Internal Erosion in Earthdams, Dikes and Levees

This study presents an experimental approach to investigate the impact of suffusion on the soil’s mechanical properties. A newly developed suffusion permeameter is used allowing separate erosion and mechanical tests. The effect of the initial density of the soil on the erosion process is examined. Thereafter, the influence of suffusion on the mechanical properties of the soil is investigated through drained and undrained monotonic triaxial compression tests. The results suggest that the shear strength of eroded soils may either decrease or increase or even may not be affected depending, among others, on the soil’s initial density. In addition, in all cases, slightly more dilative volumetric deformations seem to occur during shearing after erosion. Understanding the mechanical behavior of eroded soils depends on the combined effect of the global void ratio, the inter-granular void ratio and the final fines content after suffusion. Taking into account such a combined effect, an approach to estimate the shear strength of eroded soils is proposed.

An innovative soil reinforcement method developed since 2004 is currently being adapted for hydraulic structures in the frame of the BOREAL project. The objective is to prevent mechanical failure of the soil due to internal erosion or liquefaction. The method is based on microbially induced urea-based carbonate precipitation. Four different erosion test apparatus have been used to assess the methodology: Contact Erosion Tests, Hole Erosion Test, Jet Erosion Test and suffusion tests. All the tests results indicate a strong increase of resistance already with a small percentage of formed calcite (~2–4%). In parallel, tests were performed in a CNR laboratory at a larger scale (8 × 4 × 2.25 m). The first experiment was performed on a homogeneous Fontainebleau Sand. The second and third tests were performed on sandy gravels, which are highly representative of the soil layers that could be treated with this method. These tests validated the effectiveness of the treatment on soils with open porosity, under natural Darcy velocity up to 10−3 m/s. Finally, the fourth test dealt with the treatment of an interface between a coarse and a fine soil layer, underlying the interest of the method for remediation in the case of contact erosion problems.

The Jet Erosion Test consists of impacting a soil sample with a water jet and analyzing the evolution of the scour depth in order. It was originally designed to assess the resistance of fine soils against overflowing erosion by modelling the headcut migration. To extend this research to the overflowing erosion of mixed coarse – fine soils, EDF and geophyConsult have developed a larger JET apparatus, relying on the hypothesis, which needs to be confirmed, that some mixtures of fine and coarse soils may also have a headcut migration-type overflowing erosion process. A first test campaign has confirmed that JET tests carried out on the same soil with the original Hanson bench and the new apparatus provide comparable results. A second test campaign studied the influence of the nozzle diameter, testing the same soil successively with 6.35, 12 and 20 mm nozzle diameter. The first results confirmed that the results are not significantly influenced by this parameter. A third series of tests were related to soils that contain particles above 4.75 mm. Parallel to this experimental research, the interpretation of the test has been re-analysed, trying to understand the reason for the discrepancies observed between results obtained on French soil tests database and Hanson classification.

The WAC Bennett Dam is one of the largest embankment dams in North America, and one of the most instrumented and studied dams in the world. Following the 1996 sinkhole incident, considerable effort has been made to characterise the zoned materials of the dam with reference to susceptibility to internal erosion. Informed by the findings of this experience, we reflect upon the state-of-practice in dam engineering, and discuss some of its inherent limitations with reference to internal erosion, dam safety, and the management of aging embankment dams. We then describe the state-of-the-art for erosion of zoned earthfill by internal instability, a phenomenon whereby seepage flow removes a finer fraction of the soil gradation. Our paper addresses laboratory testing for materials characterization at the University of British Columbia. In collaboration with our industry research partner BC Hydro, we are seeking to advance the state-of-the-art with reference to a mechanics-based understanding of the phenomenon. We report insights gained from (i) a recent study using a flexible-wall permeameter, and (ii) the status of a new investigation using an advanced triaxial permeameter.

The potential for seepage induced internal erosion in a soil depends on a combination of both geometric and hydromechanical criteria, with the complex stress states that the soil is subject to being one of the determinant factors. This preliminary study investigates under small-scale laboratory conditions, the influence of seepage flow under complex stress states on the behaviour of a gap-graded soil including change in volume, permeability, stress-strain response and particle size distribution. The study uses a stress path triaxial permeameter capable of assessing gap-graded soils under complex triaxial stress states undergoing internal erosion via downward seepage. The results obtained from the preliminary analysis of seepage flow testing under constant stress condition indicate that erosion of fines reduces the permeability of gap-graded samples and may also reduce its shear strength. Limitations of this research and future plans are also discussed.

