 Dams are designed to support the normal load of water in a reservoir. They are also designed to withstand the extreme loads caused by floods and earthquakes. When the construction of a dam is complete, the reservoir is allowed to fill with water. Depending on precipitation in the reservoir size, the reservoir may fill quickly or it may take several years. The first time a reservoir is filled, engineers like the reservoir to fill slowly because it gives a dam structure time to adjust to the water load. If water is pumped into the reservoir from somewhere else, the initial filling can be controlled more accurately. The dam's first filling is closely observed to make sure that the dam is adjusting safely and performing according to the design. Concrete dams attain an equilibrium condition relatively quickly after the reservoir is filled with water. Embankment dams, however, take a longer time, possibly several years to reach equilibrium. This is because settling inside the dam and seepies through the embankment take a longer time to develop. So an embankment dam reaches an equilibrium state more slowly. The terms rapid reservoir filling and rapid reservoir withdrawal, commonly called drawdown, need to be defined for embankment dams. The rate that seepage patterns develop in a dam depend on the permeability of the materials in the embankment. If the reservoir level rises or falls faster than the embankment can initially absorb or drain off the water trapped inside, the change is called rapid. If a rapid drawdown takes place after an embankment dam has reached full equilibrium, internal loads may develop in the upstream portion of the structure that may cause the upstream slope to fail. Fortunately, this almost never happens. A rapid drawdown may also cause stability problems to develop in a natural slope surrounding the reservoir. If this happens, landslide material can damage or block outlet structures or spillways. Because of the potential damage that may occur, dam and reservoir slope should be evaluated for stability under rapid drawdown. In this program, the Federal Interagency Committee on Dam Safety is presenting Mr. John Lowe III, who will discuss rapid drawdown and present a method for evaluating the stability of slopes under rapid drawdown conditions. Mr. Lowe, a member of the National Academy of Engineers, has worked in geotechnical engineering his entire career. He did his graduate studies at MIT and in 1945 he joined NAP Engineering Company, which later became TIPPET's Abbott McCarthy Stratton, commonly called TAMS. As head of their soil and rock engineering department, he established their first soils laboratory. He eventually became a partner, retiring in 1983. During his career, he has worked on numerous large civil works projects, including ports, highways, airports and dams. His experience with dams includes active participation from early feasibility stage through design and construction into operation and surveillance. In fact, his role in Tarbella Dam in Pakistan, the world's largest embankment dam, is an example of his long-term involvement and broad experience. Also, the use of an excess of one million cubic yards of roller-compacted concrete at Tarbella exemplifies his role in technology development and application. He has served as secretary of the U.S. National Committee of the International Society of Soil Mechanics and Foundation Engineers and as president of the U.S. Committee on Large Dams. For many years, he was a member of the Soil Mechanics Advisory Board of the Corps of Engineers, Office Chief of Engineers. Contemporary members of the board were Arthur Casagrande, Ralph Peck and Stanley Wilson, among others. His experience includes working on dam projects in the Mediterranean and Middle East, the Far East, South America and the United States. For his contributions in Morocco, the king decorated him as commander of the Order of Alawit. Since his retirement from Tams in 1983, Mr. Lowe has been a consultant on a number of projects around the world. Some of his assignments included working in Bulgaria, Chile, Malawi, Jordan, Greece and the United States. May I present Mr. Lowe. The purpose of this video is to present a description of the principles governing proper stability analysis of embankment dams under the condition of rapid drawdown. Drawdown is considered rapid when insignificant water content change takes place in the soil along the critical sliding surface during the period of rapid drawdown. The term sudden drawdown is also frequently used for this type of loading since it implies that there is no time for water content change to occur during the loading. There are two other types of rapid loading to which embankment dams may be subjected. Seismic loading is a classic case. During the very short duration of an earthquake there obviously is no time for any water content change to take place. Even in a soil of intermediate permeability. Similarly, for low permeability soils no significant water content change occurs during a period of rapid heightening of a dam. Heightening may be done in order to increase the freeboard of an existing dam or may occur when a dam is built in stages and the stages are widely separated in time. I will talk primarily about the method of stability analysis for the rapid drawdown case. Methods for analysis of the seismic and rapid heightening cases are identical in principle. In all three cases there is the prerequisite condition that the soil in the dam must be subjected to a particular reservoir level for an appreciable period of time. This time period is necessary first for a steady pattern of seepage to develop through the dam. Then time is required for the soil to come to equilibrium under the earth and water forces occurring in the dam under the steady pattern of seepage. Earth dams