OTC-27623-MS

The State of Knowledge of Pipe-Soil Interaction for On-Bottom Pipeline

Design

D. J. White, University of Western Australia; E. C. Clukey, Jukes Group; M. F. Randolph, University of Western
Australia; N. P. Boylan, Norwegian Geotechnical Inst.; M. F. Bransby, Fugro AG; A. Zakeri and A. J. Hill, BP; C.
Jaeck, Cathie Associates

Copyright 2017, Offshore Technology Conference

This paper was prepared for presentation at the Offshore Technology Conference held in Houston, Texas, USA, 1–4 May 2017.

This paper was selected for presentation by an OTC program committee following review of information contained in an abstract submitted by the author(s). Contents of
the paper have not been reviewed by the Offshore Technology Conference and are subject to correction by the author(s). The material does not necessarily reflect any
position of the Offshore Technology Conference, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Offshore Technology Conference is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgment of OTC copyright.

Abstract
The paper reviews recent advances in the understanding of pipe-soil interaction, and provides a updated
knowledge on best practices for on-bottom pipeline design. Since the late 1990s, major programs of research
work have been undertaken to develop appropriate models for pipe-soil interaction for seabed pipelines
in challenging environments and operating at high temperature and pressure, to mitigate design issues
associated with geohazards, hydrodynamic stability and thermal expansion management. Project-specific
programs of work have been extended into industry-wide Joint Industry Projects, and operating pipelines
are now providing field observations to validate and refine the design analyses. Much of this new knowledge
is now maturing into best practices that can be presented in codes and standards. This paper synthesises that
work, and provides recommendations of methodologies suited to codification that will guide future projects.
The paper has been authored by a team of practitioners and researchers that comprise a Technical Panel
working under the API/ISO geotechnical committee, and the paper sets out some of our views on future
additions to the API/ISO codes.
Recent advances in the treatment of pipe-soil interaction in pipeline design cover a range of aspects,
including (i) quantifying subaqeous flow (submarine slide) geohazards - slide runout behaviour, pipeline
impact loads and pipe deformation, (ii) predicting pipeline embedment, including the effects of the lay
process, and through-life changes due to sediment transport, (iii) modelling axial pipe-soil interaction,
including the strong influence of drainage and consolidation on soft soils, (iv) modelling lateral pipe-soil
interaction, including cyclic effects such as the growth of soil berms beside the pipe, (v) modelling scour
and self-burial, in regions of hydrodynamic activity, and the resulting changes in pipeline stability.
Many of these effects are complex, involving temporal changes in seabed bathymetry and soil strength.
However, they can also offer significant design efficiencies, providing a motivation to capture them
accurately. For example, self-burial of a pipeline through seabed mobility may lead to an improvement in
stability that reduces the requirement for weight coating or secondary stabilization works. Also, long-term
changes in seabed friction due to consolidation following each cycle of expansion and contraction may
lead to a progressive stabilization, reducing the need for anchoring. This paper includes examples where
2 OTC-27623-MS

it has been possible for methods emerging from research to be applied in practical design, validated by
observations from the laboratory or from operating pipelines.
Many aspects of modern methods for pipe-soil interaction analysis are reaching a level of maturity that
allows a consensus to be reached on best practices for design. This will unlock consistent and efficient
approaches for future pipeline systems, and for management of existing systems.

Introduction

Context. This paper has been prepared as part of the activities of a Technical Panel chaired by Dr. Ed Clukey
working under the common API/ISO geotechnical committee, which is composed of the API Geotechnical
Resource Group SC2/RG7 and the ISO TC67/SC7/WG10 group. These function as a common committee
and their main respective geotechnical codes are API RP2GEO (2014) and ISO 19901-4 (2016). These
codes are becoming progressively consistent and harmonized, and are being expanded to include advice
on pipeline geotechnics.
This paper provides the state-of-knowledge that should represent the basis for future updates to the
pipeline-seabed interaction (PSI) advice within these codes, and explains the background and motivation
for these updates.
Current API/ISO code guidance on pipeline geotechnics. The current guidance in these codes regarding
pipeline geotechnics was drafted for API RP2GEO (2014) and has been adopted in virtually identical form
in ISO (2016). The guidance is mainly informative and relates predominantly to surface-laid pipelines on
soft clays. The sub-sections are as follows:
1. Introduction, states that geotechnical advice should be sought to predict the as-laid pipeline
embedment and the resulting force-displacement responses in the axial and lateral directions,
emphasizing that both upper and lower estimates of these responses can be critical for a given limit
state
2. Actions on pipelines, are listed, including scour, seabed liquefaction, debris flow and turbidity current
impact, snag and impact loading, and also changing temperature and pressure which cause design
issues of walking and buckling.
3. Pipeline-soil interaction models are introduced with analogy to t-z (axial) and p-y (lateral) models
for pile-soil interaction, highlighting more complex features such as the soil berms that develop
progressively during large–amplitude cyclic motion.
4. The influence of drained and undrained soil behavior is outlined, with advice that designers should
compare the durations of actions (movements or loads) in order to select whether drained or undrained
conditions will prevail.
5. Vertical penetration of a pipeline into the seabed during laying is discussed qualitatively with reference
to the remoulding process during dynamic laying and the overstress created in the catenary touchdown
zone. References are given to bearing capacity solutions for undrained and drained conditions, but
noting that the dynamics of the lay process mean that the embedment will generally be greater than
predicted by static bearing capacity theory.
6. Axial soil resistance is discussed for drained conditions, in which the resistance is simply the drained
interface friction coefficient enhanced slightly by a ‘wedging’ effect around the curved pipe periphery.
It is noted that the resistance will be different in undrained conditions due to pore pressure effects,
but no quantitative advice is given.
7. Lateral soil resistance is discussed qualitatively with illustration of so-called ‘light pipe’ and ‘heavy
pipe’ responses derived from model test observations. Light pipes tend to rise to the soil surface,
whereas heavy pipes sink deeper and self-bury. Reference is made to empirical methods based on
OTC-27623-MS 3

combining ‘friction’ and ‘passive’ components of lateral resistance and also failure envelope methods
based on the combined vertical-horizontal bearing capacity.
The bulk of this advice is informative. Only one equation is specified, related to drained axial resistance.
No advice is given for buried pipelines, and no specific equations are given in the text for calculating pipeline
embedment, lateral resistance, undrained axial resistance or slide impact loading. Reference is made to 17
publications, 13 of which were produced in the past 10 years, highlighting the recent rapid advancement
of importance and understanding of this topic. However, overall the content is not as mature or ready for
practical application compared to the conventional sections of the API and ISO codes related to piles and
shallow foundations.
Motivations and drivers for updating the API/ISO code guidance. There are three drivers to update the
advice in these codes for the next revisions (due in 2018 for the API and 2020 for the ISO), and provide
more quantitative guidance on pipeline geotechnics, either within the code or by reference to other sources:
1. Advances in understanding – meaning that improved knowledge and calculation methods now exist.
2. Identified project benefits – meaning that projects are more aware of opportunities for cost and
reliability improvements via better pipe-soil interaction design practices, with this process being
achieved firstly through internal Operator communities of practice and then through revision of
external pan-industry codes.
3. Existing parallel guideline development – meaning that other organisations and Joint Industry
Projects have already completed significant efforts to reach a consensus on design methods for
pipeline-seabed interaction, which may be suitable for endorsement and adoption in the API and ISO
codes.
The various advances in understanding from recent research and project experience are outlined in the
following sub-sections, accompanied by digested summaries that identify material suitable for inclusion in
the next versions of the API and ISO codes.
Some new projects face cost and reliability challenges that can benefit from improved pipeline-seabed
interaction practices. These identified project benefits can be unlocked by the PSI process and the resulting
outcome. The best process includes (i) timely gathering of relevant site characterization data, (ii) efficient
(and early) generation of pipeline-seabed design parameters without signficnat changes throughout the
design progress, and (iii) smooth adoption of these parameters in the pipeline engineering work, minimizing
iterations or rework. The best outcome involves accurate design parameters relevant to the specific design
conditions that appropriately capture the uncertainty and variability. These are available as early as possible
in the project, and have a level of detail and complexity that is consistent with how critical pipe-seabed
interaction is for the overall pipeline design.
Some first steps towards codifying better PSI practice have been programs of industry-supported research
– such as the SAFEBUCK JIP (Atkins 2015, www.safebuck.com) – and investigations where unexpected
behavior has been observed in the field. Operators have captured this improved understanding internally,
for application to their own projects. At least three major operators have prepared internal guidance notes
or best practice documents drawing together external research and their internal observations and practices
related to pipe-soil interaction. The next step to maximize how projects can benefit from this knowledge
is to build a pan-industry consensus by migrating this knowledge into the API/ISO codes. Such pooling of
knowledge leads to a better understanding and improved practices, and maximizes the alignment between
operators and their contractors and consultants.
Benefits to projects from developing robust and reliable PSI models begin at the highest level of avoiding
loss of containment resulting from ruptured pipelines or damaged connections to other infrastructure.
Beyond that fundamental aim, the next most significant reason for progressing the subject is in reducing
4 OTC-27623-MS

CAPEX associated with oversized hardware and reducing OPEX by minimizing the need for intervention or
close monitoring due to uncertainties and low confidence in the anticipated operating response of pipeline
systems. The final category of benefits to having a well-defined approach to PSI is in the design phase
where inefficiencies can incur costs in terms of wasted engineer-hours, threats to project delivery schedules
or missed opportunities due to late resolution of design challenges.
Key motivations for development of PSI knowledge include the following: (1) Underestimation of axial
friction should be avoided, since this leads to underprediction of axial force profiles and the increased
likelihood of rogue buckles that could occur at sensitive structure locations, (2) conversely, over-estimation
of axial friction can lead to underprediction of end-expansions and hence the potential for over-stressed
end connections, (3) similarly for the lateral response, under-prediction of soil resistance can lead to over-
stressed buckle crowns and over-prediction can lead to underprediction of cyclic movements and hence
unconservative estimates of fatigue damage (or could suggest that unplanned buckles will not occur, when
there is acatully a significant likelihood that they will). These examples alone demonstrate that the notion
of conservatism is entirely design-case specific. Both high and low expected limits to the geotechnical
resistance must be quantified.
CAPEX savings that can be realized include proving the need (or otherwise) for anchor piles to prevent
pipeline walking. Such anchors may be seen as a complete solution to mitigating pipeline walking concerns,
but they need to be sized and positioned correctly to be effective. Anchors and their connection points
to pipelines are also costly to fabricate and install. Proving anchors are not needed can be a significant
design ‘win’ and depending on the uncertainty level, a middle-ground of deferring anchor installation is
increasingly common. Given that pipe walking is a progressive process there is an inherent early warning of
excessive movement provided a commitment to appropriate monitoring is made and intervention is possible.
This approach can be a cost-effective strategy. However, the approach is not always ideal since there is a
significant cost associated with manufacturing the connection points, and interventions involve significant
cost and time. Furthermore, the interaction of the anchor and the pipeline during its gradual tensioning adds
complexity to the global pipeline model, affecting the end movement and any lateral buckle and route-curve
pullout. It is therefore far preferable for a definitive decision to be made on whether anchors are required
or not, but this will often require a high level of refinement in the PSI model.
Oversized spools and/or termination structures can also be avoided if the PSI model is refined and
uncertainty is reduced. This reduces the installation complexity as well as the cost of fabrication. Large
spools designed to address the uncertainty in pipeline expansions and walking behavior can easily require
an increase in installation vessel capacity and be cumbersome to handle. Similarly, mudmats (whether
restrained or intentionally compliant, with mechanical sliders or on-seabed sliding designs) can easily
become unmanageable if conservatism is added to their geometry to account for end structure movements
that may or may not occur. Clearly, confidence in PSI models is a major contributor to resolving this issue.
As the subject matter matures, several lessons have been learned from operating pipeline systems,
and continued learnings will yield additional benefits for future designs. If the periodic pipeline surveys
during operation are of sufficient accuracy and resolution, the long-term PSI behavior can be extremely
enlightening. Time-dependent effects such as buckle stabilization, consolidation-related increases in axial
resistance and the impact of low and high-frequency cyclic variation in pipeline content density have had
observed effects on the global pipeline response. Integrity management surveys and full use of the data
they provide are invaluable, and in some projects behavior outside the planned design envelope has been
revealed. In the long term, distilling these learnings into more reliable PSI models will reduce the need for
such close monitoring.
Another benefit in improved PSI methodologies is in the achievement of more efficient design processes,
including (i) an early assessment of the criticality of the PSI model to the pipeline system design, (ii) better
planning of a fit-for-purpose geotechnical data collection scheme and (iii) earlier determination of the level
of PSI analysis complexity that is justified. Again, the optimum path is highly project-specific. There is
OTC-27623-MS 5