The assessment of the internal stability of a base soil is a fundamental aspect in the design of the filtering transitions. It can be evaluated by means of semi-empirical, theoretical and experimental methods; the assessment of internal stability of a soil using the available semi-empirical criteria can provide different results. In the paper, in order to verify and validate two methods recently proposed by the authors (Moraci et al. 2012a, b; Moraci et al. 2015), the internal stability of granular soils using different methods is evaluated. The internal stability of test soils has been firstly obtained by Kezdi, Sherard and Kenney and Lau criteria (semi-empirical methods), then it has been assessed theoretically by the method “SimulFiltr” and finally it has been verified by long term filtration tests performed in a rigid wall permeameter. The obtained results allowed us to better define the zones that constitute the new graphical method “Butterfly Wings Chart”, recently proposed by the authors.

Suffusion is the process of selective erosion of fine particles within the matrix of coarse soil particles under the effect of seepage flow. Suffusion can induce important modifications in the hydraulic and mechanical characteristics of the soil in different flow directions. Thus, to ensure the safety assessment of hydraulic earth structures, the experimental models need to match the reality of the body of dikes or dams with horizontal flow. In this paper, a new multi-directional flow apparatus is described for characterizing soil sensibility for the suffusion process and to study the effect of flow direction. A series of suffusion tests was performed using this new device with vertical flow or horizontal flow on both gap graded soils and widely graded soils. All tested specimens were subjected to a multi-stage hydraulic gradient. The comparison of the results of the new device with results obtained from another device in vertical flow was realized. The specimens have the same initial hydraulic conductivity and suffusion susceptibility classification with both devices. Furthermore, for specimens characterized by limited soil anisotropy, the suffusion susceptibilities of these soils are quite identical under vertical flow and horizontal flow. These results permit to validate the new device and the experimental method.

One of the local defects in soils due to seepage is heave. This may occur at the downstream toe of hydraulic structures in the case of upward external seepage where particles of uniform soils are subject to uplift and also upward seepage forces. The phenomena starts with gradual liquefaction of soil grains accompanied by gradual loss of shear strength followed by boiling, heave and possible backward erosion and overall collapse of the structure. The problem of heave has been addressed by numerous authors who have published relationships based on the balance of forces and also on results of laboratory experiments. In this paper the results of extensive systematic experimental research carried on glass beads subjected to upward seepage in the vertical Darcy apparatus are presented. Three glass beads’ diameters 0.2 mm, 0.5 mm and 1.0 mm with uniformity coefficient CU from 1.1 to 1.3 and different porosity n from 0.36 to 0.44 were tested. To enable statistical evaluation of the uncertainty in critical hydraulic gradient 177 individual tests have been performed. The results of measurements were analysed and compared with relationships proposed by various authors. The best agreement was provided by the well-known Terzaghi formula relating the critical gradient to the specific mass of grains and soil porosity. Based on the experimental data the uncertainty in the use of Terzaghi formula was expressed via reliability coefficients recommended for use in technical practice.

This paper describes preliminary tests from a ‘transparent soil permeameter’ that has been developed to study the mechanisms that occur during internal erosion in filter materials for embankment dams. The laboratory-based experiments utilise an optical approach where glass particles are used in place of soil, and optically matched oil is used in place of water. The refractive index matching of the fluid and solid enables a two-dimensional “slice” or plane of particles and fluid to be viewed inside the permeameter, away from its walls via a laser sheet and captured by digital camera. The developed set up has already been tested and showed that optically matched glass and oil can behave similarly to soil and water materials as used in previous laboratory testing. In this study we present a flow characterization within a refractive index matched medium made of glass beads. To this end a small amount of fluid tracers is seeded inside the fluid and the velocity field inside the porous media is obtained using PIV measurements.

Dams with core of broadly graded glacial moraines (tills) exhibit signs of internal erosion by suffusion to a larger extent than dams constructed with other types of materials, as reported by Sherard (1979). Garner and Fannin (2010) indicated that internal erosion initiates when an unfavorable combination of soil material, stress conditions and hydraulic load occur. A laboratory program, carried out at Luleå University of Technology (LTU), aims to study the effects of void ratio and hydraulic gradient on the initiation of suffusion of glacial till. It consists of suffusion tests conducted in permeameters with an inner diameter 101.6 mm and a height of 115 mm. Results show, as expected, that the hydraulic conductivity is lower with lower void ratio. Nevertheless, as the hydraulic gradient increases, the hydraulic conductivity reaches steady values. Changes in the hydraulic conductivity suggest variation in the initial void ratio due to detachment of the finer particles from the soil matrix. These fine particles start clogging the lower layers, therefore the rate of water flow decreases and so does the hydraulic conductivity. The hydraulic gradient for which the hydraulic conductivity reaches steady values is considered as the upper limit without suffusion evolved.

