Sname (2008) PDF

T echnical
&
R esearch Bulletin 5-5A
Guidelines for
Site Specific
Assessment of
Mobile Jack-Up Units
August 2008
This Bulletin contains two documents relating to the site-specific assessment of

mobile jack-up units. These documents have been produced by a joint industry-fund-
ed project (JIP). They have been published by SNAME for the benefit of naval archi-
tects and marine engineers involved with the world-wide assessment of mobile jack-
up units.
The first document is T&R 5-5 and is referred to as a "Guideline". This document
describes a general approach to site assessment which should be applied.
The second document is T&R 5-5A and is referred to as a "Recommended Practice".

This is the title selected by the JIP participants. However, T&R 5-5A must not be
construed as a recommended practice proposed by SNAME. T&R 5-5A describes
one example of a methodology that can be followed to achieve the intent of the
"Guideline".
The Society of Naval Architects

and Marine Engineers
601 Pavonia Avenue, Jersey City, New Jersey 07306
THE SOCIETY OF NAVAL ARCHITECTS AND MARINE ENGINEERS
TECHNICAL AND RESEARCH PRORAM
Technical & Research Bulletin 5-5A
This volume contains the following documents:
• T&R 5-5 – Guideline for Site Specific Assessment of Mobile Jack-Up

Units – (First Edition – May 1994)
• T&R 5-5A – Recommended Practice for Site Specific Assessment of
Mobile Jack-Up Units – (First Edition – Rev.3, August 2008)
• Commentaries to Recommended Practice for Site Specific Assessment of
Mobile Jack-Up Units – (First Edition – Rev.3, August 2008)
August 2008
GUIDELINE
FOR SITE SPECIFIC ASSESSMENT
OF MOBILE JACK-UP UNITS
FIRST EDITION – MAY 1994

Preparation of this Bulletin has been monitored by
PANEL OC-7
(SITE ASSESSMENT OF JACK-UP RIGS)
OF
THE SOCIETY OF NAVAL ARCHITECTS AND MARINE ENGINEERS

TECHNICAL AND RESEARCH PRORAM
WILLIAM T. BENNETT, CHAIRMAN
Reviewed and Approved

by
OFFSHORE COMMITTEE
William P. Stewart, Chairman
William T. Bennett Peter G. Noble

Jeremiah Daniel John A. Pritzlaff
David B. Lorenz William J. Sember
Jack Y. K. Lou N. Pharr Smith
James M. Magill Y. S. David Tein
Scott C. McClure David P. Tuturea
John A. Mercier
Philip B. Kimball
Executive Director
Guideline for Site Specific Assessment of Mobile Jack-Up Units
ACKNOWLEDGEMENTS
This GUIDELINE has been drafted by the Working Group of the Joint Industry Sponsored
project "Jack-Up Site Assessment Procedures - Establishment of an International Technical
Guideline". Technical and administrative management has been provided by Noble Denton
Consultancy Services Limited. Funding was provided by the Working Group members and the
other Participants in the study listed below:
The Working Group
Friede & Goldman W.T. Bennett (Chair)

Noble Denton M.J.R. Hoyle (Technical Secretary)
ABS D.E. Jones
Amoco C.F. Cowgill
BP R.O. Snell
Bureau Veritas J.L. Isnard
DnV P.E. Bergrem
Exxon C.R. Brinkman
Lloyds Register W.J. Winkworth
Maersk Drilling G. Kudsk
Marathon Le Tourneau C.A. Wendenburg
MSC C.J. Mommaas
Noble Denton J.C. Trickey/B.P.M. Sharples
Reading & Bates R.W. Mowell
Santa Fe C.N. Springett
Sedco-Forex J. Chevallier
Shell P.G.F. Sliggers
Other Participants
AGIP S.P.A.
ARCO Oil & Gas Co
CFEM
COGLA
Department of Energy (UK)
Elf
Enterprise Oil
Far East Levingston Shipbuilding
IADC (sponsoring members)
Maersk Olie Og Gas
Mobil
Norwegian Maritime Directorate
Norwegian Petroleum Directorate
Phillips Petroleum
Statoil
Technip Geoproduction
Texaco
The opinions or assertions of the authors herein are not to be construed as official or reflecting
the views of SNAME or any government agency.
It is understood and agreed that nothing expressed herein is intended or shall be construed to give
any person, firm or corporation any right. Remedy, or claim against SNAME or any of its
officers or members.
© Copyright 2008 by The Society of Naval Architects and Marine Engineers
Correspondence should be addressed to:
Joseph H. Rousseau
Technical Secretary to SNAME OC-7 panel
ABS Americas
16855 Northchase Drive
Houston, Texas 77060
Fax: (281) 877-6796

Phone: (281) 877-6626
TABLE OF CONTENTS
PAGE
1 INTRODUCTION
1.1 General 1
1.2 Reference Document 1
1.3 Applicability and Limitations 1
1.4 Typical Approach to Site Assessment 2
2 DATA TO BE ASSEMBLED FOR EACH SITE LOCATION
2.1 Rig Data 3

2.2 Site Data 3
2.3 Environmental Data 3
2.4 Geotechnical Data 4
3 CONFIGURATIONS
3.1 Mode of Operation 6

3.2 Airgap 6
3.3 Leg Length Reserve 6
4 LOADINGS
4.1 Loading Cases 7

4.2 Wave and Current Forces 8
4.3 Wind Forces 8
4.4 Reaction Point and Foundation Fixity 8
4.5 Storm Approach Angle 9
4.6 Centre of Gravity 9
4.7 Displacement Dependent Loads 9
4.8 Dynamic Effects 9
4.9 Other Loads 10
5 RESISTANCE
5.1 General 11
5.2 Overturning Stability 11
5.3 Foundation and Preload 11
5.4 Structural Integrity 12
5.5 Adjacent Structures 12
5.6 Other 12
GUIDELINE FOR SITE SPECIFIC ASSESSMENT

1 INTRODUCTION
1.1 General
1.1.1 This document is a GUIDELINE for the site specific structural and foundation
assessment of jack-up units. The purpose of this GUIDELINE is to identify the
factors which are likely to be the main concerns for any site assessment of a jack-
up unit. It is not to be interpreted as guidance for design or construction as there
are existing rules and regulations, both by Classification Societies and
Governmental Agencies, covering these aspects.
1.1.2 This GUIDELINE has been developed by representatives of all parts of the jack-
up industry working in a Joint Industry Project. It is intended to serve as a basic
standard and to provide a common reference when comparing the work of
different assessors. The user is advised to take due account of any Regulatory
requirements that may apply to the particular geographic area of operation.
1.2 Reference Document
1.2.1 The accompanying document entitled "RECOMMENDED PRACTICE FOR

SITE SPECIFIC ASSESSMENT OF MOBILE JACK-UP UNITS" (hereafter
referred to as the RECOMMENDED PRACTICE) provides further guidance and
recommendations on the procedures and criteria for site specific assessment. It
may be revised to account for technical developments.
1.3 Applicability and Limitations
1.3.1 An assessment should be made of the jack-up for each site location. This
GUIDELINE relates only to the assessment of the jack-up in the elevated
condition. Transportation to and from the site and moving on and moving off
location are not covered in this document.
1.3.2 Guidance on the Safety Factors that may be adopted is given in the
RECOMMENDED PRACTICE, however an owner, insurer, operator, etc., may
justify different factors in particular circumstances.
1
1.3.3 This GUIDELINE will apply to most jack-ups. It is recognized that there may be
designs and/or circumstances when certain provisions may not apply. Such
instances shall be reviewed on a case by case basis.
1.3.4 It is assumed that the jack-up is built to recognized standards, and has been
maintained as required to continue to meet those standards. Any deterioration of
the jack-up should be taken into account in the fitness for purpose site specific
assessment.
1.4 Typical Approach to Site Assessment
1.4.1 Where a jack-up is to be employed in conditions well within its design capacity
and existing calculations in accordance with the RECOMMENDED PRACTICE
are available, the site specific assessment may be undertaken by appropriate
comparisons between the parameters used in the calculations and those applicable
to the new location. Otherwise, engineering calculations of various degrees of
complexity are required to justify that the jack-up can be safely used at the
location.
2
2 DATA TO BE ASSEMBLED FOR EACH SITE LOCATION
2.1 Rig Data
2.1.1 The specific rig information that is required to perform a site specific assessment
may include:
- Rig type,
- Pertinent drawings and specifications,
- Weights,
- Preloading capability,
- Materials,
- Design parameters, any proposed deviations for the intended operation, and
- Details of any relevant modifications.
2.1.2 When modelling the jack-up structure, it is necessary to consider the

leg/guide/elevating/holding system in some detail. Because of the complexity of
modelling these structures it is often appropriate that the designer's advice is
obtained. Typically, the information required will include upper and lower guide
flexibility, stiffness of the elevating/holding system and any special details
regarding its interaction with the leg.
2.2 Site Data
2.2.1 The site data should include the location coordinates, seabed topography and
water depth referenced to a clearly specified datum (e.g., Lowest Astronomical
Tide (LAT) or Chart Datum (CD)). Note that charts derived for use by
comparatively shallow draft shipping are often not sufficiently accurate for siting
jack-ups.
2.3 Environmental Data
2.3.1 It is of prime importance to obtain appropriate wind, wave and current data for the
site evaluation with due recognition of the quality of the data. Other data are to
be evaluated as applicable e.g., tides, rate of marine growth, ice, sea and air
temperature, earthquake, etc.
3
2.3.2 It is recommended that the 50 year return extremes be used for the site specific
assessment of manned jack-ups. In some cases longer return periods may be
preferred or required. In other cases a smaller return period may be justified. For
example, if the jack-up is unmanned or can be demanned readily and the risk of
environmental pollution is considered to be sufficiently low, the use of a 10 year
return period storm is recommended. In such cases the availability and
effectiveness of early warning systems and the adequacy of the evacuation plans
should be assessed.
2.3.3 If the jack-up deployment is to be of limited duration, applicable seasonal data

may be used (for example, the 50 year return period summer storm).
2.3.4 Directionality of wind, wave and current may be considered if accurate data is
available. Where there is sufficient evidence that any of the environmental forces
at the site are directional, it may be possible to orientate the jack-up on the most
advantageous heading.
2.3.5 The extreme wind, wave and current in some parts of the world (e.g., North Sea)
are specified by the Regulatory Authorities for the geographic area. It should be
noted that these data may not take account of local variations. Improved
environmental data are continually becoming available due to the use of better
hindcasting techniques and more reliable measured data. Consequently, a specific
meteorological assessment of the site is desirable, especially if the jack-up is
loaded to near its design limitation.
2.3.6 In some areas of the world there are no adequate data, or there are variations in
existing published data. For such cases there is a need for site specific studies to
establish the meteorological criteria.
2.4 Geotechnical Data
2.4.1 Site specific geotechnical information must be obtained. The type and amount of
geotechnical data required will depend on the particular circumstances such as the
type of jack-up and previous experience of the site, or nearby sites, for which the
assessment is being performed. Such information may include shallow seismic
survey, coring data, cone-penetrometer tests, side-scan sonar, magnetometer
survey and diver's survey.
2.4.2 The site should be evaluated for the presence of shallow gas deposits.
4
2.4.3 For sites where previous operations have been performed by jack-ups of the same
basic design, it may be sufficient to identify the location of, and hazards
associated with, existing footprints and refer to previous site data and preloading
records; however, it is recommended that the accuracy of such information be
verified.
2.4.4 At sites where there is any uncertainty, corings and/or cone penetrometer tests
(CPT) data are recommended. Alternatively the site may be tied-in to such data at
another site by means of shallow seismic data. If data are not available prior to
the arrival of the jack-up on location at the site, it may be possible to take
coring(s), etc. from the jack-up before preloading and jacking to full airgap.
Suitable precautions should be taken to ensure the safety of the unit during this
initial period on location.
2.4.5 The site should be evaluated for potential scour problems. These are most likely
to occur at sites with a firm seabed composed of non-cohesive soils where the
penetration will be minimal.
2.4.6 Certain locations prone to mudslides may involve the acceptance of additional
risks. Such risks should be assessed by carrying out specialist studies of the area
on which the jack-up is to be sited.
5
3. CONFIGURATIONS
3.1 Mode of Operation
3.1.1 This GUIDELINE applies to the elevated condition. The jack-up may be used in
alternative modes at one location, for example normal operating or elevated storm
mode, tender mode or cantilever drilling/workover mode. Where more than one
mode of operation is contemplated it may be important in the site assessment to
investigate the differences in these modes (e.g., the varying airgaps required for
each) as well as the operations necessary to change from one mode to another
(e.g., skidding the cantilever in for a storm). The practicality of any required
mode change should be evaluated, and appropriate assumptions incorporated into
the site assessment calculations. Any restrictions on the operations must be
included in the operating procedures.
3.2 Airgap
3.2.1 The Airgap is defined as the distance between the underside of the hull and the
lowest astronomical tide (LAT) during operations. It is usually not practical to
change the airgap in preparation for a storm, and therefore the minimum Elevated
Storm airgap for an intended operation should be calculated based upon a suitable
return period storm. It is recommended that this return period should not be less
than 50 years, even if a lower return period is used for other purposes.
3.2.2 The jack-up may be required to operate over a fixed platform or some other
obstruction which may dictate a larger airgap. This larger airgap should be used
for the site assessment.
3.3 Leg Length Reserve
3.3.1 It is recommended that a reserve above the upper guides of 1.5 meter of leg length
or one jack stoke on hydraulic units is allowed to account for any settlement, and
to provide a contingency in case the actual penetration exceeds that predicted. A
larger reserve may be required due to the strength limitations of the top bay of the
leg or leg/hull interface considerations.
6
4. LOADINGS
4.1 Loading Cases
4.1.1 A more detailed discussion of the various loadings that must be considered for
site assessments can be found in the RECOMMENDED PRACTICE. The
following outlines the loadings to be considered in general terms:
1) Environmental Loads
a) Loading due to the extreme storm one (1) minute mean wind on hull
and exposed areas (e.g., legs) as applicable, plus
b) Loading due to extreme wave and current on legs and other submerged
structure, plus
2) Functional Loads
a) Dead loads, plus
b) Applicable live loads, plus
3) Response Effects
a) Displacement dependent loads, plus
b) Dynamic effects.
4) Other Loads
4.1.2 Wind, wave and current are typically considered to act simultaneously and from
the same direction. Directionality of wind, wave and current may be applied
when it can be demonstrated that such directionality persists for the site under
consideration.
4.1.3 For dead loads it is typical to consider the jack-up in the fully loaded condition for
structural checks and with the minimum anticipated variable load (often 50%) for
the overturning calculation. If the assessment of the jack-up shows it is marginal
in one of these conditions, consideration may be given to limiting the variable
load to a lower or higher level (depending on the critical parameter), providing the
jack-up can be successfully operated under such restrictions. Any restrictions on
the variable loads should be incorporated in the operating procedures and the rig
personnel should be properly briefed.
7
4.2 Wave and Current Forces
4.2.1 Wave and current forces on the legs and appurtenances (e.g., raw water tower)
should be computed using the Morison equation. A wave theory appropriate to
the wave height, period and water depth should be used for the determination of
particle kinematics. The derived loadings are directly affected by the current
profile chosen and the method used to modify the profile in the presence of
waves.
4.2.2 Drag and inertia coefficients valid for the flow regime and chosen wave theory
should be selected. Applicable test results may be used to select the coefficients.
The effects of raw water piping, ladders and other appendages should be
considered in the calculation of the force coefficients for the legs.
4.2.3 The effect of marine growth on the hydrodynamic loading should be considered.
Because jack-ups are mobile, opportunities are available to clean the leg if
required. Where existing marine growth is not to be cleaned or where the
operation is to last long enough for significant growth to occur the influence of
growth on the leg hydrodynamic properties should be considered. It may also be
important to consider the timing of the jack-up deployment in relation to the
marine growth season.
4.3 Wind Forces
4.3.1 Wind forces should be computed using the one (1) minute mean wind velocity
and appropriate formulae and coefficients or should be derived from applicable
wind tunnel tests. Wind forces on legs can be a dominant factor for jack-ups
operating at less than their maximum design water depth. Generally, for site
assessments, block areas are used for the hull and appendages. The maximum
value may be used for all headings or alternatively directionality may be
considered.
4.4 Reaction Point and Foundation Fixity
4.4.1 The selected reaction point at the spudcan should be specified clearly in the site
assessment. The selection of the reaction point should be based on the estimated
penetration using the geotechnical information from the site.
4.4.2 The jack-up's legs will normally be assumed to be pinned at the reaction point.
Any divergence from this assumption should be clearly stated together with the
assumptions for any moment fixity provided to the leg's cans by the soil.
4.4.3 For mat supported jack-ups the reaction is typically considered to act at the
underside of the mat bottom plating.
8
4.5 Storm Approach Angle
4.5.1 The critical storm approach angles relative to the jack-up are usually different for
the various checks that are made (e.g., strength vs. overturning checks). The
critical direction for each check should be used as appropriate.
4.6 Centre of Gravity
4.6.1 The location of the cantilever, substructure, and other significant weights should
be considered. If these differ for the Operating Condition and the Elevated Storm
survival condition, the practicality of making the changes required to achieve the
Elevated Storm survival condition should be established.
4.7 Displacement Dependent Loads
4.7.1 Loading effects that are a consequence of the displacement of the structure should
be considered in the analysis. These effects are due to the first order sway (P-
delta), and its enhancement due to the increased flexibility of the legs in the
presence of axial loads (Euler amplification).
4.8 Dynamic Effects
4.8.1 The following principles are outlined to provide an understanding of the dynamic
behaviour of a jack-up:
a) The structure will vibrate at its natural period if excited by the forces of
ocean waves.
b) The magnitude of this vibration is primarily influenced by the amount of

wave energy at or near the natural period of the structure and to a lesser
extent by the wave energy at other periods.
c) Such vibrations induce inertial loads which are the product of the mass and
acceleration of the system.
d) The total load on the system is the combination of static and inertial
components. The direct calculation and application of the inertial loads is
preferable to the application of a Dynamic Amplification Factor.
4.8.2 The dynamic response of the jack-up should always be considered, and is
particularly important for seastates having significant energy near the natural
period of the jack-up or multiples thereof.
9
4.9 Other Loads
4.9.1 In certain areas there may exist Regulatory requirements to investigate other types
of load. Examples may include:
- Boat impact,
- Earthquakes,
- Ice loading, and
- Blast over-pressure due to an explosion on an adjacent structure.
It is seldom necessary to consider such loads in combination with the maximum

environmental or functional loading.
10
5. RESISTANCE
5.1 General
The key resistance checks include:
- Overturning stability,
- Foundations and preload, and
- Structural integrity.
5.2 Overturning Stability
5.2.2 The check on the factor of safety on overturning is intended to ensure against
uplift of the windward leg(s). With an appropriate factor of safety this check may
additionally serve to ensure that there is adequate vertical load on the windward
legs to prevent the sliding of footings with small penetrations. Such sliding may
cause load redistribution and possible progressive collapse. For further details,
and applicable factors of safety, refer to the RECOMMENDED PRACTICE.
Foundation fixity should only be included in the evaluation of the overturning
factor of safety when an applicable and detailed foundation study has been made.
5.3 Foundation and Preload
5.3.1 The purpose of preloading is to develop adequate foundation capacity to resist the
extreme vertical and horizontal loadings. The jack-up should normally be capable
of preloading to exceed the maximum vertical soil loadings associated with the
assessment storm. Where there is insufficient preload capacity to meet the
extreme loadings, a lower preload may be acceptable when justified by
appropriate geotechnical calculations.
5.3.2 If the penetrations obtained are significantly different to those predicted in the site
evaluation, further investigation should be undertaken to determine the reasons
(e.g., the jack-up's location may differ from that evaluated or local anomalies may
exist below the spudcans) before proceeding to full airgap.
5.3.3 Certain jack-ups are more sensitive than others to the effects of rapid leg
penetration when going on location. The structural behaviour of jack-ups under
such conditions varies considerably between different designs. Hence some jack-
ups are more sensitive to lateral loads and are more susceptible to local damage.
It is therefore important that an in-depth understanding of the behaviour of the
subject jack-up is obtained if there is a risk of rapid leg penetration (punch-
through).
11
5.3.4 It is necessary to check the resistance to sliding. Such checks are likely to be
most critical when considering mat supported jack-ups. The windward leg(s) of
independent leg jack-ups which are subject to shallow penetrations are also likely
to be critical.
5.4 Structural Integrity
5.4.1 Strength
Checks are required to ensure that the strength complies with the acceptance
criteria. It may be possible to compare the critical leg loadings to existing
calculations in accordance with the RECOMMENDED PRACTICE. Foundation
fixity should only be included in the evaluation of the upper leg when an
applicable and detailed foundation study has been made. Where foundation fixity
may exist, the lower parts of the leg should be checked assuming an upper bound
fixity value.
Areas which are often critical on jack-up rigs are the legs at the lower guides, the
legs between guides, the lower guides, the pinions and/or rack teeth, the chocks
and/or chock supports (if chocks are fitted) and the leg to can or mat connection.
Any strength limitations with respect to maximum or minimum penetrations

and/or bearing area or amount of foundation fixity should be related to the
geotechnical information for the specific site.
5.4.2 Fatigue
Fatigue should be considered. This does not imply that a detailed assessment or
analysis will normally be required.
5.5 Adjacent Structures
5.5.1 It may be necessary to consider the interaction of the jack-up with any adjacent
structures. Possible topics to be included in the site specific assessment are the
effects of the jack-up's spudcans on the foundation of the adjacent structure and
the effects of relative motions on the drill-string, well conductor, well conductor
guides, etc.
5.6 Other
5.6.1 The assessor should be aware that there may be other characteristics and/or
peculiarities of certain designs that will impact the site specific assessment.
Additionally there may be characteristics which vary within a design class that
should be considered.
12
This page intentionally left blank.
RECOMMENDED PRACTICE

(REVISION 3 – AUGUST 2008)
Rev Issue Date Details

Rev 1 May 1997 Changes made to pages 3, 9, 15, 19, 20, 24, 36, 42, 61, 65, 66, 67, 73-77, 87, 88, 90,
95, 97, 98, 99, 104, 108, 115, 118, 120, 125, 126, 127, 133 and 136
Revised areas indicated by sidelines thus:
Rev 2 Jan 2002 Changes made to pages.
1, 76, 77, 78, 79, 80, 104, 110, 114, 123, 126, 136, 137, 141
Revised areas indicated by double sidelines thus:
Rev 3 Aug 2008 Changes made to pages 1, 2, 9, 15, 52, 75, 76, 79, 89, 99, 105, 105, 123, 110, 114,
121, 129-134, 136-138
Revised areas indicated by triple sidelines thus:
Recommended Practice for Site Specific Assessment of Mobile Jack-Up Units Page 2
Rev 3, August 2008
This document evolved through a joint industry project (JIP) sponsored by a large number of
companies who are listed on the next page. Technical and administrative management of the
project was provided by Noble Denton Consultancy Services Ltd. This document has not been
produced by SNAME although some SNAME members have participated in its production.
SNAME has, at the request of the working group for the JIP, published this document so that it
may be widely disseminated in industry. However, SNAME takes no responsibility for any of
the technical or other contents of this document. SNAME cannot provide any technical or other
support for this document. For naval architects, engineers, or any other persons using this
document, technical support is available on a fee-paying basis from American Bureau of
Shipping. The contact at American Bureau of Shipping at the time of publication of this
document is:
Mr. Joseph H. Rousseau

Technical Secretary to SNAME OC-7 panel
ABS Americas
16855 Northchase Drive
Houston, Texas 77060-6008
USA
FAX: (281) 877-6796

Phone: (281) 877-6626
Although this document is entitled "Recommended Practice for Site Specific Assessment of
Mobile Jack-up Units," it must not be construed as a recommended practice proposed by
SNAME.
© Copyright 2008 by The Society of Naval Architects and Marine Engineers

Rev 3, August 2008
ACKNOWLEDGMENTS
This RECOMMENDED PRACTICE was drafted by the Working Group of the Joint Industry
Sponsored project "Jack-Up Site Assessment Procedures - Establishment of an International
Technical Guideline". Technical and administrative management was provided by Noble
Denton Consultancy Services Limited. Funding was provided by the Working Group members
and the other Participants in the study listed below:
The Working Group

Friede & Goldman W.T. Bennett (Chair)
Noble Denton M.J.R. Hoyle (Technical Secretary)
ABS D.E. Jones
Amoco W.O. McCarron
BP R.O. Snell
Bureau Veritas L. Leblanc
DnV P.E. Bergrem
LeTourneau, Inc. (Marine Div.) J.F. Bowes (previously N.P. Smith)
Lloyds Register W.J. Winkworth
Maersk Drilling G. Kudsk
MSC C.J. Mommaas
Noble Denton J.C. Trickey/B.P.M. Sharples
Reading & Bates R.W. Mowell
Santa Fe C.N. Springett
Sedco-Forex J.P. Cahuzac
Shell P.G.F. Sliggers
U.K. Health & Safety Executive M. Birkinshaw/D. Smith/W.J. Supple
Other Participants
Phase 2 only:
AGIP S.P.A.
Norwegian Maritime Directorate
Phases 2 and 3: IADC sponsoring members:

ARCO Oil & Gas Co. Arethusa/Zapata Off-Shore Co.
Eiffel (UK) (ex CFEM) Chiles Offshore Corp.
National Energy Board - Canada (ex COGLA) Forasol
Elf Global Marine Drilling Co.
Enterprise Oil Japan Drilling Co.
Far East Levingston Shipbuilding Lauritzen Offshore
IADC (sponsoring members) Maersk Drilling
Maersk Olie Og Gas Neddrill Nederland B.V.
Mobil Noble Drilling International
Norwegian Petroleum Directorate Penrod Drilling Corp.
Phillips Petroleum Company Reading & Bates Drilling Co.
Statoil Rowan Companies, Inc.
Technip Geoproduction Saipem S.p.a
Texaco Santa Fe International
Sedco Forex Drilling Services
Smedvig A.S.
Also contributing:
KCA Drilling Group
Nabors Drilling Internat'l
Sonat Offshore Drilling Inc.
Rev 3, August 2008
CONTENTS
SECTION TITLE PAGE NO
1 INTRODUCTION 9
2 OBJECTIVES 10
3 ASSESSMENT INPUT DATA 13

3.1 Rig data
3.2 Functional Loadings
3.3 Environmental Conditions - General
3.4 Wind
3.5 Waves
3.6 Current
3.7 Water Levels and Airgap
3.8 Temperatures
3.9 Marine Growth
3.10 Leg Length
3.11 Geotechnical and Geophysical Information
3.12 Bathymetric Survey
3.13 Seabed Surface Survey
3.14 Geophysical Information - Shallow Seismic Survey
3.15 Surface Soil Samples
3.16 Geotechnical Investigations
Glossary of Terms for Section 3
4 CALCULATION METHODS –
HYDRODYNAMIC AND WIND FORCES 27
4.1 Introduction
4.2 Wind Force Calculations
4.3 Hydrodynamic Forces
4.4 Wave Theories and Analysis Methods
4.5 Current
4.6 Leg Hydrodynamic model
4.7 Hydrodynamic Coefficients for Leg Members
4.8 Other Considerations
STRUCTURAL ENGINEERING 42
5.1 General Conditions
5.2 Seabed Reaction Point
5.3 Foundation Fixity
5.4 Leg Inclination
5.5 P-Δ Effects
5.6 Structural Modeling
- Introduction
- General considerations
- Applicability and limitations
- Modeling the leg
- Modeling the hull
- Modeling the hull/leg connection
- Modeling the spudcan
Rev 3, August 2008
CONTENTS (continued)
5.7 Load Application
- Self weight, variable and drilling loads
- Wind loads
- Hydrodynamic wave-current loads
- Inertial loads due to dynamic response
- Second order effects
GEOTECHNICAL ENGINEERING 61
6.1 Introduction
6.2 Prediction of Leg Penetration During Preloading
- Analysis method
- Penetration in clays
- Penetration in silica sands
- Penetration in carbonate sands
- Penetration in silts
- Penetration in layered soils
6.3 Foundation Stability Assessment
- Approach
- Ultimate bearing capacity for vertical loading
preload check (Step 1)
- Bearing capacity/sliding check - pinned footing (Step 2a)
- Footing with moment fixity and vertical and
horizontal stiffness (Step 2b)
- Displacement Check (Step 3)
6.4 Other aspects of jack-up unit foundation performance
- Leaning Instability
- Footprint Considerations
- Scour
- Seafloor Instability
- Shallow Gas
- Spudcan-Pile Interaction
DETERMINATION OF RESPONSES 89
7.1 General
7.2 Quasi-Static Extreme Response with Inertial Loadset
7.3 Dynamic Extreme Response
- Factors Governing Dynamics
- The Structural System
- The Excitation
- The Dynamic Analysis
- The Natural Period(s)
- Inertial Loadset Approaches
- Detailed Dynamic Analysis Methods
- Acceptance Criteria
7.4 Fatigue
- General
- Fatigue life requirements
- Fatigue sensitive areas
- General description of analysis
Rev 3, August 2008
8 ACCEPTANCE CRITERIA 110

8.1 Structural Strength Check
- Introduction
- Definitions
- Factored loads
- Assessment of members (excluding stiffened and
high R/t tubulars)
- Assessment of other geometries
- Assessment of member joints
8.3 Foundation Assessment
- Step 1, Preload and sliding checks
- Step 2a, Capacity check - Pinned foundation
- Step 2b, Capacity check - With foundation fixity
- Step 3, Displacement check
- Punch-through
8.4 Horizontal Deflections
8.5 Loads in the Holding System
8.6 Hull
8.7 Structure Condition Assessment
REFERENCES 139
INDEX 141
LIST OF FIGURES
Figure 2.1 Overall flow chart for the assessment 12
Figure 3.1 Suggested current profile 19
Figure 4.1 Range and validity of different wave theories 32

for regular waves
Figure 4.2 Flow angles appropriate to a lattice leg 35
Figure 4.3 Gusset plates 37
Figure 4.4 Split tube chord and typical values for CDi 38
Figure 4.5 Triangular chord and typical values for CDi 39
Figure 5.1 Formulas for the determination of Equivalent 54

Member Properties
Figure 5.2 Leg shear force and bending moment - jack-ups 55

with a fixation system

without a fixation system and having a fixed
jacking system with opposed pinions
Rev 3, August 2008
LIST OF FIGURES (continued)

without a fixation system and having a fixed
jacking system with unopposed pinions

without a fixation system and having a floating
jacking system
Figure 5.6 Correction of point supported guide model for 59

finite guide length
Figure 6.1 Typical spudcan geometries 62
Figure 6.2 Spudcan foundation model 62
Figure 6.3 Stability factors for cylindrical 64

excavations in clay
Figure 6.4 Spudcan bearing capacity analysis 65
Figure 6.5 Spudcan bearing capacity analysis - squeezing 66

clay layer
Figure 6.6 Spudcan bearing capacity analysis - firm clay 67

over weak clay
Figure 6.7 Spudcan bearing capacity analysis - sand over 68

clay
Figure 6.8 Spudcan bearing capacity analysis - three 69

layer case
Figure 6.9 Foundation stability assessment 71
Figure 6.10 Calculation procedure to account for foundation 80

Fixity
Figure 6.11 Calculation procedure to account for foundation 81

fixity
Figure 7.1 Recommended approach to determine extreme 92

dynamic responses
Figure 7.2 Procedure for calculation of distributed inertial 101

loadset (2-D response)
Figure 8.1 Flow chart for member strength assessment 111
Figure 8.2 Typical members and components 127
Figure 8.3 Effective Length Factors 128
Figure 8.4 Chart for Determination of η 128

Rev 3, August 2008
LIST OF TABLES
Table 3.1 Foundation risks, methods for evaluation and 22

prevention
Table 4.1 Height coefficients 28
Table 4.2 Shape coefficients 29
Table 4.3 Base hydrodynamic coefficients for tubulars 36
Table 5.1 Applicability of the suggested models 48
Table 7.1 Recommended damping from various sources 102
Table 7.2 Recommendations for application of dynamic 104

analysis methods
Table 7.3 Recommendations for determining MPME 105

Rev 3, August 2008
1 INTRODUCTION
1.1 The purpose of this document is to provide a Recommended Practice (PRACTICE) for
use with the 'Guideline for Site Specific Assessment of Jack-Up Units' (GUIDELINE).
Each assessment should address the areas of this document as appropriate for the
particular jack-up and location as described in Section 1.4 of the GUIDELINE.
1.2 This document has been formulated as a result of a Joint Industry Project involving all
sections of the industry. It is not intended to obviate the need for applying sound
judgment as to when and where this PRACTICE should be utilized.
1.3 The formulation and publication of this PRACTICE is not in any way intended to impose
calculation methods or procedures on any party. It leaves freedom to apply alternative
practices within the framework of the accompanying GUIDELINE.
1.4 This PRACTICE relates only to the assessment of independent leg jack-up units in the
elevated condition. The development has been based on 3 legged truss-leg units and
caution is advised when applying the PRACTICE to other configurations. Transportation
to and from the site and moving on and moving off location are not covered in this
document.
1.5 This PRACTICE may be revised if and when more information/research results become
available.
1.6 For further details of the applicability and limitations, refer to the GUIDELINE.
1.7 This PRACTICE may be used by anyone desiring to do so, and a diligent effort has
been made by the authors to assure the accuracy and reliability of the information
contained herein. However, the authors make no representation, warranty or
guarantee in connection with the publication of this PRACTICE and hereby
expressly disclaim any liability or responsibility for loss, damage or injury resulting
from its use, for any violation of local regulations with which a recommendation
may conflict, or for the infringement of any patent resulting from the use of this
publication.
1.8 The load factors presented in Section 8 herein were determined from the reliability
analysis of a limited number of jack-up/site combinations. The load factors are
provisional pending the further evaluation of the results from a wider range of
assessments by the SNAME OC-7 panel.
Alternative values can be used when acceptable rationale is provided.

Rev 3, August 2008
2 OBJECTIVES
2.1 The principal objective of this PRACTICE is to provide acceptance criteria and
associated engineering methods that may be applied in the site specific assessment of a
jack-up to:
a) Establish the geometric suitability of the jack-up with respect to leg length, airgap and
leg penetration.
b) Establish that the jack-up is structurally adequate for its intended application.
c) Ensure that the foundation can offer suitable support to meet this objective.
d) Ensure adequate overturning stability.
2.2 This PRACTICE is applicable to the various possible modes of jack-up operation
(drilling, production, accommodation, construction, etc.) in all areas of the world. It
should be noted that different extreme environmental return periods may be appropriate
for manned and unmanned operations.
2.3 The user of this PRACTICE is advised that, in some areas of the world, the requirements
of the local regulatory bodies may be more onerous than those recommended herein.
2.4 Scope of the Assessment
2.4.1 The primary objective of the site specific assessment is to ensure the integrity of the jack-
up in the elevated condition. The assumptions incorporated into the assessment must
conform with the structural condition of the unit.
2.4.2 The assessment will normally assume that the jack-up is in sound mechanical and
structural condition and it is the responsibility of the owner to ensure that this is so. The
existence of valid documents indicating that the jack-up is presently in class by a
recognized classification society is usually sufficient to verify the mechanical and
structural condition of the jack-up to the assessor.
2.4.3 Accidental loads (dropped objects, ship impact, etc.) are not specifically addressed and
should be covered at the design stage. Furthermore, the site specific assessment
addresses the global structural integrity, hence local damage not affecting the overall
integrity is outside the scope of the PRACTICE.
2.4.4 As indicated in Section 1.4.1 of the GUIDELINE, the assessment of the jack-up may be
carried out at various degrees of complexity. These are as expanded below, at increasing
levels of complexity. The objective of the assessment is to show that the acceptance
criteria of Section 8 of this PRACTICE are met. If this is achieved by a particular level
there is no need to consider a more complex level.
1. Compare site conditions with design conditions or other existing assessments

determined in accordance with this PRACTICE.
Rev 3, August 2008
2.4.4 2. Carry out appropriate calculations according to the simple methods given in this
PRACTICE. Possibly compare results with those from existing more
detailed/complex calculations.
3. Carry out appropriate detailed calculations according to the more complex methods
given in this PRACTICE.
In all cases the adequacy of the foundation should be assessed.
An overall flow chart for the assessment is given in Figure 2.1 overleaf.
Rev 3, August 2008
Figure 2.1 - Overall flow chart for the assessment

Rev 3, August 2008
3 ASSESSMENT INPUT DATA
3.1 Rig data
3.1.1 The information that may be required to perform the assessment is outlined in Section 2.1
of the GUIDELINE.
3.1.2 The operating procedures and limitations of the jack-up should be clearly defined in the
Operating Manual. Those sections of the Operating Manual which give relevant
information and are required to perform a site assessment in accordance with this
PRACTICE are to be provided.
3.2 Functional Loadings
3.2.1 The operating and survival conditions may be treated separately, provided it is practical
to change the mode of the jack-up unit from operating to survival mode on receipt of an
unfavorable weather forecast, and appropriate procedures exist. The limits of operational
loading conditions may depend on the drilling program proposed and consideration
should be given to loadings on the conductors if supported by the jack-up.
3.2.2 For both operational and survival conditions, the following shall be defined:
a) Maximum and minimum elevated weight and weight distribution (fixed and variable
load), excluding legs. In the absence of other information the minimum elevated
weight may normally be determined assuming 50% of the variable load permitted by
the operating manual.
b) Extreme limits of center of gravity position (or reactions of the elevated weight on the
legs) for the conditions in a) above.
c) Substructure and derrick position, hook load, rotary load, setback and conductor
tensions for the conditions in a) above.
d) Weight, center of gravity and buoyancy of the legs.
3.2.3 With reference to Section 4.1.3 of the GUIDELINE, if a minimum elevated weight or a
limitation of center of gravity position is required to meet the overturning safety factor in
survival conditions, then the addition of water in lieu of variable load is permitted,
provided that:
a) Maximum allowable loadings are not exceeded.
b) Procedures, equipment and instructions exist for performing the operation.
c) The maximum variable load, including added water, is used for all appropriate
assessment checks (preload, stress, etc.).
3.3 Environmental Conditions - General
3.3.1 The environmental data required for an assessment and their application are discussed in
Section 2.3 of the GUIDELINE.
Rev 3, August 2008
3.3.2 Section 2.3 of the GUIDELINE recommends that 50 year return period extremes are
normally used, however in particular circumstances other return periods may be
appropriate.
3.3.3 Unless there is specific data to the contrary, wind, wave and current loadings shall be
considered to be those caused by the individual return period extremes acting in the same
direction and at the same time as the extreme water level. Seasonally adjusted values
may be adopted as appropriate to the duration of the operation.
Note:
Where directional and/or seasonal data are utilized, these should generally be factored so
that the data for the worst direction and/or season equals the omni-directional/all-year
data for the assessment return period.
3.4 Wind
3.4.1 The wind velocity shall be the 1 minute sustained wind for the assessment return period,
related to a reference level of 10.0m above mean sea level. The Commentary discusses
the conversion of data for averaging periods other than 1 minute to 1 minute values.
1
3.4.2 The wind velocity profile is normally taken as a power law with exponent 10 unless site
specific data indicates otherwise (see Section 4.2.2).
3.4.3 Different jack-up configurations (weight, center of gravity, cantilever position, etc.) may
be specified for operating and survival modes. In such cases, the maximum wind
velocity considered for the operating mode should not exceed that permitted for the
change to the survival mode.
3.5 Waves
3.5.1 The extreme wave height environment used for survival conditions shall, as a minimum,
be computed according to the following sub-sections based on the three-hour storm
duration with an intensity defined by the significant wave height, Hsrp, for the assessment
return period. The seasonally adjusted wave height may be used as appropriate for the
operation.
The wave height information for a specific location may also be expressed in terms of
Hmax, the individual extreme wave height for the return period, rather than the significant
wave height Hsrp. The relationship between Hsrp and Hmax must be determined accounting
for the effects of storms (longer than 3 hours) and for the additional probability of other
return period storms (see Commentary Section C3.5.1). This relationship will depend on
the site specific conditions, however Hsrp may usually be determined from Hmax using the
generally accepted relationship for non-cyclonic areas:
Hsrp = Hmax/1.86
For cyclonic areas the recommended relationship is:
Hsrp = Hmax/1.75
Rev 3, August 2008
3.5.1 Note:
The wave load can be computed either stochastically (through a random frequency or
time domain approach) or deterministically (through an individual maximum wave
approach). The scaled wave heights for the two approaches are discussed in Sections
3.5.1.1 and 3.5.1.2 respectively (see Commentary). The scaled wave heights are to be
used only in conjunction with the associated kinematics modeling recommended in
Section 4.4 and the hydrodynamic coefficients given in Sections 4.6 to 4.8.
3.5.1.1 For stochastic/random wave force calculations Airy wave theory is implied, see Section
4.4.2. To account for wave asymmetry, which is not included in Airy wave theory, a
scaling of the significant wave height should be applied to capture the largest wave forces
at the maximum crest amplitude. The effective significant wave height, Hs, may be
determined as a function of the water depth, d in meters, from:
Hs = [1 + 0.5e(-d/25)] Hsrp (d ≥ 25m)
and should be used with the wave kinematics model described in Section 4.4.2.
For water depths less than 25m a regular wave analysis should be considered.
The selection of wave period for use in stochastic/random wave force analysis is
discussed in Section 3.5.3 and the Note thereto.
3.5.1.2 For deterministic/regular wave force calculations it is appropriate to apply a kinematics

reduction factor of 0.86 in order to obtain realistic force estimates (see Commentary).
This factor may be considered to implicitly account for spreading and also the
conservatism of deterministic/regular wave kinematics traditionally accomplished by
adjusting the hydrodynamic properties.
The factor should be applied by means of a reduced wave height, Hdet. Hdet may be
determined as a function of Hmax from:
Hdet = 0.86 Hmax
The use of a factor smaller than 0.86 may be justified by analysis explicitly accounting
for the effects of three-directional spreading. However, such effects should be properly
balanced by the inclusion of second-order interaction effects between spectral wave
components.
The wave loads should be determined using an appropriate wave kinematics model in
accordance with Section 4.4.1.
In the analysis a single value for the wave period Tass, in seconds, associated with the
maximum wave may be considered. Unless site specific information indicates otherwise
Tass will normally be between the following limits:
3.44 ( H srp ) < Tass < 4.42 ( H srp )
where Hsrp is the return period extreme significant wave height in meters.
Rev 3, August 2008
3.5.2 For airgap calculations the wave crest elevation may be obtained from the formulations
of an appropriate deterministic wave theory (see Section 4.4.1) and the maximum wave
height, Hmax, from the relationship:
Hmax = 1.86 Hsrp
In Tropical Revolving Storm areas the relationship:
Hmax = 1.75 Hsrp
may alternatively be applied.
It is noted that the minimum return period recommended by the GUIDELINE for Hsrp for
airgap calculations is 50 years, even if a lower return period is used for other purposes.
3.5.3 Where the analysis method requires the use of spectral data, the choice of the analytical
wave spectrum and associated spectral parameters should reflect the width and shape of
spectra for the site and significant wave height under consideration. In cases where fetch
and duration of extreme winds are sufficiently long a fully developed sea will result (this
is rarely realized except, for example, in areas subject to monsoons). Such conditions
may be represented by a Pierson-Moskowitz spectrum. Where fetch or duration of
extreme winds is limited, or in shallow water depths, a JONSWAP spectrum may
normally be applied (see Note at the end of this Section).
The wave spectrum can be represented by the power density of wave surface elevation
Sηη(f) as a function of wave frequency by:
Sηη(f) = (16I0(γ))-1Hs2TP(TPf)-5exp(-1.25/(TPf)4)γq
[Note: An alternative formulation is given in the Commentary]

where;
q = exp(-(Tpf-1)2/2σ2) with:
σ = 0.07 for Tpf ≤ 1
σ = 0.09 for Tpf > 1 (Carter 1982, [1])
and;
Hs = significant wave height (meters), including depth correction, according
to Section 3.5.1.1
Tp = peak period (seconds)
f = frequency (Hz)
γ = peak enhancement factor
I0(γ) = is discussed below.
The above definition yields a single parameter Pierson-Moskowitz spectrum when γ = 1 and Tp =
5 ( H s ) , with Hs in meters. In this case an appropriate Tp/Tz ratio is 1.406 (see below).
Rev 3, August 2008
When considering a JONSWAP spectrum, the peak enhancement factor γ varies between
1 and 7 with a most probable average value of 3.3. There is no firm relationship between
γ, Hs and Tp. Relationships between variables for different γ according to Carter (1982)
[1] are as follows:
γ I0(γ) Tp/Tz
1 .200 1.406
2 .249 1.339
3 .293 1.295 ⎡ Alternatively: ⎤
3.3 .305 1.286 ⎢ 0.2 ⎥
⎢ I 0 ( γ ) = 1 − 0.287 Ln ( γ ) ⎥
4 .334 1.260 ⎣ ⎦
5 .372 1.241
6 .410 1.221
7 446 1.205
Unless site specific information indicates otherwise γ = 3.3 may be used.
For a given significant wave height the wave period depends on the significant wave
steepness which in extreme seas in deep water often lies within the range 1/20 to 1/16.
This leads to an expression for zero-upcrossing period Tz, related to Hsrp in meters, as
follows:
3.2 ( H srp ) < Tz < 3.6 ( H srp )
However in shallow water the wave steepness can increase to 1/12 or more, leading to a
zero-upcrossing period Tz as low as 2.8 ( H srp ) . This is because the wave height
increases and wave length decreases for a given Tz.
Note:
If a JONSWAP spectrum is applied the response analysis should consider a range of periods
associated with Hsrp based on the most probable value of Tp plus or minus one standard
deviation. However it should be ensured that the assumptions made in deriving the spectral
period parameters are consistent with the values used in the analysis. Alternatively, applicable
combinations of wave height and period may be obtained from a scatter diagram determined
from site specific measurements; in this case specialist advice should be obtained on a suitable
spectral form for the location. To avoid the need for analyses of several wave periods a practical
alternative is to use a 2 parameter spectrum with γ = 1.0 in combination with the site specific
most probable peak period.
Rev 3, August 2008
3.5.4 For stochastic/random wave force calculations, the short-crestedness of waves (i.e. the
angular distribution of wave energy about the dominant direction) may be accounted for
when site-specific information indicates that such effects are applicable. In all cases the
potential for increased response due to short-crested waves should be investigated. The
effect may be included by means of a directionality function F(α), as follows:
Sηη(f, α) = Sηη(f).F(α)
where;
α = angle between direction of elementary wave trains and dominant
direction of the short-crested waves.
Sηη(f, α) = directional short-crested power density spectrum.
F(α) = directionality function.
and, in the absence of more reliable data:
π π
F(α) = C.Cos2nα for - ≤ α ≤
2 2
where;
n = power constant
C = constant chosen such that:
π/2
∑ F(α) .dα = 1.0
-π/2
The power constant n, should not normally be taken as less than:
n = 2.0 for fatigue analysis
n = 4.0 for extreme analysis
3.5.5 Where the natural period of the jack-up is such that it may respond dynamically to waves
(Section 7.3), the maximum dynamic response may be caused by wave heights or
seastates with periods outside the ranges given in Sections 3.5.1.2 and 3.5.3. Such
conditions shall also be investigated to ensure that the maximum (dynamic plus quasi-
static) response is determined.
3.5.6 For fatigue calculations (Section 7.4), the long term wave climate may be required. For
the purposes of the fatigue analysis the long-term data may be presented deterministically
in terms of the annual number of waves predicted to fall into each height/period/direction
group. Alternatively the probability of occurrence for each seastate (characterized by
wave energy spectra and the associated physical parameters) may be presented in the
form of a significant wave height versus zero-upcrossing period scatter diagram or as a
table of representative seastates.
3.6 Current
3.6.1 The extreme wind driven surface current velocity shall be that associated with the
assessment return period wind, seasonally adjusted if appropriate. When directional
information regarding other current velocity components is available the maximum
surface flow of the mean spring tidal current and the assessment return period surge
current, seasonally adjusted if appropriate, shall be vectorially added in the down-wind
direction and combined with the wind driven surface current as indicated in Section 3.6.2.
If directional data are not available the components shall be assumed to be omni-
irectional and shall be summed algebraically.
Note: A site specific study will normally be required to define the current velocity
components.
Rev 3, August 2008
3.6.2 The current profile may be expressed as a series of velocities at certain stations from
seabed to water surface. Unless site specific data indicates otherwise, and in the absence
of other residual currents (such as circulation, eddy currents, slope currents, internal
waves, inertial currents, etc.), an appropriate method for computing current profile is (see
Figure 3.1):
VC = Vt + Vs + (Vw - Vs) [(h+z)/h], for |z| ≤ h and Vs < Vw
VC = Vt + Vs for |z| > h or Vs ≤ Vw
where;
VC = current velocity as a function of z. Note that a reduction may be

applicable according to Section 4.5.
Vt = downwind component of mean spring tidal current.
Vs = downwind component of associated surge current (excluding wind driven
component).
Vw = wind generated surface current. In the absence of other data this may
conservatively be taken as 2.6% of the 1 minute sustained wind velocity
at 10m.
h = reference depth for wind driven current. In the absence of other data h
shall be taken as 5 meters.
z = distance above still water level (SWL) under consideration (always
negative).
Figure 3.1 - Suggested current profile
3.6.3 In the presence of waves the current profile should be stretched/compressed such that the
surface component remains constant. This may be achieved by substituting the elevation
as described in Section 4.4.2. Alternative methods may be suitable, however mass
continuity methods are not recommended. The current profile may be changed by wave
breaking. In such cases the wind induced current could be more uniform with depth.
3.6.4 For a fatigue analysis, current may normally be neglected.

