12 Design of Revetments, Seawalls and Bulkheads: Bank Protection

Ref: Shore Protection Manual, USACE, 1984

Coastal Engineering Manual, USACE
EM 1110-2-1614, Design of Revetments, Seawalls and Bulkheads, USACE, 1995
Breakwaters, Jetties, Bulkheads and Seawalls, Pile Buck, 1992
Coastal, Estuarial and Harbour Engineers' Reference Book, M.B. Abbot and W.A. Price,
1994, (Chapter 27)
Topics
Definitions and Descriptions of Bank Protection and Earth Retention Structures
Design Considerations
Shoreline Use
Shoreline Form and Composition
Seasonal Variations of Shoreline Profiles
Conditions for Protective Measures
Design Water Levels
Design Wave Estimation, Wave Height and Stability Considerations
Breaking Waves
Height of Protection; Wave Runup & Overtopping
Stability and Flexibility
Bulkhead Line
Equations for Armoring and Riprap
Reserve Stability
Toe Protection
Filters
Flank Protection
Material Hazards and Problems
--------------------------------------------------------------------------------------------------------------------

Definitions and Descriptions of Bank Protection and Earth Retention Structures

Bank protection and earth retention structures differ from breakwaters mainly in that they are
constructed against land. Therefore, the earth pressure is a main design concern. These structures
could be vertical, such as bulkheads, or sloped, such as revetment, levees and dikes. Seawalls may
be vertical, but may also be curved, sloped or stepped.
Seawall
• massive structure
• primarily designed to resist wave action along high value coastal property
• either gravity- or pile-supported structures
• concrete or stone.
• variety of face shapes (figures 1)

1
¾ Curved Face - designed to accommodate the impact and runup of large waves while
directing the flow away from the land being protected. Flow strikes the wall Æ
forced along the curving face Æ falls harmlessly back to the ground, or it is recurved
to splash back seaward. Large wave forces must be resisted and redirected. This
requires a massive structure with an adequate foundation and sturdy toe protection.
¾ Stepped Face - designed to limit wave runup and overtopping. They are generally
less massive than curved-face seawalls, but the general design requirements for
structural stability are the same for this kind of structure.
¾ Combination - incorporates the advantages of both curved and stepped face seawalls.
¾ Rubble - essentially a rubble breakwater that is placed directly on the beach. The
rough surface tends to absorb and dissipate wave energy with a minimum of wave
reflection and scour.

Seawall Alternatives

2
Revetment
• facing of erosion resistant material, such as stone or concrete
• built to protect a scarp, embankment, or other shoreline feature against erosion
• major components: armor layer, filter, and toe (see figure 2)
• armor layer provides the basic protection against wave action
• filter layer supports the armor, allows water to pass through the structure and prevents the
underlying soil from being washed through the armor
• toe protection prevents displacement of the seaward edge of the revetment

Typical revetment

3
Bulkheads
• retaining walls which hold or prevent backfill from sliding
• provide protection against light-to-moderate wave action
• used to protect eroding bluffs by retaining soil at the toe and increasing stability, or by
protecting the toe from erosion and undercutting
• used for reclamation projects, where a fill is needed seaward of the existing shore
• used in marinas and other structures where deep water is needed directly at the shore.

Sheet-pile bulkhead

4
Design Considerations (excerpts from EM 1110-2-1614, USACE, 1995)
A. Shoreline Use
Some structures are better suited than others for particular shoreline uses. Revetments of
randomly placed stone may hinder access to a beach, while smooth revetments built with concrete
blocks generally present little difficulty for walkers. Seawalls and bulkheads can also create an
access problem that may require the building of stairs. Bulkheads are required, however, where some
depth of water is needed directly at the shore, such as for use by boaters.
B. Shoreline Form and Composition
1. Bluff shorelines.
Bluff shorelines that are composed of cohesive or granular materials may fail because of scour
at the toe or because of slope instabilities aggravated by poor drainage conditions, infiltration,
and reduction of effective stresses due to seepage forces. Cantilevered or anchored bulkheads can
protect against toe scour and, being embedded, can be used under some conditions to prevent
sliding along subsurface critical failure planes. The most obvious limiting factor is the height of
the bluff, which determines the magnitude of the earth pressures that must be resisted, and, to
some extent, the depth of the critical failure surface. Care must be taken in design to ascertain the
relative importance of toe scour and other factors leading to slope instability. Gravity bulkheads
and seawalls can provide toe protection for bluffs but have limited applicability where other slope
stability problems are present. Exceptions occur in cases where full height retention is provided
for low bluffs and where the retained soil behind a bulkhead at the toe of a higher bluff can
provide sufficient weight to help counter-balance the active thrust of the bluff materials.
2. Beach shorelines.
Revetments, seawalls, and bulkheads can all be used to protect backshore developments along
beach shorelines. An important consideration is whether wave reflections may erode the fronting
beach (i.e. sloped faces absorb more wave energy than vertical walls).
C. Seasonal Variations of Shoreline Profiles
Beach recession in winter and growth in summer can be estimated by periodic site
inspections and by computed variations in seasonal beach profiles. The extent of winter beach
profile lowering will be a contributing factor in determining the type and extent of needed toe
protection.
D. Conditions for Protective Measures
Structures must withstand the greatest conditions for which damage prevention is claimed in
the project plan. All elements must perform satisfactorily (no damage exceeding ordinary
maintenance) up to this condition, or it must be shown that an appropriate allowance has been made
for deterioration (damage prevention adjusted accordingly and rehabilitation costs mortised if
indicated). As a minimum, the design must successfully withstand conditions which have a 50
percent probability of being exceeded during the project’s economic life. In addition, failure of the
project during probable maximum conditions should not result in a catastrophe (i.e. loss of life or
inordinate loss of property/money).
E. Design Water Levels
The maximum water level is needed to estimate the maximum breaking wave height at the