In the present study, a numerical solution is proposed in order to quantify the impact of internal erosion on dike stability. The mathematical model, consisting of erosion equations, mixture flow equations and stress equilibrium equations, is solved numerically by the finite element method using COMSOL Multiphysics. The shear strength reduction technique is used to analyze the stability of a dike taking into account the effect of internal erosion. The variation in time and space of porosity as a consequence of internal erosion is chosen as the coupling parameter. Soil stiffness and strength are made dependent on porosity, with the material becoming weaker as porosity increases. The results show that the porosity increases significantly at the dike toe, which was explained by an erosion of this zone. Erosion at the dike toe induces alterations in the mechanical response of the medium. Since the soil strength decreases at increasing porosity, the factor of safety of the downstream slope undergoes significant reduction. This study may help to better understand how internal erosion affects embankments performance, and to better prevent instability of hydraulic structures.

The control of the granulometric stability of a fine-grained material (B) requires a correctly designed protective granular material (T) whose voids, related to the grain size distribution (GSD) and porosity, must be small enough to stop the migrating particles of B within short distances (formation of “natural filter”), and simultaneously allow a safe drainage of B to prevent the occurrence of limit states (e.g. piping, clogging, blinding), inducing in turn uncontrolled increases of interstitial pressure. The available (empirical and analytical) methods to analyze particle migration phenomena at the contact between different materials generally don’t considerer the coupled effects of the involved geometrical and hydraulic variables (e.g. GSD, porosity, volume voids distribution, permeability, piezometric gradients, seepage velocity), as well as their progressive space-time evolution. Thus, a numerical procedure allowing to simulate coupled 1D seepage and particle migration processes, by taking into account both geometrical and hydraulic governing variables, as well as their mutual dependency, has been developed and applied to carry out a detailed analysis and review of some experimental data.

Levee foundations along meandering rivers are often modeled in seepage analyses with simplified models that allow for use of simplified reliability methods. Due to the complex geomorphic environment that is often encountered in the fluvial environment and curvature alignment, levee foundation geometry can range from simple to very complex. Geomorphic features in the soil layers underlying a structure often have a significant effect on the underseepage behavior and the potential for initiating internal erosion. Based on the hypothesis that levee underseepage susceptibility comes from localized subsurface geomorphic features that interrupt the characteristic profile along that levee reach, a methodology has been developed that assesses the hydraulic effect of geomorphic features in levee underseepage reliability. The methodology consists of a response surface-Monte Carlo analysis that takes into account the uncertainty in the subsurface geometry and soil properties in assessing the seepage regime associated with the feature. The method utilizes three-dimensional steady-state finite-element underseepage analyses to develop a response surface representing the relationship between soil properties and the three-dimensional levee foundation. The response surface then serves as the driving function for reliability analyses by means of Monte Carlo simulation analyses, resulting in cumulative probability functions for either hydraulic exit gradient or factor of safety against heave. These computed probability functions represent an assessment of conditional probability of initiation of internal erosion. Results can be adjusted for curvature effects when needed. The analysis of a crevasse-splay, an abandoned channel, and a meander scroll feature found in the Sacramento River (east side) levee system in California are presented as application examples.

Backward erosion piping is a highly three-dimensional process responsible for the failure of many embankment dams and levees. Unfortunately, the majority of numerical models developed for predicting piping are two-dimensional. This study presents finite element models for backward erosion piping computations in both two- and three-dimensional domains. Analyses results indicate that the degree of concentration of flow in three-dimensional models is much more severe than in two dimensions, resulting in higher estimates of the hydraulic gradient near the upstream end of the erosion channel.

A method recently proposed for the computational modeling of backward erosion piping is applied for the numerical back-analysis of some pioneering experimental tests on physical models of cofferdams performed by Marsland (1953).

