COFFERDAMS

Contents
1. Introduction
2. Requirements of a Cofferdam
3. Planning a Cofferdam
4. Causes of Failure
5. Support Arrangements
6. Design of Cofferdams
7. Single Skin Cofferdams
8. Cofferdams for River Crossings
9. Single Skin Cofferdam Design Example
10. Design of Support System
11. Allowable Stresses for Steel Walings
12. Cofferdam Support Frames
13. Strength of Walings - Universal Columns
14. Strength of Walings - Timber Beams
15. Safe Loads for Horizontal Box Pile Struts
16. Safe Loads for Horizontal Universal Column Struts
17. Safe Loads for Horizontal Universal Compound Beam Struts
18. Safe Loads for Square Timber Struts
19. Circular Cofferdams
20. Reinforced Concrete Walings for Circular Cofferdams
21. Earth Filled Double Wall and Cellular Cofferdams
22. Double Skin/Wall Cofferdams
23. Cellular Cofferdams
24. Effect of Water Pressure
25. Flow Nets
26. Pump Sumps

1.Introduction
The purpose of a cofferdam is to exclude soil and water from an area in which it is
required to carry out construction to a depth below the surface. Total exclusion of water
is often not necessary and in some instances is not possible, but the effects of water
ingress must be taken into account in any calculations.

There are two principal approaches to cofferdam design. Single skin structures are most
common, but for very large or deep excavations and marine works, double wall or
cellular cofferdams may be preferred.

The design of a cofferdam is carried out in the same manner as a retaining wall and
consequently, the same design rules apply.
2.Requirements of a Cofferdam
The design of a cofferdam must provide the following:

 The structure must be able to withstand the loads applied to it.

 The quantity of water entering the cofferdam must be controllable by pumping.
 At any stage in the construction, the formation level must be stable and not
subject to excessive heave or boiling.
 Deflection of the cofferdam walls and bracing must not affect the permanent
structure or any existing structure adjacent to the cofferdam.
 Overall stability must be demonstrated against out of balance earth pressures
due to sloping ground or circular slip failure.
 The cofferdam must provide an appropriate size of hole to suit what is to be built
inside.
 Temporary cofferdams must be built in such a way that the maximum amount of
construction materials can be recovered.

Consideration should be given to the incorporation of piles used to form temporary

cofferdams in the permanent structure as significant savings in both time and overall
cost can be achieved.

3.Support Arrangements
The arrangement of supports to a cofferdam structure is the most critical part of a
cofferdam design. The level at which support is provided governs the bending moments
in the sheet piles and the plan layout governs the ease of working within the structure.
Whilst structural integrity is paramount, the support layout must be related to the
proposed permanent works activities causing the minimum obstruction to plant and
materials access. As a general rule simplicity should always be favoured.

Support frames should be located such that concrete lifts can be completed and the
support loads transferred to the permanent works before the frame is removed.
Clearance to starter bars for the next lift should also be considered when positioning
frames.

The clear space between frame members should be optimised to provide the largest
possible uninterrupted area without the need for excessively large structural elements.
Positioning of support members is often a matter of experience.

3.Planning a Cofferdam
The designer of a cofferdam must have an established set of objectives before
commencing the design. The sequence of construction activities must be defined in
order that the design can take into account all the load cases associated with the
construction and dismantling of the structure. From this sequence, the critical design
cases can be determined and the bending moments, shear forces and minimum
penetration depths calculated to determine the pile section, which is required to do the
job.

As part of the analysis of construction activities, the designer should undertake a risk
assessment of the effect of any deviation from the planned sequence. Such deviations
may be in the form of over excavation at any stage, inability to achieve the desired pile
penetration, installation of supports at the wrong level or the imposition of a large
surcharge loading from construction plant. If any stage in the cofferdam construction is
particularly vulnerable then contingency plans should be developed to minimise any risk
and site staff should be informed to limit the possibility of critical conditions being
realised.

The majority of cofferdams are constructed as temporary works and it may be

uneconomic to design for all possible loading cases. Decisions will have to be taken to
determine the level of risk which is acceptable when assessing the design cases; such a
situation may occur when assessing hydraulic loading on a cofferdam. Flood conditions
tend to be seasonal and provision of a cofferdam, which will exclude water at all times,
may involve a substantial increase in pile size and strength. In an extreme flood
condition the design philosophy may involve evacuation of the excavation and allowing
for the structure to be overtopped. Under these conditions, the designer must allow for
overtopping, considering the effect of the sudden ingress of water on the base of the
cofferdam and the effect that any trapped water will have on the stability of the
cofferdam when the flood subsides. Prior to the commencement of construction the site
area should be cleared to permit plant and guide frames to be set up. Excavation should
not begin until all plant and materials for supporting the piles are readily available
including pumping equipment where necessary.

4.Causes of Failure
There are many possible causes of cofferdam failure, but usually failure can be
attributed to a relatively small number of causes. These include:-

 Lack of attention to detail in the design and installation of the structure.