5
structure, the amount of runup to be expected, and the required crest elevation of the structure.
Minimum expected water levels play an important role in anticipating the amount of toe scour that
may occur and the depth to which the armor layer should extend. Consideration are:
1. Astronomical tides
2. Wind setup and pressure effects
3. Storm surge
4. Lake level effects, including regulatory works controls
F. Design Wave Estimation, Wave Height and Stability Considerations
Wave heights and periods should be chosen to produce the most critical combination of
forces on a structure with due consideration of the economic life, structural integrity and hazard
for events that may exceed the design conditions. Wave characteristics may be based on an
analysis of wave gauge records, visual observations of wave action, published wave hindcasts,
wave forecasts or the maximum breaking wave at the site. Wave characteristics derived from
such methods may be for deepwater locations and must be transformed to the structure site using
refraction and diffraction techniques as described in the SPM. Wave analyses may have to be
performed for extreme high and low design water levels and for one or more intermediate levels
to determine the critical design conditions.
Available wave information is frequently given as the energy-based height of the zeroth
moment, Hmo. In deep water, Hs and Hmo are about equal; however, they may be significantly
different in shallow water due to shoaling (Thompson and Vincent 1985). The following equation
may be used to equate Hs from energy-based wave parameters (Hughes and Borgman 1987):

Hs   d  − c1 
= exp co  s 2   ,
H mo   gT p  
 
where Tp is the period of the peak energy density of the wave spectrum and co and c1 are
regression coefficients equal to 0.00089 and 0.834, respectively, described in the SPM. A
conservative value of Hs may be obtained by using 0.00136 for co, which gives a reasonable
upper envelope for the data in Hughes and Borgman.
This equation should not be used when d gT p2 < 0.0005 or there is substantial breaking. In
shallow water, Hs is estimated from deepwater conditions using the irregular wave shoaling and
breaking model of Goda (1975, 1985) which is available as part of the Automated Coastal
Engineering System (ACES) package (Leenknecht et al. 1989). Goda (1985) recommends for the
design of rubble structures that if the depth is less than one-half the deepwater significant wave
height, then design should be based on the significant wave height at a depth equal to one-half the
significant deepwater wave height.
Wave period for spectral wave conditions is typically given as period of the peak energy
density of the spectrum, Tp. However, it is not uncommon to find references and design formulae
based on the average wave period or the significant wave period.
The wave height to be used for stability considerations depends on whether the structure is
rigid, semirigid, or flexible. Rigid structures that could fail catastrophically if overstressed may
warrant design based on H1 . Semi-rigid structures may warrant a design wave between H1 and H10..

