**8. Practical examples to illustrate the analysis of earth erosion problem caused by rapid filling and drawdown conditions in embankments**

#### **8.1 General settings**

298 Soil Erosion Studies

During the field investigation to construct an earth dam, particularly when looking for the materials that might be used to construct the embankments, it is very important to identify the existence of dispersive soils. This identification should be done first through one of the special field tests that exist for this purpose. Although the results of such tests must be verified through laboratory tests, field tests might give a good preliminary evaluation of the dispersivity of the soils under investigation. Knodel (1988) presents a good description of the most common laboratory and field tests that are used in the engineering practice to identify the dispersivity of soils. Among them are the following: a) for field: *crumb test*, *water drop test*, *dissolved sodium test*, and *turbidity test*; b) for laboratory: *crumb test*, *the double* 

For any earth construction it is necessary to investigate, by using one or more of those methods mentioned above, the existence of dispersive soils; this investigation can be carried out through soil samples obtained in open wells during the soil exploration phase. Once the construction materials have been identified, a decision to use or refuse them has to be taken. Sometimes dispersive soils might be used in earth structures if they are mixed with lime or if well designed filters and drains are installed. If for economical reason it is decided to use

a. Arching. This problem might occur in zones around conduits through the embankment, near concrete structures, and at the foundation interface. In order to avoid negative effects, special control of compaction and moisture content during construction should be taken. b. Cracking due to differential settlements caused by soil consolidation, stress

c. Soil improvement of the dispersive clay, by adding hydrated lime or non dispersive clay of medium to high plasticity. Special care should be taken in compacting soil

d. Construction control. Special standards and specifications should be used when dispersive soils are involved in the construction of earth structures, particularly those related to soil density and compaction procedures. For instance, there should not be moisture concentration while adding water to obtain the specified water content during layers' compaction. Special monitoring consideration to dams with cores containing dispersive soils, should be giving during the first filling, in order to prevent any piping or internal erosion effect. Observation instruments are particularly recommended to periodically measure water pressures, water leaks and water levels at different zones of

concentration, two or three dimensional effects, etcetera, should be avoided.

**7.2.2 Design considerations when constructing with dispersive clays** 

dispersive clays, the following conditions have to be taken into account:

adjacent to rigid structures such as conduits.

the embankment.

propagate very rapidly causing a dam failure.

**7.2.1 Field and laboratory test for dispersive soils** 

*hydrometer test*, *pinhole test* and *chemical test*.

embankment, where there exists a high hydraulic gradient; such phenomenon progresses upstream, forming a kind of pipe until it reaches the water source. With dispersive clays, however, the internal erosion is due to a deflocculation process and it might start at the upstream side where there is the water source; the tunnel-shape passage or pipe, that is formed, is propagated toward the downstream side. If such dispersive soils exist in areas where there are already some cracks, or not well compacted zones as those presented along conduits, such cracks might increase and

The wet slope of an earth dam, lake or river banks and channel slopes are frequently subjected to sudden changes of water level (increments or decrements), which modify flow conditions inside the soil mass. Flow velocities, hydraulic gradients and seepage forces might, in extreme conditions, cause soil erosion problems ranging from slight to severe, such as piping or even the total failure of the structure. These phenomena, known as *rapid filling* and *rapid drawdown*, are complex problems in which the magnitude and rate of filling or drawdown, hydraulic conductivity and porosity of materials constituting the earth structure, geometry of slope and initial boundary conditions of flow are involved (Auvinet & Lopez-Acosta, 2010). By using the *Plaxflow* algorithm (Delft University of Technology, 2007), analyses to assess how these two phenomena affect soil erosion in earth structures are carried out through a numerical modeling based on finite element method (FEM) (Auvinet & Lopez-Acosta, 2010; Lopez-Acosta et al., 2010).

#### **8.2 Analysis considering only rapid drawdown phenomenon**

In this practical example, the effect of rapid drawdown phenomenon on erosion problems in a typical embankment is analyzed (Auvinet & Lopez-Acosta, 2010). Simplified geometry of the studied domain and boundary conditions considered in this analysis are shown in Figure 13. Rate of drawdown was established at 1.1m/day. Thus, a total dewatering of 5.5m in 5 days was assumed in this analysis. In the same way, it was accepted that embankment is constituted by a homogeneous and isotropic material with hydraulic conductivity *k*=1×10- 5 m/s and porosity *n* = 0.3 (void ratio *e* = 0.43).