Seepage flow through a deformable porous medium may cause detachment, transport, and even deposition of fines particles which are initially parts of the granular skeleton. This volumetric erosive process is called suffusion. With the aim of modelling the mechanical consequences of suffusion, we propose a constitutive framework for a fully coupled poromechanical model of a suffusive soil. Considering kinematics, suffusion is modeled as a mass transfer from the solid phase to the fluid phase. We therefore introduce, into the classical poromechanical framework, a new state variable $$\phi _{er}$$ ϕ er as the suffusion induced porosity. From thermodynamics of porous media, we deduce a possible coupling between suffusion and seepage flow. In order to capture the mechanical consequences of suffusion, $$\phi _{er}$$ ϕ er is also regarded as an internal variable into a poroplastic model so that it contributes to the hardening rule. Numerical integrations of the developed model were carried out under monotonic drained loading conditions in such a way that a part of the abilities of the model are illustrated.

The flow conditions influencing the onset of contact erosion have been investigated physically using a novel setup based on the Particle Imaging Velocimetry (PIV) and numerically using a coupled DEM-LBM (Discrete Element Method-Lattice Boltzmann Method) approach. The flow conditions at the transition from base to filter material were experimentally quantified and computationally simulated resulting in a satisfactorily good agreement. An interesting and important outcome of these investigations is that the maximum flow velocity is not appearing in the base material, but in the constriction of the filter material in the transition zone from base to filter just above the surface of the base material. A generalised relationship between flow velocity and geometrical conditions at the transition zone was developed based on the Froude number. The characteristic evolution of the curve clearly shows the competition between hydraulic and geometric/mechanical conditions influencing the onset of contact erosion.

The void space of granular materials can be considered as a collection of poly-sized pore bodies separated by narrow pore throats or constrictions. Pore network models have been developed to estimate the probable path length covered by fine particles flowing through a granular filter. For this calculation two pieces of information are required: the constriction size distribution and the mean void spacing between two constrictions which corresponds to the mean pore diameter. Different assumptions have been previously made in the literature to determine this void spacing. However, they all neglect the influence of the density and thus, the estimation of this quantity remains an open issue. This paper compares different definitions for the mean pore size based on statistical analyses performed on numerical samples composed of spheres by means of the Discrete Element Method. Different sphere packings with different gradings and density states were considered and a weighted Delaunay tessellation was applied to extract the main void characteristics. In a second part, a simple formula is proposed to quickly estimate the equivalent sieve opening size of the granular filter. This value is very close to the mode value of the constriction size distribution which is related to the most represented constriction size in a granular material.

Installation of steel sheet piles at the toe of a levee has been adopted as a seismic countermeasure against liquefaction-induced lateral spreading of the levee foundation. Since installation of the sheet piles alters the seepage flow in the foundation ground, use of permeable steel sheet piles is proposed. Holes sufficient to secure the water flow in the ground on the sheet pile might induce internal erosion due to the concentrated seepage flow. In this study, development of seepage-induced internal erosion near the holes on the permeable steel sheet pile is examined by physical model test and numerical simulation. Comparison between physical model test and numerical analysis reveals that the experimental result can be captured by the numerical analysis as a whole. However, the clogging of fines in the experiment cannot be modelled, as detachment of fines from the soil skeleton and fines redeposition are not separately considered in the erosion model used.

The coarse sand barrier (CSB) is a single granular filter used to retrofit an existing structure, making it more stable against backward erosion piping. The barrier material should be chosen carefully to retain particles from the sand layer upstream of the barrier, yet provide optimal resistance against backward erosion piping. This means the particles in the barrier should be large, thus difficult to transport, and the barrier should have a high permeability in order to reduce the local hydraulic gradient inside the barrier. However, transport of the fine sand upstream of the barrier into the barrier, may result in less permeable filter cake just inside the barrier. Therefore, a column set-up was designed and experiments were conducted with various sand types, using the same materials as used in small-scale and medium-scale backward erosion piping experiments with at CSB. The aim was to compare the results of different tests and check if the criterion for the formation of a filter cake is the same as the well-known filter rules. Consequently, this paper presents the results of different column experiments. These materials were carefully selected and fulfilled the selected filter rules but one soil composition caused the filter cake formation. This indicates that to avoid filter cake formation for the conditions tested stricter rules apply than the filter rules considered here.

The coarse sand barrier (CSB) is a promising method to avoid ongoing backward erosion piping resulting in increased safety of a dike for this failure mechanism. Experiments are performed at different scales at Deltares, the Netherlands. These experiments show a significant increase in the critical head for structures with a CSB compared to structures without a CSB. The increase of critical head cannot only be ascribed to the lower erodibility of the coarser particles in the barrier, but also to the reduction of hydraulic load on these coarse particles in the barrier, resulting from the permeability contrast of barrier material and surrounding sand. To investigate the influence of a CSB on the flow pattern numerical and analytical calculations have been performed. This paper focusses on the analytical calculations. It will be shown that these can explain the increase in strength and the measured scaling effects.

