EM 1110-2-2100

1 Dec 05

Chapter 3
Stability Requirements

3-1. General

The concepts used to develop the structural stability requirements contained in this manual are to establish safety
factors or safety provisions for the three prescribed load condition categories of usual, unusual, and extreme such
that the risk of a failure is kept to an acceptably low level and such that performance objectives are achieved. The
use of three different design-load condition categories permits different safety factors or safety provisions to be
assigned to the various load conditions depending on the probability of the load condition occurring during the life
of the structure. The load conditions used in the stability analyses are described on a probabilistic basis, except the
seismic loads and large flood loads falling into the extreme category may be either probabilistic or deterministic.

3-2. Load Condition Categories

The load conditions that a structure may encounter during its service life are grouped into the load condition
categories of usual, unusual, and extreme. Associated with each category is a likelihood that the load condition will
be exceeded in a given time period. The load conditions, expressed in probabilistic terms, are provided in Table 3-1.
The structural performance and the risk of damage or failure depends not only on the likelihood of the loading
condition, but also on the safety factors or the safety provisions used, the degree of conservatism used in selecting
the foundation strength parameters and hydrological data, and the degree of conservatism inherent in the methods
used for the analysis. No attempt has been made to define the likelihood of damage or failure in probabilistic terms.
However, the use of these guidelines in conjunction with other Corps guidance will provide structures with adequate
protection against stability failure.

Table 3-1 Load Condition Probabilities

Load Condition Annual Probability (p) Return Period (tr)

Categories

Usual Greater than or equal to 0.10 Less than or equal to 10 years

Unusual Less than 0.10 but greater than or equal Greater than 10 years but less than or equal
to 0.0033 to 300 years

Extreme Less than 0.0033 Greater than 300 years

• Usual loads refer to loads and load conditions, which are related to the primary function of a structure and
can be expected to occur frequently during the service life of the structure. A usual event is a common
occurrence and the structure is expected to perform in the linearly elastic range.

• Unusual loads refer to operating loads and load conditions that are of infrequent occurrence. Construction
and maintenance loads, because risks can be controlled by specifying the sequence or duration of activities,
and/or by monitoring performance, are also classified as unusual loads. Loads on temporary structures
which are used to facilitate project construction, are also classified as unusual. For an unusual event some
minor nonlinear behavior is acceptable, but any necessary repairs are expected to be minor.

• Extreme loads refer to events, which are highly improbable and can be regarded as emergency conditions.
Such events may be associated with major accidents involving impacts or explosions and natural disasters
due to earthquakes or flooding which have a frequency of occurrence that greatly exceeds the economic

3-1
 EM 1110-2-2100
1 Dec 05

service life of the structure. Extreme loads may also result from the combination of unusual loading events.
The structure is expected to accommodate extreme loads without experiencing a catastrophic failure,
although structural damage which partially impairs the operational functions are expected, and major
rehabilitation or replacement of the structure might be necessary.

Appendix B lists the loading conditions that must be evaluated to ensure the stability of specific structure types. The
loading conditions have been taken from other USACE manuals and may have been modified to be consistent with
other provisions of this manual. When a loading condition is defined in terms of a return period (for example, the
Operational Basis Earthquake is defined as an earthquake with a return period of 144 years), the structural engineer
can determine if the load condition is usual, unusual, or extreme by referring directly to Table 3-1. When a load
condition is stated in non-probabilistic terms, (for example, pool elevation at the top of closed spillway gates, or
water to the top of a flood wall), the return period must be determined to see if that particular load condition is usual,
unusual, or extreme. In some cases, the load condition category is specifically designated based on established
practice, irrespective of any return period (for example, construction is listed as an unusual loading). The engineer
only needs to verify stability for those conditions listed in Appendix B. For example, for the unusual category, it is
not necessary to verify stability for a 300 year flood or earthquake if these are not specifically listed in Appendix B.
Definitions of common loadings for civil works projects are provided in Chapter 4, including: normal operating,
infrequent flood, maximum design flood, probable maximum flood, operational basis earthquake, maximum design
earthquake, and maximum credible earthquake.

3-3. Risk-based Analysis for USACE Flood Project Studies

USACE policy now requires the application of risk-based analysis in the formulation of flood-damage-reduction
projects. The requirements are briefly discussed in the next paragraph to familiarize the structural engineer with the
procedures used by hydrology/hydraulics (H&H) engineers use to develop the degree of protection provided by the
project (i.e., dam height, floodwall height). The structural engineer needs to coordinate with the H&H engineers to
obtain return periods for the required loading conditions to determine the load condition category from Table 3-1.

Risk-based analysis quantifies the uncertainty in discharge-frequency, elevation (stage)-discharge, and elevation-
damage relationships and explicitly incorporates this information into economic and performance analyses of alter-
natives. The risk-based analysis is used to formulate the type and size of the optimal structural (or nonstructural)
plan that will meet the study objectives. USACE policy requires that this plan be identified in every flood-reduction
study it conducts. This plan, referred to as the National Economic Development Plan (NED), is the one that
maximizes the net economic benefits of all the alternatives evaluated. It may or may not be the recommended plan,
based on additional considerations. A residual risk analysis for the NED Plan is next performed to determine the
consequences of exceeding project capacity. For any flood-protection project, it is possible that project capacity
may be exceeded sometime during its service life. Therefore, the question becomes, “If that capacity is exceeded,
what are the impacts, both in terms of economics and the threat to human life?” If the project-induced and/or
residual risk is unacceptable, and a design to reduce the risk cannot be developed, other alternatives are further
analyzed. Either a larger project, that will ensure sufficient time for evacuation, or a different type of project, with
less residual risk, should be selected to reduce the threat to life and property. For a detailed discussion of the H&H
requirements, see ER 1105-2-101 and EM 1110-2-1619.

When the type and size of the project have been selected, detailed design begins. The structural engineer, in coordi-
nation with the hydrology/hydraulic engineers, may use expected values (best estimates) of discharge-frequency and
stage-discharge curves to estimate return periods for the various prescribed structure-dependent hydrostatic load
conditions listed in Appendix B. For load conditions with prescribed water elevations, (for example, water to the top
of closed spillway gates, or water to the top of a flood wall) the headwater elevation may be used in conjunction with
the stage-discharge curve and discharge-frequency curves to estimate the annual probability and return period for the
event representing the load condition. For some projects, such as high pools at power projects, other information
such as project operating data will also be used in estimating the return period for a prescribed loading condition.
The designer then refers to Table 3-1 to determine if each particular load condition is usual, unusual, or extreme.