 Failure to take the possible range of water levels and conditions into account.
 Failure to check design calculations using information discovered during
excavation.
 Over excavation at any stage in the construction process.
 Inadequate framing (both quantity and strength) provided to support loads.
 Loading on frame members not taken into account in the design, ie walings and
struts used to support walkways, pumps etc.
 Accidental damage to structural elements not repaired.
 Insufficient penetration of piles to prevent piping or heave.
 Lack of communication between temporary and permanent works designers or
designer and site staff.
In many cases failure may result from the simultaneous occurrence of a number of the
above factors, any one of which might not have been sufficient to cause the failure
itself.

5.Support Arrangements
The arrangement of supports to a cofferdam structure is the most critical part of a
cofferdam design. The level at which support is provided governs the bending moments
in the sheet piles and the plan layout governs the ease of working within the structure.
Whilst structural integrity is paramount, the support layout must be related to the
proposed permanent works activities causing the minimum obstruction to plant and
materials access. As a general rule simplicity should always be favored.

Support frames should be located such that concrete lifts can be completed and the
support loads transferred to the permanent works before the frame is removed.
Clearance to starter bars for the next lift should also be considered when positioning
frames.

The clear space between frame members should be optimised to provide the largest
possible uninterrupted area without the need for excessively large structural elements.
Positioning of support members is often a matter of experience.

6.Design of Cofferdams
The life of the cofferdam structure must be assessed in order that the appropriate
geotechnical parameters for the soils in which the cofferdam is to be constructed can be
selected. In the majority of cases, total stress parameters for clay soils can be used as
the cofferdam is a temporary structure. However, the designer must assess the
susceptibility of any clay to rapid attainment of a drained state and if there is any doubt,
a check should be made on the final structure using effective strength parameters.

As a rule of thumb it is recommended that cofferdams, which will be in service for three
months or more, should be designed using effective stress strength parameters.

The most economic section of sheet pile to be used will be determined by the bending
moments calculated in the cofferdam design. This section must then be considered with
respect to site conditions as it may be necessary to select a heavier section and/or
higher quality of steel to ensure that the calculated penetration depth can be achieved.
This is often the case where long piles are to be driven to deep penetrations or when
hard driving conditions are expected. This topic is discussed in greater depth in
Installation.

In order to maximise the benefits in terms of the design from provision of a heavier
section or higher quality steel, it is recommended that engineers consider driving of piles
at an early stage in the planning/design process as the need for a heavy pile for driving
may eliminate one or more levels of support with a consequent reduction in design
effort.
It is recommended that a minimum surcharge load of 10 kN/m² is adopted in design
calculations to allow for the presence of plant and equipment around the edge of the
excavation during construction. However, this surcharge should be increased (locally or
globally) if heavy plant is to be located adjacent to the cofferdam or if excavated spoil is
to be stacked before removal from the site - a 1m height of spoil will result in a
surcharge load in the order of 20kN/m² .

Following the guidance of BS 8002:1994 the structural design of steel piling, walings,
struts and tie rods should be undertaken in accordance with BS 449: Part 2: 1969. It is
recommended that, due to the degree of uncertainty in the determination of tie rod
loads, an additional factor of safety is incorporated into their design as described in
Retaining Walls. Timber components should be designed in accordance with BS 5268:
Part 2: 1991.

The application of the referenced codes to sheet pile design yields the following
allowable stresses:

Allowable stresses for sheet piles in bending

Class of Work Steel Grade EN 10248: 1995

S270GP S355GP
(N/mm²) (N/mm²)

Permanent 180 230

Temporary 200 260

7.Single Skin Cofferdams

Single skin cofferdams are typically formed of sheet piles supported either by means of
internal props or external anchors. The mechanics of single skin cofferdam design are
those already outlined in Retaining Walls. The piles are considered to be simply
supported between frames and below the lowest frame (in both temporary and final
conditions) and to have either ‘fixed’ or ‘free’ earth support, depending on the type of
soil and the depth of cut-off of the piles below the excavation level.

Piles will need to be driven to such a level as to generate sufficient passive resistance
and also to control the infiltration of water into the excavation. Records should be kept
during driving for any indication of declutching of the piles. In such a case it may be
necessary to grout behind the piles in order to control seepage. Cantilever pile
cofferdams can be formed but have the same limitations as cantilever retaining walls
particularly in terms of the achievable retained height.
When the cofferdam has very large plan dimensions, but relatively shallow depth, it is
often more economical to incorporate inclined struts or external anchorages similar to
those described in the chapter on retaining walls. It should not however be forgotten
that the installation of external anchorages requires space which is outside the
cofferdam area and wayleaves may be required to install the anchors under adjacent
properties.

For a typical cofferdam with a depth exceeding 3m, a system of internal frames
constructed from either timber or steel is normally employed.

