STABILITY ANALYSIS OF LEVEES

Abstract

Flood is a natural menace, which occurs in many parts of the world. In India, some areas are severely affected by the floods. This natural menace, leads to loss of lives and properties. Therefore, it is essential to build suitable structures to save the living beings and properties from flood. Amongst the various flood control elements, levees and floodwalls are very common. Levees are amongst the world's oldest hydroengineering structures. They are constructed to control flood and safeguard land, properties, and living beings. Levees are also built to facilitate navigation, irrigation, recreation, and to prevent saltwater intrusion. Both the river and canal levees are very much similar to earthen dams of medium height. However, duration of ponding of water and fluctuation of water level in upstream side, degree of saturation of soil in the levee, etc., make river levees different from canal levees and earthen dams. Unlike earthen dams, river levees are sometimes constructed on weak foundations. For construction of levees, borrow materials are generally obtained from shallow pits nearby streams. These materials may be heterogeneous and of marginal quality for levee construction. Levees failure is common in India and in other countries. Many levees along the rivers Ganga, Yamuna, Koshi, Brahamputra and Sone in India, Mississippi in North America, Danube in Czechoslovakia, Hungary, and Yugoslavia, have faced many problems, such as: rapid rise ofwater level in upstream side, overtopping of floodwater, surface erosion, sudden drawdown of floodwater, subsurface erosion (sand boils and piping), slope failure, liquefaction in foundation during earthquake and blasting, etc. Unfavourable changes in levees arising from hydrodynamic loading can lead to (i) liquefaction in loose sands, (ii) piping in sandy gravels, and (iii) hydraulic fracturing in silts, silty clay and rock masses. The water seeping through the body of the levee and through the pervious foundation of levee may prove harmful to the stability of levee due to development of pore water pressure. During floods, two adverse phenomena may occur, one is spring, sand boil and piping either through the body or through the foundation, and the other is heaving which may lead to the failure of the hydraulic staictures. in The stability of dam or levee is that property which enables it to stay in position. A levee is stable if the resultant of all the forces acting on the levee structure does not result in movement. Stabilities of levees may be classified as construction stability, slope stability, filtration stability, stability of the levee body, stability of the subsoil, the stability of both levee and subsoil, and stability of the protecting layers. Moreover, in general, it is customary to consider only mechanical stability, even though for river and canal levees filtration stability is very important. Deformation due to seepage stress and failure should be investigated for stability analysis of levees. Considering the importance of stability of levees, the following objectives have been identified for the present work. 1. To study the stability of foundation soil, susceptible zone to sand boiling, and path of piping channel considering insitu effective stress, induced stresses due to levee structure and seepage force. 2. To find seepage characteristics for a homogeneous pervious levee on impervious base with a toe filter considering the resistance ofthe upstream face and to locate the appropriate position oftoe filter so that the capillary rise above the phreatic line is well within the downstream sloping face. 3. To design downstream cut-off wall for an impervious levee resting on infinitely deep permeable soil considering exact seepage force. 4. To compute seepage characteristics under steady and unsteady state, and to study variation ofpiezometric head, pore water pressure, and exit gradient for levee with antiseepage elements by applying numerical methods. 5. To analyze stability of levee slope considering realistic pore water pressure, dry and saturated unit weights in a levee on impervious base. 6. To study the seepage characteristics by conducting experiments for onedimensional flow through sand column mixed with varying percentage of clay, plant root, etc., and sandy gravels in order to compare the theoretical values of exit gradient, uplift pressure, etc., with the laboratory values. 7. To conduct laboratory tests on clayey soils to differentiate between dispersive and non-dispersive clays, to study mode of failure, and time of collapse for suitability of clay for construction of river and canal levees. IV The salient findings of the present study are as follows: (i) Stability Analysis of Foundation by Analytical Method Seepage flow characteristics (pressure head and direction of flow) at different points in the flow domain have been obtained using conformal mapping technique. Three cases have been considered, these are: (i) an impervious levee on a permeable soil of infinite depth, (ii) an impervious levee on a permeable soil of finite depth, and (iii) an impervious levee with a cut-off wall resting on a permeable soil of infinite depth. The flow nets have been drawn analytically. The seepage force acting at different points in the flow domain is computed using gradient method [Cedergren, 1989]. Insitu effective stress due to self-weight and induced stresses due to levee structure and pondage water on upstream slope