often the potential to miss opportunities for design oprtimisation or to add delay and cost through poor
survey planning that leads to inadequate seabed characterisation.
The structural design of HPHT pipelines is generally performaed within a structural reliability
framework, with likelihoods of lateral buckles forming being determined explicitly. Given this approach,
the geotechnical inputs should also be determined and expressed probabilistically (meaning, for example,
that low and high estimates should be paired with a likelihood of exceedance). Such probabilistic techniques
allow consistency between the geotechnical and structural design work and ensure that levels of risk are
rationally conveyed through geotechnical and structural aspects on to the wider project teams.
A common pitfall in HPHT pipeline design is to invoke complex numerical methods too soon in a project
when simpler approaches can steer projects in the right direction sooner and more efficiently. Finite element
analysis is critical and powerful if used at the right time and with due consideration to the number of runs
and parametric variations that are likely to yield the most value. Inefficiencies in the engineering process –
including both geotechnical analysis and pipeline structural modelling – can not only cause wasted effort,
but sometimes threaten project delivery schedules. Well-defined PSI models that appropriately address the
relevant design challenges avoid these pitfalls.
The work of the API/ISO Technical Panel has been only a small component of the overall effort expended
by industry and academia on pipeline-seabed interaction research and technology development. There are
already significant volumes of synthesized research and design guidelines on pipeline-seabed interaction in
the public domain which form a starting point for updated API/ISO code advice. These include the following
key contributions:
1. SAFEBUCK Guideline (2015): The SAFEBUCK Joint Industry Project on the design of pipelines
with lateral buckling closed in 2014 with the production of a Guideline,that underwent a detailed
review and revision process overseen by the 13 Operator and Contractor participants as well as
DNV. The Guideline features a 51-page Appendix on pipe-soil interaction written by Atkins and the
University of Western Australia. This Appendix provides detailed recommendations for calculating
the embedment and force-displacement responses for surface-laid pipelines, illustrated by worked
examples. Much of this advice is based on published research.
2. DNV F110 – SAFEBUCK Merged Guideline (2015): Following completion of the SAFEBUCK JIP,
a merged guideline was written by Atkins and DNV by combining the SAFEBUCK Guideline with
DNV F110 (Design of Pipelines for Global Buckling). The Merged Guideline contains a slightly
shortened version of the SAFEBUCK PSI Appendix and also includes a 28-page Appendix on pipe-
soil interaction for buried pipelines (which is based on the earlier 2007 version of DNV F110)
3. DNV F114 (Pipe-soil Interaction for Submarine Pipelines): Work is currently underway by DNV
to develop a unified pipe-soil interaction recommended practice, drawing together the elements of
pipeline geotechnics in all DNV codes (covering global buckling of on-bottom and buried pipelines
(F110), on-bottom stability (F109), pipeline protection (F107), pipeline-trawlgear interaction (F111)
and free span analysis (F105)). Industry organisations and university partners are involved in this
effort.
The DNV F110 – SAFEBUCK Merged Guideline and DNV F114 are due for first public release in 2017.

Pipe-soil interaction aspects in pipeline design

Many aspects of offshore pipeline design are influenced by geotechnical considerations and the expected
pipe-soil interaction response, including (from White & Bransby 2017):

• Route selection, to minimise risks associated with geohazards

• In-place stresses after pipe laying, due to undulations of the seabed (‘bottom roughness’)
6 OTC-27623-MS

• Hydrodynamic (on-bottom) stability of pipelines, flowlines, umbilicals and risers under wave and
current loading
• Pipeline and flowline responses to thermal and pressure-induced loading, including upheaval
buckling, lateral buckling, axial ‘walking’ and axial end-expansions
• Fatigue and overloading, where pipelines (in spans) and risers touch down at the seabed

• Installation – including selection and optimization of any burial method (trenching, ploughing,
jetting) or stabilization scheme (matresses, rockdump, anchors etc)
Some of these design aspects also apply to short lengths of pipeline known as spools, which provide
connections between subsea infrastructure and the pipelines and allow for pipeline end expansion. However,
there are also significant differences due to the 3D shape and short length of spools (with the end connections
affecting the vertical spool-soil contact force); their installation method is also different, with spools lowered
statically into position, rather than being laid dynamically from a vessel. The expected pipe-soil interaction
response must therefore be predicted both for long lengths of on-bottom or buried pipeline, and also for
any local variations in this behavior where pipelines terminate or are supported on crossings or by buckle
initiation structures; these various situations are illustrated in Figure 1.

Figure 1—Pipeline-seabed and pipeline-infrastructure interactions

Forms of pipe-soil interaction model

The key PSI parameters for design are the limiting resistances during axial or lateral pipe movement (for
both buried and on-bottom pipelines). For buried pipelines, the uplift resistance is of primary importance.
PSI can be incorporated in the structural analysis of a pipeline in various ways, which can be summarised
as follows, in order of increasing complexity (after White & Bransby 2017):
1. As a single limiting value of axial or lateral resistance (or friction factor) – a rigid-plastic response.
2. As independent force-displacement responses (non-linear ‘springs’) in the axial and/or lateral and/or
vertical directions. This is the most common method of incorporating PSI in a finite element analysis
of a pipeline, and is akin to the ‘t-z’ and ‘p-y’ load transfer methods of analysing pile response. Simple
OTC-27623-MS 7

force-displacement (or friction-displacement) responses can be bilinear (elastic perfectly-plastic), tri-

linear (with an initial peak) or piecewise linear. Additional rules can be introduced to define the cyclic
behaviour, capturing the influence of any trenches and/or soil berms that form during cyclic lateral
movement.
3. As a general vertical-lateral response model, based on plasticity theory, which is often termed a
‘macroelement’ (Zhang et al. 2002, Tian & Cassidy 2008). These models allow more complex
coupling between vertical and horizontal lateral loads compared to independent horizontal and vertical
springs.
4. Through explicit modelling of the soil continuum, pipe and pipe-soil interface (e.g. Konuk & Yu
2007, Martin et al. 2013). This is computationally expensive for a typical pipeline buckling or
stability analysis, although efficiencies can be made by simulating the soil domain via widely-spaced
perpendicular slices (Martin et al. 2013).
Approaches (3) and (4) are predominantly research tools and are used only occasionally in projects.
Future revision of the API/ISO codes will likely focus on quantitative advice for approaches (1) and (2),
while permitting the more complex approaches to be used if desired.
Pipe-soil interaction forces vary depending on the drainage condition, just like all other geotechnical
capacities. The general advice in API RP2GEO (2014) for shallow foundations is that "undrained
calculations are typically adopted where no drainage, and hence no dissipation of excess pore pressure,
occurs during loading. This may occur as a result of the rate of loading or the impermeable nature of the
soil. In contrast, drained calculations are typically adopted where no excess pore pressures arise during
loading." (Section 7, Stability of shallow foundations, sub-section 7.2, Principles). The same logic applies
for pipeline-seabed interaction, and calculation methods should be divided into undrained and drained
approaches rather than separating the recommendations into sand and clay conditions. Furthermore, partially
drained conditions can be relevant and approaches should be able to account for the resulting soil behavior.

Predicting initial pipeline embedment

For pipelines laid directly on the seabed, rather than being trenched, the embedment depth has a significant
influence on the axial and lateral resistance, as discussed later. In relatively strong soils, such as sands
and even rocky outcrops, embedment will be a small fraction of the diameter and (in the case of sands)
will be subject to change during the operational life. Typically intermittent zones of sediments build up
(hence ‘embedment’) and spanning occur, changing locations due to sediment transport during storms, but
generally with the average embedment tending to increase with time (Leckie et al. 2015). Periodic surveying
of the pipeline for such seabed conditions is essential in order to monitor the condition of the pipeline, and
hence assess the stability of both the pipeline and the surface seabed sediments.
For softer seabed conditions such as fine-grained clays, carbonate muds and silts, the pipeline embedment
not affected by sediment transport but is affected by conditions during laying, which is generally assumed
to be an undrained event. Analytical solutions for the vertical penetration resistance of a pipe in fine-grained
material (Murff et al. 1989, Aubeny et al. 2005) form the basis from which to estimate embedment. However,
embedment due to the (buoyant) pipeline weight alone must be adjusted (a) for the local force concentration
that occurs within the so-called touchdown zone (where the pipeline catenary emanating from the lay vessel
reaches the seabed); and (b) additional dynamic effects due to motion of the lay vessel. These can lead to
embedment depths that are an order of magnitude greater than that due to the pipeline weight alone, but in
addition the effect of dynamic motions can lead to significant variability along a pipeline route, depending
on the metocean conditions during lay (Westgate et al. 2010, 2012). Methods to assess these different aspects
are summarized below.
Static penetration resistance. It is customary to express the vertical resistance V of a cylinder penetrating
into fine-grained soils (i.e. undrained conditions) as the sum of two resistance terms, one related to the shear
8 OTC-27623-MS

strength at invert level suinvert and one representing the buoyancy force due to the submerged unit weight,
γ', of the displaced soil:

(1)

where D is the pipeline diameter, w the pipe invert penetration and A’ is the embedded cross-sectional area
of the pipe. The constants c and b vary with the shear strength profile and the pipe-soil roughness (Aubeny
et al. 2005, Randolph & White 2008a, Chatterjee et al. 2012a). However, values of c ~ 6 and b ~ 0.25 are
commonly assumed, and sometimes a cut-off function is applied close to the surface (for w/D < 0.1). The
buoyancy factor, fb, represents the enhancement of the buoyancy force due to the soil heave that occurs
around the pipe, and a value of fb = 1.5 is typically appropriate (Merifield et al. 2009, Chatterjee et al. 2012a).
In soft soils (i.e. low strength gradients compared with the effective unit weight γ’ of the soil), the
buoyancy term can represent a significant fraction of the penetration resistance, so needs to be taken into
account, as detailed by Chatterjee et al. (2012a).
Force concentration. For a pipeline of submerged weight (per unit length) W', the maximum vertical
load Vmax within the touchdown zone is a function of the pipeline bending rigidity EI, the lay tension T0
(horizontal component, as relevant near the seabed) and a seabed stiffness k (ratio of penetration resistance
to penetration (V/w) in units of modulus or kN/m/m). An approximate expression for this relationship is
(Randolph & White 2008)

(2)

This expression is accurate in the range T01.5 / ((EI)0.5 W') > 1, based on numerical analysis (Randolph
& White 2008a).
Dynamic lay effects. The most uncertain aspect of predicting as-laid pipeline embedment is in allowing for
the effects of dynamic motions of the pipeline during lay. There are three primary phenomena that contribute
to increased penetration relative to the static lay conditions:
1. dynamic motions lead to remoulding of the seabed sediments, reducing the strength and hence
increasing penetration for a given vertical load;
2. enhancement of the vertical load above that given by equation (2)) due to dynamic amplification;
3. ploughing actions under lateral components of the motion, physically scraping material away to each
side of the pipe.
All three of these depend on the metocean conditions, and (at least for the first and third) the length of
time that each section of the pipe is within the zone. The relevant time, and hence number of motion cycles,
may be estimated from the lay rate, which will typically range from 30 to 300 m/hr, and the length of the
‘mobile’ part of the touchdown zone. This is commonly taken as the characteristic length, λ = (EI/T0)0.5,
with typical values in the range 20 to 40 m. During continuous pipe-lay, with, say, two pipe sections (24 m
total) welded at a time, the time within the touchdown zone is thus comparable with the welding sequence
time. Delays due to issues in the firing line, or poor metocean conditions, can lead to much greater times
within the touchdown zone, resulting in deeper local embedment of the pipe, while rapid lay (such as during
the final lay-down of the suspended catenary) will lead to reduced embedment (Westgate et al. 2012).
For projects where pipeline axial and lateral resistance, and hence estimation of embedment, is critical,
centrifuge model tests may be conducted using seabed sediment sourced from the pipeline route, and detailed
modelling of the ‘design’ dynamically enhanced pipeline forces and motions of the pipe during lay (Gaudin
OTC-27623-MS 9

& White 2012). Such tests allow sensitivity studies in order to gauge best, lower and upper estimates of
pipeline embedment, although it is vital that sample reconstitution is performed carefully. The characteristics
of the model soil – such as the water content, strength and sensitivity – must be compared to the in situ
material and any differences taken into account during the interpretation.
Detailed calculation models for estimating pipeline embedment, using data from cyclic full-flow
penetrometer testing and taking account of the estimated pipe motions and numbers of cycles, have been
proposed (Westgate et al. 2013). Even without such detailed analysis, a base-level approach comprises the
following steps:
1. Evaluate the submerged pipeline weight and the maximum force Vmax arising from the (static) lay
conditions (see equation (2))).
2. Evaluate the pipeline embedment using equation (1)), or equivalent solutions, but using the fully
remoulded shear strength profile for the seabed sediments.
This approach has been validated using field observations of pipeline embedment. The SAFEBUCK
JIP assembled a database of >50 pipelines and found that this method gave a mean ratio of measured
to predicted average embedment of 1.06, using the best estimate remoulded soil strength (Atkins 2014).
However, there are some additional effects shown by field observations. The approach underestimates the
embedment during periods of high swell or downtime during lay, while overestimating it during benign
conditions or during final lay-down of the pipeline catenary (Westgate et al. 2012, 2013). Essentially the
assumption of fully remoulded conditions compensates approximately for ignoring the effects of dynamic
amplification of the pipe force and lateral ‘ploughing’ of soil from beneath the pipe.
Methods for predicting pipeline embedment in coarse-grained soils where drained conditions prevail
are less mature. In these conditions, general theoretical solutions are more difficult to generate due to the
additional influence of soil density and dilatancy as well as the friction angle. Experimental observations
show an approximately linear relationship between penetration resistance divided by pipe diameter and
embedment (Zhang et al. 2002), when a pipe is pushed vertically into a bed of soil. It is common to estimate
the slope of this response directly from the CPT profile (e.g. Westgate et al. 2012), rather than attempting
to link the response to soil strength or stiffness properties. However, often the long-term embedment of
a pipeline laid on coarse-grained soils is controlled by sediment transport, since such materials are more
commonly found in shallow waters, where seabed currents and wave action are significaint – see the later
section on seabed mobility.
In coarse-grained soil conditions, field observations of pipeline embedment remain valuable to help
develop and validate prediction methods of embedment. In fine-grained soils, field observations of
embedment can also be used to confirm and narrow design ranges of embedment. In some situations, a
re-analysis of the pipeline response may be justified based on the observed embedment to confirm the in-
service response, and finite element analyses can include longitudinal profiles of friction factor that are
derived from the observed embedment (e.g. Sriskandarajah et al. 2015).

Modelling axial pipe-soil interaction

The axial resistance between an on-bottom pipeline and the seabed is controlled by the effective stress
friction – which may be influenced by any excess pore pressure – acting around the pipe-soil interface.
This friction is mobilized each time the pipeline expands or contracts or moves axially due to any other
action. Examples of the axial force-displacement response observed in experiments on clay are shown in
Figure 2. The observations shown in Figure 2a are from a 120 mm diameter model pipe tested on soft
Onsoy clay in a large test tank at NGI in Oslo (Smith & White 2014). The observations in Figure 2b are
from a deployment offshore West Africa of the Fugro SMARTPIPE in situ tool, which comprises a 225 mm
diameter model pipe attached to a seabed frame (Hill et al. 2008, Ballard et al. 2013a,b, White et al. 2010).
10 OTC-27623-MS

Since pipeline expansions are typically on the order of hundreds of millimetres, it is the axial resistance at
large displacements (the ‘residual’ resistance) that is the main focus in design. However, the brittle peak
that is sometimes observed at small movements can also influence the global pipeline response.

Figure 2—Observed axial pipe-soil response during pipe movement on soft clay (a) Large scale
model test (Smith & White 2014) (b) In situ Fugro SMARTPIPE tests (Ballard et al. 2013a,b)

The results in Figure 2 are illustrative of the behavior observed across a wide range of model tests
performed over the past decade (e.g. Jewell & Ballard 2011, White et al. 2011, Ballard et al. 2013a,b, Boylan
et al. 2014). This work has led to a relatively simple framework for describing axial pipe-soil resistance
which is illustrated schematically in Figure 3:

Figure 3—General model for residual axial pipe-soil resistance

• Axial pipeline movement involves shear failure at or close to the pipe-soil interface. The submerged
pipe weight, W’ controls the normal effective stresses at the pipeline–soil interface. The integrated
normal contact stresses around the pipeline periphery exceed the vertical contact force due to the
curved shape of the pipe surface. The total normal force may be expressed as N = ζW’ where the
wedging factor, ζ, can be estimated for a given embedment from an assumed stress distribution
around the pipe periphery. If the pipe is idealized as a line load on an elastic medium, ζ ranges
OTC-27623-MS 11

from 1 to 1.27 as the embedment increases from zero to half a diameter (White & Randolph 2007).
Numerical analyses give similar results (Krost et al. 2011).
• In drained conditions, the steady (‘residual’) axial resistance is controlled by the large-displacement
interface friction angle, δres, enhanced by the wedging effect. Laboratory tests show that δres varies
with stress level and interface roughness (Najjar et al. 2007, Hill et al. 2012, White et al. 2012,
Figure 3c). The drained residual axial resistance, FA,res,d, expressed as a friction factor, FA,res,d/W', is

(3)

• In undrained conditions, a simple model can be used on soft normally-consolidated (or lightly
over-consolidated) clay, where it is assumed that the preconsolidation pressure of the surficial soil
contacting the pipe is controlled by the pipe weight (and that this weight has been applied to the
seabed for sufficient time for the soil to consolidate under this load). Using this assumption, a
SHANSEP or Cam clay-type model (Muir Wood 1990) allows the undrained residual axial friction
factor, FA,res,u/W’ to be written as

(4)

• The parameter Rnc = (su-int-res/σ'no)nc is the normally-consolidated interface undrained strength ratio,
at large displacements. This parameter is akin to the strength ratio for normally consolidated clay,
but applicable to the interface (and at large relative displacements) rather than soil-soil failure. The
index parameter m is equivalent to the plastic volumetric strain ratio in Cam clay or the SHANSEP
over-consolidation index. The apparent overconsolidation ratio, OCR, is the ratio between the
previous maximum pipe weight (e.g. when flooded for hydrotesting), W'max and the current pipe
weight, W'. This reflects the level of over-consolidation and therefore enhanced strength created by
any previous elevated pipe weight, assuming that elevated weight is sustained for sufficient time
for full excess pore pressure dissipation. The parameter m is typically in the range 0.5-1.
• The interface model parameters depend on the soil and interface type and can all be measured in
low stress interface shear box tests (Hill et al. 2012, White et al. 2012, Boukpeti & White 2017),
or via shallow penetrometer devices (Yan et al. 2011, Yan et al. 2016). The drained parameter, δ,
can be measured alternatively via tilt table tests (Najjar et al. 2007).
• The relevant drainage condition depends on the duration of the pipe movement. In clay it is common
that the drainage period exceeds the duration of any startup or shutdown event, but is shorter
than the intervening operational period, although this depends on the consolidation coefficient,
cv. In clean sands it is common that the drainage occurs more rapidly than the mobilization of
axial resistance, so fully drained conditions apply throughout. More information on timescales of
drainage during axial pipe movement is given by Randolph et al. (2012) and Yan et al. (2014). If
the drainage process occurs concurrently with continuous pipe movement, it leads to a transition
in axial resistance from undrained to drained values that can be quantified by the period required
for 50% of this transition, t50 (Figure 3b).
• For soft clays, the drainage between movements leads to contraction and an increase in undrained
strength. This process of ‘consolidation hardening’ causes the axial resistance to rise towards the
drained value over several cycles. This effect is clear in Figure 2a and 2b, has been observed in
other model tests (White 2014) and in situ SMARTPIPE tests (Ballard et al. 2013a,b), as well as
being replicated in coupled numerical analysis (Yan et al. 2014). Field observations of pipeline
walking and expansion behavior, while complex to interpret, appear to confirm a build up in axial
resistance over the operating life of pipelines laid on soft clays and silts (Hill & White 2015, Peek
et al. 2017).
12 OTC-27623-MS

• The same process from episodes of shearing and consolidation is seen in over-consolidated
sediments. This is especially the case with light pipes and/or cases where the near surface soils have
a high shear strength crust in about the top 1m of sediment (especially when created artificially for
model tests, via consolidation of the clay bed using a surcharge). These cases mobilize negative
pore pressures during axial sliding and then swell and soften during the intervening consolidation
period. This leads to ‘consolidation softening’, as the resistance converges towards the drained
limit, as shown in physical model tests reported by Boylan et al. (2014).
• The axial mobilisation response is non-linear, as shown in Figure 2, but is commonly modelled
as elastic-perfectly plastic in practice. The distance to fully mobilize the resistance is observed to
be smaller in drained conditions than in undrained conditions, and some results suggest that the
mobilisation distance also varies with the rate of movement (White et al. 2010).
• A peak in resistance is commonly observed for the first cycle of movement on clay, but decays
with further cycles of movement (Figure 2).
This general framework is consistent with the approach adopted in SAFEBUCK (2015), DNV (2015)
and DNV (2017), and provides a basis for future API/ISO recommended practices. A key principle of the
models outlined above is that the limiting axial resistance is defined entirely in terms of soil parameters that
can be routinely measured in the laboratory or in situ. This is preferable to previous code guidance with
‘hard wired’ parameters that represent the soil-interface friction coefficient. In reality these parameters vary
with soil and pipe coating type.