6
Flexible structures are usually designed for Hs or H10. Stability coefficients are coupled with these
wave heights to develop various degrees of damage, including no damage.
Available wave gauge and visual observation data for use by designers is often sparse and
limited to specific sites. In addition, existing gauge data are sometimes analog records which have
not been analyzed and that are difficult to process. Project funding and time constraints may prohibit
the establishment of a viable gauging program that would provide sufficient digital data for reliable
study. Visual observations from shoreline points are convenient and inexpensive, but they have
questionable accuracy, are often skewed by the omission of extreme events, and are sometimes
difficult to extrapolate to other sites along the coast.
For wave hindcasts and forecasts, designers should use the simple methods in ACES
(Leenknecht et al. 1989) and hindcasts developed by the U.S. Army Engineer Waterways
Experiment Sta-tion (WES) (Resio and Vincent 1976-1978; Corson et al. 1981) for U.S. coastal
waters using numerical models. These later results are presented in a series of tables for each of the
U.S. coasts. They give wave heights and periods as a function of season, direction of wave approach,
and return period; wave height as a function of return period and seasons combined; and wave
period as a function of wave height and approach angle. Several other models exist for either
shallow or deep water.
G. Breaking Waves
Wave heights derived from a hindcast should be checked against the maximum breaking
wave that can be supported at the site given the available depth at the design still-water level and the
nearshore bottom slope. Design wave heights will be the smaller of the maximum breaker height or
the hindcast wave height.
For the severe conditions commonly used for design, Hmo may be limited by breaking wave
conditions. A reasonable upper bound for Hmo is given by
(H mo )max = 0.10 L p tanh (k p h ) , where Lp and kp are the wave length and wave number
determined for Tp and at depth h.
H. Height of Protection
When selecting the height of protection, one must consider the maximum water level, any
anticipated structure settlement, freeboard, and wave runup and overtopping.
Elevation of the structure is perhaps the single most important controlling design factor and
is also critical to the performance of the structure. Numerous seawall failures can be directly and
indirectly attributed to inadequate elevations.
Elevation with reference to mean lower low water (MLLW) is determined by the following
equation:
he = δt + δs + δw + H + F
δt = spring tidal range.
δs = design storm surge.
δw = wave setup.
H = design wave height.
F = freeboard.

7
 The MLLW level is a local property and varies from location to location with reference to
the chart datum. In the United States the US Geodetic datum, known as NGVD is the standard.
Around other parts of the world the datum are different. It is important to obtain such datums from
the local government.
Freeboard
F
Reflected wave height
H
Wave set-up
δw
Storm surge he

δs

Mean high spring tide

δt
MLLW
Chart Datum
D
Dredge line
S
Scouring depth
Note from the diagram that below MLLW, the chart datum, dredge level and scour depth
must be considered. In addition, settlement may be important.
Sometimes for practical reasons, the elevation is set below the calculated design value and
wave overtopping will occur during a storm. Under this condition, the designer must understand the
effects and consequences of allowing overtopping and make adequate provisions to counter these
effects and consequences. This often means partial or total loss of structures supported by the upland
soil and partial or total loss of seawalls.

8
I. Wave Runup & Overtopping
Runup is the vertical height above the still-water level (SWL) to which the uprush from a
wave will rise on a structure. Note that it is not the distance measured along the inclined surface.
a. Rough slope runup. Maximum runup by irregular waves on riprap-covered revetments
may be estimated by (Ahrens and Heimbaugh 1988)
Rmax aξ
= , where Rmax is the maximum runup, a and b are regression coefficients (1.022
H mo 1 + bξ
2πH mo
and 0.24, respectively) and ξ is the surf zone parameter found by ξ = tan θ , where θ
gT p2
is the slope of the revetment.
A more conservative value for Rmax is obtained by using 1.286 for a in the equation.
Maximum runups determined using this more conservative value provide a reasonable upper limit to
the data from which the equation was developed.
Runup estimates for revetments covered with materials other than riprap may be obtained
with the rough slope correction factors in Table 1 (Table 2-2 in EM 1110-2-1614). Table 1 was
developed for earlier estimates of runup based on monochromatic wave data and smooth slopes. To
use the correction with the irregular wave rough slope runup estimates of the above equation,
multiply Rmax obtained from the equation for riprap by the correction factor listed in the table and
divide by the correction factor for quarry-stone. For example, to estimate Rmax for a stepped 1:1.5
slope with vertical risers, determine Rmax and multiply by (correction factor for stepped
slope/correction factor for quarrystone) (0.75/0.60) = 1.25. Rmax for the stepped slope is seen to be
25 percent greater than for a riprap slope.
b. Smooth slope runup. Runup values for smooth slopes may be found in design curves in the
SPM. However, the smooth slope runup curves in the SPM were based on monochromatic wave
tests rather than more realistic irregular wave conditions. Using Hs for wave height with the design
curves will yield runup estimates that may be exceeded by as much as 50 percent by waves in the
wave train with heights greater than Hs . Maximum runup may be estimated by using the rough slope
equation and converting the estimate to smooth slope by dividing the result by the quarrystone rough
slope correction factor.
c. Runup on walls. Runup determinations for vertical and curved-face walls should be made
using the guidance given in the SPM.

