophic dam breach would follow. The GPRP further categorically overlooks the fact that horizontal failure planes cannot possibly represent the highest risk of instability irrespective of whether the analysis is based on the Limit Equilibrium Mode (LEM) or on the Progressive Failure Mode.

(2) The stress/strain deformation properties of the porous soils in the North Spur have not been made available. Only strength properties, related to fully drained conditions, have been given. How stresses relate to simultaneous deformations under undrained (or partially undrained) conditions have not been defined in any way. Such relationships are crucially essential for any up-to-date analysis of slope stability.

(3) A high risk of North Spur instability has been found related to impoundment.  A series of investigatory calculations have been made, based on deformation properties from similar landslides and on a wide variety of assumed input data for possible critical failure surfaces. The results of these analyses indicated a safety factor far below 1.

The peer review does not address the above three points. It gives a good view of the general conditions but also contains misconceptions, erroneous considerations and refutable comments indicating that the earlier reports by Bernander have not been fully understood by the panel members.

As no up-to-date analysis of the stability of the North Spur has been provided, our conclusion is that an independent group of experts, appointed by government, should be entrusted with this important task.

The poster presents the risks for a progressive landslide in a natural dam. The stability will be critical when the water level is raised after the building of a hydro power plant, Bernander (2016), Dury (2017). The analysis is based on a finite difference method developed by Stig Bernander (2011), Bernander et al.(2016)

The following issues will be discussed:  

- Material properties

- Risk for liquefaction

- Three possible failure surfaces: one horizontal, one inclined and one curved

- Failure riska for different material propeties

- The need to check the real properties of the soil

The poster presents a new Spreadsheet developed by Robin Dury (2017) to simplify the use of the Finite Difference Method developed by Stig Bernander et al (2011, 2016).

It includes:

- Material Properties

- Finite Difference Method

- Progressive failure process with five phses

- Discussion

- References

The differences are outlined in landslide analysis between the classic limit equilibrium method with assumed plastic properties of the soil and a progressive analysis applying softening material properties.

The risk for failure is studied in the dam at the North Spur riverbank ridge at Muskrat Falls in Churchill River in Labrador, Newfoundland, Canada. A sloping failure surface is much more critical than the horizontal surfaces which have hitherto been studied. Results from new analyses have now been obtained applying softening material properties probable for the ridge. The results indicate safety factors lower than 0.5, i.e. there is a high risk that the ridge will fail if the water level is raised to the proposed level.

Three reports are appended where Stig Bernander argues in detail for the need for a proper progressive failure analysis based on measured material properties. He also proposes how such properties may be obtained and gives an example of a way to stabilize the ridge if the soil properties show a softening behaviour. Finally examples of progressive failure analyses are included using probable material properties.

A large landslide in Tuve (Gothenburg, Sweden 1977) initiated the development of a model for slope stability analysis taking the deformation-softening of soft sensitive clays into consideration. The model studies triggering agents and five phases in progressive slope failure are identified: (1) in-situ, (2) disturbance, (3) unstable ‘dynamic’, (4) transitory (or permanent) equilibrium, and (5) ‘global’ failure. The clay resistance in these phases may differ widely; mostly due to different rates of loading. Two time dependent failure criteria are defined: (i) the triggering load condition in the disturbance Phase (2), and (ii) the transitory equilibrium in Phase (4), indicating whether minor downhill displacements or a veritable landslide catastrophe will occur. The analysis explains why downhill landslides tend to spread over vast areas of almost horizontal ground further down-slope. The model has been applied to landslides in Scandinavia and Canada. Three case studies are briefly discussed. The model is a finite difference approach, where local downhill deformations caused by normal forces is maintained compatible with deviatory shear deformations above the potential (or the established) failure surface. Software and an easy-to-use spreadsheet are introduced as well as recent developments. See also Video Abstract.

The presence of sensitive clay pose a challenge when performing slope stability assessments. Because of the strain softening behavior, the validity of conventional calculation methods based on the principle of limit equilibrium (LE) are not fully valid. This paper studies downward progressive failure in long natural slopes with an aim of identifying the governing parameters and the validity range of LE methods. A FEM approach which accounts for the non-linear stress-strain curve of the material, including the post peak softening behavior, is used. Sensitivity of the analyses to variations of key parameters like in-situ shear stress at the failure plane, brittleness, stiffness of the soil mass, and geometry are investigated in terms of critical load for initiating the slide and the corresponding critical length. The results show that the capacity of the slope in terms of external actions is reduced the steeper the slope is, the more strain softening behavior the material display and the lower stiffness the overlying soil has. The initial shear stress level is identified as a highly sensitive parameter. Further, by studying variations in the critical length it is indicated that the validity of classical LE methods is limited for steep slopes in soft and very sensitive clay.

Observations from past events are used to show that the concept of progressive failure may explain translational progressive landslides and spreads — large landslides occurring in sensitive clays. During progressive failure, the strain-softening behaviour of the soil causes unstable forces to propagate a failure surface further in the slope. Translational progressive landslides generally take place in long, gently inclined slopes. Instability in a steeper upslope area is followed by redistribution of stress, which increases earth pressure further downslope. Passive failure may therefore occur in less-inclined ground, heaving the soil. Spreads are usually trigged by erosion of a deposit having a higher angle near the toe. Instability starts near the toe of the slope and propagates into the deposit, reducing earth pressure. This may lead to the formation of an active failure with dislocation of the deposit into horsts and grabens. The failure mechanism of both types of landslides is controlled by the stresses in the slope and the stress–strain behaviour of the soil. The mechanism presented explains the sensitivity of a slope to minor disturbances and the resulting high retrogressions observed for such landslides in Scandinavia and eastern Canada.

Des observations d’évènements passés sont utilisées afin de démontrer que la rupture progressive peut expliquer les glissements translationnels progressifs et les étalements — de grands glissements survenant dans les argiles sensibles. Lors d’une rupture progressive, le comportement anti-écrouissage du sol génère une force instable propageant la surface de rupture dans la pente. Les glissements translationnels progressifs surviennent généralement dans de longues pentes, légèrement inclinées. Une instabilité, générée dans le haut de la pente, est suivie par une redistribution des contraintes, augmentant la pression des terres plus bas dans la pente. Une rupture passive peut donc survenir dans la partie moins inclinée du dépôt. Les étalements surviennent dans des pentes ayant de plus fortes inclinaisons à leur base. L’instabilité est déclenchée près du bas de la pente, généralement par érosion, et réduit la pression des terres, menant à la dislocation active du sol en horsts et grabens. Le mécanisme de rupture de ces deux types de glissement est contrôlé par l’état des contraintes dans la pente et la relation contrainte-déformation du sol. Ce mécanisme de rupture explique la sensibilité d’un talus à une perturbation mineure et les grandes rétrogressions observées dans plusieurs cas Scandinaves et dans l’est du Canada.