3-2
 EM 1110-2-2100
1 Dec 05

3-4. Site Information

a. General. A proper stability analysis cannot be performed without knowing the potential planes of weakness
beneath the structure, the strength of the materials along potential planes of weakness, uplift forces that occur on the
structure or on planes of weakness, the strength of backfill materials, and all loads and load conditions to which the
structure may be subjected. Knowledge of geologic formations beneath the structure is also important in defining
seepage conditions and uplift pressures. Without adequate foundation explorations and testing, the safety factors
provided to assess stability of the structure are meaningless. Preliminary stability analyses are useful to identify
design parameters, which require special attention. In some rock foundations there may be many faults, shear zones,
and discontinuities that make it impossible to do little more than predict average shear and cohesive strengths of the
materials that make up the foundation. Use of lower bound values for foundation shear strength or upper bound
values for loads is only acceptable when it can be demonstrated that the added costs to improve the accuracy of the
strength and loading data will not lead to significant savings for the structure or foundation. Lower factors of safety
are permitted by this manual in cases where there is an abundance of information on the various foundation and
structure properties used to establish the strength parameters for the stability analysis. Conversely, higher factors of
safety are required when there is only limited information on either foundation or structure properties. Three
categories of site information, well defined, ordinary, and limited, were used in establishing safety requirements.

b. Well-defined site information. This category is restricted to use for existing projects. To qualify as well
defined, site information must satisfy the following requirements:

• Available records of construction, operation, and maintenance indicate the structure has met all performance
objectives for the load conditions experienced.
• Foundation strengths can be established with a high level of confidence.
• The governing load conditions can be established with a high level of confidence.
• Uplift pressures for design load conditions are known, or can be extrapolated for design load conditions
based on measured uplift pressure data.

c. Ordinary site information. This category applies to most new project designs. To qualify as ordinary, site
information must satisfy the following requirements:

• Foundation strengths have been established with current USACE explorations and testing procedures.
• Foundation strengths can be established with a high level of confidence.
• The governing load conditions can be established with a high level of confidence.

d. Limited site information. This category applies to those new or existing structures designated as normal
(critical structures can not be designed or evaluated based on limited site information), where either of the following
are true:

• Foundation strengths are based on limited or inadequate explorations and testing information, or
• Governing load conditions cannot be established with a high level of confidence because of insufficient
historical data on stream flow, flood potential, etc.

3-5. Critical Structures

Civil works structures, for the purpose of establishing safety factors or safety provisions for use in stability analyses,
are to be designated as either critical or normal. Structures designated as critical are those structures on high hazard
projects whose failure will result in loss of life. Loss of life can result directly, due to flooding or indirectly from
secondary effects. Loss of life potential should consider the population at risk, the downstream flood wave depth
and velocity, and the probability of fatality of individuals within the affected population. Information is provided in
Appendix H to help design engineers determine if the structure should be designated critical or normal.

3-3
 EM 1110-2-2100
1 Dec 05

3-6. Existing Structures

The safety factors provided in this manual are based on the assumption that for critical and normal structures, the
strength of the materials in the foundation and structure has been conservatively established through explorations
and testing. This may not be the case for older existing structures, or, if adequate explorations and testing were
performed, the records may not be available. When the stability of an existing structure is in question, a phased,
systematic approach to evaluating stability should be performed before any remedial actions are undertaken to
improve stability. This systematic evaluation process is described in Chapter 7. The load conditions used to
evaluate an existing structure should be carefully checked to make sure that what was considered as a usual load
condition for the original design is not, once the probabilities of the load conditions are examined, really an unusual
or extreme load condition. Evaluation of existing structures should utilize analytical methods which accurately
describe the behavior without introducing excess conservatism. When available, actual uplift pressures can be used
as a basis for evaluating the stability of existing structures.

3-7. Factors of Safety for Sliding

Analysis of sliding stability is discussed in detail in Chapter 2 and Chapter 5. A factor of safety is required in sliding
analyses to provide a suitable margin of safety between the loads that can cause instability and the strength of the
materials along potential failure planes that can be mobilized to prevent instability. The factor of safety for sliding is
defined by equation 3-1. The required factors of safety for sliding stability for critical structures and for normal
structures are presented in Tables 3-2 and 3-3, respectively.

N tan φ + cL
FSs = (3-1)
T
where

N = force acting normal to the sliding failure plane under the structural wedge.
φ = angle of internal friction of the foundation material under the structural wedge.
c = cohesive strength of the foundation material under the structural wedge.
L = length of the structural wedge in contact with the foundation.
T = shear force acting parallel to the base of the structural wedge.

Table 3-2 Required Factors of Safety for Sliding - Critical Structures

Load Condition Categories

Site Information Category Usual Unusual Extreme

Well Defined 1.7 1.3 1.1

Ordinary 2.0 1.5* 1.1*

Limited** - - -
*For preliminary seismic analysis without detailed site-specific ground motion,
use FS=1.7 for unusual and FS=1.3 for extreme. See further explanation in section 3.11 b.
**Limited site information is not permitted for critical structures

3-4
 EM 1110-2-2100
1 Dec 05

Table 3-3 Required Factors of Safety for Sliding - Normal Structures

Load Condition Categories

Site Information Category Usual Unusual Extreme

Well Defined 1.4 1.2 1.1

Ordinary 1.5 1.3 1.1

Limited 3.0 2.6 2.2

3-8. Factors of Safety for Flotation

A factor of safety is required for flotation to provide a suitable margin of safety between the loads that can cause
instability and the weights of materials that resist flotation. The flotation factor of safety is defined by equation 3-2.
The required factors of safety for flotation are presented in Table 3-4. These flotation safety factors apply to both
normal and critical structures and for all site information categories.

W S +W C + S
FS f = (3-2)
U −W G

where

WS = weight of the structure, including weights of the fixed equipment and soil above the top surface of the
structure. The moist or saturated unit weight should be used for soil above the groundwater table and the
submerged unit weight should be used for soil below the groundwater table.
WC = weight of the water contained within the structure
S = surcharge loads
U = uplift forces acting on the base of the structure
WG = weight of water above top surface of the structure.

