Double Pipe Wall

Cold Regions Science and Technology 69 (2011) 3–20
Contents lists available at ScienceDirect
Cold Regions Science and Technology

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c o l d r e g i o n s
Offshore pipeline protection against seabed gouging by ice: An overview

Paul Barrette ⁎
Canadian Hydraulics Centre, National Research Council of Canada, Ottawa, ON, Canada K1A0R6
a r t i c l e i n f o a b s t r a c t
Article history: Offshore operators in the Arctic will rely on seafloor installations, notably pipelines, to manage and transport
Received 4 January 2011 hydrocarbons. In icy waters, these structures are at risk of being damaged by gouging ice features, either
Accepted 12 June 2011 icebergs or sea ice ridges. This phenomenon generally occurs when an ice feature drifts into shallower areas
and its keel starts plowing the seabed over considerable distances. It is generally agreed that adequate
Keywords:
protection against these events can be achieved by burying the pipeline below the seafloor. The question is:
Pipelines
Iceberg
what constitutes a safe and economical burial depth for any given location? An answer to this question
Ice ridge requires adequate knowledge of material properties (soil, ice keel and pipeline), a reliable handle on the
Seafloor processes taking place during gouging and a consensus on what constitutes acceptable risks. Research on this
Scour subject has been on-going for the past few decades, along several fronts. One is by means of field studies,
Arctic including replicating gouging scenarios in a natural environment, in situ ice characterization, seabed mapping
and on-land relict gouge investigations. Another is through laboratory studies, either at single gravity or in a
centrifuge. Theoretical analyses and numerical simulations have also contributed to our current
understanding of gouging phenomena. Several research groups proposed some form of guidelines for
estimating gouging parameters—examples are presented. These methodologies are instructive in that they
represent an integrated approach to an improved understanding of gouging phenomena. They point the way
to what one may expect in terms of future guidelines to a safe and cost-effective burial depth.
Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1. Objectives and target readership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2. The three main actors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1. Subsea pipelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1.1. General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1.2. Design criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1.3. Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1.4. Hazards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1.5. Additional risks in cold waters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1.6. Examples of subsea pipelines in cold waters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2. The gouging ice features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2.1. Icebergs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2.2. Sea ice and sea ice ridges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3. The seabed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3.1. About cohesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3.2. Cohesionless seabed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3.3. Cohesive seabed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3. Gouging and its bearing on offshore engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.1. The ice gouging process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.1.1. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.1.2. As the keel meets the seabed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.1.3. Ice dynamics during gouging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.1.4. Seafloor response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
⁎ Tel.: + 1 613 993 6809.

E-mail address: pbarrette@fastmail.fm.
0165-232X/$ – see front matter. Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.coldregions.2011.06.007
4 P. Barrette / Cold Regions Science and Technology 69 (2011) 3–20
3.1.5. Some numbers on gouges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3.2. Damage to pipelines by drifting ice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.3. Design standards, rules and codes of practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4. Pipeline protection against ice gouging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.1. Ice management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.2. Shielding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.3. Trenching and burial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.4. Other options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.5. Bottom line on pipeline protection: burial depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5. Investigations on gouging dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5.1. Field studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.1.1. Gouging scenarios in a natural environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.1.2. In situ characterization of real ice features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.1.3. Seabed mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.1.4. On-land gouge relics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.2. Laboratory studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.2.1. Simulations at single gravity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.2.2. Simulations at high simulated gravity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.3. Theoretical analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.4. Numerical simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
6. Toward a design methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.1. Limit states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.2. Reliability levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.3. Methodologies for estimating gouging parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.3.1. Kioka et al. (2004) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
6.3.2. Croasdale et al. (2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
6.3.3. Younan et al. (2007) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
6.3.4. Nobahar et al. (2007a, 2007b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
6.3.5. Been et al. (2008) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
6.3.6. King et al. (2009) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
7. Summary and conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
7.1. Current status on burial depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
7.2. Filling in the knowledge gaps—pros and cons of various approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
7.3. What lies ahead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Appendix A. Standards and codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1. Introduction against ice gouging—the most cost-effective approach is to bury

them to a given depth below the seafloor. The question is: how
According to the USGS (Gautier et al., 2009), the area north of the deep, keeping in mind that the deeper, the more expensive.
Arctic Circle may enclose up to 13% and 30% of the world's undiscovered
oil and gas, respectively. Most of these resources would be in continental
shelves beneath less than 500 m of water, which make up one third of 1.1. Objectives and target readership
that area. These numbers are, of necessity, speculative but they do point
out to a vast potential, hailed as a strong incentive for future Arctic A decision on burial depth relies on a proper understanding of
offshore activity (e.g.Mørk, 2007; Randell et al., 2008). More than 400 oil ice gouging processes. These are complex, as they involve the
and gas fields were known to exist already in the Arctic by 2007, mostly interplay between ice dynamics as well as seabed and pipeline
in Northern Russia and on the North Slope of Alaska (Gautier et al., responses, along with hydro-meteorological factors. A very large
2009). Fig. 1 provides an overall perspective on world-wide distribution amount of research has addressed these issues over the last few
of offshore production sites. decades, and progress has been steady. The main purpose of this
The Arctic poses significant challenges to the design, construc- article is to provide a general, up-to-date perspective of ice gouging
tion and operation of offshore structures, primarily due to a very as it relates to offshore pipelines, with an emphasis on the North
severe winter climate, the presence of ice and a sensitive American Arctic and Canadian East Coast. Appropriate referencing
ecosystem. Of critical concern are the drifting ice features that will direct the reader to additional information. For further insight
gouge the seafloor, which represent a threat to submarine facilities, into seabed gouging by ice and other topics dealing with offshore
notably pipelines. The latter are currently seen as the most efficient engineering in the Arctic, the reader is referred to Løset et al.(2006),
method for liquid and gas delivery. Because they extend to Kenny et al. (2007), Mørk (2007), Abdalla et al. (2008), Randell et al.
considerable distances, they are vulnerable to being damaged by (2008), Vershinin et al. (2008), Konuk (2009), Palmer and Tjiawi
a gouging ice feature. Potential consequences are not trivial— (2009) and King (2011).
downtime repair costs are estimated at up to 75% of the initial This paper is intended for non-specialists, e.g. engineers and scientists
investment for the pipeline facility (Vershinin et al., 2008), and the that are new to this subject, decision-makers higher up in a business'
environmental damages caused by a spill in these remote icy areas hierarchy with little knowledge of gouging issues, graduate students, etc.
is a serious concern to the engineer, the pipeline operator, the Hopefully, it will serve as a basic starting point for this readership. As for
regulators and the local communities. As will be seen in this article, experts in seabed gouging issues, they may find this article to be a useful
a number of options may be considered to protect these lines source of recent references.
P. Barrette / Cold Regions Science and Technology 69 (2011) 3–20 5
Arctic Circle
Fig. 1. Current offshore oil and gas developments (Dean, 2010). Areas 3, 7, 8, 16 and 18 are affected by ice gouging activity. This figure does not include planned developments, such
as in the Barents/Kara Seas in Northern Russia (Vershinin et al., 2008).
Adapted from Dean, 2010 with permission from Thomas Telford Limited.
2. The three main actors Pipeline diameter is a major design specification—it has to be large
enough to avoid too much pressure drop between ends, and small
This section provides background information on the three main enough for the deployment scheme to be cost-effective. Pipeline
actors involved in gouging scenarios: the subsea pipeline itself, the diameter typically ranges from about 0.1 m for gas flowlines up to 2 m
gouging ice features and the seabed. for high oil or gas flow volumes; this parameter comes into
consideration when high temperature and/or high pressure are an
2.1. Subsea pipelines issue (Dean, 2010). Pipeline wall thickness is chosen in function of
various factors, e.g. internal pressure containment, corrosion allow-
2.1.1. General description ance and fabrication tolerance. It is typically in the order of 10 to
Pipelines (or flowlines) are an effective means to transport crude oil 50 mm. The diameter/wall thickness ratio and the nature of the
and natural gas in the offshore environment, at water depths up to and material are used as a measure of resistance to bending. Because of
exceeding 2000 m (Palmer and King, 2008). These structures are an currents and/or waves, pipeline stability on the seafloor can be
alternative to using tankers, and although initial capital investment is problematic, but not if the structure is buried. Factors to be considered
higher, the operating costs are lower. Subsea installations typically at the planning stage are material selection, line pipe qualification,
incorporate several other components: well heads, risers, infield procedures for welding, ‘limit state’ criteria, monitoring and inspect-
flowlines, umbilicals, subsea manifolds, etc. (Fig. 2). The export pipeline ing tools (Guo et al., 2005; Kenny et al., 2007; Newbury et al., 2007;
is the component that brings the resource to shore. It may extend over Palmer and King, 2008).
long distances and is thus most vulnerable to seabed gouging.
2.1.3. Operations
2.1.2. Design criteria A pipeline has an operational lifespan of over 40 years (Palmer and
Pipeline design is generally based on stress-related principles, which King, 2008). Testing prior to commissioning includes flooding, cleaning,
guide material selection and welding requirements. Internal pressure gauging, hydrostatic pressure and leak testing (Guo et al., 2005). Flow
(from the contained fluid) is typically the most important load a pipeline assurance is a procedure conducted on a daily basis to ensure adequate
will be submitted to, but external pressure (from the water column) is fluid flow from the source downstream to the selling point. This is a
also considered. Where ice gouging is the controlling design parameter, complex operation designed to prevent and control the deposition of
a stress-based approach is inadequate, since achieving the required various products (e.g. paraffin, gas hydrates, asphaltenes), as well as
strength would be too costly. Instead, as discussed later, strain-based scales in the oil and gas production system (Guo et al., 2005, p. 7). Robotic
principles are used, whereby pipeline design makes provision for a systems are used for that purpose as well as for leak detection and
certain amount of permanent- or plastic-strain (e.g.Lanan et al., 2011). potential repairs—these devices are referred to as pigs. They are
Concentrated loads, impact and excessive plastic strain (e.g. due to introduced at a given access point and are able to circulate long distances
too much bending) during construction and service, as well as under their own power. In the case of oil, for instance, temperature
upheaval buckling are other important considerations (Abdalla et al., variations (about 100–200 °C as it is pumped out from a reservoir at
2008; Guo et al., 2005; Palmer et al., 2003; Palmer and King, 2008). depth) is a key element, as it has various implications for pipeline
Fig. 2. A seafloor oil production facility typically comprises a number of elements, examples of which are shown in this figure, in this case displaying non ice-resistant structures
(taken from Dean, 2010 with permission from Thomas Telford Limited). The export pipeline, used to bring the resource to shore, is most vulnerable to ice gouging because it typically
extends over 10s of kilometers.
performance: variation in viscosity, pressure gradients, phase deposition meltwater that flows downward through holes in the ice cover streams.
and thermal expansion of the structure are examples. If extensive enough, strudels may remove seabed material from below
the pipe, causing long spans to become unsupported—see Fig. 5 in
2.1.4. Hazards Abdalla et al. (2008) (also: Lanan et al., 2008; Leidersdorf et al., 2001;
Failure incidence is a normal part of the operating lifespan of many Mørk, 2007; Nogueira and Paulin, 1999; Palmer, 2000a). Permafrost is
mechanical systems (automobile, dishwasher, space station, …)—a another potential issue (Lanan et al., 2008, 2011; Nixon, 1991), although
pipeline is not an exception (Fig. 3). Post-commissioning failure per this has yet to be fully assessed, since in general, temperature at the sea
unit time tends to be relatively high and due mainly to inherent bottom is above freezing point (Mørk, 2007). In places where
defects (e.g. material flaws, design oversights). Failures later decline permafrost does occur, particularly near shorelines (see references in
but the pipe is still subject to external damage, such as from anchors Palmer, 2000b), thaw subsidence has to be taken into account in
and severe storms (Palmer and King, 2008; Zhou et al., 2010). Failure pipeline design (e.g.Lanan et al., 2011).
incidence then rises again due to age-related issues (e.g. corrosion, It is generally agreed that the most significant risk to pipelines in
fatigue). Failure incidents can be mitigated via regular inspections and cold waters is seabed gouging by ice features (Abdalla et al., 2008;
maintenance. The overall design of subsea pipelines also has to take Konuk, 2009; Mørk, 2007; Palmer and Tjiawi, 2009; Randell et al.,
into account a number of extrinsic factors, notably seabed type, 2008; Vershinin et al., 2008). This phenomenon influences both
strength and morphology, uneven seabed (pipeline spanning two pipeline design and route selection.
high points), sandwave fields, potential for geological faulting, As for pipeline construction, it can be undertaken in the winter if
submarine landslides, tidal currents, and wave regime in shallower the water is not too deep. This minimizes environmental impact.
areas (Guo et al., 2005; Morrow and Larkin, 2007; Palmer and King, During that period of the year, wildlife activity is less and the sea ice
2008). Other hazards include commercial fishing, ship anchors, can be used as a working platform. Harsh weather conditions and
objects of a military nature (mines, live ordnance) and sinking ships. darkness in the winter have an impact on the logistics at the
construction and operational phases. Sea ice conditions vary some-
2.1.5. Additional risks in cold waters what predictably on a yearly basis, but much less so in shorter terms.
Arctic seas present additional challenges for subsea pipelines. Here may be the crux of the matter: what constitute routine
Strudel scours are seabed erosional features produced by spring procedures for subsea pipelines down south may turn out to be a
delicate planning exercise in cold regions. The success of the operation
may rely on proper management of ‘windows of opportunities’ and
ample contingency margins (Palmer, 2001).
2.1.6. Examples of subsea pipelines in cold waters

