Offshore Foundations: Technologies,

Design and Application

Master Student: Pedro Gomes Simões de Abreu
Supervisor: Dr Peter Bourne-Webb
Abstract:
The offshore oil industry started over 60 years ago, since then it evolved immensely. This
evolution was forced by the need of exploiting oil and gas reserves in more challenging regions.
The purpose of this study was to gather information about the foundation structures used in the
offshore industry, and to assess the applicability of two types of foundation in a real scenario. São
Tome & Principe was selected as the case-study for this paper because it is a member of the
Community of Portuguese Language Countries, and has recently been subjected to several studies
in its offshore region to evaluate its potential as an oil & gas supplier. This paper described the
geotechnical characterisation of the offshore of STP based on investigations performed in the
Gulf of Guinea (GoG) for more than 10 years. The results of the study were that the soil is
probably a highly sensitive clay (St=2 to 4), and the shear strength profile presents a gradient of
about 1.5 kPa/m. Another conclusion is that many sites in the GoG exhibit a greater resistance
(up to about 15 kPa) in the first 2 m, this phenomenon is called a “crust”. The paper also describes
design principles for two anchoring systems: the Torpedo Anchors and Suctions Embedded Plate
Anchors (SEPLA). For Torpedo, the results revealed that the pull-out resistance, after
reconsolidation, is expected to be 8.7 MN. Whereas, the results for SEPLA holding capacity is
expected to be 10 MN. For both systems the calculations were made for the largest of their
solutions available in the market.

1. Introduction:
The offshore oil industry started in 1947 with the As this evolution made possible the exploration of
installation of the first oil rig in just 6 m depth of water, more challenging oil and gas fields, some countries
off the coast of Louisiana in the United States. where Portuguese is the official language have gained
Nowadays there are over 7000 offshore platforms attention, and as a result some are currently under
around the world located in a large range of water investigation (e.g. Mozambique, Guinea Bissau and
depths, which are starting to exceed 2000 m. This São Tomé & Principe) , while in others investments
evolution forced a change in the concept of “deep have already been made (e.g. Brazil and Angola).
water”, as in the 1970s deep-water meant depths of 50
São Tome & Principe is a member of the group of
m to 100 m, now this concept refers to water depths
countries which belong to the Community of
around 800 m. With this, a new concept was created to
Portuguese Language Countries (CPLP) and over the
refer to water depths starting from 1000 m, “ultra-
past five years has been subject to many field tests in
deep” water.
order to quantify the potential oil & gas reserves and
to evaluate the quality of the potential extractable floating production and drilling units in operation may
product (oil) of those reserves as well. The exclusive provoke the congestion of the sea bottom due to the
economic zone (EEZ) of STP is now divided into high number of risers and mooring lines employed. In
several blocks, which are licensed to Oil & Gas this scenario, dynamically penetrated anchors (DPA),
companies, so they can develop investigation work and in particular torpedo anchors, have proven to be a
and evaluate the potential of the reserves, Figure 1. reliable alternative used in Brazilian offshore fields
(Aguiar et al., 2009). The reduced mooring line radius
Most of the recent developed projects in Brazil
employed on torpedo anchors relative to catenary
and Angola are in deep and ultra-deep water,
mooring systems with drag anchors, reduces sea
therefore, the adopted foundation systems had to be
bottom congestion, Figure 2.
anchoring systems. The majority of the EEZ of STP is
also in ultra-deep waters, ranging from 1800 m to 3000
m.
This work has the purpose of gathering
information about the foundation structures used in the
offshore industry, and the assessment of the
application of two types of foundation in the offshore
of STP. The choice of these types of foundation
systems are based on their novelty and economic
aspects. Thus, the systems evaluated are: the torpedo
anchor which have been applied in Campos Basin in
Brazil, and the suction embedded plate anchors that
are currently in use in Angola.

Figure 2 – Radius comparison between floating units linked

to conventional drag anchors and torpedo anchors, from:
http://www.hindawi.com/journals/jam/2012/102618.fig.00
2.jpg.

Torpedo anchors (TA) are the most applied type of

DPA and they have been developed by the Brazilian
oil company Petrobras. TAs are cone-tipped,
cylindrical steel pipes filled with concrete and scrap
metal. They penetrate the seabed relying on the kinetic
energy they acquire while free falling from heights of
between 30 m and 150 m above the seabed. Torpedo
anchors come in various sizes from 0.76 m to 1.07 m
in diameter, 12 m to 17 m in length, and 241 kN to 961
kN in weight. The inside of the anchor shaft is filled
with ballast to increase the weight and maintain the
centre of gravity below the centre of buoyancy for
stability. Some versions of the TA have been fitted
with 4 flukes at the trailing edge, ranging in width
from 0.45 m to 0.9 m, and 9 m to 10m long (Raie,
2009; Medeiros et al.,1997, 2001, 2002). Two
Figure 1 - EEZ of STP and block division for licensing round.
different DPAs, with and without fins are pictured in
1.1. Torpedo anchors Figure 3(a).