Table 3-4 Required Factors of Safety for Flotation – All Structures

Load Condition Categories

Site Information Category Usual Unusual Extreme

All Categories 1.3 1.2 1.1

3-9. Limits on Resultant Location

The factor of safety approach established for sliding and flotation is not appropriate for use in the evaluation of
rotational modes of failure. Rotational behavior is evaluated by determining the location of the resultant of all
applied forces with respect to the potential failure plane. This location can be determined through static analysis.
Limits on the location of the resultant are provided in Table 3-5. The entire base must be in compression for the
usual load condition, to maintain full contact between the structure and the foundation, so there is no chance for
higher uplift pressures to develop in a crack. This helps ensure linear behavior for common loading conditions. For
the unusual load case, higher uplift pressures may develop in a relatively short crack, but this would cause only
minor nonlinear behavior. For extreme load conditions on typical civil works projects, a shear or bearing failure will

3-5
 EM 1110-2-2100
1 Dec 05

occur before overturning could occur. Therefore, the resultant is permitted to be anywhere within the base, and
safety is ensured by the safety factor requirements for sliding and by the limits on allowable bearing stresses.

Table 3-5 Requirements for Location of the Resultant – All Structures

Load Condition Categories

Site Information Category Usual Unusual Extreme
100% of Base in 75% of Base in Resultant
All Categories
Compression Compression Within Base

3-10. Allowable Bearing Capacity

Allowable concrete compressive stresses and/or allowable bearing capacity values established by materials engineers
and geotechnical engineers are used as the basis for evaluating bearing modes of failure. The allowable bearing
capacity value is defined as the maximum pressure that can be permitted on soil or rock giving consideration to all
pertinent factors with adequate safety against rupture of the soil or rock mass, or movement of the foundation of such
magnitude that the structure is impaired. Bearing failure is related to the relative compressibility of the foundation
materials, the loading conditions, the geometry of the structure base, and the strength of the foundation and concrete
at the structure-foundation interface. Bearing capacity may be related to the shear capacity of the foundation
materials or to the deformability of the foundation. Information on foundation bearing analysis can be found in EM
1110-1-1905 for soils, and EM 1110-1-2908 for rock. Safety against bearing failure is generally expressed in terms
of an allowable compressive stress for concrete and an allowable bearing capacity for foundation materials. These
allowables include an adjustment, which represents a factor of safety. The allowable compressive stress and
allowable bearing capacity values are established by testing performed by materials engineers and geotechnical
engineers. Discussion on exploration and testing can be found in Chapter 2. The allowable compressive stress and
bearing capacity values established for usual load conditions can be increased for the unusual and extreme load
conditions. A 15% increase is permitted for unusual load conditions and a 50% increase is permitted for extreme
load conditions.

3-11. Seismic Stability

a. General. Traditionally, the seismic coefficient method has been used to evaluate the stability of structures
subjected to earthquake ground motions, but this method fails to take into account the true dynamic characteristics of
the structure. There have been cases where structures similar to those used on civil works projects have failed
during earthquakes because of a sliding or bearing failure. These failures for the most part are attributable to
liquefaction and soil strength degradation in the foundation or backfill materials. Seismic stability analyses should be
a performed in phases in accordance with requirements of ER 1110-2-1806. Seismic loads to be used in the first
phase analysis are provided in Chapter 5 of this manual. Structures which meet sliding stability factor of safety
requirements when evaluated by this procedure are considered to be safe and no additional seismic stability analyses
are required. Structures that fail to meet factor of safety requirements when evaluated using this procedure should
not be considered unsafe or in need of a stability retrofit. The failure to meet these requirements should only suggest
the need for other seismic coefficient and dynamic analyses to fairly assess the demands placed on the structure and
foundation during a major earthquake. From these advanced analyses engineers can determine if the displacements
and stresses experienced by the structure and foundation will place the structure at risk of a stability failure. In many
instances, it is acceptable for sliding and rocking to occur at the base of the structure during extreme earthquake load
conditions. Stability in such cases is evaluated using dynamic analysis methods, and performance is ensured by
limiting permanent displacements to acceptable levels.

b. Modified Factor of Safety. The factors of safety given in Tables 3-2 include FS=1.5 for unusual and FS = 1.1
for extreme load conditions, for ordinary site information. The ordinary site information and related factor of safety

3-6
 EM 1110-2-2100
1 Dec 05

must be used in the seismic coefficient method. These factors of safety are based on use of extreme loads with very
low probabilities of being exceeded. When factors of safety for seismic loadings are being calculated using the
coefficient method, the MCE loads are usually not based on detailed site-specific seismic data. Since the loads
would be based on less precise data, there would be greater probability that the predicted extreme loads could be
exceeded, therefore, it is appropriate to use higher factor of safety for such analyses. For such analyses, use a factor
of safety of 1.7 for unusual and 1.3 for extreme, as stated in the notes following the above table.

3-12. Mandatory Requirements

For a general discussion on mandatory requirements, see Paragraph 1-5. As stated in that paragraph, certain
requirements within this manual are mandatory. The following are mandatory for Chapter 3.

a. Load condition categories. Unless the loading condition category (usual, unusual, extreme) is specifically
designated in Appendix B, the return period range limitations specified in Table 3-1 shall be used to establish the
correct loading condition designation. When the return period for a particular loading condition can not be
established with sufficient accuracy to determine if the loading condition is usual or unusual (or unusual or extreme),
the loading condition with the more stringent safety requirements shall be used.

b. Critical structures. In accordance with section 3-5, structures on high hazard projects shall be considered
critical where failure will result in loss of life; all other structures will be classified as normal. In making the
determination of critical or normal, the engineer must follow the guidelines in Appendix G.

c. Site information. Structures shall be assigned to one of three site information categories: well-defined,
ordinary, or limited. Site information category selection shall be in accordance with the provisions of Paragraph 3-4.

d. Sliding stability. Sliding stability factors of safety shall be equal to, or greater than, the values specified in
Tables 3-2 and 3-3. The sliding stability factor of safety shall be determined using Equation 3-1.

e. Flotation stability. Flotation factors of safety shall be equal to, or greater than, the values specified in Table 3-
4. The flotation stability factor of safety shall be determined using Equation 3-2.

f. Resultant location. The location of the resultant of all forces acting on the base of the structure shall be within
the limits specified in Table 3-5.

g. Bearing pressures. Bearing pressures for usual load conditions shall be within allowable limits established by
the geologist/geotechnical engineer. Increases in allowable bearing pressures shall not exceed 15% for unusual and
50% for extreme load conditions, in accordance with the guidance in section 3-10.

h. Loading conditions. As a minimum, the loading conditions provided in Appendix B shall be satisfied in the
stability analysis.

i. Loads. Loads shall comply with the mandatory requirements of Chapters 4 and 5.