There has been very little marine pipeline construction in the
Arctic. As a result, there is limited experience in the matter to draw
from. Palmer et al. (1979) describes the underwater flowline from
Drake F-76 project, connecting the onshore facilities to “the world's
first Arctic subsea production gas well”. An example of an operational
subsea pipeline in the Alaskan Beaufort Sea is the Northstar Project,
first development in the Arctic to incorporate this structure (Lanan and
Ennis, 2001; Nogueira and Paulin, 1999) (Fig. 4). The Oooguruk oilfield,
offshore the North Slope of Alaska, is the second Beaufort Sea oil field
using a subsea flowline system, described in Lanan et al. (2008). The
Nikaitchuq flowline system is another example (Lanan et al., 2011).
Fig. 3. Incidence failure pattern (or ‘bathtub’ curve—Palmer and King, 2008) for a given These sources provide valuable information on design, construction and
system such as a pipeline network. operability in these particular cases.
Fig. 4. Depth profile of North Star subsea pipeline, including backfill (from Lanan and Ennis, 2001). The initial sea ice cover was thickened with a pump and was used as a platform for
trenching operations.
2.2. The gouging ice features exceed 100 m, with masses that can be more than 100,000 tons (Kenny
et al., 2007). According to the International Ice Patrol (IIP), as many as
Gouging ice features are generally divided into two types (Fig. 5): 1) 2200 sightings per year were reported in the past few decades, with an
Icebergs, which are fragments broken off from glaciers and ice shelves; average of about 800 per year during the last 10 years—the peak season
and 2) sea ice ridges, resulting from the formation of a sea ice cover is from March to June (Randell et al., 2008). Iceberg ice is somewhat
which are later reworked by currents and winds. A free-floating ice weaker than other freshwater ice such as lake ice. This is attributed to
feature (i.e. before it begins to gouge the seafloor) is expected to be in the existence of pre-existing flaws (Barrette and Jordaan, 2002; 2003;
hydrostatic equilibrium. In the case of an iceberg, the immersed part Jones, 2007). Despite these features, iceberg ice is still considerably
(the keel) is about 8–9 times (depending on the density and the stronger than sea ice. Furthermore, unlike first-year ridges, which are
temperature of the ice and the water) the volume above the water line generally surrounded by an ice sheet, icebergs are often isolated features.
(the sail). It may be less than that for sea ice ridges (Timco and Burden, This has implications in terms of how the ice behaves when it is gouging
1997). Gouging features in the Beaufort Sea and in the Arctic islands are the seabed since that in these circumstances, they are freer to move. Ice
mostly sea ice ridges, while icebergs are observed in the eastern Arctic islands are very large (km-scale) tabular icebergs that calve from ice
and offshore Labrador and Newfoundland. Design criteria have to shelves (glaciers extending out onto the sea. They are extreme features
consider which case is the most relevant at the location of interest. from an offshore design view point, but are also rather uncommon.
2.2.1. Icebergs
Icebergs are individual pieces of freshwater (i.e. non-saline) ice that 2.2.2. Sea ice and sea ice ridges
formed within a glacier as a result of gravity-driven creep processes. This Sea ice (Thomas and Dieckmann, 2003; Weeks, 2010) is frozen saline
ice eventually breaks off into the sea, a mechanism known as calving. On water—it differs from frozen freshwater ice, such as lake ice or glacial ice,
Canada's East coast, icebergs drift south from Greenland. Their draft may in that it is more porous and has a more complex internal structure. As a
Fig. 5. Gouging ice features: a) iceberg, b) first-year ice ridge. For comparison purposes, both features were given a similar – and arbitrary – cross-sectional shape. Closed-circle arrows
are the forces; open-circle arrows are the reactions (modified from Croasdale et al., 2005; Lever, 1986; Vershinin et al., 2008). FB: ice buoyancy, FC: water current, FF: push from
surrounding floe, FG: weight, FH: hydrodynamic drag, Fi: inertial force (from the kinetic energy), FN: normal force from seabed, FR: restrictions from surrounding ice, FS: horizontal
force due to soil resistance, FW: wind, RF: floe failure, RK: global failure of keel, RR: local failure of rubble inside keel, andRS: pitching motion due to soil interaction with seabed. Not
all forces act at once—for instance, FH may not be involved if FC is the sole driving force (since the ice feature travels with the water), but would be expected, say, if the ice starts
moving due to FW.
result, it is also mechanically weaker than its freshwater counterpart between these particles. As a result, the soil will hold itself together
(Timco and Weeks, 2010; Weeks, 2010). Sea ice forms a perennial cover even if a small stress is applied to it. Several types of inter-particle
over much of the Arctic Basin. The surface extent of this cover undergoes bonding are observed in these soils, including flocculation and
yearly cycles, with a maximum and a minimum in the winter and cementation (e.g.Al-Khafaji and Andersland, 1992, Fig. 8.24). A
summer, respectively. In the summer, presumably through global climate detailed explanation of this phenomenon is outside the scope of this
trends, its size reduction has reached such an extent that Arctic shipping paper. At this point, what can be retained is that, in the gouging
through the Canadian Arctic Archipelago is considered (Comiso, 2002; literature, a cohesionless seabed most often designates a sandy soil,
Howell and Yackel, 2004; Kubat et al., 2006). This is an issue that could while a cohesive seabed is one generally made of a fine-grained
have an impact on exploration activities in the North American Arctic. material such as clays and silts.
A sea ice ridge forms as a result of compression loading in the plane
of the ice cover, exerted by winds and currents. Typically, two expanses 2.3.2. Cohesionless seabed
of flat (or level) ice, or floes, are driven into each other, breaking into The mechanical response of a cohesionless soil is generally
fragments of various sizes along the suture line. The resulting linear consistent with a frictional-type behavior: the stress required to
features make up a ridge, which may extend several kilometers (Timco shear the material along a plane is directly proportional to the stress
et al., 2000; Vella and Wettlaufer, 2008; Weeks, 2010). Its internal normal (σ) to that plane. This relationship is generally referred to as
structure is intricate. It consists of fragments of various shapes and sizes the Mohr–Coulomb failure criterion (e.g.Al-Khafaji and Andersland,
(making up what is known as ice rubble), and various degrees of re- 1992; Das, 1994; Dean, 2010). The angle of internal friction φ,
freezing at the waterline (known as consolidation), depending on air representing the soil's strength, is a standard parameter for soil
temperature, amongst other parameters. The keel is thought to be characterization. A cohesionless seabed is usually known to have a
weaker than the sail because of buoyancy and the near melting higher permeability—the water is able to seep in or out as the material
temperature of the ice below the waterline (Croasdale et al., 2005). undergoes a volume change during a shearing event. This is known as
Multiyear ridges are thick ridges that have survived one or more a drained response, a key concept in geotechnical engineering.
summers, during which time the ice has increased in strength due to However, because ice gouging scenarios are relatively fast, an
brine drainage (e.g.Johnston et al., 2009; Timco and Weeks, 2010). undrained response can still be expected from a sandy seabed.
Multi-year ice, also known as old ice and comprising second-year ice as a
sub-classification (Johnston, 2004), is generally less common than first- 2.3.3. Cohesive seabed
year ice. Stamukhi (the plural form of stamukha) is a term used to A cohesive soil is usually associated with a low permeability. As a
designate a large mass of grounded rubble ice (e.g.Kharitonov, 2008; result, the water content does not change significantly during the
Weeks, 2010). When these begin to move as they break up or as a result gouging event. This is known as an undrained response. During the
of larger horizontal loads, they can also gouge the seafloor. shearing process, therefore, an excess pore water pressure builds up.
In an undrained response, the shear strength is a measure of the soil's
2.3. The seabed undrained strength, typically symbolized by cu. The water content of
cohesive soils is higher than that of cohesionless soils—70 to 100 wt.%
Since much of the driving energy is absorbed by the seabed, in the former, 20–30 wt.% in the latter (Dean, 2010). The soil's
adequate knowledge of it is crucial—ice gouging is, for the most part, a consolidation state also has to be considered. A cohesive seabed is
geotechnical phenomenon. The soil's mechanical properties are a normally consolidated when the current effective overburden pressure
function of composition and grain size. Composition (granite pebbles, (‘effective’ means that the pore pressure has been subtracted) is the
quartz grains, mica flakes, etc.) depends on the source rocks inland. maximum the soil has seen in its history (Al-Khafaji and Andersland,
Grain size is a function of water energy, from currents, waves and 1992; Das, 1994). It is overconsolidated when that effective pressure is
tides—the finer, less dense material will remain in suspension longer. less.
This is why gravel and sand are often found in rivers and estuaries,
sand and silt dominate in deltas and in shallow seas, and silt and clay 3. Gouging and its bearing on offshore engineering
are found further offshore and in deeper environments (e.g.Dean,
2010; Nichols, 2009). 3.1. The ice gouging process
A core extracted from the seabed at a given location typically consists
of an alternation of layers generally at the millimeter or centimeter scale. 3.1.1. Terminology
This is known as lithostratigraphy. It reflects variations in the nature of the Several terms have been used historically to describe the action of
sedimentation process, including hydrological conditions and sediment drifting ice features on seafloors. ‘Scouring’ and ‘gouging’ are the most
input over the depositional time span (years, decades or even centuries, common. ‘Furrowing’ is also used occasionally, while ‘plowing’,
depending on soil volume transported). While implicitly acknowledging ‘scoring’ and ‘scraping’ are rare occurrences. To avoid confusion, the
that there is a wide array of potential lithostratigraphies—for instance, term ‘gouging’ is preferred to ‘scouring’ in the present article, for two
see the seabed description in Lanan et al. (2008), investigators often reasons. Firstly, the latter is used to designate erosion of the sediment
simplify this issue by dividing the seabed into two types: cohesionless and bed by water currents around a structure, an important topic in ocean
cohesive. The Grand Banks offshore Newfoundland, for instance, are engineering and river hydraulics (e.g.Annandale, 2006; Sumer and
dominated by a cohesionless soil (Sonnichsen and King, 2001), such that Fredsøe, 2002). Secondly, the term ‘scour’ is also used to describe the
laboratory or numerical simulations of gouging processes representative effects of ice surface runoff (the ‘strudel scours’ described earlier). As a
of the seabed at this location are done with that soil. On the other hand, result of this widespread usage of the term ‘scour’ (or ‘scouring’ or ‘ice
the Beaufort Sea is mostly underlain by a cohesive soil (e.g. as shown by scouring’, etc.), literature searches with these terms generate a large
Fig. 2 in Barrette et al., 2009), so this will be the seabed used in gouging number of hits that have little to do with keel-seabed interaction.
simulations representative of that area. In both cases, full saturation is
generally assumed, i.e. the soil consists of solid particles and pore 3.1.2. As the keel meets the seabed
(interstitial) water. Air content is neglected. Seabed gouging occurs when an ice feature drifts into shallower
areas, typically as it approaches the shoreline. Tidal action may also be
2.3.1. About cohesion involved. During its incursion, the keel comes in contact with the
Cohesion is the resistance to shear between adjacent soil particles seabed. At that point, a complex interplay of forces takes place, a
that stick together without normal forces acting on the contact surface simplification of which is shown in Fig. 5. The purpose of this figure is
to provide an idea of what needs to be taken into consideration when 2008; Héquette et al., 1995). A gouge depth N2 m is considered an
dealing with gouging issues. ‘extreme’ event. Gouge width may vary from a few meters to 10s of
Three scenarios may be envisaged: meters—it can exceed 400 m in the Beaufort Sea (Héquette et al., 1995;
Oickle et al., 2006), where gouging is observed at a water depth less than
1. The driving force – e.g. currents, inertia (from the kinetic energy) or
30 m. But the extreme events may occur beyond water depths of 40 m
the load exerted by the surrounding floe – does not overcome the
(much deeper along the East Coast of Canada). Gouging can also have a
resistance exerted by the seabed or the keel, and the ice feature
preferred orientation—e.g. NW–SE in the Beaufort Sea (Oickle et al.,
comes to a halt.
2006), NE–SW near the Hibernia offshore Newfoundland (King et al.,
2. If the driving force is large enough, the weak link in the case of a first-
2009). According to Blasco et al. (1998), most (99%) of the seabed
year ridge may be the keel strength. In the case of an iceberg, the
Canadian Beaufort Shelf area between 8 and 25 m water depth has been
weak link may be the moment of force: the ice simply rotates to an
affected by gouging in the last 100 years.
equilibrium angle. The ice feature keeps its course but no significant
Blasco et al. (1998) report on a 50 km-long gouge that was tracked
gouging occurs if the soil strength has not been overcome during this
down in the Beaufort Sea, with a maximum depth of 8.5 m and in water
process.
depths ranging from 40 to 50 m. Its path is not rectilinear, but crosses
3. If the driving force is large enough and the keel is stronger than the
over, and varies from straight to sinuous with occasional changes in
soil, only then will the latter fail and a gouge form in the seabed
orientation. This is considered the longest and deepest gouge in this
(e.g.Clark and Zhu, 2000). This scenario is the one that is addressed
area. The age of this extreme event is estimated at 2000 ± 200 years.
in the present article. If idealized, it leads to so-called steady-state
ice gouging (e.g.Foriero, 1998), with potentially extensive gouge
3.2. Damage to pipelines by drifting ice
lengths.
Known cases of damage to submarine pipelines by ice features
3.1.3. Ice dynamics during gouging
could provide useful information for pipeline design. But to the
The main difference between the gouging dynamics expected from
author's knowledge, very little of it has been documented. Noble and
an iceberg and from a first-year ridge lies in the way the ice feature
Comfort (1982) describe damage done to a water intake pipeline in
responds to the moment of force (e.g.Fuglem et al., 1999; Matskevitch,
the Great Slave lake (North-West Territories, Canada) in the winter of
1997). Icebergs are isolated features and their keel is made from solid
1978. It was attributed to an ice ridge. Grass (1984) alludes to
freshwater ice. Depending on impact eccentricity, the ice mass is
extensive damage by ice keels to gas pipelines in Lake Erie on the
expected to rotate (pitch) into an equilibrium angle (RS in Fig. 5),
Canadian side of the border with the United States (see Lever, 2000).
known as the keel's attack angle, with a resulting energy loss. In the
Vershinin et al. (2008) refer to Alekseev et al. (2001—not consulted by
case of a first-year ridge, an ice mass that is weaker, this situation may
the author) for an account of pipeline damage due to ice ridges. The
lead eventually to what is referred to as global failure (Croasdale et al.,
two cases above are mentioned in that source, along with a third case
2005) (RK in Fig. 5). Because they are elongated features, some form of
on iceberg-induced rupturing of communication cables offshore
restrictions (i.e. a reduction in degrees of freedom) may also be
Labrador (Canada) between 1960 and 1970. Konuk (2009) also refers
expected from the surrounding ice sheet (FR in Fig. 5).
to the same examples. Vershinin et al. (2008) interpret the lack of
Another significant difference between ridges and icebergs is with
information regarding this issue to limited experience in pipeline
respect to local failure (Croasdale et al., 2005) (RR in Fig. 5). The ice
developments in cold waters.
making up the iceberg is not expected to fail during gouging, although
some measure of ablation may occur (C. Woodworth-Lynas, pers.
3.3. Design standards, rules and codes of practice
com., 2011). But a first-year ridge may get reworked at the keel–
seabed interface. This would also lead to an equilibrium attack angle,
“A code meets the need for standard practices to be documented in
but involving rubble deformation inside the keel. The behavior of a
some formal way, as a statement of minimum requirements for the
multi-year ridge during gouging may be more akin to that of an
protection of the community” (Palmer and King, 2008, p. 572). It is not
iceberg, depending on keel strength, aspect ratio and provided the
because a design meets the code that it will be safe (Palmer, 1997a).
ridge is not too restrained by surrounding floes.
Design codes are continuously being improved upon, with new
editions appearing from time to time. They are overseen by various
3.1.4. Seafloor response
national or international committees, and can be quite different even
Gouging behavior will vary according to the physical characteristics
though they address similar issues. Code philosophy varies, as pointed
and behavior of the ice feature, as well as the seafloor response (e.g.
out by Palmer and King (2008). Some set general objectives, as when
Kenny et al., 2007; Konuk et al., 2007; Nobahar et al., 2007a; Palmer,
2001; Phillips et al., 2005). As the ice keel contacts the seabed, the soil
fails and gets displaced laterally to form side berms, and ahead of it to
form a front mound (Fig. 6). The soil below the gouge base is dragged
forward, resulting in sub-gouge soil displacements down to a
considerable depth below the gouge, in the range of a few meters for
severe gouging events. A three-fold classification scheme has been
generally agreed upon to describe qualitatively the anatomy of a gouge
along a vertical cross-section (Fig. 6):
• Zone 1 encloses the soil that has been displaced by the gouging keel;
• Zone 2 is where very large soil deformation has occurred below the
gouge;
• Zone 3 contains soil that has undergone only elastic deformation
conveyed by the overlying soil, without significant displacements. Fig. 6. Simplified scheme of a keel–seabed interface parallel to keel motion (adapted
from other sources e.g.Nobahar et al., 2007b; Palmer, 1997b). The soil displaced during
gouging forms a front mound and side berms (the latter, as shown in Fig. 9, are parallel
3.1.5. Some numbers on gouges to the plane of this section, on each side of the keel). Zone 1: gouge; Zone 2: sub-gouge
Gouge marks can reach many kilometers in length. Gouge depths soil displacement; and Zone 3: no significant soil displacement. BD: pipeline burial
may be up to 5 m, although most are less than 1 m deep (e.g.Been et al., depth.
it is recommended that pipeline burial depth take into account depth

