EFFECT OF LIQUEFACTION ON STABILITY OF TAILINGS DAM

Abstract

This study presents the research investigations carried-out to assess the stability of tailings dams, with particular reference to liquefaction of the tailings in the event of an earthquake. For this study azoned dam, under construction, having three zones consisting of, from upstream to downstream, (i) impervious material, (ii) compacted tailings, and (iii) pervious random fill, has been fully analysed. The tailings dam is founded on rock. An inclined chimney drain and a connected horizontal filter are provided to keep most of the dam section on the downstream side dry. Decantation wells with horizontal drains are also provided to drain out the impounded tailings slurry. The dam is proposed to be raised in six stages up to its maximum height of37 metres by the centerline method ofraising. At every stage ofdam raising a major portion ofthe upstream slope ofthe dam, made-up of impervious material, rests on the tailings fill of the previous stage, hence the tailings fill also becomes an integral part ofthe dam. The tailings is a fine, cohesionless material and is to be impounded by the hydraulic fill method thereby giving it a minimum placement density. This tailings material has been studied experimentally to determine its liquefaction potential. Tests were conducted on the shake table, with different exciting acceleration levels and placement densities. It was observed that the tailings material was very susceptible to a reduction in effective stress and even at an exciting acceleration value of 0.025g, complete liquefaction occurred. The liquefaction potential, defined as the ratio of the induced hydraulic gradient to the critical hydraulic gradient, was 1.79 at this level of excitation. Therefore, there is no doubt that liquefaction of the tailings fill would take place even in a mild seismic event. The Indian Standard Code, 1893-1984, has divided the country into five seismic zones. The dam is situated in zone II and for this zone the recommended horizontal seismic coefficient value is 0.06. Accordingly, this value has been adopted for the stability analysis. Stability of the upstream slope has been evaluated by determining safety factors using the simplified limit equilibrium method, for non-seismic and seismic conditions. In the non-seismic condition, safety factors have been worked-out for two situations, (i) without tailings fill, on the upstream side, for the present stage of construction, that is, impoundment of tailings is only up to the level of the previous stage, and (ii) when impoundment of tailings, on the upstream side, is complete for the present stage of construction. For the seismic condition, these two situations are considered along with the consideration of no liquefaction and complete liquefaction of the tailings fill. Thus, for the upstream slope, for the height up to the present stage of construction, two shear surfaces giving the minimum safety factor have been identified for the non-seismic condition and four such shear surfaces have been identified for the seismic condition. Thus for the six stages of dam raising, a total of thirty five shear surfaces and their corresponding safety factors are reported. This is done by trial, by considering 50 to 100 slip surfaces for each case. For both the non-seismic and seismic conditions, the minimum safety factor is for the condition when the tailings fill is only up to the level of the previous stage, that is,notailings fill support is available to theraised upstream slope. In order to see the cumulative influence of the different stress-strain characteristics of the four constituent fill materials of the dam on the gravity stresses used in limit equilibrium analysis, finite element method of stress analysis has been carried out. The modulus values of the four constituent fill materials have been determined by conducting cross borehole tests at the site. Contours for stresses cr,, <r3 ,crx,cry and rxy, for the earlier mentioned thirty five cases, have been obtainedby carrying out a 2D plane strain linear elastic analysis, cr,,o-3 are the principal stresses, crx, ay are stress in the x and v directions and t , is the shear stress on the x-y plane. One set of safety factors, for J xy ii the critical slip surfaces obtained from limit equilibrium analysis, have been determined as the average of deviatoric stress ratio (cr,-cr3)f/(cr1-cr3) at points on the shear surface, where (c^-cr^is deviatoric stress at failure and (cr,-cr3) is the actual deviatoric stress at these points. Another set of safety factors have been obtained as the average ratio of shear strength crn tan </> and shear stress rxy, where normal stress crn is derived from cr and cr stresses. \ y Results from the simplified limit equilibrium analysis, for the non-seismic condition, show that the minimum safety factor of 0.975 is for the upstream slope after second stage raising of the dam and without tailings fill for this stage of construction. This safety factor increases to a value of 2.656 when tailings fill for the second stage is complete. For the other stages of construction, the safety factors varied from 1.032 to 1.269, without tailings fill for that particular stage of construction, and varied from 2.120 to 2.872, with tailings fill for that particular stage of construction. For the seismic condition, the minimum safety factor of 0.276 is obtained after second stage raising of the dam, without tailings fill for this stage of construction and considering liquefaction of the first stage tailings fill. In case liquefaction of the first stage tailings fill does not occur, the safety factor would be 0.778, which is also less than unity. Thus, if liquefaction of the tailings fill occurs, the most vulnerable situation is when the tailings fill for this stage of construction is not there for the raised height of the slope. The safety factors for such cases range from 0.276 to 0.338 for the different stages of raising of the dam. However, for the cases where the tailings fill for this stage of construction is complete and liquefaction of fill does not occur, the safety factors range from 1.714 to 1.930. Therefore, at some intermediate stage of impoundment, without liquefaction of tailings, the safety factor ofthe slope could have any numerical value between 0.778 and 1.930. in A perusal of critical slip surfaces of minimum safety factor, when liquefaction occurs in the tailings fill, shows that the slip involves only the top two stages of construction. After the failure/slip of the upstream slope, the downstream slope can not remain stable as almost a vertical scarp of about 1.5 metre height would be left at the top. The results of the analysis using the deviator stress ratio approach indicate that these safety factors are higher than those obtained from the simplified limit equilibrium analysis by about 10 to 22 percent. However, the trend is similar. The minimum safety factor for the non-seismic condition is 1.092, for the dam slope constructed up to the second stage and with tailings fill only for the first stage dam height. For the other stages of construction, the safety factors varied from 1.171 to 1.396, without tailings fill for that particular stage of construction, and varied from 2.353 to 3.303, with tailings fill for that particular stage of construction. The results for the seismic condition also show a similar trend. The minimum safety factor of 0.337, against 0.276 obtained from the limit equilibrium approach, is after second stage raising of the dam, without tailings fill for this stage of construction and considering liquefaction of the first stage tailings fill. For nonliquefied tailings it ranges between 0.895 and 1.218, without tailings fill for that stage of construction. When the tailings fill for that stage of construction is complete and liquefaction does not occur, the safety factors range from 1.954 to 2.239. However, when the tailings fill for that stage of construction is complete and liquefaction occurs, the safety factors range from 0.911 to 1.949. The results obtained from the shear strength ratio approach are also 12 to 20 percent higher than those obtained from the simplified limit equilibrium approach and the numerical values are found to be close to those obtained from the deviator stress ratio approach. Therefore, the design of the dam, under consideration, needs to be modified to have a minimum safety factor of 1.5 in non-seismic and 1.0 in seismic conditions respectively. iv From this study, it is concluded that (i) for seismic analysis, complete liquefaction of the tailings fill has to be considered, as the material is very susceptible to liquefaction, (ii) in the design, at any stage of construction, the upstream slope of the dam must not rest on the tailings fill, which is the weakest material and is susceptible to liquefaction, and (iii) the simplified limit equilibriummethod may be used to evaluate safety factors for the dam slope, impounding tailings of such hazardous materials.