Rev 3, August 2008
3.7 Water Levels and Airgap
3.7.1 The water depth at the location shall be determined and related to lowest astronomical
tide (LAT). The relationship between LAT and Chart Datum is discussed in the
Commentary.
3.7.2 The mean water level (MWL) related to the seabed shall be expressed as the mean level
between highest astronomical tide (HAT) and lowest astronomical tide (LAT) i.e.:
MWL = (HAT + LAT)/2
3.7.3 The extreme still water level (SWL) shall be expressed as a height above LAT, and shall
be the sum of;
Mean high water spring tide (MHWS)

+ 50 year extreme storm surge (see Note 1).
unless reliable data indicates that an alternative summation is appropriate.
3.7.4 When lower water levels are more onerous the minimum still water level (SWL) to be
considered in the loading calculations shall be the sum of:
Mean Low Water Spring Tide (MLWS)

+ 50 year negative Storm Surge.
3.7.5 The Airgap (see Note 2) is defined in Section 3.2 of the GUIDELINE as the distance
between the underside of the hull and LAT during operations. It shall be not less than the
sum of:
Distance of the extreme still water level (SWL), from Section 3.7.3, above LAT
+ 50 year extreme wave crest height associated with Hmax as defined in
Section 3.5.2 (see Note 1),
+ 1.5m Clearance to the underside of the hull (or any other vulnerable part
attached to the hull, if lower). See Commentary.
Notes: 1. Section 3.2.1 of the GUIDELINE recommends that values for a return period
of no less than 50 years be applied, even if a lower return period is used for
other purposes.
2. The definition of Airgap used herein differs from that used in other areas of
offshore engineering where the Clearance used here is often defined as
Airgap.
In areas subject to freak waves a higher airgap may be applicable.
3.8 Temperatures
The lowest average daily air and water temperatures shall be compared with the steel
design temperature limits of appropriate parts of the jack-up. If these are not met,
suitable adjustments should be made to the properties applied in the strength assessment.
Rev 3, August 2008
3.9 Marine Growth
Where existing marine growth is not to be cleaned between locations or where the
operation is to last long enough for significant growth to occur, the influence of growth
on the leg hydrodynamic properties should be considered as stated in Section 4.2.3 of the
GUIDELINE. Where applicable, location specific data should be obtained. In the
absence of such data, default values for thickness and distribution are given in Section
4.7.3.
3.10 Leg Length
Recommendations regarding the reserve leg length are given in Section 3.3 of the
GUIDELINE.
3.11 Geotechnical and Geophysical Information
Adequate geotechnical and geophysical information must be available to assess the

location and the foundation stability. Aspects which should be investigated are shown in
Table 3.1 and are discussed in more detail in the referenced Sections. The information
obtained from the surveys and investigations set out in Sections 3.12 to 3.16 is required
for areas where there is no data available from previous operations. In areas where
information is available it may be possible to reduce the requirements set out below by
use of information obtained from other surveys or activities in the area. See Section 2.4
of the GUIDELINE.
3.12 Bathymetric Survey
3.12.1 An appropriate bathymetric survey should be supplied for an area approximately 1

kilometer square centered on the location. Line spacing of the survey should typically be
not greater than 100 meters x 250 meters over the survey area. Interlining is to be
performed within an area 200 meters x 200 meters centered on the location. Interlining
should have spacing not exceeding 25 meters x 50 meters.
3.12.2 Further interlining should be performed if any irregularities are detected.
3.13 Seabed Surface Survey
3.13.1 The seabed surface shall be surveyed using sidescan sonar or high resolution multibeam
echosounder techniques and shall be of sufficient quality to identify obstructions and
seabed features and should cover the immediate area (normally a 1 km square) of the
intended location. The slant range selection shall give a minimum of 100% overlap
between adjacent lines. A magnetometer survey may also be required if there are buried
pipelines, cables and other metallic debris located on or slightly below the sea floor.
Rev 3, August 2008
REFERENCE
RISK METHODS FOR EVALUATION & SECTION(S)
PREVENTION
Installation problems - Bathymetric survey 3.12
Punch-through - Shallow seismic survey 3.14

- Soil sampling and other geotechnical 3.16
testing and analysis 6.2.6
Settlement under storm - Shallow seismic survey 3.14

loading/Bearing failure - Soil sampling and other geotechnical 3.16
- Ensure adequate jack-up preload capability 6.3
Sliding failure - Shallow seismic survey 3.14

- Increase vertical footing reaction
- Modify the footing(s)
Scour - Bathymetric survey (identify sand 3.12

waves) 3.15
- Surface soil samples and seabed currents 6.4.3
- Inspect footing foundations regularly
- Install scour protection (gravel bag/
artificial seaweed) when anticipated
Seafloor instability - Side scan sonar, shallow seismic 3.13

(mudslides) survey 3.14
Gas pockets/ - Digital seismic with attribute 3.14

Shallow gas analysis processing (shallow seismic) 6.4.5
Faults - Shallow seismic survey 3.14
Metal or other object, - Magnetometer and side scan sonar 3.13

sunken wreck, anchors, - Diver/ROV inspection
pipelines etc.
Local holes (depressions) - Side scan sonar 3.13

in seabed, reefs, - Diver/ROV inspection
pinnacle rocks
or wooden wreck
Legs stuck in mud - Geotechnical data 3.14

- Consider change in footings 3.16
- Jetting
Footprints of previous - Evaluate location records 3.12

jack-ups - Consider filling/modification 3.13
of holes as necessary 6.4.2
Table 3.1 - Foundation risks, methods for evaluation and prevention

Rev 3, August 2008
3.13.2 Where seabed obstructions such as pipelines and wellheads are known to be present,
sufficient information to enable safe positioning of the jack-up is required. In some cases
an ROV or diver's inspection may be required in addition to a sidescan sonar survey.
3.13.3 Seabed surface surveys can become out-of-date, particularly in areas of

construction/drilling activity or areas with mobile sediments. Good judgment should be
used regarding the applicability of all surveys, especially with regard to validity. In open
locations the maximum period for the validity of seabed surveys for debris and mobile
sediment conditions should be determined taking account of local conditions. For
locations close to existing installations seabed surveys for debris and sediment conditions
should, subject to practical considerations, be undertaken immediately prior to the arrival
of the jack-up at the location.
3.14 Geophysical Investigation - Shallow Seismic Survey
3.14.1 The principal objectives of the shallow seismic survey are:
- To determine near surface soil stratigraphy. This requires correlation of the

seismic data with (existing) soil boring data in the vicinity.
- To reveal the presence of shallow gas concentrations.
Due to the qualitative nature of seismic surveys it is not possible to conduct analytical
foundation appraisals based on seismic data alone. This requires correlation of the
seismic data with soil boring data in the vicinity through similar stratigraphy.
3.14.2 A shallow seismic survey should be performed over an approximately 1 kilometer square
area centered on the location. Line spacing of the survey should typically not be greater
than 100 meters x 250 meters over the survey area. Equipment should normally be
capable of giving detailed data to a depth equal to the greater of 30 meters or the
anticipated footing penetration plus 1.5 to 2 times the footing diameter. Further guidance
on seismic surveys is given in reference [2].
3.14.3 The survey report should include at least two vertical cross-sections passing through the
location showing all relevant reflectors and allied geological information. The equipment
used should be capable of identifying reflectors of 0.5m and thicker.
3.15 Surface Soil Samples
The site investigation should be sufficient to identify the character of the soil surface and allow
evaluation of the possibility of scour occurring. (See Commentary to Section 6.4.3)
Rev 3, August 2008
3.16 Geotechnical Investigations
3.16.1 Site specific geotechnical testing is recommended in areas where any of the following
apply:
- the shallow seismic survey cannot be interpreted with any certainty,

- significant layering of the strata is indicated,
- the location is in a new operating area,
- the area is known to be potentially hazardous.
3.16.2 A geotechnical investigation should comprise a minimum of one borehole to a depth

equal to 30 meters or the anticipated footing penetration plus 1.5 to 2 times the footing
diameter, whichever is the greater. All the layers should be adequately investigated and
the transition zones cored at a sufficient sampling rate.
The number of boreholes required should account for the lateral variability of the soil
conditions, regional experience and the geophysical investigation. When a single
borehole is made, the preferred location is at the center of the leg pattern at the intended
location.
3.16.3 "Undisturbed" soil sampling and laboratory testing and/or in-situ cone penetrometer
testing may be conducted. Other recognized types of in-situ soil testing may be
appropriate such as vane shear and/or pressure meter tests.
3.16.4 The geotechnical report should include borehole logs, cone penetrometer records (if
appropriate) and documentation of all laboratory tests, together with interpreted soil
design parameters. Design parameters should be selected by a competent person. For
the methods recommended in Section 6, the design parameters should include profiles of
undrained shear strength and/or effective stress parameters, soil indices (plasticity,
liquidity, grain size, etc.), relative density, unit weight and, where applicable, the over
consolidation ratio (OCR).
Additional soil testing to provide shear moduli and cyclic/dynamic behavior may be
required if more comprehensive analysis are to be applied or where the soil strength may
deteriorate under cyclic loading.
Rev 3, August 2008
3 GLOSSARY OF TERMS - ASSESSMENT INPUT DATA
C = Constant in expression for F(α).

d = Water depth.
f = Wave frequency.
F(α) = Directionality function
= C.Cos2nα
h = Reference depth for wind driven current.
= 5.0 m in the absence of other data.
HAT = Water depth at highest astronomical tide.
Hdet = Reduced wave height which may be used for deterministic wave force
calculations, allowing for the conservatisms of higher order wave theories.
= 1.60 Hsrp
Hmax = The individual extreme wave height for a given return period defined as the
wave height with an annual probability of exceedence of 1/return period (e.g.
the 50 year return period Hm has a 2% annual probability of exceedence).
Where local data is not available:
Hmax = 1.86 Hsrp (for non-tropic revolving storm areas),
Hmax = 1.75 Hsrp (for tropical revolving storm areas.)
When Hmax is used for airgap calculations the minimum return period for Hsrp is
recommended as 50 years, even if a lower return period is used for other
purposes.
Hs = Significant wave height (meters), including depth/asymmetry correction,
according to Section 3.5.1.1.
Hsrp = The assessment return period significant wave height for a three hour storm.
I0(γ) = Parameter depending on γ used in the expression for Sηη(f).
LAT = Water depth at lowest astronomical tide.
MHWS = Height of mean high water spring tide above LAT.
MLWS = Height of mean low water spring tide above LAT.
MWL = Mean water level related to the seabed.
n = Power constant in expression for F(α).
= 2 or 4.
q = Exponent in expression for Sηη(f).
= exp(-(Tpf-1)2/2σ2)
Sηη(f) = Power density spectrum of long crested wave surface elevation as a function of
frequency, f.
= (16I0(γ))-1Hs2Tp(Tpf)-5exp(-1.25/(Tpf)4)γq
Sηη(f,α) = Power density spectrum of short-crested wave surface elevation as a function of
frequency, f.
= Sηη(f).F(α)
SWL = Height of extreme still water level above LAT.
= MHWS + 50 year storm surge.
= MLWS + 50 year negative storm surge (if more onerous).
Tass = Wave period associated with Hmax (also used with Hdet).
Tp = Peak period associated with Hsrp (also used with Hs).
Tz = Zero-upcrossing period associated with Hsrp (also used with Hs).
VC = Current velocity as a function of z.
Vs = Downwind component of surge current.
Vt = Downwind component of mean spring tidal current.
Vw = Wind generated surface current.
= 2.6% of 1 minute sustained wind velocity at 10m, in the absence of other data.
z = Distance above still water level used in determination of VC.
Rev 3, August 2008
3 GLOSSARY OF TERMS - ASSESSMENT INPUT DATA (Continued)
α = Angle between direction of elementary wave trains and dominant direction of

short-crested waves.
γ = Peak enhancement factor used in expression for Sηη(f). For JONSWAP
spectrum varies between 1 and 7 with a most probable average value of 3.3.
σ = Constant in expression for q
= 0.07 for Tpf <= 1
= 0.09 for Tpf > 1
Rev 3, August 2008
4 CALCULATION METHODS - HYDRODYNAMIC AND WIND FORCES
4.1 Introduction
4.1.1 The models, methods and coefficients given in this Section are matched to represent a
consistent method such that the whole Section should be considered together. No force
coefficients should be used unless they correspond to a particular stated analysis method.
4.1.2 The environmental forces may be determined according to the recommendations of this
Section based on the dimensions of the members and the environmental criteria as
described in Section 3 (wind speed, wave height and period and current velocity and
profile).
4.1.3 Since differences in shape, proportions and even detail can result in considerable
differences in the resultant forces, rational data from model testing may be used by the
assessor at his discretion subject to the conditions of Section 4.7.6.
4.2 Wind Force Calculations
4.2.1 For wind load application according to Section 5.7.2, the wind force for each component
(divided into blocks of not more than 15m vertical extent), FWi, may be computed using
the formula:
FWi = Pi AWi
where;
Pi = the pressure at the center of the block.
AWi = the projected area of the block considered.
and the pressure Pi shall be computed using the formula:
Pi = 0.5 ρ (Vref)2 Ch Cs
where;
ρ = density of air (to be taken as 1.2224 kg/m3 unless an alternative value
can be justified for the location).
Vref = the 1 minute sustained wind velocity at reference elevation (normally
10m above MWL), see Section 3.4.1.
Ch = height coefficient, as given in Section 4.2.2.
Cs = shape coefficient, as given in Section 4.2.3.
Note:
The wind area of the hull and associated structures (excluding derrick and legs) may normally be
taken as the profile area viewed from the direction under consideration.
Rev 3, August 2008
4.2.2 Ch may be derived from the wind velocity profile;
VZ = Vref (Z/Zref)1/N
where;
VZ = the wind velocity at elevation Z.
Vref = the 1 minute sustained wind velocity at elevation Zref (normally 10m
above MWL), see Section 3.4.1.
N = 10 unless site specific data indicate that an alternative value of N is

appropriate.
Hence:
Ch = (VZ/Vref)2 = (Z/Zref)2/N, but always ≥ 1.0
Alternatively, the approximate coefficients shown in Table 4.1 may be applied. The
height is the vertical distance from the still water surface to the center of area of the block
considered. Blocks which have a vertical dimension greater than 15 m shall be sub-
divided, and the appropriate height coefficients applied to each part of the block.
Height Height coefficient

m Ch
0 - 15 1.00
15 - 30 1.18
30 - 45 1.30
45 - 60 1.39
60 - 75 1.47
75 - 90 1.53
90 - 105 1.58
105 - 120 1.62
120 - 135 1.66
135 - 150 1.70
150 - 165 1.74
165 - 180 1.77
180 - 195 1.80
Table 4.1 - Height coefficients
In deriving Table 4.1 the wind velocity used to obtain Ch for the block below 15.0m is the
Vref value. For all other blocks the Ch value is that for the mid-height of the block. When
using Table 4.1 the wind velocity is derived from Section 3.4.1 for a reference height of
10m above the still water.
Rev 3, August 2008
4.2.3 Shape coefficients shall be derived from Table 4.2;
Type of member or Shape coefficient

structure Cs
Hull side, (flat side) 1.0, based on total projected area
Deckhouses, jack-frame 1.1, based on the total projected

structure, sub-structure, draw- area (i.e. the area enclosed by the
works house, and other above- extreme contours of the structure)
deck blocks
Leg sections projecting above Cs = CDe as determined from
jack-frame structure and below Section 4.6, except that marine
the hull growth may be omitted. AWi
determined from De and section
length.
Isolated tubulars (crane 0.5

pedestals, etc.)
Isolated structural shapes 1.5, based on member projected

(angles, channels, box, I- area
sections)
Derricks, crane booms, flare The appropriate shape coefficient

towers (open lattice sections for the members concerned
only, not boxed- in sections) applied to 50% of the total
projected profile area of the item
(25% from each of the front and
back faces)
Shapes or combinations of shapes which do not readily fall into the
above categories will be subject to special consideration
Table 4.2 - Shape coefficients
4.3 Hydrodynamic Forces
4.3.1 Wave and current forces on slender members having cross sectional dimensions
sufficiently small compared with the wave length should be calculated using Morison's
equation. Note: Morison's equation is normally applicable providing:
λ > 5Di where;

λ = wavelength and
Di = reference dimension of member (e.g. tubular diameter)
Morison's equation specifies the force per unit length as the vector sum:
ΔF = ΔFdrag + ΔFinertia = 0.5 ρ D CD vn ⏐vn⏐+ ρ CM A u n
where the terms of the equation are described in the following.

Rev 3, August 2008
4.3.2 To obtain the drag force, the appropriate drag coefficient (CD) is to be chosen in
combination with a reference diameter, including any required additions for marine
growth, as described in Section 4.7.
The Morison's drag force formulation is:

ΔFdrag = 0.5 ρ CD D vn ⏐vn⏐
where;
ΔFdrag = drag force (per unit length) normal to the axis of the member
considered in the analysis and in the direction of vn.
ρ = mass density of water (normally 1025 kg/m3).
CD = drag coefficient ( = CDi or CDe from Section 4.6-7).
vn = relative fluid particle velocity resolved normal to the member axis.
D = the reference dimension in a plane normal to the fluid velocity vn
( = Di or De from Section 4.6-7).
Note: The relative fluid particle velocity, vn, may be taken as:
vn = un + VCn - α r n
where;
un + VCn = the combined particle velocity found as the vectorial sum of the
wave particle velocity and the current velocity, normal to the
member axis.
rn = the velocity of the considered member, normal to the member axis
and in the direction of the combined particle velocity.
α = 0, if an absolute velocity is to be applied, i.e. neglecting the
structural velocity.
= 1, if relative velocity is to be included. May only be used for
stochastic/random wave force analyses if:
uTn/Di ≥ 20
where u = particle velocity = VC + πHs/Tz
Tn = first natural period of surge or sway motion
and Di = the reference diameter of a chord.
Note:
See also Section 7.3.7 for relevant damping coefficients depending on α.
4.3.3 To obtain the inertia force, the appropriate inertia coefficient (CM) is to be taken in
combination with the cross sectional area of the geometric profile, including any required
additions for marine growth, as described in Section 4.7.
The Morison's inertia force formulation is:
ΔFinertia = ρ CM A u n
where;
ΔFinertia = inertia force (per unit length) normal to the member axis and in the
direction of u n.
ρ = mass density of water (normally 1025 kg/m3).
CM = inertia coefficient.
A = cross sectional area of member ( = Ai or Ae from Section 4.6)
u n = fluid particle acceleration normal to member.
Rev 3, August 2008
4.4 Wave Theories and Analysis Methods
4.4.1 For deterministic analyses an appropriate wave theory for the water depth, wave height
and period shall be used, based on the curves shown in Figure 4.1, after HSE [3]. For
practical purposes, an appropriate order of Dean's Stream Function or Stokes' 5th (within
its bounds of applicability) is acceptable for regular wave survival analysis.
4.4.2 For random wave (stochastic) analyses, it is recommended that the random seastate is
generated from the summation of at least 200 component Linear (Airy) waves of height
and frequency determined to match the required wave spectrum. The phasing of the
component waves should be selected at random.
The extrapolation of the wave kinematics to the free surface is most appropriately carried
out by substituting the true elevation at which the kinematics are required with one which
is at the same proportion of the still water depth as the true elevation is of the
instantaneous water depth. This can be expressed as follows:
z−ζ
z' =
1+ ζ / d
where;
z' = The modified coordinate to be used in particle velocity formulation
z = The elevation at which the kinematics are required (coordinate measured
vertically upward from the still water surface)
ζ = The instantaneous water level (same axis system as z)
d = The still, or undisturbed water depth (positive).
This method ensures that the kinematics at the surface are always evaluated from the
linear wave theory expressions as if they were at the still water level, Wheeler (1969) [4]
(see Figure C4.4.2 in the Commentary).
4.4.3 If breaking waves are specified according to Figure 4.1, it is recommended that the wave
period is changed to comply with the breaking limit for the specified height.
Rev 3, August 2008
Notes
1) None of these theories is theoretically correct at the breaking limit. Nomenclature

2) Wave theories intended for limiting height waves should be
referenced for waves higher than 0.9Hb when stream function Hmax/gTass2 = Dimensionless wave steepness
theory may underestimate the kinematics.
3) Stream function theory is satisfactory for wave loading calculations d/gTass2 = Dimensionless relative depth
over the remaining range of regular waves. However, stream Hmax = Wave height (crest to trough)
function programs may not produce a solution when applied to near Hb = Breaking wave height
breaking waves or deep water waves d = Mean water depth
4) The order of stream function theory likely to be satisfactory is Tass = Wave period
circled. Any solution obtained should be checked by comparison L = Wave length (distance between
with the results of a higher order solution. crests)
5) The error involved in using Airy theory outside its range of g = Acceleration due to gravity
applicability is discussed in the background document.
Figure 4.1 - Range and validity of different wave theories for

regular waves, (after HSE [3])
Rev 3, August 2008
4.5 Current
4.5.1 The current velocity and profile as specified in Section 3.6 shall be used. Interpolation
between the data points may be required and linear interpolation is recommended for
simplicity.
4.5.2 The current induced drag forces are to be determined in combination with the wave
forces. This is to be carried out by the vectorial addition of the wave and current induced
particle velocities prior to the drag force calculations.
4.5.3 The current may be reduced due to interference from the structure on the flow field of the
current, Taylor [5]. The current may be reduced as follows (see Commentary):
VC = Vf [1 + CDeDe/(4D1)]-1
where;
VC = the current velocity to be used in the hydrodynamic model, VC should be
not taken as less than 0.7Vf.
Vf = the far field (undisturbed) current.
CDe = equivalent drag coefficient, as defined in 4.6.5.
De = equivalent diameter, as defined in 4.6.5.
D1 = face width of leg, outside dimensions.
4.6 Leg Hydrodynamic Model
4.6.1 The hydrodynamic modeling of the jack-up leg may be carried out by utilizing 'detailed'
or 'equivalent' techniques. In both cases the geometric modeling procedure corresponds
to the respective modeling techniques described in Section 5.6.4. The hydrodynamic
properties are then found as described below:
'Detailed' model
All relevant members are modeled with their own unique descriptions for the Morison
term values with the correct orientation to determine vn and u n and the corresponding
CDD = CDiDi and CMA = CMiπDi2/4, as defined in Section 4.7.
'Equivalent' model
The hydrodynamic model of a bay is comprised of one, 'equivalent' vertical tubular
located at the geometric center of the actual leg. The corresponding (horizontal) vn and
u n are applied together with equivalent CDD = CDeDe and CMA = CMeAe, as defined in
Sections 4.6.5 and 4.6.6. The model should be varied with elevation, as necessary, to
account for changes in dimensions, marine growth thickness, etc.
Note:
The drag properties of some chords will differ for flow in the direction of the wave
propagation (wave crest) and for flow back towards the source of the waves (wave
trough). Often the combined drag properties of all the chords on a leg will give a total
which is independent of the flow direction along a particular axis. When this is not the
case it is recommended that the effect is included directly in the wave-current loading
model. If this is not possible it is recommended that:
Rev 3, August 2008
1. Regular wave deterministic calculations use a value appropriate to the flow direction
under consideration, noting that the flow direction is that of the combined wave and
current particle motion.
2. An average drag property is considered for random wave analyses which are solely
used to determine dynamic effects for inclusion in a final regular wave deterministic
calculation which will be made on the basis of 1. above.
3. The drag property in the direction of wave propagation is used for random wave
analyses from which the final results are obtained directly.
4.6.2 Lengths of members are normally taken as the node-to-node distance of the members in
order to account for small non-structural items (e.g. anodes, jetting lines of less than 4"
nominal diameter). Large non-structural items such as raw water pipes and ladders are to
be included in the model. Free standing conductor pipes and raw water towers are to be
considered separately from the leg hydrodynamic model.
4.6.3 The contribution of the part of the spudcan above the seabed should be investigated and
only excluded from the model if it is shown to be insignificant. In water depths greater
than 2.5 Hs or where penetrations exceed 1/2 the spudcan height, the effect of the spudcan
is normally insignificant.
4.6.4 For leg structural members, shielding and solidification effects should not normally be
applied in calculating wave forces. The current flow is however reduced due to
interference from the structure on the flow field, see Section 4.5.3.
4.6.5 When the hydrodynamic properties of a lattice leg are idealized by an 'equivalent' model
description the model properties may be found using the method given below:
The equivalent value of the drag coefficient, CDe, times the equivalent diameter, De, to be
used in Section 4.3.2 for CDei of the bay may be chosen as:
CDe De = De Σ CDei
The equivalent value of the drag coefficient for each member, CDei, is determined from:
CDei = [ sin2βi + cos2βi sin2αi ]3/2 CDi D i 1 i
D es
where;
CDi = drag coefficient of an individual member (i) as defined in Section
4.7.
Di = reference diameter of member 'i' (including marine growth as
applicable) as defined in Section 4.7.
De = equivalent diameter of leg, suggested as ( ∑ D i 2 l i / s
li = length of member 'i' node to node center.
s = length of one bay, or part of bay considered.
αi = angle between flow direction and member axis projected onto a
horizontal plane.
βi = angle defining the member inclination from horizontal (see Figure
4.2).
Note:
Σ indicates summation over all members in one leg bay
Rev 3, August 2008
The above expression for CDei may be simplified for horizontal and vertical members as
follows:
Vertical members (e.g. chords): CDei = CDi (Di/De)
Horizontal members: CDei = sin3α CDi (Dili/Des)
Figure 4.2: Flow angles appropriate to a lattice leg

(after DNV Class Note 31.5, February 1992, [6])
4.6.6 The equivalent value of the inertia coefficient, CMe, and the equivalent area, Ae, to be
used in Section 4.3.3, representing the bay may be chosen as:
CMe = equivalent inertia coefficient which may normally be taken as 2.0

when using Ae
Ae = equivalent area of leg per unit height = (ΣAili)/s
Ai = equivalent area of element = πDi2/4
Di = reference diameter chosen as defined in Section 4.7
For a more accurate model the CMe coefficient may be determined as:
CMe Ae = Ae Σ CMei
where;
Aili
CMei = [1 + (sin2βi + cos2βi sin2αi)(CMi - 1)]
Aes
CMi = the inertia coefficient of an individual member, CMi is defined in
Section 4.7 related to reference dimension Di.
Note:
For dynamic modeling the added mass of fluid per unit height of leg may be determined
as ρAi(Cmi - 1) for a single member or ρAe(CMe - 1) for the equivalent model, provided
that Ae is as defined above.
Rev 3, August 2008
4.7 Hydrodynamic Coefficients for Leg Members
4.7.1 Hydrodynamic coefficients for leg members are given in this Section. Tubulars,
brackets, split tube and triangular chords are considered. Hydrodynamic coefficients
including directional dependence are given together with a fixed reference diameter Di.
No other diameter should be used unless the coefficients are scaled accordingly. Unless
better information is available for the computation of wave and current forces, the values
of drag and inertia coefficients applicable to Morison's equation should be obtained from
this Section.
4.7.2 Recommended values for hydrodynamic coefficients for tubulars (<1.5m diameter) are
given in Table 4.3 based on the data discussed in the commentary.
Surface condition CDi CMi

Smooth ⎫⎪ 0.65 2.0
⎬ See Note
Rough ⎪⎭ 1.00 1.8
Table 4.3: Base hydrodynamic coefficients for tubulars
Note:
The smooth values will normally apply above MWL + 2m and the rough values below
MWL + 2m, where MWL is as defined in Section 3.7.2. If the jack-up has operated in
deeper water and the fouled legs are not cleaned the surface should be taken as rough for
wave loads above MWL + 2m. See Commentary.
4.7.3 When applicable, marine growth is to be included in the hydrodynamic model by adding
the appropriate marine growth thickness, to, on the boundary of each individual member
below MWL + 2m where MWL is as defined in Section 3.7.2 i.e. for a tubular Di =
Doriginal + 2tm. Site specific data for marine growth is preferred (see Section 3.9). If such
data are not available all members below MWL + 2m shall be considered to have a
marine growth thickness tm = 12.5 mm (i.e. total of 25 mm across the diameter of a
tubular member). Marine growth on the teeth of elevating racks and protruding guided
surfaces of chords may normally be ignored.
The effects of marine growth may be ignored if anti-fouling, cleaning or other means are
applied, however the surface roughness is still to be taken into account (see
Commentary).
Rev 3, August 2008
4.7.4 The in-line force due to gussets in any vertical plane shall be determined using a drag
coefficient:
CDi = 2.0
applied together with the projected area of the gusset visible in the flow direction, unless
model test data shows otherwise. This drag coefficient may be applied together with a
reference diameter Di and corresponding length li chosen such that their product equals
the plane area, A = Dili and Di = li (see Figure 4.3). In the equivalent model of Section
4.6 the gussets may then be treated as a horizontal element of length li , with its axis in
the plane of the gusset. CMi should be taken as 1.0 and marine growth may be ignored.
Figure 4.3: Gusset plates

Rev 3, August 2008
4.7.5 For non-tubular geometries (e.g. leg chords) the appropriate hydrodynamic coefficients
may, in lieu of more detailed information, be taken in accordance with Figures 4.4 or 4.5
and corresponding formulas, as appropriate.
Figure 4.4: Split tube chord and typical values for CDi
For a split tube chord as shown in Figure 4.4, the drag coefficient CDi related to the
reference dimension Di = D+2tm, the diameter of the tubular including marine growth as
in Section 4.7.3, may be taken as:
⎧C D 0 ; 0° < θ ≤ 20°
⎪
CDi = ⎨
⎪⎩C D 0 + ( C D1 W / D i − C D 0 )Sin 2 [(θ − 20° )9 / 7] ; 20° < θ ≤ 90°
where;
θ = Angle in degrees, see Figure 4.4
CD0 = The drag coefficient for a tubular with appropriate roughness, see Section 4.7.2.
(CD0 = 1.0 below MWL+2m and CD0 =0.65 above MWL+2m.)
CD1 = The drag coefficient for flow normal to the rack (θ = 90°), related to projected
diameter, W. CD1 is given by:
⎧⎪18. ; W / D i < 12
.
CD1 = ⎨14 . + 13 ( W / D i ) ; . <
12 W / D i < 18
.
⎪⎩2.0 ; . <
18 W / Di
The inertia coefficient CMi = 2.0, related to the equivalent volume πDi2/4 per unit length
of member, may be applied for all heading angles and any roughness.
Rev 3, August 2008
Figure 4.5: Triangular chord and typical values of CDi
For a triangular chord as shown in Figure 4.5, the drag coefficient CDi related to the
reference dimension Di = D, the backplate width, may be taken as:
CDi = CDpr(θ) Dpr(θ) / Di
where the drag coefficient related to the projected diameter, CDpr, is determined from:
⎧170. ; θ = 0°
⎪⎪195
. ; θ = 90°
CDpr = ⎨140. ; θ = 105°
⎪165
. ; θ = 180°−θ o
⎪⎩2.00 ; θ = 180°
Linear interpolation is to be applied for intermediate headings. The projected diameter,

Dpr(θ), may be determined from:
⎧⎪D cos(θ ) ; 0 < θ < θo
Dpr(θ) = ⎨W sin(θ ) + 0.5D|cos(θ )| ; θ o < θ < 180 − θ o
⎪⎩D|cos(θ )| ; 180 - θ o < θ < 180
The angle θo, where half the rackplate is hidden, θo = tan-1(D/(2W)).
The inertia coefficient CMi = 2.0 (as for a flat plate), related to the equivalent volume of
πDi2/4 per unit length of member, may be applied for all headings and any roughness.
4.7.6 Shapes, combinations of shapes or closely grouped non-structural items which do not
readily fall into the above categories should be assessed from relevant literature
(references to be provided) and/or appropriate interpretation of (model) tests. The model
tests should consider possible roughness, Keulegan-Carpenter and Reynolds number
dependence.
4.8 Other Considerations
Local load effects will normally have been addressed at the design stage. Should the
wind or current and/or wave height parameters at the location exceed those applicable at
the design stage further consideration may be required. The Commentary provides
further details and references to calculation methods.
Rev 3, August 2008
4 GLOSSARY OF TERMS - CALCULATION METHODS

HYDRODYNAMICS AND WIND FORCES
AWi = Projected area of the block considered in wind computations.

A = Cross sectional area of member.
Ae = Equivalent area of leg per unit height = (∑ Ai 1i )/s.
Ai = Equivalent area of element = π Di2 /4.
CD = Drag coefficient.
CDe = Equivalent drag coefficient.
CDi = Drag coefficient of an individual member, related to Di.
CD0 = The drag coefficient for chord at direction θ = 0°.
CD1 = The drag coefficient for flow normal to the rack, θ = 90°.
CDpr = The drag coefficient related to the projected diameter.
CM = Inertia coefficient.
CMe = Equivalent inertia coefficient.
CMi = Inertia coefficient of a member, related to Di.
Ch = Height coefficient for wind.
Cs = Shape coefficient for wind related to projected area.
d = The mean, undisturbed water depth (positive).
D = Member diameter or backplate width.
De = Equivalent diameter of leg.
Di = Reference dimension of individual leg members.
D1 = Face width of leg, outside dimensions.
Dpr = The projected diameter.
FWi = Wind force for block i.
Hs = The effective significant wave height (Section 5.5.1.3).
li = Length of member 'i' node to node center.
Pi = Wind pressure at the center of block i.
r n = Velocity of the considered member, normal to the member axis and in the
direction of the combined particle velocity.
s = Length of one bay, or part of bay considered.
tm = Marine growth thickness.
Tn = First natural period of sway motion.
Tz = The zero-upcrossing period associated with Hs.
u = Wave particle velocity.
un = Wave (only) particle velocity normal to the member.
u n = Wave particle acceleration normal to the member.
vn = Total (relative) flow velocity normal to the member.
VCn = Current velocity to be used in the hydrodynamic model, normal to member.
Vf = Far field (undisturbed) current.
Vref = One minute sustained wind velocity at elevation Zref.
VZ = Wind velocity at elevation Z.
W = Dimension from backplate to pitch point of triangular chord or dimension from
root of one rack to tip of other rack of split-tubular chord.
z = Coordinate measured vertically upward from the mean water surface.
z' = Modified coordinate to be used in particle velocity formulation.
Z = Elevation measured from the mean water surface.
Zref = Reference elevation for wind speed.
α = Indicator for relative velocity, 0 or 1.
αi = Angle defining flow direction relative to member.
βi = Angle defining the member inclination.
ΔFdrag = Drag force per unit length.
ΔFinertia = Inertia force per unit length.
Rev 3, August 2008
4 GLOSSARY OF TERMS - CALCULATION METHODS
HYDRODYNAMICS AND WIND FORCES (Continued)
ζ = The instantaneous water surface elevation (same axis system as z).

ρ = Mass density of water or air.
θ = Angle in degrees of water particle velocity relative to the chord orientation.
θo = Angle at which half rackplate of Δ chord is hidden = tan-1 (D/(2W))
λ = Wave length.
Rev 3, August 2008
5 CALCULATION METHODS - STRUCTURAL ENGINEERING
5.1 General Conditions
5.1.1 Structural calculations should be carried out in accordance with the following sections.
5.1.2 A range of environmental approach directions and storm water levels should be
considered, such that the most onerous (i.e. that leading to the extreme maximum and/or
minimum loading) is determined for each assessment check {strength of each major type
of element (chord, brace, etc.), overturning stability, foundation capacity, horizontal
deflections, holding system, etc.}.
5.1.3 In deterministic calculations the most critical wave phase position(s) should be
considered for each case identified under 5.1.2. Normally the phase giving maximum
base shear and/or overturning moment will be found critical for overturning, leeward leg
stresses, leeward leg foundations and windward leg foundations.
5.1.4 For fatigue calculations it may be necessary to determine the load or stress ranges, and
hence other phase positions may also need to be considered.
5.2 Seabed Reaction Point
For independent leg jack-up units, the reaction point for horizontal and vertical loads at
each footing shall be situated on the geometric vertical axis of the leg/spudcan, at a
distance above the spudcan tip equivalent to:
a) Half the maximum predicted penetration (when spudcan is partially penetrated), or
b) Half the height of the spudcan (when the spudcan is fully, more than fully
penetrated).
If detailed information exists regarding the soils and spudcan the position of the reaction
point may be calculated. (Brekke et al, [7])
5.3 Foundation Fixity
5.3.1 For analyses of an independent leg jack-up unit under extreme storm conditions the
foundations may normally be assumed to behave as pin joints, and so are unable to
sustain a bending moment. Analysis and practical experience suggest that this may be a
conservative approach for bending moment in the upper parts of the leg in way of the
lower guides.
5.3.2 In cases where the inclusion of rotational foundation fixity is justified and is included in
the structural analysis, it is essential that the nonlinear soil-structure interaction effects
are properly taken into account. The model should include the interaction of rotational,
lateral and vertical soil forces.
Rev 3, August 2008
5.3.3 Methods of establishing the degree of fixity of rotational restraint, or fixity, at the
footings are discussed further in Section 6.3.4 and the Commentary to Section 6. Upper
or lower bound values should be considered as appropriate for the areas of the structure
under consideration.
5.3.4 For checking the spudcans, the leg-to-can connection and the lower parts of the leg,
appropriate calculations considering soil-structure interaction shall be carried out to
determine the upper bound can moment. These areas may be checked assuming that a
percentage of the maximum storm leg moment at the lower guide (derived assuming a
pinned footing) is applied to the spudcan together with the associated horizontal and
vertical loads. This percentage would normally be not less than 50%. For such
simplified checks the loading on the spudcan may be modeled assuming that the soil is
linear-elastic and incapable of taking tension.
5.4 Leg Inclination
The effects of initial leg inclination should be considered. Leg inclination may occur due
to leg-hull clearances and the hull inclination permitted by the operating manual. Thus
the total horizontal offset due to leg inclination, OT, may be determined as:
OT = O1 + O2
where;
OT = Total horizontal offset of leg base with respect to hull.
O1 = Offset due to leg-hull clearances.
O2 = Offset due to maximum hull inclination permitted by the operating manual.
If detailed information is not available, OT should be taken as 0.5% of the leg length
below the lower guide.
The effects of leg inclination need be accounted for only in structural strength checks.
This will normally be accomplished by increasing the effective moment in the leg at the
lower guide by an amount equal to the offset OT times the factored vertical reaction at the
leg base due to dead, live, environmental, inertial and P-Δ loads.
5.5 P-Δ Effects
5.5.1 The P-Δ Effect occurs because the jack-up is a relatively flexible structure and is subject
to lateral displacement of the hull (sidesway) under the action of environmental loads.
As a result of the hull translation the line of action of the vertical spudcan reaction no
longer passes through the centroid of the leg at the level of the hull. Consequently the leg
moments at the level of the hull are increased over those arising from a linear quasi- static
analysis by an amount equal to the individual leg load P times the hull translation, D.
This additional moment will cause additional deflection over that predicted by standard
linear-elastic theory. The increased deflection is a function of the ratio of the applied
axial load to the Euler load.
Rev 3, August 2008
Furthermore the shift in the hull center of gravity due to the hull translation will increase
the overturning moment (or decrease the righting moment). Consequently the axial loads
in the leeward leg(s) will increase and the axial loads in the windward leg(s) will reduce.
The consequences of the above are:
a) Increased hull deflections (which will increase the linear-elastic P-Δ moments).
b) A redistribution of base shears (in global axes) such that the increase in lower guide
moment is reduced in the leeward leg(s) and increased in the windward leg(s).
5.5.2 An analysis using a standard linear elastic (small displacement) finite element program
will not allow for these effects. The following Sections describe techniques which may
be used to account for the P-Δ/Euler effects. The large displacement methods are the
most accurate, but require more rigorous analysis. The geometric stiffness methods are
simpler and generally of sufficient accuracy.
5.5.3 Large displacement methods;
These methods are part of a number of finite element (F.E.) programs. In such methods
the non-linear (large-displacement) solution is obtained by applying the load in
increments and iteratively generating the stiffness matrix for the next load increment
from the deflected shape (nodal deflections) of the previous increment. Some F.E.
programs offer an intermediate solution in which the deflected geometry from an initial
linear- elastic solution is used as the input to the final 'corrected' solution.
5.5.4 Geometric stiffness methods;
5.5.4.1 These methods are also available within a number of F.E. programs. A linear correction
is made to the element stiffness matrix based on the axial load present in the element.
Iteration is also required for this solution procedure.
5.5.4.2 A simplified geometric stiffness approach allows incorporation of P-Δ effects in a

standard linear-elastic F.E. program without recourse to iteration (refer to Commentary
for derivation). In this approach a correction term is introduced into the global stiffness
matrix prior to analysis. When the analysis is complete the hull deflections, leg axial
loads and leg bending moments will include the P-Δ effects. The derivation of the
method is described in appendix C5.A of the Commentary.
The correction term is: -Pg/L

where;
Pg = Total effective gravity load on legs at hull. This includes the hull weight
and weight of the legs above the hull.
L = The distance from the spudcan reaction point to the hull vertical center of
gravity.
Rev 3, August 2008
This single (negative) value is incorporated into the global stiffness matrix by attaching a
pair of orthogonal horizontal translational earthed spring elements to a node representing
the hull center of gravity and entering the negative value for each of the spring constants.
Some F.E. packages allow direct matrix manipulation.
The negative stiffness term at the hull will produce an additional lateral force at the hull
proportional to the structural deflection. The resulting (additional) base overturning
moment will be equal to the gravity load times the hull displacement.
The additional lateral load (due to the negative stiffness term) will cause an over-
prediction of the base shear (in global axes). Typically this is not critical. However, the
base shear at each leg can be reduced by an amount equal to the difference between the
total base shear and the shear due to the applied loads (both in global axes) divided by the
number of legs.
5.5.4.3 An alternative geometric stiffness approach is given below. Here the P-Δ effects are
determined by amplifying the linear-elastic displacement (excluding P-Δ) as follows:
P
Δ = δs / (1 - )
PE
where;
Δ = the approximate displacement including P-Δ.