and width of ice gouges. Others are prescriptive, as when it is
recommended that pipeline burial depth remains “not less than 1 m
along the entire pipeline length and not less than 2 m within the shore
zone” (example from Vershinin et al., 2008, p. 22). Some only include
general provisions (Vershinin et al., 2008, p. 22–23): API RP 2N 1995
states that the pipeline ‘be buried’ in order to protect it from ice
gouging; CSA S471 1992 states that the design needs ‘to account for
depth and width of ice gouges’.
According to Mørk (2007), “the growing focus on arctic oil and gas
exploration has raised the need for new standards and industry
practices […] Material selection, qualification of materials for arctic
pipelines, and design for pipelines against ice gouges are examples of
areas that require more consideration.” But an adequate understanding
of what the pipeline will be exposed to is required. According to Konuk
(2009, p. 155), “Current North American standards and regulations
provide no explicit guidance on how to incorporate the ice gouge load
conditions in the design of offshore pipelines or offer any explicit design
criteria or design acceptance limit.” This is vindicated by Lanan and
Ennis (2001), reporting on the Northstar development, who point out to
a lack of a single industry standard in this regard. These issues are
currently being addressed by updated standards, notably ISO 19906
(comment from an anonymous reviewer).
A listing of the standards and codes that are most commonly
quoted in the ice gouging literature is provided in Appendix A. The
reader is referred to Palmer (1997a) for a short but instructive
discussion on code usage as applied to pipelines, and to Jordaan (2005,
Chap. 10) for an appreciation of risk analysis and how to draw the line
between cost optimization and human safety.
4. Pipeline protection against ice gouging

Fig. 7. Means of protecting a subsea pipeline against ice gouging: a) ice management—
A gouging ice feature may exert considerable loads on the seabed: towing the ice feature away from the structure; b) shielding the pipeline with a
10 to 100 MN, according to some sources (Kenny et al., 2007; Palmer structure that is designed to absorb the impact; c) burial (BD: burial depth).
and Tjiawi, 2009). Designing a pipeline that is sufficiently strong to
withstand these loads is deemed impractical. Three main alternatives
have been envisaged (Fig. 7): 1) ice management, 2) shielding and 3) 4.3. Trenching and burial
trenching and burial.
Seabed trenching and pipeline burial are procedures that are com-
monly used in offshore engineering at lower latitudes (Cathie et al.,
4.1. Ice management 2005; Dean, 2010; Morrow and Larkin, 2007; Palmer and King, 2008).
These procedures have a number of advantages: prevention of buckling
This approach consists in towing those ice features that threaten the from thermally or pressure-driven axial expansion, protection against
structure, so as to deviate their drifting course. It is being done with hydrodynamic loading from currents and waves, thermal insulation and
icebergs, by using a floating towrope tied to one or two vessels, but also via freespan mitigation from an uneven seabed. In the Arctic, a similar
prop-washing and water cannons when dealing with smaller icebergs principle is already used for local seabed installations, such as well heads
(e.g.Eik and Gudmestad, 2010; Rudkin et al., 2005). Ice management as a and manifolds. These structures are placed at the bottom of glory holes—
means for protecting pipelines against ice gouging may not be realistic holes dug in the seabed deep enough so that gouging ice features can
considering the extensive length of the pipeline route and the resources make their way without contact with the structure. As for submarine
required for both detection and ship mobilization. pipeline protection against gouging, trenching followed by burial is
currently viewed as the best approach. The question that is now being
4.2. Shielding addressed is: what is a safe and economic burial depth, bearing in mind
the length of these structures, and the deeper the trench the more
Shielding refers to the construction of a protective structure able to expensive?
withstand a direct impact, thereby absorbing the kinetic energy that At one time, it was thought that a safe burial depth would be one that
would be delivered by an ice feature. Submarine silos and barriers would exceed the maximum expected gouge depth, so as to avoid direct
made from rocks, steel or concrete are examples (e.g.Vershinin et al., impact (e.g.Chari, 1979; Marcellus and Palmer, 1979). This meant the
2008). These are used for local installations, such as well heads, pipeline had to be buried below ‘Zone 1’ in Fig. 6. However, later
production trees and conductor casings. Flexible concrete mattressing investigations demonstrated that the soil displacement below the gouge
or rock dump berms and even tunneling can be considered for (inside ‘Zone 2’) was such that it could induce substantial pipeline
pipelines (Morrow and Larkin, 2007; Vershinin et al., 2008). Such displacements, but with a progressive decrease with depth. Generally,
structures may be used for the pipeline's shore approach, which is emplacement in ‘Zone 3’ would lead to a safe but uneconomical design.
more exposed to currents and waves, an important consideration in However, taking advantage of the structure's ability to bend, the
pipeline layout (Palmer and King, 2008). However, here again, these challenge is now to determine at what depth to leave the pipe within
methods would not be cost-effective if applied to lengthy segments of ‘Zone 2’. This means the traditional stress-based designs needed to be
a pipeline network. re-assessed in favor of structures able to sustain a large amount of
deformation (e.g.King et al., 2009; Konuk, 2009; Lanan et al., 2008; Mørk, 5. Investigations on gouging dynamics
2007; Newbury et al., 2007; Younan et al., 2007). It also means that due
consideration should be given to bringing the pipe's crown that much Since the 1970s, a very large amount of resources has been expended
closer to the gouging keels. in generating our current knowledge-base on ice gouging. The bulk of this
work was aimed at obtaining direct and indirect evidence on gouge depth
and sub-gouge deformation. These data exist in various forms: textbooks,
4.4. Other options journal articles, conference papers, workshop proceedings, graduate
theses, government and industry reports, etc. The data in these sources
Since trenching costs increase rapidly with burial depth, this procedure were generated in a number of ways, notably through 1) field studies, 2)
may end up costing more than the pipeline material (Nobahar et al., simulations in a laboratory, and 3) numerical (computational) modeling.
2007a; Palmer and Tjiawi, 2009). This is partly because the current The complementarity of these approaches is summarized in Table 1. This
trenching technology normally handles depths not exceeding about 2 m. section should provide the reader with an idea of the ‘tools’ available to
Beyond this, dredging, a slow and expensive process, becomes necessary extend our knowledge-base on gouging issues, and the resources they
(Palmer and Tjiawi, 2009). This results in a high sediment input into the involve.
water column, which may become an environmental issue.
To make this option economically more viable, it was suggested to
incorporate a pipe-in-pipe design into the trenching-and-burial Table 1
approach (Konuk, 2009; Konuk and Yu, 2007a). Alternatively, ice Complementarities between the various types of investigations conducted on seabed
gouging.
management could be used in conjunction with that approach, so as to
reduce gouging frequency and, therefore, reduce the probability of Nature of Description Advantages Drawbacks
encounter (King et al., 2009). A novel approach is put forth in Palmer investigation
and Tjiawi (2009): the lay-out of a ‘weak layer’ above the pipeline, Field Data collection The focus is on the Logistically
with the purpose of reducing the forces and soil displacements that from large size ‘real thing’. expensive,
are transmitted downward. This option, covered by a U.S. patent features in their resource-
natural intensive.
application, could also allow a reduction in pipeline burial depth.
environment.
Seabed Information Critical No info on
mapping gathering from information on gouging ice
4.5. Bottom line on pipeline protection: burial depth ships on gouging gouging depth, feature or on sub-
activity in a given width, length and gouge
area, involving frequency deformation.
Subsea pipeline protection against ice gouging is best achieved by repetitive
burying the structure below the seafloor in order to escape the keel's mapping.
deleterious effects. The ultimate question is: what is a safe and eco- In-situ ice properties Measurements of Best way to obtain Instrumentation
nomical burial depth? The answer is not straight-forward. It can only be physical and realistic and event
mechanical information on monitoring are
obtained if adequate information is available on the materials involved
properties of realice features, challenging.
in a gouging scenario (ice, soil, pipeline). Further, the gouging regime to ice features. notably ice
which a pipeline will likely be exposed during its operational lifespan, rubble.
and what constitutes acceptable risks, also need to be factored in a Gouging Enactment of Real ice on real Instrumentation
design scheme. Fig. 8 shows a schematic framework with each of these simulations gouging events. seabed, large and event
events with monitoring are
components. realistic forces. challenging.
A proper appreciation of what this challenge entails can be best Relict On-land study of On-foot access to No information
achieved by looking at the nature of the investigations conducted so far on gouges sub-gouge unique dataset on forces involved
ice gouging dynamics, with a focus on recent work. This will enable the deformation from from large-scale during the
old (1000s years) event on gouging event or
reader to better understand how that information is generated and help
gouging events. subgouge on the pre-
decide where available resources should be directed in the future. deformation. gouged seabed.
Laboratory Artificially Simulations more Uncertainty about
produced data in easily how
a fully controlled implemented and representative of
environment. instrumented. real scenarios.
Simulations at single Instrumented Fairly sizeable Vertical stress
gravity simulations in a simulations can distribution in the
basin or flume, at be achieved. seabed not
normal gravity. representative of
real scenario.
Centrifuge Instrumented Vertical stress Scaling issues,
simulations simulations at a distribution in the length of drainage
controlled seabed is paths, Coriolis
centripetal representative of forces, need to be
acceleration. real scenario. addressed.
Theoretical Using known Low resource Simplification of
physical requirements. highly complex
principles to phenomena.
Fig. 8. General framework summarizing the main aspects involved in the determination
derive the overall
of burial depth. Example of material properties are keel strength, soil strength and
gouging process.
pipeline specifications. A gouging event, either in nature or in a physical or numerical
Numerical Computer Provides Requires rigorous
simulation, incorporates ice, soil and pipeline responses. In numerical solutions, these
simulations of information on all validation against
events should be mutually compatible (known as ‘coupling’ and here indicted by
gouging salient aspects of quality data from
double-headed open arrows). Gouge data can be an event output (generated by the
scenarios. gouging, physical
gouging event), an event input (field data used to verify/validate a gouging simulation)
including sub- simulations.
and a source of information about gouging (e.g. frequency). Reliability level is the
gouge
selected probability of occurrence (e.g. 1000 events per year or 10− 3 yr− 1); an example
deformation.
of a limit state is the maximum allowable pipeline strain.
5.1. Field studies 5.1.2. In situ characterization of real ice features