Modelling lateral pipe-soil interaction

Lateral pipe–soil resistance during break-out and large amplitude cyclic movement is influenced by the
seabed geometry around the pipeline (normally simplified as the local pipe embedment, w) immediately
prior to pipeline loading, the vertical load acting on the pipeline (W'; usually equal to the self-weight) and
the soil properties.
Two characteristic types of large amplitude lateral response are typically observed depending on the ratio
of the pipeline weight to the seabed strength (characterized by W'/suD for undrained conditions):
– For ‘light’ pipes (with values of W'/suD below about two and typically also for drained conditions),
the pipeline tends to rise after breaking out from its initial position. As the pipe starts to rise and push
soil in front of it, a break-out soil resistance (Hbrk) is generated (Figure 4a). As the pipeline continues
to rise, the lateral resistance reduces from the break-out value to a residual resistance (Hres) where the
pipeline sweeps horizontally with a berm of soil being pushed ahead of the pipe. Subsequent cycles of
lateral movement lead to a steady increase in the restraint provided by the increasing size soil berms
deposited at the end of previous cycles and any further embedment into the seabed (Figure 4b) which
is important for accurate modelling of pipeline fatigue (Bruton et al. 2007).
– For values of W'/suD greater than about two (‘heavy’ pipes), the pipeline typically moves downwards
after the initial break-out resistance is mobilized. This downward movement, coupled with the growth
of a soil berm ahead of the pipe, leads to a steady increase in the lateral resistance (Figure 4b).
OTC-27623-MS 13

Figure 4—Large amplitude cyclic lateral response for light-pipe conditions (after White & Bransby 2017)

The potential range of Hbrk and Hres (and associated mobilization displacements, ybrk and yres) depends on
the range of geotechnical conditions along a given section of a pipeline as well as the range in expected
embedment and the embedment mechanism. Both high and low estimates of resistance can govern the
critical pipeline response. Calculations of embedment and lateral resistance must therefore consider the
full range of seabed and embedment conditions. This can be achieved deterministically (e.g. by combining
‘best’ or ‘worst’ combinations of parameters in individual calculations), or probabilistically (by considering
the statistical distribution of input variables and peforming calculations explicitly to determine the range of
resistances – e.g. via a Monte Carlo method: White & Cathie 2010, Bransby et al. 2011). Deterministic and
probabilistic approaches require an efficient calculation method to evaluate of Hbrk and Hres for the relevant
combinations of pipeline embedment, pipeline weight (at the time of lateral movement – which may be
water- or product-filled), and seabed geotechnical properties.
Early methods developed for calculating lateral seabed resistance (e.g. Hbrk and Hres) were empirical
correlations calibrated to model tests (e.g. Verley & Sotberg 1992, Verley & Lund 1995, Bruton et al. 2006,
White & Cheuk 2010). These methods are currently encapsulated in recent codes (e.g. DNV-RP-F109 and
the SAFEBUCK Guideline) and generally express Hbrk and Hres as the sum of components that vary with
pipe weight (‘friction’) and soil strength (‘passive’). However, since the expressions evolved from distinct
sets of model tests they can give widely differing predictions for the same conditions. They may be reliable
only for the conditions for which the model tests were performed, in particular (for undrained conditions)
pipe weight relative to soil strength, W'/suD, soil sensitivity, St, strength, su, and strength ratio (su/γ'D). Many
of the expressions do not include all the governing soil parameters so by definition will be unreliable for
other seabed conditions.
The SAFEBUCK JIP collated a database of centrifuge and large-scale laboratory tests of undrained lateral
pipe-soil behavior which comprised 67 measurements of Hbrk and 44 measurements of Hres (many tests did
not involve sufficient displacement to mobilized a clear value of Hres). These results were used to calibrate
the following expressions (White & Cheuk 2010):

(5)
14 OTC-27623-MS

(6)

The expression for Hbrk is a non-linearised version of the ‘friction + passive’ format, with an additional
term to define the passive resistance arising from the soil weight. It is limited by its format of summing
individual terms that have no interaction. Outside of the database range the expression diverges from
theoretical solutions (and indeed other subsequent model tests that were not part of the original database).
As a result, Equation 5 is reasonably reliable close to the original data, and overlies closely the theoretical
yield envelopes for the same conditions. However, it can over or underpredict Hbrk for lighter or heavier
pipes, where it also diverges from the yield envelopes. The expression for Hres includes a dependency on the
embedment prior to breakout. This is evident in the model test database, albeit among significant scatter,
and can be attributed to deeper pipes pushing a larger initial berm of soil across the seabed, causing Hres
to be higher.
An alternative and more robust approach for predicting Hbrk uses yield envelopes (or interaction diagrams)
in vertical and horizontal resistance space that bound the allowable loads for a given set of pipeline and
soil parameters (e.g. Randolph & White 2008b, Merifield et al. 2008). The approach is similar to the
methodology now accepted as an alternative for shallow foundation capacity assessment in the API/ISO
codes. They have the advantage of being rigorously based on plasticity limit analysis – consistent with
codified methods for shallow foundation capacity – and can properly capture the interaction between vertical
pipe load (or self-weight) and lateral resistance. This is an improvement over the ‘friction + passive’ methods
which diverge from theory when considering very light (e.g. buoyancy-coated) or heavy pipes.
Yield envelopes from exhaustive limit analysis parametric studies for simple strength profiles (constant
and proportional to depth) are presented by Martin & White (2012), extending earlier studies (Randolph
& White 2008b, Merifield et al. 2008). The results are commonly used in look-up table format for
design purposes (White et al. 2015). For more complex strength profiles, project-specific analyses can be
performed, e.g. using finite element limit analysis software.
To assess the validity of the empirical and yield envelope approaches, laboratory model tests (e.g.
Langford et al. 2007, Meyer et al 2016), centrifuge tests (e.g. Dingle et al. 2008, White & Gaudin 2008),
SMARTPIPE seabed model tests (Hill et al. 2012, Ballard et al. 2013), numerical methods (e.g. Chatterjee
et al. 2012b) and the back-analysis of operating pipelines (e.g. Westgate et al. 2012) can be used.
Forthcoming revisions to the API/ISO codes will require a choice to be made between empirical model
test-based expressions for Hbrk, and the use of yield enevlopes in undrained conditions. A useful parallel
is the equivalent advice in the API/ISO codes for shallow foundation capacity. This advice has pivoted
towards yield envelope methods in recent years for reasons of rigour, accuracy and practicality, and the
PSI advice could leap directly to yield envelopes of the sort set out by Merifield et al. (2008) and Martin
& White (2012). To achieve this, it will be necessary to document some validation of the yield envelope
approach relative to the available laboratory, centrifuge and field test evidence, equivalent to the validation
process that led to the empirical expressions for Hbrk prior to their adoption in the SAFEBUCK Guideline
(which is contained in White & Cheuk 2010).
For drained conditions, no general theoretical solutions yet exist for the prediction of Hbrk. Empirical
methods (e.g. Verley & Sotberg 1992) are likely to remain in common use, but it is often forgotten that
these methods have limited applicability due to the specificity of the underlying model test databases. For
example, the Verley & Sotberg model for Hbrk and Hres in sand is based predominantly on model tests with
embedment of w/D < 0.2, and normalized pipe weight of V/γ'D2 < 0.4. Some projects have found that this
range is outside the design conditions.
Project-specific numerical models – typically plane strain finite element analyses – of the pipeline
breakout response are becoming more straightforward to perform to mitigate the uncertainties and
OTC-27623-MS 15

limitations associated with other approaches. In addition, many large pipeline projects have performed
project-specific laboratory, centrifuge or field pipe-soil interaction tests on site-specific soil samples, to
validate or refine the adopted PSI calculation methods.
Beyond the prediction of Hbrk and Hres, further refinement of the lateral PSI model may be of value to
projects. The following list summarises aspects of PSI analysis that have been adopted in recent projects
to reduce critical areas of uncertainty or to address conditions where standard prediction models may not
be appropriate.
a. Site-specific soil profiles. For unconventional burial states (e.g. because sediment mobility leads
to large embedments) or for unusual seabed strength profiles, site-specific numerical analysis may
be required to define Hbrk and Hres, rather than rely on published envelopes for specific (constant or
depth-proportional) soil strength profiles. These analyses can be used either to evaluate Hbrk directly
or to produce site-specific V-H failure envelopes that are then used in a Monte Carlo probabilistic
assessment of Hbrk. Large deformation finite element analysis has also been used to calculate load-
displacement responses, but this requires careful parameter selection – e.g. via calibration to site-
specific centrifuge tests – in particular in relation to strain softening and remolding of the disturbed
soil.
b. Partial drainage. Given the wide range of potential loading durations or lateral displacement rates
(e.g. from fast hydrodynamic break-out events to slow lateral expansions) and wide range of seabed
soil types, the response is not always either fully undrained or fully undrained. In such cases a
conservative approach is to consider the extremes of the drained or undrained soil responses when
selecting low and high estimates. A less conservative approach is to consider explicitly the likely range
of drainage conditions when performing lateral friction factor calculations. The effect of uncertain
drainage condition can be wrapped into a probabilistic Monte Carlo analysis of Hbrk and Hres.
c. Spatial averaging. Field observations of pipe lay embedment show that embedment varies spatially
along a pipeline. This, alongside the geological variability along a pipeline, can be considered in
selecting the design ranges by considering different lengths of influence (e.g. Westgate et al. 2014,
Bransby et al. 2015, Westgate et al. 2016). For example, the length over which an unplanned buckle
could form is much shorter than the length over which the axial force that would drive this buckle is
mobilized. Therefore, the local variability in embedment should have a strong influence on the design
lateral resistance for unplanned buckles, whereas for axial resistance the variations in friction due
to local undulations in embedment will average out. As a result, the design range of axial resistance
should consider the averaging or ‘smearing’ of variations from local embedment, but the design range
for lateral resistance should not. The uncertainty in axial resistance form other effects – such as
drainage – would not be eliminated by this effect and should remain.
d. PSI for specific buckle initiators. The pipeline embedment and vertical loading close to buckle
initiators (e.g. in the touch-down zone on either sider of a structural buckle initiator, or along a
buoyancy module) can be very different to that elsewhere along the pipeline. Consequently, the use of
‘free-field’ lateral friction factors (i.e. based on a vertical pipe-soil load equal to the pipe self-weight,
and the average as-laid embedment) is inappropriate, leading either to unsafe or overconservative
design. This is explored in more detail by Bransby et al. (2017).
e. Cyclic lateral buckling. For cyclic lateral buckling, SAFEBUCK provides design guidance based
on the results of model tests donated to the JIP. This data is limited to a range of soil types and the
conditions examined in the model tests. It can be advantageous to perform site- and project-specific
model testing in order to reduce the uncertainty of the cyclic lateral bucking parameters compared
to the SAFEBUCK guidance.
16 OTC-27623-MS

f. Spools. Spools are installed differently to pipelines and are subject to different loading conditions
when tied in to structures and during operation. These differences should be considered explicitly
when providing values of lateral (and axial and vertical) seabed resistance (Bransby et al. 2017).
g. Consideration of scour/sediment mobility. This is discussed in the following section.