δs = the linear-elastic first order hull displacement.
P = the average axial load in the leg at the hull (i.e. the total leg load at the hull
divided by the number of legs).
PE = Euler buckling load of an individual leg.(See Section 7.3.5 for general
formulation).
Corrections can then be made to a global linear-elastic solution by manually adding P-Δ
moments to the results. The P-Δ moments are computed using the amplified deflection, Δ, and
P's adjusted to account for this. (This approach is not strictly valid because it ignores the fact
that the deflection of all the legs at the hull must be approximately equal. The imposition of this
constraint will lead to a redistribution of the global base shear between the legs.) Ignoring the
redistribution will generally be conservative for leeward leg(s) and their foundation loads and
non- conservative for windward leg(s) and their foundation loads.
Rev 3, August 2008
5.6 Structural Modeling
5.6.1 Introduction
It is important that the structural model accurately reflects the complex mechanism of the jack-
up. For most jack-up configurations the load distribution at the leg-hull interface is not amenable
to manual calculation, therefore, it is necessary to develop a Finite Element (F.E.) computer
model. A number of different modeling techniques can be used to depict the jack-up structure.
The recommended techniques are summarized below and their applicability and limitations are
discussed in more detail in Section 5.6.3.
a) Fully detailed model of legs and hull/leg connections with detailed or representative
stiffness model of hull and spudcan.
b) Simplified lower legs and spudcans, detailed upper legs and hull/leg connections with
detailed or representative stiffness model of hull.
c) Equivalent stiffness model of legs and spudcans, equivalent hull/leg connection

springs and representative beam-element hull grillage.
d) Detailed leg (or leg section) and hull/leg connection model.
Section 5.6.3 and Table 5.1 outline the limitations of the various modeling techniques and
should be referenced to ensure that the selected models address all aspects required for a
specific assessment.
5.6.2 General Considerations
In the elevated condition the most heavily loaded portion of the leg is normally between
the upper and lower guides and in way of the lower guide. The stress levels in this area
depend on the design concept of the jack-up. A specific jack-up design concept can be
described by the combination of the following components (see Commentary Figure
C5.5):
a) With or without fixation system,
b) Fixed or floating jacking system,
c) Opposed or unopposed pinions.
In units having fixation systems the transfer of moment between the leg and the hull is
largely by means of a couple due to vertical loads carried from the chord into the fixation
or jacking system.
Where a fixed or floating jacking system is fitted (and there is no fixation system) the transfer of
moment between the leg and the hull is partly by means of a couple due to horizontal loads
carried from the chords into the upper and lower guides. In this case and when the chord/guide
contact occurs between bracing nodes significant local chord bending moments are normal.
Rev 3, August 2008
If the jacking system has unopposed pinions local chord moments will arise due to:
- the horizontal pinion load component (due to the pressure angle of the
rack/pinion).
- the vertical pinion load component acting at an offset from the chord neutral axis.
The modeling of the various design aspects is critical and recommended modeling
techniques are outlined in the following sections. The Commentary provides detailed
information regarding the combination of the above three components for current jack-up
units.
5.6.3 Applicability and Limitations
It is most desirable to fully model the jack-up when assessing its structural strength.
Very often assumptions and simplifications such as equivalent hull, equivalent leg, etc.
will be made in the process of building the model. In view of this, various levels of
modeling described in a) through d) below may be used. It should be noted that some of
these methods may have limitations with respect to the accuracy of assessing the
structural adequacy of a jack-up and when simplified models, such as those described in
(c) and (d) are used it may be appropriate to calibrate against a more detailed model.
a) Fully detailed 3-leg model
The model consists of 'detailed legs', hull, hull/leg connections and spudcans modeled in
accordance with 5.6.4(a),5.6.5, 5.6.6 and 5.6.7, respectively. The results from this model
can be used to examine the preload requirements, overturning resistance, leg strength and
the adequacy of the jacking system or fixation system.
b) Combination leg 3-leg model
The model consists of a combination of 'detailed leg' for the upper portion of legs and
'equivalent leg' for the lower portion of the legs modeled in accordance with 5.6.4. The
hull, hull/leg connections and spudcans are modeled in accordance with 5.6.5, 5.6.6 and
5.6.7 respectively. The results from this model can be used to examine the preload
requirements, overturning resistance, leg strength and the adequacy of the jacking system
or fixation system.
c) Equivalent 3-stick-leg model
The model consists of 'equivalent legs' modeled in accordance with 5.6.4(b), hull
structure modeled using beam elements in accordance with 5.6.5, leg to hull connections
modeled in accordance with 5.6.6 and spudcans modeled as a stiff or rigid extension to
the equivalent leg. The results from this model can be used to examine the preload
requirements and overturning resistance. This model may also used to obtain the
reactions at the spudcan or internal forces and moments in the leg at the vicinity of lower
guide for application to the 'detailed leg' and hull/leg model (d) which should be used to
assess the strength of the leg in the area between lower and upper guides.
Rev 3, August 2008
d) Single detailed leg model
The model consists of a 'detailed leg' or a portion of a 'detailed leg' modeled in

accordance with 5.6.4(a), the hull/leg connection modeled in accordance with 5.6.6 and,
when required, the spudcan modeled in accordance with Section 5.6.7. This model is to
be used in conjunction with the reactions at the spudcan or the forces and moments in the
vicinity of lower guide obtained from Model (c). The results from this model can be used
to examine the leg strength and the adequacy of the jacking system or the fixation system.
Applicability (see Note 1)

I II III IV V VI VII
Global Leg Pinion/ Hull
Model Global Overturning Foundation Leg Member Fixation Element
Type Loads Checks Checks Loads Loads System Loads Loads
a Y Y Y Y Y Y 2
b Y Y Y Y Y Y 2
c Y Y Y Y - - -
d - - - - Y Y -
Legend:
Y = Applicable
- = Not applicable
Notes:
1. Large displacement and dynamic effects to be included where appropriate.
2. VII, hull stresses will only be available from more complex hull models.
Table 5.1 - Applicability of the suggested models
5.6.4 Modeling the Leg
The leg can be modeled as a 'detailed leg', an 'equivalent leg' or a combination of the two.
The 'detailed leg' model consists of all structural members such as chords, horizontal,
diagonal and internal braces of the leg structure and the spudcan (if required). The
'equivalent leg' model consists of a series of colinear beam elements (stick model)
simulating the complete leg structure. It is recommended that the leg model(s) be
generated in accordance with the following:
a) 'Detailed Leg' Model
The coordinates of the joints for this model are to be defined by the intersection of the
chord and brace centerlines. For joints where there is more than one brace, it is unlikely
that there will be one (1) common point of intersection between the braces and chord. In
this instance, it is usually sufficient to choose an intermediate point between the
chord/brace centerline intersections. Gusset plates normally need not be included in the
structural leg model, however their effects may be taken into account in the calculation of
member and joint strength checks.
Rev 3, August 2008
b) 'Equivalent Leg' Model
The leg structure can be simulated by a series of colinear beams with the equivalent cross
sectional properties calculated using the formulas indicated in Figure 5.1 or derived from
the application of suitable 'unit' load cases (see Commentary C5.5) to the 'Detailed Leg'
model described in 5.6.4 (a). Where such a model is used, detailed stresses, pinion loads,
etc. will be derived either directly or indirectly from a 'detailed model'.
c) 'Combination Leg' model
To facilitate obtaining detailed stress, pinion loads, etc. directly, a 'detailed leg' model can
be generated covering the region between the guides, and extending at least 4 bays below
and, where available, at least 4 bays above this region. The remainder is then modeled as
an 'equivalent leg'. Care is required to ensure an appropriate interface and consistency of
boundary conditions at the connections. The 'detailed leg'/'equivalent leg' connection
should be modeled so that the plane of connection remains a plane after the leg is bent.
Note:
The leg stiffness used in the overall response analysis may account for a contribution
from a portion of the rack tooth material. Unless detailed calculations indicate otherwise,
the assumed effective area of the rack teeth should not exceed 10% of their maximum
cross sectional area. When checking the capacity of the chords the chord properties
should be determined discounting the rack teeth.
5.6.5 Modeling the Hull
The hull structure should be modeled so that the loads can be correctly transferred to the
legs and the hull flexibility is represented accurately. Recommended methods are given
below:
a) Detailed Hull Model

The model can be generated using plate elements in which appropriate directional
modeling of the effect of the stiffeners on the plates should be included. The
elements should be capable of carrying in-plane and, where applicable, out-of plane
loads.
b) Equivalent Hull Model

Alternatively, the hull can be modeled by using a grillage of beams. Deck, bottom,
side shell and bulkheads can be used to construct the grillage. The properties of the
beam can be calculated based on the depth of the bulkheads, side-shell and the
'effective width' of the deck and bottom plating. Attention should be paid to the in-
plane and torsional properties due to the continuity of the deck and bottom structures.
Rev 3, August 2008
5.6.6 Modeling the Hull/Leg Connection
The hull/leg connection modeling is of extreme importance to the analysis since it

controls the distribution of leg bending moments and shears carried between the upper
and lower guide structures and the jacking or fixation system. It is therefore necessary
that these systems are properly modeled in terms of stiffness, orientation and clearance.
For the 'Equivalent 3-stick-leg model' a simplified derivation of the equivalent leg-hull
connection stiffness may be applicable.
For jack-ups with a fixation system, the leg bending moment will be shared by the upper
and lower guides, the jacking and the fixation systems. Normally the leg bending
moment and axial force due to environmental loading are resisted largely by the fixation
system because of its high rigidity. Depending on the specified method of operation, the
stiffnesses, the initial clearances and the magnitude of the applied loading a portion of the
environmental leg loading may be resisted by the jacking system and the guide structures.
Typical shear force and bending moment diagrams for this configuration are shown in
Figure 5.2.
For jack-ups without a fixation system, the leg bending moment will be shared by the
jacking system and guide structure. For a fixed jacking system, the distribution of leg
moment carried between the jacking system and guide structure mainly depends on the
stiffness of the jacking pinions. Typical shear force and bending moment diagrams for
this design are shown in Figures 5.3 and 5.4.
For a floating jacking system, the distribution of leg bending moment carried between the
jacking system and guide structure depends on the combined stiffness of the shock pads
and pinions. Typical shear force and bending moment diagrams for this design are
shown in Figure 5.5.
The hull/leg connection should be modeled considering the effects of guide and support
system clearances, wear, construction tolerances and backlash (within the gear-train and
between the drive pinion and the rack).
The following techniques are recommended for modeling hull/leg connections (specific
data for the various parts of the structure may be available from the designers data
package):
Detailed modeling
a) Upper and Lower Guides - The guide structures should be modeled to restrain the
chord member horizontally only in directions in which guide contact occurs. The
upper and lower guides may be considered to be relatively stiff with respect to the
adjacent structure, such as jackcase, etc. The nominal lower guide position relative to
the leg may be derived using the sum of leg penetration, water depth and airgap. It is
however recommended that at least two positions are covered when assessing leg
strength: one at a node and the other at the midspan. This is to allow for
uncertainties in the prediction of leg penetration and possible differences in
penetration between the legs.
Rev 3, August 2008
The finite lengths of the guides may be included in the modeling by means of a
number of discrete restraint springs/connections to the hull. Care is required to
ensure that such restraints carry loads only in directions/senses in which they can act.
Alternatively the results from analyses ignoring the guide length may be corrected, if
necessary, by modification of the local bending moment diagram to allow for the
proper distribution of guide reaction, see Figure 5.6.
b) Jacking Pinions - The jacking pinions should be modeled based on the pinion
stiffness specified by the manufacturer and should be modeled so that the pinions can
resist vertical and the corresponding horizontal forces. A linear spring or cantilever
beam can be used to simulate the jacking pinion. The force required to deflect the
free end of the cantilever beam a unit distance should be equal to the jacking pinion
stiffness specified by the manufacturer. The offset of the pinion/rack contact point
from the chord neutral axis should be incorporated in the model.
c) Fixation System - The fixation system should be modeled to resist both vertical and
horizontal forces based on the stiffness of the vertical and horizontal supports and on
the relative location of their associated foundations. It is important that the model can
simulate the local moment capacity of the fixation system arising from its finite size
and the number and location of the supports.
d) Shock Pad - Floating jacking systems generally have two sets of shock pads at each
jackcase, one located at the top and the other at the bottom of the jackhouse.
Alternatively shock pads may be provided for each pinion. The jacking system is free
to move up or down until it contacts the upper or lower shock pad. In the elevated
condition, the jacking system is in contact with the upper shock pad and in the transit
condition it is in contact with the lower shock pad. The stiffness of the shock pad
should be based on the manufacturer's data and the shock pad should be modeled to
resist vertical force only. It should also be noted that the shock pad stiffness
characteristics may be nonlinear.
e) Jackcase and associated bracing - The jackcase and associated bracing should be
modeled based on the actual stiffness since it has direct impact on the horizontal
forces that the upper guide can resist.
Note:
Where the hull is not modeled it is normally suitable to earth the base of the jackcase and
associated bracing, the foundations of the fixation system and the lower guide structures
at their connections to the hull.
Simple modeling
f) For applications such as those described in Section 5.6.3 c) (Equivalent 3-stick-leg

model) a simplified representation of the hull to leg connection is required. In this
instance the rotational stiffness may be represented by rotational springs and, where
applicable, horizontal and vertical stiffnesses by linear springs. Where these are
derived from a more detailed modeling, as described above, it is important that
suitable loading levels (typical of the cases to be analyzed) are selected so that the
effects of clearances, etc. do not dominate the result. Hand calculations may also be
applicable. See Section C5.5 in the Commentary.
Rev 3, August 2008
5.6.7 Modeling the Spudcan
When modeling the spudcan, rigid beam elements are considered sufficient to achieve an
accurate load transfer of the seabed reaction into the leg chords and bracing in the area
between upper and lower guides. It should be noted that, due to the sudden change in
stiffness, rigid beams can cause artificially high stresses at the leg to spudcan
connections. Hence the modeling and selection of element type should be carefully
considered when an accurate calculation of chord stresses is required in this area.
For a strength analysis of the spudcan and its connections to the leg it may be appropriate
to develop a separate detailed model with appropriate boundary conditions.
5.7 Load Application
The assessment follows a partial factor format. The partial load factors are applied to
loads as defined in other sections (i.e. they are load factors, NOT load-effect factors).
The jack-up response is non-linear, and hence the application of the combined factored
loads will not in general develop the same result as the factored combination of
individual load effects.
For typical jack-up assessments, the time-varying nature of the wave loading will
amplify the static responses and must be considered. The extreme response can be
assessed either by a quasi-static analysis procedure (Section 7.2) including an inertial
loadset (Section 7.3.6) or by a more detailed dynamic analysis procedure (Section 7.3.7).
In the former case (quasi-static analysis including an inertial loadset), the load factors
should be directly applied to the appropriate combinations of quasi-static environmental
loading and inertial loadsets. In the latter case (detailed dynamic analysis), alternative
methods can be used when acceptable rationale is provided.
The loads and load effects to be included in the analysis, with their designators used in
Section 8 in ( ), comprise:
a) Self weight and non-varying loads (D), variable and drilling loads (L).
b) Wind loads (E).
c) Hydrodynamic wave-current loads (E).
d) Inertial loads due to dynamic response (Dn).
e) Second order effects (associated with D,L,E & Dn).
These are discussed in turn below.

Rev 3, August 2008
5.7.1 Self weight, variable and drilling loads

Depending on the initial positions of the legs with respect to guide clearances, and the
operation of the jacking and fixation systems (if fitted), the distributed hull loading and
stiffness will lead to hull sagging which may impose bending moments on the legs which
remain present for the remainder of the period on location. Such moments should be
considered in the site assessment analyses, and will be larger in shallow waters where the
leg extension below the hull is small and consequently the leg bending stiffness is higher.
To correctly capture these effects the hull loads should be applied to the model in such a
manner as to represent their correct vertical and horizontal distribution. If dynamic
analyses are to be performed all weights should be represented by means of masses
together with vertical gravitational acceleration. It is generally appropriate to apply these
masses by means of factored element self-weight with additional correction masses
applied as necessary to obtain the correct total mass and center of gravity. Alternatively,
it may be sufficient to apply point masses at the node points of the model.
It is noted that an F.E. model with distributed hull stiffness and loading will incorporate
hull sag effects if the hull and variable gravity loading is 'turned on' with the unit defined
in its initially undeflected shape at the operating airgap. It should be verified that the
amount of hull sag moment arising is applicable, given the operating procedures
pertaining to the unit. It may be necessary to apply corrections to the final results for any
discrepancies in the hull sag induced loadings. Further guidance is given in Section 5.3.3
of the Commentary.
5.7.2 Wind loads
The wind loading on the legs above and below the hull may be applied as distributed or
nodal loads. Where nodal loads are used a sufficient number of loads should be applied
to reflect the distributed nature of the loading and it should be ensured that the correct
total shear and overturning moment is applied on each leg. Similarly the wind loading on
the hull and associated structure may be applied as distributed or nodal loads. The
application should ensure the correct total shear and overturning moment is applied to the
hull.
5.7.3 Hydrodynamic wave-current loads
The wave-current loading on the leg and spudcan structures above the mudline may be
applied as distributed or nodal loads. Where nodal loads are used the application should
ensure the correct total shear and overturning moment on each leg, and reflect the
distributed nature of the loading.
5.7.4 Inertial loads due to dynamic response
When the dynamic approach (see Section 7) leads to the explicit determination of an
inertial loadset, this should be applied to the hull model. In simpler dynamic approaches
the inertial load may be represented by a single lateral point loading acting at the hull
center of gravity, or by a number of point loads applied to other parts of the hull having
the same line of action. In more complex approaches a more complete distributed load
vector may be applied to the hull and legs.
5.7.5 Second order effects

Methods for including P-Δ effects are described in Section 5.5.
Rev 3, August 2008
Figure 5.1: Formulas for the determination of equivalent member properties;(After DNV Class
Note 31.5 1992 [6] (corrected)
Rev 3, August 2008
Figure 5.2: Leg shear force and bending moment - jack-ups with a fixation system
Rev 3, August 2008
Figure 5.3: Leg shear force and bending moment - jack-ups without a fixation system
and having a fixed jacking system with opposed pinions
Rev 3, August 2008
Figure 5.4: Leg shear force and bending moment - jack-ups without a fixation system
and having a fixed jacking system with unopposed pinions
Rev 3, August 2008
Figure 5.5: Leg shear force and bending moment - jack-ups without a fixation system and
having a floating jacking system
Rev 3, August 2008
Figure 5.6: Correction of point supported guide model for finite guide length
(After DNV Class Note 31.5, 1992 [6])
Rev 3, August 2008
5 GLOSSARY OF TERMS - STRUCTURAL ENGINEERING
A = Equivalent axial area of a leg for stiffness calculations.

ACi = Area of chord including a contribution from the rack teeth (see note to Section
5.6.4.)
AD = Axial area of an inclined brace.
AQi = Equivalent shear area of a leg face.
AQy = Equivalent shear area of a leg in y direction.
AQz = Equivalent shear area of a leg in z direction.
AV = Axial area of a brace perpendicular to the chords.
d = Length of inclined brace or face to face distance between chords for lattice
structures without inclined braces.
D = Self weight and non-varying loads.
Dn = Inertial loads due to Dynamic response.
E = Environmental loads.
h = Distance between chord centroids.
h = Length of guide.
IB = Second moment of area of 'brace'.
IG = Second moment of area of longitudinal girder.
IT = Equivalent torsional constant of leg about longitudinal axis.
IY = Equivalent second moment of area of leg about y-y axis for stiffness calculations.
Iz = Equivalent second moment of area of leg about z-z axis for stiffness calculations.
L = Variable loads.
L = Distance from the spudcan reaction point to the hull vertical center of gravity.
N = Number of bays, used in determination of equivalent shear area AQ.
OT = Total horizontal offset of leg base with respect to hull
= O1 + O2
O1 = Offset of leg base with respect to hull due to leg-hull clearances.
O2 = Offset of leg base with respect to hull due to maximum hull inclination permitted by
the operating manual.
P = Average axial load in the legs at the hull (total leg load divided by number of legs).
P = Guide reaction.
PE = Euler buckling load of an individual leg.
Pg = Total effective gravity load on legs at hull, including the hull weight and weight of
legs above hull.
s = Leg bay height (distance between brace nodes).
δs = Linear elastic first order displacement of hull.
Δ = Approximate hull displacement including P-Δ effects
= δs /(1 - P/PE)
v = Poissons ratio for the material
= 0.3 for steel.
Rev 3, August 2008
6 CALCULATION METHODS - GEOTECHNICAL ENGINEERING
6.1 Introduction
6.1.1 Section 6 addresses three groups of geotechnical areas of concern which are discussed in
the following subsections:
6.2 Prediction of footing penetration during preloading.
6.3 Jack-up foundation stability after preloading.
6.4 Other aspects of jack-up foundation performance during or after preloading.
6.1.2 Where geotechnical analyses are performed they should be based on geotechnical data
obtained from a site investigation incorporating soil sampling and/or in-situ testing (see
Section 3.16).
6.1.3 Uncertainties regarding the geotechnical data should be properly reflected in the
interpretation and reporting of analyses for which the data are used.
6.1.4 The majority of spudcans are effectively circular in plan but other spudcan geometries
are not uncommon. Typical spudcan designs are illustrated in Figure 6.1. The bearing
capacity formulas given in this section are consistent with 'circular' spudcan footings
without skin-friction on the leg. Due consideration should be given to the tapered
geometry of most spudcans for bearing capacity assessment.
Note: Terms which are not defined in the text may be found in the Glossary to this
Section.
6.2 Prediction of Footing Penetration During Preloading
6.2.1 Analysis Method
The conventional procedure for the assessment of spudcan load/penetration behavior is

given in the following steps:
1. Model the spudcan.
2. Compute the vertical bearing capacity of the footing at various depths below seabed
using closed form bearing capacity solutions and plot as a curve.
3. Enter the vertical bearing capacity versus footing penetration curve with the specified
maximum preload and read off the predicted footing penetration.
For conventional foundation analyses the spudcan can often be modeled as a flat circular
foundation. The equivalent diameter is determined from the area of the actual spudcan
cross section in contact with the seabed surface, or where the spudcan is fully embedded,
from the largest cross sectional area. Foundation analyses are then performed for this
circular foundation at the depth (D) of the maximum cross sectional area in contact with
the soil. (See Figure 6.2). Alternative shapes, e.g. tubular legs, should be treated as
appropriate.
Rev 3, August 2008
Figure 6.1: Typical spudcan geometries
Figure 6.2: Spudcan foundation model

Rev 3, August 2008
The depth of spudcan penetration is usually defined as the distance from the spudcan tip
to the mudline. It is therefore necessary to correct for this when referring to the
analytical foundation model.
The possibility of soil back-flow over the footing should be considered when computing
bearing capacity. In very soft clays complete back-flow may occur whereas in firm to
stiff clays and granular materials, where limited footing penetration may be expected, the
significance of back-flow diminishes.
Back-flow in clay may be assumed not to occur if:
Nc us
D≤
γ'
where, in this case, cus is taken as the average undrained cohesive shear strength over the
depth of the excavation, N is a stability factor and γ' is the submerged unit weight of the
soil.
Conservative stability factors in uniform clays, as a function of excavation depth and

diameter, are summarized in Figure 6.3. Alternative stability factors are given in the
Commentary. For spudcan penetration analyses it is recommended that conservative
criteria are used and the excavation depth be considered as the depth to the maximum
spudcan bearing area.
Both the bearing capacity analyses and the above back-flow analysis are based on simple
solutions developed for other geotechnical purposes or foundation conditions. These
differences should be recognized and are discussed further in the Commentary.
The equations given in the following sections may be considered with or without soil
back-flow over the footing. The additional load from back-flow on the footing increases
the maximum penetration. In general two cases can be distinguished:
- Immediate back-flow
- Hole side walls collapse after the installation phase.
For deeply penetrated footings the effect of side wall collapse after preloading will be to
significantly reduce the ultimate vertical bearing capacity of the foundation. Where
relevant this phenomenon should be considered.
For spudcan penetration analyses the ultimate vertical bearing capacity, FV, may be
determined at a series of spudcan penetration depths according to the criteria given in
Sections 6.2.2 to 6.2.6.
Rev 3, August 2008
Figure 6.3: Stability factors for cylindrical excavations in clay
6.2.2 Penetration in Clays
The ultimate vertical bearing capacity of a foundation in clay (undrained failure in clay,
φ = 0) at a specific depth can be expressed by:
FV = (cu.Nc.sc.dc + po')A.
The maximum preload is equal to the ultimate vertical bearing capacity, FV, taking into
account the effect of backflow, Fo'A, and the effective weight of the soil replaced by the
spudcan, γ'V (see Commentary) i.e.:
VLo = FV - F'oA + γ'V
See Figures 6.2 and 6.4 and note that the terms - F'oA + γ'V should always be considered
together.
It is recommended that the value of undrained cohesive shear strength, cu, is taken as the
average value over a distance B/2 from beneath the level where the maximum spudcan
diameter is in contact with the soil. (Refer to the Commentary).
The bearing capacity formula given above has been empirically derived for surface
foundations and does not account for foundation roughness, shape (conical for most
spudcans) or the effects of increased shear strength with depth. These factors are taken
into account in a method provided in the Commentary.
Note: It is recognized that the bearing capacity of a soil may reduce when subjected to cyclic
loading. (Refer to the Commentary.)
Rev 3, August 2008
Figure 6.4: Spudcan bearing capacity analysis
6.2.3 Penetration in Silica Sands
The ultimate vertical bearing capacity of a circular footing resting in silica sand or other
granular material can be computed by the following equation;
FV = (0.5 γ'B Nγ sγ dγ + po' Nq sq dq)A
The maximum preload is equal to the ultimate vertical bearing capacity, FV, taking into
account the effect of backflow, Fo'A, and the effective weight of the soil replaced by the
spudcan, γ'V (see Commentary) i.e.:
See Figures 6.4 and note that the terms -F'oA + γ'V should always be considered together.
Typically observed load-penetration data for large diameter spudcans suggest that
reduced friction angles may be applicable for this analysis method. To account for this
it is appropriate to reduce the laboratory derived φ by 5°. Further recommendations on
the selection of φ values are given in the Commentary together with a discussion
regarding the use of alternative bearing capacity factors.
6.2.4 Penetration in Carbonate Sands
Penetrations in carbonate sands are highly unpredictable and may be minimal in strongly
cemented materials, or large, in uncemented materials. Extreme care should be exercised
when operating in these materials. Further discussion regarding these soil conditions is
provided in the Commentary.
6.2.5 Penetration in Silts
It is recommended that upper and lower bound analyses for drained and undrained
conditions are performed to determine the range of penetrations. The upper bound
solution is modeled as a loose sand and the lower bound solution as a soft clay. Cyclic
loading may significantly affect the bearing capacity of silts. See discussion in
Commentary.
Rev 3, August 2008
6.2.6 Penetration in Layered Soils
Three basically different foundation failure mechanisms are considered in spudcan

predictions in layered soils:
1. General shear.
2. Squeezing.
3. Punch-through.
The first failure mechanism occurs if soil strengths of subsequent layers do not vary
significantly. Thus an average soil strength (either cu or φ) can be determined below the
footing. The footing penetration versus foundation capacity relationship is then
generated using criteria from Sections 6.2.2 through 6.2.5.
Criteria for the other two failure mechanisms (squeezing and punch-through) are given
below. The last condition is of particular significance since it concerns a potentially
dangerous situation where a strong layer overlies a weak layer and hence a small
additional spudcan penetration may be associated with a significant reduction in bearing
capacity.
6.2.6.1 Squeezing of clay
On a soft clay subject to squeezing overlaying a significantly stronger layer (Figure 6.5),
the ultimate vertical bearing capacity of a footing given by Meyerhof [8] is:
For no back-flow conditions:

bB 12 . D
FV = A{(a + + ) cu + po'} ≥ A{Nc sc dc cu + po'}
T B
and for full back-flow conditions:
bB 12 . D
FV = A{(a + + ) cu} + Vγ' ≥ A{Nc sc dc cu} + Vγ'
T B
where the following squeezing factors are recommended:
a = 5.00
b = 0.33
and cu refers to the undrained shear strength of the soft clay layer.
Figure 6.5: Spudcan bearing capacity analysis - squeezing clay layer

Rev 3, August 2008
It is noted that the lower bound foundation capacity is given by general failure in the clay
layer (right hand side of equation), and that squeezing occurs when B ≥
3.45T(1+1.1D/B). The upper bound capacity (for T<<B) is determined by the ultimate
bearing capacity of the underlying strong soil layer.
Comment on the limits included in the above relationships is provided in the

Commentary.
6.2.6.2 Punch-through: Two clay layers
The ultimate vertical bearing capacity of a spudcan on the surface of a strong clay layer
overlying a weak clay layer can be computed according to Brown [9]:
H
FV = A (3 cu,t + Nc sc cu,b) ≤ A Nc sc cu,t
B
See Figure 6.6.
For the evaluation of punch-through potential for deep footings, and to achieve
compatibility with the equations used for homogeneous clays, the following equations are
recommended:
For no back-flow conditions:

H D+H
FV = A {3 cu,t + Nc sc (1 + 0.2 ) cu,b + po'} ≤ A(Nc sc dc cu,t + po')
B B
and for full back-flow conditions:
H D+H
FV = A [3 cu,t + Nc sc (1 + 0.2 ) cu,b] + γ'V ≤ A Nc sc dc cu,t + γ'V
B B
Figure 6.6: Spudcan bearing capacity analysis - firm clay over weak clay
Rev 3, August 2008
6.2.6.2 It is noted that the condition (firm clay over soft clay) can also be "man-made" as in some
clays artificial crusts can form during delays in the installation procedure. Caution is
therefore required in situations where soil sampling/testing is performed from a jack-up
prior to preloading.
6.2.6.3 Punch-through: Sand overlying clay
The ultimate vertical capacity of a spudcan on a sand layer overlying a weak clay layer
can be computed using:
For no back-flow:
H
FV = FV,b - A Hγ' + 2 (Hγ' + 2 p'o) Kstanφ A
B
and for full or partial back-flow:
H
FV = FV,b - A Hγ' - A I γ' + 2 (Hγ' + 2 p'o) Kstanφ A
B
where;
FV,b is determined according to Section 6.2.2 assuming the footing bears on the
surface of the lower clay layer, with no back-flow.
See Figure 6.7.
The coefficient of punching shear, Ks, depends on the strength of both the sand layer and
the clay layer. For practical purposes a lower bound for the term Ks tanφ, applicable to
the onset of punch-through, can be approximated by:
Ks tanφ ≈ 3cu/Bγ'
An alternative analysis method is described in the Commentary.
Figure 6.7: Spudcan bearing capacity analysis - sand over clay

Rev 3, August 2008
6.2.6.4 Three Layered Systems
The foundation bearing capacity for a spudcan resting on three soil layers can be
computed using the squeezing and punch-through criteria for two layer systems. Firstly
the bearing capacity of a footing with diameter B resting on top of the lower two layers is
computed. These two layers can then be treated as one (lower) layer in a subsequent two
layer system analysis involving the (third) upper layer. For further explanation see
Figure 6.8.
Analysis 2
QV
Layer 1
Analysis 1
QV
Layer 2
Layer 3
Use 2 layer bearing capacity procedures for both analyses
Analysis 2 Analysis 1
Layer 1 over (Layer 2 and 3) Layer 2 over layer 3
Figure 6.8: Spudcan bearing capacity analysis - three layer case

Rev 3, August 2008
6.3 Foundation Stability Assessment
6.3.1 Approach
The overall foundation stability may be assessed using a phased method with three steps
increasing in order of complexity (See Figure 6.9):
- Step 1 Preload and Sliding Check (Section 6.3.2). The foundation capacity check
is based on the preloading capability. Sliding of the windward leg is also
checked. Loads from pinned footing analysis.
- Step 2 Bearing Capacity Check.
Step 2a Bearing capacity check (Section 6.3.3), based on resultant loading, assuming
a pinned footing. (see Section 5.3.1). Also check sliding.
Step 2b Bearing capacity check (Section 6.3.4), including rotational, vertical and
translational foundation stiffness.
- Step 3 Displacement Check (Section 6.3.5). The displacement check requires the
calculation of the displacements associated with an overload situation
arising from Step 2b.
Any higher level check need only be performed if the lower level check fails to meet the
foundation acceptance criteria given in Section 8.3.
The following sections give details regarding the three phased acceptance procedure.
However, there are certain aspects which are not covered in these sections which may
require further consideration. Some of the more common ones are listed below:
- Soils where the "long term" (drained) bearing capacity is less than the "short term"
(undrained) capacity. This may be the case for overconsolidated cohesive soils (silts
and clays) with significant amounts of sand seams.
- Where soil back-flows over the spudcan after the preload installation phase, (silts,
clays).
- If a reduction of soil strength due to cyclic loading occurs. This can be of particular
significance for silty soils and/or carbonate materials.
- If an increase in spudcan penetration occurs, due to cyclic loading, where a potential

punch-through exists.
- In soils with horizontal seams of weak soils located beneath the spudcan it is
recommended that the lateral bearing capacity/sliding stability of the foundation is
verified.
If any of the above circumstances exist further analysis is required.
In the case of partial spudcan embedment, (e.g. sandy soils), additional footing
embedment may result in a considerable increase in bearing capacity.
Rev 3, August 2008
Figure 6.9: Foundation stability assessment

Rev 3, August 2008
6.3.2 Ultimate bearing capacity for vertical loading - Preload Check (Step 1)
Except as discussed in 6.3.1, when the horizontal load is small, the ultimate vertical
bearing capacity under extreme conditions is assumed to be the same as the maximum
footing load during preloading, (VLo). The minimum requirements for VLo are given in
Section 8.3.1.3 or 8.3.2 as applicable.
6.3.3 Bearing Capacity/Sliding Check - Pinned footing (Step 2a)
A reduction in vertical bearing capacity, FV, of a footing occurs when it is simultaneously

subjected to horizontal loading, QH, and moment loading, QM. The latter is ignored in
Step 2a analyses as the footings are considered to be pinned.
The vertical/horizontal capacity envelope, FVH, for sands and clays may be generated
according to the following criteria, however, further discussion with regard to the
analytical applicability is provided in the Commentary.
6.3.3.1 Ultimate Vertical/horizontal bearing capacity envelopes for spudcan footings in sand:
The general ultimate vertical/horizontal bearing capacity envelope for jack-up footings in
sand is as follows:
FVH = A (0.5 γ' B Nγ sγ iγ dγ + po'Nq sq iq dq)
During the preloading phase it may be assumed that no horizontal load acts on the
foundation and that the ultimate vertical bearing capacity of the soil is in equilibrium with
the applied footing installation load, VLo. The applied footing installation load should
include the effect of back-flow and spudcan buoyancy i.e. VLo = FV - Fo'A + γ'V. In this
instance the inclination factors assume values of unity and the remaining terms may be
defined.
Substituting for iq and iγ the appropriate relationship may be written for generation of the
foundation capacity for combined vertical and horizontal loading as:
FVH = A {0.5 γ' B Nγ sγ dγ [1 - (FH/ FVH)*]m+1 + po'Nq sq dq [1 - (FH/FVH)*]m}
This may be solved by the use of assumed values for (FH/FVH) designated (FH/FVH)*. For
example use (FH/FVH)* = 0.00, 0.04, 0.08, 0.12, etc. For these values corresponding FVH
values may be determined.
The correct FH values may then be determined as FVH and (FH/FVH)* are known, e.g. for
(FH/FVH)* = 0.12, FH* = 0.12 FVH*.
The corrected horizontal capacity, FH, may then be given as:

FH = FH* + 0.5γ' (kp - ka) (h1 + h2) As
The sliding capacity envelope of a footing in sand is given by:

FH = FVHtanδ + 0.5γ' (kp - ka) (h1 + h2) As
where δ is the steel/soil friction angle which for a flat plate, δ = φ - 5°, and for a rough
surfaced conically shaped spudcan δ = φ.
Rev 3, August 2008
6.3.3.2 Ultimate vertical/horizontal bearing capacity envelopes for spudcan; footings in clay
The general equation for the horizontal and vertical bearing capacity envelopes for
footings in clay is as follows:
FVH = A [(Nc cu sc dc ic) + po' Nq sq iq dq]
Substituting for the inclination factors for a circular footing the equation may be written
as:
FVH = A {Nc cu sc dc [1 - (1.5FH*/NcAcu)]
+ po' Nq sq (1 - FH*/FVH)1.5 dq}
The ultimate bearing capacity envelope under inclined loading may be determined by
substituting values of FVH and solving for FH*.
FH may then be given as:

FH = FH* + (cuo + cul)As
Footing sliding capacity in clay:
When 0 ≤ QV ≤ 0.5 FV the sliding capacity in clay may be conservatively assumed
constant, determined by:
FH = Acuo + (cuo + cul)As
6.3.3.3 Ultimate vertical/horizontal bearing capacity envelopes for spudcan for spudcan footings
on layered soils.
The above formulas (Sections 6.3.3.1 through 6.3.3.2) can also generally be used to make
a conservative estimate of the ultimate FVH-FH relationship for layered soils by
considering failure through the weakest zones in such a soil profile.
The bearing capacity of layered soils may be determined using the principles of limiting
equilibrium analysis or the finite element method.
6.3.3.4 Settlements resulting from exceedence of the capacity envelope
Vertical settlement and/or sliding of a footing can occur if the storm load combination is
in excess of the (FVH-FH) resistance envelope computed for the spudcan at the penetration
achieved during installation. Such settlements can result in a gain of (FVH-FH) bearing
capacity, e.g. in silica sands. However, the integrity of the foundation may decrease in
the situation where a potential punch-through exists, e.g. where dense sand overlies soft
clay. More thorough analyses are required for complex and/or potentially dangerous
foundation conditions of the type listed in Section 6.3.1.
6.3.4 Footing with moment fixity and vertical and horizontal stiffness (Step 2b)
Foundation fixity is the rotational restraint offered by the soil supporting the foundation.
The degree of fixity is dependent on the soil type, the maximum vertical footing load
during installation, the foundation stress history, the structural stiffness of the unit, the
geometry of the footings and the combination of vertical and horizontal loading under
consideration.
Rev 3, August 2008
Inclusion of foundation fixity in an assessment incorporates a check on bearing capacity

in terms of vertical and horizontal (sliding) capacities. The amount of rotational fixity is
not directly involved in a checking equation, but it serves to modify the forces
(beneficially) in both the foundation and structure. The bearing and sliding checks are
performed implicitly through the use of the yield function and explicitly through the
bearing capacity and sliding checks described in Section 6.3.3.
Uncertainties in soil properties should be considered when including fixity in
assessments. Where data reliability is uncertain, an upper/lower bound sensitivity
analysis should be performed.
For performing structural analysis, horizontal and vertical spring stiffnesses should be
included in addition to the rotational stiffness (see Section 5.3). The springs should be
applied to the spudcan support point as defined in Section 5.2. The calculation of fixity
should be based on factored environmental loading including dead, live, environmental,
inertial and P-Δ loads.
6.3.4.1 Calculation procedures accounting for moment fixity – See also 6.3.4.6
The interaction of vertical, horizontal and rotational forces has been modeled based on a
plasticity relationship (References C6 [48] through [52]). The plasticity relationship can
account for moment softening at high load levels, unloading behavior and work-
hardening effects. This type of foundation modeling is preferable if foundation fixity is
to be included directly in a time-domain analysis.
For a pseudo-static analysis, a simplified application of this full plasticity analysis is
described in this section. This simple approach can be used to create moment loads on
the spudcan by inclusion of a simple linear rotational spring to generate moments at the
spudcan. The moment thus induced on the spudcan is limited to a capacity based on the
yield interaction relationship among vertical (QV), horizontal (QH) and moment (QM)
loads acting at the spudcan.
This simple procedure is described in the following steps:
1. Include vertical, horizontal and (initial) rotational stiffnesses (linear springs) to the
analytical model and apply the gravity and factored metocean and inertial loading.
2. Calculate the yield interaction function value using the resulting forces at each
spudcan. For extreme wave analysis, the result will likely indicate the force
combination falls outside the yield surface. In this case, reduce the rotational
stiffness (arbitrarily) and repeat the analysis.
3. Continue with step 2 until the force combination at each spudcan lies essentially on
the yield surface. If the moment is reduced to zero, and the force combination is still
outside the yield surface, then a bearing failure (either vertical or horizontal) is
indicated.
4. If a force combination initially falls within the yield surface, the rotational stiffness
must be further checked to satisfy the reduced stiffness conditions in Section 6.3.4.3.
The following sections are applicable to traditional spudcan designs. Information on

spudcans fitted with skirts can be found in references C6 [48] through [51].
Rev 3, August 2008
6.3.4.2 Ultimate Vertical / horizontal / rotational capacity interaction; function for spudcan
footings in sand and clay
For shallow embedment for both sand and clay, the yield interaction is defined by the
following expression:
2 2
⎡ FHM ⎤ ⎡ FM ⎤ ⎡ FVHM ⎤ ⎡ FVHM ⎤
⎢ ⎥ +⎢ ⎥ − 4⎢ ⎥ ⎢1 − ⎥=0
⎣ H Lo ⎦ ⎣ M Lo ⎦ ⎣ V Lo ⎦⎣ V Lo ⎦
where VLo is taken to be equal to the vertical spudcan load achieved during preloading
and HLo and MLo are defined as follows:
For sand:
H Lo = (C1 / C 2 )(VLo / 4)⎫
= 0.12VLo ⎪
⎪
⎬ with C1 = 0.3, C 2 = 0.625
M Lo = C1VLo B / 4 ⎪
= 0.075VLo B ⎪⎭
For clay: HLo = cuoA + (cuo + cul) As

MLo = 0.1VLoB
Note that in the above expression for the yield surface, if a load combination
(QV,QH,QM) satisfies the equality then (QV,QH,QM) = (FVHM, FHM, FM). The load
combination (QV,QH,QM) lies outside the yield surface if the left-hand side is greater
than zero. Conversely, the load combination lies inside the yield surface if the left-hand
side is less than zero.
The expression for the yield surface can be re-written to give the maximum spudcan
moment as a function of the horizontal and vertical loads. Thus, for a given vertical and
horizontal load combination which, with zero moment, lies inside the yield surface given
above, the maximum moment at a spudcan cannot exceed the value defined below.
0.5
⎧⎪ ⎡ Q ⎤ 2 ⎡ Q ⎤ 2 ⎡ Q ⎤ 2 ⎫⎪
FM = MLo ⎨16 ⎢ V ⎥ ⎢1 − V ⎥ − ⎢ H ⎥ ⎬
⎪⎩ ⎣V Lo ⎦ ⎣ V Lo ⎦ ⎣ H Lo ⎦ ⎪⎭
The equation above only applies when:
0.0 < QV < VLo
⎡ Q ⎤⎡ Q ⎤
QH < 4 H Lo ⎢ V ⎥ ⎢1 − V ⎥
⎣VLo ⎦ ⎣ VLo ⎦
Embedded footings in clay achieve greater moment and sliding capacities as compared
to shallow penetrations in clay. For fully or partially penetrated spudcans, the yield
surface at FVHM/VLo<0.5 can be expressed as:
2 2
⎡ FHM ⎤ ⎡ FM ⎤
⎢ ⎥ + ⎢ ⎥ - 1.0 = 0
⎣ f 1 H Lo ⎦ ⎣ f 2 M Lo ⎦
where;
⎡Q ⎤
f1 = α + 2(1 - α) ⎢ V ⎥
⎣ VLo ⎦
Rev 3, August 2008
f2 = f1 where suction (i.e. uplift resistance) is available,
⎡Q ⎤ ⎡ Q ⎤
= 4 ⎢ V ⎥ ⎢1 − V ⎥ where suction cannot be relied upon
⎣ VLo ⎦ ⎣ VLo ⎦
α = 1.0 for soft clays

= 0.5 for stiff clays
α accounts for the degree of adhesion. Engineers may want to consider α values within
the range 0.5-1.0 depending on site specific soil data, spudcan/soil interface roughness,
etc. An α value less than 0.5 may be considered for situations such as a hard clay at the
surface. In this case, the standard form of the yield surface should be considered.
Thus, for a given vertical and horizontal load combination which, with zero moment, lies
inside the yield surface given above, the maximum moment at a spudcan for a clay
foundation with QV/VLo<0.5 cannot exceed the value defined below:
0.5
⎧⎪ ⎡ Q ⎤ 2 ⎫⎪
FM = f 2 M L 0 ⎨1 − ⎢ H ⎥ ⎬
⎩⎪ ⎣ 1 Lo ⎦ ⎭⎪
fH
The equation above only applies when:
0.0 < QV < VLo
QH < f1 H Lo
There is no existing data for deeply embedded footings in sand. The application of the
yield surface calibrated to shallow penetrations will likely be conservative for the deep
penetration case.
6.3.4.3 Estimation of rotational, vertical, and horizontal stiffness
An initial estimate for rotational stiffness, K3, which is applicable for a flat spudcan
without embedment (Winterkorn [10]) under relatively low levels of load is given
below:
GB 3
K3 = , flat spudcan with no embedment
3(1 − ν )
Values for K3 for other cases are given in the Commentary. The selection of the shear
modulus, G, is discussed in the Commentary. An upper or lower bound value should be
selected as appropriate for the analysis being undertaken.
For clays susceptible to cyclic degradation (OCR ≥ 4) the soil rotational stiffness,
calculated from the degraded static soil properties, may be multiplied by a factor of 1.25,
Anderson [18].
If the load combination of (QV,QH,QM) lies outside the yield surface, the linear rotational
stiffness at the spudcan must be reduced until the load combination lies on the yield
surface. The reduction in stiffness is arbitrary and requires iterative analyses.
It should be noted that if the initial load combination (QV,QH,QM) lies outside the yield
surface, the final value of the rotational stiffness is determined only by the requirement
that the generated moment at the spudcan falls on the yield surface.
If the load combination of (QV,QH,QM) lies inside the yield surface, the initial estimate
of rotational stiffness should be reduced by a factor, fr. The reduction factor is equal to
Rev 3, August 2008
unity when the moment and horizontal forces are zero. It is given by the following
expression (Svanø, [56]):
fr = (1- rf ) + 0.1e100(rf −1)
where rf is the failure ratio defined by:

0.5
⎧⎪⎡ Q ⎤ 2 ⎡ Q ⎤ 2 ⎫⎪
⎨⎢ ⎥ +⎢ ⎥ ⎬
H M
⎪⎩⎣ Lo ⎦
H ⎣ Lo ⎦ ⎪⎭
M
rf =
⎡ Q ⎤⎡ Q ⎤
4⎢ V ⎥ ⎢1 − V ⎥
⎣VLo ⎦ ⎣ VLo ⎦
Note that rf > 1.0 implies that the load combination (QV,QH,QM) lies outside the yield
surface. Under such conditions, the reduced stiffness factor is not applicable.
For fully embedded foundations in clays at vertical load ratio FVHM/VLo < 0.5, the failure
ratio may be expressed as:
0.5
⎧⎪ ⎡ Q ⎤ 2 ⎡ Q ⎤ 2 ⎫⎪
rf = ⎨ ⎢ H ⎥ + ⎢ M ⎥ ⎬
⎩⎪ ⎣ 1 Lo ⎦ ⎣ 2 Lo ⎦ ⎭⎪
fH fM
where f1 and f2 are as defined in Section 6.3.4.2 above, but replacing FVHM with QV.
Vertical and horizontal stiffnesses can be estimated from the elastic solutions for a rigid
circular plate on an elastic half-space (assuming no embedment):
2GB
Vertical stiffness, K1 =
(1 − ν )
16GB(1 − ν )
Horizontal stiffness, K2 =
(7 − 8ν )
Advice on the selection of appropriate values for G may be found in the Commentary.
6.3.4.4 Extension of the yield surface for additional penetration
On seabeds of silica sands, conical spudcans which are not fully seated may show a
plastic moment restraint due to further penetration. The effect may be taken into
account for legs with QV/VLo > 0.
The moment capacity Mp associated with further penetration is estimated as the

minimum of MPS and MPV, calculated as follows (Svanø [56]):
MPS = 0.075 B VLo(D/B)3
MPV = 0.15 B FVHM
in which B is the plan diameter of the effective contact area after preload, and D is the
plan diameter of the contact area when the spudcan is fully seated.
The combined capacity should be checked against the modified yield function:
2 2
⎡ FHM ⎤ ⎡ FM ⎤ ⎡ FVHM ⎤ ⎡ FVHM ⎤
⎢ ⎥ +⎢ ⎥ − 4⎢ ⎥ ⎢1 − ⎥=0
⎣ H Lo ⎦ ⎣M P ⎦ ⎣ VLo ⎦ ⎣ VLo ⎦
For additional penetration of spudcans in clay, references C6 [49] and [52] provide
work-hardening modifications to the yield surface equations. Updated stiffnesses are
determined through plasticity principles.
Rev 3, August 2008
6.3.4.5 Deep Footing Penetration
For deep footing penetrations, typically experienced in soft clay conditions, the
calculation of foundation fixity may be augmented with the inclusion of the lateral soil
resistance on the leg members due to soil back-flow over the spudcan. This lateral soil
resistance is effectively added to the rotational elastic stiffness of the spudcan (as
determined in Section 6.3.4.3), (Brekke [7]).
The lateral soil resistance may be modeled based on concepts proposed by Matlock [17]
for lateral soil resistance of piles. The jack-up leg may be modeled as an equivalent pile
for purposes of determining "p-y", or load-deflection curves.
The diameters of the individual members (i.e., leg chords and braces) give appropriate
characteristic dimensions for determining the p-y curves. The p-y curve for each
member is summed to form a p-y curve for the entire leg. Only one face of each leg
should be assumed to be in contact with the soil and contribute to lateral resistance.
Given a set of p-y curves for the leg, the lateral force-deflection along the entire
embedded leg section is thus determined. Typically, equivalent springs at each bay
elevation are used to simplify the calculations.
6.3.4.6 Calculation procedures accounting for moment fixity – further details

Structural analyses should account for rotational, horizontal and vertical stiffnesses at all
spudcans. The jack-up is then acceptable if the following conditions are met:
1. Structural conditions satisfy acceptance criteria outlined in Section 8.1.
2. Factored foundation loads QV, QH satisfy, as applicable, the bearing capacity criteria
in Sections 8.3.2 or 8.3.1.5.
3. Factored foundation loads QV, QH, QM satisfy the appropriate unfactored yield
surface criterion from Section 6.3.4.2 or 6.3.4.4. Factored foundation loads
exceeding this requirement are permitted provided that the soil-structure interaction
model adopted accurately captures the expansion of the foundation yield surface
after first yield, and that the large-displacement effects of associated structural
displacements are taken into account.
4. The analysis ensures load & displacement compatibility between the foundation and
the structure.
5. The location is not prone to, or is protected from, scour so that the assumed fixity is
assured.
Fixity may be included in the response simulation in three ways (Refer to Figure 6.11
below):
1. By conservatively considering effects of changes to seabed boundary reactions only
and ignoring any reduction in the dynamic response with pinned footings. In this
approach quasi-static analyses are used in the iterations of the procedure given in
Section 6.3.4.1 to derive the foundation rotational and horizontal secant stiffnesses
with loadings obtained from the pinned foundation case including dynamics. This
approach is not applicable if the inclusion of fixity brings the natural period closer
to the wave period.
2. By considering linearised fixity in SDOF or more detailed dynamic calculations and
then carrying out a final quasi-static analysis with non-linear fixity using the
procedure of Section 6.3.4.1. If this approach is adopted, care should be taken to
ensure that the natural period with fixity does not fall at a cancellation point in the
Rev 3, August 2008
wave force transfer function (Sections 7.3.5.2, 7.3.5.4, C7.4 & Fig C7.1). Typically
the initial linearised rotational stiffness for the dynamic analysis may be taken as
80% of value determined from the formulation in the first paragraph of Section
6.3.4.3. When this stiffness is adjusted to avoid wave force cancellation, the
adjusted value may lie anywhere between 0% and 100% of the value from Section
6.3.4.3.
This simplified approach does not capture the temporary reductions in stiffness
which occur during plasticity events, but also does not capture the increased
damping that is associated with these events; these two effects are considered to be
largely self-cancelling. Given that care is taken to avoid wave force cancellation
effects, it is considered that the dynamic response will be determined at a level
which is either realistic or conservative.
For further discussion of approaches which may be used to avoid cancellation and
reinforcement effects refer to the Commentary Section C7.4.
3. By considering the effects of the foundation fixity on both the dynamic response
and the seabed reactions. This approach is more complete and may require a
complex iterative calculation procedure. The following outline procedure may be
adopted:
a) Use a time-domain dynamic analysis to determine structural
response and foundation loadings at each time step.
b) Compute the foundation behaviour using a non-linear elasto-plastic
model, such that at each time step the plastic and elastic portions of
the behaviour are captured. If desired, this model may include
hysteresis. This will likely require an iterative procedure.
c) When plasticity occurs, the responses will be influenced by the
load history. Consideration should be given to ensuring that the
methodology used to determine the extreme values provides stable
results. In cases where the analysis is intended to provide final
results (rather than DAF’s for application in subsequent analysis
step) it may be appropriate to perform analyses for differing wave
histories, and then determine the extremes from a procedure such
as that given in C7.B.2.3.
Rev 3, August 2008
Figure 6.10: Calculation procedure to account for foundation fixity

Rev 3, August 2008
Rev 3, August 2008
6.3.5 Displacement Check (Step 3)

Structural model with nonlinear soil response included
When a Step 2 assessment results in an overload situation, Step 3 may be used to

calculate the associated displacements and rotations from a full nonlinear load-
displacement foundation model. The procedure should account for the load redistribution
resulting from the overload and displacement of the spudcan(s). The displacements
derived from the analysis should be checked against the allowable displacements of the
spudcans and should satisfy the following requirements:
- The spudcan vertical and horizontal displacements should not lead to

unacceptable overturning or strength checks.
- The resulting rotation of the unit should neither exceed the limitations defined by
the operating manual nor lead to the possibility of contact with any adjacent
structure.
Rev 3, August 2008
6.4 Other Aspects of Jack-Up Unit Foundation Performance
6.4.1 Leaning Instability
Leaning instability of jack-ups can occur during preloading operations in soft clays where
the rate of increase in bearing capacity with depth is small. In deep water a potentially
unsafe condition (comparable to a punch-through situation) may occur. However, the
potential for such incidents may be discounted if appropriate installation procedures are
adopted. These may, for example, include preloading the footings individually.
Further discussion on leaning instability is included in the Commentary.
6.4.2 Footprint Considerations
The seabed depressions which remain when a jack-up is removed from a location are
referred to as 'footprints'. The form of these features depends on several factors such as
the spudcan shape, the soil conditions, the footing penetration achieved and the method
of extraction. The shape, and the time period over which the form will exist, will also be
affected by the local sedimentary regime.
The positioning of spudcans very close to, or partially overlapping, footprints is not
recommended. The difference in resistance between the original soil and the disturbed
soil in the footprint area and/or the slope at the footprint perimeter, may cause the
spudcans to slide towards the footprint. The resulting leg displacements could cause
severe damage to the structure and, at worst, could lead to catastrophic failure. The
situation could be complicated by the proximity of a fixed structure or wellhead.
The following two operational sequences may be considered:
a) Installation of an identical jack-up design to that previously used at a particular

location:
If a jack-up with identical footing geometry to the unit previously used is to be

installed, the re-positioning should not cause problems provided that the jack-up is
located in exactly the same position as for the previously installed unit. Thus the
footings would lie in the existing footprints. It is therefore necessary to ensure that
reliable records are obtained of the exact location of existing footprints in relation to
the well/jacket.
If the new spudcan positions are not located directly over the footprints sliding of the
legs may occur with the potential consequences described above.
b) Installation of a jack-up of different design to that previously used at a particular

location:
It is unlikely for two jack-up designs to have similar footing geometries. It is

therefore probable that it will not be possible to locate the spudcans exactly within the
existing footprints. However, it may be possible to carefully position the jack-up on a
new heading, and/or with one footing located over a footprint with the others in virgin
soil, to alleviate the potential for spudcan sliding. Again reliable records of existing
footprint locations (and depths) are required.
Rev 3, August 2008
Where it is not possible to locate the jack-up to avoid spudcan-footprint interaction
special attention is required to minimize the potential sliding problem. Consideration
may be given to infilling the footprints with imported materials. The material selection
should recognize the potential for material removal, by scour, and the differences of
material stiffness.
Further discussion is included in the Commentary.
6.4.3 Scour
Scour may occur when a footing or other object is installed on the seabed, and its
presence causes increased local current velocities. The phenomenon is usually observed
around spudcans which are embedded to a shallow level in granular materials at locations
with high current velocities.
Scour may partially remove the soil from below the footing, resulting in a reduction of
the ultimate bearing capacity of the foundation and any seabed fixity. This is normally a
gradual process and the effects of the reduced bearing capacity may not be apparent until
during storm loading when (rapid) downward movement of the leg may occur. The
effects of scour are potentially more severe when it occurs at a location where a potential
for punch-through exists.
There is no definitive procedure for the evaluation of scour potential and emphasis must
usually be placed on previous operational experience. Further guidance is given in the
Commentary.
If scour is recognized to be a potential problem, then preventative measures should be

implemented. These should be adopted on a trial basis and include:
a) Gravel dumping prior to installation provided the selected gravel gradation will not
cause damage to the jack-up footing.
b) Installation of artificial seaweed.
c) Use of stone/gravel dumping, gravel bags or grout mattresses after installation.

Rev 3, August 2008
6.4.4 Seafloor Instability

Seafloor instability may be caused by a number of mechanisms which may be interactive
or act independently. The most frequent types of instability result in large scale mass
movement, in the form of mudslides or slope failures. Such phenomena are often
associated with deltaic deposits, and it is recommended that the advice of local experts is
obtained when such situations are encountered.
Liquefaction, or cyclic mobility, occurs when the cyclic stresses within the soils cause a
progressive build up of pore pressure. The pore pressure within the profile may build up
to a stage where it becomes equal to the initial average vertical effective stress.
Foundation failure may result depending on the extent of pore pressure developed.
Such failures may be manifested as continued foundation settlements or large scale
failure of the soil mass as described above. In areas where liquefaction is known to be
possible its potential must be assessed.
For further guidance refer to the Commentary.
6.4.5 Shallow Gas
Gas in soils may originate from biogenic degradation or thermogenic diagenesis.
Gas charged sediments may result in hazards during site investigation soil borings,
reduced bearing capacity, unpredictable foundation behavior (due to seabed depressions
or gas accumulations under the spudcans) and complications with shallow drilling
operations, including blowouts.
The presence of gas charged sediments may be identified by geophysical digital high
resolution shallow seismic surveys using attribute analysis techniques.
Any gas concentration should be avoided if it is located above the primary casing shoe
level (generally 20 inch or 18.75 inch diameter casing) or the conductor pipe shoe level
which are determined during the drilling program design. This is because neither of these
holes are drilled under BOP control and, therefore, there is a risk of seabed cratering
around the well which could result in the undermining of the footings in the event of a
blow out.
Of lesser risk is the potential for gas migration from depth to the surface outside the
casing. Although this occurrence is uncommon the potential should not be discounted.
Rev 3, August 2008
6.4.6 Spudcan - Pile Interaction
For jack-ups located in close proximity to pile-founded platforms, soil displacements
caused by the spudcan penetration will induce lateral loading into the nearby piles. The
amount of soil displacement will depend on the spudcan proximity (spudcan edge-to-pile
distance), the spudcan diameter and penetration. If the foundation materials comprise
either a deep layer of homogeneous firm to stiff clay or sand and if the proximity of the
spudcan to the pile is greater than one spudcan diameter, then no significant pile loading
is expected. When the proximity is closer than one spudcan diameter, then analysis by
the platform owner is recommended to determine the consequences of the induced pile
loading.
Guidance regarding the analytical procedures available for assessing these spudcan
induced pile loads is given in the Commentary.
Rev 3, August 2008
6 GLOSSARY OF TERMS - CALCULATION METHODS, GEOTECHNICAL
ENGINEERING
A = Spudcan effective bearing area based on cross-section taken at uppermost part

of bearing area in contact with soil (see Figure 6.2).
As = Spudcan laterally projected embedded area.
a = Bearing capacity squeezing factor.
au = Adhesion.
B = Effective spudcan diameter at uppermost part of bearing area in contact with the
soil (for rectangular footing B = width).
b = Bearing capacity squeezing factor.
cu = Undrained cohesive shear strength at D + B/4 below mudline.
cu1 = Undrained cohesive shear strength at spudcan tip.
cuo = Undrained cohesive shear strength at maximum bearing area (D below
mudline).
cus = Undrained cohesive shear strength at D/2 below mudline.
cu,b = Undrained cohesive shear strength - lower clay below spudcan.
cu,t = Undrained cohesive shear strength - upper clay below spudcan.
C1 =⎫
C2 = ⎬⎭ Constants used in computation of H Lo and VLo for sand.
dc = Bearing capacity depth factor.
= 1 + 0.4 (D/B) for D/B ≤ 1.
= 1 + 0.4 arctan (D/B) for D/B > 1.
dq = Bearing capacity depth factor.
= 1 + 2tanφ(1- sinφ)2 D/B for D/B ≤ 1
= 1 + 2tanφ(1- sinφ) arctan(D/B) for D/B > 1
2
dγ = Bearing capacity depth factor = 1.

D = Distance from mudline to spudcan maximum bearing area.
f1 = Factor used in yield surface equation for embedded footings on clay.
f2 = Factor used in yield surface equation for embedded footings on clay.
fr = Reduction factor on stiffness.
Fo' = Effective overburden pressure due to back-flow at depth of uppermost part of
bearing area.
FH = Horizontal foundation capacity.
FHM = Horizontal foundation capacity in combination with moment.
FV = Vertical foundation capacity.
FV,b = Ultimate vertical bearing capacity assuming the footing bears on the surface of
the lower (bottom) clay layer with no back-flow.
FVH = Vertical foundation capacity in combination with horizontal load.
FVHM = Vertical foundation capacity in combination with horizontal and moment load.
FM = Moment capacity of foundation.
Gv = Shear Modulus for vertical loading.
Gh = Shear Modulus for horizontal loading.
Gr = Shear Modulus for rotational loading.
h = Distance from rotation point to reaction point.
h1 = Embedment depth to the uppermost part of the spudcan, (if not fully embedded
= 0).
h2 = Spudcan tip embedment depth.
H = Distance from spudcan maximum bearing area to weak strata below.
HLo = (C1/C2)(VLo/4), C1 = 0.3, C2 = 0.625 (sand)
= Acuo +(cuo + cu1)As (clay)
ic = Inclination factor (for φ = 0).
= 1 - mFH/AcuNc
Rev 3, August 2008
6 GLOSSARY OF TERMS - CALCULATION METHODS, GEOTECHNICAL
ENGINEERING (Continued)
iq = Inclination factor.
= (1 - FH/FVH)m
iγ = Inclination factor.
= (1 - FH/FVH)m+1
I Height of soil column above spudcan.
ka = Active earth pressure coefficient (for cu = 0) = tan2(45-φ/2)
kp = Passive earth pressure coefficient = 1/ka
K1,K2,K3 = Stiffness factors.
Ks = Coefficient of punching shear.
L = Foundation length, for circular foundation L=B.
For strip footing - inclination in direction of shorter side.
= (2 + B/L)/(1 + B/L)
m= For strip footing - inclination in direction of longer side.
= (2 + L/B)/(1 + L/B)
For circular footing = 1.5
MLo = C1VLoB/4, C1 = 0.3 (sand)
= 0.1VLoB (clay)
MP = moment capacity associated with further spudcan penetration under
environmental loading (equal to minimum of MPS and MPV).
MPS = moment capacity when further spudcan penetration leads to fully seated spud
conditions.
MPV = moment capacity under further spudcan penetration, when the actual vertical
force is too low to reach fully seated conditions.
n = Iteration factor, ≥ 2.
N = Stability factor.
Nc = Bearing capacity factor (taken as 5.14).
Nq = Bearing capacity factor = eπtanφtan2(45 + φ/2)
Nγ = Bearing capacity factor = 2(Nq + 1)tanφ
po' = Effective overburden pressure at depth, D, of maximum bearing area.
QH = Applied factored horizontal load.
QM = Applied factored moment load.
QV = Applied factored vertical load.
rf = Failure ratio.
sc = Bearing capacity shape factor = (1 + (Nq/Nc)(B/L))
sq = Bearing capacity shape factor = 1 + (B/L)tanφ
sγ = Bearing capacity shape factor = 1 - 0.4(B/L)
( = 0.6 for circular footing under pure vertical load).
T Thickness of weak clay layer underneath spudcan.
V = Volume of soil displaced by spudcan.
VLo = Maximum vertical foundation load during preloading.
α = Adhesion factor = 1.0 for soft clays, = 0.5 for stiff clays.
δ = Steel/soil friction angle - degrees, (φ-5≤δ≤φ).
δν = Vertical displacement of foundation.
δh = Horizontal displacement of foundation.
γ' = Submerged unit weight of soil.
θ = Foundation rotation - radians.
φ = Angle of internal friction for sand - degrees.
ν = Poisson's ratio.
Rev 3, August 2008
7 CALCULATION METHODS - DETERMINATION OF RESPONSES
7.1 General
7.1.1 The response of a jack-up unit is determined by combining the applied factored loading
with a structural model to determine the internal forces in the members and the reactions
at the foundations. These internal forces and reactions are compared with the factored
resistances available to take up these loads to determine the safety of the unit. The loads
consist of fixed loads (self weight and non-varying loads) and variable loads (see
Section 3.2) together with hydrodynamic and wind loadings (see Section 4). The
structural modeling is described in Section 5. The foundation resistance is described in
Section 6. Section 8 provides the structural resistance and a methodology to check the
adequacy of the various resistances to the acceptance criteria.
7.1.2 Two aspects of the response are to be distinguished and assessed separately. These are:
a) The extreme response. The maximum calculated response to the design environment
occurring at a particular instant in time, which is compared with the acceptance
criteria. See Sections 7.2 and 7.3.
b) Fatigue. The cumulative effect of stress/strain cycling, which is used to estimate the
fatigue lives of steel components (see Section 7.4).
7.1.3 For typical jack-up assessments, the time-varying nature of the wave loading will amplify
the quasi-static responses and must be considered. The extreme response can be assessed
either by a quasi-static analysis procedure (Section 7.2) including an inertial loadset
(Section 7.3.6) or by a more detailed dynamic analysis procedure (Section 7.3.7).
7.1.4 The dynamic amplification of the quasi-static response may not be significant for a given
set of location parameters. The magnitude of the dynamic response is primarily
influenced by the amount of wave energy at or near the natural period of the jack-up.
The distribution of wave energy is at a maximum at the peak wave period and reduces for
other periods. Thus the single most important parameter in the determination of the
dynamic amplification of responses is the separation of the natural period of the jack-up
from the peak period of the wave spectrum. Generally a large separation will produce a
small dynamic amplification. As the separation decreases, the dynamic amplification
will increase. These conclusions may be modified by effects such as wave-load
cancellation and wave-current induced harmonics.
7.1.5 For many applications, the dynamic amplification may be determined using a simple, but
empirical, method. This simple method is detailed in Section 7.3.6.1. Caution is advised
when relying solely on results using this simple method. Specific guidance on the
limitations of the method is given in Section 7.3.6.1.
Because of its simplicity, the method detailed in Section 7.3.6.1 is recommended for an
initial evaluation of the dynamic amplification. If the dynamic amplification is
determined to be relatively small (see Section 7.3.6.1), or, if acceptance criteria are met,
then random dynamic analysis is not required.
Rev 3, August 2008
7.1.6 For many applications the dynamic effects may be included through the addition of an
inertial loadset (see Section 7.3.6.1) to the environmental loads in a quasi-static analysis
procedure. In this approach the inertial loadset may be determined using a simplified
model of the jack-up. An appropriate detailed model of the jack-up may then be used to
determine the detailed responses when the inertial loadset is applied together with the
quasi-static environmental loads.
7.1.7 Appropriate combinations of gravity loads, wave/current loads and wind loads shall be
applied as required by the acceptance criteria in Section 8. Load application is described
in Section 5.7. Section 5.1 requires that the analysis is carried out for a range of
environmental headings with respect to the unit such that the most onerous loading(s) for
each major type of element in the structural system is(are) determined. The checks
cover:
Load Component
Limit State Check Section Response Parameters(s)1 L note 2 Dn
D E
min max
Strength of 8.1 Element load vectors3 Y Y4 Y Y
elements
Overturning 8.2 Overturning moment 5 5 Y Y
stability Stabilizing moment Y Y
Foundation 8.3
capacity:
- preload 8.3.1 Vertical leg reaction Y Y Y Y
- sliding 8.3.1 Vertical & Horizontal Y Y Y Y
leg reactions
- bearing 8.3.2/3 Vertical, Horizontal Y Y6 Y6 Y Y
(& moment) leg reactions
- displacement 8.3.4 Leg footing displacements Y Y6 Y6 Y Y
and reactions
Horizontal 8.4 Hull displacement. Y Y6 Y6 Y Y
deflection
Holding system 8.5 Holding system loads vectors Y Y6 Y6 Y Y
loads
where D, L, E and Dn are defined in the glossary at the end of section 7.
Notes:
1. In all instances the responses are evaluated including the effects of deformation under dead loads (hull
sag) and large displacement (P-Δ) effects.
2. Placed at most onerous center of gravity position.
3. The effects of leg offset to be added after global response analysis (see Section 5.4).
4. Consider minimum live (variable) load if this is more onerous.
5. Must be included in response calculation so P-Δ effects are included.
6. Worst case combination required.
7.2 Quasi-Static Extreme Response with Inertial Loadset
7.2.1 The most common method of analysis adopted for the determination of extreme
responses is the deterministic, quasi-static wave analysis. Such an analysis shall be
carried out in accordance with all relevant requirements of Sections 3 to 6. The
maximum wave loading is determined by 'stepping' the maximum wave through the
structure. The maximum wave is defined in Section 3.5.1.2 and the methodology for
calculating the wave loading is described in Section 4.3. Various methods for
determining the inertial loadset are given in Section 7.3.6. Load cases and combinations
are discussed in Section 7.1.7.
Rev 3, August 2008
The spudcan-foundation interface should normally be modeled as a pin joint. The
inclusion of a degree of fixity is to be justified on a case by case basis. If foundation
fixity is included it should generally be represented by a combination of horizontal,
vertical and rotational springs (which may be coupled) at the spudcan, rather than by a
rotational spring alone. (See also Sections 5.3, 6.3.4 and 7.3.5.2).
7.3 Dynamic Extreme Response
7.3.1 Factors Governing Dynamics
Dynamic amplification of the structural response must be taken into account (see Figure
7.1).
Determination of dynamic response requires the incorporation of two separate items in

the analysis:
a) The dynamic characteristics of the structural system formed by the jack-up on its
foundation,
b) The characteristics of the environmental excitation.
7.3.2 The Structural System
7.3.2.1 The characteristics of the structural system are governed by the following aspects:
a) The mass and mass distribution of the jack-up.

This includes structural mass, mass of equipment and variable load on board, added
mass due to the surrounding water and marine growth (if applicable), etc. The
magnitudes and effective centers of mass of the various mass contributions are to be
accurately determined.
b) The overall (global) structural stiffness.

This includes stiffness contributions from bending, shear deformation and axial
straining of the legs, the leg to hull connections, the hull and the spudcan-foundation
interface (if applicable).
c) The damping.
Damping contributions arise from the structural components and their connections,
the water surrounding the legs and the soil underneath/around the spudcans. For
further discussion of damping refer to Section 7.3.7.
7.3.2.2 The jack-up on its foundation represents a multi degree-of-freedom system. If the
dynamic behavior is to be investigated in some detail it should also be modeled as such.
The model may contain a number of nonlinear elements, notably the leg to hull
connections and the spudcan-foundation interfaces. The influence of gravity (P-Δ/Euler)
on the effective sway stiffness should be considered (see Section 5.5).
Rev 3, August 2008
Figure 7.1: Recommended approach to determine extreme dynamic responses

Rev 3, August 2008
Due to the fact that the mass of the hull dominates the mass distribution, the global
dynamic behavior of the jack-up may in some cases be determined from an idealized
single degree-of-freedom system (see Section 7.3.6.1).
Structural modeling at various levels of complexity is discussed in Section 5.6.
7.3.3 The Excitation
7.3.3.1 The characteristics of the environmental excitation are controlled by the fluctuating
nature of the environmental factors - wind, current and waves. Currents change slowly
compared with the natural periods at which jack-ups may oscillate and may hence be
considered to be a steady phenomenon. Variations in wind velocity cover a wide range
of periods, but the main wind energy is associated with periods which are considerably
longer than the natural periods of jack-up oscillations. Therefore, in connection with
jack-ups, the wind may generally be represented as a steady flow of air. The periods of
waves typically lie between some 2-3 sec and some 20 sec. Since typical jack-up natural
periods fall within this range, the primary source of excitation is from waves.
Sea waves are generally not regular but random in nature unless swell is predominant.
This has important implications which should be considered for both the dynamic
excitation and the resulting dynamic response.
As waves and currents interact these two environmental factors should be considered in
combination when generating time varying hydrodynamic drag forces according to
Section 4.3.
7.3.3.2 For the simplified dynamic analysis method of Section 7.3.6.1 based on a regular-wave
deterministic quasi-static analysis the wave period is chosen to be 0.9Tp where Tp is the
peak period of the wave spectrum for the extreme sea state.
For random analyses (see Sections 7.3.6.2 and 7.3.7) the most probable peak period (Tp)
of the wave spectrum for the extreme seastate will normally be selected when a 2
parameter Pierson-Moskowitz spectrum is used (Hs and Tp from site specific data and
γ = 1 in Section 3.5.3). If a JONSWAP spectrum is used it is recommended that the peak
period is considered to vary between plus and minus one standard deviation from the
most probable peak period (Tp).
Where the jack-up is sensitive to the wave period it is recommended that the range
described in Section 3.5.1.2 or 3.5.3 is investigated as appropriate.
In a deterministic calculation waves with a period close to the natural period of the jack-
up will give the largest dynamic amplification. It is therefore recommended that the
wave associated with the highest natural period of the jack-up is also investigated.
Rev 3, August 2008
7.3.4 The Dynamic Analysis
A flow chart indicating the recommended dynamic analysis approach is shown in Figure
7.1.
An initial estimate of the dynamic amplification can be obtained using the empirical
methods described in Section 7.3.6.1.
Techniques for performing random dynamic analyses can be categorized as frequency

domain methods or time domain (simulation) methods or hybrids thereof; see Section
7.3.7.
7.3.5 The Natural Period(s)
7.3.5.1 The natural period of the jack-up on its foundation in the fundamental (or first) mode of
vibration is an important indicator of the degree of dynamic response to be expected.
The first and second vibrational modes are nearly always the surge and sway modes. The
natural periods of these vibrational modes are usually close together; which of the two is
the higher depends on which direction is less stiff. Where the period varies with
environmental heading, care should be taken that the period used is applicable to the
environmental direction being considered in the analysis. The third vibrational mode is
normally a torsional mode, the three-dimensional effects of which may be important, in
particular for environmental attack directions where the legs and hence wave loads are
not symmetric about the direction of wave propagation.
7.3.5.2 If available, a finite element structural model containing the mass and stiffness properties
of the jack-up may be used to obtain the various natural periods and mode shapes. This
model should include the stiffness of the legs, hull and hull/leg connections according to
Sections 5.6.4 to 5.6.6. If a finite element model containing only stiffness properties is
available, then the global sway stiffness for the required headings may be determined by
applying lateral unit loads to the hull.
Normally the foundation will be considered pinned. This assumption may however be
unconservative for situations in which:
1. The structure natural period is within a cancellation region of the base shear transfer
function (see Commentary Section C7.4).
2. Significant foundation nonlinearities are expected at higher loading levels typical of

dominant wave frequencies but not at lower loading levels typical of inertial
frequencies.
If either of these situations occur, and detailed foundation modeling is not available, it is
recommended that the DAF's be calculated with fixity included and are then applied to a
pinned model for response calculations.
Where the foundation stiffness is included, lateral and vertical translational springs
should be included together with the rotational springs. In any case the limitations on
foundation loading according to Section 6.3.4 must be verified.
Rev 3, August 2008
7.3.5.3 If such a capability is not available, the fundamental mode period may be estimated from
the system described by:
- an equivalent mass representing the mass of the jack-up and its distribution as
referred to in Section 7.3.2; the equivalent mass is equal to the mass of the hull plus a
contribution from the mass of the legs, including added mass, and is located at the
center of gravity of the hull.
- an equivalent spring representing the combined effect of the various stiffnesses

mentioned in Section 7.3.2.
The period is determined from the following equation applied to one leg:
Tn = 2π ( M e / K e )
where;
Tn = highest (or first mode) natural period.
Me = effective mass associated with one leg.
M hull M
= + Mla + lb
N 2
Mhull = full mass of hull including maximum variable load.
N = number of legs.
Mla = mass of leg above lower guide (in the absence of a clamping mechanism)
or above the center of the clamping mechanism.
Mlb = mass of leg below the point described for Mla, including added mass for
the submerged part of the leg ignoring spudcan. The added mass may be
determined as Aeρ(CMe - 1) per unit length of one leg (for definitions of
Ae and CMe see Section 4.6.6); ρ = mass density of water.
Ke = effective stiffness associated with one leg (for derivation, refer to
Commentary).
⎡ P⎤
⎢1 − ⎥
3EI ⎣ PE ⎦
= *
L3 ⎡ 3L 12Fg I ⎧ EI L ⎫ 3(EI) 2 ⎤
⎢ − 2 ⎨ + ⎬ − ⎥
⎢1 − 4 AFv Y ⎩ K rs 2 ⎭ Fr LK rs K rh + 7.8I ⎥
⎢ ⎧ EI EI ⎫ A s Fh L2 ⎥
⎢ ⎨ + L+ ⎬ ⎥
⎢⎣ ⎩ K rs Fr K rh ⎭ ⎥⎦
When the soil rotational stiffness Krs at the spudcan-foundation interface is zero this may
be re-written:
⎡ P⎤
⎢1 − ⎥
3EI ⎣ PE ⎦
= 3 *
L ⎡ 12Fg I 3EI 7.8I ⎤
⎢1 + 2 + + ⎥
⎣ AFv Y Fr LK rh A s Fh L2 ⎦
Krs = rotational spring stiffness at spudcan-foundation interface.

Rev 3, August 2008
Krh = rotational stiffness representing leg to hull connection stiffness (see below).
Fr = factor to account for hull bending stiffness.
1
=
⎧ YK rh ⎫
⎨1 + ⎬
⎩ 2EI H ⎭
IH = representative second moment of area of the hull girder joining two legs
about a horizontal axis normal to the line of environmental action.
E = Young's modulus for steel.
A = axial area of one leg (equals sum of effective chord areas, including a
contribution from rack teeth - see Note to Section 5.6.4).
As = effective shear area of one leg (see Figure 5.1).
I = second moment of area of the leg (see Figure 5.1), including a contribution
from rack teeth (see Note to Section 5.6.4).
Y = distance between center of one leg and line joining centers of the other two
legs (3 leg unit).
= distance between windward and leeward leg rows for direction under
consideration (4 leg unit)
Fg = geometric factor.
= 1.125 (3 leg unit), 1.0 (4 leg unit)
Fv = factor to account for vertical soil stiffness, Kvs, and vertical leg-hull
connection stiffness, Kvh (see below).
1
=
⎧ EA EA ⎫
⎨1 + + ⎬
⎩ LK vs LK vh ⎭
Fh = factor to account for horizontal soil stiffness, Khs, and horizontal leg-hull
connection stiffness, Khh (see below).
1
=
⎧ EA s EA s ⎫
⎨1 + + ⎬
⎩ 2.6LK hs 2.6LK hh ⎭
L = length of leg from the seabed reaction point (see Section 5.2.1) to the point
separating M1a and M1b (see above).
P = the mean force due to vertical fixed and variable loads acting on one leg.
M hull g
=
N
g = acceleration due to gravity.
Rev 3, August 2008
PE = Euler buckling load of one leg.

= α2EI
α = the minimum positive non-zero value of αL satisfying:
⎧ (K rs + K rh )αEI ⎫
tan(αL) = ⎨ ⎬
⎩ (αEI) − (K rs K rh ) ⎭
2
Thus:
when Krs = 0 and Krh = ∞, αL = π/2 and hence:
π 2 EI
PE =
4L2
when Krs = ∞ and Krh = ∞, αL = π and hence
π 2 EI
PE =
L2
The hull to leg connection springs, Krh, Kvh and Khh represent the interaction of the leg
with the guides and supporting system and account for local member flexibility and
frame action. They should be computed with respect to the point separating M1a and M1b,
as described above. The following approximations may be applied:
Khh = ∞
Kvh = effective stiffness due to the series combination of all vertical pinion or
fixation system stiffnesses, allowing for combined action with shock-
pads, where fitted.
Unit with fixation system:
Krh = combined rotational stiffness of fixation systems on one leg.
= Fnh2kf
where;
Fn = 0.5, three chord leg; = 1.0, four chord leg
h = distance between chord centers.
kf = combined vertical stiffness of all fixation system components on one
chord.
Unit without fixation system:
Krh = rotational stiffness allowing for pinion stiffness, leg shear deformation
and guide flexibility.
kud2
= Fnh2kj +
1 + (2.6k u d / EA s )
where;
h = distance between chord centers (opposed pinion chords) or pinion pitch
points (single rack chords).
kj = combined vertical stiffness of all jacking system components on one
chord.
d = distance between upper and lower guides.
ku = total lateral stiffness of upper guides with respect to lower guides.
As = effective shear area of leg.
7.3.5.3 The above equations for estimating the fundamental natural period are approximate and
ignore the following effects:
- more realistic representation of possible fixity at the spudcan-foundation interface in
the form of (coupled) horizontal, vertical and rotational spring stiffnesses.
- three dimensional influences of the system as compared with the two-dimensional
single leg model.
Rev 3, August 2008
7.3.5.4 Due to uncertainty in the parameters affecting the natural period the calculated natural
period(s) will also be uncertain. The natural period(s) used in the dynamic analysis
should be selected such that a realistic but conservative value of the dynamic response is
obtained for the particular application envisaged. Care should be taken to ensure that the
maximum dynamic amplification is not selected as coincident with a cancellation period
causing minimum environmental loading. The potential for increased response due to
shortcrested waves should be considered (see Section 7.3.7.5). For further details refer to
the Commentary Section C7.4 and Figure C7.1.
7.3.6 Inertial Loadset Approaches
In inertial loadset approaches the dynamic response is represented in a global quasi-static

response model by either a distributed inertial loadset or an equivalent point load applied
at the hull center of gravity. The inertial loadset may be derived from the simple
approach described in Section 7.3.6.1 or from the more complex methods discussed in
Sections 7.3.6.2 and 7.3.6.3.
7.3.6.1 The classical SDOF analogy
This representation assumes that the jack-up on its foundation may be modeled as an
equivalent mass-spring-damper mechanism; see Section 7.3.2. The (highest) natural
period of the vibrational modes may be determined as described in Section 7.3.5. The
torsional mode and corresponding three-dimensional effects cannot be included in this
representation.
The single degree-of-freedom (SDOF) method is fundamentally empirical because (1) the
wave-current loading does not occur at the mass center and (2) the loading is non-
periodic (random) and non-linear.
It should also be noted that all global and detailed response parameters are not equally
amplified. The method described below will generally lead to a reasonable
approximation of the jack-up's real behavior and has been calibrated against more
rigorous methods. The following cautions are noted when using the SDOF method:
1. If the ratio of the jack-up natural period to the wave excitation period, Ω, is less than
0.5 and the current is 'relatively small' the SDOF method should give reasonably
accurate results when compared to a more rigorous analysis.
2. If Ω is greater than 0.5, the relative position of the jack-up natural period within the
base shear transfer function should be checked. If the natural period falls near a
wave force peak, then the SDOF method may be unconservative because it ignores
forcing at other than the full wave excitation period. Note that the calculation of
natural periods should include a range of periods to account for a reasonable
estimate of foundation fixity (see Section 7.3.5.2).
3. The SDOF method may be unconservative for cases with relatively high currents. If
the results of the assessment are close to the acceptance criteria further detailed
analysis is recommended.
Rev 3, August 2008
The ratio of (the amplitudes of the) dynamic to the quasi-static response as a function of
frequency (ω) or period (T) of steady state, periodic and sinusoidal excitation is
calculated as the classical dynamic amplification factor (DAF):
1
DAF =
[(1 − Ω ) + ( 2ζΩ ) 2 ]
2 2
where;
Wave Excitation frequency ω
Ω = =
Jack − up natural frequency ω n
Jack - up natural period Tn
= =
Wave excitation period T
ζ = Damping ratio or fraction of critical damping
= (% Critical Damping)/100, ≤ 0.07.
T = 0.9Tp.
Tp = most probable peak wave period.
Tn = the jack-up natural period as derived in 7.3.5.
The damping parameter ζ in this model represents the total of all damping contributions
(structural, hydrodynamic and soil damping). For the evaluation of extreme response
using the SDOF method a value not exceeding 0.07 is recommended.
The calculated DAF from the SDOF method is used to estimate an inertial loadset which
represents the contribution of dynamics over and above the quasi-static response in
accordance with Figure 7.1. This inertial loadset should be determined as follows and
applied at the hull (center of gravity) in the down-wind direction:
Fin = (DAF - 1) BSAmplitude
where;
Fin = Magnitude of the inertial loadset for use in conjunction with the
SDOF method.
BSAmplitude = Amplitude of quasi-static Base Shear over one wave cycle.
= (BS(Q - S)Max - BS(Q - S)Min)/2
BS(Q - S)Max = Maximum quasi-static wave/current Base Shear.
BS(Q - S)Min = Minimum quasi-static wave/current Base Shear.
Note: The above equation is part of a calibrated procedure and should not be altered. A
more general inertial loadset procedure, using the results from random analysis,
is described in Section 7.3.6.3.
7.3.6.2 Other SDOF approaches
An alternative use of the SDOF method is to apply the entire DAF function for all
frequencies (periods), rather than a single point DAF at one frequency. This method
reflects the random wave plus current excitation more correctly. Execution of this
procedure is as per the relevant parts of Section 7.3.7.
Rev 3, August 2008
7.3.6.3 Inertial loadset based on random analysis
The inertial loadset may be derived from random frequency or time domain analysis
according to the recommendations of Section 7.3.7. The inertial loadset should be such
that it increases the responses of the deterministic quasi-static analysis by the same ratios
as those determined between the random quasi-static (zero mass) analysis and the random
dynamic analysis (see Figure C7.B.1) In such cases the structural model (used for
dynamic analysis) may be simplified and does not need to contain all the structural
details, but will nevertheless be a multi degree-of-freedom model. The approach to the
modeling and determination of the inertial loadset is described further in the
Commentary, Section C7.B.2.
The inertial loadset can be determined to model the effect of dynamic amplification in a
more realistic manner as required. The simplest alternative uses a single point force to
match inertial overturning moment effects as shown in the Commentary, Section C7.B.2.
However the use of a distributed inertial loadset is considered more representative and
will therefore provide a more accurate description of the component dynamic
amplification effects as well as global response amplification. The distribution of the
loadset is based on the fundamental sway modes and mass distribution. Note that the use
of a distributed inertial loadset is recommended for units where a significant proportion
of the total mass (including fluid added mass) acts at a location other than the hull center
of gravity. The mathematical procedure for calculation of the distributed loadset is given
in Figure 7.2. A brief description of the calculation process is as follows:
Step 1
Perform random response analysis using a wave attack direction along the selected main
axis (x or y) and establish the global response dynamic amplification factors for base
shear and overturning moment, whereby the dynamic amplification factors are defined as
DAF3 = MPMEdyn/MPMEstatic.
Step 2
Establish a set of two simultaneous equations using combinations of 2-D mode shapes,
nodal masses and unknown modal scalar, which match the inertial base shear and
moment along the selected main axis. Solve this equation set to determine the two modal
multipliers.
Step 3
Establish the (2-D) inertial loadset Fin by a combination of the selected structural mode
shapes (ϕ1, ϕ2), scalar multipliers (α, β) and nodal masses (M), i.e. Fin = α ϕ1M + β ϕ2M.
Rev 3, August 2008
Figure 7.2 - Procedure for calculation of distributed inertial loadset

(2-D response)
7.3.7 Detailed Dynamic Analysis Methods
Fully detailed random dynamic analysis will be necessary indicated in Figure 7.1.
Random dynamic analysis may be performed in the time or in the frequency domain.
7.3.7.1 The waves may be modeled as a linear random superposition model which is fully
described by the wave spectrum (see Section 3.5.3). The statistics of the underlying
random process are gaussian and fully known theoretically. An empirical modification
around the free surface may be needed to account for free surface effects. This, together
with the fact that drag forces are a nonlinear (squared) transformation of wave
kinematics, makes the hydrodynamic force excitation always nonlinear. As a result, the
random excitation is non-gaussian. The statistics of such a process are generally not
known theoretically, but the extremes are generally larger than the extremes of a
corresponding gaussian random process. For a detailed investigation of the dynamic
behavior of a jack-up the non-gaussian effects must be included. A number of
procedures for doing this are presented in the Commentary.
7.3.7.2 The spudcan-foundation interface should normally be modeled as a pin joint in the
absence of justifiable site-specific foundation fixity information, but see Section 7.3.5.2.
If foundation fixity is included, it should be represented by a combination of horizontal,

vertical and rotational springs. Coupling of the springs is preferable. In any case the
limitations on foundation loading according to Section 6.3.4 must be verified.
7.3.7.3 When the random displacements of the submerged parts are small and the velocities are
significant with respect to the water particle velocities the damping is not well
represented by the relative velocity formulation in Morison's equation, which will tend to
overestimate the damping and underpredict the response. A criterion for determining the
applicability of the relative velocity formulation is given in Section 4.3.2.
Rev 3, August 2008
7.3.7.4 Table 7.1 summarizes appropriate percentages of global critical damping for the various
damping sources which should be summed to provide the total global damping as a
percentage of critical damping.
Damping source Global damping not to exceed

(% of critical damping)
Structure, holding
system, etc. 2%
Foundation 2% or 0%(1)
Hydrodynamic 3% or 0%(2)
Notes: 1. Where a non-linear foundation model is adopted the hysteresis foundation

damping will be accounted for directly and should not therefore be included
in the global damping.
2. In cases where the relative velocity formulation may be used (α = 1 in

Section 4.3.2) the hydrodynamic damping will be accounted for directly and
should not therefore be included in the global damping.
Table 7.1 - Recommended damping from various sources
7.3.7.5 The effects of directionality and wave spreading may be considered in any dynamic
analysis. It is recommended that a comparison be made between the Base Shear Transfer
Function (BSTF) for the chosen 2-D (long crested/unspread) analysis direction and the 3-
D (short crested/spread) BSTF to determine whether the selected direction is
unconservative. Optimally the direction of the 2-D seastate should be chosen to obtain a
match with the 3-D BSTF for the entire wave spectrum. If this is not possible the match
between the spread and unspread BSTFs should be good at the natural period.
A 3-D BSTF, H3D, can be generated from a set of 2-D BSTFs, H2D, by the following
expression:
2π
∫ [H (ω , θ )] cos 2 n (θ )dθ
2
H3D(ω) = 2D
0
where:
ω = Wave excitation frequency