Gouge depth is strongly dependant on the material properties
Field studies may be further divided into four categories, each with (notably mechanical strength) of the soil and the ice. For instance, a
its own objectives: 1) to extract information from gouging scenarios strong ice feature is expected to gouge deeper, all other parameters
in a natural environment; 2) to conduct in situ studies of real ice being equal. Because the internal structure of icebergs is compara-
features; 3) to obtain information on gouges through seabed tively simple, adequate knowledge of this ice's properties can be
mapping; and 4) to obtain information on sub-gouge deformation obtained by testing it in a laboratory environment (e.g.Jones, 2007).
by documenting on-land relict ice gouges. This is not the case for ice ridges, whose internal structure is complex
and not well understood. Amongst the uncertainties are the extent
and strength of the consolidated layer, the extent of refrozen rubble,
5.1.1. Gouging scenarios in a natural environment and the thermal and salinity gradients in the keel. According to
Real seabed gouging scenarios – i.e. the ‘real thing’ – offer an ideal Croasdale et al. (2005), “artificially created ice rubble can never hope
knowledge-gathering platform, but they involve significant technical to be representative of real ice rubble” (p. 25). The solution is to
and logistical challenges. Firstly, an area has to be identified where conduct mechanical testing on real ice rubble in the field. Sizeable test
gouging is known to occur—the exact location of the incursion is hard rigs are required for that purpose. The outcome of such tests cannot be
to predict. Secondly, once an ice feature has been identified as a representative of all ice features, given their variability in nature, or
gouging candidate (based on its drift path), it has to be closely replicate the internal stress states of the ridges (Konuk et al., 2007).
monitored (drift speed, size/mass, response to the gouging event), But it does provide information – an order of magnitude – that can be
and instrumented (for instance, with motion sensors for driving force used as guidance for designing gouging simulations in the laboratory,
estimates). Thirdly, since soil conditions are likely to be unique to this or for validating numerical models.
particular location, accurate in situ seabed characterization (soil Croasdale et al. (2001b) and Heinonen and Määttänen (2001) report
properties, gouge morphology) would be required. This is not easily on that kind of work. Both set-ups comprised hydraulically driven rams
achieved in the field. An example of a real seabed gouging scenario is to either punch the ice downward or horizontally, while logging loads
the Dynamics of Iceberg Grounding and Scouring (DIGS) experiment and indenter displacements. Other examples are documented by
on the Labrador continental shelf, conducted in 1985 (Hodgson et al., Høyland et al. (2002), Liferov et al. (2002) and Liferov and Høyland
1988; Lever et al., 1991; Woodworth-Lynas et al., 1991). Motion (2004).
sensor packages were deployed on four different icebergs using a
helicopter. Reported iceberg masses range from 2 to 19 million tons, 5.1.3. Seabed mapping
with keel drafts varying from 107 m to more than 170 m. Hydro- Seabed mapping allows an appreciation of gouging activity.
meteorological data were recorded. Seabed interactions, calving and Historically, this has been done by conducting geophysical surveys at
rolling events were captured. Gouges in water depths exceeding a given geographical location, using ship-borne instruments such as
100 m were characterized visually and seabed mapping and profiling single- or multiple-channel echo sounders, side-scan sonars and sub-
were performed with seismic systems and echo sounders. Visual bottom profilers (Fig. 9). The information that can be collected includes
documentation was also done from a manned submersible, used to gouge depth, width, length, orientation, density and frequency.
collect core samples. Recommendations for improving operational Documentation on this subject is extensive (e.g.Croasdale et al.,
efficiency are included in Hodgson et al. (1988). 2001a; King, 2011; Leidersdorf et al., 2001; McKenna et al., 2003;
A seabed gouging event can also be ‘simulated’ in the field. This is Poplin et al., 2002; Sonnichsen et al., 2005, and references therein).
done by artificially supplying the driving force to an ice feature by A challenge with determining gouging depth is the rate of
some mechanical means. The downside with this approach is that, due sediment infilling: over the years, on-going sediment deposition
to a limitation in the amount of mechanical energy that can be causes a reduction in depth between (a) when the gouging event
generated, the ice features used in the gouging enactment are occurred and (b) when the survey is done (Kenny et al., 2007; King,
considerably smaller than in real scenarios (though still more sizeable
than in a laboratory). But this is offset by improved control and
instrumentation. The work done by Shapiro and Metzner (1987)
comes to mind, but it was mostly focused on determining the friction
coefficients of large ice features sliding on a natural beach. This study
was done in Barrow (Alaska) using two large slabs of sea ice
(10,600 kg and 12,000 kg). Each one was hauled along short in-
crements (b2 m) with a bulldozer, while recording the load with a
dynamometer.
Liferov and Høyland (2004) reported on field-based simulations
of a gouging event with natural ice on a real seabed. Two simulations
(one in 2002, the other in 2003) were carried out in Spitsbergen
(Norway) on a soft silty clay. The ice was manually cut from the
parent ice sheet, which was about 0.9 m in thickness. The final size of
the blocks was about 0.15 × 0.25 × 0.45 m. Ridge dimensions were
about 4 m in length, 4 m in width and between 1 and 2 m in height
(between keel bottom and sail top). The rubble at the water line was
allowed to freeze during a few weeks, with a resulting consolidated
layer of 0.7 m in both tests. The driving force was provided by a
heavy loader (2002) and a pneumatic winch (2003). During the
gouging events, the pulling force and vertical and horizontal Fig. 9. Example of ship-borne instrumentation used in the past for seabed mapping
displacements were monitored. Other instrumentation included purposes (adapted from Dean, 2010; Poplin et al., 2002) (not to scale). The side-scan
sonar was to determine gouge orientation, number and width, while the single-beam
temperature probes and underwater video. In order to keep track of sonar provided bathymetry. A gouge profile with side berms is indicated, assuming
keel behavior, ‘failure indicators’ were specially designed to record gouging direction was perpendicular to the plane of the diagram. Note how sediment
internal shearing. infill contributes to reducing the measured gouge depth.
2011; Palmer and Niedoroda, 2005; Poplin et al., 2002)(Fig. 9). Gouge control the simulation (vertical loads, driving force, etc.), resulting in a
infilling thus introduces a non-conservative bias in the data. Although more complete data set (loads, displacements, gouge geometry, pre-
a correction factor can be applied (e.g.Kenny et al., 2007; King et al., and post-test soil analyses, etc.). A second advantage is reduced costs.
2009), this issue is generally addressed by mapping the seabed Information generated with laboratory simulations, such as sub-
repeatedly over a number of years. Such a procedure allows estimates gouge deformation, can complement that from real-scale gouge relics
of infilling rates. Moreover, repetitive mapping helps to distinguish (Woodworth-Lynas et al., 1998, see their Table 1). But laboratory
young gouges from old ones and to determine gouging frequency, a work also introduces a significant unknown in terms of how
parameter that can be represented by the frequency of crossings of a representative these much smaller, artificial events are of real-scale
line. For instance, to guide design of the 9.5 km pipeline (in 8–10 m scenarios. This is addressed by scaling up the laboratory results, a
water depth), repetitive mapping was conducted for the Northstar standard procedure in physical modeling (e.g.Altaee and Fellenius,
development in the Beaufort Sea over a four year period—from 1996 1994; Muir Wood, 2004; Novak et al., 2010; Zufelt and Ettema, 1996).
to 1999 (Leidersdorf et al., 2001). A total of 48 gouges were detected: Physical modeling consists in converting laboratory data to the
16 in the first year, 8, 3 and 21 in the following years, with a mean level of the prototype (the real-scale event) so that the former
depth of 0.4 m and a resolution of 0.1 m. The total length of gouge correctly reproduces the latter. This normally implies that dimen-
features over an area per year may also be used as a measure of sions, velocities and forces in the model are representative of those in
gouging frequency. A mean annual pipeline gouge crossing rate of the prototype. But in complex phenomena such as a seabed gouging
3 × 10 − 4 km − 1 yr − 1 was determined by King et al. (2009) from their scenario, full similarity is not easily achievable, such that compro-
analysis on the East Coast. A third alternative is the frequency of gouge mises are made. Scaling numbers, such as Froude, Peclet, Fourier and
features and their average length over an area. Hence, two gouges other dimensionless parameters are sometimes used in gouging data
250 m in length and one gouge 500 m in length would present the analysis (Been et al., 2008; Kioka et al., 2001; Schoonbeek and
same hazard. Allersma, 2006). In many instances (e.g.Barker and Timco, 2002, 2003;
A potentially revealing field survey was proposed for gouge depth Barrette and Timco, 2008; Stava et al., 2008; Vikse et al., 2007), scaling
determination (Palmer, 1997a; 2001). It would consist in probing, at a analyses are not done. The reason is that these test programs were
location of interest, the seabed strength with a standard tool (e.g. a cone designed specifically for the validation or calibration of numerical
penetrometer) and see if a sudden increase in seabed strength can be models, an equally relevant rationale. (It may still be instructive to
detected. One would expect such a break to represent the maximum conduct a physical modeling exercise on these data.) For a discussion
depth at which soil remolding from recent gouging events has occurred. on the complementarity of these two approaches, the reader is
That break would thus delimit the base of Zone 1 in Fig. 6. referred to Scott (1988) and Randolph and House (2001).
Information on gouge depth is a useful parameter in its own right, There have been a significant number of laboratory investigations
but to date seabed mapping has provided little indication of what on ice gouging. The ice keels are generally represented by rigid
happens below the gouge. Since sub-gouge soil displacement is an indenters as opposed to real ice. The attack angles for any given test
important parameter for design purposes, one must rely on other program are most often chosen on an arbitrary basis—overall, this
ways of investigating this parameter. parameter spans the full 0–90 degree spectrum. It is thought to affect
the soil response but very little is known about what this angle is in a
5.1.4. On-land gouge relics real gouging event. Depending on test set-up, the keel's freedom of
Sub-gouge deformation, a critical element when assessing pipeline motion also varies, from fully constrained to allowable along all six
burial depth (as seen earlier), cannot be effectively studied in a seabed degrees of freedom. The test bed often includes an embedded pipeline.