The effects of fluid-pipe-seabed interaction – sediment transport and scour

The traditional approaches of pipeline design separate the fluid-pipe-soil interaction into two separate
interactions, namely:
i. Fluid-pipe interaction: for assessing hydrodynamic forces on the pipeline.
ii. Pipe-seabed interaction: for determining forces and resistances between the pipeline and the seabed,
assuming that the pipe embedment is modified only by movement of the pipeline, and not by fluid-
seabed interaction.
However, it has long been known that fluid-pipe-seabed interaction can modify the seabed condition
through processes such as scour, sediment mobility and liquefaction (e.g. Palmer, 1996). Current
understanding of this tripartite fluid-pipe-seabed interaction, and how consideration of it in design can be
beneficial for projects, has developed greatly in recent years.
Seabed sediments become mobile when the shear stresses arising from bottom currents and waves (in
sufficiently shallow water depths) at the seabed exceed the critical shear stress of the sediment. When a
pipeline is resting on the seabed, exposed to the bottom currents, and the seabed is erodible, the action of
bottom currents (which will be increased locally to the pipeline) may lead to scour of the seabed beneath
the pipeline resulting in suspended free spans of the pipeline (Sumer & Fredsøe 2002). The suspended
sections of the pipeline may sag into the scour hole, or sufficient length of the pipeline may be undermined
to cause the remaining span shoulders to fail in bearing due to the pipe weight, resulting in lowering of the
pipeline relative to the seabed. If the pipeline sags or embeds further into the seabed, the altered topography
may cause deposition of sediment around the pipe, leading to partial or even full burial of the pipeline. In
other areas where the span lengths are short or the pipeline is light, the pipeline may not embed further.
Instead, sections of the pipeline may remain in span above the seabed. Observations of scour and pipeline
spanning, and its evolution over time (e.g. Pinna et al. 2003, Leckie et al. 2015) provide a useful insight
into this process. Figure 5 shows a pipeline section that was monitored over several years. Over a four year
period (Fig. 5a to Fig. 5b), the pipeline section has become lower relative to the seabed far from the pipeline
(far-field) while locally scour holes and sections of pipeline spanning can be observed. Fig. 5c shows a
different pipeline which is routed from deep to shallow water (left to right, and increasing KP on Fig. 5c),
transitioning from a fine-grained to a coarse seabed. Significant time-dependent changes in embedment are
evident in the coarse-grained sediments with a net increase in embedment. This has led to an increase in
hydrodynamically stability that could be quantified using conventional design approaches with the updated
pipeline embedment as an input. The changed embedment profile may also affect the response to thermal
expansion.
OTC-27623-MS 17

Figure 5—Field evidence of the changing burial of pipelines due to local scour and seabed
mobility (a) Pipe A, 2002, 6 months after laying, (b) Pipe A, 2006, 4 years after laying (Leckie
et al. 2015), (c) Pipe B, variation in embedment over several years (Leckie et al. 2016)

Changes in pipeline embedment along a pipeline and its local spatial variability can affect both the
hydrodynamic stability and also the response to thermal expansions. Borges-Rodriguez et al. (2013) and
Leckie et al. (2016, 2017) describe how increases in pipeline embedment due to scour and self-burial
can lead to enhanced pipeline stability, through hydrodynamic shielding and increased seabed resistance.
However, scour holes and spanning occur in areas of the pipeline where thermal expansions are managed
through controlled buckling; the two issues may interact. Scour holes may increase the likelihood of
unplanned buckles while local increases in pipeline embedment may contribute to a reduced tolerability of
planned buckles. Hence, for pipeline design in regions where the seabed is subject to scour and sediment
mobility, it can be necessary to consider the impact of these processes prior to start-up and during the
operational life of the pipeline. In considering these processes, a designer may utilise observations from
nearby pipelines if they exist in a similar metocean setting and the properties of the pipeline are similar.
Borges-Rodriguez et al. (2013) describe a process where observations of existing pipelines are used to
incorporate the effects of sediment mobility into the pipe-soil interaction parameters (‘friction factors’) used
in design. Figure 6 (Bransby et al. 2014) outlines an overall design flow where assessment of sediment
mobility and observations from existing pipelines (if available) are incorporated into the assessment of pipe-
soil interaction parameters.
18 OTC-27623-MS

Figure 6—Approach to PSI evaluation including sediment mobility effects (Bransby et al., 2014)

In some situations, physical modelling of fluid-pipe-seabed interaction in facilities such as the O-Tube
at The Univerisity of Western Australia (An et al., 2013) has been used to model sediment moblility and
pipe lowering to calibrate design calculations of pipeline stability in various different situations. Jas et al.
(2012) describes a project from Australia's North West Shelf where this form of physical modelling has been
used to assess the effects of sediment mobility on pipeline stability. Also, the STABLEpipe JIP (Griffiths
et al, 2010, Draper et al. 2014) has conducted similar testing to develop guidelines to quantify the effect of
sediment mobility on pipeline stability. These are being developed in partnership with DNV, with the aim
of evolving the current treatment of fluid-pipe-seabed interaction in the existing DNV code F109.
The above approaches are in various stage of development and are not yet encapsulated in the design
codes.

Damping from pipeline-seabed interaction in freespan analysis

Pipeline free spans that exceed a certain length require assessment for fatigue potential. The stiffness
and damping provided by the soil supports can have a significant influence on the fatigue rate. DNV
Recommended Practice F105 (DNV 2002) provides advice on the assessment of free span behavior.
Recommendations of soil damping ratio are provided based on soil classification (clay – for three
consistencies from very soft to hard; sand – loose, medium or dense) and free span length. Damping ratios
from 0.5 – 3% are specified, increasing with soil stiffness and reducing with span length. Unfortunately
the code does not indicate the origin of these values, and provides no references to the underlying reports
or publications.
In contrast to these recommended levels of damping, recent work by Templeton et al. (2016) has shown
very significant damping levels for a partially embedded pipe on the seafloor. Their results were obtained by
modeling a 12" (0.3 m) diameter pipe embedded 0.25D, varying the amplitudes and frequencies of motion
to determine both hysteretic and radiation damping ratios. The amplitude-dependent hysteretic damping
component was isolated by performing analyses where dynamic interial effects were not considered,
whereas the frequency dependent radiation damping was isolated by considering very small amplitude
displacements (± 0.002D). Additional analyses considered both types of damping and a combination rule
proposed to determine the sum of both components (Figure 7).
OTC-27623-MS 19

Figure 7—Damping ratios for 1-ft. diameter pipe resting on seafloor, from Templeton et al. (2015)

The results indicated very significant and greater levels of damping versus what is usually considered
in design, including the advice in DNV (2002). For example, for loading periods from 0.1 to 1 s (1 to 10
Hz) the damping ratio from radiation damping alone was about 20%. For very long loading periods (100
s) where radiation damping is negligible the damping ratio ranged from about 22 to 38% for displacements
± 0.006D to ± 0.04D. The highest damping ratio of about 47% was achieved at a frequency of 10 Hz and
a displacement of ± 0.04D
Preliminary sensitivity analyses have been performed to assess the potential impact of these high damping
ratios on the fatigue of a pipeline span. The pipe diameter considered was 18 in. (0.46 m) while the span
length was about 50 m. The results show significantly reduced displacements at the shoulders and mid-point
of the span for higher damping levels. In addition, about a 15 to 25% reduction in the stress range at the
mid-span was observed for damping ratios greater than 20%. This level of stress reduction would increase
fatigue life or allow longer acceptable span lengths.
The contrast between these recent observations, and the much lower damping ratios recommended
in DNV (2002) requires resolution, ideally via reference to the primary sources, be they model tests or
numerical analyses. The next revision of the API/ISO geotechnical codes offers an opportunity to reconcile
these differences and introduce new guidance that potentially offers significant design benefit through
increased damping and improve fatigue performance.
20 OTC-27623-MS

Subaqueous clay-rich density flow simulation and consequence analysis

General
This section is principally intended to provide a guide to the various approaches available for simulation
of subaqueous clay-rich density flows and evaluation of the forces resulting upon contact with submarine
pipelines. It is not intended to be a detailed manual or code setting out all those matters that ought to
be considered in each case. Rather, the purpose has been to enable engineers to understand the various
processes involved in subaqueous density flows and to summarize appropriate methods to simulate these
highly complex natural phenomena numerically and to quantify interaction forces with on-bottom pipelines.
Submarine density flows are important sediment-transporting mechanisms and play a major role in
the evolution of deep-sea fans and deltas. The triggering mechanisms, flow dynamics, interaction with
the surrounding water, depositional processes and consequences of failures have been studied by many
researchers. The complexity of subaqueous density processes flow from initiation to deposition, combined
with post-depositional consolidation and soft-sediment deformation, often make it difficult to interpret
the characteristics of the original flow from the sedimentary record, and therefore to create appropriate
models to address engineering problems (Zakeri et al. 2012). This has led to considerable confusion
of nomenclature in the literature. Mulder & Alexander (2001) provide a simple yet comprehensive and
encompassing classification of sedimentary density flows, based on physical flow properties and grain-
support mechanisms, and briefly discuss the likely characteristics of the deposited sediments.

Flow Simulation
Early attempts involved Work-Energy Theorem analysis (Heim 1932), where a landslide movement was
considered analogous to a block sliding on a curved path, characterized by constant frictional resistance at
the base of the block. This approach formed the basis for subaerial density flow simulation (Hungr 2005).
Since then, a number of analytical methods have been developed over the years that attempt to simulate
subaerial and subaqueous density flows. They range from simple few-parameter dynamic models in a two
dimensional (2D) space to more advanced approaches involving sophisticated continuum fluid and/or soil
mechanics, smooth particle hydrodynamics (SPH) and other discrete material methods etc. over irregular
three dimensional (3D) terrain. Some of the models presented for simulation of subaqueous debris flows
take root in those developed for subaerial density flows.
For clay-rich debris flow simulation, depth-averaged Eulerian schemes (continuum fluid dynamics
based), seem to provide a reasonable approximation and averaging of the processes involved. A number
of approaches have been developed and programmed (Norem et al. 1990, Mei & Liu 1987, Jiang et al.
1993, Huang & Garcia 1998, Imran et al. 2001). These approaches require characterization of debris flow
through rheological models. Various studies have demonstrated that the flow characteristics and behavior
of clay-rich debris can be described using constitutive rheological models developed for yield stress non-
Newtonian fluids (e.g. Locat & Demers 1988, Coussot & Piau 1994, Coussot & Boyer 1995, Coussot et
al. 1996, Coussot 1997, Locat 1997). Despite inherent limitations of the Eulerian scheme with regards to
modeling effects such as break-up and retrogressive failure, the approach can yield satisfactory results with
direct rheology measurements and engineering judgment for practical application (Niederoda et al. 2003a,
b, 2006, De Blasio et al. 2004, Elverhoi et al. 2005, Issler et al. 2005).