3-7
 EM 1110-2-2100
1 Dec 05

Chapter 4
Loads and Loading Conditions

4-1. General

Previously, stability criteria was provided in separate manuals for each type of structure. Those manuals listed all of
the load cases (loading conditions), which had to be investigated as part of the stability analysis. Those loading
conditions are now summarized in tables provided in Appendix B of this manual. The tables list the loading
condition and give a classification as usual, unusual, or extreme, as defined in Table 3-1. Following each table are
brief descriptions of the loading conditions. The loading conditions have been revised in some cases for general
consistency with the provisions of this manual, especially to comply with current practice for flood and seismic
loadings. This chapter defines most of the types of loads that are combined to form each loading condition.
However, soil loads are defined in Chapter 2 for multiple wedge sliding analyses and in Chapter 5 for single wedge
sliding analyses.

4-2. Construction

Based on past practice, construction loading conditions shall be classified as unusual, regardless of duration.

4-3. Water Loading Conditions

a. General. All water loading conditions should be based on hydrologic information, which gives median water
elevations in terms of return periods. A typical flood hazard curve is illustrated in Figure 4-1. Curves for both
headwater and coincident tailwater will be necessary to determine the water loads for dams and navigation locks.
Hydraulic engineers commonly use the 90-percent confidence level hazard curve when determining flood protection
requirements. However, for stability analysis, structural engineers require median flood hazard curves, which can
also be provided by the hydraulic engineers. Based on the information presented in Figure 4-1 a flood pool elevation
equal to 21 meters (68.9 feet) would be used to determine the maximum unusual loading.

30

90 % C onfidence Level
21 meters
20
Pool
Elevation Median
(Meters)
10
300 years

U susal U nususal Extrem e

0
1 10 100 1000 10,000
Return Period (Y ears)

Fig ure 4-1 Floo d H aza rd Curv e

b. Coincident pool. Coincident pool represents the water elevation that should be used for combination with
seismic events. It is the elevation that the water is expected to be at or below for half of the time during each year.

4-1
 EM 1110-2-2100
1 Dec 05

c. Normal operation. In the past, a normal operation loading condition has been used to describe loadings with
various probabilities of occurring, including rare events with long return periods. To be consistent with Table 3-1,
normal operating conditions are now defined as maximum loading conditions with a return period of no more than
10 years (annual probability of 10%). For certain floodwalls, this means that there might be no water loads on the
structure for normal operation. For hydropower dams, the pool will be fairly high for normal operation, while for
some flood-control dams, the pool will be low for normal operation. For navigation projects, the maximum loading
for normal operation might correspond to the usual navigation pool, combined with the lowest tailwater expected
with a 10-year return period. Water loads defined by the normal operation loading condition are sometimes
combined with other types of events (such as barge impacts).

d. Infrequent flood. The infrequent flood (IF) represents flood pool or water surface elevations associated with
events with a return period of no greater than 300 years (annual probability of 0.33%), making the IF an unusual
loading per Table 3-1. This loading condition replaces loadings such as water to top of spillway gates and water to
spillway crest previously used for the design and evaluation of gated and ungated spillways. It also replaces the
design flood (top of wall less freeboard) used for the design and evaluation of floodwalls. In limited cases, historical
hydrologic data may be inadequate to determine the 300-year water elevations with reasonable certainty. In such
cases, traditional loading conditions such as water to top of spillway gates, water to spillway crest, and design flood
shall be considered unusual events and evaluated in addition to the IF event.

e. Probable maximum flood. The probable maximum flood (PMF) is one that has flood characteristics of peak
discharge, volume, and hydrograph shape that are considered to be the most severe reasonably possible at a
particular location, based on relatively comprehensive hydro-meteorological analyses of critical runoff-producing
precipitation, snow melt, and hydrologic factors favorable for maximum flood runoff. The PMF load condition
represents the most severe hydraulic condition, but because of possible overtopping and tailwater effects, it may not
represent the most severe structural loading condition, which is represented by the maximum design flood described
below. Therefore, the PMF condition will not necessarily be examined for structural stability.

f. Maximum design flood. The maximum design flood (MDF) is the designation used to represent the maximum
structural loading condition (as judged by the minimum factor of safety) and must be determined for each structure
or even for each structural element. MDF may be any event up to PMF. For floodwalls, MDF is usually when the
water level is at or slightly above the top of the wall. Overtopping from higher water levels would result in rising
water levels on the protected side, thus reducing net lateral forces. The same situation may be true for dams, but
often significant overtopping can occur without significant increases in tailwater levels. The design engineer must
consult with the hydraulics engineer to explore the possible combinations of headwater and tailwater and their
effects on the structure. Some elements of dam outlet works (such as chute walls or stilling basins) are loaded
differently from the main dam monoliths. For such elements, different flow conditions will produce maximum
structural loading. When it is not obvious which loading will produce the lowest factor of safety, multiple loadings
should each be investigated as a possible MDF. Since sliding is the most likely mode of failure for most gravity
structures, MDF can usually be judged by determining maximum net shear forces. However, due to variable uplift
conditions, a loading with smaller shears could result in the lowest factor of safety. Once the MDF is determined, it
should be classified as usual, unusual, or extreme per Table 3-1, based on its return period.