having undergone real scale gouging events, because of limited access A steady-state gouging regime is not always achieved. Some of the
and observational methods. This is where investigations of on-land complexities found in nature, such as stratified soils, trench backfill
gouge relics can be helpful. These are ice gouges on the bottom of over the pipeline, are often not considered. Additional insights on
ancient water bodies, such as glacial lakes and inland seas that formed gouging simulations in the laboratory can be found elsewhere (Konuk,
from the melt water of retreating ice caps several thousand years ago. 2009; Konuk et al., 2007; Palmer and Niedoroda, 2005).
These water bodies have since disappeared, but some of the gouges Laboratory investigations are divided into two types: those at
were preserved; they were trenched and investigated (Eden and single gravity (on the floor of a laboratory) and those at high simulated
Eyles, 2002; Eyles and Clark, 1988; Woodworth-Lynas et al., 1996; gravity (in a centrifuge) (Fig. 10).
Woodworth-Lynas and Guigné, 1990).
Relict gouges are viewed as modern-day analogs. The soil deforma-
tion data obtained from this type of investigations are thought to be 5.2.1. Simulations at single gravity
representative of real scale events involving very large icebergs. These Laboratory simulations are said to be done at normal or single
studies provide insights on amount and extent of sub-gouge soil gravity to distinguish them from those done inside a centrifuge. The
deformation during such events, which are difficult to obtain in other keel–seabed interaction event is conducted inside a flume or basin on
ways. The drawback with this type of work, however, is that it only a laboratory floor (Fig. 10b). The size of these test set-ups ranges from
provides part of the story. The mass of the ice feature that produced the meter to decameter scale. Different methods are used to build the soil
relict gouge is unknown, so there is no information on the energy it was bed, from a simple manual transfer operation (or with a front-loader, for
delivering to the seabed during its interaction. The soil state before the larger volumes), followed by compaction if required, to bring the soil to
gouging event began is not known either. Because the seabed got re- the desired density level. This is done either manually or with a tool such
worked during the interaction, its properties (e.g. density for a as a plate-packer. More elaborate soil lay-out systems include a ‘raining’
cohesionless seabed; undrained shear strength for a cohesive seabed) procedure for sand, to ensure uniformity (e.g.Paulin et al., 1991).
may have changed significantly as a result of gouging. Without this Actuators may be driven in a number of ways – e.g. hydraulically,
information—ice mass and initial seabed properties, the data on sub- pneumatically and electrically – with various capacities, depending on
gouge deformation gathered from relict gouges is of limited value. keel size, vertical load, seabed strength, etc. Profiles and markers in the
seabed can then be examined to document gouge geometry and sub-
5.2. Laboratory studies gouge deformation. Recent sources documenting simulations at single
gravity include Barker and Timco (2002, 2003), Ishikawa et al. (2005),
A laboratory environment offers clear advantages over the field to Vikse et al. (2007), Been et al. (2008), Barrette and Timco (2008) and
conduct ice gouging simulations. The first one is the ability to better Stava et al. (2008).
Fig. 10. Ideally, laboratory simulations have to correctly reproduce the vertical stress profile in the real scale events, shown in (a). Although this is not possible in a test set-up at
normal gravity (b), it can be achieved in a centrifuge (c). This simple principle is the main reason why centrifuge technology has achieved such recognition in geotechnical
engineering. The photograph in (b) is the set-up documented in Barrette and Timco (2008). The centrifuge shown in (c) is C-Core's (Phillips et al., 1994).
5.2.2. Simulations at high simulated gravity water drag and the work associated with seabed displacement during
Centrifuge simulations take place on a platform affixed at the gouging. Been et al. (1990) and Croasdale et al. (2005) specifically
extremity of an arm designed to spin around a vertical axis (e.g.Phillips, address the intricacies of soil deformation ahead of the keel, assuming
1995) (Fig. 10c). An increase in spinning rate results in a higher two-dimensional plane strain and relying on the plasticity theory.
centripetal acceleration on the platform. The amount of acceleration is Kioka et al. (1999, 2004) provide a similar analysis. Choi and Lee
usually expressed in terms of a ‘g’ factor—for instance, 100 g is (2003) further discuss this type of analyses. Such studies have
equivalent to one hundred times the Earth's gravitational acceleration, generated information on how gouge and soil displacements vary as a
or 981 m/s2. In this example, the geometric scaling factor n would be function of kinetic energy, attack angle and other parameters. That
100. The payload surface area for centrifuge testing is small, typically information can then be compared with physical simulations or be
within 1 m 2. This makes seabed preparation less labor-intensive used to check outputs from numerical simulations.
(compared to that for the larger model done at single gravity). It also
limits the size and amount of instrumentation that can be deployed
inside the model. Actuator control and event monitoring are done 5.4. Numerical simulations
remotely, since the test set-up is ‘in flight’ while the simulation takes
place. These set-ups must be well endowed with lights and cameras. Over the years, computational modeling has become a highly
Centrifuge simulations have a critical advantage over those at single effective tool for simulating complex events such as seabed gouging.
gravity in that they allow the model to properly simulate the stress One of the great benefits of this approach is that it has the potential to
inside the soil column. The principle is as follows: if a model 1/n the size address all essential aspects of the gouging scenario: keel behavior,
of the prototype is submitted to an inertial acceleration n times that of soil response (including sub-gouge deformation) and soil–pipeline
the earth's gravity, the stresses due to self weight will be identical to the interaction. The principle of numerical modeling is based on solving a
stresses in the prototype at homologous points (Fig. 10). The centrifuge set of governing equations that usually address the conservation of
also has limitations. These include scaling of particle size, linked with the mass and momentum. A constitutive response establishes a link
formation of shear zones (Palmer et al., 2003), permeability and path between stresses and strains (or strain rates). Isotropic elasticity and
length for pore pressure dissipation. Other considerations, such as the plastic yield criteria (e.g. Mohr–Coulomb and Cam clay model) are
Coriolis forces, have also been raised (e.g.Konuk, 2009; Konuk et al., well known examples of these responses, but others are used,
2007; Yang et al., 1996). Recent sources documenting simulations at depending on applications, material type and interaction conditions
high simulated gravity include Phillips et al. (2005), Allersma and (e.g. strain rate, stress levels, interstitial fluid effects). There are
Schoonbeek (2005) and Schoonbeek et al. (2006). several approaches for the numerical solutions e.g. finite element, and
finite difference methods. There are also a number of ways to deal
5.3. Theoretical analyses with large deformations, e.g. remeshing and Lagrangian advection. All
have advantages and disadvantages. Fig. 11 is an example of a model
Geotechnical phenomena can also be approached by means of output from Sayed and Timco (2009). Ice gouging simulations were also
theoretical treatments (e.g.Muir Wood, 2004). Because they only reported by other investigators (e.g.Abdalla et al., 2008; Fredj et al.,
address the essential mechanisms involved, this type of analyses 2008; Kenny et al., 2005; Konuk and Gracie, 2004; Konuk and Yu, 2007b;
provide relatively rapid solutions, which can guide more elaborate Nobahar et al., 2007b; Serré, 2011; Stava et al., 2008). Differences do
endeavors, such as numerical or physical modeling. For ice gouging exist between the various numerical models and these are discussed
issues, the early work by Chari et al. (1980) comes to mind. It is based elsewhere (e.g.Fredj et al., 2008; Konuk, 2009; Konuk et al., 2007; Liferov
on the balance between the kinetic energy of the ice feature plus et al., 2007; Sayed and Timco, 2009).
Fig. 11. Examples of an output from a numerical model after 5 s from the start of the simulation (Sayed and Timco, 2009). Left: particle positions in the soil, right: contour plot of
pressure distribution.
A numerical model represents a simplification of the phenomenon The limit state design procedure needs to take into account all
being modeled, in that it attempts to capture the mechanisms that are applicable limit state conditions. For pipelines exposed to ice gouging,
the most relevant. The constitutive equations have to be reasonably it is generally acknowledged that the amount of bending as a result of
representative of the material behavior. Computational efficiency is also soil displacement is a key limit state. Pipelines are robust structures—
important. Moreover, to be accurate, the model needs to be carefully they can withstand some measure of distortion (Palmer and King,
validated against quality information (material parameters, gouging 2008). But that distortion becomes objectionable if it locally induces
data, etc.) from other sources, which are not always available. Other excessively high fluid velocity or pressure drop, or prevents the
challenges include very large soil deformation (not easily simulated passage of intelligent pigs. Internal blockage by wax or hydrates is
with conventional numerical approaches), pore water flow in the soil another example of serviceability issues. Excessive bending and
(consolidation and seepage history), coupling of contrasting material concurrent localized stress in the structure may ultimately lead to a
behavior (e.g. the interaction between a deformable keel and soil), leak, thereby defining an ultimate limit state. Direct impact between
effects of trench backfill and residual (post-construction) pipeline the ice keel and the structure may be considered an accidental limit
stresses at the soil–pipeline interface (Kenny et al., 2007; Konuk, 2009; state if a gouge reaches the pipeline. This may occur, for instance,
Liferov et al., 2007; Muir Wood, 2004; Sayed and Timco, 2009). because gouging depth was underestimated or trenching at some
As computer processing power improved over time, so did model point along the length of the structure failed to reach the design
output visualization (i.e. what the simulation looks like on a computer depth. More information about these issues may be found elsewhere
screen). Nowadays, numerical simulations results come forth as (e.g.Lanan and Ennis, 2001; Nobahar et al., 2007a; Palmer, 2001;
highly dynamic 2D or 3D animations with sophisticated color Palmer and King, 2008; Palmer and Niedoroda, 2005; Younan et al.,
schemes. As pointed out in Muir Wood (2004), computer simulations 2007).
may ‘look’ realistic, but these simulations will be misleading if the
model is not properly designed and validated.
6.2. Reliability levels
Simply put, a structure is expected to fail if a demand D exceeds