As offshore exploitation moves to water depths of Torpedoes can easily reach velocities of 25 m/s to 35
around 3000 m, new technologies have had to be m/s at the seabed after being released from a height of
developed in order to reduce installation costs, and 20 m to 40 m above the seabed, allowing tip
facilitate construction. Moreover, the high number of penetrations up to 3 times the anchor length and
holding capacities after consolidation that are
expected to be in the range of 5 to 10 times the weight
of the anchor (Randolph et al., 2005).

Figure 3 – Dynamically penetrating anchors (a) Torpedo

anchor with fins and without fins (Medeiros, 2002); (b)
installation of 4 flukes torpedo anchor (Medeiros, 2002; Figure 4 – Full scale torpedo pile and releasing situation,
O’Loughlin et al., 2004) Lieng et al. (1999).

The installation procedure for DPA has developed The main reason for using this type of anchor solution
from its original method. Instead of using only one is its simplicity and speed of installation. With regard
anchor-handling vessel (AHV) to lower the anchor to to the equipment required for installation, the torpedo
a predetermined height above the seabed, using the anchor installation is depth-independent. Moreover,
permanent mooring line, now two AHV are used. The torpedo piles are cost-effective throughout fabrication,
installation process was modified to minimize the transportation, and installation. Fabrication is easy and
effect of drag force on the mooring line on the free inexpensive due to the simple design of the torpedo
falling motion of the anchor. Accordingly, the anchor anchors. The cost of transportation is low because the
is lowered using an installation wire from the first compact design of the torpedo anchor allows more
AHV while the second AHV holds the permanent anchors to be transported per voyage of the AHV than,
mooring line that is attached to the anchor and forms a for example, suction caissons. Also, the installation is
loop. A remote release system is used at the end of economical because an external source of energy is not
installation wire to release the anchor (Araujo et al., required for installation and a quick installation is
2004). A chain segment is recommended for the lower possible using one or two AHVs and limited use of
portion of the mooring line because model tests of the ROVs. Finally, the predicted holding capacity is less
anchor installation (Lieng et al., 2000) have shown dependent on the precise evaluation of the soil shear-
that chain drag does not reduce the velocity of the strength profile. Since higher strength profiles reduce
anchor during free fall. Figure 3(b) demonstrates the the penetration and lower strength profiles increase
lowering of two model scale torpedo anchors to penetration, therefore the holding capacity is mainly a
position them before free-fall releasing. A full scale function of the kinetic energy obtained during free
torpedo pile and the situation immediately prior to TA falling. Nevertheless, torpedo anchors have the
release, in which it is possible to see the loop between disadvantage of the uncertainty in verticality of the
the permanent mooring line and the installation line, is anchor, which affects the holding capacity
illustrated in Figure 4. (O’Loughlin et al., 2013; Raie, 2009).
1.2. Suction embedded plate anchors
A new system, called a suction embedded plate anchor
(SEPLA), was developed to overcome the problems of
the conventional plate anchor (e.g. VLA), achieving
greater and more precise depth location below the
seabed (Dove et al., 1998; Wilde et al., 2001).
The SEPLA uses a suction caisson (or “follower”) to
embed a rectangular plate anchor, providing a known
initial penetration depth for the anchor, at a specified
geographical location. The components of a SEPLA SEPLA installation accuracy represents a great
are illustrated in Figure 5. improvement over that for drag embedment anchors,
however two questions emerge (these questions are
applied to all offshore plate anchors such as VLAs).
Firstly, the caisson penetration and anchor keying
provokes a disturbance in the soil mass around the
SEPLA, which leads to a decrease of the soil strength
in the region. Secondly, when keying is being initiated,
a loss of embedment depth occurs. While, the first
question can be solved as the soil strength is largely
Figure 5 – Components of a suction embedded plate anchor
recovered over time by soil reconsolidation, the
(Gaudin et al., 2006). second problem cannot because loss of embedment
depth is permanent. This is a very important issue,
SEPLA installation consists of 3 steps: caisson since SEPLA capacity significantly depends on its
penetration, caisson retraction, and anchor keying. embedment depth when the soil has increasing
These steps are shown schematically in FIGURE 6. strength with depth (which is a typical in the offshore
First, the caisson with a plate anchor slotted vertically environment). Therefore, it becomes very important
in its base is lowered to the seafloor and penetrated to accurately estimate the loss of embedment depth
into the soil under its dead weight until the skin during the keying process. This estimate can then be
friction and end-bearing resistance of the soil on the factored into the design; Wilde et al. (2001) report
caisson equal the caisson’s dead weight. The vent upward movements ranging between 0.5 and 1.7 times
valve on the top of caisson is then closed and the water the plate height, which is a wide range when plate
trapped inside is pumped out. The ensuing differential heights of 2.5 m to 4.5m are used in practice.
pressure at the top drives the caisson to the design
depth. The plate anchor is then released and the water Even though the undrained capacity of plate anchors
has been extensively investigated by means of
is pumped back into the caisson, causing the caisson
to move upward, leaving the plate anchor in place in a analytical and experimental methods; for SEPLA,
vertical orientation. The caisson is retracted from the there are a limited number of reported studies and
seabed and prepared for the next installation. As the therefore the keying process is not yet well
understood. However, Song el al. (2009) present a
anchor chain is tensioned, it cuts into the soil.
Simultaneously, the anchor line applies a load to the theoretical model to predict the trajectory and
anchor’s offset padeye causing it to rotate or “key”. In corresponding capacities of SEPLA during the keying
order to achieve the maximum mobilized capacity, the process based on empirical and plastic limit analysis.
plate must be as close to perpendicular to the direction 2. Geotechnical site conditions
of loading as possible (Yang et al., 2011).
São Tomé & Principe is a group of islands situated on
the Gulf of Guinea, the island of Principe is the nearest
to the site where possible oil exploration is more
likely. Figure 7 shows the location of STP and the
surrounding geology, it is also possible to see two red
lines which one of them refer to the schematic cross
sections presented in Figure 8. The cross section
extend from Principe Island to Nigeria.