The design should be undertaken in stages to reflect accurately the construction

process. Typically the sequence of operations would be to excavate and dewater to just
below top frame level then install the first frame; this procedure being repeated for each
successive frame. In the case of cofferdams in water, it should be noted that the
stresses occurring during dewatering and installation of frames may be considerably in
excess of those in the completed cofferdam. For cofferdams in water it may be
preferable to employ Z-profile piles as the interlocks are located in the zone of maximum
stress and therefore tend to tighten as the load is applied. If U-profile piles are
employed it is advisable to investigate the use of some form of proprietary interlock
sealant. When a cofferdam is to be used solely for the purpose of excluding water and
the depth of soil to be excavated is only nominal, it is often more efficient to install all
the framing under water before commencing dewatering. The table below shows the
optimum spacing of frames for this method of construction. The spacing results in
approximately equal loading on the second and successive frames. The table below also
indicates the maximum spacing between the top and second insert frame for different
pile sections.

Maximum Depth, h, from first to second waling (top waling

at water level)
The above dimensions are based on an allowable working stress of 180N/mm² . For
other stresses the dimensions may be approximated by the equation:-

h³ =1.56 x Z x pbc

NB - For depths exceeding 7m waling loads may become excessive

8.Cofferdams for River Crossings

The following cofferdam arrangements may be employed when a pipeline is to be laid
under a river bed and it is not possible to close off the waterway.
  Construction of cofferdams in river banks

Cofferdams for River Crossings

The following cofferdam arrangements may be employed when a pipeline is to be laid
under a river bed and it is not possible to close off the waterway.
  Construction of cofferdams in river banks

9.Single Skin Cofferdam Design Example

The following example is based upon the same soil conditions as used for the earth
pressure calculation example included in the Retaining Walls Chapter. The earth
pressures in the final condition are those calculated in that example.

In this example, the analysis of the sheet piles assumes the presence of a hinge at the
lower prop position to make the problem statically determinate. The problem can, using
this assumption, be treated as two single propped retaining walls.

Before placing the top frame, excavation will have had to be carried out and the piles
will therefore act in cantilever mode. The net pressure diagram for that case is given in
the figure below.
The top frame is assumed to be 1m below the ground surface. We will assume in this
example that the lower frame is positioned 5.5m below the ground surface.

The excavation depth required to install the lower frame is therefore 6.1m.

The net pressure diagram for this condition is given below.

Pp at 6.1m below ground level in sand and gravel = 0x6.0 = 0.0 kN/m²
Pp at 11m below ground level in sand and grave = (53.9x6.0)+49.0 = 372.4 kN/m²
Pp at 11m below ground level in firm clay = 102.9+(2.3x65) = 252.4 kN/m²
Pp at 13m below ground level in firm clay = 142.9+(2.3x65) = 292.4 kN/m²

Net Pp at 11m below ground level in sand and gravel = 372.4-133.2 = 239.2 kN/m²
Net Pp at 11m below ground level in firm clay = 252.4-98.2 = 154.2 kN/m²
Net Pp at 13m below ground level in firm clay = 292.4-138.2 = 154.2 kN/m²
In this condition, the piles are supported by their embedment below the excavation level
at 6.1m depth and the top frame. This depth is sufficient to produce a fixed earth
support to the toes of the piles. Hence the effective span of the piles may be taken as
the distance from the top waling and the point ‘0’ on the diagram.

Taking moments of the active pressures about the point ‘0’:

kNm/m
69.6 x 1.1 x 0.5 x 0.73 = 27.94
68.7 x 2.19 x 0.5 x 1.83 = 137.66
27.1 x 2.19 x 0.5 x 2.56 = 75.97
27.1 x 1.51 x 0.5 x 3.79 = 77.55
12.0 x 1.51 x 0.5 x 4.30 = 38.96
26.2 x 1.2 x 0.5 x 5.2 = 81.74
10.2 x 1.2 x 0.5 x 5.6 = 34.27
10.2 x 1.2 x 0.5 x 6.4 = 39.17
3.3 x 1.2 x 0.5 x 6.8 = 13.46
= 526.72
The load on the top frame = 526.72/6.2 = 84.95 kN/m

Zero shear occurs at 4.66m below ground level. (Where the area of the active pressure
diagram above the level equals the top frame load)

Take moments about and above the level of zero shear:

kNm/m
84.95 x 3.66 = 310.92
-41.35 x 0.75 x 0.5 x 0.25 = - 3.88
-27.1 x 0.75 x 0.5 x 0.5 = - 5.08
-27.1 x 1.51 x 0.5 x 1.25 = -25.58
-12.0 x 1.51 x 0.5 x 1.76 = -15.95
-26.2 x 1.2 x 0.5 x 2.66 = -41.82
-10.2 x 1.2 x 0.5 x 3.06 = -18.73
-10.2 x 1.2 x 0.5 x 3.86 = -23.62
-3.3 x 1.2 x 0.5 x 4.26 = - 8.43
167.83

Maximum bending moment in the temporary condition = 167.83 kNm/m

Lower frame is installed 5.5m below ground level.

Excavate to final level.