6. Toward a design methodology
capacity C (i.e. D N C). If both D and C were single values, then system
predictability would be relatively straight-forward. But in many
Ultimately, guidelines are required to help an operator plan pipeline
engineered structures, this is not the case. The reason is that the
deployment. Several research groups have come up with design method-
system includes a number of imponderables: material defects, welding
ologies for that purpose. This is now discussed. But firstly, a brief
quality, bolt tightening, etc. A probabilistic approach is then required,
definition of two key concepts is provided: limit states and reliability
which is typically expressed as an annual probability, measured in
levels.
orders of magnitude per year (e.g.Jordaan, 2005, Chap. 10). Palmer and
King (2008, p. 519) quote DNV regulations that indicate a target
probability of failure of 10 − 5 per pipeline per year, taking into account
6.1. Limit states
various factors, such as safety class, fatigue and accidental limit states.
This is equivalent to one failure for 1000 pipelines operating during
The limit state concept is important inasmuch as it relates to what
100 years.
constitutes an unacceptable damage to the structure. Offshore design
The reader is referred to Palmer and King (2008, chap. 16) for further
standards consider four kinds (Dean, 2010; Palmer and King, 2008):
discussion on the reliability theory and its relevance to pipelines, and to
Jordaan (2005) for a more rigorous treatment of concepts such as target
• Ultimate limit state: This is an extreme event but which compro-
probabilities, factor of safety and reliability.
mises the structure's integrity.
• Serviceability limit state: This is a less severe but more common
event—it lies within the structure's operational range. 6.3. Methodologies for estimating gouging parameters
• Accidental limit state: This is an event that is conceivable and control-
lable but which is not meant to happen. Following are examples of recent efforts aimed at improving our
• Fatigue limit state: Resulting from cyclic loading. understanding of ice gouging phenomena. The purpose of this section is
to provide a general description of what each of these studies consist of into a probabilistic formulation that considers two limit states: 1) keel–
and how they can contribute to this aim. As will be seen, some are rather pipeline contact and 2) exceedence of 0.2% strain upon limited contact
limited in scope, while others go a long way toward an answer on burial (with soil and/or keel failure). The methodology's key output (shown in
depth. Their value rests on the validity/quality of the input parameters their Fig. 10) is an annual non-performance profile for both limit states,
and the effectiveness of the chosen theory or algorithms for numerical along the full length of the pipeline.
processing. More information on these aspects is included in each
source, as well as on parameter uncertainties, error margins and how 6.3.3.2. Assessment. This study allows an appreciation of a few important
these factors were dealt with. These methodologies are from recent factors relevant to a pipeline planning scheme. For instance, it considers
sources and are presented for an informative purpose only—further the consequences of failing to achieve the target burial depth along
evaluation is beyond the scope of this paper and is left to the reader. certain segments of the structure. It also draws attention on a potential
leeway in design that may be gained by allowing the soil and the
6.3.1. Kioka et al. (2004) pipeline to absorb some of the energy delivered by the keel. The
purpose of this investigation was to evaluate the risks of allowing ice–
6.3.1.1. Description. Kioka et al.'s (2004) model derives gouge depth structure contact (as opposed to providing guidance on burial depth).
for ice features whose driving energy is provided by the surrounding This methodology includes some simplifying assumptions, e.g. gouge
ice sheet along with wind and current. Gouging is assumed to stop direction at right angle to pipeline axis and keel attack angle neglected.
once the resistance mobilized by the seabed becomes high enough.
The ice feature is a ridge whose geometry is hypothetical, but whose 6.3.4. Nobahar et al. (2007a, 2007b)
internal dimensions are interrelated (based on a known model).
Frequency distributions (log-normal, beta) and deterministic values 6.3.4.1. Description. Nobahar et al. (2007a) look at the failure
were assigned to the most influential parameters (as indicated by a probability anywhere along the 10-km length of a hypothetical
sensitivity analysis)—e.g. keel draft, ice thickness, attack angle. Using a pipeline, for various diameter/wall thickness ratios. The procedure
theoretical model, a gouge depth distribution was generated, along factors in pipeline properties and assumes a soil with a 30 kPa
with gouge length as a function of keel draft. undrained shear strength. Two limit states are considered: 1) 90%
pipeline yield stress, and 2) pipeline compressive strain criteria. For
6.3.1.2. Assessment. This methodology mainly addresses gouge depth. the latter case, FEA was used to determine the effects of sub-gouge soil
Material parameters are assumed. Soil–pipeline interaction and sub- displacements on pipeline behavior. This analysis provides evidence
gouge deformation are not considered. as to why stress-based criteria are not appropriate for displacement-
controlled loading events.
6.3.2. Croasdale et al. (2005)
6.3.4.2. Assessment. This methodology is not unlike that of King et al.
6.3.2.1. Description. Croasdale et al. (2005) put forth an ‘Ice Gouging (2009). It addresses all factors except for the ice keel, the nature and
Model’, meant to determine the maximum gouge depth for a range of behavior of which is not considered. The authors report that it has
soil and keel properties. This model specifically addresses gouging from been used ‘successfully’ in several offshore projects.
first-year ice ridges. It is based on the authors' theoretical model
describing the soil behavior ahead of an advancing inclined plate, the 6.3.5. Been et al. (2008)
latter representing the ice keel, and using the Mohr–Coulomb failure
criterion. Some assumptions are made on keel strength, geometry, 6.3.5.1. Description. Been et al. (2008) present a sub-gouge displace-
bearing capacity and soil properties and these inputs take the form of ment model reasonably consistent with results obtained from both
probabilistic distributions. The keel moves up an inclined seabed, and single gravity and centrifuge data. This exercise was for a cohesive
the forces are calculated for each displacement increment. Keel failure, seabed with keel attack angles of 15, 30 and 45°. In order to compare
either global or local, during gouging is allowed but is compensated by the two data sets on an even level (since they were produced with two
adjusting the level of the consolidated layer. The keel stops when the different test set-ups), they use a non-dimensional number that
reaction force from the seabed exceeds either the drifting force or the incorporates the parameters thought to affect sub-gouge deformation
strength of the ice cover surrounding the gouging feature. The most most: undrained shear strength, soil stiffness, gouge depth and width,
influential parameters were found to be soil strength, keel attack angle keel angle and the soil–keel interface strength.
and ice strength.
6.3.5.2. Assessment. The nature of this model is empirical, meaning that
6.3.2.2. Assessment. If reliable information is available on ice keel and soil it does not adhere to a theory or is not derived from numerical
properties for the area of interest, these can be input into the model and algorithms (e.g. FEA). As such, it is only valid for the conditions that
gouge depths can be estimated. If these data are not available, a range apply to the test set-ups used to produce the data sets. The keel–soil
for each parameter can be used as input into the methodology and a interface shear strength is assumed to be the same for all tests. The
probabilistic distribution on gouging activity may be derived. Soil– model specifically addresses sub-gouge deformation.
pipeline interaction and sub-gouge deformation are not addressed.
6.3.6. King et al. (2009)
6.3.3. Younan et al. (2007)
6.3.6.1. Description. The methodology presented by King et al. (2009)
6.3.3.1. Description. The methodology devised by Younan et al. (2007) is applied to a hypothetical pipeline offshore Newfoundland in an area
generates a ‘reliability framework’ for a hypothetical 9 km-long pipeline, where good quality information on gouging events is available (the
taking into account pipeline properties (length, diameter, wall thickness, data set includes 950 gouges and 40,000 cross-sections). It looks at the
yield strength and elasticity modulus), bathymetry, gouging frequency failure probability of the pipeline (anywhere along its 60-km length)
and variations in burial depth (from a target value of 1 m). The pipeline for burial depths varying from 0 to 4 m. The seabed is assumed to
capacity was determined through finite element analysis (FEA), the consist of clay with a 100 kPa undrained shear strength. The
output of which was used to establish a ‘non-performance’ limit of 0.2% procedure factors in pipeline properties (length, diameter, wall
strain. This information, along with probability distributions for ice keel thickness, material) and gouging frequency (3 × 10 − 4 km − 1 yr − 1,
parameters (strength, bearing factor, gouge depth and width), were fed equivalent to 58 crossings per year over the structure's full length).
Two limit states are considered: 1) keel–pipeline direct contact and 2) Acknowledgements
pipeline compressive and tensile strain exceeding 2.11% and 2.5%,
respectively. For the latter case, FEA was used to determine effects of The work presented in this paper was funded by Canada's Program of
sub-gouge soil displacements on pipeline behavior. A failure return Energy Research and Development (PERD). Comments from G. Timco
period of 1000 years and 10,000 years were considered acceptable are gratefully acknowledged. M. Sayed is also thanked for providing
when near an unmanned and manned structure, respectively. helpful feedback. Generous input from three anonymous reviewers
contributed significantly to improving the manuscript.
6.3.6.2. Assessment. This analysis addresses pipeline failure due to both Appendix A. Standards and codes
direct impact and excessive strain due to soil displacement. The
authors point out to additional issues that would need to be looked Following is a listing of some of the standards and codes quoted in the
into: refinements on strain limits, non-orthogonal gouge crossings, scientific literature on seabed ice gouging (from Croasdale, 2000;
keel shape and attack angle. The input to this methodology comprises Croasdale et al., 2001a; Dean, 2010; Kenny et al., 2004; King et al.,
real information (as opposed to educated guesswork) and incorpo- 2009; Lanan and Ennis, 2001; Morrow and Larkin, 2007; Nobahar et al.,
rates all major components of the interaction, except for the ice keel, 2007a; Randell et al., 2008; Vershinin et al., 2008; K. Been, pers. comm.).
the nature and behavior of which is not considered. This listing is not up to date. But it allows a glimpse of what needs to be
considered in subsea pipeline design and construction. A sizeable
appendix in Palmer and King (2008) is dedicated to this topic. Acronyms
7. Summary and conclusion
are: the American Petroleum Institute (API), the American Society of
Civil Engineers (ASCE), the American Society of Mechanical Engineers
7.1. Current status on burial depth
(ASME), the British Standards (BI), the Canadian Standards Association
(CSA), Det Norske Veritas (DNV) and the International Organization for
It is generally agreed that adequate protection of subsea pipelines
Standardization (ISO).
against ice gouging can be provided by burying these structures into
the seabed. As to what constitutes a safe and economical burial depth, • API 5L: Specification for Line Pipe
there is no short answer to that question. Ice gouging is a complex • API RP 2A: Planning, Designing and Constructing Fixed Offshore
phenomenon. It relies on a proper understanding of the mechanisms Platform
taking place in the soil around the pipe during gouging, of those • API RP 2N: Planning, Designing and Constructing structures and
affecting the ice feature, and of the pipeline behavior itself. Pipelines for Arctic API RP 14J: Recommended Practice for Design and
Hazards Analysis for Offshore Production Facilities
• API RP 17A: Recommended Practice for Design and Operation of Subsea
7.2. Filling in the knowledge gaps—pros and cons of various approaches Production Systems
• API RP 17C: Recommended Practice on TFL (Through Flowline) Systems
An adequate awareness of expected gouging activity in the area of • API RP 17D: Specification for Subsea Wellhead and Christmas Tree
interest is critical. This is why a significant amount of resources have Equipment
been spent on seabed mapping, to determine gouge frequency and • Conditions
geometry (notably depth and width). However, information of soil • API RP 1111: Design, Construction, Operation and Maintenance of
displacements below the gouge, which is what dictates design Offshore Hydrocarbon Pipelines
requirements, cannot be readily obtained from ship-borne instru- • ASCE: Guidelines for the Seismic Design of Oil and Gas Pipeline Systems,
mentation. Moreover, the nature of the gouging features is unknown. 1984
Physical modeling at a reduced scale, either at single gravity or in a • ASCE: Guidelines for the design of buried steel pipe, 2001
centrifuge, provides further insights on sub-gouge displacements. It is • ASME B31.8: Gas Transmission and Distribution Piping Systems
generally less expensive than field projects but the applicability of the • BS 8010: Code of Practice for Pipelines Part 3. Pipeline Subsea: Design,
data generated during these studies to real-scale scenarios needs to be Construction and Installation
carefully assessed. Theoretical studies provide an overall representa- • CSA S471: General Requirements, Design Criteria, The Environment