The paper presents a multidisciplinary analysis carried out for the characterization and monitoring of a levee in Bozen Province, North Italy. The study treats a small section of the Adige river embankments, interested in the recent past by moderate piping phenomena and subjected to some subsequent interventions for the risk mitigation. The data acquired with an Electrical Resistivity Tomography (ERT) investigation and an optical fiber distributed temperature sensing (DTS) are compared to boreholes information, laboratory tests and piezometers measurements. They provided a multi-dimensional characterization of the levee and of the close subsoil water-meadows, possible piping preferential paths. Since the presence of more permeable lenses within the silty matrix characterizing the levee foundations, the subsequent 2D seepage analysis was carried out with the Boolean Stochastic Generation (BoSG) method, which randomly generates lenses with specific rheological properties within a matrix with another set of parameters. The soil configurations that are more congruent with the monitoring data were selected within a pool of 360 simulations, providing information about the probable seepage mechanism within the levee and the reliability of the interventions.

Due to several triggering causes, e.g. heterogeneity of the grain size of quarried materials, inappropriate compaction, discontinuities of displacements, dynamic effects, particles migration and internal erosion processes and animal actions, permeability defects (p.d.) are practically unavoidable in earthen structures. Their detection, usually carried out through the monitoring of seepage flow and erosion phenomena, is needed to evaluate the safety evolution of these structures and prevent exceeding serviceability (change of discharge, turbid water) or ultimate (local or global instabilities, piping, structural collapses) limit states. To this purpose, distributed thermal monitoring systems, based on optical fibers sensors, are becoming more popular. The variations in measured temperature values are not however immediately linkable to the effects of p.d., due to the ‘coupling’ of seepage flow process and heat transport mechanism through materials and foundation soils. To better understand the mutual dependence between these coupled processes and to investigate the effects of a p.d. due to internal erosion, several numerical simulations have been run. The analysis of a large scale test on a monitored experimental dyke is finally carried out.

In the assessment of water defences for macrostability and backward erosion piping the design hydraulic load mostly refers to steady state conditions in equilibrium with the maximum river water level. In this contribution, we show selected results of coupled hydromechanical numerical analyses of a paradigmatic Dutch case, which demonstrate that this assumption leads to high overestimation of the true hydraulic loads at the toe of the water defence embankment. The hydraulic resistance of the bed of the river and the deformability of the foundation layers introduce a decay in the pore pressure time history, which largely reduces the action on the hydraulic protection structure. The finite element model was developed to assist in the assessment of an innovative solution based on passive wells as a measure to reduce the risk for macrostability and piping. It was calibrated on available pore pressure measurements in the foundation of critical sections of the dykes of the river Lek in the Netherlands under the daily tidal action. The model was used to determine the distribution of pore pressure expected in the subsoil of the dykes for the design maximum load. The calibration stage of the model is specifically interesting to the aim of evaluating the reduction of the input pore pressure due to the hydromechanical resistance of the geotechnical system.

The study will present a further salient advance of 3-D modelling and its application in the efficient design of hydraulic barriers intended to prevent sand boil formation. The approach to hydraulic analysis was carried out using FEMWATER (Lin et al. 1997) a 3-D Finite Element model while data have been processed by the graphical interface of the Groundwater Modelling System (GMS 2017 software). The area that arose our interest due to the reoccurrence in the past 70 years of several impressive sand boils, is located along a stretch of the Po river, Italy, between the villages of Boretto (RE) and Pieve Saliceto (RE). Recorded sand boils formed after flood events in 1951, 1994 and 2000 (reactivated in 2014), each one appearing downstream from the previous one and ever closer to a freeway fly-over, without though a significant change in distance from the levee toe. In order to mitigate sand boil risk formation one of the solutions to be studied was: build a new cut-off wall along the same stretch of river embankment.

The presence of sand boils near the toe of a levee indicates that backward erosion piping, an internal erosion process, has initiated at a given location. As a primary failure mechanism for levees, backward erosion piping is commonly evaluated as part of levee risk analysis. Currently, sand boil observations are subjectively incorporated into risk analyses through expert judgement. This study presents an approach in which sand boils are statistically assessed as point pattern data. Results indicate that density estimates may provide a convenient means to compute initiation probabilities for backward erosion piping, and cluster analysis may provide a means of estimating characteristic levee segment lengths. Further research is needed to explore the generality of these concepts under various geologic conditions.