consist of the following components. Slope protection on the upstream face, an upstream shell, an impervious zone, a chimney drain, a horizontal drainage blanket, a downstream shell, and downstream slope protection. In some embankment dams the impervious zone of the dam extends practically to the upstream face of the dam. Such dams do not have a pervious upstream shell. The impervious zone would be covered by a filter layer and then a layer of slope protection. The critical surface for stability failure is located at very shallow depth inside the impervious zone as shown in the figure. The undrained shear strength of the impervious material is a minimum at this location. During a long period of hold of the reservoir at high level, the soil at the upstream face of the impervious zone comes to equilibrium under the submerged unit weight of the overlying filter and slope protection. Deeper into the impervious zone the soil would have greater overburden load upon it and also would experience the effect of seepage forces between it and the upstream face of the impervious zone. Upon drawdown the weight of the filter and the slope protection layers would increase from submerged weight to drained weight. This increase amounts to practically a doubling of the weight of these layers and thereby a doubling of the tendency of the filter and slope protection layers to slide down the face of the impervious zone. Many embankment dams have a central impervious zone and an upstream relatively pervious shell. Again the critical surface for sliding is just below the upstream face of the impervious zone. This is the location where the impervious zone experiences the lowest confining pressures during high reservoir level. The failure surface then extends as a low angle plane through the upstream shell and slope protection. If the upstream shell does not drain significantly the driving force increases from that due to the submerged weight of the shell to that due to the saturated weight. If the upstream shell is free draining during drawdown the driving force increases from that due to submerged weight to that due to drained weight of the shell. In both cases practically a doubling of the driving force occurs. If drainage of the upstream shell is still taking place at the end of drawdown stability of the area of the upstream slope in the vicinity of the drawdown level of the reservoir where water is seeping out of the upstream shell should be evaluated. Occasionally the impervious core of a dam is connected to an impervious upstream blanket. In this case the critical surface for sliding extends upstream just below the surface of the upstream blanket. In the case of seismic loading instead of the increase in driving load being due to an increase in the unit weight of the material upstream of the impervious zone the increased driving load generally considered is a horizontal load equal to the total weight of the upstream material times an appropriate horizontal earthquake acceleration. At times a vertical load equal to the total weight of the upstream material times an appropriate vertical earthquake acceleration is considered to act concurrently. There are plans to heighten cardinal fly ash retention dam too. This picture was taken of the existing dam from the right abutment. The fly ash pond is on the left. In the distance the intake tower for the low level outlet can be seen. The white floating material adjacent to the upstream face are tiny hollow glass spheres collected with the fly ash. The upstream slope of the dam is covered with rock slope protection and the downstream slope with grass. The crest of the dam is 200 feet above stream bed. The dam has an upstream shell of clay mine spoil, an upstream sloping impervious core of clay, a transition zone of clay mine spoil, a chimney drain, and a downstream shell of clay mine spoil. The clay mine spoil is only slightly less plastic than the clay used for the core. The level of the reservoir at the time of heightening is expected to be elevation 903. The top seepage line is indicated on the cross section for this elevation. The proposed cross section for the heightening is shown here. Roller compacted concrete is proposed for the upstream portion of the upstream slope and minimizes the amount of overlay required over the original downstream slope. The clay core and chimney drain are extended upward for the heightening. The downstream shell is overlaying with mine spoil. A possible critical surface for stability failure is the circular surface shown here. If failure were to occur, the portion of the dam above the circle would slide upstream past the portion below the circle. Sharing would occur along the entire circular surface. The soil below the groundwater table in the existing dam will have come to equilibrium under the weight of the overlying material of the existing dam. Upon rapid heightening of the dam, the additional driving forces imposed by this material will have to be resisted by its undrained shear strength. In order to carry out stability analysis of a dam under any of the three rapid loading conditions, it is necessary to know the shear strength of the soil at all points along the failure surface being investigated. The preferred type of shear test is the triaxial test. In this test, a cylindrical specimen of soil is encased in a rubber membrane. Here is a picture of a giant triaxial specimen. The specimen is 24 inches in diameter and 62 inches high. The large diameter was necessary for testing the impervious material used in tabella dam. The material was widely spaced from 4 inch maximum size to 30 inch, 30 percent passing the 200 sieve. The tabella tests are the only ones that I know of that have been performed on such widely graded impervious material. The rubber membrane is attached to a pedestal at the base and