4-4. Uplift Loads

Uplift loads have significant impact on stability. Sliding stability, resultant location, and flotation are all aspects of a
stability analysis where safety can be improved by reducing uplift pressures. Since uplift pressures are directly
related to flow paths beneath the structure, uplift pressure distribution may be determined from a seepage analysis.
Such an analysis must consider the types of foundation and backfill materials, their possible range of horizontal and
vertical permeabilities, and the effectiveness of cutoffs and drains. Techniques for seepage analysis are discussed in
EM 1110-2-1901, EM 1110-2-2502, Casagrande (1937), Cedergren (1967), Harr (1962), and EPRI (1992). Seepage
analysis techniques to determine uplift pressures on structures include flow nets, finite element methods, the line-of-
creep method, and the method of fragments. Uplift pressures resulting from flow through fractures and jointed rock,
however, are poorly understood and can only be accurately known by measurements taken at the point of interest.
Joint asperities, changes in joint aperture, and the degree to which joints interconnect with tailwater influence uplift
pressures and pressure distribution. Uplift pressures are site-specific and may vary at a given site due to changes in

4-2
 EM 1110-2-2100
1 Dec 05

geology. Uplift pressures can be reduced through foundation drainage, or by various cutoff measures such as grout
curtains, cutoff walls, and impervious blankets. Uplift pressures should be based on relatively long-term water
elevations. Short duration fluctuations, such as from waves or from vibrations due to high velocity flows, may be
safely assumed to have no effect on uplift pressures. Uplift pressures to be used for stability analysis of new
structures are covered in Appendix C. The conservative uplift pressures used for the design of new structures may be
significantly higher than those the actual structure may experience during its lifetime. For this reason, the use of
actual uplift pressures for the evaluation of existing structures is permitted under the provisions discussed in Chapter
7. However, the engineer should be aware that in some instances the actual uplift may not be reflected by uplift cell
readings. Since uplift measurement devices only capture a snapshot of a given part of the foundation, they should be
used with caution, based on an overall evaluation of the foundation.

4-5. Maintenance Conditions

The return periods for a maintenance condition loading may be greater or less than 10 years, but based on past
experience maintenance has been designated as an unusual load condition. The classification as an unusual loading
is based on the premise that maintenance loadings take place under controlled conditions and that the structure
performance can be closely monitored during maintenance.

4-6. Surge and Wave Loads

a. General. Surge and wave loads are critical in analyzing the stability of coastal protection structures but
usually have little effect on the stability of inland structures. Wave and water level predictions for the analysis of
structures should be based on the criteria presented in the Shore Protection Manual 1984, EM 1110-2-1612, and EM
1110-2-1614. Design forces acting on the structure should be determined for the water levels and waves predicted
for the most severe fetch and the effects of shoaling, refraction, and diffraction. The methods recommended for
calculation of wave forces are for vertical surfaces. Wave forces on other types of surfaces (sloping, stepped,
curved, etc.) are not sufficiently understood to recommend general analytical design criteria. In any event, the
structural engineer should consult with a coastal engineer in establishing wave forces for the design of critical
structures.

b. Wave heights. Wave heights for design are obtained from the statistical distribution of all waves in a wave
train and are defined as follows:

HS = average of the highest one-third of all waves

H1 = 1.67 HS = average of highest 1 percent of all waves

Hb = height of wave which breaks in water depth db

c. Non-breaking waves. When the water depth is greater than approximately 1.5 times the wave height , waves
do not break. The H1 wave shall be used for the non-breaking condition. Design pressures for non-breaking waves
shall be computed using the Miche-Rudgren method. Whenever the maximum stillwater level results in a non-
breaking condition, lower stillwater levels should be investigated for the possibility that shallow water may produce
breaking wave forces, which are larger than the non-breaking forces.

d. Breaking waves. Waves break when the steepness of the wave and the bottom slope at the front of the
structure have certain relationships to each other. It is commonly assumed that a wave will break if the water depth
is not greater than 1.3 times the wave height. Study of the breaking process indicates that this assumption is not
always valid. The height of the breaking wave and its breaking point are difficult to determine, but breaker height
can equal the water depth at the structure, depending on bottom slope and wave period. Detailed determination of
breaker heights and distances for a sloping approach grade in front of the structure are given in the Shore Protection
Manual 1984. Design breaking wave pressure should be determined by the Minikin method presented in EM 1110-2-
1614. Breaking-wave impact pressures occur at the instant the vertical face of the wave hits the structure and only
when a plunging wave entraps a cushion of air against the structure. Because of this dependence on curve geometry,
high impact pressures are infrequent against prototype structures; however, they must be recognized as possible and

4-3
 EM 1110-2-2100
1 Dec 05

must be considered in design. Also, since the impact pressures caused by breaking waves are of very short duration,
their importance in design against sliding and rotational instability may be questionable relative to longer lasting,
smaller dynamic forces.

e. Broken waves. Broken waves are those that break before reaching the structure, but near enough to have
retained some of the forward momentum of breaking. The design breaker height in this case (Hb) is the highest wave
that will be broken in the breaker zone. Design wave forces for this height should be determined by the method
presented in Chapter 7 of the Shore Protection Manual (1984).

4-7. Earthquake Loading Conditions

a. Seismic Load Conditions. Earthquake loads are used to represent the inertial effects attributable to the
structure mass, the surrounding soil (dynamic earth pressures), and the surrounding water (hydrodynamic pressures).
Design earthquakes shall comply with requirements of ER 1110-2-1806, based on the following seismic events.

• Operational basis earthquake (OBE). The OBE is considered to be an earthquake that has a 50 percent
chance of being exceeded in 100 years (or a 144-year return period).

• Maximum design earthquake (MDE). The MDE is the maximum level of ground motion for which a
structure is designed or evaluated. For critical structures the MDE is the same as the maximum credible
earthquake (MCE). Generally, the probabilistically determined MDE for other structures is an earthquake
that has a 10 percent chance of being exceeded in a 100-year period (or a 950-year return period).

• Maximum Credible Earthquake. The MCE is defined as the greatest earthquake that can reasonably be
expected to be generated on a specific source, on the basis of seismological and geological evidence. The
MCE is based on a deterministic site hazard analysis.