θ = Angle between 2-D BSTF and dominant direction of 3-D BSTF
n = Power constant of spreading function
≥ 2.0 for fatigue analysis
≥ 4.0 for extreme analysis
A simple approximation to the incorporation of wave spreading into inertial load

calculations is to perform a 2-D analysis with the wave approach angle which is between
the two approach angles which give the maximum and minimum forces at the
cancellation and reinforcement points (see Figure C7.1 in the Commentary).
Rev 3, August 2008
7.3.7.6 Tables 7.2 and 7.3 respectively identify the most important factors associated with each
type of analysis method and with each approach to determining the extreme responses.
Further details of the methods are provided in the Commentary.
7.3.8 Acceptance Criteria
The results of a dynamic extreme response analysis shall be assessed against the
acceptance criteria described in Section 8. The required load factors should be
introduced when combining the component loads into total load combinations.
Rev 3, August 2008
Method Recommendations
Frequency Consider linearization assumptions with
Domain respect to:
- wave-current loading (quadratic dependence
on particle velocity and finite wave ht).
- structural non-linearity.
Generate random sea from at least 200
components and use divisions of equal
frequency.
Note: fewer frequency components may be used
provided that the divisions are shown to be
sufficiently small around the wave period,
the natural period & periods associated
with reinforcement and cancellation.
Time Generate random sea from at least 200 components

Domain and use divisions of generally equal energy. It is
recommended that smaller energy divisions are used
in the high frequency portion of the spectrum, which
will generally contain the reinforcement and
cancellation frequencies. Each wavelet should be
taken to disperse with its own linear dispersion
relationship [12]
Check validity of wave simulation:
- correct mean wave elevation
- standard deviation = (Hs /4) ± 1%
- -0.03 < skewness < 0.03
- 2.9 < kurtosis < 3.1
- Max crest elevation = (Hs/4)√{2ln(N)} -5% to
+7.5%
where N is the number of cycles in the time series
being qualified, N ≈ Duration / Tz
Integration time-step less than the smaller of:
Tz/20 or Tn/20
where;
Tz = the zero-upcrossing period of the wave
spectrum
Tn = the jack-up natural period
(unless it can be shown that a larger time-step leads
to no significant change in results)
Avoid transients in 'run-in' (≥100 secs).
Ensure simulation length OK for method chosen to
determine the Most Probable Maximum Extreme
(MPME) response(s).
Note: The MPME is defined in Table 7.3
Table 7.2 - Recommendations for application of dynamic

analysis methods (see Commentary)
Rev 3, August 2008
Method Recommendations
General Define the Most Probable Maximum Extreme
(MPME) as the extreme with a 63% chance of
exceedence (typically this is the mode or
highest point on the probability density
function (PDF)).This is approximately
equivalent to the 1/1000 highest peak level
in a 3-hour storm.
Frequency Use mean & standard deviation to determine
Domain drag-inertia parameter and use Figure C7.B.6
in Commentary Section C7.B.2.1.
Time Use mean & standard deviation to determine
Domain drag-inertia parameter and use Figure C7.B.5
or Figure C7.B.6 in Commentary Section
C7.B.2.1. Simulation time of at least 60
minutes usually required to obtain stable
standard deviation.
or
Fit Weibull distribution to distribution,
for 3-hour probability level. Take results
as average of MPME's from ≥ 5 simulations.
Each input wave simulation to be of
sufficient length for recommendations of
Table 7.2 to be met (usually at least 60
minutes). See Commentary C7.B.2.2.
or
Use multiple 3-hour simulations and use
Gumbel distribution on the extreme from each
simulation. Sufficient simulations (usually
at least 10) are required to obtain stable
MPME of responses. See Commentary C7.B.2.3.
or
Use Winterstein's Hermite polynomial model,
with improvements by Jensen if Kurtosis > 5.
Simulation of sufficient duration to provide
stable skewness and kurtosis of responses
(normally in excess of 180 minutes). See
Commentary Section C7.B.2.4.
Table 7.3 - Recommendations for determining MPME (see Commentary)

Rev 3, August 2008
7.4 Fatigue
7.4.1 General
The fatigue of jack-ups should be considered for all new locations and operations. Jack-
ups are mobile structures, generally operating in a wide range of water depths, therefore
the location of the fatigue sensitive areas may vary (see Section 7.4.3). This means that
fatigue damage at any member/joint or other component may not occur equally
throughout the life of the unit and tends to complicate the fatigue problem.
If the original analysis carried out for the unit demonstrates that lives of critical
components are adequate then a unit may not require a separate analysis if on location for
a period of less than one year provided that adequate proof from a recent inspection exists
showing that the unit is behaving as originally predicted.
If no original analysis and/or inspection proof is in existence then a separate analysis may
be required for all operations in excess of one year. In extreme cases six months may be
more appropriate if this period contains the rough winter season. Alternatively a recent
assessment inspection, or proof that such an inspection (including detailed NDT) has
been carried out may serve as a demonstration of the adequacy of the unit.
7.4.2 Fatigue Life Requirements
A fatigue analysis, if undertaken, should ensure that all structural components have
(remaining) fatigue lives of more than the greater of four times the duration of the
assignment or 10 years. Different (reduced) fatigue life requirements may be justified for
certain items on a case by case basis where structural redundancy or ease of access for
inspection and repair permit.
7.4.3 Fatigue Sensitive Areas
All structural members subject to fatigue loading are to be checked in the analysis, with
emphasis on the following areas, which are likely to be the most critical. However, other
areas should also be studied if they are potentially more critical:
a) The leg members and joints in the vicinity of the upper and lower guides for the
operating leg/guide location(s).
b) The rack teeth of the chord.
c) The leg members and joints adjacent to the waterline.
d) The jack-frame/jackhouse and associated areas of the hull.
e) The leg members and joints in the vicinity of the leg to spudcan connection.
f) The spudcan to leg connection.
Records of inspections, damage and repair for the unit may provide guidance in the
selection of critical areas.
Rev 3, August 2008
As mentioned the fatigue analysis should consider all loading conditions that may
occurduring the period under consideration and for items c) through f) the cumulative
damage due to transit loadings should also be included.
7.4.4 General Description of Analysis
Suitable approaches to the analysis may be found in reference [13]. Equivalent

approaches may be applied.
Rev 3, August 2008
7 GLOSSARY OF TERMS - DETERMINATION OF RESPONSES
A = Equivalent axial area of a leg (see Figure 5.1), including contribution from rack
teeth (see note to Section 5.6.4).
As = Effective shear area of one leg.
BS = Base Shear.
d = Distance between upper and lower guides.
D = Self weight and non varying loads.
DAF = Dynamic Amplification Factor.
Dn = Inertial loads due to Dynamic response.
E = Environmental loads.
E = Young's modulus for steel.
Fg = Geometric factor
= 1.125 (3 leg unit), 1.0 (4 leg unit)
Fh = Factor to account for horizontal soil stiffness, Khs, and horizontal leg-hull
connection stiffness, Khh.
Fin = Magnitude of inertial loadset.
Fn = 0.5, three chord leg; = 1.0, four chord leg
Fr = Factor to account for hull bending stiffness.
Fv = Factor to account for vertical soil stiffness, Kvs, and vertical leg-hull connection
stiffness, Kvh.
g = Acceleration due to gravity.
h = Distance between chord centers or pinion pitch points.
Hdet = The wave height to be used for deterministic waveforce calculations, allowing for
conservatisms in the theoretical predictions of higher order wave theories.
= 1.60 Hsrp
Hmax = The maximum deterministic wave height.
= 1.86 Hsrp, generally.
= 1.75 Hsrp, in Tropical Revolving Storm areas.
Hs = Significant wave height (meters), including depth/asymmetry correction,
according to Section 3.5.1.1.
Hsrp = The assessment return period significant wave height for a 3 hour storm.
H2D = 2-D base shear transfer function.
H3D = 3-D base shear transfer function.
I = Second moment of area of the leg (see Figure 5.1) including contribution from
rack teeth (see note to Section 5.6.4).
IH = Representative second moment of area of the hull girder joining two legs about a
horizontal axis normal to the line of environmental action.
kf = Combined vertical stiffness of all fixation system components on one chord.
kj = Combined vertical stiffness of all jacking system components on one chord.
ku = Total lateral stiffness of upper guides with respect to lower guides.
Ke = The effective stiffness associated with one leg.
Khh = Horizontal stiffness of leg-hull connection, generally infinite.
Khs = Horizontal stiffness at the spudcan-foundation interface.
Krh = Rotational stiffness representing the leg-hull connection.
Krs = Rotational stiffness at the spudcan-foundation interface.
Kvh = Vertical stiffness of leg-hull connection.
Kvs = Vertical stiffness at the spudcan-foundation interface.
L = Variable loads.
L = Length of leg from the seabed reaction point (see Section 5.2.1) to the point
separating M1a and M1b.
Rev 3, August 2008
7 GLOSSARY OF TERMS - DETERMINATION OF RESPONSES (Continued)
M = Nodal masses.
Me = Effective mass associated with one leg.
Mhull = Full mass of hull, including variable load.
M1a = Mass of a leg above lower guide (in the absence of a clamping mechanism) or
above the center of the clamping mechanism.
M1b = Mass of leg below the point described for M1a, including added mass for the
submerged part of the leg.
MPME = Most Probable Maximum Extreme response(s). The extreme response with a
63% chance of exceedence; approximately equal to the 1/1000 highest peak
level in a 3-hour storm.
n = Power constant of spreading function.
≥ 2.0 for fatigue analysis.
≥ 4.0 for extreme analysis.
N = Number of legs.
N = Number of cycles.
P = The mean force due to vertical dead weight and variable load acting on one
leg.
M hull g
=
N
PE = Euler buckling load of one leg.
= α2EI
T = 0.9 Tp.
Tass = Wave period associated with Hmax (also used with Hdet).
Tn = Natural period of jack-up (subject to the precautions of Section 7.3.5.4).
Tp = Peak period associated with Hsrp (also used with Hs).
Tz = Zero-upcrossing period of the wave spectrum.
Y = Distance between center of one leg and line joining centers of the other two
legs (3 leg unit).
= Distance between windward and leeward leg rows for direction under
consideration (4 leg unit).
α = The minimum positive non-zero value of αL satisfying:
⎧ ( K rs + K rh )αEI ⎫
tan (αL) = ⎨ ⎬
⎩ (αEI) − ( K rs K rh ) ⎪⎭
2
α = Scalar multiplier used in establishing 2-D Fin.

β = Scalar multiplier used in establishing 2-D Fin.
ϕ1,ϕ2 = Structural mode shapes.
Ω = ω/ωn = Tn/T.
ρ = Mass density of water.
θ = Angle between 2-D BSTF and dominant direction of 3-D BSTF.
ω = Wave excitation frequency = 2π/T.
ωn = Jack-up natural frequency = 2π/Tn.
ζ = Damping ratio or fraction of critical damping.
Rev 3, August 2008
8 ACCEPTANCE CRITERIA
The acceptance checks in the following sections cover:

- Structural strength (Section 8.1),
- Overturning stability (Section 8.2),
- Foundation capacity (preload, bearing, sliding displacement and punch-through)
(Section 8.3),
- Horizontal deflections (Section 8.4),
- Loads in the holding system (Section 8.5),
- Loads in the hull (Section 8.6) and
- The condition of the unit (Section 8.7).
Meeting acceptance criteria implies that the factored resistance is equal to or greater than
the internal forces or reactions due to the application of the factored loads.
8.1 Structural Strength Check
Note: Figure 8.1 provides a flowchart for member strength assessment.
8.1.1 Introduction
8.1.1.1 Code Basis

The main basis for the structural strength check is the AISC 'Load and Resistance Factor
Design (LRFD) Specification for Structural Steel Buildings' [14]. The AISC LRFD
specification has been interpreted and, in some cases, modified for use in the assessment
of mobile jack-up unit structures. Interpretation of the code has been necessary to enable
a straight-forward method to be presented for the assessment of beam-columns of non 'I'
section. Development of the code has been necessary in two areas as described below:
a) A method has been established for dealing with sections constructed of steels with
different material properties.
b) A method has been established for the assessment of beam columns under biaxial
bending to overcome a conservatism which has been identified in the standard AISC
LRFD equations.
Rev 3, August 2008
Figure 8.1: Flow chart for member strength assessment

Rev 3, August 2008
One particular type of member geometry which is not covered at all by AISC LRFD is
the high R/t ratio tubular which usually has ring frame and/or longitudinal stiffeners.
Recommendations for checking such members are given in Section 8.1.5 where the user
is referred to an applicable code and guidance is given on suitable load and resistance
factors.
The resistance factors used in the AISC LRFD specification have been adopted.
In addition to checking the strength of members, it may be necessary to check the

strength of joints between members. Recommendations for joint checking are given in
Section 8.1.6 where the user is referred to an applicable code and guidance is given on
suitable load and resistance factors.
8.1.1.2 Limitations
The structural strength check assessment described here is limited by the following
criteria:
a) The geometry of structural components and members, as defined in 8.1.2, must fall
reasonably within the categories described in that section.
b) In accordance with AISC LRFD Specification, Chapter A Para. A5, the minimum
specified yield stress of the strongest steel comprising the components and members
should not exceed:
- 65 ksi (448 MN/m2) if (elasto-)plastic structural analysis is used to determine the

member loads. For slender geometries plastic structural analysis is precluded,
even if the yield stress is below 65 ksi.
- 100 ksi (690 MN/m2) if elastic structural analysis is used to determine the
member loads.
For higher strength steels within the holding system, refer to Section 8.5.
It should also be noted that the assessment has been tailored towards the types of analysis
normally carried out for jack-ups. The detailed recommendations which follow focus
particularly on closed section brace and chord scantlings in truss type legs.
Geometries outside the limits of Sections 8.1.2 - 8.1.4 may be checked in accordance
with the recommendations of Section 8.1.5.
Notes:
1. Of necessity, many of the equations presented in Section 8.1 are dimensional. Such
equations are quoted firstly in metric units (MN, m, MN/m2 etc.) and then in { } in
North American imperial units (kips, inches, ksi, etc.).
2. Where the member geometry may contain components of part-tubular shape it is

appropriate to consider their dimensions in terms of radius and thickness (rather than
diameter and thickness), and hence relevant equations have been converted to this
format.
Rev 3, August 2008
3. The AISC LRFD source equations/text are identified between [ ].
4. The terms in the equations are defined where they appear. A glossary is also
provided at the end of Section 8.
8.1.2 Definitions
8.1.2.1 Structural Members and Components
a) Structural Members
For the purposes of strength assessment, it is necessary to consider the structure as

comprised of structural members. Typically each structural member could be represented
by a single finite element in an appropriate finite element model of the structure.
Examples of members would include braces and chords in truss type legs, box or tubular
legs and plating which forms a piece of structure for which the properties can readily be
calculated.
The strengths of structural members are to be assessed according to Section 8.1.4 with the
exception of structural members exceeding any of the following provisions which should
be assessed according to Section 8.1.5.
i) A plain tubular with R/t > 44,815/Fy

{Imperial: 6,500/Fy}
[Table A-F1.1]
ii) Any tubular with ring stiffeners with or without longitudinal stiffeners.
iii) Tubulars with longitudinal stiffeners where;

R/t > 11,375/Fy {Imperial: 1,650/Fy}
[Table B5.1]
b) Structural Components
A structural component is defined as a part of a structural member (see Figure 8.2).

Typically, structural components are pieces of plating or tubulars such as the plates, split-
tubulars and rack pieces forming a jack-up chord, or the stringers on a panel. Note that it
is not always appropriate to consider fundamental structural parts as components. A
plain tubular, for example is better analyzed as a member. A component should not
consist of more than one material.
8.1.2.2 Stiffened and Unstiffened Components
A component which is stiffened along both edges is denoted a stiffened component. A

component which is supported along only one edge is denoted an unstiffened component.
Typically all the components forming parts of chord sections may be regarded as
stiffened.
Rev 3, August 2008
8.1.2.3 Compact, noncompact and slender sections
Steel sections are divided into compact sections, noncompact sections and sections with
slender compression elements. Compact sections are capable of developing a fully
plastic stress distribution before the onset of local buckling. Noncompact sections can
develop the yield stress in compression components before local buckling occurs, but will
not resist inelastic local buckling at the strain levels required for a fully plastic stress
distribution. Slender compression components buckle elastically before the yield stress is
achieved.
Where a distinction is required between these categories, appropriate limiting slenderness

ratios have been stipulated.
8.1.3 Factored Loads
Factored loads in structural components and members are to be determined in accordance

with the previous sections, using the most onerous condition for each structural
component or member.
Each structural component or member meet the acceptance criteria for the member loads
(i.e. axial load, moments and, if applicable, shears and torsion) resulting from the
application to the jack-up of the factored load Q (as described in 5.7) where;
Q = γ1.D + γ2.L + γ3(E + γ4.Dn)
and
γ1 = 1.0
γ2 = 1.0
γ3 = 1.15 (provisional - see Section 1.8)
γ4 = 1.0
D = The weight of structure and non-varying loads including:

- Weight in air including appropriate solid ballast.
- Equipment.
- Buoyancy.
- Permanent enclosed liquid.
L = The maximum variable load (gravity adds to environmental loads) or

minimum variable load (gravity opposes environmental loads) positioned
at the most onerous center of gravity location applicable to extreme
conditions as specified in Section 3.2.
E = The load due to the assessment return period wind, wave and current
conditions (including associated large displacement effects).
Dn = The inertial loadset which represents the contribution of dynamics over

and above the quasi-static response as described in Section 7.3.6
(including associated large displacement effects).
Rev 3, August 2008
8.1.4 Assessment of Members

- excluding stiffened and high R/t ratio tubulars
8.1.4.1 General interaction equations
Each structural member within the scope of Section 8.1.2 shall satisfy the following
conditions:
If Pu/φaPn > 0.2

1
η η
8 ⎡⎧ M uex ⎫ ⎪⎧ M uey ⎪⎫ ⎤
η
Pu
+ ⎢⎨ ⎬ +⎨ ⎬ ⎥ ≤ 1.0 [Eq. H1-1a]
φa Pn 9 ⎢⎩ φb M nx ⎭ ⎪⎩ φb M ny ⎪⎭ ⎥
⎣ ⎦
else
Pu ⎡⎧ M ⎫η ⎧⎪ M ⎫⎪η ⎤ η
+ ⎢⎨ uex ⎬ + ⎨ ⎬ ⎥ ≤ 1.0
uey
[Eq. H1-1b]
2φa Pn ⎢⎩ φb M nx ⎭ ⎪⎩ φb M ny ⎪⎭ ⎥
⎣ ⎦
where;
Pu = applied axial load
Pn = nominal axial strength determined in accordance with Section 8.1.4.2
(tension) and 8.1.4.3 (compression).
Muex,Muey = effective applied bending moment determined in accordance with Section
8.1.4.4 (tension) and 8.1.4.5 (compression).
Mnx,Mny = nominal bending strength determined in accordance with Section 8.1.4.6.
φa = Resistance factor for axial load = 0.85 for [Eq. E2.1] compression and
0.90 for tension [Eq. D1.1].
φb = Resistance factor for bending = 0.9 [Ch. F1.2]
η = Exponent for biaxial bending, a constant dependent on the member cross
section geometry, determined as follows:
i) For purely tubular members, η = 2.0
ii) For doubly symmetric open section members, η = 1.0
iii) For all other geometries, the value of η may be determined by

analysis as described in Section 8.1.4.7 but shall not be less than 1.0.
In lieu of analysis, a value of η equal to 1.0 may be used.
The interaction equations can be used in a reduced form if one or two of the three load
ratio terms in the equation are zero.
Alternatively, the more complex interaction formulations given in Section C8.1.4.7 of

the Commentary may be used where applicable.
Rev 3, August 2008
8.1.4.2 Nominal Axial Strength of a Structural Member in tension Pn
For a member comprising more than one component, the nominal tensile strength lies
between the maximum individual tensile strength of any one component, and the sum of
all the individual tensile strengths.
The nominal tensile strength of a tension component shall be the lower value from the
following equations:
a) Pni = FyiAi
b) Pni = 5 6 FuiAi
where;
Ai = area of component
Fyi = specified minimum yield stress of component (or specified yield strength
where no yield point exists)
Fui = specified minimum tensile (ultimate) strength of component
Pni = component nominal axial tensile strength
This assumes that for members in jack-up units the net section is equal to the gross
section [Eq's. D1.1 and D1.2].
The total member nominal tensile strength shall be:

Pn = FminΣAi
with the resistance factor
φt = 0.90 [Eq. D1.1]
where Fmin is the smallest value of Fyi or 5 6 Fui of all the components.
Note: If for any component the nominal strength is significantly different from the
nominal strengths of other components, the formulation above may be conservative and
alternative rational methods may be applied. An example is given in the Commentary.
8.1.4.3 Nominal Axial Strength of a Structural Member in Compression Pn
So long as local buckling of the components of a member is not the limiting state, the
member can be treated for global loads only. Should local buckling dominate, the loads
in the components must be considered. Therefore, in determining the nominal axial
strength of a member in compression, a local buckling check must first be applied.
Check: Local buckling
The structural components which make up the cross section of a compact or noncompact
section must satisfy the following criteria [Table B5.1]:
i) For rectangular components stiffened along both edges

bi/ti ≤ 625/ ( Fyi − Fr )
{Imperial: bi/ti ≤ 238/ ( Fyi − Fr ) }
Rev 3, August 2008
ii) For rectangular components stiffened along one edge

bi/ti ≤ 250/ ( Fyi ) {Imperial: bi/ti ≤ 95/ ( Fyi ) }
iii) For tubular sections
R/ti ≤ 11380/Fyi {Imperial: R/ti ≤ 1650/Fyi}
where;
bi = width of a rectangular component
ti = thickness of a rectangular component or tube wall
R = outside radius of the tube or tubular component
Fr = residual stress due to welding (114 MPa, {16.5 ksi})
Members containing rectangular and tubular sections which do not meet this criteria are
considered to be slender and are treated in 8.1.4.3 b) for local buckling.
a) Strength assessment for Compact and Noncompact Sections
The nominal axial strength of a structural member subject to axial compression and
within the above stipulated restrictions regarding cross section shall be determined from
the following equations:
Pn = A Fcr [Eq. E2.1]
λc2
Fcr = (0.658 ) Fyeff For λc ≤ 1.5 [Eq. E2.2]
⎧ 0.877 ⎫
Fcr = ⎨ 2 ⎬ Fyeff For λc > 1.5 [Eq. E2.3]
⎩ λc ⎭
where;
A = gross area of section (excluding rack teeth of chords)
1
Kι ⎧ Fyeff ⎫ 2
λc = ⎨ ⎬ for max. Kι/r from all directions [Eq. E2.4]
rπ ⎩ E ⎭
ι = unbraced length of member:
- face to face for braces
- braced point to braced point for chords
- longer segment length of X-braces (one pair must be in tension, if not
braced out of plane)
r = radius of gyration, based on gross area of section.
E = material Young's modulus (200,000 MN/m2 {29,000 ksi}).
Fyeff =effective material yield stress, to be taken as the minimum of (specified)
yield stress or 5/6 (ultimate stress) of all components in the member unless
rational analysis shows that a higher value may be used.
K = effective length factor. Figure 8.3 provides generally recommended values for
K. For the specific case of jack-up truss legs, the value of K shall be taken as
follows [Table C-C2.1], unless alternative values are shown applicable by
rational analysis:
Assumed boundary conditions
Chord members 1.0 pinned-pinned
K-Braces & span breakers 0.8 between pinned-pinned
X-Braces 0.9 and fully built-in
Complete legs 2.0 pinned-sliding
Rev 3, August 2008
b) Strength Assessment for Members with Slender Components
The nominal axial strength of a structural member subject to axial compression and
outside the restrictions for a) above shall be determined from the following equations.
Pn = A Fcr
where;
2
Fcr = Q(0.658Qλc )Fyeff for λc Q ≤ 1.5 [Eq. A-B5-11]
⎧ 0.877 ⎫
Fcr = ⎨ 2 ⎬ Fyeff for λc Q > 1.5 [Eq. A-B5-13]
⎩ λc ⎭
where λc is defined in Section 8.1.4.3 a) and Q is determined from the
following:
i) For members comprising entirely of stiffened components [A-B5.3.b

and A-B5.3.c]:
Q = Qa
where;
Qa = Ae/A [Eq. A-B5-10]
and
Ae is the section effective area found from:
Ae = Σ bei ti (excluding rack teeth of chords)
with
856t i ⎧⎪ 170 ⎫⎪
bei = ⎨1 − ⎬ ≤ bi
f i ⎪⎩ ( b i / t i ) f i ⎪⎭
326t i ⎧⎪ 64.9 ⎫⎪
{Imperial: bei = ⎨1 − ⎬ ≤ bi }
f i ⎪⎩ ( b i / t i ) f i ⎪⎭
[Eq A-B5-7]
and fi is the calculated elastic stress in the component where, for the
analysis, the member area is based on the actual cross sectional area but
with elastic section modulus and radius of gyration based on effective
area.
ii) For members comprising of stiffened and unstiffened components

[A-B5.3.b and A-B5.3.c]:
Q = Qa Qs
where Qa is determined from Section 8.1.4.3 b) i) but with the additional

check that fi for the stiffened component must be such that the maximum
compressive stress in the unstiffened component does not exceed φcFcr
with Fcr defined in Section 8.1.4.3 b) with Q = Qs and φc = 0.85 or
φbFyeffQs with φb = 0.90.
Rev 3, August 2008
Qs is the lowest value for all components in the member which are
stiffened along one edge determined from the following:
For 250/ Fy < bi/ti < 460/ Fy

{Imperial: 95/ Fy < bi/ti < 176/ Fy }
Qs = 1.415 - 0.00166(bi/ti) Fy
{Imperial: 1.415 - 0.00437(bi/ti) Fy } [Eq. A-B5-3]
For bi/ti ≥ 460/ Fyi {Imperial: bi/ti ≥ 176/ Fyi }
Qs = 137,900/[Fyi(bi/ti)2]
{Imperial: Qs = 20,000/[Fyi(bi/ti)2]} [Eq. A-B5-4]
Note: The implication of this section is that the critical components in

the member will be the unstiffened components. If these buckle, then
the assumed buckling lengths and hence strengths for the stiffened
components will then be wrong, hence invalidating the original
assumptions. This assumes that the unstiffened components are placed
in the member to reduce the buckling length of the major components.
iii) For members comprising a tube alone and:
11,375/Fy < R/t < 44,815/Fy

{Imperial: 1,650/Fy < R/t < 6,500/Fy}
3790 2 550 2
Q = + {Imperial: Q = + }
Fy ( R / t ) 3 Fy ( R / t ) 3
[Eq. A-B5-9]
8.1.4.4 Effective Applied Moment for Members in Tension; Mue (Muex,Muey)
In many cases, the effective applied moments used in the interaction equations will not be
equal to applied moments obtained in a structural analysis. This can be due to the type of
structural model and /or the effective length effect on buckling. The following
procedures shall be followed for the determination of the effective applied moment.
The effective applied moment for a member under axial tension shall be taken to be equal
to the applied moment from an analysis including global P-Δ effects and accounting for
local loading.
Rev 3, August 2008
8.1.4.5 Effective Applied Moment for Compression Members; Mue (Muex,Muey)
The effective applied moment for a member under axial compression shall be taken to be:
Mue = B Mu [Eq. H1.2]
where;
Mu is the applied moment determined in an analysis which includes global P-
Δ/hull-sway effects and accounts for local loading. When eccentricity is not
incorporated in the model, the equation for Mue should be modified to include pue
due to the eccentricity, e, between the elastic and plastic neutral axes. Note: When
the member considered represents the leg the requirement to include P-Δ effects in
the global analysis means that the provisions of ii) below apply.
and
i) Where the individual member loads are determined from a first order linear elastic
analysis i.e. the equilibrium conditions were formulated on the undeformed
structure, (For example a linear analysis of a detailed truss type leg, using external
loads determined from a second order analysis of a simplified global model):
Cm
B = ≥ 10
. [Eq. H1-3]
(1 − Pu / PE )
where:
PE = (π2r2AE)/(Kι)2 with K ≤ 1.0 and PE is to be calculated for the plane of
bending. A is defined in Section 8.1.4.3 a) and r is the radius of gyration
for the plane of loading.
Cm = a coefficient whose value shall be taken as follows [Ch. H1.2a]:
i) For members not subject to transverse loading between their

supports in the plane of bending Cm = 0.6 - 0.4 (M1/M2) [Eq. H1-4]
where M1/M2 is the ratio of the smaller to the larger moments at the
ends of that portion of the member unbraced in the plane of bending
under consideration. M1/M2 is positive when the member is bent in
reverse curvature, negative when bent in single curvature.
ii) For members subjected to transverse loading between their supports,

the value of Cm can be determined from rational analysis. In lieu of
such analysis, the following values may be used:
For members whose ends are restrained against
sidesway Cm = 0.85
For members whose ends are unrestrained against
sidesway Cm = 1.0
ii) Where the individual member loads are determined from a second order analysis
i.e. the equilibrium conditions were formulated on the elastically deformed structure
so that local P-Δ loads were also included in the analysis:
B = 1.0
Rev 3, August 2008
8.1.4.6 Nominal Bending Strength; Mn (Mnx, Mny)
The calculation of nominal bending strength is based on the plastic properties of the
section. The practice allows for hybrid sections built up from components of different
yield strengths. Standard techniques shall be applied to obtain a section plastic moment
in the absence of axial load, Mp, based on the individual component values which are the
lesser values of Fyi and 5/6 Fui (an example is given in the Commentary).
Lateral torsional buckling and local buckling of components must be considered.
If both tensile and compressive yielding occur during the same load cycle, it shall be
demonstrated that the structure will shake down without fracture.
Check: Lateral torsional buckling (Not applicable to tubulars)

The cross sectional geometry of a member subjected to bending shall be examined for
susceptibility to the limit state of lateral torsional bucking. The member cross section
must satisfy the following criteria for compact sections for the nominal bending strength
to be assessed under Sections 8.1.4.6 a) or 8.1.4.6 b).
Lb/ry ≤ 25860 ( JA) / M p {Imperial: Lb/ry ≤ 3750 ( JA) / M p }

[Table A-F1.1]
where;
Lb = Laterally unbraced length; length between points which are either braced
against lateral displacement of the compression flange or braced against
twist of the cross section.
ry = Radius of gyration about the minor axis.
A = Cross sectional area.
J = Torsional constant for the section
Sections which do not satisfy this criteria are susceptible to lateral torsional buckling and
are treated as having slender compression components as in Section 8.1.4.6 c).
Check: Local buckling

The cross sectional geometry of a member subjected to bending is to be examined for
susceptibility to the limit state of local bucking. If local buckling is deemed to be the
limit state, the nominal bending strength shall be reduced in accordance with the
following paragraphs. Members with particularly slender components are covered in
Section 8.1.4.6c).
For this check it is necessary to identify web components and flange components. This
can be done by visual inspection, with knowledge of the major and minor axes. For
example, in a split-tubular, opposed rack chord, the rack plate would be a suitable web
component, and the split tubulars flanges. For a teardrop chord, the rack and side plates
would be web components, and the back plate the flange. In cases of doubt, components
shall be checked as both web and flange.
Rev 3, August 2008
a) Compact Sections
For members in which all the components sections satisfy the following [Table B5.1]:
bi/ti ≤ λp
where;
λp = 500/ ( Fyi ) {Imperial: λp = 190/ ( Fyi ) }
bi/ti ≤ λp
where;
λp = 170/ ( Fyi ) {Imperial: λp = 65/ ( Fyi ) }
2R/t ≤ λp
where;
λp = 14270/Fyi {Imperial: λp = 2070/Fyi}
The nominal bending strength is given by the plastic bending moment of the whole
section
Mn = Mp [Eq-A-F1-1]
where Mp is derived as discussed above.
Note: Where significant plastic hinge rotations are required the section must remain
stable after rotation through an appreciable angle. In such cases, to achieve this
requirement, the limitations of ii) and iii) above should be reduced to:
ii) λp = 135/ ( Fyi ) {Imperial: λp = 52/ ( Fyi ) }
iii) λp = 11000/Fyi {Imperial: λp = 1600/Fyi}
b) Noncompact Sections
For members in which all the components do not satisfy the previous criteria but satisfy
the following [Table B5.1]:
bi/ti ≤ λr
where;
λr = 625/ ( Fyi − Fr ) {Imperial: λr = 238/ ( Fyi − Fr ) }
Fr = 114 MN/m2 {16.5 ksi} residual stress
bi/ti ≤ λr
where;
λr = 278/ ( Fywj − Fr ) {Imperial: λr = 106/ ( Fywj − Fr ) }
Fywj = web component yield stress.
Fr = 114 MN/m2 {16.5 ksi} residual stress.
2R/t ≤ λr
where;
λr = 61850/Fyi {Imperial: λr = 8970/Fyi}
Rev 3, August 2008
The nominal bending strength is given by an interpolation between the plastic bending
moment and the limiting buckling moment:
⎪⎧ λ − λ p ⎪⎫
Mn = Mp - (Mp- Mr) ⎨ ⎬ [Eq. A-F1.3]
⎪⎩ λ r − λ p ⎪⎭ h
where;
Mp = Section Plastic Moment.
h = subscript referring to the component which produces the smallest value of
Mn .
λ = b/t or 2R/t as applicable for component h.
λp is determined for component h from 8.1.4.6 a).
λr is determined for component h from 8.1.4.6 b).
Mr is the limiting buckling moment of the section defined as follows:
For bending of non-tubular sections about the major axis, the lesser of
Mr = Fl S (flange buckling) [Table A-F1.1]
Mr = ReFyfjS (web buckling) [Table A-F1.1]
where;
Fl = the smaller of (Fyfj – Fr) and Fywj
S = minimum section elastic modulus for plane of bending under
consideration.
For bending of non-tubular sections about the minor axis;

Mr = FyfjS (flange buckling) [Table A-F1.1]
For bending of tubular sections: [Table A-F1.1]

⎧ 2068 ⎫ ⎧ 300 ⎫
Mr = ⎨ + Fy ⎬S {Imperial: Mr = ⎨ + Fy ⎬S }
⎩R/t ⎭ ⎩R / t ⎭
Fyfj = yield stress of flange component.
Re = hybrid girder reduction factor [A-G2]
= 1.0 if components are of the same material
otherwise:
= [12 + ar (3m – m3)] / (12 + 2 ar) ≤1.0
ar = ratio of total web area to area of compression flange.
m = ratio of web component yield stress to flange component yield stress
which gives smallest value of Re.
c) Slender Sections
The nominal bending strength of members including components which do not satisfy the
above criteria for compact and noncompact sections or for lateral torsional buckling shall
be determined in accordance this section.
Rev 3, August 2008
The nominal bending strength of a member is given by the limiting flexural bending
moment:
Mn = S Fcr
where S is the elastic section modulus for the plane of bending under consideration and
Fcr is the lowest value from (where appropriate):
i) Doubly symmetric members (lateral torsional buckling)
12
C b X1 2 ⎧ X1 2 X 2 ⎫
Fcr = 6.895 ⎨1 + ⎬ [Table A-F1.1(b)]
λ ⎩ 2λ2 ⎭
12
C X 2 ⎧ X1 2 X 2 ⎫
{Imperial: Fcr = b 1 ⎨1 + ⎬ }
λ ⎩ 2λ2 ⎭
where;
Cb = 1.75 + 1.05(M1/M2) + 0.3(M1/M2)2 ≤ 2.3 where M1 is the smaller and M2
the larger end moment in the unbraced member; M1/M2 is positive when
the moments cause reverse curvature.
X1 = (π/S) ( EGJA / 2)
X2 = (4Cw/Iy)(Sx/GJ)2
E = Modulus of elasticity (200,000 MN/m2 {29,000 ksi}).
G = Shear modulus of elasticity (77,200 MN/m2 {11,200 ksi}).
J = Torsion constant for section.
A = Cross-sectional area (excluding rack teeth).
Iy = Second moment of area of section about minor axis.
Sx = Elastic section modulus for major axis bending.
Cw = Warping constant.
λ = Lb/ry
ry = Radius of gyration about the minor axis
ii) Singly symmetric members (lateral torsional buckling) [Table A-F1.1(c)]
393,000C b
Fcr = {B1 + (1 + B2 + B1 2 ) } ( I y J ) ≤ Fy
SL b
57,000C b
{Imp'l: Fcr = {B1 + (1 + B2 + B1 2 ) } ( I y J ) ≤ Fy}
SL b
where;
⎧⎪ ⎛ I c ⎞ ⎫⎪⎧ h ⎫⎧ I y ⎫ 2
1
B1 = 2.25 ⎨2⎜⎜ ⎟⎟ − 1⎬⎨ ⎬⎨ ⎬

⎪⎩ ⎝ I y ⎠ ⎪⎭⎩ L b ⎭⎩ J ⎭
⎧⎪ ⎛ I c ⎞ ⎫⎪⎧ h ⎫ 2 ⎧ I c ⎫
B2 = 25 ⎨1 − ⎜⎜ ⎟⎟ ⎬⎨ ⎬ ⎨ ⎬
⎪⎩ ⎝ I y ⎠ ⎪⎭⎩ L b ⎭ ⎩ J ⎭
h = web depth.
Ic = second moment of area of compression flange about the section minor axis
Cb = as for doubly symmetric sections.
Rev 3, August 2008
iii) Doubly and singly symmetric members (flange local buckling)
[Table A-F1.1(g)]
Fcr = 77,220/λ2 {Imperial: Fcr = 11,200/λ2}
where;
λ = bi/ti for flange(s)
iv) For tubular members (Local buckling) [Table A-F1.1]

Fcr = 33,610/(R/t) {Imperial: Fcr = 4,875/(R/t)}
where;
R = radius of tubular
t = wall thickness of tubular
8.1.4.7 Determination of η for non-tubular sections
The general interaction equation requires that applied bending moments are resolved into
components in two perpendicular axes (X,Y). For elaborate sections such as chords,
these axes may be selected on the basis of section geometry and not on load incidence.
Therefore neither of these axes need be coincident with the angle of load. The use of the
exponent η is necessary to ensure that the effective nominal bending strength of the
section is not significantly influenced by this choice of axes.
To determine a suitable value of η the following process is applied:

1. For angles q = 0°, 30°, 45°, 60°, 90°, to the X-axis, obtain the allowable bending
strengths Mnq.
2. Assume loads incident at angle q=30°, the limiting bending moment in the absence of
axial load is Muq = Mnq. When the section is non-compact Mnq is a function of Mr and
a suitable analysis in line with that of Section 8.1.4.6 b), should be applied in
determining Mr for non-principal axes.
3. Resolve this limiting moment into limiting components M'uex and M'uey about the X
and Y axes:
M'uex = Muq cos q
M'uey = Muq sin q
4. The interaction ratio for the (X,Y) axis pair is required to give the same result as if the
X-axis were lined up with the q-axis:
1
⎧⎪⎛ M ' ⎞ η ⎛ M ' uey ⎞ η ⎫⎪ η M uq
⎨⎜ ⎟ + ⎜⎜ ⎟⎟ ⎬ =
uex
i.e.
⎪⎩⎝ M ⎠ ⎝ M ny ⎠ ⎪ M nq
nx
⎭
Since Mnx and Mny are by definition Mnq for q = 0° and 90° respectively and so are
known, the only unknown in the above identity is η. This can be determined from
the graph in Figure 8.4 or by numerical means if preferred. Figure 8.4 is based on
the ratio Muq/Mnq being equal to unity, and will produce conservative results when
axial loads are present.
5. Step 4 yields a value of η suitable for loads from 30° to the X-axis. Steps 2 to 4 are
repeated for q = 45° and 60° to obtain a range of values of η.
6. The value of η for use in subsequent assessment shall be the least of the above
determined values, but not less than 1.0.
Rev 3, August 2008
This method includes some approximation. Since bending will not be along or
perpendicular to a plane of symmetry, deflection will not necessarily be at the same angle
as the applied moment. This effect is second order.
Note: An alternative, more detailed approach, involving modified interaction equations

is presented below for a number of typical chord configurations.
8.1.4.8 Plastic Interaction Curve Approach
Alternatively, interaction equations and curves for generic families of chords are
presented in Figures C8.1.8 - C8.1.11 in the Commentary. These are taken from Dyer
[19] and based on the interaction approach proposed by e.g. Duan & Chen [20]. It
should be noted that the curves and equations are based on axial load applied at the
'center of squash' which is defined as the location at which the axial load produces no
moment on the yielded section. For chords without two axes of symmetry (triangular
and tubular with offset rack) this is offset from the elastic centroid when the section is
comprised of materials of differing yield strengths. Before a section is checked it is
necessary to correct as appropriate moments by the axial load times the offset distance
between the elastic centroid (used in the structural analysis) and the 'center of squash'.
This offset, together with other geometric data for the members of each family of
chord is presented in Tables C8.1.1 to C8.1.4 in the Commentary. The effective
applied moment may then be calculated from:
Muex = Bx(Mux + Pu.ey)
Muey = By(Muy + Pu.ex)
The interaction equations are based on ultimate capacity. It is therefore necessary to
introduce the required resistance factors. This is achieved by defining:
Py = F1.φa.Pn
Mpx = F2.φb.Mnx
Mpy = F2.φb.Mny
where; F1 = 1.0, unless alternative values are justified by analysis.
F2 = 1.0, unless alternative values are justified by analysis.
The ratio of Pu/Py, Muex/Mpx and Muey/Mpy shall be determined for the condition under
consideration. The user should then enter the plastic interaction curves with the
Muex/Mpx and Muey/Mpy ratios. The allowable value for Pu/Py may then be determined.
A measure of the interaction ratio can then be obtained as the ratio between the actual
and allowable values of Pu/Py.
The user should note that the equations for sections with only one axis of symmetry
depend on the sign of the moment about the Y-Y axis (given in the Figures). The sign
convention should be observed with care.
The equations are based on lower bound data from each family of chord shape and will
therefore tend to be conservative. More accurate results will be obtained from the
individual consideration of the chord in question.
[NOTE: At present Figures C8.1.8 - C8.1.11 in the Commentary cover only fully
plastic section strength considerations, and their use for a beam-column member is
based on the assumption that the member being evaluated is sufficiently short/compact
that elasto-plastic stability (buckling at large strains) is not a consideration. Violating
this assumption may lead to errors on the unsafe side. Updated information covering
elasto-plastic stability may be generated in the future, and should preferentially be
used for member evaluations.]
Rev 3, August 2008
8.1.5 Assessment of other member geometries
It is recommended that other member geometries are assessed using the relevant
provisions of AISC LRFD [14] or, for stiffened or high R/t ratio shell members, the DNV
Rules for fixed offshore installations in conjunction with the DNV Classification note on
Buckling Strength Analysis of Mobile Offshore Units [15].
For these geometries, the nominal strength/resistance factors shall be the same as given in
the relevant codes, but the load cases and factored loads should be determined in
accordance with Section 8.1.3 rather than using the factors in the reference.
8.1.6 Assessment of member joints
It is recommended that the assessment of joints of members which form a truss structure
be carried out in accordance with AISC LRFD [14] or API LRFD [16] as appropriate for
the joint under consideration. The factored loads should be determined in accordance
with Section 8.1.3, rather than using the factors in the references.
Figure 8.2: Typical members and components

Rev 3, August 2008
Figure 8.3: Effective Length Factors (from AISC-LRFD [14])
Figure 8.4: Chart for Determination of η

Rev 3, August 2008
8.2.1 For independent leg jack-ups the assumed overturning axis shall be the most critical axis
passing through any two spudcan reaction points as defined in Section 5.2.
8.2.2 The overturning moment shall be calculated from the components of environmental
loading, resolved normal to the overturning axis, times the vertical distance from the
point of action of the component to the overturning axis.
The overturning stability should meet the overturning requirements of 8.2.3, based on
the overturning moment MO resulting from the application to the jack-up of the factored
load Q described in 8.1.3.
8.2.3 The unit shall be shown to satisfy the following overturning requirements:
MO ≤ φ1.MD + φ2.ML + φ3.MS
where;
MD = The stabilizing moments due to weight of structure and non-varying loads
(at the displaced position resulting from the factored loads - see note)
including:
- Weight in air including appropriate solid ballast.
- Equipment.
- Buoyancy.
- Permanent enclosed liquid.
ML = The stabilizing moment due to the most onerous combination of
minimum variable load and center of gravity applicable to extreme
conditions as specified in Section 3.2 (at displaced position - see note).
MS = The stabilizing moments due to seabed foundation fixity (these shall not
be taken into account unless specific calculations for the location and the
spudcan concerned show that a significant contribution from seabed
fixity may be expected).
φ1 = R.F. for dead load moments (MD) = 0.95
φ2 = R.F. for live load moments (ML) = 0.95
φ3 = R.F. for seabed moments (MS) = 0.95
Note: It may be convenient to consider the reduction in dead and live load stabilizing
moment caused by the displacement resulting from the factored loads as an
increase in the overturning moment, rather than as a reduction in the stabilizing
moment.
Rev 3, August 2008
8.3 Foundation assessment

The foundation assessment shall be carried out in a step-wise manner until the
requirements of the current stage are satisfied when it is not necessary to proceed further.
The philosophy is described in Section 6.3 and shown in Figure 6.9.
8.3.1 Step 1 - Preload and Sliding checks
Step 1a - Preload check
8.3.1.1 A preload check shall be used to verify the adequacy of the leeward leg foundation. The
acceptance criteria for the windward leg are discussed in Section 8.3.1.5.
8.3.1.2 The preloading capability should be checked for the vertical leg reaction Qv caused by
the following factored loads resulting from the application to the jack-up of the factored
load Q described in 8.1.3.
8.3.1.3 The preload capacity shall be shown to be sufficient to satisfy the following
requirements:
Qv ≤ φp.VLo
where;
VLo = Vertical leg reaction during preloading
φp = R.F. for foundation capacity during preload
= 0.9 (see Commentary)
Rev 3, August 2008
8.3.1.4 In dense sands (i.e. with maximum bearing area not mobilized) and in clayey soils the
preload check may be applied if the leeward leg horizontal reaction QH < 0.1VLo (with
QH determined in accordance with Section 8.3.1.5). For a spudcan fully embedded in
sand the preload check may be applied if the leeward leg horizontal reaction QH <
0.03VLo. In all other cases a pinned condition bearing capacity check of the foundation
shall be carried out in accordance with Section 8.3.2 (see Commentary).
8.3.1.5 Step 1b - Sliding Resistance - Windward Leg(s)
a) The sliding capacity of the windward leg(s) should be checked for the horizontal leg
reaction QH in association with the vertical leg reaction Qy, both resulting from the
application to the jack-up of the factored load Q described in 8.1.3.
b) The foundation shall be shown to satisfy the following capacity requirements:

QH ≤ φHfc.FH
where;
FH = foundation capacity to withstand horizontal loads when load QV is acting
φHfc = R.F. for horizontal foundation capacity (see Commentary).