FIGURE 6 –SCHEMATIC OF SEPLA INSTALLATION (YANG ET AL., 2011).

 Figure 9 - Deep-water sediments physical properties on Gulf
of Guinea (Puech, 2004)

2.2. Shear strength profiles

Sultan et al. (2007) reported some studies from the

continental slope of Nigeria. Even though the water
depth where this study was executed ranged between
Figure 7 – São Tomé & Principe location and surrounding
only 1100 m and 1250 m, the study area was actually
geology, Courtesy of Agencia Nacional do Petroleo of STP
very close to Principe Island as shown in Figure 10.
from: http://www.stp-
The shear strength profile based on laboratory
eez.com/DownLoads/Posters/3_STP_RegionalGeol.pdf . geotechnical tests is represented in Figure 11. This
profile has a similar development to the Puech (2004)
proposal of a gradient of 1.5 kPa/m, therefore in future
calculations this is the gradient that will be used.
However, the shear strength profile does not start from
the zero, therefore it is assumed to be 5 kPa in the first
3 m, and after that assumes the proposed 1.5 kPa/m
gradient (blue line Figure 11).

Figure 8 – Cross section from Nigeria cost to Principe Island,

Courtesy of Agencia Nacional do Petroleo of STP from:
http://www.stp-
eez.com/DownLoads/Posters/3_STP_RegionalGeol.pdf .

2.1. Index Properties of GoG sediments

The sediments present in GoG deep-waters are

characterised by very high water contents, values
which can be between 150% and 250% at the seabed
and gradually decrease with depth, Figure 9(a). Puech
(2004) suggests soil unit weights starting from 12-13
kN/m3 at seabed and increasing to 13-15 kN/m3 below Figure 10 - Nigerian continental slope study area, Sultan et
6-8m, Figure 9(b). al. (2007).
 The design process for Torpedo anchors is rather
complex due to the difficulties of predicting the anchor
embedment and set-up after installation. These two
factors have a direct effect on the anchor capacity and
they depend on the geometry and characteristics of the
anchor, as well as the soil properties such as undrained
shear strength and coefficient of consolidation
(horizontal and vertical).
3.1.1. Adopted anchor geometry
The main problem with torpedo anchors is the lack of
field experience, especially outside Brazil. Therefore,
the geometry suggested in this text will be based on
that used in the Albacora Leste Field (FPSO P-50), a
FPSO unit in water depth of 1400 m with required
capacity of 7500 kN (Araujo et al., 2004).
The torpedoes used for the FPSO P-50 were type T-
98, which are illustrated in Figure 12 and Figure 13,
and had the following characteristics (Brandão et al.,
2006):

 Total mass of 98 tons

 Diameter of 1.07 meters
 Length of 17 meters
Figure 11 - Shear strength profile of Nigerian continental
 Four stiffener wings (flukes): 0.9 m wide x
slope obtained in laboratory geotechnical tests and its 10m long.
comparison with shear strength profiles proposed by others.