Fig. 13. Simplified geometry and boundary conditions of the studied embankment (Auvinet & Lopez-Acosta, 2010)

From results of analyses (Auvinet & Lopez-Acosta, 2010), Figure 14 shows, for a typical time interval during drawdown (*t*=4 d), the free surface line which separates unsaturated material (upper part) from saturated material. Variation of this free surface, called *desaturation line* (for drawdown), obtained at different time intervals during rapid drawdown is illustrated in Figure15. Other authors prefer to call it *phreatic line* (Huang & Jia, 2009; Lam & Fredlund 1984; Lam et al., 1987). It must be underlined that this free surface line is not rigorously a flow line since velocity vectors cross it (see Figs. 16a and b). In the same way, results from analysis demonstrate that during drawdown, when water surface descends, large velocities are generated at the contact between the level of water and the slope (which are proportional to hydraulic gradients, and consequently to seepage forces in that zone); these velocities can

Internal Erosion Due to Water Flow Through Earth Dams and Earth Structures 301

Figure 17 shows the maximum exit hydraulic gradient (*imax*=0.499) reached at the toe of slope at the end of the rapid drawdown (*t*=5 d). Additionally, Figure 18 illustrates how the free surface line and velocity vectors change due to the placement of a horizontal drain inside the embankment (Lezama, 2010). The most conspicuous difference can be observed in the reduction of hydraulic gradient at the toe of slope at the end of rapid drawdown (*t*=5 d), from *imax*=0.499 to *imax*=0.25. This demonstrates the usefulness of placing drains in order to

Desaturation line

Desaturation line

Fig. 17. Hydraulic gradients (magnitude) at the end of rapid drawdown (*t*=5 d) (Lezama,

Fig. 18. Changing in velocity vectors and reduction of hydraulic gradient (magnitude) at the end of rapid drawdown (*t*=5 d) due to the placement of a horizontal drain into the levee

This example focuses on studying the effects on soil erosion due to transient flow within a levee as water level of a river increases and decreases because of the rain cycles in a tropical region. Simplified geometry of studied domain including foundation soil of the levee is illustrated in Figure 19. Properties of materials are specified in Table 3 (Lopez-Acosta et al.,

7 6.7m

4.0m

100.0m

Fig. 19. Simplified geometry and material number of the studied domain (Lopez-Acosta et

10.4m

**8.3 Analysis considering both rapid filling and drawdown phenomena** 

reduce soil erosion problems in earth structures.

*imax* = 0.25 Horizontal drain

2010)

*imax* = 0.499

(Lezama, 2010)

2010).

4 5 6

al., 2010)

1 2 3

facilitate local "piping" of material of these regions and jeopardize slope stability. The existence of this maximum flow velocity close to the slope and under the level of water can be observed in Figures 16a and 16b (Auvinet & Lopez-Acosta, 2010).

Fig. 14. Variation of degree of saturation for *t* = 4 d (345200 s) (Auvinet & Lopez-Acosta, 2010)

Fig. 15. Variation of *desaturation line* at different time intervals during rapid drawdown (Auvinet & Lopez-Acosta, 2010)

Fig. 16. Velocity vectors (magnitude) for two different time intervals during rapid drawdown (Auvinet & Lopez-Acosta, 2010)

facilitate local "piping" of material of these regions and jeopardize slope stability. The existence of this maximum flow velocity close to the slope and under the level of water can be

Free surface

Fig. 14. Variation of degree of saturation for *t* = 4 d (345200 s) (Auvinet & Lopez-Acosta,

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 26.0 28.0 30.0

Fig. 15. Variation of *desaturation line* at different time intervals during rapid drawdown

*Vmáx*=2.46×10-6 *m/s*

(a) *t* = 1 d (86000 s) Desaturation line

Fig. 16. Velocity vectors (magnitude) for two different time intervals during rapid

Desaturation line (b) *<sup>t</sup>* = 4 d (345200 s)

*Vmáx*=4.66×10-6 *m/s*

Distance *X* (m)

observed in Figures 16a and 16b (Auvinet & Lopez-Acosta, 2010).

2010)

t=1.6 h t=1 d t=2 d t=3 d t=4 d t=5 d

(%)

(Auvinet & Lopez-Acosta, 2010)

drawdown (Auvinet & Lopez-Acosta, 2010)

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

Values of

*H* (m) Figure 17 shows the maximum exit hydraulic gradient (*imax*=0.499) reached at the toe of slope at the end of the rapid drawdown (*t*=5 d). Additionally, Figure 18 illustrates how the free surface line and velocity vectors change due to the placement of a horizontal drain inside the embankment (Lezama, 2010). The most conspicuous difference can be observed in the reduction of hydraulic gradient at the toe of slope at the end of rapid drawdown (*t*=5 d), from *imax*=0.499 to *imax*=0.25. This demonstrates the usefulness of placing drains in order to reduce soil erosion problems in earth structures.

Fig. 17. Hydraulic gradients (magnitude) at the end of rapid drawdown (*t*=5 d) (Lezama, 2010)

Fig. 18. Changing in velocity vectors and reduction of hydraulic gradient (magnitude) at the end of rapid drawdown (*t*=5 d) due to the placement of a horizontal drain into the levee (Lezama, 2010)