= 0.80 (effective stress - sand/drained).
= 0.64 (total stress - clay/undrained).
Rev 3, August 2008
8.3.2 Step 2a - Capacity check - pinned foundation
8.3.2.1 The bearing capacity of the leeward leg should be checked for the leg reaction vector
QVH of vertical and horizontal leg reaction resulting from the application to the jack-up
of the factored load Q described in 8.1.3.
8.3.2.2 The leeward leg foundation shall be shown to satisfy the following capacity
requirements:
QVH ≤ φVH.FVH
where;
FVH = foundation capacity to withstand combined vertical and horizontal loads
taken as a vector from the still water load vector in the same direction as
QVH.
φVH = R.F. for foundation capacity (see Commentary).

= 0.90 - Maximum bearing area not mobilized.
= 0.85 - Penetration sufficient to mobilize maximum bearing area.
8.3.2.3 The windward leg foundations should be checked according to the requirements of
Section 8.3.1.5.
Rev 3, August 2008
8.3.3 Step 2b - Capacity check - with foundation fixity
8.3.3.1 The foundation capacity of the leeward and windward legs should be checked for the leg
reaction vector, including vertical and horizontal leg reaction and the associated can
moment, QVHM, resulting from the application to the jack-up of the factored load Q
described in 8.1.3.
8.3.3.2 The leg reaction vector QVHM shall be checked to satisfy the yield surface as defined in
6.3.4.
8.3.3.3 The windward and leeward leg foundations shall also be shown to satisfy the bearing
capacity and sliding capacity requirements of 8.3.2.
Rev 3, August 2008
8.3.4 Step 3 - Displacement check
If the factored loads on any footing exceed the factored capacity discussed above a
further assessment may be performed in order to show that any additional settlements
and/or the associated additional structural loads are within acceptable limits. See Section
6.3.5.
8.3.5 Punch-through
The selection of factors of safety against punch-through should be made using sound
engineering judgment, accounting for the accuracy of the available soil data and the
magnitude of any possible sudden penetration (see Commentary).
When the possibility of punch-through exists during the installation and preloading
phases it may be applicable to consider the magnitude of possible sudden penetration in
comparison with the structural capability of the unit to resist punch-through.
If the possibility of punch-through remains once the unit has been installed on location
and elevated to the operational airgap the evaluation should account for long term effects
(e.g. cyclic degradation).
8.4 Horizontal Deflections
When working close to or over a platform the assessor shall, if required by the platform
owner, provide the extreme deflections of the jack-up to the platform owner (see Section
5.5.1 of the GUIDELINE).
8.5 Loads in the Holding System
8.5.1 The holding system (elevation and/or fixation system) is deemed to be the system which
forms the load path connecting the hull to the legs.
8.5.2 The loads in the holding system shall not exceed those specified by the manufacturers,
unless the basis of the limitations and the equivalent reference stress levels are stated,
when the loads in the holding system resulting from the application to the jack-up of the
factored load Q described in 8.1.3 may be compared with the ultimate capacity
multiplied by a R.F. (φ) of 0.85.
8.5.3 he stresses in the structural members connecting the holding system to the hull shall be in
accordance with the requirements of Section 8.1.
8.6 Hull
8.6.1 It is assumed that the jack-up hull is designed and built to the structural/scantling
requirements of a recognized Classification Society and carries a valid Class Certificate.
8.6.2 For jack-ups where 8.6.1 does not apply it shall be shown that the hull has adequate
strength to withstand appropriate combinations of dead load, variable load, environmental
load, deflections, preload conditions and dynamics effects.
Rev 3, August 2008
8.7 Structure Condition Assessment
The objective of the site specific assessment is to ensure an appropriate level of structural
reliability of the jack-up in the elevated condition. To achieve this, account must be
taken of any deterioration in the jack-up structure (see Section 1.3.4 of the GUIDELINE).
The condition of the structure is the responsibility of the owner and is deemed to be
satisfactory if the jack-up has valid class certification as described in Section 2.4.2.
Normally the owner can thus provide the assessor with all the information required to
satisfy the structure condition requirement.
In special cases (usually at the option of the operator), an on site structural inspection
may be required to assess the condition of the jack-up. Guidance for such an on site
structural inspection is given in the Commentary. In the event that the results of this
inspection reveal deterioration of the structure, due account of such deterioration shall be
taken into account in the assessment.
Rev 3, August 2008
8 GLOSSARY OF TERMS - ACCEPTANCE CRITERIA
ar = Ratio of total web area to area of compression flange.

A = Cross sectional area of a member (excluding rack teeth).
Ae = Section effective area (excluding rack teeth).
Ai = Area of a component in a member.
bei = Effective width of a component.
bi = Width of a rectangular component.
B,B1,B2 = Factors used in determining Mu for combined bending and compressive axial
load.
Bx,By = Moment amplification factors.
Cb = Bending coefficient dependent on moment gradient.
Cm = Coefficient applied to bending term in interaction formula for prismatic
members dependent upon column curvature caused by applied moments.
Cw = Warping constant.
D = Dead load vector due to the self-weight of the structure and non-varying loads.
Dn = The load vector due to the inertial loadset which represents the contribution of
dynamics over and above the quasi-static response (including associated large
displacement effects).
e , ex , ey = Eccentricity between elastic and plastic neutral axes.
E = Load due the to assessment return period wind, wave and current conditions
(including associated large displacement effects).
E = Modulus of elasticity (200,000 MN/m2 {29,000 ksi}).
fi = Component compressive stress.
Fcr = Critical stress.
FH = Foundation capacity to withstand horizontal loads when QV is acting.
Fmin = The smaller value of Fyi and (5/6)Fui of all the components (in a member).
Fr = Residual stress due to welding (114 MN/m2).
FVH = Foundation capacity to withstand combined vertical and horizontal loads.
FVHM = Foundation capacity to withstand combined vertical, horizontal and moment
loads.
Fy = Minimum specified yield stress or specified yield strength.
Fyh = Minimum yield stress or specified yield strength of component with highest b/t
ratio.
Fyeff = Effective material yield stress for consideration of axial buckling.
Fyi = Minimum specified component yield stress or specified yield strength.
Fywj = Minimum specified web yield stress or specified yield strength.
Fyfj = Minimum specified flange yield stress or specified yield strength.
Fui = Component material ultimate strength.
G = Shear modulus of elasticity.
h = Subscript referring to the component which produces the smallest value of Mn.
h = Web depth.
Ic = Second moment of area of compression flange.

I = Second moment of area of section.
Ix = Second moment of area of section about major axis.
Iy = Second moment of area of section about minor axis.
J = Torsional constant for the section.
Rev 3, August 2008
8 GLOSSARY OF TERMS - ACCEPTANCE CRITERIA (continued)
K = Effective length factor.

ι = Unbraced length of member; face to face for braces, braced point to braced
point for chords.
L = The load vector due to the maximum or minimum variable load positioned at
the most onerous center of gravity location applicable to extreme conditions.
Lb = Laterally unbraced length; length between points which are either braced
against lateral displacement of compression flange or braced against twist of the
cross section.
m = Ratio of web component yield stress to flange component yield stress which
gives smallest value of Re.
MD = Stabilizing moment due to self weight.
ML = Stabilizing moment due to most onerous combination of variable load and

center of gravity.
MS = Stabilizing moment due to seabed foundation fixity.
Mn, Mnx,
Mny = Nominal bending strength.
Mnq = Allowable bending strength about axis q.
MO = Factored overturning moment.
Mp = Section plastic moment.
Mpx = Plastic moment capacity about member x-axis.
Mpy = Plastic moment capacity about member y-axis.
Mr = Limiting buckling moment of section.
Mu = Applied moment determined in an analysis which includes global P-Δ effects
and accounts for local loading.
Mue, Muex,
Muey = Effective applied bending moment.
M'uex, M'uey = Limiting components of applied bending moment.
Muq = Assumed limiting bending moment about axis q in absence of axial load.
M1 = Smaller end moment of a member.
M2 = Larger end moment of a member.
PE = Euler buckling strength.
Pu = Applied axial load.
Pn = Nominal axial strength.
Pni = Component nominal axial strength.
Py = Axial yield strength.
q = Angle of load heading with respect to defined X axis.
Q = Factored load vector.
Q = Full reduction factor for slender compression components.
Qa = Reduction factor for slender stiffened compression components.
QH = Factored horizontal leg reaction.
Qs = Reduction factor for slender unstiffened compression components.
QV = Factored vertical leg reaction.
QVH = Factored leg reaction vector of vertical and horizontal loads.
QVHM = Factored leg reaction vector of vertical, horizontal and moment loads.
r = Radius of gyration.
rx = Radius of gyration about the major axis.
ry = Radius of gyration about the minor axis.
R = Outside radius of the tube or tubular component.
Re = Hybrid girder reduction factor.
Rev 3, August 2008
8 GLOSSARY OF TERMS - ACCEPTANCE CRITERIA (continued)
S = Elastic section modulus.

Sx = Elastic section modulus for major axis bending.
Sy = Elastic section modulus for minor axis bending.
t = Thickness of tubular member or tubular section.
ti = Thickness of a rectangular or tubular component..
VLo = Vertical leg reaction during preloading.
X1,X2 = Beam buckling factors.

Zi = Component plastic modulus.
γ = Load factor.
γ1 = Load factor for dead load vector.
γ2 = Load factor for variable load vector.
γ3 = Load factor for environmental load vector.
γ4 = Load factor for inertial load vector due to dynamic response.
λ,λc = Column slenderness parameter.
λp = Limiting slenderness parameter for compact component.
λr = Limiting slenderness parameter for noncompact component.
η = Exponent for biaxial bending.
φ = Resistance factor.
φa = Resistance factor for axial load.
φb = Resistance factor for bending.
φc = Resistance factor for axial load (compression).
φHfc = Resistance factor for foundation to withstand horizontal loads when QV is
acting.
φp = Resistance factor for foundation during preload.
φt = Resistance factor for axial load (tension).
φVH = Resistance factor for foundation to withstand combined vertical and horizontal
loads.
φVHM = Resistance factor for foundation to withstand combined vertical, horizontal and
moment loads.
φ1 = Resistance factor for dead load moments (MD).
φ2 = Resistance factor for live load moments (ML).
φ3 = Resistance factor for seabed moments (MS).
Rev 3, August 2008
REFERENCES
1 Carter D.J.T. (1982), 'Estimation of Wave Spectra from Wave Height and Period', I.O.S.
Report No. 135.
2 Compiled by P. Walker (1990), The UKOOA Surveying and Positioning Committee

'Technical Notes for the conduct of Mobile Drilling Rig Site Surveys (Geophysical and
Hydrographic)'.
3 Health and Safety Executive, Petroleum Engineering Division (1990), 'Offshore

Installations; guidance on design and construction' (and subsequent amendments).
4 Wheeler J.D. (1969) 'Method for Calculating Forces Produced by Irregular Waves',
Proceedings 1st Offshore Technology Conference, Houston. (OTC 1006).
5 Taylor P.H. (1991), 'Current Blockage - Reduced Forces on Offshore Space-Frame

Structures' Proceedings 23rd Offshore Technology Conference, Houston. (OTC 6519).
6 Det Norske Veritas, Classification Note 31.5, 'Strength Analysis of Main Structures of
Self-Elevating Units', February 1992.
7 Brekke J.N., Murff J.D., Campbell R.B., and Lamb W.C., (1989) 'Calibration of Jackup
Leg Foundation Model Using Full-Scale Structural Measurements', Proceedings 21st
Offshore Technology Conference, Houston. (OTC 6127).
8 Meyerhoff G.G. and Chaplin T.K., 'The Compression and Bearing Capacity of Cohesive
Layers', Br. J. Appl. Phys, No 4, 1953.
9 Brown J.D., and Meyerhoff G.G., 'Experimental study of Bearing Capacity in Layered
Soils', Proc. 7th ICSMFE, Vol 2, 1969.
10 Winterkorn H.F. and Fang H.Y. (1975) 'Foundation Engineering Handbook', Van Nostrand
Reinbhold Company.
11.1 Noble Denton & Associates Limited (1987), 'Foundation fixity of jack-up units, Joint
Industry Study', Volumes I and II.
11.2 Noble Denton & Associates Limited (1988), 'Foundation fixity of jack-up units, Joint
Industry Study, extra work'.
12 Sarpkaya T. and Isaacson M. (1981), 'Mechanics of Wave Forces on Offshore Structures',

Van Nostrand Reinhold Company.
13 Det Norske Veritas, Classification Note 30.2, 'Fatigue Strength Analysis for Mobile
Offshore Units', August 1984.
14 American Institute of Steel Construction, 'Specification for Structural Steel Buildings -

Load and Resistance Factor Design', September 1986.
Rev 3, August 2008
REFERENCES (Continued)
15 Det Norske Veritas, 'Rules for Classification of Fixed Offshore Installations', July 1989,
Part 3, Chapter 1, Section 6B and associated Class Note 30.1, 'Buckling strength analysis of
Mobile Offshore Units', October 1987.
16 American Petroleum Institute, 'Draft Recommended Practice for Planning, Designing and
Constructing Fixed Offshore Platforms - Load and Resistance Factor Design'
(RP 2A-LRFD), First Edition, December 1989.
17 Matlock H. (1970), "Correlations for Design of Laterally Loaded Piles in Soft Clay",
Proceedings Offshore Technology Conference (OTC 1204).
18 Andersen K.H. (1992), "Cyclic effects on Bearing Capacity and Stiffness for a Jack-up
Platform on Clay", NGI Oslo report 913012-1, Rev 1.
19 Dyer A.P., "Plastic Strength Interaction Equations for Jack-Up Chords", MSc Thesis, Dept
of Mechanical Engineering, Univ. of Sheffield, Nov. 1992.
20 Duan L., Chen W.-F., "A Yield Surface Equation for Doubly Symmetrical Sections",
Engineering Structures, Vol 12, pp. 114-119, April 1990.
Rev 3, August 2008
INDEX
Subject Page(s)
ACCEPTANCE CRITERIA 10, 110 - 135

dynamic extreme response 102
foundation assessment 130
holding system loads 134
horizontal deflections 134
hull 134
overturning stability 129
punch-through 134
structural strength check 110
structure condition assessment 135
ADDED MASS 35, 91, 95, 100
AIRGAP 16, 20
AIRY WAVE THEORY 15, 31
AISC-LRFD CODE 110, 112, 113
ASSOCIATED WAVE PERIOD 15
AXIAL AREA
chord 49
leg 54
AXIAL LOAD
at leg/hull connection 50
due to P-Δ 44
AXIAL STRENGTH
compact and noncompact sections 117
slender sections 118
structural member in compression 116
structural member in tension 116
BASE SHEAR TRANSFER FUNCTION 90,94,98,99,102,103
BATHYMETRIC SURVEY 21, 22
BEARING CAPACITY 61-82

bearing capacity check, foundation stability 72, 132
penetration in carbonate sands 65
penetration in clays 64
penetration in layered soils 66-69
penetration in silica sands 65
penetration in silts 65
settlements resulting from exceedence of capacity envelope 73
soil back-flow 63
BENDING MOMENT
bending moment diagrams for leg 55-58
bending moment due to foundation 42
Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
BENDING STRENGTH –121-125

compact sections 122
noncompact sections 122-123
slender sections 123-125
BIAXIAL BENDING EXPONENT –125-128
BOREHOLE INVESTIGATION 24
BREAKING WAVES 19, 31
CENTER OF GRAVITY 13, 14, 114, 129-134
COMPACT SECTIONS
definition 114
nominal axial strength 116
nominal bending strength 121
CONE PENTROMETER TESTING 24
CREST ELEVATION 15, 20
CURRENT 18-19, 33
drag forces 29-30, 33
environmental excitation 93
other considerations 39
load application 53, 89
profile 19, 33
stretching 19
structure interference 33, 34
surface current 18
surge 18, 19
tide 18, 19
velocity 18, 19, 33
DETAILED LEG MODELING

hydrodynamic 33-39
structural 46-52
DETERMINISTIC ANALYSIS 31,42

dynamic wave analysis 93
extreme response determination 89
hydrodynamic modeling 34
wave height for 15
wave theories 31
DIRECTIONALITY 14
directionality function for spreading 17
effects on dynamic response 102
DISPLACEMENT CHECK, FOUNDATION ASSESSMENT 70, 81, 134

Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
DRAG COEFFICIENTS 33-39

equivalent drag coefficient 34-35
gusset drag coefficient 37
split tube chord drag coefficient 38
triangular chord drag coefficient 39
tubular drag coefficient 36
DRAG FORCE 30, 33
DYNAMIC AMPLIFICATION 89,91,93,98-100

dynamic amplification factor 98, 99
DYNAMIC ANALYSIS 89-107

application of dynamic analysis methods 104
closed-form frequency domain analysis 98
damping 90, 98, 101
dynamic amplification 89,91,93,94,98
dynamic amplification factor 98,99
environmental excitation 93
equivalent mass-spring-damper system 95,98
extreme response 89,92,104,105
fixation system 46, 48, 49-51,55-58, 97
frequency domain method 94, 100,101,104,105
inertial loadset 52, 53, 89-91, 98- 101
JONSWAP spectrum 16-17, 93
maximum response 18
most probable maximum 100,101,104,105
natural period 18,90,93,94-99
nonlinear elements 91
Pierson-Moskowitz spectrum 16, 93
random analysis 90,93,94,100-103
regular wave (deterministic) analysis 93
single degree of freedom analogy 98,99
structural system 91
damping 91,98,101
masses 91,95
springs 95
stiffness 91,94
time domain methods 94,101,104,105
EFFECTIVE LENGTH 117

effective length factor 117, 128
ENVIRONMENTAL DATA 14-21

(see also WIND, WAVE and CURRENT)
directionality 14, 17
return period 10, 14
return period for airgap 15, 20
ENVIRONMENTAL EXCITATION 93
EQUIVALENT LEG MODELING

hydrodynamic 33
structural 47, 49,54
Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
EXTREME (MPME) RESPONSE 89,92,104
EXTREME STILL WATER LEVEL 20, 42
FACTORED LOADS 110

for foundation stiffness determination 74
foundation checks 130-134
overturning check 129
FACTORED RESISTANCE 110

foundation checks 130 - 134
overturning check 129
FATIGUE 88, 106, 107

analysis 42, 107
environmental data for 17-18, 19
life requirements 106
sensitive areas 106
FIXATION SYSTEM 46
modeling 48, 49-51
rotational and vertical stiffness 97
shear force and bending moment diagrams 55-58
FIXITY 42-43
degree of fixity 43, 91, 94
foundation capacity with 133
horizontal and vertical stiffness 75, 96
rotational foundation fixity (stiffness) 42-43, 73-75, 76, 95
FOOTPRINTS 22, 79
FOUNDATION ASSESSMENT 130-134

capacity check 132
foundation fixity 133
pinned foundation 132
displacement check 134
horizontal leg reaction 133
preload check 130
sliding resistance 131
vertical leg reaction 125
FOUNDATION ANALYSIS 61-85

bearing capacity 61-81, 130-134
displacement check 70, 81,134
moment fixity 73-81
footprints 22,83
leaning instability 83
other considerations 70
partial spudcan embedment 70
preloading penetration 61-69
preloading check 70-72, 130-131
Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
FOUNDATION ANALYSIS (Continued)

scour 22, 84
seafloor instability 85
shallow gas 22, 85
sliding check 70,72-73, 131
spudcan-pile interaction 85
FREQUENCY DOMAIN ANALYSIS 92, 94, 101-103, 104,105

closed-form frequency domain analysis 100
GEOTECHNICAL ANALYSIS 61-85

leg penetration 61-69
analysis method 61-63
carbonate sands 65
clay 64
layered soils 66-69
silica sands 65
silts 65
spudcan geometries 61, 62
spudcan foundation model 62
GEOTECHNICAL SURVEYS 21-24

bathymetric survey 21, 22
borehole investigation 24
cone penetrometer testing 24
geotechnical investigation 24
seabed surface survey 21-23
shallow seismic survey 22, 23
soil sampling 22
use of geotechnical data 61
GUIDES 46, 49-51, 59
GUSSETS 37, 48
HEIGHT COEFFICIENT FOR WIND LOADING 28
HOLDING SYSTEM LOADS 46, 49, 134
HORIZONTAL DEFLECTIONS 134
HORIZONTAL LEG REACTION, FOUNDATION ASSESSMENT 131-133
HULL
acceptance criteria 134
detailed hull model 49
equivalent hull model 49
functional loads 13
loading 52
HULL/LEG CONNECTION MODELING 46, 47, 49-51

equivalent system 49, 51
fixation system 50, 51
guides 49-51, 59
Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
HULL/LEG CONNECTION MODELING (Continued)

jackcase 51
jacking system 50, 51
modeling considerations 49-51
shock pads 51
HULL MODELING 49
HYDRODYNAMIC COEFFICIENTS 36-39

gussets 37
marine growth 21, 30, 36
non-tubulars 38
other shapes 39
rough tubulars 36
smooth tubulars 36
split tube chord 38
triangular chord 39
tubulars 36
HYDRODYNAMIC LOADS 29-31, 53

deterministic/regular wave analysis 15, 31
drag force 30
fluid-structure interaction 30, 98
inertia force 30-31
Morison's equation 29, 30, 31, 33, 36
slender members 29
stochastic/random wave height/spectra 15, 16-17, 31
wave kinematic extrapolation 31
INERTIA
inertia coefficients 35, 36, 37, 38-39
inertia force (wave) 30
inertial loadset 52, 53, 89, 90,98-101
INTERACTION EQUATIONS FOR MEMBER CHECKS 115, 119, 125
JACKING SYSTEM 46, 49-51,55-58
LARGE DISPLACEMENT ANALYSIS (see NON-LINEAR ANALYSIS)
LEANING INSTABILITY 79
LEG/CAN CONNECTION 43
LEG/HULL CONNECTION (see HULL/LEG CONNECTION)
LEG INCLINATION 43
Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
LEG MODELING
hydrodynamic
added mass 35, 90, 94
buoyancy 13
detailed leg modeling 33-34
drag coefficients 33-39
equivalent leg modeling 33-35
inertia coefficients 35, 36-37, 38-39
member lengths 34
non-structural items 34
shielding 34
solidification 34
spudcan modeling 34
structural
combination leg modeling 49
detailed leg modeling 46-52
equivalent leg modeling 47-49, 54
member lengths 48
single detailed leg model 48
spudcan modeling 52
LEG PENETRATION
analysis method 61-63
carbonate sands 65
clay 64
layered soils 66
silica sands 65
silts 65
LEG RESERVE 21
LOAD AND RESISTANCE FACTOR DESIGN, STRUCTURAL STRENGTH 110-113, 127
LOADS
application to structural model 52-53
combinations 89
current 53, 89
hull (functional) 13, 52
hydrodynamic 29-31, 53
inertial loadset 52, 53, 89, 90,98-101
P-Δ 43-45, 52, 53
wind 27-29, 53
LOCAL BUCKLING 116,121
MARINE GROWTH 21, 30, 36
MASS-SPRING-DAMPER SYSTEM 95,98-100
MAXIMUM HEIGHT, WAVE 14
MEAN WATER LEVEL 20
MINIMUM STILL WATER LEVEL 20

Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
MOMENT
bending moment capacity 121-125
bending moment due to foundation fixity 42
bending moment diagrams for leg 55-58
can moment 129, 133
effective applied moment
members in compression 120
members in tension 119
hull sag moment 52-53
leg/hull connection moment 46, 49-51
lower guide moment due to leg inclination 43
overturning moment 90,100,129
P-Δ moment 43-45
second moment of area (legs) 54
MORISON'S EQUATION 29, 30, 31, 33, 36
MOST PROBABLE MAXIMUM 100,101,104,105
NATURAL PERIOD 90,93,94-99
NONLINEAR MODELING METHODS 44, 77

non-linear elements 87
NON-STRUCTURAL ITEMS, LEG MODELING 34
NON-TUBULARS, HYDRODYNAMIC COEFFICIENTS 38-39
OVERTURNING STABILITY 10, 129

axis 124
moment 44, 45, 89, 129
P-Δ 43-45
geometric stiffness modeling methods 44
linear-elastic displacement amplification 45
loads 52, 53
manual addition of P-Δ moments 45
non-linear modeling methods 44
PERIOD
natural 18,90, 93, 94-99
return 14, 15, 20
wave
associated 15
peak 16-17, 90,99
zero-upcrossing 17
PINIONS46-47,49-51,55-58,134
PLASTIC ANALYSIS 112

plastic moment 121,122
plastic stress distribution 114
Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
PRELOAD 90,110
foundation assessment 130, 131
foundation stability 70-72
leg penetration during 61-69
PUNCH-THROUGH 134
QUASI-STATIC ANALYSIS 44, 87, 90,93,98-100
RACK TEETH
fatigue 100
stiffness due to 49
marine growth on 36
RANDOM WAVE ANALYSIS (see STOCHASTIC ANALYSIS)
REFERENCE LEVEL, WIND 14, 27, 28
REGULAR WAVE ANALYSIS (see DETERMINISTIC ANALYSIS)
RESERVE LEG LENGTH 21
RESISTANCE FACTORS
foundations 130 31
holding system 134
overturning 129
structural members 115
RESPONSE ANALYSIS 89-107
RETURN PERIOD 14, 15, 20
SCOUR 22, 80
SEABED REACTION POINT 42
SEABED SURFACE SURVEY 21
SEAFLOOR INSTABILITY 81
SHALLOW GAS 22, 81
SHALLOW SEISMIC SURVEY 22, 23
SHAPE COEFFICIENTS FOR WIND LOADING 29
SHEAR FORCE DIAGRAMS 55-58
SHIELDING 34
SHOCK PADS 50, 51
SHORTCRESTEDNESS (WAVE SPREADING) 17-18, 98, 102,103

Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
SIGNIFICANT WAVE HEIGHT 14, 15, 17
SINGLE DEGREE OF FREEDOM ANALOGY 98,99
SLENDER SECTIONS
hydrodynamic loads 29
structural considerations 114,118,123-125,127
SLIDING CHECK 70, 72-73, 130-131
SMOOTH VALUES, HYDRODYNAMIC COEFFICIENTS FOR TUBULARS 36
SOIL SAMPLING 23, 24
SOLIDIFICATION 34
SPECTRUM (WAVE) 16-18

JONSWAP 16-17, 92
Pierson-Moskowitz 16-17,92
SPLIT TUBE CHORD, HYDRODYNAMIC COEFFICIENTS 38
SPUDCAN
modeling 34, 52
partial spudcan embedment 70
spudcan foundation model 62
spudcan geometries 61, 62
spudcan-pile interaction 86
STIFFNESS
due to chord rack 49
for natural period estimation 95-98
geometric stiffness modeling methods 44
STOCHASTIC ANALYSIS
dynamic analysis 93,94,100,101,104,105
hydrodynamic modeling 34
kinematic extrapolation 31
wave height for (scaled) 15
wave spectra 16-17
wave theory 31
STORM DIRECTIONS, RANGE OF 42
STRUCTURAL ANALYSIS 42-53

fatigue analysis 18, 42, 107
foundation fixity 42-43, 73-81
general conditions 42
leg inclination 43
load application 52-53
P-Δ effects 43-45
range of storm directions 42
response analysis 88-105
seabed reaction point 42
structural modeling 46-52
Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
STRUCTURAL MEMBERS
definitions 113
structural strength check 110-127
STRUCTURAL MODELING 46-52

combination 3-leg model 47
combination leg model 49
detailed 3-leg model 47
detailed leg model 48
equivalent 3-stick-leg model 47
equivalent leg model 49
fixation systems 46, 51
general considerations 46-47
jacking systems 46, 51
jack-case and bracing 51
leg-hull connection modeling 49-51
model applicability 47
pinions 47, 51
rack tooth stiffness 49
shock pad 51
single detailed leg model 48
spudcan modeling 52
STRUCTURAL STRENGTH CHECK 110-127

AISC code 110-112, 127
axial strength
structural member in compression 116-119
structural member in tension 116
bending strength 121-125
biaxial bending exponent 125, 128
compact sections 114
axial strength 116-119
effective applied moment
members in compression 120
members in tension 119
effective length factors 117, 128
factored loads 110-114
factored resistance 110, 115, 127
interaction equations 115
local buckling 116-119,121-125
limitations 112,113
load and resistance factor design 110,111,127
member joints 127
noncompact sections 122
axial strength 117
bending strength 122
other geometries 127
resistance factors 112,115, 127, 134
Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
STRUCTURAL STRENGTH CHECK (Continued)

slender sections 114
axial strength 118,119
structural components - stiffened and unstiffened 113
structural members 113
structural strength check 110-127
torsional buckling (lateral) 121
yield stress 112,114,117
STRUCTURAL SYSTEM (FOR DYNAMICS) 91-93

damping 91, 102
masses 91,95
stiffness 94, –95-98
STRUCTURE CONDITION ASSESSMENT 10, 135
STRUCTURE INTERFERENCE, CURRENT 33, 34
TEMPERATURE - AIR AND WATER 20

THREE LEG MODEL
combination 3-leg model 47
detailed 3-leg model 47
equivalent 3-stick-leg model 47
TIME DOMAIN ANALYSIS 94,101,104,105
TRIANGULAR CHORD, HYDRODYNAMIC COEFFICIENTS 39
TUBULARS, HYDRODYNAMIC COEFFICIENTS 36
VERTICAL LEG REACTION, FOUNDATION ASSESSMENT 130-133
WATER LEVEL 20
chart datum 20
extreme still water level 20
lowest astronomical tide 20
mean water level 20
minimum still water level 20
WAVES 14-18
Airy wave theory 15, 31
breaking waves 19, 31
crest elevation 15, 20
directionality function 17, 102
extreme wave height 14
freak waves 20
kinematic extrapolation 31
maximum height 14, 15
period
associated 15
peak 16-17
return 14, 15, 20
zero-upcrossing 16-17
Rev 3, August 2008
INDEX (Continued)
Subject Page(s)
WAVES (Continued)
shortcrestedness (spreading) 17, 97, 102, 103
significant height 14
significant height (scaled) 15, 17
spectrum
JONSWAP 16-17, 93
Pierson-Moskowitz 16-17, 93
steepness 17
WEIGHT
center of gravity 13, 14, 114, 129 - 133
minimum elevated weight 13
WIND 14, 27-29

force calculation 27-29
height coefficient 28
load application 53
profile 14, 28, 29
reference level 14, 27-28
shape coefficient 29
velocity 14, 27-28
YIELD STRESS 112,114,117
YIELD SURFACE, FOUNDATION FIXITY 74-75
ZERO-UPCROSSING WAVE PERIOD 16-17

This page intentionally left blank.
COMMENTARIES TO
RECOMMENDED PRACTICE

(REVISION 3 – AUGUST 2008)
Rev Issue Date Details

Rev 1 May 1997 Changes made to pages 11, 27, 30, 80, 86, 128, 155, 171, 176, and 191.
Revised areas indicated by sidelines thus:
Rev 2 Jan 2002 Changes made to pages 112,114, 131, 138, 147, 152, 153, 164, 176
Revised areas indicated by double sidelines thus:
Rev 3 Aug 2008 Changes made to pages 9,115 – 119, 127,131, 164, 180
Revised areas indicated by triple sidelines thus:
Note that page numbers listing above changes in Rev.1 and 2 may no longer be
accurate due to insertion of material in Rev.3.
Commentaries to Recommended Practice for Site Specific Assessment of Page 2
Mobile Jack-Up Units Rev 3, August 2008
PREAMBLE
These Commentaries to the Recommended Practice for Site Specific Assessment of Mobile Jack-
Up Units (PRACTICE) have been written to provide background information, supporting
documentation, and additional or alternative calculation methods as applicable.
The reader should recognize that the information presented herein should only be taken in
conjunction with the PRACTICE and that the cautions and limitations discussed in
Section 1 of the PRACTICE apply.
CONTENTS
C3 COMMENTARIES TO ASSESSMENT INPUT CONDITIONS 12
C3.3 ENVIRONMENTAL CONDITIONS - GENERAL 12
C3.4 WIND 12
C3.5 WAVES 13
C3.5.1 Determining Wave Heights for Regular and Irregular
Wave Analysis
C3.5.1.1 Significant Wave Height for Stochastic
Irregular Waves Analysis.
C3.5.1.2 Wave Height for Regular Wave Analyses
C3.5.3 Alternative formulation for wave spectrum.
C3.5.4 Spreading
C3.7 WATER LEVELS AND AIRGAP 20
Glossary of terms for Section C3 21
References for Section C3 22
C4 COMMENTARIES TO CALCULATION METHODS –

HYDRODYNAMIC AND WIND FORCES 23
C4.1 INTRODUCTION 23
C4.2 WIND FORCE CALCULATIONS 23
C4.3 HYDRODYNAMIC FORCES 24

C4.3.1 General
C4.3.2 Drag forces
C4.3.3 Inertia forces
C4.4 WAVE THEORIES 27

C4.4.1 General
C4.4.2 Regular wave analysis
C4.4.3 Irregular wave analysis
C4.5 CURRENT 31
C4.5.1 General
C4.5.2 Combination with wave particle velocities
C4.5.3 Reduction of current by the actuator disc formula
C4.5.4 Current stretching
C4.6 LEG HYDRODYNAMIC MODEL 32

C4.6.1 General
C4.6.2 Length of members
C4.6.3 Spudcan
C4.6.4 Shielding and Solidification
C4.6.5 Equivalent drag coefficient
C4.6.6 Equivalent Inertia coefficient
CONTENTS (Continued)
C4.7 HYDRODYNAMIC COEFFICIENTS FOR LEG MEMBERS 34

C4.7.1 General
C4.7.2 Hydrodynamic Coefficients for Tubulars
C4.7.2.1 General
C4.7.2.2 Literature survey and recommended values
C4.7.2.3 Marine Growth dependence
C4.7.2.4 Definition of relevant parameters
C4.7.2.5 Dependence on roughness
C4.7.2.6 Keulegan Carpenter number dependence
C4.7.2.7 Reynold's number dependence
C4.7.3 Marine Growth Thickness
C4.7.4 Hydrodynamic Coefficients for Brackets
C4.7.5 Hydrodynamic Coefficients for Chords
C4.7.5.1 Split tube chords
C4.7.5.2 Triangular chords
C4.7.6 Other shapes
C4.8 OTHER CONSIDERATIONS 58
Appendices to Section C4
C4.A Example of Equivalent Model Computations 66
C4.B Comparison cases to assess implications of PRACTICE 70
formulation
C4.C Comparison of test results for chords 75

STRUCTURAL ENGINEERING 78
C5.2 GENERAL 78
C5.3 GLOBAL RESPONSE 79
C5.4 DISCUSSION OF THE LEG-HULL CONNECTION 80
C5.5 DETERMINATION OF PROPERTIES FOR

EQUIVALENT MODELLING OF LEG AND LEG-HULL CONNECTION 82
C5.6 LOAD APPLICATION 84
C5.7 EVALUATION OF FORCES 84
C5.A Derivation of alternative geometric stiffness formulation for P-Δ effects 90

GEOTECHNICAL ENGINEERING 98
C6.2 PREDICTION OF FOOTING PENETRATION DURING PRELOADING 98

C6.2.1 Analysis method for leg penetration prediction
C6.2.2 Penetration analysis for clays
C6.2.3 Penetration analysis for silica sands
C6.2.4 Penetration analysis for carbonate sands
C6.2.5 Penetration in silts
C6.2.6 Penetration analysis for layered soils
C6.2.6.1 Squeezing of clay
C6.2.6.3 Punch-through : Dense sand over soft clay
C6.2.7 Summary
C6.3 FOUNDATION STABILITY ASSESSMENT 111

C6.3.3 & C6.3.4 Bearing capacity for inclined loading
C6.4 OTHER ASPECTS OF JACK-UP UNIT INSTABILITY 121

C6.4.1 Leaning instability
C6.4.2 Footprint considerations
C6.4.3 Scour
C6.4.4 Seafloor instability
C6.4.6 Spudcan-pile interaction

DETERMINATION OF RESPONSES 128
C7.2 QUASI-STATIC EXTREME RESPONSE WITH INERTIAL LOADSET 128
C7.3 CONSIDERATIONS AFFECTING THE DYNAMIC RESPONSE 128
C7.4 SELECTION OF APPROPRIATE EXCITATION PERIOD 130
C7.5 METHODS FOR DIRECT DETERMINATION OF THE

DYNAMIC RESPONSES 131
C7.A Derivation of jack-up stiffness equation 134
C7.B Details of appropriate dynamic analysis methods 145
C7.B.1 Analysis methods
C7.B.1.1 Frequency domain methods
C7.B.1.2 Time domain methods
Appendices to Section C7 (continued)

C7.B.2 Methods for determining the MPM 145
C7.B.2.1 Use of drag-inertia parameter (or equivalent) 147
determined from mean and standard deviation
of a frequency or time-domain analysis.
C7.B.2.2 Fit Weibull distribution to results of a 148
number of time-domain simulations to
determine responses at required probability
level and average the results.
C7.B.2.3 Fit Gumbel distribution to histogram of peak 150
responses from a number of time-domain
simulations to determine responses at
required probability level.
C7.B.2.4 Apply Winterstein's Hermite polynomial method 151
to the results of time domain simulation(s).
C8 COMMENTARIES TO ACCEPTANCE CRITERIA 164
C8.0 BACKGROUND TO PARTIAL LOAD FACTORS 164

C8.0.1 General
C8.0.2 Fundamental Question
C8.0.3 Solution
C8.0.3.1 Probabilistic Description of Input
C8.0.3.2 Limit States
C8.0.3.3 Response Model
C8.0.3.4 Safety Index vs. Safety Factor
C8.0.3.5 Reference or Target Safety Level
C8.0.3.6 Derivation/Calibration of Safety Factors
C8.1 STRUCTURAL STRENGTH CHECK 167

C8.1.1 Introduction
C8.1.2 Definitions
C8.1.3 Factored loads
C8.1.4 Assessment of members - excluding stiffened and high
D/t ratio tubulars
C8.1.4.1 General interaction equations
C8.1.4.2/3 Nominal Axial Strength
C8.1.4.4/5 Effective Applied Moment
C8.1.4.6 Nominal Bending Strength
C8.1.5 Assessment of members - other geometries
C8.3 FOUNDATION ASSESSMENT 186
C8.7 STRUCTURE CONDITION ASSESSMENT 188

C8.7.1 Introduction
C8.7.2 Scope of condition assessment
C8.7.3 Condition monitoring

LIST OF FIGURES
C3.5.1 Comparison of wave crest elevation predicted skewness and 16

observed data at 70m in the North Sea.
C4.3.1 Oscillating drag coefficient vs. motion amplitude to diameter 26
ratio x0/D for given reduced velocities.
C4.4.1 Range of validity of different wave theories. 29
C4.4.2 Surface elevation, and velocity profiles for deterministic 30
regular waves.
C4.4.3 Linearization w.r.t. wave heights. 30
C4.7.1 Comparison between measured and computed forces on a pile up 42
to free surface
C4.7.2 Drag coefficient for rough cylinders at high Reynold's number 44
C4.7.3 Drag coefficient for post critical Reynolds numbers for rough 44
cylinders.
C4.7.4 Effect of roughness on drag coefficient and vortex shedding 45
frequency for post-critical Reynolds numbers.
C4.7.5 Recommended values for the drag coefficient as function of 45
relative roughness.
C4.7.6 Drag coefficient dependence on KC number. 47
C4.7.7 Drag coefficient dependence on KC-number for clean cylinders 47
of the Ocean Test Structures.
C4.7.8 Drag coefficient dependence on KC-number for barnacle covered 48
cylinders of the Ocean Test Structure.
C4.7.9 Recommended drag coefficient dependence on KC for cylinders 48
in waves, at high Reynolds numbers.
C4.7.10 Suggested Reynolds dependence in existing guidance. 49
C4.7.11 Reynolds dependence of drag coefficient in test results. 49
C4.7.12 Reynolds dependence of drag coefficient. 50
C4.7.13 Recommended values for Reynolds dependence for different 50
values of relative roughness, KC>40.
C4.7.14 Definition of directions and dimensions for a split tube 53
chord.
C4.7.15 Drag coefficient at 90° related to the rack width W. 53
C4.7.16 Alternative interpolation formulations fit to data. 54
C4.7.17 Comparison with some current practices for regular wave 54
analysis. W/D = 1.24 and the scaling regular/irregular =0.7,
valid below MWL+2.0m.
C4.7.18 Definition of dimensions and angles for a triangular chord. 56
C4.7.19 Drag coefficients for basic sections in uniform flow. 57
C4.7.20 Comparison between TEES test results and the PRACTICE. 57
C4.A.1 Model of a bay for test purposes 66
C4.A.2 Square bay with triangular chords 68
C4.C.1 Comparison of PRACTICE formulation with model tests, ratio 75
W/D = 1.08
C4.C.2 Comparison of PRACTICE formulation with model tests, ratio 75
W/D = 1.10
C4.C.3 Comparison of PRACICE formulation with model tests, ratio 76
W/D = 1.13
C4.C.4 Comparison of PRACTICE formulation with model tests, rack 76
W/D = 1.18
C4.C.5 Comparison of PRACTICE formulation with model tests, rack 77
W/D = 1.24
LIST OF FIGURES (continued)
C5.1 Responses/reactions from first order analyses 85

C5.2 P-Δ and leg inclination effects 85
C5.3 Contribution of second order effects to first order responses 86
C5.4 Representative leg-hull connection 86
C5.5 Leg-hull connection component combinations 87
C5.6 Guide clearances 87
C5.7 Jacking system backlash 87
C5.8 Types of leg guide arrangement 88
C5.9 Unopposed and opposed pinion arrangements 88
C5.A.1 Analysis model 94
C5.A.2 Load application 94
C6.1 Comparison of bearing capacity analytical procedures for 98
shallow foundations and jack-ups.
C6.2 Stability factors for cylindrical excavations in clay. 98
C6.3 Conical footing bearing capacity - problem definition and 101
notation.
C6.4 Depth of failure zone in sand 107
C6.5 Spudcan bearing capacity analysis - sand over clay - load 109
spread method.
C6.6 Foundation bearing failure modes. 110
C6.7 Vertical/horizontal load envelopes for footings in clay. 112
C6.8 Foundation combined vertical/horizontal loading on sand 112
- comparison of design criteria and observed data.
C6.9 Vertical/horizontal load envelopes for footings in sand. 112
C6.10 Normalised initial shear modulus as a function of
Plasticity Index, Ip, for 11 different clays.
Figure 10.2 from Anderson [55] 116
C6.11 Vertical load-displacement curves for leeward and windward 119
legs.
C7.1 Periods for wave force cancellation and reinforcement as a 133
function of leg spacing.
C7.A.1 Graphical solution of equation (24) 142
C7.B.1 Part 1 - Procedure for determining inertial loadset 153
C7.B.1 Part 2 - Procedure for determining (distributed) inertial 154
loadset
C7.B.2 Time domain procedure for determining mean and standard 155
deviation.
C7.B.3 Frequency domain procedure for determining mean and standard 156
deviation.
C7.B.4 Procedure for estimating the extreme response. 157
C7.B.5 Procedure for determining the mpm-factor of the static 160
response.
C7.B.6 Ratio CR of most probable maximum to standard deviation as a 162
Function of drag-inertia parameter K for N = 1000 peaks.
C7.B.7 Comparison between the normalized spectra Sη(ω), Sφ(ω) and SPM(ω) 162
C8.0.1 Link between safety factor and safety index 166
C8.1.1 Stress-strain curve - ultimate strength much bigger than 171
yield strength.
C8.1.2 Stress-strain curve - yield strength close to ultimate strength. 171
C8.1.3 Stress-strain curves for two component member for which 171
addition of nominal strengths is permissible.
C8.1.4 Stress-strain curves for two component member in which one 172
component fractures before the other is loaded to its nominal
strength.
C8.1.5 Stress-strain curves for components of example member. 172
C8.1.6 Example hybrid chord section. 174
C8.1.7 Fully plastic stress distribution. 174
C8.1.8 Interaction equations/curves for tubular chords with double 177
central racks.
C8.1.9 Interaction equations/curves for split tubular chords with 180
double central racks.
C8.1.10 Interaction equations/curves for tubular chords with offset 182
double racks.
C8.1.11 Interaction equations/curves for triangular chords with 184
single racks.
LIST OF TABLES
C3.5.1 Regular Wave Analysis Normalized Results, CDeDe = 5.13 over the 14
full depth.
C3.5.2 Scaling Factor γd on loads to comply with Airy Wheeler in 15
Irregular Seas.
C4.7.1 Survey of Relevant Literature on CM- and CD-values for 37
Tubulars.
C4.7.2 Recommended Roughness Values for Tubulars. 43
C4.7.3 Recommended Hydrodynamic Coefficients for Tubulars. 43
C4.7.4 Comparison of Drag Coefficients for Simple Sections and Chord 56
CDpr Evaluated From tests.
C4.A.1 Computations of Equivalent model for heading 0° to be used in 66
Site Assessment for z < MWL +2m, chord W/D = 1.13.
C4.A.2 Computations of Equivalent model for heading 0° to be 67
Compared with Model Test Results, Chord W/D = 1.13, Model Scale
1:4.264.
C4.A.3 Computations of Equivalent model for heading 30° to be 67
Compared with Model Test Results, Chord W/D = 1.13, Model Scale
1:4.264.
C4.A.4 Square Bay with Triangular Chords, Equivalent Model to be 68
used in Site Assessment z < MWL + 2m.
C4.A.5 Square Bay with Triangular Chords, Equivalent Model to be 69
used in Comparison with Test Results, Model Scale 1:4.256.
C4.B.1 Comparison including wave height scaling, water depth = 30m, 72
Hsrp = 10m.
C4.B.2 Comparison including wave height scaling, water depth = 90m, 73
Hsrp = 14m.
LIST OF TABLES (continued)
C5.A.1 Verification of Simple Procedure for P-Δ Effect with Exact 93

Solution. Wave Loading Case.
C5.A.2 Verification of Simple Procedure for P-Δ Effect with Exact 93
Solution. Wind Loading Case.
C6.1 Nc' factors as a function of embedment, rate of increase of 103
shear strength with depth and roughness, cone angle 30°.
C7.1 Recommended combinations of the structural system and 132
environmental excitation models for a dynamic analysis.
C8.1.1 Data for tubular chords with double central racks. 179
C8.1.2 Data for split tubular chords with double central racks. 181
C8.1.3 Data for tubular chords with offset double racks. 183
C8.1.4 Data for triangular chords with single racks. 185
C3 COMMENTARIES TO ASSESSMENT INPUT CONDITIONS
C3.3 ENVIRONMENTAL CONDITIONS - GENERAL
The PRACTICE does not permit the use of full joint probability (assessment return
period) environmental data. Nevertheless some account of joint probabilities is permitted
as noted below:
- Seasonally adjusted data may be used if appropriate (Section 3.3.1).