Table 1 are summarized the adopted values for the

different soil characteristics that were used in the
evaluation of the SEPLA and torpedo anchors.

Figure 12- Schematic longitudinal section drawing of the T-

Table 1 – Adopted values for different soil characteristics and
98, Brandão et al. (2006).
references from which they were based on.

3. Design of anchor solutions

Since the water depths in the zone of proposed

offshore development near São Tomé & Príncipe
range from 1800 m to 3000 m, future installed
facilities must be floating platforms and hence the
foundation systems in the seabed will be resisting
tensile forces instead of compression. Therefore, the
only types of foundation solution suitable for this
region are anchoring systems.
3.1. Torpedo anchors Figure 13 – Photos of the T-98 body sections welding and its
final adjustments, Brandão et al. (2006).
 3.1.2. Impact Velocity and free-fall Height
It is advised to use drop heights above the seabed
between 30 m and 150 m, which usually result in
impact velocities from 0.5 to 0.33 times the terminal
velocity (Medeiros, 2002). Thus, for this study, an
impact velocity of 40m/s is considered. Using 𝒂 =
𝟏
. 𝑪𝑫 . 𝝆𝒘 . 𝑨𝒑 . 𝒗𝟐 . 𝒎 Eq. 1 and adopting 𝐶𝐷 = 0.33
𝟐
as proposed by Fernandes et al. (2005) for torpedo
anchors, the acceleration needed to achieve an impact
velocity of 40m/s after free-release is 7.4 m/s-2.
𝟏
𝒂 = 𝟐 . 𝑪𝑫 . 𝝆𝒘 . 𝑨𝒑 . 𝒗𝟐 . 𝒎 Eq. 1

Using the equations for conservation of mechanical

energy, determine the height needed to achieve the
chosen impact velocity as the anchor reaches the
seabed:
1 Eq. 2
𝐸𝑐 = . 𝑚. 𝑣 2 = 0.5 × 98000 × 402
2
𝐸𝑝 = 𝑚. 𝑎. ℎ = 98000 × 7.4 × ℎ Eq. 3

𝐸𝑐 = 𝐸𝑝 ⟹ 𝒉𝒇𝒓𝒆𝒆−𝒇𝒂𝒍𝒍 = 𝟏𝟎𝟖. 𝟒 𝒎 Eq. 4

Figure 14 – Comparison of centrifuge and field test

3.1.3. Tip Embedment Depth embedment data, O’Loughlin et al. (2013).

O’Loughlin et al. (2013) after gathering penetration

From Figure 14 it is possible to see the good
data from worldwide field tests and comparing them
agreement between the formulated curve and the
to centrifuge tests of equivalent prototype scale
dataset. So, the prediction for STP using the
models, were able to propose a relationship that
O’Loughlin et al. (2013) proposed relationship, would
predicted penetration depth with reasonable accuracy
be:
for this very large dataset that encompassed a wide
range of anchor masses, geometries and impact 𝑧𝑒 𝐸𝑡𝑜𝑡𝑎𝑙
1/3
≈( Eq. 5
velocities, Figure 14. 𝑑𝑒𝑓𝑓 4 )
𝑘. 𝑑𝑒𝑓𝑓
2
𝐸𝑡𝑜𝑡𝑎𝑙 = 𝐸𝑐 + 𝐸𝑝 = 1/2. 𝑚. 𝑣𝑖𝑚𝑝𝑎𝑐𝑡 + 𝑚′ . 𝑔. 𝑧𝑒 Eq. 6
1
2
0,5. 𝑚. 𝑣𝑖𝑚𝑝𝑎𝑐𝑡 + 𝑚′ . 𝑔. 𝑧𝑒 3
𝑧𝑒+1 ≈( ) × 1.17 Eq. 7
1.5 × 1.174
= 39.24 𝑚
The two grey lines traced in the graph (Figure 14)
bound almost every result, additionally the grey line
below the original equation line intersects the field test
result from T-98 (Brandão et al., 2006). Therefore, this
ratio may vary 13% from the real depth, so the
penetration achieved is expected to be between 36 m
and 43 m.
 3.1.4. Pull-out capacity
Offshore Site São Tomé & Príncipe
The torpedo anchor resistance was determined by the
Tip embedment depth (m) 40
sum of three components: shaft resistance, reverse en
Average undrained shear
bearing resistance and weight of the torpedo pile. strength gradient (kPa)
1.5z
Average undrained shear
47.25
strength on shaft (kPa)
3.2. SEPLA
Adhesion factor (α) 0.7
The functional requirement of SEPLA is to resist the
𝜸𝒔 (kN/m3) 14
specified maximum factored mooring line load, while
avoiding significant displacements, both in the Rs (kN) 4880
direction of the applied load or vertically. SEPLA
Rb (kN) 3470
holding capacity is related primarily to three basic
aspects, which must be defined for the design of the Vertical pile resistance (kN) 8720
solution, those aspects are: Table 2 - Soil properties, penetration and vertical
resistance of Campos Basin in Brazil and STP offshore.
 Anchor plate area
 Undrained shear strength 3.2.2. Keying
 Penetration depth
The anchor keying process promotes two negative
3.2.1. Anchor geometry effects in the plate holding capacity. First, it induces
an upward movement on the anchor during the plate
The components of a typical SEPLA are illustrated in rotation, hence reducing the embedment depth; and
second, the soil in the immediate vicinity of the plate
, the geometry and the characteristics proposed for
Offshore Site São Tomé & Príncipe
SEPLA applied offshore of STP are:
Tip embedment depth (m) 40
 Plate area: 10 m length by 4.5 m width
Average undrained shear
 Shank: 2.5 m high (padeye eccentricity) and strength gradient (kPa)
1.5z