Earthquake-generated inertial forces associated with the OBE are unusual loads. Those associated with the MDE are
extreme loads. Earthquake loads are to be combined with other loads that are expected during routine operations,
and should not be combined with other infrequent events such as flood loads. Seismic loads should be combined
with coincident pool, which is defined as the elevation that the water is expected to be at or below for half of the
time during each year.

b. Analytical Methods. Several analytical methods are available to evaluate the dynamic response of structures
during earthquakes: seismic coefficient, response spectrum, and time-history. These methods are discussed in
reference ER 1110-2-1806. The current state-of-the-art method uses linear-elastic and nonlinear finite element time
history analysis procedures, which account for the dynamic interaction between the structure, foundation, soil, and
water. The seismic coefficient method, although it fails to account for the true dynamic characteristics of the
structure-water-soil system, is accepted as a semiempirical method for determining if seismic forces control the
design or evaluation, and to decide if dynamic analyses should be undertaken. The information in the following
paragraphs describes the differences between the seismic coefficient method and dynamic analysis methods. Figure
4-2 illustrates the differences in the inertial and hydrodynamic earthquake loads obtained by the two different
methods. The seismic coefficient used for the preliminary seismic stability evaluation of concrete hydraulic
structures should be equal to 2/3 the effective peak ground acceleration (EPGA) expressed as a decimal fraction of
the acceleration of gravity. The EPGA can be obtained by dividing the 0.30 second spectral acceleration, for the
return period representing the design earthquake, by a factor of 2.5 The 0.30 second spectral acceleration is
obtained from the spectral acceleration maps in Appendix D of ER 1110-2-1806.

c. Inertia force due to structure mass. In the seismic coefficient approach, the inertial force is computed as the
product of the mass of the structural wedge (including the soil above the heel or toe and any water contained within
the structure) and the seismic acceleration. This may also be expressed as the weight of the structural wedge times
the seismic coefficient, expressed as a fraction of gravity.

Fh = m a = kh W (4-1)

4-4
 EM 1110-2-2100
1 Dec 05

where: Fh = horizontal component of the inertial force (a similar equation can be used for vertical component)
m = mass of structural wedge
a = seismic acceleration
W = gross weight of structural wedge (including soil above the heel and toe, and water contained within
the structure)
kh = seismic coefficient = a / g
g = acceleration of gravity

Figure 4-2 Idealized earthquake loads

The horizontal component of the inertial force is assumed to act at the center of mass of the structure, based on the assumption
that the structure is a rigid body. In actuality, almost all structures have some flexibility, and the use of the rigid body concept
often under estimates the magnitude of the inertial force. The location of the horizontal inertial force is also related to the
flexibility of the structure, and usually acts at a location higher than the center of mass. However, because of the cyclic nature
of earthquake loads, there is little probability of a rotational-stability related failure.

d. Inertial effects of soil. Backfill material adjacent to a structure will induce inertial forces on the structure during an
earthquake. See Chapter 5 and Appendix G for information on soil loads due to earthquakes.

e. Effects of water. Water that is above the ground surface and adjacent to, or surrounding a structure will increase the
inertial forces acting on the structure during an earthquake. The displaced structure moves through the surrounding water
thereby causing hydrodynamic forces to act on the structure. The water inside and surrounding the structure alters the dynamic
characteristics of the structural system, increasing the periods of the fundamental modes of vibration and modifying the mode
shapes. In seismic coefficient methods, the hydrodynamic effects are approximated by using the Westergaard method
(equation 4-2) (Westergaard 1933). The hydrodynamic force can either increase or decrease the water force, depending on
direction of seismic acceleration. Figure 4-3 illustrates hydrodynamic pressures based on the Westergaard method.

PE = (7/12) kh γw h2 (4-2)

4-5
 EM 1110-2-2100
1 Dec 05

where: PE = hydrodynamic force per unit length

kh = horizontal seismic coefficient
γw = unit weight of water
h = water depth.
The hydrodynynamic force is added algebraically to the static water pressure force to get the total water force on the structure.
The pressure distribution is parabolic and the line of action for the force PE is 0.4 h above the ground surface. The Westergaard

Figure 4-3 Hydrodynamic Forces for Freestanding Water

method assumes the structure is rigid and the water is incompressible. Since most structures are flexible, this method can lead
to significant error. For free-standing intake towers, the hydrodynamic effects are approximated by adding mass to the
structure to represent the influence of the water inside and surrounding the tower. Engineers using the seismic coefficient
approach for stability analyses should be aware of the limitations and the simplifying assumptions made with respect to
hydrodynamic pressures and their distribution on the structure.

4-8. Other Loads

a. Impact. Impact loads for locks and dams on navigation systems are due to the structures being struck by barges.
These loads can be quite large and for some structures, such as lock guide walls, control the stability analyses. Where impact
loads must be considered, refer to EM 1110-2-2602.

b. Ice. Loads due to ice are usually not critical factors in the stability analysis for hydraulic structures. They are more
important in the design of gates and other appurtenances. Ice damage to gates is quite common, but there is no known case of
a dam failure due to ice. Where ice loads must be considered, refer to EM 1110-2-1612.

c. Debris. Debris loads, like ice loads, are usually of no consequence in stability analyses. However, they may be
critical for the design of gates and floodwalls.

d. Hawser pull. Hawser pulls from barges are significant in the stability analysis for lock guide walls, mooring
facilities, and floodwalls. Where hawser pulls must be considered, refer to EM 1110-2-2602.

4-6
 EM 1110-2-2100
1 Dec 05

e. Wind. Wind loads are usually small in comparison to other forces, which act on civil works structures. Therefore,
wind loads should usually be ignored. For structures such as coastal flood walls where wind might cause instability, or for
structures under construction, wind pressures should be based on the requirements of ASCE 7

f. Silt. Silt accumulation can occur upstream of dams. Not all dams will be susceptible to silt accumulation and the
structural engineer should consult with hydraulic engineers to determine if silt buildup is possible, and to what extent it may
accumulate over time. Silt loads should be included in the loading conditions indicated in Appendix B. Horizontal silt
pressure is assumed to be equivalent to that of a fluid weighing 1362 kg/m3 (85 pcf). Vertical silt pressure is determined as
if silt were a soil having a wet density of 1922 kg/m3 (120 pcf). These values include the effects of water within the silt.

4-9. Mandatory Requirements.

For a general discussion on mandatory requirements, see Paragraph 1-5. As stated in that paragraph, certain requirements
within this manual are mandatory. The following are mandatory for Chapter 4.

a. Load Conditions. Stability shall be satisfied for all load conditions listed in Appendix B.

b. Maintenance. Maintenance load conditions shall be classified as unusual.

c. Water loading conditions. Water loadings shall be based on hydrologic analyses giving median water elevations in
terms of return periods. Water elevations for various load conditions shall be as follows:

• Coincident pool shall be the elevation that the water is expected to be at or below for half of the time during each
year. This water loading shall be used in combination with seismic loads.