9
 Table 1 Rough Slope Runup Correction Factors (Carstea et al. 1975)
Slope RelativeSize Correction Factor
ArmorType
(cot θ) H/Kra,b r
Quarrystone 1.5 3 to 4 0.60
Quarrystone 2.5 3 to 4 0.63
Quarrystone 3.5 3 to 4 0.60
Quarrystone 5 3 0.60
Quarrystone 5 4 0.68
Quarrystone 5 5 0.72
Concrete Blocksc Any 6b 0.93
d
Stepped slope with vertical risers 1.5 1 ≤ Ho'/Kr 0.75
d
Stepped slope with vertical risers 2.0 1 ≤ Ho'/Kr 0.75
d
Stepped slope with vertical risers 3.0 1 ≤ Ho'/Kr 0.70
d
Stepped slope with rounded edges 3.0 1 ≤ Ho'/Kr 0.86
Concrete Armor Units
Tetrapods random two layers 1.3 to 3.0 - 0.45
Tetrapods uniform two layers 1.3 to 3.0 - 0.51
Tribars random two layers 1.3 to 3.0 - 0.45
Tribars uniform one layer 1.3 to 3.0 - 0.50
a
Kr is the characteristic height of the armor unit perpendicular to the slope. For quarrystone, it is
the nominal diameter; for armor units, the height above the slope.
b
Use Ho' for ds/Ho' > 3; and the local wave height, Hs for ds/ Ho' ≤ 3. (Ho' is the unrefracted
deepwater wave height)
c
Perforated surfaces of Gobi Blocks, Monoslaps, and concrete masonry units placed hollows up.
d
Kr is the riser height.

d. Overtopping. It is generally preferable to design shore protection structures to be high

enough to preclude overtopping. In some cases, however, prohibitive costs or other considerations
may dictate lower structures than ideally needed. In those cases it may be necessary to estimate the
volume of water per unit time that may overtop the structure.
(1) Wave overtopping of riprap revetments may be estimated from the dimensionless
equation (Ward 1992)
Q
Q′ = = Co exp(C1 F ′ + C 2 m )
2
gH mo

where non-dimensional freeboard, F ′ = F (H mo Lo )

2 1/ 3
and F = structure freeboard
m = cotangent of the revetment slope, cot θ
Co, C1, C2 = regression coefficients equal to 0.4578, -29.45, 0.8464 respectively
The coefficients listed above were determined for dimensionless freeboards in the
range 0.25 < F' < 0.43, and revetment slopes of 1:2 and 1:3.5.
(2) Overtopping rates for seawalls are complicated by the numerous shapes found on the
seawall face plus the variety of fronting berms, revetments, and steps. Information on

10
 overtopping rates for a range of configurations is available in Ward and Ahrens (1992).
For bulkheads and simple vertical seawalls with no fronting revetment and a small
parapet at the crest, the overtopping rate may be calculated from
Q  F
= Co exp C1 F ′ + C 2 
2
gH mo  ds 

where ds = depth at structure toe and F' is defined above

Co, C1, C2 = 0.338, -7.385, -2.178 respectively
For other configurations of seawalls, Ward and Ahrens (1992) should be
consulted, or physical model tests should be performed.
(alternate equations are contained in the Pile Buck Manual and SPM)
Since onshore winds increase the overtopping rate at a barrier. The overtopping rate may be
adjusted by multiplying a wind correction factor given by:
 h- ds 
k ′ = 1.0 +W f  + 0.1 sin θ
 R 
where Wf is a coefficient depending on wind speed, and θ is the structure slope (90 deg. for a vertical
wall). For a wind speed of 60 mi/hr or greater, Wf =2.0 should be used.
J. Stability and Flexibility
Structures can be built by using large monolithic masses that resist wave forces or by
using aggregations of smaller units that are placed either in a random or in a well-ordered array.
Examples of these are large reinforced concrete seawalls, quarrystone or riprap revetments, and
geometric concrete block revetments. The massive monoliths and interlocking blocks often
exhibit superior initial strength but, lacking flexibility, may not accommodate small amounts of
differential settlement or toe scour that may lead to premature failure. Randomly placed rock or
concrete armor units, on the other hand, experience settlement and readjustment under wave
attack, and, up to a point, have reserve strength over design conditions. They typically do not fail
catastrophically if minor damages are inflicted. Final design will usually require verification of
stability and performance by hydraulic model studies. For larger wave heights, model tests are
preferable to develop the optimum design.
J. Bulkhead Line
The Bulkhead Line is the position of the structure. Federal and state regulations and/or local
ordinances, sometimes, impose restrictions as to the location and the position of the structures.
Therefore, this line must be determined by the combined factors of:
a. Intended purposes.
b. Site characteristics.
c. Regulations.