at an angle of 60ᵒ with the fluke. Average undrained shear

47.25
strength on shaft (kPa)
 Anchor thickness: 0.20 m
Adhesion factor (α) 0.7
 Anchor weight: 50 tons
 Installation penetration: 24 m 𝜸𝒔 (kN/m3) 14

Rs (kN) 4880

Rb (kN) 3470

Vertical pile resistance (kN) 8720

anchor is remoulded, therefore reducing the soil
strength (Randolph et al., 2005). Even though, this
latter effect may be recovered as the soil
reconsolidates, the loss of embedment is permanent.
As clay deposits in Gulf of Guinea are typically
characterized by an increasing strength profile with
depth, any loss in embedment will correspond to a
non-recoverable loss in potential anchor capacity.
Figure 15 – Typical SEPLA with keying flap, Wang et al. (2012). Song et al. (2009) ran several centrifuge tests and
developed large deformation finite-element (FE)
analyses. From the results of those studies, they found
that the loss in anchor embedment during anchor
keying may be expressed in terms of a non-
dimensional anchor geometry factor, which is a
function of the eccentricity of the padeye, angle of Wilde et al. (2001) and DNV-RP-E302 (2002) have
loading, and the net moment applied to the anchor at similar results, and the major source of their difference
the stage where the applied load balances the anchor is the usage of a partial safety factor by DNV-RP-E302
weight. Using eq. (8) taking into account all the anchor (2002), for that reason a partial factor of 1.4 is also
geometry measurements and anchor submerged unit applied in Wilde et al. (2001) method to obtain a
weight. The initial moment M0 corresponding to zero design holding capacity. For the design holding
net vertical load on the anchor is given by eq. (9). capacity in Merifield et al. (2001) method is used a
∆𝑧𝑒 0.2
global safety factor of 2.0.
= Eq. 8
𝐵 𝑒 𝑡 0.3 𝑀0 0.1 Contrary to Merifield et al. (2001), these two
( )( ) ( )
𝐵 𝐵 𝐴𝐵𝑠𝑢
procedures do not take into account the self-weight of
𝑀0 = (𝑓 + 𝑊 ′ 𝑎 )𝑒 − 𝑓𝑒𝑓 − 𝑊′𝑎 𝑒𝑤 Eq. 9 the plate anchor in their resistance. This conservative
decision means not taking into consideration 500 kN
∆𝑧𝑒 = 4.19 𝑚 ≈ 4 𝑚
of resistance, which represents 4 to 5% of the anchors
Therefore, for this case study, it is plausible to say that holding capacity.
the SEPLA embedment depth after installation is
about 20 m. As a result, is believed to be more reliable to
use DNV-RP-E302 (2002) method, since it seems to
3.2.3. Plate Holding capacity have a better calibration of its empirical reduction and
safety factors. Thus, plate anchor capacity adopted is
There are several suggested procedures for evaluating
9.5 MN, determined after DNV-RP-E302 (2002) plus
the plate holding capacity in clay, and in this work
the anchor self-weight, 0.5 MN, therefore the
this estimation will be made according to three
predicted plate anchor capacity in this case study is
different procedures, which are:
predicted to be 10 MN.
 Wilde et al. (2001)
4. Discussion of results
 Merifield et al. (2001)
 DNV-RD-E302 (2002) Both of the considered anchor solutions are likely to
be economically competitive alternatives to
The predicted plate anchor capacity according to the conventional offshore anchors, for application in STP.
three different methods is expressed in Table 3. The In Table 4 and Table 5 are enumerated, respectively,
Merifield et al. (2001) procedure returned an ultimate the advantages and disadvantages of both SEPLA and