• Normal operation loading shall represent maximum loads with a 10-year return period.

• Infrequent flood shall represent maximum loads with a 300-year return period.

• Maximum design flood shall be the maximum structural loading up to PMF.

d. Uplift loads. Uplift loads shall be calculated per the requirements of Appendix C, as follows:

• Uplift pressures shall be calculated based on an approximate seepage analysis and shall be applied over the full area
of the base of the structure, or the failure plane under investigation.

• When a loss of contact is calculated to occur at the heel of the structure, full uplift pressure due to headwater shall be
assumed to exist in this area. This provision does not apply to earthquake loading conditions.

• The maximum assumed effectiveness of drainage systems, cutoff wall systems, and combined drain and cutoff wall
systems shall be 50%.

• Where overflow results in significant velocities and causes hydraulic jump and retrogression, tailwater pressures used
in uplift calculations shall be reduced as described in Appendix C.

e. Earthquake loads.

• Earthquake loads shall be based on design earthquakes specified in ER 1110-2-1806.

• Structural inertia loads shall be calculated using the seismic coefficient method.

• Hydrodynamic loads shall be calculated using Westergaard’s formula.

• Soil loads for seismic events shall be calculated per the requirements of Chapter 5.

4-7
 EM 1110-2-2100
1 Dec 05

Chapter 5
Soil Forces and Single-Wedge Sliding Analysis

5-1. General

Chapter 4 described various loading conditions and specific loads, except for soil loads. Chapter 5 describes soil
loads and explains how to use various loads in a single-wedge stability analysis. The methods presented in this
chapter are intended to produce reasonably conservative estimates of soil forces acting on a structure. This manual
only addresses normal soil conditions, other conditions such as swelling soils require special studies. The definitions
of terms that will be used throughout this chapter are as follows:

• Single wedge. The single wedge is the wedge to which forces are applied, i.e., the structure itself, which is
referred to as the structural wedge.

• Applied driving forces. Driving forces are defined as those lateral forces whose primary influence is to
decrease structural stability. The side of the structure upon which these forces are applied will be called the
driving side. Uplift and downdrag are also treated as applied forces.

• Applied resisting forces. Resisting forces are defined as those lateral forces whose primary influence is to
increase structural stability. The side of the structure upon which these forces are applied will be called the
resisting side. The resisting side is on the opposite side of the structural wedge from the driving side. The
difference between the driving and resisting forces is transferred to the foundation by the structural wedge.

• Reactions. The shear and the normal force between the foundation and the base of the single wedge are
reactions, which are necessary to place the structure in static equilibrium, they are not included in the
applied forces.

5-2. Single-Wedge Stability Analyses

a. Basic requirements. For the single-wedge analysis, the engineer must calculate the driving and resisting soil
forces, lateral water forces, and uplift and apply them to the structural wedge. The vertical drag force discussed in
Appendix F may also be included, if the requirements stated in the appendix are satisfied. Using these forces and the
weight of the structural wedge, the engineer can determine the magnitude, location, and slope of the resultant acting
at the base of the structural wedge. The resultant will be used to evaluate sliding stability, and the location of the
resultant will be used to determine the potential for partial loss of contact between the structure and the foundation
materials. The resultant and its location will also be used to evaluate the bearing capacity of the foundation
materials. The forces applied to the structural wedge will also be used for design of the structural elements (e.g.,
shears, moments and axial loads in the base and stem of a retaining wall).

b. Soil forces. Lateral soil forces acting on the single wedge should be calculated using the minimum required
factor of safety against sliding to obtain the developed soil strength parameters φd and cd. The use of developed
parameters results in a force larger than the active soil force on the driving side, and a force smaller than the passive
soil force on the resisting side These developed parameters shall be used in Equations 5-3 through 5-15 to calculate
the lateral soil forces acting on the driving side, and used in Equations 5-16 through 5-22 to calculate the lateral soil
forces acting on the resisting side. Soil forces due to seismic events are discussed in Section 5-5. If there is a
significant difference between the calculated and minimum required safety factors, it may be appropriate to re-
evaluate the developed soil strength parameters used to determine the soil forces. The values for the developed soil
strength parameters shall be determined as:

⎛ tan φ ⎞ c
φ d = tan - 1 ⎜⎜ ⎟⎟ and cd = (5 - 1, 5 - 2)
⎝ FS s ⎠ FS s

where FSs = required factor of safety against sliding

5-1
 EM 1110-2-2100
1 Dec 05

φ = nominal angle of internal friction of the soil

c = nominal cohesive strength of the soil

c. Sliding. The resultant can be resolved into components parallel and normal to the base plane of the
structural wedge. The sliding factor of safety is calculated as follows:

N tan φ + c L
FS s = (5-3)
Τ

where: N = the component of the resultant normal to the base

T = the component of the resultant parallel to the base

L = length of base in compression

If the safety factor is equal to or greater than the required safety factor, the sliding stability criterion is satisfied.
Note that this calculated safety factor might not be equal to the minimum factor used to determine developed soil
strength parameters. If there is a significant difference between the calculated and minimum required safety factors,
it may be appropriate to re-evaluate the developed soil strength parameters used to determine the soil forces. The
resultant location shall be used to determine the length of the base that is in compression, and cohesion shall not be
effective on that part of the base that is not in compression. If there is any loss of contact, uplift forces should be re-
evaluated. However, for seismic events, cyclic loading periods are so short that the time is not sufficient for uplift
pressures to change, therefore, uplift should not be increased due to loss of contact for seismic loads.

5-3. Soil Pressures and Forces

a. Active soil pressures.

(1) Cohesionless backfill. Cohesionless materials such as clean sand are the recommended backfill for most
structures. Large-scale tests (Terzaghi 1934; Tschebatarioff 1949; Matsuo, Kenmochi, and Yagi 1978) with
cohesionless backfills have shown that lateral pressures are highly dependent on the magnitude and direction of wall
movement. The minimum lateral pressure condition, or active soil pressure, develops when a structure rotates about
its base and away from the backfill an amount on the order of 0.001 to 0.005 radians. As the structure moves,
horizontal stresses in the soil are reduced, and vertical stresses due to backfill weight are resisted by increasing shear
stresses until shear failure is imminent (Figures 5-1 and 5-2).