The depth of cut off is insufficient to produce ‘fixed earth support’. Check for ‘free earth
support’ considering that the piles are simply supported at the bottom frame position
due to the assumption of a hinge at the support position.
Moments about and below the lower frame position:

kNm/m
57.3 x 0.6 x 0.5 x 0.2 = 3.44
68.7 x 0.6 x 0.5 x 0.4 = 8.24
69.6 x 1.8 x 0.5 x 1.2 = 75.17
93.0 x 1.8 x 0.5 x 1.8 = 150.66
93.0 x 1.46 x 0.5 x 2.89 = 198.89
= 436.40

Assume that the piles will be 11.75m long. Take moments of net passive pressures
about and below the lower frame:

kNm/m
102.4 x 1.62 x 0.5 x 4.96= 411.4
116.4 x 0.75 x 5.875= 512.89
= 924.29
Factor of safety against movement of pile toes

= 924.29
436.40
= 2.12

Depth of cut off required to give a factor of safety of 1:

411.4 + 116.4x (5.5 + x/2) = 436.4

Simplifying x² + 11x - 0.43 = 0.0

Solving for x gives x = 0.04m

Pp when x=0.04:

102.4 x 1.62 x 0.5= 82.94

116.4 x 0.04= = 04.66
= 87.60kNm/m

Zero shear occurs 7.7m below ground level.

(This is found by calculating the level at which the area of the active pressure diagram
below the line equals the calculated total passive pressure P p working up the diagram
from the net zero pressure point).

Take moments of pressures about and below the point of zero shear:

kNm/m
90.4 x 0.2 x 0.5 x 0.07 = 0.63
93.0 x 0.2 x 0.5 x 0.13 = 1.21
93 x 1.48 x 0.5 x 0.69 = 47.49
-102.4 x 1.62 x 0.5 x 2.76 = -228.93
-116.4 x 0.04 x 3.32 = - 15.46
= -195.06

The maximum bending moment in the sheet piles is 195.06 kNm/m (Final condition)

Take moments of the active pressures about and above the lower frame:

kNm/m
57.3 x 1.59 x 0.5 x 0.53 = 24.14
27.1 x 1.59 x 0.5 x 1.06 = 22.84
27.1 x 1.51 x 0.5 x 2.09 = 42.76
12.0 x 1.51 x 0.5 x 2.6 = 23.56
26.2 x 1.2 x 0.5 x 3.50 = 55.02
10.2 x 1.2 x 0.5 x 3.90 = 23.87
10.2 x 1.2 x 0.5 x 4.70 = 28.76
3.3 x 1.2 x 0.5 x 5.10 = 10.1
= 231.05

The load on the top frame in the final condition = 231.05/4.5 = 51.34kN/m

To calculate the load on the lower frame in the final condition, take moments of the
entire load system about the level where the factor of safety against rotational failure =
1 (ie 11.04m below ground level)

kNm/m
-116.4 x 0.04 x 0.02 = - 0.09
-102.4 x 1.62 x 0.5 x 0.58 = -48.11
93.0 x 1.48 x 0.5 x 2.65 = 182.37
93.0 x 1.8 x 0.5 x 3.74 = 313.04
69.6 x 1.8 x 0.5 x 4.34 = 271.86
68.7 x 2.19 x 0.5 x 5.67 = 426.53
27.1 x 2.19 x 0.5 x 6.40 = 189.92
27.1 x 1.51 x 0.5 x 7.63 = 156.11
12.0 x 1.51 x 0.5 x 8.14 = 73.75
26.2 x 1.2 x 0.5 x 9.04 = 142.11
10.2 x 1.2 x 0.5 x 9.44 = 57.77
10.2 x 1.2 x 0.5 x 10.24 = 62.67
3.3 x 1.2 x 0.5 x 10.64 = 21.07
-51.34 x 10.04 = -515.45
1,333.55

The load on the lower frame = 1,333.55/5.54 = 240.71kN/m

Summary of Results

Maximum bending moment in the temporary condition = 167.83 kNm/m

Maximum bending moment in the final condition = 195.06 kNm/m

Section modulus required

= 195.06 x 1000 x 100/200 x 100 = 975.3 cm³/m

Note that in the example given above it has been assumed that the construction is
temporary works and a working stress in the steel piles of 200N/mm² has been used.

The recommended sections for this construction are therefore:

 LX 12 (z=1208 cm³/m), or
 Frodingham 2N (z=1161 cm³/m)

both in S270GP steel

10.Design of Support System

The choice of timber or steel for cofferdam bracing is governed by the external loads
which the piles must withstand, taking into account the most severe conditions possible
and the internal dimensions of the cofferdam. The loads on the walings are obtained
from consideration of the same conditions used to obtain the bending moments in the
piles. It is acceptable to use good quality timber such as Douglas Fir or Pitch Pine for the
lighter loadings. However, such timbers must be free from knots, splits, sloping grain
and similar defects which could lead to possible failure.

Heavier loadings will necessitate the provision of steel members, ie universal beams
suitably reinforced by stiffeners where necessary to safeguard against buckling.
Information regarding the resistance to bending and shear of timber and steel beams is
given in the table below.