a cap at the top. Both are equipped with porous stones to permit water to flow into or out of the specimen as desired during testing. Poor water pressure measuring devices were installed in the specimen. The leads to these devices are evident at the bottom of the specimen. A pressure chamber is put around the specimen for triaxial testing. And all around pressure up to 470 pounds per square inch can be applied to the specimen in the chamber. Port holes were provided in the steel walls of the chamber so that the specimen could be observed during testing. A piston which passes through the top of the chamber is used to apply an axial load to the test specimen. The piston is not visible in the photograph. For conservative representation of field conditions, test specimens must be fully saturated. To accomplish this, the chamber pressure and the pore water pressure in the specimen must be built up concurrently and at a slow rate. When a small increase in chamber pressure is reflected by an equal increase in pore water pressure, the specimen is considered to be 100% saturated. That is, no air bubbles are present in the pore water. Usually a chamber pressure of about 100 pounds per square inch is required to accomplish this fully saturated condition. Testing then proceeds by increasing the chamber pressure and applying a load to the piston to represent the desired major and minor principal stresses for consolidation. The stress imposed by the piston is termed the deviator stress. After consolidation, the deviator stress is increased to failure of the specimen without permitting any drainage. When the major principal stress is greater than the minor principal stress during consolidation, the test is termed anisotropically consolidated undrained triaxial test. The abbreviation is ACU. When the major principal stress and minor principal stress are the same during consolidation, the specimen is isotropically consolidated. The abbreviation is CU for such a consolidated undrained test. Casagrande called the CU test an R test. Shear strength is determined by performing laboratory tests on representative samples of embankment material. These tests must be carried out so as to duplicate field conditions as closely as possible. Before the rapid loading, the soil had come to equilibrium under the stresses which occur in the dam prior to the rapid loading. That is, it had been consolidated under these stresses. As an illustration, the major and minor principal stresses acting upon the critical failure surface for a central core dam with upstream impervious blanket at time of consolidation and immediately after rapid drawdown are shown here. The major and minor principal stresses before drawdown are indicated as sigma 1c and sigma 3c. After drawdown, both sigma 1c and sigma 3c increase to values about two times their values at the time of consolidation. The increase in sigma 3 merely adds to the back pressure in the laboratory testing. Since the applied black pressure has already resulted in 100% saturation of the test specimen, there is no need to increase the applied sigma 3 stress. Thus the chamber pressure used for consolidation is not increased. The deviator stress, however, is increased until failure occurs. For failure surfaces having a factor of safety of 1.5 to 2 before drawdown, the ratio of major principal stress at time of consolidation sigma 1c to minor principal stress sigma 3c usually has a magnitude of about 2. Because the driving force weight is about double during rapid drawdown, the major and minor principal stresses also about double. A more diagram is a convenient way to represent the stresses existing at a point on the failure surface in the field or in the laboratory triaxial test specimen. A typical more diagram representing stresses at the time of consolidation are shown here. The more diagram is a plot of shear stress as ordinate versus normal stresses as the obsessive. The normal stress and the shear stress occurring at a point on the failure surface at time of consolidation can be obtained directly from a stability analysis for the condition before drawdown. The shear force on the failure surface at time of consolidation is designated as tau fc and the corresponding normal force as sigma fc. The obliquity of the resultant force on the failure surface is alpha. From these two stresses and the friction angle mobilized at the time of failure, phi sub m, the major and minor principal stresses and their ratio can be determined. The mobilized friction angle is the arc tan of the tangent of the effective friction angle phi prime divided by the factor of safety. The concept in limit equilibrium stability analysis is that if the shear strength of the soil is divided by the factor of safety, sliding along the failure surface is on the point of occurring. The angle theta that the failure plane makes with the major principal stress plane can then be computed as 45 degrees plus phi sub m over 2. This angle is shown on the sketch of a test specimen. We're going on a more diagram knowing phi, the major and minor principal stresses can be computed using the geometry of the more diagram. The angle between the lines from the plotted point of tau fc and the major and minor principal stress points form a right angle with each other. The resulting formula for the ratio of major principal stress to minor principal stress is as follows. Case of c defined as sigma 1c over sigma 3c can be computed from the formula 1 plus 2 tan alpha over sin 2 theta divided by 1 minus tan alpha divided by tan theta. Alpha in this case is equal to the arc tan of tau fc divided by sigma fc. The more circles representing the consolidation stresses and the failure stresses from a typical anisotropically consolidated triaxial test are shown here. The shear and normal stresses on the failure plane