(2) Cohesive backfill. For situations where cohesive backfill is unavoidable, solutions are included herein for
soil pressures involving both frictional and cohesive soil strength parameters (φ and c). Where cohesive backfill is
used, two analyses (short-term and long-term) are usually required in order to model conditions that may arise during
the life of the structure. Short-term analyses model conditions prevailing before pore water pressure dissipation
occurs, such as the end-of-construction condition. Unconsolidated-undrained test parameters, which yield a
relatively high cohesion value and a low or zero friction value, are appropriate for short-term analyses. Long-term
analyses model conditions prevailing after shear-induced pore pressures have dissipated. For long-term analyses,
consolidated-drained test parameters are appropriate. These tests usually yield a relatively high value for internal
friction and a low or zero value for cohesion.

b. Passive soil pressures. If a structure is moved toward the backfill, lateral soil pressures increase and shear
stresses reverse direction, first decreasing and then increasing to a maximum at failure (Figure 5-4). Full
development of passive pressure requires much larger structure rotations than those required for the active case, as
much as 0.02 to 0.2 radians (Figures 5-1 and 5-2). However, the rotation required to develop one-half of the passive
pressure is significantly less, as little as 0.005 radians. The designer must be certain that soil on the resisting side of

5-2
 EM 1110-2-2100
1 Dec 05

any structure will always remain in place and not be excavated or eroded before its effect is included in the stability
analyses.

c. At-rest soil pressure. If no structural movement occurs, then the at-rest condition exists.

d. Design soil pressures - driving side. In practice, the active and passive soil pressure conditions seldom
exist. Hydraulic structures are designed using conservative criteria that results in relatively stiff structures.
Structures founded on rock or stiff soils usually do not yield sufficiently to develop active pressures. Even for
foundations capable of yielding, experiments with granular backfill (Matsuo, Kenmochi, and Yagi 1978) indicate
that following initial yield and development of active pressures, lateral pressures may in time return to greater
values. Another reference (Casagrande 1973) states that the gradual buildup of the backfill in compacted lifts pro-
duces greater-than-active pressures, as do long-term effects from vibrations, water level fluctuations, and
temperature changes. For these reasons and because large rotations are required for the development of passive
pressures, soil pressures on both the driving side and the resisting side of the single wedge will be estimated by using
the developed soil strength parameters, as defined in paragraph 5-2. These parameters are then used to calculate the
equivalent-fluid soil pressure coefficients (K).

(1) General wedge method for equivalent fluid pressure coefficients. Lateral soil forces are assumed to
act parallel to the top surface of driving side wedges when the surface slopes downward toward the structure.
Equivalent fluid-pressure coefficients are calculated as follows:

⎛ + C 12 + 4 C 2 ⎞⎟
α = tan -1 ⎜ C 1 (5-4)
⎜ 2 ⎟
⎝ ⎠

where α = the critical slip plane angle for the soil wedge (see Appendix E for a derivation of α)

⎛ 2V ⎞ ⎛ 2 cd ⎞
t + ⎜⎜ ⎟⎟(1+ tan 2 φ d ) tan 2 β + ⎜⎜ ⎟r
γ ( h2 - d c2 ⎝ γ (h + d c ) ⎟⎠
= ⎝
)⎠
C2
A

⎛ 2V ⎞ ⎛ 2 cd ⎞
A = tan φ d + tan δ - ⎜⎜ 2 ⎟
⎟(1 + tan 2 φ d ) + ⎜⎜ ⎟⎟ r
⎝γ ( h - dc )⎠ ⎝ γ (h + d c ) ⎠
2

r = 1 - tanδ tanφd - tanβ (tanδ + tanφd)

s = tanβ + tanφd + tanδ (1 - tanβ tanφd)

t = tanφd - tanβ - (tanδ + tanβ) tan2φd

φ = soil internal friction parameter

φd = developed internal friction parameter

c = soil cohesion parameter

cd = developed cohesion parameter

ß = top surface slope angle, positive when slope is upward when moving away from the structure (When
the top surface of the backfill is broken, solutions for α may be obtained by using analogous positive and
negative strip surcharges.)

5-3
 EM 1110-2-2100
1 Dec 05

δ = wall friction angle. When β is positive, δ = β. When β is zero or negative δ = 0. Vertical shear
(drag), as discussed in Appendix F, shall not be used to calculate the value of α or equivalent fluid
soil pressure coefficients. However, drag may be used in addition to lateral soil pressures when the
requirements of Appendix F are satisfied.

γ = average unit weight of soil (moist weight above water table, buoyant weight below)

V = strip surcharge

h = height of vertical face of soil wedge

dc = depth of cohesion crack in soil (should always be assumed filled with water when calculating lateral
forces)

The equivalent fluid-pressure coefficients are:

1 - tan φ d cot α
K=
cos δ [(1 - tan δ tan φ d ) + ( tan φ d + tan δ ) tan α ]

and for soils that possess cohesive properties as well as internal friction:

1
Kc=
2 cos α ( tan α - tan β ) [1 - tan δ tan φ d + ( tan δ + tan φ d ) tanalpha ]
2

The equation for the depth of a cohesive crack is:

2 Kc c
dc=
⎛ tan α ⎞
K γ ⎜⎜ ⎟⎟
⎝ tan α - tan β ⎠

And the total lateral soil force is calculated as:

1 ⎛ tan α ⎞
P= Kγ ⎜⎜ ⎟⎟ (h - d c )2 + K V tan α
2 ⎝ tan α - tan β ⎠

Examples are presented in Appendix D.

When any of the variables in the above equations are not present in a particular problem, they are set equal to zero,
thereby simplifying the equations. Figure 5-3 illustrates a wedge containing cohesionless soil, showing the methods
used to calculate the lateral force, and the soil pressure at any point on the vertical face of the wedge. Figure 5-4
shows similar information for a wedge consisting of a cohesive soil. Figure 5-5 shows the method used to determine
the pressure distribution for a strip surcharge (a line load V).

When ß is greater than φd a solution for α cannot be obtained from Equation 5-4 because the number under the
radical will be negative, making the square root indeterminate. However, when ß is equal to φd, Equation 5-4 will
give a value for α equal to ß and φd. When α = φd = ß, the total lateral soil force, for a granular soil not supporting a
strip surcharge, is:

5-4