Impact on Pipelines
A number of techniques have been proposed to assess the forces arising from pipeline-density flow
interaction including situations at the onset of a landslide. The problem has been investigated from two
perspectives: historically a fluid dynamics approach and more recently a geotechnical approach that also
allows for fluid drag (Zakeri 2009, Randolph & White 2012). Both approaches model the rheology of
the flow material as a non-Newtonian material where the mobilized shear stress or strength is strain-rate
OTC-27623-MS 21

dependent, typically captured using a power law or Herschel-Bulkley model (Herschel & Bulkley 1926).
This may be expressed as

(7)

where τy (or su0) is the yield stress or strength at very low or zero strain rate , K is a (dimensional)
‘consistency’ parameter, η is a dimensionless viscosity parameter, while n and β express the strength
dependency on strain rate. The alternative forms reflect the different fluid mechanics or geotechnical
nomenclature but in other respects achieve the same objective of a rate dependent mobilized ‘strength’ of
the flow material. In principle, the yield strength may be a function of cumulative shear strain (to model
strain softening), for example to simulate the transition from landslide initiation to density flow runout, but
for pipeline impact a fully remoulded value is often appropriate.
In the fluid dynamics approach, the interaction forces are quantified in terms of drag coefficients that
are a function of a non-Newtonian Reynolds number (RenN = ρv2/τ, where ρ is the material density, v the
velocity and τ the mobilized shear stress). Relationships proposed by Zakeri (2009b) for the normal (Fθ=90)
and longitudinal (frictional, Fθ=0) forces, for flow impacting a pipeline at an angle θ to the pipe axis, are:

where

(8)

The constant terms in the drag coefficients represent asymptotic values at high Reynolds numbers, while
the second term leads to rather high values of resulting force at low velocities (tending to infinity as v
reduces to zero). The CD-90 parameters in Eq. 8 are for a pipe fully immersed in the flow. Flume experiments
showed that the drag forces can be 20 to 30% less if the pipe remains on the seafloor with flow passing only
over the top of the pipe (Zakeri et al., 2008a).
The geotechnical approach attempts to provide sensible estimation of forces at low velocities, while
retaining the asymptotic behavior at high flow velocities. Randolph & White (2012) proposed expressions
for the normal and frictional force components of

(9)

where CD, Np-90 and fa,0 are nominally fixed values (independent of Reynolds number) although may be
affected by the thickness of the density flow relative to the pipe diameter (i.e. whether the pipe is fully
immersed). For general impact angles, a yield envelope approach to evaluate the normal and lateral force
coefficients was adopted, represented by

(10)

with exponents of p = 1, q = 3 and r = 0.7 derived to match the numerical data from Zakeri (2009b). For a
pipeline fully engulfed in a thick density flow, values of CD = 0.6/sinθ (taking v as the normal component,
vsinθ, in Eq. (9)), Np-90 = 11.9 and fa,0 =1.4. Both the latter values are theoretical values for a rough pipe.
In both approaches, key parameters, in particular the drag coefficients, derived either experimentally or
from numerical studies, show some discrepancies which need resolution. The expression for CD-90 in the fluid
22 OTC-27623-MS

mechanics approach (Eq. (8)) is a compromise between experimental and numerical data. Similarly, best-
fit experimental values of Np-90 and CD reported by Sahdi et al. (2014), for conditions where a fast moving
pipe segment was buried 2.5D below the surface of weak soil, were 7.35 and 1.06, with the former value
consistent with theoretical solutions for relatively shallow embedment (Martin & White 2012). A recent
numerical study by Liu et al. (2015), extending the work of Zakeri (2009b), confirmed the overall approach
and values of CD, Np-90 and fa,0 proposed by Randolph & White (2012), but recommended exponents in Eq.
(10) of p = 3.4, q = 2.4 and r = 0.4, which lead to ~20% higher impact force components under inclined
attack.
An important principle in the above approaches is that the flow velocity v should be interpreted as a
relative velocity between the moving density flow (or soil) and the pipeline. This is important for conditions
where a density flow might engulf a pipeline and carry it forward until such point where equilibrium is
achieved between the forces exerted on the pipeline by the decelerating density flow and the structural forces
induced in the deformed pipeline. In principle, the geotechnical approach above may be applied to pipelines
buried within a slow-moving landslide in addition to submarine density flows. In both cases appropriate
values of operational shear strength need to be assessed, taking account of the strain rate dependency of
soil stiffness and shear strength (e.g. Biscontin & Pestana 2001, Dayal & Allen 1975, Diaz-Rodriguez et
al. 2009, Lunne & Anderson 2007, Boukpeti et al. 2012, Zakeri & Hawlader 2013). It is recommended that
suitable strain rate formulations and associated parameters needed to estimate su,op are determined through
laboratory testing, covering an appropriate range of shear strain rate.
The structural response of a pipeline to impact by a density flow or buried within a landslide needs to
consider the lateral and axial passive resistance either side of the the slide zone, taking account of vertical
droop or heave of the pipeline. Simple parametric solutions (e.g. Randolph et al. 2010) are useful for
estimating potential failure of the pipeline, which typically occurs by full-bore tensile failure unless the slide
zone is relatively narrow. This aspect is confirmed by pipeline failures in slow-moving landslides (Sweeney
et al. 2004). Such solutions may also be integrated into depth-averaged analysis of run out behavior to allow
estimation of distances over which the pipeline may be carried by the density flow and of the interaction
forces and structural tensile and bending stresses in the pipeline at equilibrium conditions (White et al. 2016,
Boylan & White 2017). This integration of the slide and pipeline responses leads to reduced estimates of
the pipeline deformation and induced strains for conditions where the debris flow reduces velocity within
the distance over which the pipe is displaced or conditions where the velocity at the position of the pipeline
is less than the maximum velocity of the flow (Boylan & White 2017).
Another consideration is potential for vertical cyclic stresses arising from vortex shedding during an
impact on suspended pipelines (spanning pipelines) or river crossings where a segment of a pipe is exposed
to torrents (sediment laden density flows). Zakeri et al. (2008a) observed vortex shedding occurring behind
the pipe in their flume experiments for RenN> 20. Measurements were complemented with numerical
simulations and preliminary relationship between Strouhal number and RenN has been established (Zakeri
et al., 2009). Further investigation on this subject is encouraged.
An area requiring investigation is development of mitigative and control measures to increase
survivability of a pipeline against a density flow impact event. Measures may be case-specific; however,
there is a need for development of fit-for-purpose innovative solutions. Preliminary works involve
investigating protective berms for laid-on-seafloor pipelines in shallow waters or fjord crossing and
introduction of a concept for deepwater suspended pipelines (Zakeri et al., 2008b). Solutions can include
introduction of slack or use of buoyancy modules. Further research in this area is encouraged as there is
significant room to advance the state of knowledge and introduce innovative practices.
Glide blocks and out-runner blocks are an intact piece of seabed material that still carry the strength
properties of the failed soil mass. Therefore, the rate corrected geotechnical approach is also suitable for
such cases. Zakeri et al. (2012) conducted a series of physical experiments in a geotechnical centrifuge to
OTC-27623-MS 23

quantify the drag force on a submarine pipeline caused by a glide block or an out-runner block impact normal
to the pipe axis (0.04 < velocity < 1.3 m/s; 4 < su < 8 kPa). Zakeri and Hawlader (2013) later complemented
the centrifuge test results using CFD and provided a geotechnical-based approach for glide block and out-
runner block impact on pipelines. It should be noted that for impact situations where RenN is near unity (i.e.
slow impact or creeping flow), both the Zakeri and Hawlader (2013) geotechnical approach and Zakeri et
al. (2009) approach yield similar results.

Conclusions
Recent advances in the understanding of pipe-soil interaction are maturing and a consensus on best
practices for on-bottom pipeline design is emerging. Design guidelines are needed for the modern generation
of deep-water pipelines, in particular the problems posed by high pressure, high temperature (HPHT)
pipelines. There is also the demand for efficiencies in the design of primary and secondary stabilization
for hydrodynamic stability. This paper has provided a synthesis of recent advances and current industry
best practice, and has set out the state-of-knowledge for future revisions to the API/ISO codes. The aim
has been to unlock consistent and efficient approaches for future pipeline systems, and for management of
existing systems.

Acknowledgements
The authors acknowledge the assistance of Charles de Brier (Fugro GeoConsulting Belgium) in supplying
data for Figure 2.
The lead author is supported via the Shell EMI Chair in Offshore Engineering at the University of Western
Australia (UWA), and the third author is supported via the Fugro Chair in Geotechnics at UWA.

References
An, H., Luo, C., Cheng, L. & White, D.J. (2013). A new facility for studying ocean-structure–seabed interactions: the
O-tube. Coast. Eng. 82, 88–101.
API (2014). Geotechnical and Foundation Design Considerations. Recommended Practice RP2 GEO. American
Petroleum Institute.
Atkins (2015). SAFEBUCK JIP - Safe Design of Pipelines with Lateral Buckling: Design Guideline, Report No. 5087471 /
01 / C, dated 15 January 2015.
Atkins (2014). SAFEBUCK JIP - Feedback from Operating Pipelines. Report No: 5087471.002.RP01.02. 10 Nov. 2014
Aubeny, C.P., Shi, H. and Murff, J.D. (2005). Collapse loads for a cylinder embedded in trench in cohesive soil,
International Journal of Geomechanics, ASCE, 5(4):320–325.
Ballard, J.-C., de Brier, C., Stassen, K. & Jewell, R. (2013a). Observations of pipe-soil response from in-situ
measurements. Offshore Technology Conference, Houston, USA. Paper OTC 24154.
Ballard, J.-C., van den Abeele, F. & de Brier, C., (2013b). Influence of axial response on pipeline end expansion and
walking. 6th International Pipeline Technology Conference, Ostend, Belgium. Paper S30-01.
Biscontin, G. and Pestana, J. M. (2000). Influence of Peripheral Velocity on Undrained Shear Strength and Deformability
Characteristics of a Bentonite-Kaolinite Mixture. Department of Civil and Environmental Engineering, University of
California, Berkeley
Borges Rodriguez A., Bransby M.F., Finnie I.M., Low H.E. & White D.J. (2013). Changes in pipeline embedment
due to sediment mobility: observations and implications for design. Proc. Int. Conf. on Offshore Mech. and Arctic
Engineering. Paper OMAE-2013-11425
Boukpeti N. & White D.J. (2017). Interface shear box tests for assessing axial pipe-soil resistance. Géotechnique
67(1):1–13
Boukpeti, N., White, D. J. & Randolph, M. (2012). Strength of fine-grained soils at the solid–fluid transition.
Geotechnique, 62(3):213 -226.
Boylan, N.P., White, D.J. & Brunning, P. (2014). Seabed friction on carbonate soils: Physical modelling of axial pipe-soil
interaction. Proc. Offshore Technology Conference. Paper OTC25287
Boylan, N.P. & White, D.J. (2017). Depth-averaged numerical modelling of submarine slide run-out in softening soil.
In review.
24 OTC-27623-MS