Note: When seasonal data are specified, the data should not be divided into periods
of less than one month and the values so calculated should generally be factored such
that the extreme for the most severe period equals the all-year value for the required
assessment return period.
- Where directional data are available, these may be considered (Section 3.3.1).
Note: When directional data are specified, the data should normally not be divided
into sectors of less than 30° and the directional values so calculated should generally
be factored such that the extreme for the most severe sector equals the omni-
directional value for the required assessment return period and season where
applicable. In certain areas 30° sectors may be inappropriate; caution should be
exercised if an assessment heading falls marginally outside a sector with higher data.
- The downwind (vector) component of the maximum surface flow of the mean spring
tidal current is specified rather than the maximum spring tidal current (Section 3.6.1).
- Site specific information may be used to determine an appropriate combination of

wind driven and surge currents (Section 3.6.1).
C3.4 WIND
The PRACTICE selects the 1 minute sustained wind for determining the wind loadings
on the jack-up. In some instances the wind data will be supplied only for an alternative
averaging period. The conversion to the 1 minute sustained value can not be uniquely
defined as the conversion can be a function of various parameters, including the wind
speed itself. In the absence of site specific data the following formula may be applied
[1], providing that the design storm is of longer duration than the supplied averaging
period (the supplied averaging period may exceed the storm duration in areas of the
world where the extreme winds are due to squalls, thunderstorms, etc.):
t
Vref = Vs[1 - 0.047ln( ref )]
ts
where;
Vref = wind velocity for reference averaging period required by PRACTICE
(1 minute).
Vs = wind velocity for supplied averaging period, ts.
tref = averaging period required by PRACTICE (1 minute).
ts = averaging period for supplied wind velocity.
C3.5 WAVES
C3.5.1 Determining Wave Heights for Regular and Irregular Wave Analysis
The wave heights utilized by the PRACTICE for wave load calculations are related to the
return period significant wave height for a three-hour storm, Hsrp. The PRACTICE
however recognizes that this data may not always be available to the assessor and
therefore provides relationships between Hsrp and Hmax, the individual extreme wave
height for the assessment return period with an annual probability of exceedance of
1/return period. The assessment return period is normally taken as 50 years in which
case Hmax(50) is the wave height with a 2% annual probability of exceedance.
Hsrp and the associated period are normally determined through a direct extrapolation of
measured or hindcast site specific significant wave heights. Hmax may be determined
either from an extrapolation of the distribution of individual wave heights over the
assessment return period or by the application of a multiplication factor to Hsrp.
It is noted that the 'extreme wave height' of a regular wave, Hmpm, determined from a 3-
hour storm segment is the most probable maximum (MPM) wave height, defined as the
distance from the extreme crest to the following trough. Using this definition, the MPM
wave height from the 3-hour storm segment is given by:
Hmpm = 1.68 Hsrp
This relationship is confirmed by the data of [2] for individual storms. However, Hmpm
must not be confused with Hmax and must not be used to determine the value of Hsrp on
which an assessment is based. This is because Hmax includes site specific considerations
of potentially longer durations of storms (including build up and decay) and the
additional probability contributions of other return period storms (i.e. 20, 30, 40, 100-
year, etc., return period storms). Consequently the ratio Hmax/Hsrp is larger than the ratio
Hmpm/Hsrp.
A consequence of the site specific nature of the derivation of Hmax is that there is no
unique relationship between Hmax and Hsrp applicable to all areas of the world. Thus, if a
specified return period maximum wave height is given at a particular location there is no
consistent way to derive Hsrp without knowledge of how the maximum (Hmax) wave
height was derived originally.
Average factors between Hsrp and Hmax have been derived for a North Sea and a Gulf of
Mexico location for a 50-year return period. Without further information, the North Sea
factors can be generalized to any non-tropical revolving storm area and the Gulf of
Mexico factors can be generalized to tropical revolving storm areas. These factors are:
Environmental Conditions Hmax/Hsrp

Tropical revolving storms 1.75
Non-tropical storms 1.86
The Dean's stream function/Stoke's fifth order theories predict higher peak than trough
amplitudes, increasing the maximum velocities and the wetted surface compared with the
Airy theory. In Figure C4.4.2 the difference in the profiles is illustrated. Using the same
specified wave height this difference may be seen in terms of the overturning moment,
base shear or deck displacement.
A number of computations were performed to determine the differences due to wave

kinematics on selected Jack-up designs. Some results are summarized in Tables C3.5.1
and C3.5.2. See also Appendix C4.B.
Table C3.5.1 Regular wave analysis normalized results,

CDeDe = 5.13 over the full water depth
Theory Water Wave Crest Base Overt. Dean's

depth H:T amp. shear moment overturning/
m m:sec m MN MNm other
Airy
Const. 30 15/14 7.5 3.577 91.607 1.74
Airy
Wheeler 15/14 7.5 3.266 82.782 1.93
Stoke's
fifth 15/14 10.22 5.211 156.16 1.02
Dean's
stream 15/14 10.42 5.243 159.45 1.
Airy 15/14 7.5 2.916 160.83 1.12
Const. 70 28/16 14.0 14.121 677.69 1.44
Airy 15/15 7.5 2.563 138.80 1.30
Wheeler 28/16 14.0 13.446 636.53 1.53
Stoke's 15/14 8.41 3.171 180.80 1.00
fifth 28.16 19.17 18.264 976.62 1.00
Dean's 15/14 8.41 3.161 180.30 1.
stream 28/16 19.33 18.136 972.54 1.
In Table C3.5.2 the deterministic analysis is based on application of a various Hmax to
Hs relationships. The Stochastic analysis refers to extreme values determined from
time domain analyses by fitting a three parameter Weibull distribution to the response
peaks and reading the extreme as the 0.999 fractile, approximating a three hour storm
extreme.
The results show dependence on the choice of wave kinematics differing with wave
height.
Table C3.5.2 Scaling factor γd on loads to comply with
Airy Wheeler in irregular seas, [11].
BASE SHEAR: Airy Airy Airy Stokes

Wheeler No stretch Constant fifth
Stochastic
irregular seas 1.00 1.03 0.83 -
Deterministic 1) 0.79 0.84 0.69 0.66
regular waves 2) 0.66 0.69 0.56 0.66
3) 0.71 0.75 0.61 0.66
4) - - - 0.92
OVERTURNING MOMENT:
Stochastic
irregular seas 1.00 1.10 0.79- -
Deterministic 1) 0.81 0.93 0.69 0.66
regular waves 2) 0.67 0.76 0.56 0.66
3) 0.72 0.83 0.61 0.66
4) - - - 0.93
Water depth 110m, Hs = 13.0m, Tp = Tass = 17.0 sec, uniform current V = 0.4 m/s
1) Hmax = 1.86Hs
2) crest as Stokes
3) Hmax = 1.86Hs * 1.07 except Stokes.
4) Hmax = 1.60Hs (PRACTICE recommendation)
Wheeler stretching basis for normalized results, i.e.:
Airy Wheeler stochastic load = γd (other load)
C3.5.1.1 Significant Wave Height for Stochastic Irregular Waves Analysis.
Only Airy theory is currently applicable together with a stochastic irregular seas
analysis, and in Section 4.4 the Wheeler stretching is recommended for describing the
kinematics to the instantaneous surface.
It is accepted that the increasing assymetry described by higher order theories such as
Stokes is appropriate. The asymmetry can also be seen in recorded data as skewness of
the waves, as shown in Figure C3.5.1. Since Airy theory has certain limitations, a
practical way to compensate for the assymetry is to increase the significant wave height
used as input to the force computations. In order to show that a scaling of significant
wave height is appropriate, and to determine the absolute values of the scaling factors,
one needs to decide which theory is correct at a given wave condition. Based on the
good fit to test results in wave tank measurements, [6], the Wheeler stretching is found
to be a best fit. However, due to the asymmetry of wind generated ocean waves in
shallow water, this agreement is judged to be valid only for large water depths. In [7] it
is also indicated that a higher peak than trough is appropriate.
Here it is assumed that the significant wave height should have a scaling factor close to
1.0 for Wheeler stretching at 110m using irregular wave analysis. At shallower water
depths a scaling factor in excess of 1.0 should be due to the wave asymmetry.
In [8] a scaling of wave crests is suggested based on the Stokes wave profiles.
Comparisons are made both with data for North Sea conditions (d = 70 m), see also
Figure C3.5.1, and shallow waters (d ≈ 5.0 m) in the Baltic Sea implying that this may
be a general model. A correction proportional to wave steepness is deduced which
shows fair agreement with the data.
Figure C3.5.1 - Comparison of wave crest elevation predicted skewness

and observed data at 70 m in the North Sea [8]
The crest height correction formula may be simplified neglecting the higher order terms
to be [8]:
ηs/η ≈ 1. + 0.6 α3 + 0.5 (α4 - 3)
where;
ηs = crest elevation by Stokes

η = crest elevation by Airy
α3 = 2.5 D2 Hs / Tp2 , α4 - 3 ≈ (1.6α3)2 : Skewness & kurtosis relations
D2 = coth (kd) [1 + 3/(2sinh2(kd)] : Depth attenuation
k = (2π/T) /g
2
: Wave number
The data and the model indicate that the skewness, α3, is about 0.08 - 0.2 for large
seastates at 70 m water depths giving a correction of 1.05-1.12 on the crest height
compared with a linear model. The forces on a Jack-up structure increase
proportionally as the square (or more) of the elevation. Applying a correction for the
square of the bias in wave crest the correction for 70m should be in the range 1.10-1.25,
depending on wave steepness.
By combining the above suggested formulae a correction for the Wheeler stretching in
a stochastic analysis may be deduced as:
Hs = 1.0 + 1.5 D2 Hsrp / Tp2
The D2 factor includes a dependence on the wave number for individual; waves. This
is not suitable for the purpose of inclusion in the PRACTICE, since there is no unique
wave number for a seastate.
The elevation is not the only parameter to be considered and others; are:
- the depth attenuation over water depth,
- the profiles are not similar in horizontal directions,
- and forces at some distance lose correlation.
This gives a different scaling than that deduced from the wave crest height only.
Based on the above a significant wave height for stochastic/irregular wave analysis
using Airy waves and Wheeler stretching is recommended as:
Hs = [1 + 10 Hs/Tp2 exp(-d/25)]Hsrp
This removes the direct link to the Stokes profile as suggested in [8], but contains the
linear dependence on steepness and a depth dependence with an exponential decay.
Further, by inserting the limited range of wave steepness specified in Section 3.5 the
scaling may be further simplified. Assuming a peak enhancement factor of γ = 3.5, the
peak period may be approximated as Tp/Tz = 1.3, giving a range for
0.046 < Hs/Tp2 < 0.057 for all areas. A ratio Hs/Tp2 = 0.05 is therefore introduced, such
that the significant wave height is recommended as:
Hs = [1 + 0.5 exp(-d/25)] Hsrp
The scaling factor should be limited to a water depth above, say 25m.
A similar scaling on wave height for Airy/Wheeler stretching is currently being applied
indirectly in design specifications, [9], where it is stated that the wave heights
according to Airy should be two times the peak amplitude predicted by the Stokes wave
profile.
The above scaling is an approximation. It would be more correct to account for the
wave asymmetry directly in the generation of the sea surface elevation by, for example,
the methods indicated in [8]. The significant wave height Hsrp could then be applied
directly.
Scaling for other stretching techniques combined with Airy waves may be deduced for
stochastic, irregular waves and based on computational comparisons for different wave
heights and water depths. However, this will not give exactly the same force profile
over the leg and discrepancies in force prediction will occur. Such scaling is therefore
not included in the PRACTICE.
For computational comparisons using this wave height scaling, see also Appendix
C4.B.
C3.5.1.2 Wave Height for Regular Wave Analyses
The selection of wave height to be applied in a particular analysis approach (regular or

irregular waves) is recommended based on matching the loads resulting from the
combination of the wave height and kinematics models, as recommended in Section 4.4.
The scaling of wave heights is introduced as an alternative to the scaling of drag
coefficients, using the wave height relation Hmax = 1.86Hs.
For regular wave analyses the wave asymmetry is properly accounted for, but the
irregularity of the sea surface and the wave spreading may not be modeled properly. As
indicated in Tables C3.5.1 and C3.5.2 a reduction factor is required to give similar forces
as predicted by an irregular seas simulation if Hmax = 1.86Hs. In [3] a reduction of the
drag coefficient by a factor 0.7 is chosen and in [4] a reduction of wave kinematics is
chosen. Classification societies generally specify lower CD values than specified in
Section 4 and these apply to regular wave analyses.
Considering that the computations with regular waves are made with a kinematics model
that has been documented in [5] to be somewhat conservative a reduction factor is
appropriate to arrive at realistic force estimates.
Accepting that a scaling factor on kinematics is applicable, a practical way of

implementing this in the PRACTICE is to reduce the wave height to be used for force
computations in regular wave analyses. This may be more practical than using a factor
on kinematics as most software on the market does not include such a scaling factor.
Equivalent wave heights are suggested as:
Hdet = 1.60Hsrp
The scaling factors on kinematics may be implemented assuming that the load effect is
proportional to wave height to the power 2.2, remembering that CD's should not be
scaled.
As a comparison with previous practices the relationship Hdet ≈ 1.60Hsrp may also be
compared with the reduction of CD by a factor 0.7 as recommended in [3] in combination
with the wave height Hmax = 1.86Hs. By assuming that load effects are proportional to
the ratio of wave heights to the power 2.2, the scaling becomes (1.60/1.86)2.2 = (0.86)2.2 =
0.72, indicating that this is not lower than current practice. The computational results of
Table C3.5.2 indicate also that scaling of 0.66 would give similar static forces as the
irregular seas simulation at large water depths. See also Appendix C4.B for a
comparison of the computational results, related to other practices.
C3.5.2 The wave heights specified in the PRACTICE for use in airgap determination will be
generally applicable. Special consideration may, however, be required in areas subject to
Freak Waves or where the 1.5m clearance will not be adequate to cover the increase in
wave height associated with higher return period waves. It should be noted that certain
regulatory bodies require the use of higher return period waves (e.g. 10,000 years) for the
determination of airgap requirements.
C3.5.3 Alternative formulation for wave spectrum
The following alternative, and rather restrictive, representation of the wave spectrum by
the power density of wave surface elevation Sηη(f) as a function of wave frequency may
be used:
Sηη(f) = αg2(2π)-4(f)-5exp(-1.25/(Tpf)4)γq
where;
α = equilibrium range parameter = 0.036 - 0.0056Tp/ H m 0 2
g = acceleration due to gravity
q = exp(-(Tpf-1)2/2σ2)
σ = spectral peakwidth parameter = 0.07 for Tpf <= 1
= 0.09 for Tpf > 1
Hm0 = estimate of Hs significant wave height (meters)
Tp = spectral peak period (seconds)
f = frequency (Hz)
γ = peak enhancement factor
= exp(1/0.287[1-0.1975αTp4/Hm02])
The above definition yields a Pierson-Moskowitz spectrum when γ = 1 and Tp = 5√(Hs)

with Tp in seconds and Hs in meters.
C3.5.4 Spreading
The PRACTICE provides a formulation which may be used to incorporate the effects of
wave spreading in the analysis. The power constants recommended [10] imply that the
extreme seastate is close to long-crested, and that there is therefore little angular
distribution of wave energy about the mean direction.
It should be noted that where significant spreading exists it may be non-conservative to
assume a long-crested sea.
In [4] a reduction formula is suggested which reduces the velocity by a factor 'primarily
accounting for wave spreading':
ured/u = √[(2n+1)/(2n+2)]
where;
n = the exponent in the cos2nθ spreading function at Tp,
u = the computed velocity for long crested waves,
ured = the reduced horizontal velocity.
For a range of the spreading exponent, 2 < n < 3, the range of the scaling is 0.91 <
ured/u < 0.94. This corresponds to a reduction of the forces by a factor ranging from
0.833-0.875. To use such a spreading factor in reducing overall forces on a structure is
debatable, and especially so for jack-up structures. There may be cases where the
inclusion of the spreading in irregular seas results in higher forces for some headings.
If the leg spacing corresponds to a wave period, inducing opposing wave forces for
different legs coinciding with the first resonance period, the forces will in fact be
amplified when spreading is included. For jack-ups where the resonance period may
often be as high as 4-7 sec., the effect of wave spreading is believed to reduce forces.
However, the size of the reduction is dependent on the structure.
C3.7 WATER LEVELS AND AIRGAP
The PRACTICE references water depths to lowest astronomical tide (LAT). In some
instances the water depth may be referenced to Chart Datum. It is modern practice for
these reference levels in hydrographic surveys to be the same, however caution should
be exercised when using older data or navigation charts and the relation of Chart
Datum to LAT should be checked and any necessary corrections applied.
See also the Section C3.5.2 regarding wave heights for airgap determination.
GLOSSARY OF TERMS FOR SECTION C3
CDeDe = equivalent drag coefficient times effective diameter.

d = water depth.
D2 = depth attenuation.
f = frequency (Hz).
g = acceleration due to gravity.
H = wave height.
Hdet = reduced wave height which may be used in deterministic/regular wave force
calculations.
Hmax = maximum wave height for a given return period; used for airgap calculations.
Hmpm = wave height associated with Hsrp equivalent to the height between the extreme
crest and the following trough.
Hmo = estimate of Hs significant wave height (meters).
Hs = scaled significant wave height to be used in irregular seas simulation (meters).
Hsrp = significant wave height for assessment return period.
k = wave number.
n = the exponent in the Cosnθ spreading function at Tp.
q = exp(-(Tpf-1)2/2σ2)
Snn(f) = power density of wave surface elevation as a function of wave frequency.
tref = wind averaging period required by PRACTICE (1 minute).
ts = wind averaging period for supplied wind velocity.
T = wave period (seconds).
Tp = peak period in wave spectrum (seconds).
Tz = zero-upcrossing period of wave spectrum (seconds).
u = the computed velocity for long crested waves.
ured = the reduced horizontal velocity.
V = current.
Vref = wind velocity for reference averaging period required by PRACTICE (1 minute).
Vs = wind velocity for supplied averaging period, tu.
α = equilibrium range parameter = 0.036 - 0.0056Tp / H m 0 2
α3 = skewness.
α4 = kurtosis.
γ = peak enhancement factor = exp(1/0.287[1-0.1975αTp4/Hm02]), for Tp in seconds
and Hm0 in meters.
γd = scaling of drag forces.
ηs = crest elevation by Stokes.
η = crest elevation by Airy theory.
σ = spectral peakwidth parameter = 0.07 for Tpf <= 1
= 0.09 for Tpf > 1
REFERENCES FOR SECTION C3
1 Det Norske Veritas, Classification Notes No 30.5, 'Environmental Conditions and

Environmental Loads', Høvik, March 1991.
2 Heideman J.C. and Schaudt K.J., 'Recommended Equations for Short-term Statistics of
Wave Heights and Crest Heights', 1 April 1987.
3 'Practice for the Site Specific Assessment of Jack-up Units', By Marine Technology
division, SIPM, EDP-5, The Hague, May 1989.
4 American Petroleum Institute, proposal for an update of the API-RP2A, 'Hydrodynamic

Force Guidelines for U.S. Waters', received 6 February 1992.
5 L. Skjelbreia and J.A. Hendricksen, 'Fifth-order Gravity Wave Theory', Proceedings of

Seventh Conference on Coastal Engineering, 1961, pp. 184-196.
6 J.E. Skjelbreia, G. Berek, Z.K. Bolen, O.T. Gudmestad, J.C. Heideman, R.D. Ohmart, N.
Spidsoe and A. Torum, 'Wave Kinematics in Irregular Waves', OMAE, Stavanger, 1991.
7 Health and Safety Executive, Petroleum Engineering Division, 'Offshore Installations:

Guidance on Design, Construction and Certification', London, 1990.
8 S.R. Winterstein, E.M. Bitner-Gregersen and K. Ronold, 'Statistical and Physical Models
of Nonlinear Random Waves', OMAE, Volume II, Safety and Reliability, Stavanger, 1991,
pp.23-31.
9 O.J. Andersen, E. F`rland and S. Haver, 'Design Basis, Environmental Conditions,

Statfjord', Statoil Report no. F&U-ST 88007, Stavanger, April 25, 1988.
10 S. Haver, 'On the Modelling of Short Crested Sea for Structural Response Calculations',
EurOMS, Trondheim, 20-22 August 1990.
Other project reports and related technical notes:
11 D. Karunakaren, 'Scaling of Hydrodynamic Loads According to Computational Models',

Technical memo no. 710762, SINTEF, Trondheim, July, 1991.

HYDRODYNAMIC AND WIND FORCES
C4.1 INTRODUCTION
The main objective of this Section is to provide documentation of the numbers,

methods and formulations of the Section 4 of the PRACTICE.
This Section is limited to considering Jack-Up specific methods for wind loading on
legs and hulls and hydrodynamic forces acting on the legs under the action of waves
and current. Typical jack-up leg designs consist of legs with an open lattice frame
structure with typical member dimensions of 0.25-1.0m in diameter. A special feature
are the racks fitted to the chord elements for jacking purposes. The fact that jack-ups
are mobile will also limit the marine growth.
The models, methods and coefficients for computing the forces are considered together
in the development of PRACTICE Section 4, and represent a consistent method such
that the whole Section should be considered in its entirety. This means that no
coefficients should be taken from this Section or Section 4 of the PRACTICE unless
the corresponding method is applied.
The Section is organized such that the main sub-sections have the same numbers as the
corresponding section in the PRACTICE. This means that Section C4.2 in this report
corresponds to section 4.2 in the PRACTICE and so on.
C4.2 WIND FORCE CALCULATIONS
The wind force acting on each block of the jack-up is obtained by multiplying the
pressure (which accounts for the elevation and shape of the block - see C4.2.2 and
C4.2.3 respectively) by projected area. The total wind force on the jack-up can then be
obtained by summing the wind forces over all the blocks. Shielding effects are not
normally included in the calculation, except that the wind area of the hull and
associated structures (excluding derrick and legs) may normally be taken as the profile
area viewed from the direction under consideration.
The wind speed varies with height since the boundary layer friction (which in increased
by the roughness of the sea surface) retards the wind near the sea surface. The lower
layers then retard those above them, resulting in increasing velocity above the sea level,
until the retarding forces reduce to zero.
A wind profile is normally used to represent the variation of wind speed with respect to
height. The PRACTICE recommends a power law of 10 (N = 10) to represent the
wind. The wind speed measured at 10m above the mean sea level is normally used as
the reference in defining the wind speed profile. Alternatively, the height coefficients
(Ch) listed in Table 4.1 can be used to determine the wind speed at various heights.
Where a block has a vertical extent of more than 15m, it is recommended that it is sub-
divided and the appropriate height coefficients are applied to each part of the block.
The shape coefficients for various typical components of a jack-up are given in Table
4.2. Items with 'solid' faces are treated as individual blocks. A different approach is
used for open lattice structures, such as derricks, crane booms, helideck support
structure, flare booms and raw water towers, etc. Here Table 4.2 recommends the use
of 50% of the total projected profile area of the item (e.g. 50% of the product of the
derrick width overall and the vertical extent of block under consideration) in
association with the appropriate shape coefficient for the isolated shapes comprising the
lattice.
For leg structures, the equivalent hydrodynamic coefficients on lattice legs may be
taken from Section 4.6. These will generally be the same as those for clean legs in
large velocities and long waves and hence the smooth values are generally
recommended.
C4.3 HYDRODYNAMIC FORCES
C4.3.1 General
Jack-up leg sections are complex structures, usually made of slender members. The
best engineering tool available for computation of hydrodynamic forces is Morison's
equation. However, the limitations of Morison's equation should be recognized. For
single large diameter members/legs, which may be an alternative to lattice legs, more
appropriate theories and formulations for the inertia forces should be applied.
MacCamy and Fuch's [60] corrections on the inertia coefficients of vertical elements
may be an alternative for those structures.
A limitation on the application of Morison's equation to predict wave loads is

implemented. The limitation is set to:
λ > 5Di (4.3.1)
where;
λ = wave length and
Di = reference dimension of individual leg members (within a lattice leg).
The above limitation implies that the members should be small compared with the
waves.
Morison's equation [30] is an empirical relation given by a drag term plus an inertia
force term as:
ΔF = ΔFdrag + ΔFinertia = 0.5 ρ CD D | ux | ux + ρ CM (πD2/4) u x (4.3.2)
where;
CD = the drag coefficient.
CM = the inertia coefficient.
ux, u x = the horizontal water particle velocity and acceleration.
D = the tubular diameter.
ρ = the density of fluid surrounding the tubular.
The above equation was established to be used for vertical circular cylinders in waves,
but has later been modified and generalized to account for current, inclined members
and relative velocity and acceleration. These extensions are further defined for use in
the PRACTICE and discussed in the following sections.
C4.3.2 Drag forces
For the drag part of the equation the extension from Morison's original formula is made
as:
ΔFdrag = 0.5 ρ CD D | vn | vn (4.3.3)
where vn is now introduced as the relative particle velocity normal to the local member
axis including current, taken as:
vn = un + VCn - α r n (4.3.4)
where;
un + VCn = the combined particle velocity from wave and current by vectorial
summation normal to the member considered.
r n = the velocity of the considered member normal to its axis and in the
direction of the combined particle velocity.
α = 0, if an absolute velocity is to be applied, i.e. neglecting the structural
velocity.
= 1, if relative velocity is to be included. May only be used for
stochastic/random wave force analyses if:
Ured = uTn/Di ≥ 20.
where;
u = particle velocity,
Tn = first natural period of surge or sway motion
Di = the reference diameter of a chord.
In the above definition of combined velocity, current is included. This should be
acceptable as the member does not distinguish between the velocity due to current or
wave motions. The backflow of the wake is different in combined wave and current
fields, (KC dependence) but this has a small influence on the prediction of the largest
force in an extreme wave for single members of diameters typical for jack-ups, see
Section C4.7.2.6.
For inclined members the above definition implies that the procedure to arrive at the
force components is first to determine the particle velocity component normal to the
member axis, then determine the force normal to the member axis and thereafter to
determine the force components in the global directions. This implies that the force
component along the member is neglected.
On the inclusion of the relative velocity there has been some reluctance to directly
accept the extension to the original Morison's equation. Intuitively the extension
should be correct using the same argument as for current forces as the member only
experiences the flow field passing locally. However, the displacements of the members
are quite small and there has been few data to support such an extension as pointed out
in [34]. In [55] test results show that for small amplitude motions the damping may be
overpredicted when the relative velocity is included. However, for a typical jack-up,
with member diameters less than 1m and natural periods around 5.0 seconds, the
sensitivity to member displacement is not large because the parameter Ured = uTn/Di ≈
20 or more in an extreme sea state, see Figure C4.3.1. In addition, the Christchurch bay
test results show that the relative velocity formulation gives good prediction of the in-
ine loading [44], 'correctly predicting the important hydrodynamic damping at the
resonant frequency'. From this it may be concluded that the relative velocity
formulation is probably applicable for jack-up structures. A limitation is introduced to
avoid any significant overprediction of damping.
The reduced velocity Ured may be computed for a wave height equal to the significant
wave height and using the first natural period normally corresponding to the fixed
condition soil parameters. In practical cases it is suggested to evaluate Ured for a
majority of members close to the sea surface, and to include relative velocity either for
all or no members.
The relative velocity formulation is in effect similar to the inclusion of damping

reaction forces. All predictions of damping are uncertain, and compared with other
damping estimates the relative velocity formulation is judged to be reasonably well
estimated. This additional damping from the relative velocity formulation should be
considered when choosing the structural/proportional damping coefficient. A low
structural damping should be considered when the relative velocity is included.
A procedure to combine the forces on several individual members into one member
with equivalent diameter and drag coefficient to be used with the horizontal water
particle velocities is discussed in Section 4.6.
Figure C4.3.1 Oscillating drag coefficient vs. motion amplitude to

diameter ratio xo/D for given reduced velocities [55]
C4.3.3 Inertia forces
These forces are not dominant for extreme loads of typical jack-up lattice legs. A more
comprehensive model could be applied to include relative accelerations (noting that in
this case the added mass should not be included in the structural model).
In the RP the formulation is given as:

ΔFinertia = ρ CM (πD2/4) u n (4.3.5)
where;
ΔFinertia = normal force per unit length of member (in this case the member is
vertical and the force horizontal).
ρ = density of fluid surrounding the tubular.
D = diameter of the member.
u n = water particle acceleration normal to the member.
This implicitly defines how to treat inclined members. However, for inclined members
the horizontal force may alternatively be determined by accounting for the inclination
on the added mass part of the inertia force, but not on the Froude-Krylov part of the
force. The horizontal inertia force is hence computed as:
ΔFinertiaH = ρπD2/4 [(CM-1)sin2βi + 1] u n (4.3.6)
where βi is the angle between the particle acceleration and the element orientation as
defined in Figure 4.2. It should be noted that the vertical particle acceleration will also
provide a horizontal component on inclined braces. For global force calculations this
will generally be unimportant as the loadings on different braces at different angles will
tend to cancel out.
C4.4. WAVE THEORIES
C4.4.1 General
In general there are two different computational methods with corresponding suitable
wave theories;
- Deterministic regular wave analysis, and

- Stochastic irregular or random wave analysis.
For the deterministic regular wave analysis all formulated wave theories may be chosen
from a mathematical point of view. For shallow waters however, the choice of wave
theory is limited to those properly predicting wave asymmetry and the corresponding
change in wave kinematics.
For the stochastic irregular wave analysis only the linear Airy wave theory, or
variations of Airy theory are suitable. Airy wave theory does not fully describe the
wave kinematics behavior since this wave theory implies symmetric waves, which are
not always applicable for shallow water. This will limit the application of this type of
analysis to deeper and intermediate water depths and is considered further in Section
3.5.1, see Appendix C4.B.
C4.4.2 Regular wave analysis
Currently there are a number of wave theories that are applied in the analysis of jack-up
platforms. In most cases the deterministic computations are performed using Stokes
fifth order [42] or Dean Stream function [40] theories. The Dean Stream function
theory shows the best fit to test results, [40,41], for shallow water waves. The
difference in overall forces from these two wave theories will however be small at large
to intermediate water depths and for low wave steepnesses.
Figure C4.4.1 is included in the PRACTICE in a linear scale to guide the selection of
the appropriate wave theory for deterministic analyses. Only the Dean stream and
Stokes wave theory are recommended here in order to limit the range of possible
choices, reducing the scatter in wave force predictions.
C4.4.3 Irregular wave analysis
For the irregular waves analysis, Airy's wave theory is the only possible choice using
the principle of sum of independent wave components as implied in standard irregular
seas time domain simulation and frequency domain solutions. For both the Dean
Stream and Stokes wave theories there are implicit phase dependencies between wave
components at different frequencies.
To account for changes in wetted surface a modification of the Airy wave theory is
required, introducing the surface elevation as a parameter in the kinematics. A number
of such stretching methods have been proposed in literature. One simple method, the
Wheeler stretching method [37], compares well with test results in model tank
measurements [38]. Even for the Wheeler stretching method there exist different
variations. The chosen definition is that originally suggested in [37], to substitute the
true elevation at which the kinematics are required with one which is at the same
proportion of the mean water depth. This can be expressed by:
z−ζ
z' = (4.4.1)
1+ ζ / d
where;
z = The elevation at which the kinematics are required. (Coordinate measured
vertically upward from the mean water surface)
z' = modified coordinate to be used in particle velocity formulation
ζ = The instantaneous water level (same axis system as z)
d = the mean, or undisturbed water depth (positive)
This method causes the kinematics at the surface to be evaluated from linear theory
expressions as if they were at the mean water level.
If a frequency domain analysis is to be applied in extreme response predictions, it is
recommended to use linearization with respect to a finite wave height, Hmax defined in
3.5.1, however damping should be linearized using a lower wave height. Stochastic
linearization implies the use of a unit wave height and when combined with the
assumption of Gaussian statistics the extreme response may be underpredicted, see
Figure C4.4.3. For fatigue computations stochastic linearization, [51], is recommended
as fatigue damage is not dominated by the extreme wave heights, however
consideration should be given to local loads arising from the finite wave height.
Notes
1) None of these theories is theoretically correct at the breaking limit. Nomenclature
2) Wave theories intended for limiting height waves should be referenced for
2
waves higher than 0.9Hb when stream function theory may underestimate Hmax/gTass = Dimensionless wave steepness
2
the kinematics. d/gTass = Dimensionless relative depth
3) Stream function theory is satisfactory for wave loading calculations over Hmax = Wave height (crest to trough)
the remaining range of regular waves. However, stream function programs Hb = Breaking wave height
may not produce a solution when applied to near breaking waves or deep d = Mean water depth
water waves. Tass = Wave period
4) The order of stream function theory likely to be satisfactory is circled. Any L = Wave length (distance between
solution obtained should be checked by comparison with the results of a crests)
higher order solution. g = Acceleration due to gravity
5) The error involved in using Airy theory outside its range of applicability is
discussed in the background document.
Figure C4.4.1 : Range of validity of different wave theories [52,58]

Figure C4.4.2 : Surface elevation, and velocity profiles for

deterministic regular waves
Figure C4.4.3 : Linearization w.r.t. wave heights [47]

C4.5. CURRENT
C4.5.1 General
The current specified for a specific site is to be included as specified in section 3.6 of
the PRACTICE. Interpolation between the data points may be required and linear
interpolation is recommended for simplicity.
C4.5.2 Combination with wave particle velocities

It should be emphasized that the wave and current velocities are to be treated together,
as a sum of separate force contributions will significantly underestimate the
hydrodynamic forces.
C4.5.3 Reduction of current by the actuator disc formula

The current velocity will be reduced due to the presence of the structure in the current
flow field. An estimate of the reduction of the steady flow velocity may be found by
[53]:
VC/Vf = [1 + Σ(CDiDi)/4W]-1 ≥ 0.7 (4.5.1)
where; VC = the reduced current velocity to be used in analysis.
Vf = the observed far field current.
CDi = the drag coefficient of an element i.
Di = the element diameter of element i.
W = the width of the structure.
Several limitations of the above relation are discussed in [53,56] and a lower limit to
the reduction of the current velocity is suggested to be 0.7.
The above equation contains a sum of CDi and diameters Di, but is not explicit with
respect to inclined members. The summation ΣCDiDi is similar to the computation of
the equivalent drag coefficient and diameter, CDeDe, in Section 4.6 of the PRACTICE,
where member inclination is accounted for. Since the equation should be considered
for separate groups of elements [53], it suggested to apply the formula for each leg and
use the following format:
VC = Vf [1 + CDeDe/(4D1)]-1 (4.5.2)
where; VC = the current velocity to be used in the hydrodynamic model, VC
should not be taken less than 0.7Vf.
Vf = the far field (undisturbed) current.
CDe = equivalent drag coefficient, as defined in 4.6.6.
De = equivalent diameter, as defined in 4.6.6.
D1 = face width of leg, outside dimensions.
For structures where the hydrodynamic geometry varies significantly with depth, the
blockage factors can be computed for different depths. In view of the reduced drag
above MSL (due to lack of marine growth) it may be appropriate to calculate current
blockage for the stretched part of the current above MSL separately.
C4.5.4 Current stretching

It is suggested to let the profile follow the surface elevation by changing the coordinate
system similarly to that of the Wheeler stretching defined by equation 4.4.1. The
current profile is recommended in Section 3.6.2.
C4.6. LEG HYDRODYNAMIC MODEL
C4.6.1 General
The hydrodynamic modeling of the leg of a jack-up may be carried out by utilizing
either 'detailed' or 'equivalent' techniques. In both cases the geometric orientation of
the elements are accounted for. The hydrodynamic properties are then found as
described below:
'Detailed Model'
All relevant members are modeled with their own unique descriptions for the Morison
term values and with correct orientation to determine vn and u n and the corresponding
drag coefficient times diameter CDD = CDiDi and inertia coefficient times sectional area
CMA = CmiπDi2/4, as defined in Section 4.7.
'Equivalent Model'
The hydrodynamic model of a bay is comprised of one, 'equivalent', vertical tubular to
be located at the geometric center of the actual leg. The corresponding (horizontal) vn
and u n are to be applied with equivalent CDD = CDeDe and CMA = CMeAe, given in
4.6.5 and 4.6.6. The model should be varied with elevation, as necessary, to account
for changes in dimensions, marine growth thickness, etc.
C4.6.2 Length of members
Lengths of members are normally to be taken as node-to-node distance of the members,
in order to account for small non-structural items.
C4.6.3 Spudcan
A criteria for considering the spudcan is suggested such that the effect of the wave and
current forces on the spudcan may normally be neglected at deep water or deep
penetrations. However, there may be special cases with e.g. large spudcans in
combination with high currents that should be considered also outside the suggested
criteria.
C4.6.4 Shielding and Solidification
Shielding is normally neglected for computations of the hydrodynamic model as
presented herein. The shielding is dependent on KC- and Re- numbers. Since it is
difficult to quantify shielding, and shielding in waves is less than in constant flow
[45,62], shielding is neglected. The same criteria are used for solidification as for
shielding such that both effects should be considered if advantage is taken due to
shielding in wave and current loads.
According to [56] shielding is recommended to be neglected for S/D ≤ 4 for an array of
elements, where S is the outer diameter of the array and D is the diameter of individual
elements. This is also considered in [45].
If information on shielding is obtained from experiments, care should be taken to
distinguish between shielding and the effect discussed in C4.5.3. These effects are
different, but could possibly be confused in tests on small models in large tanks.
Solidification is an increase of wave forces due to interference from objects 'side by
side' in the flow field. This is normally not included in the hydrodynamic coefficient
formulation for jack-ups since shielding is also neglected. Jack-up rigs are usually
space frame structures with few parallel elements in close proximity so that this effect
is usually not important.
In [45] solidification effects are quantified for two elements and for a group of
elements. The drag coefficient may increase 100% if two tubulars are placed side by
side, or be reduced for a group of elements, e.g. a conductor array, where shielding is
also present.
The effect is less than 10% in the worst direction and is therefore suggested omitted in
the PRACTICE, when:
As/At < 0.5
where;
As = sum of projected areas for all members in the considered plane
At = the total projected envelope area of the considered plane.
Solidification should be considered if shielding is included.
C4.6.5 Equivalent drag coefficient
In order to comprise the information on drag forces for individual members of a lattice
leg into an equivalent vertical member over the bay length s, a fixed diameter and a
directional dependent drag coefficient is specified. This model accounts for the
geometrical orientation of the individual members. In this model the principle of no
shielding and no blockage is assumed.
The equivalent diameter is recommended such that the inertia coefficient normally will
follow without any further computations. The equivalent drag coefficient, CDe, times
the equivalent diameter, De, is specified. If another reference diameter De is preferred,
the product of CDeDe should in any case be equal to that specified in Section 4.6.5 of
the PRACTICE. The expression for CDei may be simplified for horizontal and vertical
members as follows:
- Vertical member (e.g. a chord) : CDei = CDi (Di/De)
- Horizontal member : CDei = sin3αi CDi (Dili/Des)
C4.6.6 Equivalent Inertia coefficient
The equivalent value of the inertia coefficient, CMe, and the equivalent area, Ae, to be
used in Section 4.3.3, representing the CMA chosen as:
CMe = may normally be taken as equal to 2.0 when using Ae
= 1.0 for flat plates (brackets).
Ae = equivalent area of leg per unit height = (Σ Ai li)/s.
Ai = equivalent area of element = π Di2/4.
Di = reference diameter as defined in Section 4.7.
The reference diameters Di and corresponding area of member Ae, are chosen such that
the use of an inertia coefficient CMe = 2.0 or CMe = 1.0 is consistent with the inertia
forces for chords and brackets respectively. A conservatism is present since the inertia
coefficient for rough tubulars is set to 1.8 and there is no reduction of forces for
inclined members. For normal lattice leg designs the conservatism will not play any
significant role as the drag forces are dominant. The inertia force will also be
dominated by chords due to their larger diameter, such that the conservatism is judged
to be insignificant for extreme wave forces.
If, however, a more accurate model is wanted an alternative is given using the
individual member inertia coefficients, as specified in Section 4.7 of the PRACTICE,
and including the effect of inclined members. The CMe coefficient is then determined
by the summation shown in Section 4.6.6 of the PRACTICE. This model is in closer
agreement with the 'detailed model'. It should be stressed that the coefficients must be
defined together with their reference dimensions Di.
As comments to this formulation the following may be observed:
- for horizontal members with flow along the length axis the inertia coefficient is:
CMei = 1.0
- for a vertical rough tubular the inertia coefficient will be:
CMei = 1.8
- for other vertical members the inertia coefficient will be:
CMei = 2.0
- for other flat plates (brackets) the inertia coefficient will be:
CMei = 1.0
C4.7 HYDRODYNAMIC COEFFICIENTS FOR LEG MEMBERS
C4.7.1 General
The coefficients determined herein are based on tests where the particle velocities and
accelerations are measured simultaneously as the forces, usually in a controlled
environment. This is the logical way to determine the loading coefficients. However,
the important result in engineering is the overall forces predicted by the Morison's
equation over the Jack-up legs. Since some wave theory has to be applied, which does
not perfectly predict the wave particle motions in all cases, additional scaling is
suggested in Section 3.5 of the PRACTICE, see also Appendix C4.B. This is important
to consider when reading this chapter as the stated coefficients may be somewhat larger
than those applied in other recommendations or classification rules.
C4.7.2 Hydrodynamic Coefficients for Tubulars
C4.7.2.1 General;
There exists a wealth of data on hydrodynamic coefficients (drag and inertia

coefficients) for tubulars, mainly from model tests. A number of model tests have been
performed in wind tunnels, others in oscillating water environment or in steady water
flow, while (to our knowledge) only a few model tests have been performed in a wave
environment. In addition a few full scale tests have been reported.
In the following section (Section C4.7.2.1-7) an overview is given of the literature that
has been applied for the purpose of recommending values for the hydrodynamic
coefficients of jack-up platforms.
Before choosing the appropriate hydrodynamic coefficients for tubular parts of jack-up
platforms the following questions have to be answered:
- Are the coefficients to be used for a fatigue analysis or an ultimate strength
analysis?
- Are the tubular parts smooth or rough, and if they are rough what is the
roughness to be applied?
The parameters to be considered in determining the hydrodynamic coefficients are;
UmT
Keulegan-Carpenter number KC =
D
UD
Reynolds number Re =
ν
k
Relative roughness =
D
where;
k = roughness height
D = diameter
Um = maximum orbital particle velocity
T = wave period
U = flow velocity at the depth of the considered element.
ν = kinematic viscosity of water
(ν ≈ 1.4 x 10-6 m2/sec, t = 10°C)
Concerning the first question above, it is important to determine the range of Reynolds
numbers and Keulegan-Carpenter numbers of interest. Both the drag coefficient CD
and the inertia coefficient CM are dependent on the Reynolds number and the
Keulegan-Carpenter number. In the ultimate strength case one is interested in the CD
and CM coefficients in relatively long and steep waves, i.e. wave steepness S = Hs/λ in
the range 1/10-1/15. A typical ultimate strength case may for example be, a tubular
with diameter D = 0.3 m standing in a seastate with average zero-upcrossing period Tz
= 10 secs.
(λ = 156 m) and significant wave height Hs = 13.0 m. The representative water particle
velocity for this wave will be:
Hsπ
UW = = 4.1 m/s.
Tz
Assuming a current velocity UC of about 1.0 m/s, the total water particle velocity will
be U = UW+UC = 5.1 m/s. This results in the following Reynolds number and
Keulegan-Carpenter numbers (close to the water surface):
UD UTz
Re = = 1.1 106, KC = = 170
ν D
This means that in the ultimate strength case we are dealing with high KC-numbers and
post-critical Re-numbers. Sarpkaya (see for example [4]) uses a parameter β = Re/KC
to describe the test environment. In the ultimate strength environment described above,
the value of β is approximately 6500.
A typical fatigue case may for example be the same tubular in a seastate with Tz = 6
secs. (λ = 56 m) and Hs = 5.6 m. In this case the representative water particle velocity
will be UW = Hsπ/Tz = 2.9 m/s. In the fatigue case, current is not part of the water
particle velocity, which is to be applied. This results in the following Reynolds number
and Keulegan Carpenter number (close to the water surface):
Re = 0.62 106, KC = 58.
This means that post-critical Re-numbers and relatively high KC-numbers are also to
be dealt with in the fatigue case. Sarpkaya's β parameter has a value β = 10860 for the
described fatigue case.
It may be concluded that, in general, for jack-up tubulars, the following ranges of Re-
numbers and KC-numbers will be of interest:
- Re-numbers: Ultimate Strength, roughly from Re ≈ 1.0x106 - 4.5 x 106

Fatigue, roughly from Re ≈ 0.5x106 - 1.0 x 106
- KC-numbers: Ultimate Strength, KC > 100
Fatigue, KC ≈ 25 - 60
Since quite a large amount of the literature survey is dealing with papers written by
Sarpkaya, the following range of Sarpkaya's β-parameter may be regarded to be of
interest: β ≈ 6000 - 20000 (depending on the KC-number).
The answer to the second question concerning the roughness of the tubulars will
depend largely on type of paint used and the smoothness of the steel surface, whether
the tubular is new or has been in the water for quite some time (marine growth), or
whether the tubular mainly stays in air, etc.
Smooth cylinders are defined as cylinders having a roughness k/D < 0.0001, while
rough cylinders are assumed to have a roughness k/D > 0.004 (i.e. highly rusted steel
k/D ≈ 0.005-0.01). Marine roughness due to marine growth implies a roughness in the
range k/D ≈ 0.01-0.15.
C4.7.2.2 Literature Survey and Recommended Values
In Table C4.7.1 a survey result is presented of relevant literature with respect to inertia
coefficients (CM) and drag coefficients (CD) for tubulars. Of course, there exists more
relevant literature than that presented in Table C4.7.1, but it should give a reasonably
representative overview.
Table C4.7.1 : Survey of Relevant Literature on CM and CD values for Tubulars
Source Geometric Shape Re-Number KC-number CD CM Comments

Keulegan Smooth Cylinder 0.1-0.3 105 25-50 1.3-1.5 1.3-1.8 Sub-Critical
5
Carpenter 0.1 10 >100 1.0-1.2 2.4-2.6 and Critical
1958 [1] Flow. Low
Re-numbers.
Sarpkaya Smooth Cylinder >0.5 106 20-40 0.6-0.7 1.7-1.8 Post-Crit.
1976 [2] >0.7 106 60-100 0.6-0.7 1.7-1.9 Oscillating
Flow.
Rough Cylinders
Sand k/D = 0.005 >0.5 106 20-40 1.5-1.7 1.2-1.4 Post-Crit.
k/D = 0.01 1.6-1.8 1.2-1.4 Oscillating
Roughened k/D = 0.02 1.7-1.9 1.1-1.3 Flow.
Sand k/D = 0.005 >0.5 106 60-100 1.4-1.6 1.5-1.7 Post-Crit.

k/D = 0.01 1.5-1.6 1.4-1.6 Oscillating
Roughened k/D = 0.02 1.6-1.7 1.4-1.6 Flow.
Hogben Smooth Cylinder >1.0 106 >25 ≈0.6 ≈1.5 Post-Crit.
et al. Flow.
1977 [3]
Rough Cylinders Post-Crit.
Survey k/D = 0.0002 >1.0 106 >25 0.6-0.7 Flow.
Paper k/D = 0.002 >0.5 106 >25 ≈1.0
State of k/D = 0.01 >0.5 106 >25 ≈1.0
the Art k/D = 0.05 >0.1 106 >25 ≈1.25 1977
Sarpkaya Smooth Cylinder >0.1 106 25-40 0.6-0.8 1.5-1.7 Critical -
et al. Super-Crit.
1982 [4] Oscillating
Rough Cylinder >0.1 106 25-40 1.5-1.7 1.0-1.2 Flow.
k/D = 0.01 β = 4000.
Sarpkaya Smooth Cylinder ≈0.1 106 25-40 0.7-0.8 1.5-1.7 Critical -
et al. ≈0.15 106 60 0.6-0.65 1.5-1.6 Super-Crit.
Rough Cylinder ≈0.1 106 25-40 1.4-1.5 1.4-1.6 Flow.
k/D = 0.01 ≈0.15 106 60 1.4-1.5 1.5-1.6 β = 2500.
Sarpkaya Rough Cylinder 0.1-0.2 106 25-40 1.4-1.5 1.0-1.3 Critical -
et al. k/D = 0.01 Super-Crit.
1985 [6] ≈0.21 106 50 1.4-1.5 1.2-1.3 Oscillating
Flow.
β = 4200.