The coarse sand barrier is considered as a promising measure to prevent backward erosion piping from causing failure of embankments. A pipe is allowed to progress backwards until it encounters the coarse sand barrier, which prevents it from progressing unless the head difference over the embankment is significantly increased. A three stage experimental programme supported by groundwater flow modelling is carried out to investigate the feasibility of this method. The hypothesis is that the strength of the barrier is characterised by a local gradient at the interface between the barrier and the pipe. Major questions are: can the horizontal gradient as measured in laboratory tests be used to characterise the strength of the barrier material, over which distance should a horizontal gradient be determined, and is this distance the same for models at different scales? This paper presents the background theory and demonstrates the effects using scale dependent criteria. Preliminary results of small- and medium-scale experiments are used to compare the two approaches.

Suffusion is one of the main internal erosion processes in earth structures and their foundations. The assessment of this phenomenon can be difficult since in-situ geotechnical properties of soils are variable and uncertain. By means of a case study, this paper aims at presenting a general method to assess the suffusion potential of compacted impervious cores of zoned embankment dams. First, the suffusion susceptibility of the compacted layers forming the analysed dam core is estimated from four soil parameters that can be easily measured in situ or in laboratory during construction. Second, the saturated hydraulic conductivity of the compacted layers is evaluated based on the amount of fines content and on available construction data. Moreover, the power dissipated by seepage flow is inferred based on the saturated hydraulic conductivity and simplified fluid boundary conditions. Finally, the combined consideration of erosion resistance index and dissipated energy allows the identification of zones characterized by a relatively larger suffusion potential.

A recent study (Aielli et al. 2017) about sand boil reactivations in the major Po river banks (Italy) has demonstrated how a structured database of historic information can significantly contribute to better understanding of seepage phenomena and provide a practical tool for safe management of hydraulic works. The primary aim of the present work is to enhance the information collected in the database (DB), which cover the whole course the Po river, to the beginning of the delta. In particular, an improved estimation of the critical height (i.e. height of the water level in the river which causes a sand boil reactivation) for sand boil phenomena is discussed herein. The maximum height without reactivation and the minimum height with reactivation have been evaluated for each recorded sand boil. The DB information and processing of the available data allow defining how the values of the critical heights are distributed along the course of the river. The DB information, connected to the alert thresholds and to the early warning system, can provide an indication of the possible reactivations of the piping phenomena in advance, enabling efficient coordination of the emergency actions against backward erosion piping progression.

Backward erosion piping is a failure mechanism which involves the formation of shallow pipes in a sandy foundation layer and is considered to be a major risk for levees. For understanding this mechanism and the development of prediction models, laboratory experiments are essential. In addition, due to scale effects and heterogeneity in field conditions, field observations and case histories are indispensable for validation of models and delineation of piping sensitive conditions. However, both experiments and field observations are often not easily utilized for this purpose. Piping experiments have been conducted in various research programmes, countries, and in a variety of configurations making the experiments difficult to compare due to inconsistent observations and differing configurations. Case histories are often poorly documented and like experiments, described in different sources and different levels of detail, due to which their full potential is often not reached. Given the importance of experimental and field data for the prediction of backward erosion piping, a need exists for a centralized organization of data. Two different databases are presented here, for laboratory experiments and field observations respectively, each combined with a web application for viewing and exporting the data. The laboratory experiment database is populated with 332 experiments. The field observation database is currently populated with 3 failure cases and 2840 sand boils located in the Netherlands and the United States. Future work will focus on a more complete population of the databases, user-friendliness of the web viewer, and analysis of the gathered data for improvement of prediction models.

When comparing different literature sources, differences in the terminology concerning soil deformation due to seepage may be found. The distinctions are caused by language barrier (Slavic and Germanic language groups), by the view on the mechanisms and by the way of understanding individual soil deformation processes. These differences in terminology are related to the naming of the phenomenon itself or to translation issues. Clear terminology is important when comparing stability criteria applied in different countries. Here “western” and “eastern” schools may be distinguished. This study provides an overview on terminology for soil deformation due to seepage in English, French, German, Russian, Polish and Czech languages, some suggestions are presented with discussion. The paper is an initial phase for further comparison of criteria for individual soil deformation modes.