holding capacity 50% higher than the others, and this torpedo anchors.
may be explained by the two facts. Firstly, one
parameter of the evaluated situation is outside of the Torpedo anchor (DPA)
SEPLA advantages
advantages
range of the theoretical solutions. Secondly, Merifield
 Cost of anchor element
et al. (2001) do not use any empirical reduction factor  Simple and economic to
is the lowest of all the
fabricate
whereas the other two methods do; 0.70 is used by deep-water anchors.
 Simple to design.
Wilde et al. (2001) and 0.75 in DNV-RP-E302 (2002).  Uses proven suction
caisson installation  Accurate to position
methods. with no requirements
Procedure Ultimate Holding Design Holding for proof loading.
 Provides an accurate
 Rapid installation
suggested by: Capacity (MN) Capacity (MN) measure of embedment
and position of the  Robust and compact
Wilde et al. anchor. design makes handling
13 9.2 and installation simple
(2001)  Design based on well-
developed procedures and economic with only
one Anchor Handling
Merifield et for plate anchors.
Vessel (AHV) and no
18 9.0  Experience in the Gulf
al. (2001) ROV.
of Guinea
DNV-RP- Table 4– SEPLA and torpedo anchor advantages.
13.3 9.5
E302 (2002)
Torpedo anchor (DPA)
Table 3 - Plate anchor holding capacities in the idealized STP SEPLA disadvantages
disadvantages
offshore conditions according to three different methods.
  Patented installation In order to relate the foundation technologies
 Patented installation
method. discussed with a real scenario, a specific region was
method.
 No experience outside chosen to assess the viability of two different
 Installation time
Brazil.
greater than for a foundation solutions. Proposed offshore developments
 Lack of documented
caisson. near São Tomé & Príncipe were chosen for the case
installation and design
 Requires keying and study, because it is a region that has recently gathered
methods with
proof loading. attention from the oil & gas industry, and has not been
verification agencies.
 Requires a ROV.
 Unknown orientation subjected to any platform construction yet.
 Limited field load tests.
once embedded.
Table 5 – SEPLA and torpedo anchor disadvantages Water depths in the zone of proposed offshore
development near São Tomé & Príncipe range from
Finally, when comparing the resistance obtained by 1800 m to 3000 m; thus, future installed facilities must
each foundation element, the SEPLA has a greater be floating platforms and hence the foundation
advantage. After applying safety factors in the systems in the seabed will be resisting tensile forces
proposed SEPLA, solution the predicted resistance is instead of compression. Therefore, the only types of
about 10 MN. For torpedo anchors the ultimate foundation solution suitable for this region are
resistance is predicted to be 8.7 MN, however the use anchoring systems. Anchoring systems that were
of a safety factor of 2 is advised (Eltaher et al., 2003), selected for evaluation in this case are those for which
which reduces the design resistance to about 4.4 MN. fewer studies and investigation have been made, but
Thus, double the number of torpedoes will be required on the other hand are likely to be the most economic
to provide the same load resistance as a single plate systems – in this case, torpedo anchors and suction
anchor. embedded plate anchors were evaluated.