Allowable Stresses
The application of BS 449 : Part 2 : 1969 to the design of steel walings in universal
beams and columns, channels and compound girders composed of universal beams, or
plated channels yields the following allowable stresses:

Action and Material Steel Grade to BS EN 10210:1995

S275 (N/mm²) S355 (N/mm²)

Bending <40mm 180 230

<63mm 230

<100mm 165 215

Shear in <40mm 110

unstiffened <63mm 140

web <100mm 100 130

12. Cofferdam Support Frames

Timber Framing
Steel Framing

13.Strength of Walings -- Universal Columns

The table below is based on S275 steel with an allowable bending stress of 180N/mm²
and an allowable shear stress in unstiffened webs of 110N/mm² . Webs should also be
checked for buckling and bearing.

When the waling is also subject to axial load the allowable bending moment should be
reduced to ensure that the sum of the stress ratios does not exceed unity.

Elastic Modulus 'X-X' Allowable BM 'X-X'

Serial Size (kg) Shear Value kN
axis cm³ axis kNm

254 x 146 x 31 351 63 166

305 x 165 x 40 560 100 200

356 x 171 x 45 687 123 270

406 x 178 x 60 1063 191 353

457 x 152 x 67 1263 227 453

457 x 191 x 74 1458 262 452

533 x 210 x 92 2072 372 592

610 x 229 x 113 2874 517 741

610 x 305 x 149 4111 740 794

14. Strength of Walings -- Timber Beams

The table below is based upon BS 5268: Part 2: 1991. Permissible bending moments
have been calculated for Grade SS, Class SC4 timber in the wet exposure condition
under medium term loading. For differing exposure/loading conditions the designer must
refer to the code. When the waling is also subject to axial load the allowable bending
moment should be reduced to ensure that the sum of the stress ratios does not exceed
the permissible values given in the code.

Basic Size Elastic Modulus cm³ Permissible BM kN

mm ' X-X' axis 'Y-Y' axis 'X-X' axis 'Y-Y' axis

150x150 562 562 4.2 4.2

150x200 1000 750 7.5 5.6

150x300 2250 1125 16.8 8.4

200x200 1333 1333 10 10

250x250 2604 2604 19.5 19.5

300x300 4500 4500 33.7 33.7

350x350 7145 7145 53.6 53.6

400x400 10666 10666 80 80

15.Safe Loads for Horizontal Box Pile Struts

Note: Table is in accordance with BS 449 for S270GP steel. Self weight and a 10kN
accidental load at the centre of the strut has been allowed for when used with x- x axis
vertical. For additional or alternate loading normal to the span of the strut, calculations
must be made to establish safe axial load.

16.Safe Loads for Horizontal Universal Column Struts

Note: Table is in accordance with BS 449 for S275 steel. Self weight and a 10kN
accidental load at the centre of the strut has been allowed for when used with web
vertical. For additional or alternate loading normal to the span of the strut, calculations
must be made to establish safe axial load.

17.Safe Loads for Horizontal Universal Compound

Beam Struts
 Note : Table is in accordance with BS 449 for S275 steel. Self weight and a 10kN
accidental load at the centre of the strut has been allowed for when used with web
vertical. For additional or alternate loading normal to the span of the strut, calculations
must be made to establish safe axial load.

18.Safe Loads for Square Timber Struts

Note:- Table is in accordance with BS8004 for SS Grade Class SC4 timber in the medium
term condition. Self weight has been allowed for. Particular attention must be paid to
straightness of the strut, eg by limiting bow to approximately 1/ 300 of the length.

19. Circular Cofferdams

The table in Product Information gives the approximate minimum diameters of cofferdams
constructed in LX, Larssen and Frodingham sheet piling; the approximate maximum driving
lengths are also given.

The table is intended as a guide as the minimum diameter will depend upon several other factors
such as type of ground, length of piles and penetration required.

Smaller diameters can be achieved by introducing individual bent corner piles.

On site it is usual to pitch the whole circle before driving, the piles being driven in stages as the
hammer works its way several times around the circumference.

Earth pressures are calculated as for straight-sided cofferdams and the piles may be supported by
circular ring beams, instead of wailings and struts, thus leaving the central area clear of
obstruction. The ring beams are normally subjected to axial loads only, which are calculated as
follows:

Axial load (kN) = waling load (kN/m) x radius of cofferdam (m)

The following table gives an indication of suitable sizes of reinforced concrete ring beams for
various cofferdam diameters and loadings.

Reinforced Concrete Walings for Circular Cofferdams

Diameter of Safe Waling Load (kN per metre run)

Cofferdam (metres)
Size of Waling 'd' x 'b' in mm and number of reinforcing bars

450 x 300 600 x 400 750 x 500 900 x 600 1050 x 700
8 no. 10 no. 8 no. 8 no. 10 no.
12φbars 12φbars 16φbars 20φbars 20φbars

5 473 840

10 236 420 656

15 158 280 438 630

20 210 328 473 643

25 263 378 515

 30 315 429

35 368

Note: The number and size of reinforcing bars given in the table is based on the minimum area of
steel for column construction given by table 3.27 of BS8110: Part 2: 1985 and assumes the use of
High Yield steel (fy = 460N/mm² ) reinforcing bars to BS4449: 1988.