11
Equations for Armoring and Riprap
1. Armor or riprap weight (Hudson's equation):
γaH 3
W50 =
K D (SG − 1) cot α
3

Stones within the cover layer can range from 0.75 to 1.25 W as long as 50 percent
weigh at least W and the gradation is uniform across the structure’s surface. The above
equation can be used for preliminary and final design when H is less than 5 ft and there is no
major overtopping of the structure. For larger wave heights, model tests are preferable to
develop the optimum design. The equation is for monochromatic waves and calculated armor
weights should be verified during model tests using spectral wave conditions.

Table 2, Suggested Values for Use In Determining Armor Weight (Breaking Wave
Conditions) (from EM 1110-2-1614, Table 2-3)
Armor Unit n1 Placement Slope (cot θ) KD
Quarrystone
Smooth rounded 2 Random 1.5 to 3.0 1.2
Smooth rounded >3 Random 1.5 to 3.0 1.6
Rough angular 1 Random 1.5 to 3.0 Do Not Use
Rough angular 2 Random 1.5 to 3.0 2.0
Rough angular >3 Random 1.5 to 3.0 2.2
Rough angular 2 Special2 1.5 to 3.0 7.0 to 20.0
Graded riprap3 24 Random 2.0 to 6.0 2.2
Concrete Armor Units
Tetrapod 2 Random 1.5 to 3.0 7.0
Tripod 2 Random 1.5 to 3.0 9.0
Tripod 1 Uniform 1.5 to 3.0 12.0
Dolos 2 Random 2.0 to 3.05 15.06
1
n equals the number of equivalent spherical diameters corresponding to the median stone weight that would fit within
the layer thickness.
2
Special placement with long axes of stone placed perpendicular to the slope face. Model tests are described in Markle
and Davidson (1979).
3
Graded riprap is not recommended where wave heights exceed 5 ft.
4
By definition, graded riprap thickness is two times the diameter of the minimum W50 size.
5
Stability of dolosse on slope steeper than 1 on 2 should be verified by model tests.
6
No damage design (3 to 5 percent of units move). If no rocking of armor (less than 2 percent) is desired, reduce KD by
approximately 50 percent.

12
 1/ 3
W 
2. Layer thickness: r = nk ∆   , typically n = 2
 γa 
2/3
Na  P  γ a 
3. Armor units per surface area: = nk ∆ 1 −  
A  100  W 
4. Graded riprap layer thickness. The layer thickness for graded riprap must be
• at least twice the nominal diameter of the W50 stone ( D = (W γ ) )
1/ 3

• at least 25 percent greater than the nominal diameter of the largest stone
• and should always be greater than a minimum layer thickness of 1 ft (Ahrens 1975).
 W 
1/ 3
 W100 
1/ 3

∴ rmin = max 2.0 50 min  , 1.25  , 1 ft 
  γ r   γr  

where rmin is the minimum layer thickness perpendicular to the slope.

Greater layer thickness' will tend to increase the reserve strength of the revetment against
waves greater than the design. Gradation (within broad limits) appears to have little effect on
stability provided the W50 size is used to characterize the layer. The following are suggested
guidelines for establishing gradation limits (from EM 1110-2-1601) (see also Ahrens 1981):
(1) The lower limit of W50 stone (W50min) should be selected based on stability requirements
using Hudson's equation.
(2) The upper limit of the W100 stone (W100max) should equal the maximum size that can be
economically obtained from the quarry but not exceed 4 times W50min.
(3) The lower limit of the W100 stone (W100min) should not be less than twice W50min.
(4) The upper limit of the W50 stone (W50max) should be about 1.5 times W50min.
(5) The lower limit of the W15 stone (W15min) should be about 0.4 times W50min.
(6) The upper limit of the W15 stone (W15max) should be selected based on filter requirements
specified in EM1110-2-1901. It should slightly exceed W50min.

13
Reserve Stability
A well-known quality of randomly placed rubble structures is the ability to adjust and resettle
under wave conditions that cause minor damages. This has been called reserve strength or reserve
stability. Structures built of regular or uniformly placed units such as concrete blocks commonly
have little or no reserve stability and may fail rapidly if submitted to greater than design conditions.
Values for the stability coefficient, KD, allow up to 5 percent damages under design wave
conditions. If the armor stone available at a site is lighter than the stone size calculated using the
wave height at the site, the zero-damage wave height for the available stone can be calculated, and a
ratio with the site’s wave height can be used to estimate the damage that can be expected with the
available stone. The H/HD=0 values found in the SPM are for breakwater design and non-breaking
wave conditions and include damage levels above 30 percent. Due to differences in the form of
damage to breakwaters and revetments, revetments may fail before damages reach 30 percent. The
values should be used with caution for damage levels from breaking and non-breaking waves.
Information on riprap reserve stability can be found in Ahrens (1981). Reserve stability
appears to be primarily related to the layer thickness although the median stone weight and structure
slope are also important.