In other words, the resistance obtained per unit weight Geotechnical characterization of the zones offshore
10000 𝑘𝑁
of the anchor element is higher for SEPLA, 50 𝑡𝑜𝑛𝑛𝑒𝑠 = from São Tomé & Príncipe was based on
4350 𝑘𝑁 investigations performed in the Gulf of Guinea over
200 𝑘𝑁⁄𝑡𝑜𝑛𝑛𝑒; than for the torpedo anchor, =
98 𝑡𝑜𝑛𝑛𝑒𝑠 more than 10 years. This large database on the
44 𝑘𝑁⁄𝑡𝑜𝑛𝑛𝑒. behaviour of deep-water sediments made possible the
assumption of several soil parameters. The soil in the
Even though, both torpedo and suction embedded
Gulf of Guinea typically is a highly sensitive clay (St
plate anchors have smaller unit costs in comparison
= 2 to 4), and one of the most important parameters is
with the other types of anchoring systems, one of them
the shear strength profile, it was apparent that it would
is more suitable than the other to be recommended for
have a positive gradient with depth of about 1.5 kPa/m.
use in this case-study. The best solution is the one
It also became clear that many sites in this region
which has more unit resistance per foundation
exhibit a greater resistance (up to about 15 kPa) in the
element, the design and installation method is more
first 2 m, this phenomenon is called a “crust”, and no
reliable, research has provided a more fundamental
unique or convincing explanation has been proposed
basis for understanding the method, and if possible,
for its existence so far. Other important properties,
should already been used in the Gulf of Guinea or a
which are not well defined yet, are the interface soil-
similar geotechnical region. Taking into account the
steel friction resistance and the set-up effects.
above criteria, the more appropriate solution for this
case study is considered to be the Suction Embedded The selected torpedo anchor to be employed in this
Plate Anchor. case study would be the same that was used in
Albacora Lest Field in Brazil Basin by Petrobras, and
5. Conclusion
it is the T-98 torpedo. The design of the torpedo anchor
This thesis has presented an outlined of existing was based on the design of simple cylindrical pile. Its
platform and associated foundation technologies used pull-out capacity comes from three different sources:
in offshore developments, and the most important shaft friction resistance, self-weight of the pile and
aspects involved in their design. Also, geotechnical reverse end bearing capacity. The shaft friction
characterization issues relating to the offshore resistance is the fraction that gives the greatest
environment such as topography, seabed composition contribution, thus the correct assessment of the
and geohazards were discussed. adhesion factor (α) plays an important role.
The assessment of the torpedo anchor free penetration amalgamation of various published data from
into the soil is also important to recognise the torpedo the Gulf generally.
embedment, and therefore the soil shear strength along  Understanding of the origin and
the torpedo shaft. Following O’Loughlin et al. (2013), characterisation of the near seabed “crust”
after releasing the torpedo from 108 m above the particular to this region and the effect this may
seafloor, it should reach a velocity of about 40 m/s have on foundation installation processes.
before impacting with the soil and the torpedo should  The interface friction resistance between the
penetrate about 40 m into the soil. An adhesion factor, soil and the steel elements in the short and
α=0.7 was considered, which corresponds to full long-term.
reconsolidation of the soil in the vicinity of the pile
after being remoulded by the torpedo penetration, Since there is no experience of torpedo anchors in the
consequently the pull-out resistance is expected to be Gulf of Guinea, it would be of great interest to develop
8.7 MN. in situ model scale tests to study the behaviour of the
torpedo during penetration of the soil, and its
The SEPLA solution considered is a 4.5 m x 10 m resistance in the short and long-term as well. As this
plate anchor, which is proposed by Wilde et al. (2001) in situ tests are very expensive, it would be very
for permanent installations. Plate anchor keying interesting to evaluate the influence of the near seabed
induced loss of embedment was calculated according “crust” on the torpedo penetration, using for that
to Song et al. (2009) and is expected to be about 0.81B computational programs or laboratorial tests.
(B is the anchor breadth), i.e. approximately 4 m.
A study to evaluate the forces that the platforms will
The holding capacity of the SEPLA is provided by the be subjected in the offshore of São Tomé & Principe
end bearing resistance plus the self-weight of the would be of great interest, because depending on those
anchor. This capacity was calculated according to forces the foundation solutions may be more or less
three different design procedures: economic, i.e. number and size of the foundation
elements.
 Wilde et al. (2001)
 Merifield et al. (2001) 7. References
 DNV-RP-E302 (2002)
Aguiar, C. S., de Sousa, J.R., and Ellwanger, G. B.
The most conservative procedure proved to be from (2009), “Análise da Interação Solo-Estrutura
DNV-RP-E302 (2002), with a holding capacity of de Âncoras do Tipo Torpedo para
about 9.5 MN. However, this do not take into account Plataformas Offshore”, PhD Thesis, UFRJ.
the self-weight of the anchor and includes a partial
Araújo, J. B., Machado, R. D., and Medeiros Jr., C.
resistance factor, therefore an additional 0.5 MN may
be added to provide a total resistance of 10 MN. P. (2004), “High Holding Power Torpedo