have been determined by assuming soil to be elastic. Considering these stresses and stress due to seepage force, normal and tangential stresses on two orthogonal planes, i.e. vertical and horizontal planes at a point are computed and the Mohr's circle drawn. Mohr-Coulomb failure envelope is drawn for foundation soil with known effective cohesion (q) and angle of internal friction ((jjj). If Mohr's circle of stress touches the Mohr's failure envelope, it is presumed that soil at that point is critical to sand boiling and piping. If the Mohr's circle of stress at a point in the flow domain does not touch the Mohr-Coulomb failure envelope then soil is stable against piping. Using this concept, the stability of foundation in respect of sand boiling and piping of an existing ring levee along rivers Ganga, Sone and Punpun, near Patna (India) has been evaluated. (ii) Seepage Through a Levee An approximate tractable solution to the problem of seepage through a levee with a toe drain on an impervious base is obtained by considering resistance ofentire portion of upstream side, using method offragments. The function ofa filter is to control seepage through an earth dam or a levee. Filter is also essential to keep down the top flow line as low as possible to increase dry zone in downstream side in the levee for greater stability. Provision offilter nearer to the upstream side results in, higher seepage losses and increment in required filter length. If the filter is located nearer to the downstream toe, though seepage gets reduced, saturated zone is increased resulting in reduction of dry zone. Therefore, in the present study, the performance of a filter has been studied and its appropriate location has been suggested. The study has been extended for a zoned levee having an upstream rock-filled part. It is found that Casagrande [1940] correction accounting for resistance of the upstream portion of a levee is not enough. The capillary saturation above phreatic surface governs the position of filter and required length of filter. (iii) Design of Sheet Pile Cut-off Wall Considering exact seepage forces and applying the laws of equilibrium to a trial wedge, active and passive earth pressures acting on a rigid sheet pile cut-off wall of an impervious levee have been determined. The resultant thrust on the cut-off wall has been obtained by subtracting the passive thrust from combined active thrust and unbalanced water force as suggested by Terzaghi [1954]. The weight of the levee has been considered in the computation of active earth thrust. The design procedure of cut off wall has been explained. (iv) Stability Analysis of Foundation by Numerical Methods For complex foundation conditions, steady state and transient confined seepage through the foundation of a levee with antiseepage elements has been analyzed by finite difference method [FDM], and finite element method [FEM] using software ANSYS. From the study conducted, it has been observed that the flow characteristics for steady state condition obtained by numerical methods i.e. FEM compare well with those obtained by analytical methods. Performance of relief well in reducing pore water pressure in foundation soil on the downstream side has been investigated. Effect of relief well on exit gradient has been also studied. It is observed that relief well is very effective in reducing exit gradient. (v) Stability Analysis of Slope A Roorkee based software SARC.FOR has been modified and used to study the stability of the downstream slope of a homogeneous levee with and without filter considering actual pore water pressure, dry and saturated unit weight ofsoil above and below phreatic line respectively, whereas software SANC.FOR has been used for heterogeneous and zoned levee. It is observed that the provision of afilter ensures high stability ofdownstream slope ofthe levee and the stability ofthe downstream slope ofa vi levee decreases as the filter is shifted towards the downstream toe. Furthermore, stability of the downstream slope of a ring levee existing in Patna (India) has been investigated using modified software and the levee slope has been found to be stable. (vi) Experimental Investigation In view that the experimental studies are needed to understand the mechanism of sand boiling and piping, seepage failure tests have been conducted in laboratory for the following types of soil: (i) Yamuna river sand mixed with various percentage of clay, (ii) sand with plant root. Piping tests have been performed on sandy gravels. Pinhole tests and soil security tests have been also carried out for clayey soils collected from a site near by Ganga canal levee at Roorkee (India). Uplift pressure tests have been conducted on river sands. The experimental findings are as below: 1. Clay and plant root are very effective in preventing sand boiling and piping. 2. When stability index (H/F)ml„ is less than 1.0 and uniformity coefficient (C«) is greater than 12, piping of fine particles occurs at very small hydraulic gradient. Here F is mass fraction smaller than D; H is mass fraction between D and AD, and D is the any chosen particle size. 3. Clay tested from nearby Ganga canal levee at Roorkee is of non-dispersive nature. 4. The uplift pressure at the base of hydraulic structures bears a good correlation between theoretical and experimental values during the initiation of sand boiling and piping. Thus in this study, stability aspects of levees have been studied from different perspectives. VI1