Bransby, M.F., Borges Rodriguez, A., Zhou, H., Low, H.E. & White, D.J. (2014). Sediment mobility effects on seabed
resistance for unburied pipelines. Proc. Offshore Technology Conference. Paper OTC25287
Bransby M.F., Ramm, M., Zhou, H., Low, H-E., White, D.J. (2017). Pipe-Soil Structure Interaction: Best practices for
spools, buckle initiators, end connections and sand waves. Proc. Offshore Technology Conference, Houston. Paper
OTC-27689
Bransby M.F., White D.J., Low H-E. & Westgate Z. (2015). Accurate descriptions of pipe-soil interaction to facilitate the
design of high temperature pipelines. Proc. Offshore Pipeline Technology Conference Asia.
Bruton, D.A.S, White, D.J., Cheuk, C.Y., Bolton, M.D. & Carr, M.C. (2006). Pipe-soil interaction behaviour during lateral
buckling, including large amplitude cyclic displacement tests by the SAFEBUCK JIP, Proc. Offshore Technology
Conf., paper OTC 17944, Houston, Texas, May 2006
Bruton, D., Carr, M. & White, D.J. (2007). The influence of pipe-soil interaction on lateral buckling and walking of
pipelines: The SAFEBUCK JIP, Proc. Intl. Conf. on Offshore Site Investigation and Geotechnics, SUT OSIG, London,
U.K., Sepy. 2007, 133–150
Chatterjee, S., Randolph, M.F. and White, D.J. (2012a). The effects of penetration rate and strain softening on the vertical
penetration resistance of seabed pipelines. Géotechnique, 62(7):573–582.
Chatterjee S., White D.J. & Randolph M.F. (2012b). Numerical simulations of pipe-soil interaction during large lateral
movements on clay. Géotechnique, 62(8):693–705
Coussot, P. (1997). Mudflow Rheology and Dynamics, Rotterdam, Balkema.
Coussot, P. & Boyer, S. (1995) Determination of yield stress fluid behaviour from inclined plane test. Rheologica Acta,
34(6):534–543.
Coussot, P. & Piau, J. M. (1994) On the behaviour of fine mud suspensions. Rheologica Acta, 33(3):175–184.
Coussot, P., Proust, S. & Ancey, C. (1996). Rheological interpretation of deposits of yield stress fluids. J. Non-Newtonian
Fluid Mechanics, 66(1):55–70.
Dayal, U. & Allen, J. H. (1975). The effect of penetration rate on the strength of remolded clay and sand samples. Canadian
Geotechnical Journal, 12(3):336–348.
De Blasio, R., Elverhoi, A., Issler, D., Harbitz, C. B., Bryn, P., and Lien, R. (2004) Flow models of natural debris flows
originating from overconsolidated clay materials. Marine Geology 213(1-4):439–455.
Diaz-Rodriguez, J. A., Martinez-Vasquez, J. J. and Santamarina, J. C. (2009) Strain-rate effects in Mexico City soil.
Journal of Geotechnical and Geoenvironmental Engineering. 135(2):300–305.
Dimakis, P., Elverhoi, A., Hoeg, K., Solheim, A., Harbitz, C., Laberg, J. S., Vorren, T. O., and Marr, J. (2000) Submarine
slope stability on high-latitude glaciated Svalbard-Barents Sea margin. Marine Geology, 162(2-4): 303–316.
Dingle, H.R.C, White, D.J. & Gaudin, C., (2008). Mechanisms of pipe embedment and lateral breakout on soft clay. Can.
Geotech. J. 45(5): 636–652.
DNV (2015) Merger of SAFEBUCK and DNV RP F110: Merged Guideline. Dated 24 August 2015.
DNV (2017) Pipe-soil interaction for submarine pipelines. Recommended Practice F-114 (currently undergoing review
hearing). DNVGL
Elverhoi, A., Issler, D., De Blasio, F. V., Ilstad, T., Harbitz, C. B., and Gauer, P. (2005) Emerging insights into the dynamics
of submarine debris flows. Natural Hazards and Earth System Sciences, 5(5): 633–648.
Frankenmolen S., Peek R., Carr M., MacRae I., White D.J., Rimmer J. & Ang S-Y. (2017) Pipe-Clamping Mattress (PCM)
to Stop Flowline Walking. Proc. Offshore Technology Conference, Houston, Paper OTC27815.
Gaudin C. & White D.J. (2012). Centrifuge modelling to support the design of subsea pipelines. Proc. GeoCongress 2012,
San Francisco, ASCE STP225 2402-2411.
Heim, A. (1932) Bergsturz und Menschenleben, Zurich, Fretz and Wasmuth Verlag,.
Herschel, W. H. and Bulkley, R. (1926). Konsistenzmessungen von Gummi-Benzollo¨sungen. Kolloid Z. 39, 291–300.
Hill, A. J. & Jacob, H. (2008). In-situ measurement of pipe–soil interaction in deep water. Proc. Offshore Technology
Conference, Houston, Texas, USA, Paper OTC 19528.
Hill A., White D.J., Bruton D.A.S., Langford T., Meyer V., Jewell R. & Ballard J-C. (2012). A new framework for axial
pipe-soil interaction illustrated by a range of marine clay datasets. Proc. Int. Conf. on Offshore Site Investigation and
Geotechnics. SUT, London. 367–377
Hill A. & White D.J. (2015). Pipe-soil interaction: Recent and future improvements in practice. Proc. Int. Symp. on
Frontiers in Offshore Geotechnics (ISFOG).387–392
Huang, X. and Garcia, M. H., (1998). A Herschel Bulkley Model for Mud Flow Down a Slope. Journal of Fluid Mechanics,
374(1): 305–333
Hungr, O., (2005) Landslide Risk Management: Proceedings of the International Conference on Landslide Risk
Management, Vancouver, Canada, 31 May – 3 June 2005, London, Taylor & Francis,.
Imran, J., Harff, P. and Parker, G. (2001) A numerical model of submarine debris flow with Graphical User Interface.
Computers & Geosciences. 27(6): 717–729.
OTC-27623-MS 25

ISO (2016) Petroleum and natural gas industries - Specific requirements for offshore structures — Part 4: Geotechnical
and foundation design considerations. ISO 19901-4. International Standards Organisation.
Jas, E., O'Brien, D., Fricke, D., Gillen, A., Cheng, L., White, D. & Palmer, A. (2012). Pipeline stability revisited. Journal
of Pipeline Engineering, 4th Quarter, pp. 259–268.
Jewell, R.J., and Ballard, J.-C. (2011). Axial pipe-soil interaction – a suggested framework. Proc. Offshore Technology
Conference, Houston, Paper OTC 22010.
Jiang, L. and Leblond, P. H. (1993) Numerical modeling of an underwater Bingham plastic mudslide and the waves which
it generates. Journal of Geophysical Research-Oceans. 98(C6): 10303–10317.
Konuk, I. & Yu, S. (2007). Continuum FE Modeling of Lateral Buckling: Study of Soil Effects. Proceedings of
OMAE07, 26th International Conference on Offshore Mechanics and Arctic Engineering, San Diego, California. Paper
OMAE2006-29376
Langford, T.E., Dyvik, R. and Cleave, R. (2007) Offshore pipeline and riser geotechnical model testing: Practice and
interpretation. Proc. Intl. Conf. on Offshore Mechanics and Arctic Engineering, paper OMAE2007-29458, San Diego,
California.
Leckie, S., Draper, S., White, D.J., Cheng, L. & Fogliani N. (2015). Lifelong embedment and spanning of a pipeline on
a mobile seabed. Coastal Engineering. 95:130–146.
Leckie S., Mohr H., Draper S., McLean D., White D.J. & Cheng L. (2016). Sedimentation-induced burial of subsea
pipelines: observations from field data and laboratory experiments. Coastal Engineering 114:137–158
Leckie S., Draper S., White D.J., Cheng L., Griffiths T. & Fogliani A. (2017). Observed changes to the stability of a seabed
pipeline caused by seabed mobility. Coastal Engineering In press.
Liu, J., Tian, J. and Yi, P. (2015) Impact forces of submarine landslides on offshore pipelines. Ocean Engineering
95:116–127.
Locat, J., (1997) Normalized rheological behaviour of fine muds and their flow properties in a pseudoplastic regime.
Debris-Flow Hazards Mitigation: Mechanics, Predication and Assessment, Water Resources Engineering Division,
American Soc. Civil Engineers (ASCE).
Locat, J. and Demers, D. (1988) Viscosity, yield stress, remolded strength, and Liquidity Index relationships for sensitive
clays. Canadian Geotechnical Journal, 25(4):799–806.
Lunne, T. and Andersen, K. H. (2007) Soft clay shear strength parameters for deepwater geotechnical design. Proc. 6th Int.
Offshore Site Investigation and Geotechnics Conf.: Confronting New Challenges and Sharing Knowledge, London,
UK.
Martin, C.M., Kong, D. & Byrne, B.W. (2013). 3D analysis of transverse pipe–soil interaction using 2D soil slices.
Géotechnique Letters 3 (3):119–123
Mei, C. C. and Liu, K. F. (1987) A Bingham-Plastic model for a muddy seabed under long waves. Journal of Geophysical
Research-Oceans, 92(C13): 14581–14594.
Merifield, R.S., White, D.J. and Randolph, M.F. (2008). Analysis of the undrained breakout resistance of partially
embedded pipelines, Géotechnique, 58(6): 461–470
Meyer, V., Langford T., and White, D.J. (2016) Physical modelling of pipe embedment and equalization in clay.
Géotechnique, 66 (7), p. 602–609
Muir Wood, D.M. (1990) Soil behaviour and critical state soil mechanics. Cambridge Press.
Mulder, T. and Alexander, J. (2001) The physical character of subaqueous sedimentary density flows and their deposits.
Sedimentology, 48(2):269–299.
Mulder, T., Savoye, B. and Syvitski, J.P.M. (1997). Numerical modeling of a mid-sized gravity flow: The 1979 Nice
turbidity current. (dynamics, processes, sediment budget and seafloor impact). Sedimentology. 44:305–326.
Murff, J.D., Wagner, M.A. and Randolph, M.F. (1989). Pipe penetration in cohesive soil. Geotechnique 39(2): 213–229.
Najjar, SS., Gilbert, R.B., Liedtke, E.A., McCarron, B. & Young, A.G. (2007). Residual shear strength for interfaces
between pipelines and clays at low effective normal stresses. ASCE Journal of Geotechnical and Geoenvironmental
Engineering. 133(6): 695–706.
Niedoroda, A.W., Reed, C., Das, H., Hatchett, L., and Perlet, A.B. (2006) Controls of the behavior of marine debris flows.
Norwegian Journal of Geology. 86: 265–274.
Niedoroda, A.W., Reed, C.W., Hatchett, L., Jeanjean, P., Driver, D., Briaud, J.-L., and Bryant, W. (2003) Bottom currents,
deep sea furrows, erosion rates, and dating slope failure - Induced debris flows along the Sigsbee Escarpment in the
deep Gulf of Mexico. Offshore Technology Conference (OTC), Houston, USA.
Niedoroda, A.W., Reed, C.W., Hatchett, L., Young, A., Lanier, D., Kasch, V., Jeanjean, P., Orange, D., and Bryant, W.,
(2003). Analysis of past and future debris flows and turbidity currents generated by slope failures along the Sigsbee
Escarpment in the deep Gulf of Mexico Offshore Technology Conference, Houston, Paper OTC 15162.
26 OTC-27623-MS