at time of consolidation are tau fc and sigma fc. The major and minor principal stresses corresponding to these shear and normal stresses are sigma 1c and sigma 3c. The total major and minor principal stresses at the time of failure are shown as sigma 1f and sigma 3f. The pore water pressure which develops in the specimen during loading to failure is uf. The intergranular or effective stresses in the specimen at failure are sigma 1f prime equals sigma 1f minus uf and sigma 3f prime equals sigma 3f minus uf. The failure circle for effective stresses is shown by the dash line. The point where the effective stress circle is tangent to the effective friction angle is the shear strength of the material at failure. Since the normal stress on the failure plane at the time of consolidation, sigma fc is known from the stability analysis before drawdown. It is convenient to plot the shear strengths against this normal stress. This is the circled point on the diagram. A positive pore pressure indicated as uf developed during the testing to failure. Similar more circles for a test specimen at smaller consolidation stresses are shown here. The circle for consolidation and the total and effective stress circles at failure are shown. In contrast to the previous test, the pore water pressure which developed during testing to failure however was negative or tensile. The tensile pore water pressure moves the effective more circle to the right. The circle point represents the shear strength plotted against the normal stress on the failure plane at the time of consolidation, sigma fc. A case where zero pore water pressure occurs at failure is shown here, where the total and effective circles at failure are identical. Also, the minor principal stress at consolidation, sigma 3c, is the same as the minor effective principal stress at failure, sigma 3f. The objective of these tests has been to determine a relationship between undrained shear strength s and normal stress on the failure plane at the time of consolidation, sigma fc. The value of this normal stress is obtained directly from the stability analysis made for before drawdown condition. One other condition must be known as well. That condition is the ratio of the major and minor principal stresses at the time of consolidation, k sub c. Usually, this ratio is about two. The design curve for determining the shear strength values to use in the after drawdown stability analysis consists of two straight lines. For higher sigma f sub c values where positive pore pressures occur at failure, the actual laboratory test results are used as indicated by the line labeled k sub c equals two. Where tensile pore pressure occurs at failure, the strength is reduced to that consistent with zero pore water pressure. It does not seem reasonable to rely on soil strength developed due to tensile pore water pressure. All pore water has gas dissolved in it. If tensile pore water pressure or a significant reduction in compressive pore water pressure occurs, gas probably will come out of solution and effectively destroy any strength attributed to the tensile pore water pressure. Although this k sub c ratio is usually about two, it is convenient to have test results for other k sub c ratios so that the results for various k sub c ratios can be determined by interpolation. Usually isotropically consolidated undrained triaxial shear tests, CU tests are also performed. For purposes of interpolation, it is necessary to plot the results of such CU tests in the same manner as that just described for an isotropically consolidated ACU test. An example of such plotting is shown here. Under isotropic consolidation, sigma 1c equals sigma fc equals sigma 3c. The total stresses at failure are shown by the solid line circle and the effective stresses at failure by the dash line circle. The shear strength is plotted against the normal stress on the failure plane at the time of consolidation, sigma fc. This is in contrast to the usual procedure of plotting the shear strength against the total normal stress on the failure plane at the time of failure, sigma ff, marked by x on the drawing. The undrained shear strength of a soil is directly dependent upon the normal stress on the failure plane and the ratio of major to minor principal stresses at the time of consolidation. It is a consistent and logical way to define the shear strength. The results of ACU tests at a principal stress ratio k sub c equal to 2 performed on the core material for the Emborca Sao Dam in Brazil is shown here. Also shown are the s versus sigma f sub c lines for k sub c equal 1. The steep initial lines represent the condition of no strength being attributed to tensile pore water pressure. The effective stress failure line is shown dashed and for an effective friction angle 5 prime of 35 degrees, its value k sub c equal 3.69. K sub c values for actual cases generally fall in the range of 1.7 to 2.3. Strength values for k sub c values somewhat above and somewhat below k sub c equal 2 can be determined by interpolation with the k sub c equal 1 and the k sub c equal 3.69 lines. It is recommended that the shear strength used to prepare the above mentioned plots of strength versus normal stress on the failure plane at time of consolidation be strength obtained at 15% strain in the triaxial test. Sometimes a higher, that is a peak value, occurs at a much lower percentage strain. Since the assumption in the limit equilibrium stability analysis assumes that the entire failure surface is at the point of failure, the use of the shear strength at 15% strain appears more reasonable than use of a peak value. When the toe of the failure surface experiences enough strain to be at the point of failure, the upper part of the surface will have experienced much more strain due to deformation of the sliding