Based on a criteria that involved the resistance, Pile – Results for the First Long Term
installation process, experience, knowledge and Application”, Proceedings of the ASME 23rd
reliability of both anchoring systems, is was
OMAE Conference, 51201, Vancouver.
considered that the use of suction embedment plate
anchor systems would be more appropriate offshore
Brandão, F. E. N., Henriques, C. C. D., Araújo, J. B.,
from São Tome & Principe.
Ferreira, O. C. G. & Amaral, C. D. S. (2006).
6. Further Research
“Albacora Leste field development: FPSO P-
In terms of geotechnical issues associated with this
50 mooring system concept and installation”.
case study, there is still need for further studies, in
particular: Proc. Offshore Technol. Conf., Houston, TX,

 Site specific characterisation of the seabed in paper OTC 18243.

the São Tomé & Principe region, instead of
the broad characterisation of the Gulf of
Guinea made in this study and based on the
DNV-RP-E302 (2002). “Design and Installation of Medeiros Jr., C. J., Hassui, L. H. Machado, R. D.,
Plate Anchors in Clay- Recommended (1997). “Pile for Anchoring Floating
Structures and process for Installing the
Practice”, Det Norsk Veritas, 43 pages. Same.” United States Patent Number
6,106,199.
Dove, P., Treu, H., and Wilde, B. (1998). “Suction
Medeiros, C.J. (2001). “Torpedo anchor for deep
embedded plate anchor (SEPLA): A new
water”. Proc. Deepwater Offshore
anchoring solution for ultra-deep water Technology Conf., Rio de Janeiro.
mooring.” Proc., D.O.T. 10th Int. Conf. and Medeiros, C.J. (2002). “Low cost anchor system for
exhibition, Deep Offshore Technology, flexible risers in deep waters”. Proc. Annual
Offshore Technology Conf., Houston, Paper
London.
OTC 14151.
Eltaher, A., Rajapaska, Y., AND Chang, K. T. Merifield, R. S., Sloan, S. W. & Yu, H. S. (2001).
(2003). “Industry trends for design of “Stability of plate anchors in undrained clay”
anchoring systems for deepwater offshore Géotechnique 51, No. 2, 141-153.
structures”. Proc., Offshore Technology
O’Loughlin, C. D., Randolph, M. F., Richardson, M.
Conf., OTC No. 15265. (2004). “Experimental and Theoretical
Studies of Deep Penetrating.” Proceedings of
Fernandes, A. C., dos Santos, M. F., Araújo, J. B., the 36th Annual OTC, Texas, May 3-6, 2004,
Almeida, J. C. L., Diniz, R. and Matos, V. Paper nº. OTC 16841.
(2005). “Hydrodynamic aspects of the O’Loughlin, C. D., Richardson, M., Randolph, M. F.
torpedo anchor installation”. Proc. & Gaudin, C. (2013). “Penetration of
Dynamically Installed Anchors in Clay”.
International Conference on Offshore Géothecnique, 63(11), 909-919.
Mechanics and Artic Engineering, Halkidiki,
Puech, A., Dendani, H., Nauroy, J-F. and Meunier, J.
Greece, OMAE 2005-67201.
(2004). “Characterisation of gulf of guinea
Lieng, J. T., Hove, F., Tjelta, T. I. (1999). “Deep deepwater soils for geotechnical engineering:
Penetrating Anchor: Subseabed Deepwater Successes and Challenges”. 21-22 Octobre
Anchor Concept for Floaters and Other 2004, Seatech week Colloque
Installations.” Proceedings of the 9th “Caractérisation in situ des sols marins”
International Offshore and Polar Engineering Brest, France.
Conference, Brest, France, 30 May- 4 June
Raie, M. S., & Tassoulas, J. L. (2009). “Installation
1999, Vol. I, pp. 613-619. of torpedo anchors: numerical modeling”.
Journal of geotechnical and
Lieng, J. T., Kavli, A., Hove, F., Tjelta, T.I. (2000). geoenvironmental engineering, 135(12),
“Deep Penetrating Anchor: Further 1805-1813.

Development, Optimization and Capacity Randolph, M., Cassidy, M., Gourvenec, S., &
Erbrich, C. (2005, September). “Challenges
Verification.” Proceedings of the 10th
of offshore geotechnical engineering”. In
International Offshore and Polar Engineering Proceedings of the international conference
Conference, Seattle, Washington, 28 May- 2 on soil mechanics and geotechnical
June 2000, pp 410-416.
 engineering (Vol. 16, No. 1, p. 123). AA keying flap”. Ocean Engineering, 59, 107-
BALKEMA PUBLISHERS. 116. ISO 690.
Song, Z., Hu, Y., O’Loughlin, C., & Randolph, M. F.
Wilde, B., Treu, H., and Fulton, T. (2001). “Field
(2009). “Loss in anchor embedment during
testing of suction embedded plate anchors”.
plate anchor keying in clay”. Journal of
Proc., 11th Int. Offshore and Polar
geotechnical and geoenvironmental
Engineering Conf., Int. Society of Offshore
engineering, 135(10), 1475-1485.
and Polar Engineers, Mountain View, CA,
Sultan, N., Voisset, M., Marsset, T., Vernant, A.M., 544–551.
Cauquil, E., Colliat, J.L. & Curinier, V.
Yang, M., Aubeny, C. P., & Murff, J. D. (2011).
(2007). “Detection of free gas and gas
“Behaviour of suction embedded plate
hydrate based on 3D seismic data and cone
anchors during keying process”. Journal of
penetration testing: An example from the
Geotechnical and Geoenvironmental
Nigerian continental slope” .Marine Geology,
Engineering, 138(2), 174-183.
240, 235–255.

Wang, D., Gaudin, C., & Randolph, M. F. (2013).

“Large deformation finite element analysis
investigating the performance of anchor