Niedoroda, A. W., Reed, C. J., Badalini, M., Kruse, G., Mullee, G., Parker, G., and Forristall, G. (2000). Developing
engineering design criteria for deep water turbidity currents. 19th International OMAE Conference, Energy for the
New Millennium, New Orleans, Louisiana, USA
Norem, H., Locat, J. and Schieldrop, B. (1990) An approach to the physics and the modeling of submarine flowslides.
Marine Geotechnology, 9(2): 93–111.
Palmer, A. (1996). A flaw in the conventional approach to stability design of pipelines, Offshore Pipeline Tech.
Conference, Amsterdam.
Pinna, R., Weatherald, A., Grulich, J., Ronalds, B.F. (2003). Field observations and modelling of the self-burial of a North
West Shelf pipeline. Proc. International Conference on Offshore Mechanics and Arctic Engineering ASME 591–599.
Randolph, M.F., Seo, D. and White, D.J. (2010). Parametric solutions for slide impact on pipelines. Journal of
Geotechnical and Geoenvironmental Engineering, ASCE, 136(7), 940–949.
Randolph, M.F. and White, D.J. (2008a). Pipeline embedment in deep water: processes and quantitative assessment. Proc.
Offshore Technology Conf., Houston, Paper OTC 19128.
Randolph, M.F. and White, D.J. (2008b). Upper bound yield envelopes for pipelines at shallow embedment in clay,
Géotechnique. 58(4) 297–301
Randolph, M. F. and White, D. J. (2012) Interaction forces between pipelines and submarine slides — a geotechnical
viewpoint. Ocean Engineering, 48:32–37.
Randolph, M.F., White, D.J. & Yan, Y. (2012). Modelling the axial soil resistance on deepwater pipelines. Géotechnique,
62(9):837–846
SAFEBUCK (2015) Safe design of pipelines with lateral buckling. Design Guideline. Document 5087471/C 15 January
2015. Atkins.
Sahdi, F., Gaudin, C., White, D. J., Boylan, N., and Randolph, M. F., (2014) Centrifuge modelling of active slide–pipeline
loading in soft clay. Geotechnique, 64(1):16–27.
Smith, V.B. & White D.J. (2014). Volumetric hardening in pipe-soil interaction. Proc. Offshore Technology Conference,
Asia. Kuala Lumpur. Paper OTC-24856-MS
Srikandarajah, T., Roberts, G., Tanscheit, P., Solano, R.F., Hansen, A. & Antunes, A.R. (2015) Forward predictions for life-
of-field assessments based on the observed lateral buckling behavior of operating deepwater pipelines. Proc. Offshore
Technology Conference, Houston. Paper OTC25961
Sumer, B.M., Fredsøe, J. (2002). The Mechanics of Scour in the Marine Environment. World Scientific Publishing,
Singapore.
Sweeney, M., Gasca, A., Garcia Lopez, M. & Palmer, A. (2004). Pipelines and landslides in rugged terrain: a database,
historic risks and pipeline vulnerability. Proc. Int. Conf. on Terrain and Geohazard Challenges Facing Onshore Oil
and Gas Pipelines, Thomas Telford, London, 641–659.
Templeton, J.S.III, Barrett, M.J., and Clukey, E.C. (2015). Finite element analysis of soil damping for pipeline spanning.
Proc. 3rd International Synposium for Frontiers in Offshore Geotechnics (ISFOG), Oslo
Tian, Y. & Cassidy M.J. (2008) Modeling of pipe–soil interaction and its application in numerical simulation. International
Journal of Geomechanics 8(4):213–229
Toniolo, H., Lamb, M. and Parker, G. (2006) Depositional turbidity currents in diapiric minibasins on the continental
slope: formulation and theory. Journal of Sedimentary Research, 76:783–797.
Verley, R. and Sotberg, T. (1992). A soil resistance model for pipelines placed on sandy soils. Proc. Intl. Conf. on Offshore
Mechanics and Arctic Engineering, OMAE, Calgary, 5-A:123–131
Verley, R. and Lund, K.M., (1994). A soil resistance model for pipelines placed on clay soils. Proc. Intl. Conf. on Offshore
Mechanics and Arctic Engineering, OMAE, Copenhagen, 5:225–232
Westgate, Z.J., Randolph, M.F., White D.J. and Li, S. (2010). The influence of seastate on as-laid pipeline embedment:
a case study. Applied Ocean Research. 32(3):321–331.
Westgate, Z.J., White, D.J. and Randolph, M.F. (2012). Field observations of as-laid pipeline embedment in carbonate
sediments. Géotechnique 62 (9), 787–798.
Westgate, Z.J., White, D.J. and Randolph, M.F. (2013). Modelling the embedment process during offshore pipe laying on
fine-grained soils. Canadian Geotechnical Journal, 1, 15–27.
Westgate Z.J. & White D.J. (2015). Quantifying spatial variability in pipeline embedment for subsea pipeline design. Proc.
Int. Conf. Offshore Mech. and Arctic Engng. OMAE2015-42292
Westgate Z.W., Haneberg, W. & White D.J. (2016). Modelling the spatial variability in as-laid embedment for HPHT
pipeline design. Canadian Geotechnical Journal. 53(11): 1853–1865.
White, D.J. and Gaudin, C. (2008). Simulation of seabed pipe-soil interaction using geotechnical centrifuge modelling,
Proc. 1st Asia-Pacific Deep Offshore Technology Conf., DOT, Perth, Australia, December 2008
White D.J. & Bransby M.F. (2017). Pipeline-seabed interaction. Chapter in Encyclopedia of Marine & Offshore
Engineering, Wiley.
OTC-27623-MS 27

White D.J., Campbell M.E., Boylan N.P. & Bransby M.F. (2012). A new framework for axial pipe-soil interaction
illustrated by shear box tests on carbonate soils. Proc. Int. Conf. on Offshore Site Investigation and Geotechnics. SUT,
London. 379–387
White D.J., Bolton M.D., Ganesan S.A., Bruton D., Ballard J-C., Langford T. (2011). SAFEBUCK JIP: Observations
from model testing of axial pipe-soil interaction on soft natural clays. Proc. Offshore Technology Conference, Houston
Paper OTC 21249.
White D.J., Hill A.J., Westgate Z. & Ballard J-C (2010). Observations of pipe-soil response from the first deepwater
deployment of the SMARTPIPE. Proc. 2nd Int. Symp. on Frontiers in Offshore Geotechnics. Perth. 851–856
White, D.J., Westgate, Z., Ballard, J.C., de Brier, C., & Bransby, F. (2015). Best practice geotechnical characterization
and pipe-soil interaction analysis for HPHT pipeline design. Proc. Offshore Technology Conference, Houston. Paper
OTC 26026.
White D.J., Randolph M.F., Gaudin C., Boylan N.P., Wang D., Boukpeti N., Zhu H. & Sahdi F. (2016). The impact
of submarine slides on pipelines: Outcomes from the COFS-MERIWA JIP. Proc. Offshore Technology Conference,
Houston. Paper OTC27034
White D.J. & Randolph M.F. (2007). Seabed characterisation and models for pipeline-soil interaction. Proc. 17th Int.
Offshore & Polar Engng. Conference, Lisbon, Portugal. CD 11pp
White D.J. (2014). Axial pipe-soil interaction framework. Report to Atkins and the SAFEBUCK Joint Industry Project
Phase III/GEO. UWA Geo Report 13684v2 176pp.
Yan Y., White D.J. & Randolph M.F. (2010). Investigations into novel shallow penetrometers for fine-grained soils. Proc.
2nd Int. Symp. on Frontiers in Offshore Geotechnics. Perth. 321–326
Yan, Y., White D.J. & Randolph M.F. (2014). Cyclic consolidation and axial friction on seabed pipelines. Géotechnique
Letters 4:165–169
Yan Y., White D.J. & Randolph M.F. (2016). Elastoplastic consolidation solutions for scaling from shallow penetrometers
to pipelines. Canadian Geotechnical Journal, in press.
Zakeri, A., Høeg, K. and Nadim, F., (2008a). Submarine debris flow impact on pipelines - Part I: Experimental
investigation. Coastal Engineering. 55:1209–1218.
Zakeri, A., Hoeg, K. and Nadim, F. (2008b). Submarine debris flow impact on pipelines: Drag forces, mitigation and
control. Offshore Technology Conference (OTC). Houston, Texas, Paper OTC19173.
Zakeri, A., (2009a). Review of the State-of-the-Art: Drag forces on submarine pipelines and piles caused by landslide
or debris flow impact. J. Offshore Mechanics and Arctic Engineering, American Society of Mechanical Engineers
(ASME), 131 DOI: 10.1115/1111.2957922.
Zakeri, A. (2009b). Submarine debris flow impact on suspended (free-span) pipelines: Normal and longitudinal drag
forces. Ocean Engineering, 36:489–499.
Zakeri, A., (2009c). Corrigendum to "Submarine debris flow impact on pipelines – Part II: Numerical Analysis". Coastal
Engineering, 56(1):1.
Zakeri, A., Høeg, K. and Nadim, F. (2009). Submarine debris flow impact on pipelines - Part II: Numerical Analysis.
Coastal Engineering, 56:1–10.
Zakeri, A., Si, G., Marr, J. D. G., and Hoeg, K. (2009). Experimental investigation of subaqueous clay-rich flows, generated
turbidity and sediment deposition. Submarine Mass Movements and Their Consequences, Springer Science + Business
Media B.V. 95–106.
Zakeri, A., Hawlader, B. and Chi, K. (2012) Drag forces caused by submarine glide block or out-runner block impact on
suspended (free-span) pipelines. Ocean Engineering. 47:50–57.
Zakeri, A. and Hawlader, B. (2013) Drag forces caused by submarine glide block or out-runner block impact on suspended
(free-span) pipelines— numerical analysis. Ocean Engineering. 67:89–99.
Zhang, J., Stewart, D.P. & Randolph, M.F. (2002). Modeling of shallowly embedded offshore pipelines in calcareous sand.
J. Geotechnical and Geoenvironmental Engineering, ASCE, 128(5): 363–371