tion of the interaction event, and may be used to roughly validate and Loads
information obtained by other means. Numerical simulations are • CSA S472: Foundations
becoming increasingly powerful. They are not yet capable of • CSA S473: Steel Structures
reproducing all the intricacies of a gouging scenario (as physical • CSA S474: Concrete Structures
models do). But they can address all salient aspects of the ice–seabed • CSA Z245.1: Steel Line Pipe
interaction. Numerical simulations rely on quality data from other • CSA Z662: Oil and Gas Pipeline Systems
sources, against which they have to be carefully validated, but that are • DNV OS-F101: Submarine Pipeline Systems
not always available. • DNV RP F107: Risk Assessment of Pipeline Protection
• DNV RP-F109: On-Bottom Stability Design of Submarine Pipelines
• ISO 13623: Petroleum and natural gas industries—Pipeline transportation
7.3. What lies ahead
systems
• ISO 16708: Petroleum and Natural Gas Industries – Pipeline
Methodologies devised by various research groups capture some or
Transportation Systems – Reliability-Based Limit State Methods
most of the major components involved in a gouging scenario. These
• ISO 19900: General Requirements for Offshore Structures
methodologies are instructive in that they represent an integrated
• ISO 19906: Arctic Offshore Structures
approach to an improved understanding of gouging phenomena. They
point the way to what one may expect in terms of future guidelines to
a safe and cost-effective burial depth. Such guidelines will only be References
possible when all relevant information gaps are filled in: 1) adequate Abdalla, B., Jukes, P., Eltaher, A., Duron, B., 2008. The technical challenges of designing
knowledge of material properties, 2) a reliable handle on the processes oil and gas pipelines in the Arctic. OCEANS 2008 IEEE Proceedings, Quebec City,
Canada, pp. 1–11.
taking place during gouging and 3) a consensus on what constitutes Al-Khafaji, A.W., Andersland, O.B., 1992. Geotechnical Engineering and Soil Testing.
acceptable risks. Saunders College Publishing, N.Y. 695 pp.
Alekseev, Y.N., Afanasyev, V.P., Litonov, O.E., Mansurov, M.N., Truskov, P.A., 2001. Ice Gautier, D.L., Bird, K.J., Charpentier, R.R., Grantz, A., Houseknecht, David W., Klett, T.R.,
Technical Aspects of the Marine Oil and Gas Fields Exploration. St. Petersburg, Moore, T.E., Pitman, J.K., Schenk, C.J., Schuenemeyer, J.H., Sørensen, K., Tennyson, M.E.,
Russia. 280 pp. Valin, Z.C., Wandrey, C.J., 2009. Assessment of undiscovered oil and gas in the Arctic.
Allersma, H.G.B., Schoonbeek, I.S.S., 2005. Centrifuge modeling of scouring ice keels in clay. Science 324, 1175–1179.
Proceedings of the 15th International Offshore and Polar Engineering Conference. The Grass, J.D., 1984. Ice scour and ice ridging studies in Lake Erie. Proceedings of the 7th
International Society of Offshore and Polar Engineers (ISOPE), Seoul, pp. 404–409. International Symposium on Ice. Association of Hydraulic Engineering and Research
Altaee, A., Fellenius, B.H., 1994. Physical modeling in sand. Canadian Geotechnical (IAHR), Hamburg, pp. 221–236.
Journal 31, 420–431. Guo, B., Song, S., Chacko, J., Ghalambor, A., 2005. Offshore Pipelines. Elsevier, New York.
Annandale, G.W., 2006. Scour Technology: Mechanics and Engineering Practice. 281 pp.
McGraw-Hill, New York. 420 pp. Heinonen, J., Määttänen, M., 2001. Full-scale testing of ridge keel mechanical properties
Barker, A., Timco, G., 2002. Laboratory experiments of ice scour processes: rigid ice in Loleif-project. Proceedings of the 16th International Conference on Port and
indentor. Cold Regions Science and Technology 35, 195–206. Ocean Engineering under Arctic Conditions (POAC), Ottawa, pp. 1435–1444.
Barker, A., Timco, G., 2003. Laboratory experiments of ice scour processes: buoyant ice Héquette, A., Desrosiers, M., Barnes, P.W., 1995. Sea ice scouring on the inner shelf of
model. Cold Regions Science and Technology 36, 103–114. the southeastern Canadian Beaufort Sea. Marine Geology 128, 201–219.
Barrette, P.D., Jordaan, I.J., 2002. Healed cracks in iceberg ice. Journal of Glaciology 48 (163), Hodgson, G.J., Lever, J.H., Woodworth-Lynas, C.M.T., Lewis, C.F.M., 1988. The dynamics
587–589. of iceberg grounding and scouring (DIGS) experiment and repetitive mapping of
Barrette, P.D., Jordaan, I.J., 2003. Pressure–temperature effects on the compressive the eastern Canadian continental shelf. Environmental Studies Research Funds
behavior of laboratory-grown and iceberg ice. Cold Regions Science and Report No. 094. . (2 vols.), 315 pp.
Technology 36, 25–36. Howell, S.E.L., Yackel, J.J., 2004. A vessel transit assessment of sea ice variability in the
Barrette, P., Timco, G., 2008. Ice scouring in a large flume: test set-up and preliminary Western Arctic, 1969–2002: implications for ship navigation. Canadian Journal of
observations. Proceedings of the 8th International Conference on Ships and Marine Remote Sensing 30 (2), 205–215.
Structures in Cold Regions (ICETECH), Banff, pp. 1–8. Høyland, K.V., Liferov, P., Moslet, P.O., Løset, S., Bonnemaire, B., 2002. Medium scale
Barrette, P.D., Timco, G.W., Palmer, A., 2009. Sliding resistance of grounded ice on clay in modelling of ice ridge scouring of the seabed — part II: consolidation and physical
a laboratory: potential implications for offshore design. Cold Regions Science and properties. Proceedings of the 16th International Symposium on Ice. Association of
Technology 58, 1–14. Hydraulic Engineering and Research (IAHR), Dunedin, pp. 94–100.
Been, K., Kosar, K., Hachey, J., Rogers, B.T., Palmer, A.C., 1990. Ice scour models. Ishikawa, R., Kioka, S., Kubouchi, A., Saeki, H., 2005. Soil deformation within the seabed
Proceedings of the 9th International Conference on Offshore Mechanics and Arctic due to ice-scouring. Proceedings of the 20th International Symposium on Okhotsk
Engineering (OMAE). The American Society of Mechanical Engineers (ASME), Sea and Sea Ice, Mombetsu, Hokkaido, pp. 193–200.
Houston, pp. 179–188. Johnston, M., 2004. Properties of second-year and multi-year ice during freeze-up. CHC-NRC
Been, K., Sancio, R.B., Ahrabian, D., Deltares, W.V.K., Croasdale, K., Palmer, A., 2008. Technical Report, CHC-TR-024. 26 pp.
Subscour displacement in clays from physical tests. Proceedings of the 5th Johnston, M., Masterson, D., Wright, B., 2009. Multi-year ice thickness: knowns and
International Pipeline Conference (IPC). American Society of Mechanical Engineers unknowns. Proceedings of the 20th International Conference on Port and Ocean
(ASME), Calgary, Canada. Engineering under Arctic Conditions (POAC), Luleå, Sweden.
Blasco, S.M., Shearer, J.M., Myers, R., 1998. Seabed scouring by sea-ice: scouring process Jones, S.J., 2007. A review of the strength of iceberg and other freshwater ice and the
and impact rates: Canadian Beaufort Shelf. Proceedings of the 1st Ice Scour and effect of temperature. Cold Regions Science and Technology 47, 256–262.
Arctic Marine Pipelines Workshop, 13th International Symposium on Okhotsk Sea Jordaan, I.J., 2005. Decisions under Uncertainty: Probabilistic Analysis for Engineering
and Sea Ice. Mombetsu, Hokkaido, pp. 53–58. Decisions. Cambridge University Press. 672 pp.
Cathie, D.N., Jaeck, C., Ballard, J.-C., Wintgens, J.-F., 2005. Pipeline geotechnics—state-of- Kenny, S., Bruce, J., King, T., McKenna, R., Nobahar, A., Phillips, R., 2004. Probabilistic
the-art. In: Gourvenec, S., Cassidy, M. (Eds.), Frontiers in Offshore Geotechnics: design methodology to mitigate ice gouge hazards for offshore pipelines.
ISFOG 2005. Taylor & Francis Group, London, pp. 95–114. Proceedings of the 5th International Pipeline Conference (IPC). American Society
Chari, T.R., 1979. Geotechnical aspects of iceberg scours on ocean floors. Canadian of Mechanical Engineers (ASME), Calgary, Canada.
Geotechnical Journal 16, 379–390. Kenny, S., Phillips, R., Clark, J., Nobahar, A., Clark, J.I., 2005. PRISE numerical studies on
Chari, T.R., Peters, G.R., Muthukrishnaiah, K., 1980. Environmental factors affecting subgouge deformations and pipeline/soil interaction analysis. Proceedings of the
iceberg scour estimates. Cold Regions Science and Technology 1, 223–230. 18th International Conference on Port and Ocean Engineering under Arctic
Choi, K., Lee, J.-H., 2003. A comparative study of ice scour–seabed interaction models. Conditions (POAC), Potsdam, N.Y, pp. 43–52.
Proceedings of the 18th International Symposium on Okhotsk Sea and Sea Ice, Kenny, S., Palmer, A.C., Been, K., 2007. Design challenges for offshore pipelines in Arctic
Mombetsu, pp. 113–122. environments. Exploration and Production Oil & Gas Review 6 (2).
Clark, J.I., Zhu, F., 2000. Comparison of ice strength and scour resistance. Proceedings of Kharitonov, V.V., 2008. Internal structure of ice ridges and stamukhas based on thermal
the 2nd Ice Scour and Arctic Marine Pipelines Workshop, 15th International drilling data. Cold Regions Science and Technology 52 (3), 302–325.
Symposium on Okhotsk Sea and Sea Ice, Mombetsu, Hokkaido, pp. 141–149. King, T., Phillips, R., Barrett, J., Sonnichsen, G., 2009. Probabilistic pipeline burial analysis
Comiso, J.C., 2002. A rapidly declining perennial sea ice cover in the Arctic. Geophysical for protection against ice scour. Cold Regions Science and Technology 59 (1),
Research Letters 29 (20) 17-1. 58–64.
Croasdale, K.R., 2000. Study of iceberg scour & risk in the Grand Banks region. CHC-NRC King, T., 2011. Protection of pipelines from ice gouging. Journal of Pipeline Engineering
Technical Report, 31–26. 177 pp. 10 (2), 115–120.
Croasdale, K., Brown, R., Campbell, P., Crocker, G., Jordaan, I., McKenna, R., King, T., Kioka, S., Abe, S., Sasaki, H., Saeki, H., 1999. Mechanical model of ice scour. Proceedings
Myers, R., 2001a. Iceberg risk to seabed installations on the Grand Banks. of the 15th International Conference on Port and Ocean Engineering under Arctic
Proceedings of the 16th International Conference on Port and Ocean Engineering Conditions (POAC), Helsinki, pp. 537–544.
under Arctic Conditions (POAC), Ottawa, pp. 1019–1028. Kioka, S., Yasunaga, Y., Matuo, Y., Watanabe, Y., Saeki, H., 2001. Medium-scale model
Croasdale, K.R., Bruneau, S., Christian, D., Crocker, G., English, J., Metge, M., Ritch, R., tests on ice gouge. Proceedings of the 16th International Conference on Port and
2001b. In-situ measurements of the strength of first-year ice ridge keels. Ocean Engineering under Arctic Conditions (POAC), Ottawa, pp. 1029–1038.
Proceedings of the 16th International Conference on Port and Ocean Engineering Kioka, S., Kubouchi, A., Ishikawa, R., Saeki, H., 2004. Application of the mechanical model
under Arctic Conditions (POAC), Ottawa, pp. 1445–1454. for ice scour to a field site and simulation method of scour depths. Proceedings of the
Croasdale, K., Comfort, G., Been, K., 2005. Investigation of ice limits to ice gouging. 14th International Offshore and Polar Engineering Conference. The International
Proceedings of the 18th International Conference on Port and Ocean Engineering Society of Offshore and Polar Engineers (ISOPE), Toulon, pp. 891–898.
under Arctic Conditions (POAC), Potsdam, pp. 23–32. Konuk, I., Gracie, R., 2004. A 3-dimensional Eulerian finite element model for ice scour.
Das, B.M., 1994. Principles of Geotechnical Engineering. PWS Publishing Company, Proceedings of the 5th International Pipeline Conference (IPC). American Society of
Boston. 672 pp. Mechanical Engineers (ASME), Calgary, Canada.
Dean, E.T.R., 2010. Offshore Geotechnical Engineering — Principles and Practice. Konuk, I., Liferov, P., Løset, S., 2007. Challenges in modelling ice gouge and pipeline
Thomas Telford, Reston, VA, U.S.A. 520 pp. response. Proceedings of the 19th International Conference on Port and Ocean
Eden, D.J., Eyles, N., 2002. Case study of a relict iceberg scour exposed at Scarborough Engineering under Arctic Conditions (POAC). Dalian University of Technology
Bluffs, Toronto, Ontario: implications for pipeline engineering. Canadian Geotech- Press, Dalian, pp. 760–773.
nical Journal 39, 519–534. Konuk, I., Yu, S., 2007a. A pipeline case study for ice scour design. Proceedings of
Eik, K., Gudmestad, O.T., 2010. Iceberg management and impact on design of offshore the 26th International Conference on Offshore Mechanics and Arctic Engineer-
structures. Cold Regions Science and Technology 63, 15–28. ing (OMAE). The American Society of Mechanical Engineers (ASME), San Diego,
Eyles, N., Clark, B.M., 1988. Storm-influenced deltas and ice scouring in a late pp. 163–169.
Pleistocene glacial lake. Geological Society of America Bulletin 100, 793–809. Konuk, I., Yu, S., 2007b. Continuum FE modeling of lateral buckling: study of soil effects.
Foriero, A., 1998. Steady-state ice scouring of the seabed. Proceedings of the 1st Ice Proceedings of the 26th International Conference on Offshore Mechanics and Arctic
Scour and Arctic Marine Pipelines Workshop, 13th International Symposium on Engineering (OMAE). The American Society of Mechanical Engineers (ASME), San
Okhotsk Sea and Sea Ice, Mombetsu, Hokkaido, pp. 83–99. Diego, pp. 1–8.
Fredj, A., Comfort, G., Dinovitzer, A., 2008. A case study of high pressure/high Konuk, I., 2009. Arctic pipeline design challenges and current practices — ice scour.
temperature pipeline for ice scour design using 3D continuum modeling. Proceedings of the 28th International Conference on Ocean, Offshore and Arctic
Proceedings of the27th International Conference on Offshore Mechanics and Arctic Engineering (OMAE). The American Society of Mechanical Engineers (ASME),
Engineering. American Society of Mechanical Engineering (ASME), Estoril, Honolulu, Hawaii, pp. 155–162.
Portugal, pp. 563–572. Kubat, I., Gorman, R., Collins, A., Timco, G., 2006. Climate change impact on northern
Fuglem, M., Muggeridge, K., Jordaan, I., 1999. Design load calculations for iceberg impacts. shipping regulations. Proceedings of the 7th International Conference on Ships and
International Journal of Offshore and Polar Engineering 9 (4), 298–306. Marine Structures in Cold Regions (ICETECH), Banff, pp. 1–8.
Lanan, G.A., Ennis, J.O., 2001. Northstar offshore Arctic pipeline design and construction. Palmer, A., 2000b. Gouging in the context of critical issues of Arctic offshore pipeline