Toe Protection
Toe protection is supplemental armoring of the beach or bottom surface in front of a structure
which prevents waves from scouring and under-cutting it. Factors that affect the severity of toe scour
include wave breaking (when near the toe), wave runup and backwash, wave reflection, and grain-
size distribution of the beach or bottom materials. Toe stability is essential because failure of the toe
will generally lead to failure throughout the entire structure. Specific guidance for toe design based
on either prototype or model results has not been developed. Some empirical suggested guidance is
contained in Eckert (1983).
a. Revetments. The revetment toe often requires special consideration because it is subjected
to both hydraulic forces and the changing profiles of the beach fronting the revetment.
(1) Design procedure. Toe protection for revetments is generally governed by hydraulic
criteria. Scour can be caused by waves, wave-induced currents, or tidal currents. For
most revetments, waves and wave-induced currents will be most important. For
submerged toe stone, weights can be predicted based on:
γ H3
Wmin = 3 S
N s (SG − 1)
3

where Ns is the design stability number for rubble toe protection in front of a vertical
1− K  h 
wall, and is the maximum of N s = 1.3 1 / 3  t + 1.8 exp − 1.5
(1 − K )2 ht  or N =
 s
 K H  K 1/ 3 H 
2kht
1.8, where K = sin 2 kBt (ACES Technical Ref, USACE, ch. 4). For toe
sinh 2kht
structures exposed to wave action, the designer must select either Hudson's equation for
rubble mound breakwaters which applies at or near the water surface or the equation

14
 above. (Note: The above equation yields a minimum weight and Hudson's equation
yields a median weight.) When the toe protection is for scour caused by tidal or riverine
currents alone, the designer is referred to EM 1110-2-1601, Hydraulic Design of Flood
Control Channels. Virtually no data exist on currents acting on toe stone when they are a
product of storm waves and tidal or riverine flow. It is assumed that the scour effects are
partially additive. In the case of a revetment toe, some conservatism is provided by using
the design stability number for toe protection in front of a vertical wall as suggested
above.
(2) Suggested toe configurations. The Figure 4 illustrates possible toe configurations. Other
schemes known to be satisfactory by the designer are also acceptable. Designs I, II, IV,
and V are for up to moderate toe scour conditions and construction in the dry. Designs III
and VI can be used to reduce excavation when the stone in the toe trench is considered
sacrificial and will be replaced after infrequent major events. A thickened toe similar to
that in Design III can be used for underwater construction except that the toe stone is
placed on the existing bottom rather than in an excavated trench.

Figure 1, Example Revetment Toe Configurations

b. Seawalls and bulkheads. Design of toe protection for seawalls and bulkheads must consider
geotechnical as well as hydraulic factors. Cantilevered, anchored, or gravity walls each depend on
the soil in the toe area for their support. For cantilevered and anchored walls, this passive earth
pressure zone must be maintained for stability against overturning. Gravity walls resist sliding
through the frictional resistance developed between the soil and the base of the structure.
Overturning is resisted by the moment of its own weight supported by the zone of bearing beneath

15
the toe of the structure. Possible toe configurations are shown in Figure 5.

Figure 2, Seawall and Bulkhead Toe Protection

(1) Seepage forces. The hydraulic gradients of seepage flows beneath vertical walls can
significantly increase toe scour. Steep exit gradients reduce the net effective weight of
the soil, making sediment movement under waves and currents more likely. This seepage
flow may originate from general groundwater conditions, water derived from wave
overtopping of the structure, or from precipitation. A quantitative treatment of these
factors is presented in Richart and Schmertmann (1958).
(2) Toe apron width. The toe apron width will depend on geotechnical and hydraulic factors.
The passive earth pressure zone must be protected for a sheet-pile wall as shown in
Figure 6. The minimum width, B, from a geotechnical perspective can be derived using
the Rankine theory as described in Eckert (1983). In these cases the toe apron should be
wider than the product of the effective embedment depth (de) and the coefficient of
passive earth pressure for the soil ( cot(45-φ/2) ). Using hydraulic considerations, the toe
apron should be at least twice the incident wave height for sheet-pile walls and equal to
the incident wave height for gravity walls. In addition, the apron should be at least 40
percent of the depth at the structure, ds. The greatest width predicted by these
geotechnical and hydraulic factors should be used for design. In all cases, undercutting

16
 and unraveling of the edge of the apron must be minimized.