Contents

20.Reinforced Concrete Walings for Circular

Cofferdams
The tabulated safe loads are based on:

1. The permissible compressive stress in the concrete not exceeding 8.8N/mm²,

assuming the use of C35 concrete.
2. W = 1.5 E I / (105 r³)

Where:
W = Waling Load in kN/m
 E = Young’s modulus for concrete
= 25000N/mm²
I = Moment of Inertia about ‘x-x’ in cm4
r = Radius of cofferdam in metres

3. The cofferdam diameter divided by the width of the beam (d) < 35

The above basic formula is based on Timoshenko’s work wherein the formula is given
as:

Wu = (k.E.I) / (R³.105) kN/m

where Wu is the ultimate waling load and k is a factor, the value of which is dependent
on the stiffness of the retained medium. 3 is the value for water, eg in a marine
cofferdam built to facilitate construction on the sea bed. Progressively higher values are,
in theory, applicable for weak/medium/strong soils. However, it is common to use the
value of 3, to which a factor of safety of 2 is applied. Hence the value of 1.5 in the basic
formula.

The ring beam can tolerate very little distortion from a true circle before the onset of
catastrophic instability. Hence the empirical rule:

D/d < 35

where d is the depth of the ring beam, ie the difference between the inner and outer
radii of the beam and D is the diameter of the cofferdam (ie the diameter of the inner
face of the piles). If the sheet piles deflect to any great extent then the load in the
walings will be concentrated at the top or bottom of the waling and will impart torsion
into the beam. This condition should also be checked in the design.

21.Earth Filled Double Wall and Cellular Cofferdams

Earth filled cofferdams are self supporting gravity structures, either parallel-sided
double-wall cofferdams or cellular coffer-dams. The stability of both types is dependent
on the properties of the fill and the soil at foundation level as well as on the
arrangement and type of the steel sheet piling. Typical uses are as dams to seal off dock
entrances temporarily so that work below water level can be carried out in the dry and
in the construction of permanent walls for land reclamation, quays, wharves and
dolphins.

22.Double Skin/Wall Cofferdams

Double wall cofferdams comprise two parallel lines of sheet piles connected together by
a system of steel walings and tie rods at one or more levels. The space between the
walls is generally filled with granular material such as sand, gravel or broken rock.
The exposed or inner pile wall is designed as an anchored retaining wall while the outer
line of piles acts as the anchorage. U or Z-profile sheet piles are appropriate to this form
of construction.

The wall as a whole should be analysed as a gravity structure and, in order to achieve
adequate factors of safety against overturning and sliding, the width should be not less
than 0.8 of the retained height of water or soil. When determining the resistance to
sliding, the coefficient of friction m along the base may be taken as 0.5 to 0.6 but
cannot exceed tanf of the strata. Resistance to sliding per unit run of wall = γ‘.h.μ. W
plus the passive resistance in front of the wall, where W is the width of the structure.

Transverse bulkheads should be provided to form strong points at the ends and at
intermediate positions to assist construction and confine any damage which might occur.
The strong points may comprise a square or rectangular cell tied in both directions.

The water regime both inside and outside the structure is critical. It is recommended
that weepholes are provided on the inner side of the structure near the bottom of the
exposed portion of pile to permit free drainage of the fill material reducing the pressures
on the inner wall and preventing a decrease in the shear strength of the fill with time.
As complete drainage of the fill is not always practical, allowance should be made for
any water pressure acting on the piles.
It is essential that clay or silt is not used as fill material and any material of this type
occurring above the main foundation stratum within the cofferdam must be removed
prior to fill being placed.

The piles must be driven into the soil below excavation or dredge level to a sufficient
depth to generate the required passive resistance. In this condition the structure will
deflect towards the excavated side and the lateral earth pressure on the landward side
may be taken as active. When cohesionless soils occur at or below excavation level, the
penetration of the piling must also be sufficient to control the effects of seepage. The
bearing capacity of the founding strata should be checked against the weight of the
structure and any super-imposed loading.

The presence of rock at excavation level makes this type of cofferdam unsuitable,
unless:

 The rock is of a type that will allow sheet piles to be driven into it to an adequate
penetration (see Installation).
 Tie rods can be installed at a low level (probably underwater).
 A trench can be preformed in the rock into which the piles can be placed and
grouted.
 The pile toes can be pinned with dowels installed in sockets in the rock.

If the piles are driven onto hard rock, or to a nominal depth below dredged level, the
resistance to overturning and sliding should be developed by base friction and
gravitational forces alone. In this condition the lateral earth pressure on the landward
side will be in a condition between at rest and active, depending on the amount of
deflection.

The internal soil pressures acting on the outer walls should be taken as 1.25 times
active values.

23.Cellular Cofferdams
Cellular cofferdams are constructed using straight web sheet piles. (See figure) The piles
are interlocked and driven to form closed cells which are then filled with cohesionless
material. Provided that the material on which they are to be founded is solid they
require only nominal penetration to be stable. Pile penetration will assist in the
resistance of any lateral loads occurring during the construction phase in the vulnerable
period before the fill has been placed and the cell has become inherently stable.