Proceedings of the 33rd Offshore Technology Conference. Offshore Technology development. Proceedings of the 2nd Ice Scour and Arctic Marine Pipelines
Conference (OTC), Houston, pp. 621–628. Workshop, 15th International Symposium on Okhotsk Sea and Sea Ice. Mombetsu,
Lanan, G.A., Cowin, T.G., Hazen, B., Maguire, D.H., Hall, J.D., Perry, C., 2008. Oooguruk Hokkaido, pp. 5–10.
offshore Arctic flowline design and construction. Offshore Technology Conference Palmer, A., 2001. Are we ready to construct submarine pipelines in the Arctic? Journal of
(OTC), pp. 1180–1192. Petroleum Technology 53 (4), 55–56.
Lanan, G.A., Cowin, T.G., Johnston, D.K., 2011. Alaskan Beaufort Sea pipeline design, Palmer, A.C., White, D.J., Baumgard, A.J., Bolton, M.D., Barefoot, A.J., Finch, M., Powell, T.,
installation and operation. Proceedings of the Offshore Technology Conference, Faranski, A.S., Baldry, J.A.S., 2003. Uplift resistance of buried submarine pipelines:
Houston. comparison between centrifuge modelling and full-scale tests. Geotechnique 53 (10),
Leidersdorf, C.B., Hearon, G.E., Hollar, R.C., Gadd, P.E., Sullivan, T.C., 2001. Ice gouge and 877–883.
strudel scour data for the Northstar pipelines. Proceedings of the 16th International Palmer, A., Niedoroda, A., 2005. Ice gouging and pipelines: unresolved questions.
Conference on Port and Ocean Engineering under Arctic Conditions (POAC), Proceedings of the 18th International Conference on Port and Ocean Engineering
Ottawa, pp. 145–154. under Arctic Conditions (POAC), Potsdam, pp. 11–22.
Lever, J., 1986. Iceberg dynamics of the DIGS experiment. In: Lewis, C.F.M., Parrott, D.R., Palmer, A.C., King, R.A., 2008. Subsea Pipeline Engineering. Pennwell, Tulsa, USA. 624 pp.
Simpkin, P.G., Buckley, J.T. (Eds.), Ice Scour and Seabed Engineering — Proceedings Palmer, A., Tjiawi, H., 2009. Reducing the cost of protecting pipelines against ice
of a Workshop on Ice Scour Research: Environmental Studies Revolving Funds, gouging. Proceedings of the 20th International Conference on Port and Ocean
Report No. 049, Calgary, Canada, pp. 138–142. Engineering under Arctic Conditions (POAC), Luleå, Sweden.
Lever, J.H., Bass, D.W., Lewis, C.F.M., Klein, K., Diemand, D., Dyke, M., 1991. Iceberg/seabed Paulin, M.J., Lach, P.R., Poorooshasb, F., Clark, J.I., 1991. Preliminary results of physical
interaction events observed during the DIGS experiment. Journal of Offshore model tests of ice scour. Proceedings of the 11th International Conference on Port
Mechanics and Arctic Engineering 113 (1), 74–87. and Ocean Engineering under Arctic Conditions (POAC), St. John's, pp. 1017–1036.
Lever, J.H., 2000. Assessment of Millennium Pipeline Project Lake Erie Crossing: Ice Phillips, R., Clark, J.I., Paulin, M.J., Meaney, R., Millan, D.E.L., Tuff, K., 1994. Canadian
Scour, Sediment Sampling, and Turbidity Modeling. Cold Regions Research and national centrifuge centre with cold regions capabilities. In: Leung, C.F., Lee, F.H.,
Engineering Laboratory (CRREL), Hanover, USA. ERDC/CRREL TR-00-13, 56 pp. Tan, T.S. (Eds.), Centrifuge 1994. A.A Balkema, Rotterdam, pp. 57–62.
Liferov, P., Løset, S., Moslet, P.O., Bonnemaire, B., Høyland, K.V., 2002. Medium scale Phillips, R., 1995. Centrifuge modelling: practical considerations. In: Taylor, R.N. (Ed.),
modelling of ice ridge scouring of the seabed — part I: experimental set-up and Geotechnical Centrifuge Technology, pp. 34–60. U.K., London.
basic results. Proceedings of the 16th International Symposium on Ice. Association Phillips, R., Clark, J.I., Kenny, S., 2005. PRISE studies on gouge forces and subgouge
of Hydraulic Engineering and Research (IAHR), Dunedin. deformations. Proceedings of the 18th International Conference on Port and Ocean
Liferov, P., Høyland, K.V., 2004. In-situ ice ridge scour tests: experimental set up and Engineering under Arctic Conditions (POAC), Potsdam, pp. 75–84.
basic results. Cold Regions Science and Technology 40, 97–100. Poplin, J.P., Wang, A.T., Leidersdorf, C.B., 2002. Ice gouge measurement strategy.
Liferov, P., Shkhinek, K.N., Vitali, L., Serre, N., 2007. Ice gouging study — actions and International Journal of Offshore and Polar Engineering 12 (1), 34–39.
effects. Proceedings of the 19th International Conference on Port and Ocean Randell, C., Morgan, V., Ralph, F., 2008. Protection and risk mitigation strategies for
Engineering under Arctic Conditions (POAC). Dalian University of Technology subsea infrastructure in ice environments. Proceedings of the Offshore Technology
Press, Dalian, pp. 774–786. Conference, Houston, pp. 658–666.
Løset, S., Shkhinek, K.N., Gudmestad, O.T., Høyland, K.V., 2006. Actions from Ice on Randolph, M.F., House, A.R., 2001. The complementary roles of physical and computational
Arctic Offshore and Coastal Structures. Krasnodar, St. Petersburg. 271 pp. modelling. International Journal of Physical Modelling in Geotechnics 1, 1–8.
Marcellus, R.W., Palmer, A.C., 1979. Shore crossing techniques for offshore pipelines in Rudkin, P., Young, C., Barron, P., Timco, G., 2005. Analysis and results of 30 years of
Arctic regions. Proceedings of the 5th International Conference on Port and Ocean iceberg management. Proceedings of the 18th International Conference on Port and
Engineering under Arctic Conditions (POAC), Trondheim, pp. 201–215. Ocean Engineering under Arctic Conditions (POAC), Potsdam, pp. 595–604.
Matskevitch, D.G., 1997. Eccentric impact of an ice feature: linearized model. Cold Sayed, M., Timco, G.W., 2009. A numerical model of iceberg scour. Cold Regions Science
Regions Science and Technology 25, 159–171. and Technology 55, 103–110.
McKenna, R., King, T., Brown, R., Bruce, J., Phillips, R., 2003. Ice scour risk to offshore Schoonbeek, I.S.S., Allersma, H.G.B., 2006. Centrifuge modelling of scouring ice keels in
pipelines — making the most of available data. Proceedings of the 17th clay. Proceedings of the 6th International Conference on Physical Modelling in
International Conference on Port and Ocean Engineering under Arctic Conditions Geotechnics (ICPMG), pp. 1291–1296.
(POAC), Trondheim, pp. 689–698. Schoonbeek, I.S.S., Xin, M.X., Van Kesteren, W.G.M., Been, K., 2006. Slip line field solutions
Mørk, K., 2007. The challenges facing Arctic pipelines — design principles for extreme as an approach to understand ice subgouge deformation patterns. Proceedings of the
conditions. Offshore 67 (9). 16th International Offshore and Polar Engineering Conference. The International
Morrow, D.R., Larkin, P.D., 2007. The challenges of pipeline burial. Proceedings of the Society of Offshore and Polar Engineers (ISOPE), San Francisco, pp. 628–633.
17th International Offshore and Polar Engineering Conference (ISOPE), Lisbon, Scott, R.F., 1988. Physical and numerical models. In: Craig, J., James, D.A., Schofield, A.
Portugal, pp. 900–907. (Eds.), Centrifuges in Soil Mechanics, pp. 103–118. The Netherlands, Rotterdam.
Muir Wood, D., 2004. Geotechnical Modelling. Spon Press, New York. 504 pp. Serré, N., 2011. Numerical modelling of ice ridge keel action on subsea structures. Cold
Newbury, B.D., Hukle, M.W., Crawford, M.D., Lwig, D.B., 2007. Welding engineering for Regions Science and Technology. doi:10.1016/j.coldregions.2011.02.011.
high strain pipelines. Proceedings of the 17th International Offshore and Polar Shapiro, L.H., Metzner, R.C., 1987. Coefficients of friction of sea ice on beach gravel.
Engineering Conference. The International Society of Offshore and Polar Engineers Journal of Offshore Mechanics and Arctic Engineering 109, 388–390.
(ISOPE), Lisbon, pp. 2996–3000. Sonnichsen, G.V., King, E.L., 2001. Surficial sediments, Grand Bank, Offshore
Nichols, G., 2009. Sedimentology and Stratigraphy. Wiley-Blackwell, Malden, MA, U.S.A.. Newfoundland. National Research Council of Canada, Ottawa. CHC-NRC Technical
432 pp. Report, 31-27, 30 pp.
Nixon, J.F., 1991. Thaw — subsidence effects on offshore pipelines. Journal of Cold Sonnichsen, G.V., King, T., Jordaan, I., Le, C., 2005. Probabilistic analysis of iceberg
Regions Engineering 5 (1), 28–39. scouring frequency based on repetitive seabed mapping, offshore Newfoundland
Nobahar, A., Kenny, S., King, T., McKenna, R., Phillips, R., 2007a. Analysis and design of and Labrador. Proceedings of the 18th International Conference on Port and Ocean
buried pipelines for ice gouging hazard: a probabilistic approach. Journal of Engineering under Arctic Conditions (POAC), Potsdam, N.Y, pp. 85–94.
Offshore Mechanics and Arctic Engineering 129, 129–228. Stava, I., Nystrøm, P.R., Vikse, N., Gudmestad, O.T., Liferov, P., Grønli, J., 2008. Small scale
Nobahar, A., Kenny, S., Phillips, R., 2007b. Buried pipelines subject to subgouge model tests of ice gouge in soft sandy silt. Proceedings of the 27th International
deformations. International Journal of Geomechanics 7 (3), 206–216. Conference on Offshore Mechanics and Arctic Engineering (OMAE). The American
Noble, P.G., Comfort, G., 1982. Damage to an underwater pipeline by ice ridges. In: Society of Mechanical Engineers (ASME), Estoril, pp. 1–10.
Frederking, R.M.W., Pilkington, G.R. (Eds.), Proceedings of Workshop on Sea Ice Sumer, B.M., Fredsøe, J., 2002. The mechanics of scour in the marine environment. Advanced
Ridging and Pile-up, Snow and Ice Subcommittee, Associate Committee on Series on Ocean Engineering, 17. World Scientific, Hackensack, N.J., U.S.A. 550 pp.
Geotechnical Research, National Research Council Canada, Technical Memorandum Thomas, D.N., Dieckmann, G.S., 2003. Sea ice — An Introduction to its Physics,
No.134, pp. 248–284. Chemistry. Blackwell Science, Biology and Geology. 355 pp.
Nogueira, A.C., Paulin, M., 1999. Limit state design for Northstar offshore pipeline. Timco, G.W., Burden, R.P., 1997. An analysis of the shapes of sea ice ridges. Cold Regions
Offshore Magazine 59 (9), 146–196. Science and Technology 25, 65–77.
Novak, P., Guinot, V., Jeffrey, A., Reeve, D.E., 2010. Hydraulic Modelling — An Introduction. Timco, G.W., Croasdale, K., Wright, B., 2000. An overview of first-year sea ice ridges.
Spon Press, New York. 616 pp. CHC-NRC Technical Report, HYD-TR-047. 157 pp.
Oickle, E.J., Blasco, S.M., Shearer, J.M., 2006. Extreme ice-scouring processes on the Timco, G.W., Weeks, W.F., 2010. A review of the engineering properties of sea ice. Cold
Canadian Beaufort Shelf caused by sea-ice pressure ridge keels. Atlantic Geology 42 Regions Science and Technology 60, 107–129.
(1), 104. Vella, D., Wettlaufer, J.S., 2008. Explaining the patterns formed by ice floe interactions.
Palmer, A.C., Brown, R.J., Baudais, D.J., Masterson, D.M., 1979. Design and installation of Journal of Geophysical Research 113, C11011.
an offshore flowline for the Canadian Arctic Islands. Proceedings of the 11th Vershinin, S.A., Truskov, P.A., Liferov, P.A., 2008. Ice Action on Seabed and Subsea
Offshore Technology Conference. Offshore Technology Conference (OTC), Houston, Structures. PPC Russkaya kniga, Moscow, Russia. 224 pp.
pp. 765–772. Vikse, N., Gudmestad, O.T., Nystrøm, P.R., Liferov, P., 2007. Small scale model tests on
Palmer, A., 1997a. Pipeline codes or pipeline cookbooks? Journal of Pipeline Engineering 6, subgouge soil deformations. Proceedings of the 26th International Conference on
69–74. Offshore Mechanics and Arctic Engineering (OMAE). The American Society of
Palmer, A., 1997b. Geotechnical evidence of ice scour as a guide to pipeline burial depth. Mechanical Engineers (ASME), San Diego, pp. 1–6.
Canadian Geotechnical Journal 34, 1002–1003. Weeks, W.F., 2010. On Sea Ice. University of Alaska Press, Fairbanks. 664 pp.
Palmer, A., 2000a. Are we ready to construct submarine pipelines in the Arctic? Proceedings Woodworth-Lynas, C.M.T., Guigné, J.Y., 1990. Iceberg scours in the geological record:
of the 32nd Offshore Technology Conference. Offshore Technology Conference (OTC), examples from glacial Lake Agassiz. Geological Society Special Publication 53,
Houston, pp. 737–744. 217–233.
Woodworth-Lynas, C.M.T., Josenhans, H.W., Barrie, J.V., Lewis, C.F.M., Parrott, D.R., 1991. Yang, Q.S., Poorooshasb, H.B., Lach, P.R., 1996. Centrifuge modeling and numerical
The physical processes of seabed disturbance during iceberg grounding and simulation of ice scour. Soils and Foundations 36 (1), 85–96.
scouring. Continental Shelf Research 11 (8–10), 939–961. Younan, A.H., Hamilton, J.M., Weaver, J., 2007. Ice Gouge Reliability of Offshore
Woodworth-Lynas, C., Nixon, D., Phillips, R., Palmer, A., 1996. Subgouge deformations and Arctic Pipelines. The American Society of Mechanical Engineers (ASME), San
the security of Arctic marine pipelines. Proceedings of the 28th Offshore Technology Diego, pp. 1–7.
Conference. Offshore Technology Conference (OTC), Houston, pp. 657–664. Zhou, C., Li, G., Dong, P., Shi, J., Xu, J., 2010. An experimental study of seabed responses
Woodworth-Lynas, C.M.T., Phillips, R., Clark, J.I., Hynes, F., Xiao, X., 1998. Verification of around a marine pipeline under wave and current conditions. Ocean Engineering.
centrifuge model results against field data: results from the pressure ridge ice scour doi:10.1016/j.oceaneng.2010.10.006.
experiment. Proceedings of the 1st Ice Scour and Arctic Marine Pipelines Zufelt, J.E., Ettema, R., 1996. Model ice properties. Cold Regions and Research Engineering
Workshop, 13th International Symposium on Okhotsk Sea and Sea Ice, Mombetsu, (CRREL), 96-1, 19 pp.
Hokkaido, pp. 123–138.

Double Pipe Wall

Uploaded by

Copyright:

Available Formats

Double Pipe Wall

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Double Pipe Wall

Uploaded by

Copyright:

Available Formats

Cold Regions Science and Technology 69 (2011) 3–20

Contents lists available at ScienceDirect

Cold Regions Science and Technology

Offshore pipeline protection against seabed gouging by ice: An overview

⁎ Tel.: + 1 613 993 6809.

3.1.5. Some numbers on gouges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

1. Introduction against ice gouging—the most cost-effective approach is to bury

2.1.6. Examples of subsea pipelines in cold waters

it is recommended that pipeline burial depth take into account depth

4. Pipeline protection against ice gouging

5.1. Field studies 5.1.2. In situ characterization of real ice features

Simply put, a structure is expected to fail if a demand D exceeds

You might also like