Figure 3, Toe Aprons for Sheet-Pile Bulkheads

(3) Toe stone weight. Toe stone weight can be predicted based on the equations in section
a(1) above. A design wave between H1 and H10 is suggested.

Filters
A filter is a transitional layer of gravel, small stone, or fabric placed between the underlying
soil and the structure. The filter prevents the migration of the fine soil particles through voids in the
structure, distributes the weight of the armor units to provide more uniform settlement, and permits
relief of hydrostatic pressures within the soils. For areas above the waterline, filters also prevent
surface water from causing erosion (gullies) beneath the riprap.
a. Graded rock filters. The filter criteria can be stated as:
d15 filter d15 filter
< 4 to 5 <
d 85 soil d15 soil
where the left side of the equation is intended to prevent piping through the filter and the
right side provides for adequate permeability for structural bedding layers. This guidance
also applies between successive layers of multi-layered structures. Such designs are
needed where a large disparity exists between the void size in the armor layer and the
particle sizes in the under-lying layer.
b. Riprap and armor stone underlayers. Underlayers for riprap should be sized as:
d15 armor
<4
d 85 filter
where the stone diameter, d, can be related to the stone weight, W, through the layer
thickness equation setting n equal to 1.0. This is more restrictive than the previous
equation and provides an additional margin against variations in void sizes that may occur
as the armor layer shifts under wave action. For large riprap sizes, each underlayer should
meet the above condition and the layer thickness' should be at least 3 median stone
diameters. For armor and underlayers of uniform-sized quarrystone, the first underlayer
should be at least 2 stone diameters thick, and the individual units should weigh about
one-tenth the units in the armor layer. When concrete armor units with KD > 12 are used,

17
 the underlayer should be quarrystone weighing about one-fifth of the overlying armor
units.
c. Plastic filter fabric selection. Selection of filter cloth is based on the equivalent opening
size (EOS), which is the number of the U.S. Standard Sieve having openings closest to the
filter fabric openings. Material will first be retained on a sieve whose number is equal to
the EOS.
(1) For granular soils with less than 50 percent fines (silts and clays) by weight (passing a
No. 200 sieve), select the filter fabric:
EOS sieve
≤1
d 85 soil
(2) For other soils, the EOS should be no larger than the openings in a No. 70 sieve.
Furthermore, no fabric should be used whose EOS is greater than 100, and none
should be used alone when the underlying soil contains more than 85 percent material
passing a No. 200 sieve. In those cases, an intermediate sand layer may provide the
necessary transition layer between the soil and the fabric.
(3) Finally, the gradient ratio of the filter fabric is limited to a maximum value of three.
That is, based on a head permeability test, the hydraulic gradient through the fabric
and the 1 in. of soil adjacent to the fabric (i1) divided by the hydraulic gradient of the 2
in. of soil between 1 and 3 in. above the fabric (i2) is:
i1
Gradient Ratio = ≤3
i2
d. Filter fabric placement. Experience indicates that synthetic cloths can retain their strength
even after long periods of exposure to both salt and fresh water. To provide good
performance, however, a properly selected cloth should be installed with due regard for
the following precautions.
(1) Heavy armor units may stretch the cloth as they settle, eventually causing bursting of
the fabric in tension. A stone bedding layer beneath armor units weighing more than 1
ton for above-water work (1.5 tons for underwater construction) is suggested and
multiple underlayers may be needed under primary units weighing more than 10 tons.
(2) The filter cloth should not extend seaward of the armor layer; rather, it should
terminate a few feet landward of the armor layers.
(3) Adequate overlaps between sheets must be provided. For lightweight revetments this
can be as little as 12 in. and may increase to 3 ft for larger underwater structures.
(4) Sufficient folds should be included to eliminate tension and stretching under
settlement. Securing pins with washers is also advisable at 2-to 5-ft intervals along the
midpoint of the overlaps.
(5) Proper stone placement requires beginning at the toe and proceeding up the slope.
Dropping stone can rupture some fabrics even with free falls of only 1 ft. Greater drop
heights are allowable underwater.
Flank Protection
Flank protection is needed to limit vulnerability of a structure from the tendency for erosion