mass. Also, if a peak value were to be used, one would have to be sure that the structure of the soil in the test specimen was truly representative of the structure of the soil in the dam. Thus far we have talked about anisotropically consolidated undrained triaxial tests. Another type of anisotropy is the possible difference in shear strength characteristics in the horizontal direction in the core of the dam in the field as compared to its strength characteristics in the vertical direction. Generally, this possible anisotropy which could result from construction is neglected. Stability analysis is generally carried out by the limit equilibrium method, usually by computer. The basic concept is to determine the factor of safety by which to divide the shear strength of the materials along the failure surface so that sliding is imminent all along the failure surface. In effect, this concept attributes all of the factor of safety to uncertainty regarding the shear strength of the materials. There is also some uncertainty regarding whether the critical failure surface selected is actually the most critical one. And there is some uncertainty in the method used for carrying out the limit equilibrium analysis. The most uncertainty, however, is in connection with the shear strength so that ascribing all of the factor of safety to the shear strength is not too unreasonable. The limit equilibrium analysis is generally carried out by the method of slices. The sliding mass is divided into a series of vertical slices as shown here. Usually, 10 to 15 slices are sufficient. The forces acting on each slice are determined by a trial and error procedure. The forces acting on the two end slices, slice 1 and slice 10, and on a typical central slice are shown here. The forces for slice 1 are the total weight of the slice, W, the water force on the right-hand side of the slice, UR, the water force on the bottom of the slice, UB, the normal earth force on the bottom of the slice, N, the shear force on the bottom of the slice, S, and finally, the earth force on the right-hand side of the slice, ER. Similar forces, W, U left, U right, UB, N, S, E left and E right occur in slice 4. Water forces and earth forces occur on both the left-hand and the right-hand sides of the slice. Slice 10 has no right-hand side and thus no ER. These forces can be analyzed by constructing a force polygon for each slice. Such a polygon for slice 4 is shown. The total weight of the slice is drawn. When the water forces are the left-hand and the right-hand slices and on the bottom are added, the remaining portion of the weight force is the submerged weight. If the drawing of this force polygon starting at slice 1, then the left-hand side earth force will be known and can be drawn. The direction of the P-force makes an angle of Phi sub M with the normal or perpendicular direction to the base of the slice. Phi sub M is the mobilized friction angle. A line in this direction can be drawn from the end of the east of L force. A direction for the E right force is assumed and a line in that direction is drawn from the top of the weight force. The intersection of P and E right closes the polygon. E right then becomes the E left for slice 5. A closed polygon for the slice indicates that the sum of the vertical components of forces on the slice add up to zero and that the sum of the horizontal components of the forces on the slice also add up to zero. The two assumptions made are the direction of the earth forces on the sides of the slice and the factor of safety. The factor of safety F determines the obliquity of the earth force on the base, Phi sub M. For the polygon shown, the shear strength on the base was assumed to be due entirely to friction. If part of the shear strength on the base is due to cohesion, an additional shear force equal to the unit cohesion times the length of the base of the slice and divided by the factor of safety should be added to the frictional shear force shown in the polygon. The check as to whether the assumed factor of safety is the correct one for the sliding mass consists in seeing whether when the equal but opposite force to the lateral earth force determined for the right side of slice 9 is applied to the left side of slice 10, it causes closure of the polygon for slice 10. In this connection, it is convenient to plot the force polygons for the 10 slices so that the earth forces on the sides of the slices overlap. For example, the earth force on the right hand side of slice 3 is equal and opposite to that on the left hand side of slice 4. The drawing showing the overlapping of the force polygons for the 10 slices is shown here. In this case, the force polygon for slice 10 closes perfectly and therefore the assumed factor of safety for this particular set of polygons is the correct factor of safety. A different factor of safety will result if different assumptions are made regarding the earth forces on the sides of the slices. The direction assumed for these forces may vary from horizontal to an obliquity, definitely less than the obliquity, which would occur should sliding failure take place on the sides of the slices. The obliquity may be assumed to be the same on all the interfaces between the slices or it may be assumed to vary from interface to interface. An example developed by Geoslope International shows how the factor of safety varies with a consistent obliquity assumed for all the interfaces. The abscissa scale is the tangent of the assumed obliquity. The corresponding obliquity angle is given also. The ordnance scale is the factor of safety. For zero degrees obliquity, that is a horizontal earth force on the sides of the slices, the factor of safety is 0.93 and for an obliquity of 38.7 degrees it is 1.18. Generally an obliquity of about 22 degrees would be reasonable. The