Cellular cofferdam structures are used to retain considerable depths of water or

subsequently placed fill. The straight web pile section and particularly the interlocks,
have been developed to resist the circumferential tension, which is developed in the
cells, due to the radial pressure of the contained fill and, at the same time permit
sufficient angular deflection to enable cells of a practical diameter to be formed. In
cellular construction no bending moments are developed in the sheet piles, which
enables the steel to be disposed of in such a manner to enable the maximum tension to
be developed across the profile. The sections have therefore very little resistance to
bending and are not suitable for normal sheet pile wall construction. Since the straight
web sections themselves are capable of retaining the fill, walings and tie rods are not
required. U or Z-profile piles are not suitable for this application as they are not
designed to act in tension in this manner.

The radial pressure ‘p’ at any depth 'h' is Ko.γ'.h plus the net hydrostatic pressure at that
depth. Ko is the ‘at rest’ earth pressure coefficient and is obtained from the equation Ko
= 1 - sinφ. The effective or buoyant density of the fill should be used below the
contained water level. In some circumstances, eg when using hydraulic fill, the
contained water level may be at the top of the structure during and immediately after
filling.

Circumferential tension = p x radius of cell.

The stability of the cofferdam is dependent, on the tensile strength of the piles, the size
and shape of the cell and the properties of the fill and the foundation soil. The granular
fill within the cells provides stability to the whole structure by resisting shear forces
which are set up in it when external forces are applied. To provide adequate factors of
safety the average width of the wall should not be less than 0.8 of the retained height.
The stability calculations should include checks for sliding, overturning, bursting of
interlocks, web tension, internal shear of fill, tilting and foundation failure. The ratio of
ultimate interlock strength against the maximum calculated circumferential tension
should not be less than 2. The disturbing forces on the landward side will be derived
from the active state due to the flexibility of the cellular structure. When determining
the resistance to sliding, the coefficient of friction m between the fill and the founding
stratum may be taken as 0.5 to 0.6 but cannot exceed tanf of the fill. Resistance to
sliding per unit run of wall = γ‘.h.μ.W a where Wa is the average width of the cofferdam.

If the structure is to be founded on clay the strata must be hard or very stiff and the
resistance to sliding should be checked for shear failure within the clay. When founding
on fine sand, a berm may be necessary on the inner side of the wall in order to prevent
loss of stability by boiling of the natural stratum due to seepage. With a sand or gravel
foundation penetration of the piles must be sufficient to prevent seepage affecting the
structural stability. Rock provides a suitable base for cellular construction and
penetration is generally not required for stability. Some limited penetration is desirable
in order to prevent loss of fill and water under the wall, otherwise some other means of
sealing the base should be investigated such as placing mass concrete. Whatever the
type of soil at foundation level all soft clays or silts must be removed prior to filling.
Where there is a significant depth to the founding strata consideration should be given
to removing the overlying material to reduce the driving depth.

All the piles should be driven to the same nominal toe level to permit the transfer of
stresses around the structure. If this is not possible eg due to the presence of a sloping
rock platform, it will be necessary to relieve the load from any piles below the level of
continuity. This may be achieved by the formation of a horizontal concrete infill wedge;
on steeply sloping surfaces such a wedge may need to be dowelled into the rock. Where
the rock profile is irregular, consideration should be given to levelling the surfaces prior
to construction.

Adequate drainage must be provided near the bottom of the piles on the exposed face
by means of weepholes, together with filters where necessary, to improve stability.
Complete drainage may not be possible, however, so the effect of water on the weight
of the infill and pressures on the piles must be investigated.

The usual maximum length in which straight web sections are provided is 20m. Longer
lengths can be supplied by prior arrangement and splice details can be supplied on
request. Due to their low bending strength, however, special care is required when
handling straight web piles. Lifting beams should be employed when moving the piles
horizontally between storage positions. Piles exceeding 15m in length require two or
even three points of support when being raised to the vertical pitching position.

An efficient, robust arrangement of temporary works and templates is fundamental to

the success of a cellular cofferdam scheme. A minimum of two templates is generally
required often placed at low and high water in a tidal location. During pitching it is
essential that pile verticality is maintained. Driving plant should be lightweight and easy
to handle to avoid causing damage to the piles and facilitate repositioning. In difficult
driving conditions some form of driving assistance such as water jetting, should be
investigated. Guidance on temporary works arrangements and the construction
sequence is available on request.
Frodingham Straight Web Section - Dimensions and Properties

Note: *Including 4 No T-Junction Piles.

All Dimensions to centre point of pile web. The practicability of radii less than 3.75m is
subject to site conditions and guidance should be sought from the Corus Construction
Centre. The dimensions given above are for the unloaded condition. Allowance should
be made for a barrelling effect during filling as the tensile force is developed in the piles.

A cellular cofferdam is a flexible structure; horizontal movements of the piles and

vertical movement of the fill should be expected. As pitched, the pile interlocks are not
stressed and a barrelling effect, producing movements of 150mm, is not uncommon
during filling. Horizontal movements at the top of high cofferdams can be up to 500mm
and fill settlements of 150mm are not unusual.