18
to continue around its ends. Return sections are generally needed at both ends to prevent this. Sheet-
pile structures can often be tied well into existing low banks, but the return sections of other devices
such as rock revetments must usually be progressively lengthened as erosion continues. Extension of
revetments past the point of active erosion should be considered but is often not feasible. In other
cases, a thickened end section, similar to toe protection, can be used when the erosion rate is mild.
Material Hazards and Considerations
Corrosion is a primary problem with metals in brackish and salt water, particularly in the
splash zone where materials are subjected to continuous wet-dry cycles. Mild carbon steel, for
instance, will quickly corrode in such conditions. Corrosion-resistant steel marketed under various
trade names is useful for some applications. Aluminum sheet-piling can be substituted for steel in
some places. Fasteners should be corrosion-resistant materials such as stainless or galvanized steel,
wrought iron, or nylon. Various protective coatings such as coal-tar epoxy can be used to treat
carbon steel. Care must always be taken to avoid contact of dissimilar metals (galvanic couples). The
more active metal of a galvanic couple tends to act as an anode and suffers accelerated corrosion.
Concrete should be designed for freeze-thaw resistance (as well as chemical reactions with
salt water), as concrete may seriously degrade in the marine environment. Guidance on producing
suitable high quality concrete is presented in EM 1110-2-2000 and Mather (1957).
Timber used in marine construction must be protected against damage from marine borers
through treatment with creosote and creosote coal-tar solutions or with water-borne preservative
salts (CCA and ACA). In some cases, a dual treatment using both methods is necessary. Specific
guidance is included in EM 1110-2-2906.
The ultraviolet component of sunlight quickly degrades untreated synthetic fibers such as
those used for some filter cloths and sand-bags. Some fabrics can completely disintegrate in a matter
of weeks if heavily exposed. Any fabric used in a shore protection project should be stabilized
against ultraviolet light. Carbon black is a common stabilizing additive which gives the finished
cloth a characteristic black or dark color in contrast to the white or light gray of unstabilized cloth.
Even fabric that is covered by a structure should be stabilized since small cracks or openings can
admit enough light to cause deterioration.
Abrasion occurs where waves move sediments back and forth across the faces of structures.
Little can be done to prevent such damages beyond the use of durable rock or concrete as armoring
in critical areas such as at the sand line on steel piles.
At sites where vandalism or theft may exist, construction materials must be chosen that
cannot be easily cut, carried away, dismantled, or damaged. For instance, sand-filled fabric
containers can be easily cut, small concrete blocks can be stolen, and wire gabions can be opened
with wire cutters and the contents scattered.

Additional References
Ahrens, J. P. 1975 (May). “Large wave tank tests of riprap stability,” CERC Technical
Memorandum 51, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
Ahrens, J. P. 1981 (Dec). “Design of riprap revetments for protection against wave attack,”
CERC Technical Paper 81-5, U.S. Army Engineer Waterways Experiment Station,

19
Vicksburg, MS.
Ahrens, J. P. 1981 (Dec). “Design of riprap revetments for protection against wave attack,”
CERC Technical Paper 81-5, U.S. Army Engineer Waterways Experiment Station,
Vicksburg, MS.
Ahrens, J. P. and Heimbaugh, M. S. 1988 (May). “Approximate upper limit of irregular
wave runup on riprap,“ Technical Report CERC-88-5, U.S. Army Engineer Waterways
Experiment Station, Vicksburg, MS.
Carstea, D., et al. 1975. “Guidelines for the environmental impact assessment of small
structures and related activities in coastal bodies of water,” Technical Report MTR-6916,
The Mitre Corp., McLean, VA.
Eckert, J. W. 1983. “Design of toe protection for coastal structures,” Coastal Structures ’83
ASCE Specialty Conference, 331-41.
Goda, Y. 1985. Random seas and design of maritime structures. University of Tokyo Press.
Herbich, J.B. 1991, Coastal Engineering Handbook
Mather, B. 1957 (Jun). “Factors affecting the durability of concrete in coastal structures,”
CERC Technical Memorandum 96, U.S. Army Engineer Waterways Experiment Station,
Vicksburg, MS.
Richart, F. E., Jr., and Schmertmann, J. H. 1958. “The effect of seepage on the stability of
sea walls.” Sixth International Conference on Coastal Engineering, 105-28.
Sorensen, R.M. 1997, Basic Coastal Engineering
Ward, D. L. 1992 (Apr). “Prediction of overtopping rates for irregular waves on riprap
revetments,” CERC Miscellaneous Paper 92-4, U.S. Army Engineer Waterways Experiment
Station, Vicksburg, MS.
Ward, D. L., and Ahrens, J. P. 1992 (Apr). “Overtopping rates for seawalls,” CERC
Miscellaneous Paper 92-3, U.S. Army Engineer Waterways Experiment Station, Vicksburg,
MS.
Weggel, J. R. 1972. “Maximum breaker height for design,” Journal, Waterways, Harbors and
Coastal Engineering Division, American Society of Civil Engineers 98 (WW4), Paper 9384.

EM 1110-2-1601, Hydraulic Design of Flood Control Channels, USACE

20