factor of safety at this obliquity is 1.06. Earth forces on the sides of the slices are absolutely necessary in order that full shear resistance be developed along the entire failure surface. The driving moment causing a stability failure comes from the weight of the material overlying the upper part of the failure surface, the shaded areas on the drawing. There is no net driving moment for the unshaded part of the sliding mass. The moments of the areas on either side of a vertical line dropped from the center of the circle balance each other. The earth forces on the sides of the slices modify the earth forces on the bases of the slices so that every part of the sliding surface carries its fair share of the load. Full equilibrium of each slice requires that. The sum of the horizontal components of the forces equals zero. The sum of the vertical components of the forces equals zero. And the sum of the moments of the forces equals zero. In the solution by force polygons previously described, only the sum of horizontal components and the sum of vertical components equals zero. Alternatively, the slices method of analysis can be carried out so that the sum of the moments and the sum of the vertical components both equals zero. If this is done, then a different relationship between factor of safety, f sub m, and the obliquity of earth forces on the sides of the slices is obtained. A graph of such a relationship is shown by this dash line. For comparison, the factor of safety determined by the force polygon method, f sub f, is shown by the solid line. Obviously, there is very much less variation of the moment of factor of safety with obliquity of the earth forces on the sides of the slices than variation of the force factor of safety with obliquity. The Morgenstern price and the Spencer methods of analysis calculate the factor of safety which occurs at the intersection of the two factors of safety lines. Also shown is the point representing the factor of safety as calculated by Bishop's simplified method. He uses the moment equilibrium method and assumes zero obliquity of the earth forces on the sides of the slices. Since the moment factor of safety is not much affected by the assumed obliquity, the Bishop's simplified method gives a close approximation of the factor of safety. The method, however, only applies to circular failure surfaces, while the other two methods can be used for failure surfaces of irregular shape. In the Morgenstern price and Spencer methods, if a composite failure surface is used, a circle may be used to approximate the surface, and the center of this circle used for the point about which moments are summed. This completes my discussion of how I think stability analyses for the rapid drawdown case and the other rapid loading cases should be carried out. Let me summarize the main points. Rapid loading type of stability analysis is required when saturated soils in an embankment have come to equilibrium and are fully consolidated under conditions prevailing before the rapid loading, and then cannot drain to any significant amount during the rapid loading. Rapid loading cases occur for seismic loading, rapid drawdown of the reservoir, and rapid heightening of a dam. Immediately before the application of rapid loadings, the soil along this critical sliding surface is consolidated under anisotropic stress conditions. That is, the major principal stress is greater than the minor principal stress. The ratio of major to minor principal stress usually is about 1.8 to 2. Standard triaxial tests where all around equal pressures are used for consolidation are not suitable to properly determine the shear strength of the soil on the failure surface at the time of rapid loading, anisotropically consolidated triaxial. The anisotropic consolidation conditions prior to rapid loading can be estimated from a stability analysis performed for the conditions prior to rapid loading. This stability analysis provides the normal stress and the shear stress on the failure surface. From these two stresses, the ratio of major to minor principal stress prior to rapid loading can be computed. Alternatively, the ratio of major to minor principal stress can also be obtained from a finite element analysis representing the consolidation condition preceding the rapid loading. The shear strength to use at each point along the failure surface for the rapid loading stability analysis is the strength indicated by anisotropically consolidated triaxial tests. Consolidation is under the stress conditions that exist before rapid loading. At low normal stress, however, any strength due to tensile port water pressures is neglected. Proper ACU triaxial testing requires that the test specimens be 100% saturated. Back pressures of about 100 pounds per square inch generally are required to achieve such saturation. Unfortunately, many soil testing laboratories are not equipped to provide such back pressure. Their test results for CU as well as ACU triaxial tests are therefore questionable. To be applicable, the rapid loading stability analysis which I have described requires two conditions to be met by the impervious soil along the critical failure surface. A, where below ground water table, the impervious soil has become fully saturated and is 100% consolidated under the before drawdown conditions. And B, no significant drainage occurs in the soil during the period of rapid loading. In storage reservoirs, it is unusual to meet these two conditions. It requires that the reservoir be at relatively high level for a long period of time and then is drawn down rapidly because of a shortage of water or in order to permit work to be done in the reservoir area. Pump storage projects meet the rapid drawdown condition but may not meet the condition of saturation and consolidation under high reservoir.