The properties of the Frodingham Straight Web Section SW-1A are shown here. The
range of standard corners and junctions is also indicated. Non-standard corners and
junctions can be fabricated on request. The figure below illustrates the most common
type of cellular construction in circular cells.
24.Effect of Water Pressure
The stability of a cofferdam can be adversely affected by the action of water pressures
on the soils at formation level to the extent that collapse may occur. In granular soils,
excess water pressure causes piping and in cohesive or very tightly packed soils, heave
results.

Piping occurs when the pressure on the soil grains due to the upward flow of water is so
large that the effective stress in the soil approaches zero. In this situation, the soil has
no shear strength and assumes a condition which can be considered as quicksand which
will not support any vertical load. This is obviously a very dangerous situation for
personnel operating in the cofferdam and will also affect the passive resistance afforded
to the cofferdam wall by the soil. The likelihood of piping for a given cross section can
be predicted by the construction of a flow net which will allow the engineer to calculate
the exit hydraulic gradient. Comparison of the calculated value to the critical hydraulic
gradient will indicate the factor of safety against piping; for clean sands this should
generally lie between 1.5 and 2.0. Care should be taken when designing circular
cofferdams and at the corners of rectangular structures where the three dimensional
nature of the situation is more critical than in the case of a long wall.

The factor of safety against piping can be increased by installing the piles to a greater
depth thereby increasing the flow path length and reducing the hydraulic gradient. The
table below gives an approximate guide to the safe minimum cut-off depth for a
cofferdam constructed in medium uniform cohesionless soil, when the toes of the piles
do not penetrate impermeable soils and excess water is pumped from sumps at
excavation level.

Safe Minimum Cut Off Depth

Width of Depth of Cut-

Cofferdam 'W' off 'D'

2H or more 0.4H

H 0.5H
0.5H 0.7H

Base heave can occur in cohesive or very tightly packed granular material, if the force
exerted by water pressure acting on a block of material inside the cofferdam exceeds
the bulk weight of the block. The likelihood of heave can be assessed using a flow net to
calculate the average water pressure acting on the line drawn between the toes of the
piles and converting this to an uplift force on the soil. This can then be compared to the
weight of the soil plug within the cofferdam.

The flow of water into a cofferdam may also be reduced by lowering the ground water
level by means of well points outside the cofferdam. Alternatively, flow into the
cofferdam can be reduced by pumping from well points located inside the cofferdam at
pile toe level. It should however be remembered that, when the stability or ease of
operation of a cofferdam involves pumping, reliability of the pumps is of paramount
importance and back-up capacity must be available to cope with any emergencies.

25.Flow Nets
The preparation of flow nets is a useful tool, as it not only allows the Engineer to
calculate the water pressures in a particular situation, but also provides a visual
representation of the flow regime in the soil. The shape and complexity of a flow net is a
function of the homogeneity and permeability of the soil and the following notes indicate
how a net can be drawn for uniform soil conditions and permeability.

 A scaled drawing of the problem should be produced.

 A datum level should be marked on the cross section either at an impermeable
boundary or at a suitable level below the cofferdam.
 The flow criteria must be determined.
 External water level.
 Internal water level.
 Centre line of cofferdam (this is the axis of symmetry).
 Lines of flow must be parallel to the cofferdam walls and the impervious datum.
Using the above as guidelines, the net is produced from flow lines and equipotential
lines (a stand pipe at any point on an equipotential line would register the same height
H above datum level). These are at right angles to each other and form approximate
squares. This process is very much trial and error but practice will enable the flow net to
be produced with a reasonable degree of speed and accuracy.

To calculate the pore water pressure ‘u’ at any point (using the example above)

 Calculate the potential head ‘H’ at the desired point (note that the potential head
drop is always the same between successive equipotential lines once a square
net has been formed)

H = H1 - (H1 - H2 ) x n / Nd

where

n = number of equipotential drops to the point being considered

Nd = total number of drops

Hence at a point A,

H = H1 - (H1 - H2 ) x 2 / 10

 At any point H = u / γw + z
 where

u = pore water pressure

γw = density of water

z = height of point above datum

AsH, γw and z are known, u can be calculated,

u = (H-z) x γw

26.Pump Sumps
Although a sheet pile wall can prevent the ingress of water into an excavation, it is not
possible to give any guarantee that a cofferdam will be watertight. In order to deal with
any water which enters the excavation it is often desirable to install a drainage system
which can channel water to a sump from which the water can be pumped away.

As the hydraulic gradient adjacent to the corner of a cofferdam is at its largest, it is

advisable to place any sumps at excavation level as far as possible from any corner and
wall.

It should not be forgotten that pumps are able to remove soil as well as water and a
suction hose laid in the bottom of a cofferdam can disturb the base of the excavation
with subsequent movement of the wall if the hose is badly located. Consideration should
be given to forming a sump using a perforated drum into which the hose can be fixed to
limit damage.