According to Sarpkaya, available data with current + oscillatory flow substantiate the fact that
drag coefficients obtained from tests at sea in general will be smaller than those obtained under
laboratory conditions.
Sarpkaya Smooth Cylinder 0.2-0.3 106 20-25 0.6-0.7 1.6-1.8 Super-Crit.
Flow.
β = 11240.
6
Sarpkaya Smooth Cylinder >0.5 10 25-40 0.6-0.8 1.5-1.8 Post-Crit.
1985 [8], >0.5 106 >50 0.6-0.7 1.6-1.8 Oscillating
1986 [9] Flow.
Survey Rough Cylinder >0.5 106 25-40 1.4-1.8 1.2-1.4

Articles k/D = 0.02 >0.5 106 >50 1.4-1.6 1.3-1.5
Nath Smooth Cylinder ≈0.5 106 ∞ 0.4-0.5 Super-Crit.
1982 [10] Steady Flow.
Rough Cylinder
k/D = 0.02 ≈0.5 106 ∞ 0.9-1.0 Post-Crit.
k/D>0.1 ≈0.5 106 ∞ 1.0-1.2 Steady Flow.
Smooth Cylinder 0.15-0.2 106 15-25 0.3-0.6 0.8-1.4 Super-Crit.

Oscillating
Rough Cylinder Flow.
k/D = 0.02 0.15-0.2 106 15-25 0.6-1.0 0.4-1.0
k/D>0.1 0.15-0.2 106 15-25 1.0-2.0 0.8-2.3
There is very large scatter in the data presented by Nath.

Bearman Smooth Cylinder 0.15-0.5 106 ≈20 0.6-0.7 1.4-1.5 Super/Post-
et al. Crit. Flow,
1985 [11] Regular
Waves.
The authors present results for random waves as well, but it is difficult to draw any conclusion
from these results.
Kasahara Smooth Cylinder 0.5-1.0 106 20-40 0.5-0.6 1.6-1.8 Post-Crit.
et al. Oscillating
1987 [12] Rough Cylinder Flow.
6
k/D = 0.0083 0.5-1.0 10 20-40 1.1-1.4 1.3-1.7 Large Scatter
≈50 1.1-1.2 1.6-2.3 in CM-values.
k/D = 0.0042 0.5-1.0 106 20-40 0.9-1.3 1.3-2.1
≈50 0.9-1.1 1.6-2.1
Chaplin Smooth Cylinder ≈0.2 10 6
≈20 0.6-0.7 1.4-1.5 Super/Post-
1988 [13] Crit. Oscilla-
ting Flow.

Davies Smooth Cylinder >0.5 106 ≈20 0.6-0.7 1.5-1.6 Post-Crit.
et al. Flow, Reg.
1990 [14] Waves.
Smooth Cylinder >0.5 106 ≈18 0.5-0.7 1.5-1.7

Post-Crit.
Flow. Ran-
dom Waves.
The authors conclude that for smooth cylinders and KC>4, drag and inertia coefficients in
periodic waves may be used to represent average CD- and CM- values in random waves.
Roden- Smooth Cylinder >1.0 106 ∞ ≈0.6 Post-Crit.
busch Rough Cylinder Steady Flow
et al. k/D = 0.02 >1.0 106 ∞ 0.9-1.1 (Steady Tow).
1983 [15]
Smooth Cylinder >1.0 106 >30 0.6-0.7 1.6-1.7 Post-Crit.
Rough Cylinder Oscilla-
k/D = 0.02 >1.0 106 >30 1.4-1.5 1.1-1.3 ting Flow
(Forced
Motion).
Smooth Cylinder >1.0 106 20-40 0.6-0.8 1.4-2.0 Post-Crit.

Rough Cylinder Random Flow
k/D = 0.02 >1.0 106 20-40 1.0-1.8 1.0-1.9 (Forced
Motion).
Smooth Cylinder >1.0 106 60-90 0.6-0.8 1.5-1.7 Post-Crit.

Rough Cylinder Random Flow
k/D = 0.02 >1.0 106 60-90 1.1-1.4 1.0-1.4 (Forced
Motion).
The random tests show relatively large spread especially for the lower KC-numbers (20-40).
Roden- Smooth Cylinder >1.0 106 >60 0.65-0.75 1.5-1.7 Post-Crit.
busch Random Flow
et al. Rough Cylinder (Forced
6
1986 [16] k/D>0.0005 >0.5 10 >60 1.1-1.3 1.1-1.5 Motion).
Theopha- Rough Cylinder Post-Crit.
natos k/D = 0.005 >0.8 106 0.95-1.05 Steady Flow
et al. k/D = 0.0095 1.0-1.1 (Steady Tow).
1989 [17] k/D = 0.025 1.15-1.25 => Sand Rough.
-----------------------------------------------------------------------
k/D = 0.049 1.15-1.25
k/D = 0.098 1.3-1.4 => Pyramids
-----------------------------------------------------------------------
k/D = 0.067 1.2-1.3 => Mussels
Klopman "Rough" Cylinder Post-Crit.
et al. k/D = 0.00012 ≈0.5 106 ≈15 0.6-0.9 1.3-1.6 Random Waves.
1990 [18]
6
Heideman Smooth Cylinder >0.2 10 15-30 0.5-1.2 Post-Crit.
et al. >0.2 106 >30 0.6-0.8 1.2-1.9 Random Waves.
1979 [19] Rough cylinder Ocean Test
6
k/D≈0.03-0.05 >0.2 10 15-30 0.9-1.8 Structure.
>0.2 106 >30 0.8-1.3 0.9-1.7
Large spread for lower KC-numbers (< 30). Authors state that for large KC-numbers (> 30),
smooth cylinder CD approaches an asymptote CD = 0.68, while rough cylinder CD approaches an
asymptote CD = 1.0. Tests performed in an ocean environment.
Nath Rough Cylinder Barnacles: Post-Crit.
1988 [20] k/D = 0.073 ∞ 0.95 Steady Flow.
k/D = 0.104 ∞ 0.98-1.2
Artificial Hard Fouling: Post-Crit.

k/D = 0.078 ∞ 0.98-1.2 Steady Flow.
Wolfram Rough Cylinder Mussels: Post-Crit.
& Theo- k/D = 0.075 ∞ 1.22 Steady Flow.
phanatos k/D = 0.085 ∞ 1.26
1990
[21] Mixed Hard Fouling:
k/D = 0.076 ∞ 1.11
Kelp:
k/D = 1.25 ∞ 1.51
k/D = 2.5 ∞ 1.69
Sea Anemones:
k/D = 0.16 ∞ 1.35
Roshko Smooth Cylinder >3.5 106 ∞ 0.65-0.75 Post-Crit.
1961 [22] Steady Flow
Wind Tunnel.
Miller Smooth Cylinder >3.0 106 ∞ 0.60-0.65 Post-Crit.
1976 [23] Steady Flow
Rough Cylinder Wind Tunnel.
k/D = 0.0004 >3.0 106 ∞ ≈0.80
Sand k/D = 0.0009 0.8-0.9
Roughened k/D = 0.0014 0.8-0.9
k/D = 0.0021 0.9-1.0
k/D = 0.0031 1.0-1.1
k/D = 0.0050 1.0-1.1
Pearl k/D = 0.015 >0.5 106 ∞ ≈1.1 Post-Crit.
Barley k/D = 0.023 ≈1.1 Steady Flow
k/D = 0.044 1.1-1.2 Wind Tunnel
Dried k/D = 0.042 >0.5 106 ∞ 1.1-1.2
Peas k/D = 0.063 1.2-1.4
Pearcey Smooth Cylinder >3.5 106 ∞ ≈0.6 Post-Crit.
et al. Steady Flow
1982 [24] Rough Cylinder Wind Tunnel
k/D = 0.0004 >2.0 106 ∞ ≈0.8
k/D = 0.0014 ≈0.88
k/D = 0.0028 ≈0.92
In addition to the literature review presented in Table C4.7.1, an interesting and useful
overview of existing literature is presented in a survey report prepared by Advanced
Mechanics Engineering Limited for the Health and Safety Executive [25].
The literature review presented in Table C4.7.1 shows that the test results at different
facilities agree reasonably well with respect to the drag coefficients for smooth
cylinders in post-critical flow. The majority of tests show CD values between 0.6 and
0.7, both for the lower KC range for fatigue (25-60) and the higher KC range for
ultimate strength. The suggested CD value for smooth tubular elements (k/D < 0.0001)
in post-critical flow is therefore chosen to be CD = 0.65.
For rough cylinders the spread between the individual tests with respect to CD values is
considerably larger. Especially Sarpkaya [2] operates with very high post-critical CD
values for rough cylinders. It should be noted that none of the values obtained by the
other authors referenced in Table C4.7.1 support the Sarpkaya values in the post-
critical region. The differences between individual tests may partly be due to the
different types of post-critical flow (different test conditions) and to the non-uniform
definition of roughness used by the different authors.
One should also bear in mind that the wave particle velocities decrease with increasing
depth below the water surface, which might mean a transition from the post-critical
regime to the super-critical or even critical regime. This will result in a reduction in CD
values for smooth cylinders (although in the lower Re-number part of the critical
regime it may result in an increase in CD values, but here the water particle velocities
are so low that the resulting contribution to the overall drag force will be significantly
smaller than the contributions higher up on the cylinder). For the rough cylinders the
critical regime occurs at lower Re-numbers and there is no reduction in the drag
coefficient in the super-critical regime. For large roughnesses an increase in the drag
coefficient has in fact been reported in this regime [3, 32].
Based on the literature survey presented in Table C4.7.1 and the discussion above, the
drag coefficient for rough cylinders (roughness k/D>0.004) is chosen equal to CD = 1.0,
both for the ultimate strength and the fatigue cases.
C4.7.2.3 Marine Growth dependence
Rust and hard marine growth has been found to behave in essentially the same manner
as artificial hard roughness, but a surface with hard marine growth behaves quite
differently from a surface with soft marine growth. Another point of consideration is
that different types of marine growth on a submerged tubular may dominate at different
depths below the sea surface.
The use of anti-fouling coating will at least delay the development of marine growth
but after a few years the anti-fouling coating becomes less effective. Regularly
cleaning of the tubulars is another possible way to limit the development of marine
growth. In Table 4.3, Section 4.7.2 of the PRACTICE, it is assumed that severe marine
growth is not allowed. This is in accordance with the operational profile of mobile
jack-up rigs, with cleaning of legs at intervals preventing severe marine growth.
The main contribution to forces is in the surface region, such that the extension of the
marine growth below the surface zone is not important for the overall forces. The paint
will in addition be somewhat roughened when exposed to the salt water for a longer
period. Above the marine growth region the use of values for a smooth cylinder has
been recommended. This is mainly based on the fact that the marine growth will be
limited to the region below MWL + 2m, limiting the roughness above this region.
In addition measurements also indicate that the wave forces in ocean waves are less
than predicted by use of a constant CD [39, 31], see also Figure C4.7.1
Based on this it is recommended that the value CD for a smooth surface (CD = 0.65) is
generally used for the legs above MWL + 2m and the value for a rough surface below
MWL + 2m (CD = 1.0), as stated in Table 4.4 of the PRACTICE.
Figure C4.7.1 Comparison between measured and computed forces on a

pile up to free surface [39, 31]
C4.7.2.4 Definition of relevant parameters
The drag coefficient (CD) for tubulars, may be considered as a function of roughness
(k/D), Keulegan-Carpenter number (KC) and Reynolds number (Re) as an alternative to
Table 4.3. This explicit dependence is intended to be used in cases where there is more
detailed knowledge, first of all on the roughness and in addition on the flow conditions
around the members at a specific site. A definition of these governing parameters are
included in section C4.7.2.1.
C4.7.2.5 Dependence on roughness
The roughness may be accounted for explicitly if the roughness is documented to be of

an intermediate value compared with the smooth and rough k/D values assumed above.
Recommended values for the roughness, k, may be found from table C4.7.2.
Table C4.7.2 Recommended Roughness Values for Tubulars [45]
Surface k (meters)
Steel, new uncoated 5.0E-5
Steel, painted 5.0E-6
Steel, highly rusted 3.0E-3
Marine growth 5.0E-3 - 5.0E-2
Several authors have presented, in graphical form, the CD dependence on the relative
roughness k/D at post-critical Re-numbers. Figure C4.7.2 presents a graph from Miller
[23], showing the variation of CD with varying k/D based on several model experiments
at post-critical Re numbers. Figure C4.7.3 and Figure C4.7.4 show similar graphs
presented by respectively Wolfram et al [21] and Pearcey et al [28]. Based on the
available data with respect to the dependence of CD on k/d, the expressions presented in
Equation (4.7.1) have been proposed to describe this dependence for the purpose of the
PRACTICE. The drag coefficient CDi may then be obtained from Equation (4.7.1):
⎧C Dsmooth = 0.65 ; k / D < 0.0001
⎪
CDi(k/D) = ⎨C Dsmooth *( 2.36 + 0.34 Log10 ( k / D )) ; 0.0001 < k / D < 0.004
⎪⎩C Drough = 10
. ; 0.004 < k / D
(4.7.1)
A graphic representation of Equation (4.7.1) is shown in Figure C4.7.5. With respect to
the inertia coefficients for smooth cylinders, all the references from Table C4.7.1 report
post-critical CM values lower than the asymptote CM = 2.0. The CM values lie mainly in
the range 1.6 - 1.7. However the question is whether (in general) some inertia
contribution has been included in the drag forces used for the CD determination. This
would mean that the CD values are slightly overestimated and the CM values slightly
underestimated. At the same time, since both fatigue and ultimate strength imply
Keulegan-Carpenter numbers >25, it is the drag dominated region which is of most
interest and the chosen CM values are not really critical. Based on this argument the
inertia coefficient for smooth cylinders in the post-critical regime is set equal to the
asymptotic value CM = 2.0.
The CM values for rough cylinders, are in general reported to be slightly lower than the
CM values for smooth cylinders. Based on the same argument as used for the smooth
cylinders, the inertia coefficient for rough cylinders in the post-critical regime is set
equal to CM = 1.8.
A summary of the recommended values for the hydrodynamic coefficients for tubulars
is given in Table C4.7.3.
Table C4.7.3 : Recommended Hydrodynamic Coefficients for Tubulars
Tubular CDi CMi

Smooth (k/D<0.0001) 0.65 2.0
Rough (k/D>0.004) 1.0 1.8
Intermediate k/D Equation 4.7.1 2.0
Figure C4.7.2 Drag coefficient for rough cylinders at high Reynold's number, [23]
Figure C4.7.3 Drag coefficient for post critical Reynolds numbers

for rough cylinders, [21]
Figure C4.7.4 Effect of roughness on drag coefficient and vortex shedding

frequency for post-critical Reynolds numbers, [28]
Figure C4.7.5 Recommended values for the drag coefficient as function of

relative roughness
Keulegan-Carpenter number dependence.
In post-critical conditions, for KC-numbers lower than say 30-40, there seems to be
some dependence of the drag coefficient on the KC-number, at least for rough
cylinders. For smooth cylinders this KC-dependence is more uncertain. The
Christchurch Bay Tower (CBT) results for a clean cylinder reported by Bishop [26], for
example, show this dependence for smooth cylinders, and so do the results reported
from the Ocean Test Structure (OTS) [19]. Wolfram and Theophanatos [21], and the
SSPA results reported by Rodenbusch and Gutierrez [15], do not show this dependence
for smooth cylinders.
For rough cylinders in post-critical conditions, the KC-dependence of the drag

coefficient for KC-numbers lower than say 30-40, seems to be a more generally
observed trend, as in [15, 19, 27] amongst others.
It must be emphasized that for decreasing KC-numbers (<30) the (post-critical)

conditions will gradually be more inertia dominated and less drag dominated, implying
an increasing uncertainty in the reported CD- values.
Figure C4.7.6 shows CD as a function of the KC-number for cylinders in waves from
[52]. Figures C4.7.7 and C4.7.8 show in a similar way CD as a function of KC-number
for respectively a clean cylinder and a rough barnacle covered cylinder of the Ocean
Test Structure (OTS [19]) as presented in [25].
Based on the discussion above and the results reported in the literature the explicit KC-
dependence presented in Equation (4.7.2) may be included in the computations in
addition to the roughness dependence:
⎧⎪145
. ; KC < 10
CDi(KC,k/D) = CDi(k/D) * ⎨2 / ( KC − 5) ;
0 .2
10 < KC < 37 (4.7.2)
⎪⎩10
. ; 37 < KC
A graphic representation of Equation (4.7.2) is given in Figure C4.7.9.
Equation (4.7.2) should also be used for smooth cylinders, in spite of the uncertainty
with respect to the KC-dependence. However, for low KC- values the choice of CD-
value is less critical due to the transition to inertia dominated conditions. Furthermore
using Equation (4.7.2) for the entire roughness range (from smooth to rough) results in
a more uniform and easier to use way to handle the KC-dependence.
An inertia coefficient CM = 2.0 for smooth cylinders and CM = 1.8 for rough cylinders
is suggested for use if KC-dependence is used for the drag coefficients at low KC-
numbers.
Figure C4.7.6 Drag coefficient dependence on KC number, [52]
Figure C4.7.7 Drag coefficient dependence on KC-number for clean

cylinders of the Ocean Test Structures, [19, 25]
Figure C4.7.8 Drag coefficient dependence on KC-number for barnacle covered

cylinders of the Ocean Test Structure, [19, 25]
Figure C4.7.9 Recommended drag coefficient dependence on KC for

cylinders in waves, at high Reynolds numbers
Reynold's number dependence
As previously discussed the drag coefficient is dependent on the Re-number and this
has been reported by several authors [3, 32] (Figures C4.7.11 and C4.7.12) and is
reflected in some guidance on load computations, e.g. [29, 33] (Figure C4.7.10).
However, the use of test results reducing CD in the critical region is not relevant for
practical purposes as the roughness k/D<1/100000 implied in the curve for smooth
cylinders is not applicable for jack-up structures. The change in the Reynolds
dependence with respect to roughness is quite large and it is therefore not possible to
recommend one single curve for this dependence. The recommended set of curves
shown in Figure C4.7.13 are mainly based on a functional fit to the test results
presented in [32] and in addition the drag coefficient in the critical regime is set to
minimum of CD = 0.45. Test results have indicated lower CD values, but only in the
ideal conditions of test facilities. A recommended set of curves are given in Figure
C4.7.13, complying with the roughness dependence of Figure C4.7.5 at large Reynolds
numbers. Using the curve for roughness k/D = 0.01 there is no reduction below CD =
1.0 for Reynolds numbers above 105, which supports the use of a constant CD in the
PRACTICE.
Figure C4.7.10 Suggested Reynolds dependence in existing guidance, [33]
Figure C4.7.11 Reynolds dependence of drag coefficient in test results, [3]

Figure C4.7.12 Reynolds dependence of drag coefficient, [32]
Figure C4.7.13 Recommended values for Reynolds dependence for different

values of relative roughness, KC>40
C4.7.3 Marine Growth Thickness
In addition to the effect on the roughness the effective diameter should be increased to
account for marine growth. Here it is recommended to increase the radius by 12.5 mm
(i.e. diameter increased by 25mm) over the full water depth for tubulars.
C4.7.4 Hydrodynamic Coefficients for Brackets

These are treated as flat plates implying that a CDi = 2.0 and CMi = 1.0 is to be applied
(considering the area associated with CM as that of the circumscribed circle). For
convenience the PRACTICE proposes that brackets are included as horizontal members
in the hydrodynamic model to assure proper directional dependence of forces.
Shielding may be considered according to Figure 4.3 of the PRACTICE.
C4.7.5 Hydrodynamic Coefficients for Chords
C4.7.5.1 Split tube chords

In considering the test results for split tube chord data the following aspects are
considered:
- The tests are almost always performed with a smooth cylinder section.
- Most of the tests are performed in stable current or wind conditions.
The first aspect is the most important, also indicated by the test results evaluation of the
tubulars presented in Section C4.7.2. The drag coefficient for smooth tubulars is about
CDsmooth = 0.65, while for rough cylinder this increases to CDrough = 1.0.
The second aspect on stable flow conditions should not affect the results very much as
the KC values are very high in extreme load evaluations. However, the test result in
[63] showed higher values in waves than in stable current. This has not yet been fully
explained, but comparisons made in [25] indicates that test results in waves from flume
experiments overpredict the drag forces. A number of test results [62] have been
considered, from wind tunnel tests and towing in water, when evaluating hydrodynamic
coefficients for split tube chords.
The drag coefficient is first estimated for the directions 0° and 90° as defined in Figure
C4.7.14. The drag coefficients for 0° are dominated by the tubular part and no
particular effect of the rack on the drag coefficient is seen from the tests. That is, for
typical dimensions of the tubular diameter and rackplate thickness t, Di/t >> 1.0, tests
show values of about CD ≈ 0.65. This indicates that the drag coefficients chosen for the
tubular are also valid for the split tube chord for the 0° direction. In order to be
consistent with the roughness dependence of the drag coefficient for tubulars, the drag
coefficient in the marine growth region is increased due to roughness to CDrough = 1.0
for θ = 0°.
For the 90° direction the drag coefficient should be similar to that of a flat plate for
large W/Di ratios, CDplate = 2.0. However, test results seem to indicate that the CD
values for this direction referring to the mean rack width W, are, on average, about 1.8,
see Figure C4.7.15. The suggested drag coefficient in the PRACTICE is therefore set
to be 1.8 for small W/Di ratios, increasing to 2.0 for large W/Di ratios. The
interpolation between these two numbers is based on engineering judgment.
The drag coefficient for the wave flow normal to the rack, related to the rack width W,
is recommended as:
⎧⎪18. W / D i < 12
.
CD1 = ⎨14. + W / 3D i 12
. < W / D i < 18
. (4.7.3)
⎪⎩2.0 . < W / D i < 2.0
18
For the interpolation between the directions 0° and 90° a number of formulations are
available, but since there were a number of test results available, a best fit of a new
formulation was decided.
The following interpolation formula were found to fit the data best, see Appendix C4.C,
and at the same time be flexible with respect to the drag coefficient for rough and
smooth surfaces at 0°:
⎧C D 0 θ < 20°
CDi = ⎨C + ( C W / D − C )sin 2 [(θ − 20° )9 / 7] 20° < θ < 90° (4.7.4)
⎩ D0 D1 i D0
where;
CD0 = is the drag coefficient for the chord at θ = 0° and is to be taken as that
of a tubular with appropriate roughness, see Section C4.7.2 i.e. CD0 =
0.65 above MWL + 2.0m and CD0 = 1.0, below MWL+2.0m. Possible
dependence on KC and Re numbers as for a tubular.
CD1 = The drag coefficient for flow normal to the rack (θ = 90°), related to
the projected diameter (the rack width W).
Explicit dependence on KC may be taken according to Figure C4.7.9,

but is normally not relevant for extreme loading conditions.
Dependence on k/D or Re may normally be neglected.
The above formulation was derived based on the assumption that the chord behaves
like a tubular up to a direction where the rack enters the flow field, and from there and
up to 90° the chord acts as a flat plate.
In addition to the above formulation two other formulations were tested as shown in
Figure C4.7.16. Equation 4.7.4 gave an excellent fit with the observed drag
coefficients for a smooth tubular and is therefore recommended to be used for split tube
chords.
Interpolation formula similar to those used in [29] and [46] are compared in Figure
C4.7.17 with Equation 4.7.4, for a regular wave analysis. There is some difference in
the direction close to 90°, but the number of test results behind the PRACTICE
formulation is believed to justify the change.
Figure C4.7.14 Definition of directions and dimensions for a split tube chord
Figure C4.7.15 Drag coefficient at 90° related to the rack width W

Figure C4.7.16 Alternative interpolation formulations fit to data
Figure C4.7.17 Comparison with some current practices for regular wave analysis [29], [46].
W/D = 1.24 and the scaling regular/irregular = 0.7, valid below MWL + 2.0m
For the inertia coefficient the theory [32] indicates CM = 2.0 for a smooth tubular
related to the projected diameter and CM = 1.0 for an isolated flat plate. However,
when the plate is considered in conjunction with the split tube sections CM = 2.0 is
appropriate for the combined section. For a rough tubular as indicated in Table 4.4 of
the PRACTICE, the inertia coefficient should be about 1.8 in the marine growth region.
However, since the inertia forces will not contribute much to the extreme forces on the
legs, the inertia coefficient was set to 2.0 related to the width of the chord measured
over the tubular, i.e. at 0°. This is a simple solution and will be conservative in the
direction of the tubular for a rough surface and unconservative in the direction of the
rack and on average correct. This formulation will also be consistent with the
simplified modeling of the leg section where the reference diameter Di is the dimension
D and using CMe = 2.0.
For large rack to diameter ratios W/D, it may however be considered appropriate to
modify (reduce) the inertia coefficient such that it accounts more correctly for the
combination of the contributions from the flat plate and tubular components.
C4.7.5.2 Triangular chords
For triangular chords (Figure C7.4.18) little test data are available. Some currently
applied formulae for drag coefficients of more basic sections were therefore used in
addition to the test results to improve the background for the actual chosen values. Drag
coefficients related to two typical shapes are given in [45], as shown in Figure C4.7.19,
for a triangular box section and two plates mounted normally on each other. A
triangular chord is a combination of these cross sections. The numbers at different
directions are compared in Table C4.7.4. The drag coefficients were determined by
vectorial summation of drag forces in direction 1 and 2 according to Figure C4.7.19.
To relate the drag coefficient to a fixed dimension Di = D the back plate width is
chosen. A fixed dimension and directional dependent drag coefficient is convenient for
modeling purposes. The drag coefficient related to this fixed diameter may be
computed as:
CDi = CDpr(θ) * Dpr(θ) / Di

where;
CDpr(θ) = the drag coefficient referenced to the projected diameter.
⎧170. ; θ = 0°
⎪⎪195
. ; θ = 90°
= ⎨140 . ; θ = 105°
⎪165
. ; θ = 180°−θ o
⎪⎩2.00 ; θ = 180°
Dpr(θ) = the projected diameter of the chord determined as:
⎧D cos( θ ) ; 0 < θ < θo
⎪
Dpr(θ) = ⎨W sin( θ ) + D / 2 cos( θ ) ; θ o < θ < 180 − θ o (7.9)
⎪⎩D cos(θ ) ; 180 − θ o < θ < 180
θo = the angle where half the backplate is hidden behind the rackplate,
determined as θo = tan-1(D/2W).
Table C4.7.4 Comparison of drag coefficients for simple sections

and chord CDpr evaluated from tests.
θ ⊥ /45/ Δ /45/ PRACTICE
CDpr: 0 1.7 1.3 1.70

45 2.5 1.8 1.825
90 2.2 1.95
135 1.5 1.3 1.50
180 2.0 1.8 2.00
(W/D = 1.1)
As a basis for the suggested drag coefficients the results available from TEES [67] and
DHL [61,62] were considered together with the recommendations in [46]. The drag
coefficients recommended in the PRACTICE are compared with the TEES test results
in Figure C4.7.20.
The inertia coefficient CMi = 2.0 may be applied for all directions, related to the
equivalent volume of πDi2/4 per unit length, where Di = D, the backplate dimension.
This assumes that the outline cross sectional area is approximately πDi2/4. If the rack
width is not of similar size to the backplate dimension, a more detailed consideration of
the inertia coefficient should be made if the loadings on the leg are not drag dominated
i.e. if the results are sensitive to the choice of inertia coefficient.
Explicit dependence on KC may be taken according to Figure C4.7.9, but is normally

not relevant for extreme loading conditions. Dependence on k/D or Re may normally
be neglected.
Figure C4.7.18 Definition of dimensions and angles for a triangular chord
C4.7.6 Other shapes
For other shapes, or groups of elements, see e.g. [45].

Figure C4.7.19 Drag coefficients for basic sections in uniform flow [45]
Figure C4.7.20.a : Marathon LeTourneau 116C
Figure C4.7.20.b : Marathon LeTourneau Gorilla
Figure C4.7.20 Comparison between TEES test results and the PRACTICE
C4.8 OTHER CONSIDERATIONS
For the most critical individual leg members the possibility of local vortex induced
vibrations should be evaluated. This check will normally be covered at the design
stage. However, if the site conditions of wind or current and/or wave height exceed
those used for design such a check may be required. This is because vortex induced
vibrations may lead to very high local stresses and a major contribution to fatigue
loading.
Vortex induced resonance will not normally be occur if:

S < 0.2 for tubulars
< 0.145 for flat plates
where;
S = [(fi Di)/vn], the Strouhal number
vn = flow velocity normal to the member
Di = diameter of member
fi = fundamental vibration frequencies of member (in Hz).
Further information and bounds for S above which vortex shedding will not occur may
be found in [45] and [59].
A = cross sectional area of member.
Ae = equivalent area of leg per unit height = (Σ Ai li )/s.
Ai = equivalent area of element = π Di2 / 4.
As = sum of projected areas for all members in the considered plane.
At = total projected envelope area of the considered plane.
CD = drag coefficient.
CDo = The drag coefficient for chord at direction θ = 0°.
CD1 = The drag coefficient for flow normal to the rack, θ = 90°.
CDe = equivalent drag coefficient.
CDei = equivalent drag coefficient of member i
= CDi (Di/De) (vertical member)
= sin3αi CDi (Dili/Des).
CDi = drag coefficient of an individual member, related to Di.
CDpr(θ) = drag coefficient to the projected diameter.
CDrough = drag coefficient for a rough member.
CDsmooth = drag coefficient for a smooth member.
Ch = height coefficient for wind.
CMe = equivalent inertia coefficient.
CMei = equivalent inertia coefficient of member i
= 1.0, horizontal member with flow along the length axis
= 1.8, vertical rough tubular with flow normal to the length axis
= 2.0, other vertical members with flow normal to the length axis
= 1.0, flat plates (brackets).
CMi = inertia coefficient of a member, related to Di.
d = the mean, undisturbed water depth (positive).
D = member diameter.
De = equivalent diameter of leg bay.
Di = reference dimension of individual leg members.
Dl = face width of leg, outside dimensions.
Dpr(θ) = projected diameter of the chord.
fi = fundamental vibration frequencies of the member.
Hs = significant wave height.
k = roughness height.
k/D = relative roughness.
KC = Keulegan-Carpenter number.
li = length of member 'i' node to node.
N = constant in wind velocity power law.
rn = velocity of the considered member, normal to the member axis and in the direction
of the combined particle velocity.
Re = Reynolds number.
s = length of one bay, or part of bay considered.
S = Strouhal's number.
S = average wave steepness.
S = outer diameter of an array of tubulars.
T = wave period.
Tn = first natural period of sway motion.
Tz = zero-upcrossing period.
u = particle velocity.
un = liquid particle velocity normal to the member.
u n = liquid particle acceleration normal to the member.
ux, u x = horizontal water particle velocity and acceleration.
U = flow velocity at the depth of the considered element.
UC = particle velocity by current.
Um = maximum orbital particle velocity.
GLOSSARY OF TERMS FOR SECTION C4 (Continued)
Ured = reduced particle velocity for regular waves, Ured = uTn/D.

UW = particle velocity by waves.
vn = total flow velocity normal to the member.
VC = the reduced current velocity to be used in analysis.
VC = the current velocity to be used in the hydrodynamic model, VC should not be
taken less than 0.7Vf.
VCn = current velocity normal to member used in the hydrodynamic model.
Vf = far field (undisturbed) current.
W = dimension from backplate to pitch point of triangular chord or dimension from
root of one rack to tip of other rack of split -tubular chord.
W = the width of the structure.
xo = motion amplitude used in consideration of the applicability of the relative
velocity formulation.
z = coordinate measured vertically upward from the mean water surface.
z' = modified coordinate to be used in particle velocity formulation.
Z = elevation measured from the mean water surface.
α = indicator for relative velocity, 0 or 1.

αi = angle defining flow direction relative to member.
β = Re/KC parameter to describe the test environment.
βi = angle defining the member inclination.
ΔFdrag = drag force per unit length.
ΔFinertia = inertia force per unit length.
ΔFinertiaH = horizontal inertia force per unit length.
λ = wave length.
ν = kinematic viscosity.
θ = angle in degrees for waves relative to the chord orientation.
θo = angle where half the backplate is hidden behind the rackplate.
ρ = mass density of water or air.
ζ = the instantaneous water surface elevation (same axis system as z)
ζo = the wave crest (same axis system as z).
1 Keulegan, G.H., Carpenter, L.H., 'Forces on Cylinders and Plates in an Oscillating Fluid',
Journal of Research of the National Bureau of Standards, Volume 60, No. 5, May 1958.
2 Sarpkaya, T., 'In-Line and Transverse Forces on Smooth and Sand-Roughened Cylinders in
Oscillatory Flow at High Reynolds Numbers', Naval Postgraduate School, Report NPS-69
SL 76062, 1976.
3 Hogben, N., Miller, B.L., Searle, J.W., Ward, G., 'Estimation of Fluid Loading on Offshore
Structures', Proc. Institution Civil Engineers, Part 2, 1977.
4 Sarpkaya, T., 'Wave Forces on Inclined Smooth and Rough Circular Cylinders', Offshore
Technology Conference, Paper OTC 4227, 1982.
5 Sarpkaya, T., Bakmis, C., Storm, M.A., 'Hydrodynamic Forces from Combined Wave and
Current Flow on Smooth and Rough Circular Cylinders at High Reynolds Numbers',
Offshore Technology Conference, Paper OTC 4830, 1984.
6 Sarpkaya, T., Storm, M.A., 'In-Line Force on a Cylinder Translating in Oscillatory Flow',
Applied Ocean Research, Volume 7, No. 4, 1985.
7 Sarpkaya, T., 'Force on a Circular Cylinder in Viscous Oscillatory Flow at Low Keulegan-
Carpenter Numbers', Journal of Fluid Mechanics, Volume 165, 1986.
8 Sarpkaya, T., 'Past Progress and Outstanding Problems in Time-Dependent Flows about
Ocean Structures', Proc. of Separated Flow around Marine Structures, The Norwegian
Institute of Technology, Trondheim, Norway, 1985.
9 Sarpkaya, T., 'On Fluid Loading of Offshore Structures - After Ten Years of Basic and
Applied Research', Offshore Operations Symposium, 9th ETCE, New Orleans, 1986.
10 Nath, J.H., 'Heavily Roughened Horizontal Cylinders in Waves', Proceedings of BOSS,

1982.
11 Bearman, P.W., Chaplin, J.R., Graham, J.M.R., Kostense, J.K., Hall, P.F., Klopman, G.,
'The Loading on a Cylinder in Post-Critical Flow Beneath Periodic and Random Waves',
Proceedings of BOSS, 1985.
12 Kasahara, Y., Koterayama, W., Shimazaki, K., 'Wave Forces Acting on Rough Circular
Cylinders at High Reynolds Numbers', Offshore Technology Conf., Paper OTC 5372,
1987.
13 Chaplin, J.R., 'Loading on a Cylinder in Uniform Oscillatory Flow: Part I - Planar

Oscillatory Flow', Applied Ocean Research, Vol.10, No. 3, 1988.
14 Davies, M.J.S., Graham, J.M.R., Bearman, P.W., 'In-Line Forces on Fixed Cylinders in
Regular and Random Waves', Society for Underwater Technology, Volume 26:
Environmental Forces on Offshore Structures and Their Prediction, 1990.
15 Rodenbusch, G., Gutierrez, C.A., 'Forces on Cylinders in Two-Dimensional Flows', Report

BRC13-83, Shell Development Co., 1983.
REFERENCES FOR SECTION C4 (Continued)
16 Rodenbusch, G., Källström, C., 'Forces on a Large Cylinder in Random Two-Dimensional

Flows', Offshore Technology Conference, Paper OTC 5096, 1986.
17 Theophanatos, A., Wolfram, J., 'Hydrodynamic Loadings on Macro-Roughened Cylinders

of Various Aspect Ratios', Journal of Offshore Mechanics and Arctic Engineering, Volume
111, No. 3, 1989.
18 Klopman, G., Kostense, J.K., 'The Loading on a Vertical Cylinder in Random Waves at
High Reynolds Numbers', Water Wave Kinematics, pp. 679-699, 1990.
19 Heideman, J.C., Olsen, O.A., Johansson, P.I., 'Local Wave Force Coefficients', Civil
Engineering in the Oceans IV, ASCE, 1979.
20 Nath, J., 'Biofouling and Morison Equation Coefficients', Proceedings 7th. International
Conf. on Offshore Mechanics and Arctic Engineering, 1988.
21 Wolfram, J., Theophanatos, A., 'Marine Roughness and Fluid Loading', Society for
Underwater Technology, Volume 26: Environmental Forces on Offshore Structures and
Their Prediction, 1990.
22 Roshko, A., 'Experiments on the Flow Past a Circular Cylinder at Very High Reynolds
Number', Journal of Fluid Mechanics, Volume 10, Part 3, 1961.
23 Miller, B.L., 'The Hydrodynamic Drag of Roughened Circular Cylinders', Transactions

RINA, 1976.
24 Pearcey, H.H., Cash, R.F., Salter, I.J., 'Flow Past Circular Cylinders: Simulation of Full-
Scale Flows at Model Scale', NMI Report R131, 1982.
25 'Roughness and Vortex Shedding Effects for Cylinders in Flume and Real Sea Waves',
Report for the Health and Safety Executive by Advanced Mechanics Engineering Limited,
1991.
26 Bishop, J.R., 'An Analysis of Peak Values of Wave Forces and Particle Kinematics from
the Second Christchurch Bay Tower', NMI Report R180, 1985.
27 Bishop, J.R., 'Wave Force Experiments at the Christchurch Bay Tower with Simulated
Hard Marine Fouling', Report No. OTI 89 541, HMSO, 1989.
28 Pearcey, H.H., Matten, R.B. and Singh, S., 'Fluid Forces for Cylinders in Oscillatory Flow
Waves and Currents when Drag and Inertia Effects Are Present Together', BMT Report to
the Health and Safety Executive OT-0-86-011, 1986.
29 Det Norske Veritas Rules for the Classification of Fixed Offshore Installations.
30 J.R. Morison, M.P O'Brien, J.W. Johnson, S.A. Schaaf, 'The Forces Exerted by Surface
Waves on Piles', J. of Petr. Techn., American Inst. of Mining Engrs., Vol. 189, 1950, p149-
154.
31 S.K. Chakrabarti, 'Hydrodynamics of Offshore Structures', Springer Verlag, Berlin,

Hedelberg, 1987.
32 O.M. Faltinsen, 'Sea Loads on Ships and Offshore Structures', Cambridge University Press,
Trumpington Street, Cambridge, 1990.
33 BSI Code of Practice No. 3, Chapter 5, Part 2, 'Wind Loads', September 1972
34 J.H. Vugts, 'A Review of Hydrodynamic Loads on Offshore Structures and Their
Formulation', BOSS'79, Imperial College, London, England, August 1979
35 E.J. Laya, J.J. Connor and S.Shyam Sunder, 'Hydrodynamic Forces on Flexible Offshore
Structures', Journal of Engineering Mechanics, Vol. 110, No.3, 1984.
36 S.R. Winterstein, E.M. Bitner-Gregersen and K. Ronold, 'Statistical and Physical Models
of Nonlinear Random Waves', OMAE, Volume II, Safety and Reliability, Stavanger, 1991,
pp.23-31.
37 J.D. Wheeler, 'Method for Calculating Forces Produced by Irregular Waves', OTC, paper
no. 1006, Dallas, Texas, 1969.
38 J.E. Skjelbreia, G. Berek, Z.K. Bolen, O.T. Gudmestad, J.C. Heideman, R.D. Ohmart, N.
Spidsoe and A. Torum, 'Wave Kinematics in Irregular Waves', OMAE, Stavanger, 1991.
39 R.G. Bea and N.W. Lai, 'Hydrodynamic Loadings on Offshore Platforms', OTC paper no
3064, May, 1978, pp. 155-168.
40 R.G. Dean and P.M. Aagaard, 'Wave Forces: Data Analysis and Engineering Calculations',
Journal of Petroleum Technology, March 1970, 105-119.
41 J.R. Chaplin and T.P Flintham, 'Breaking Wave Forces on Tubulars', 3rd International
Jack-up Conference, City University, London, September 1991.
42 L. Skjelbreia and J.A. Hendricksen, 'Fifth-order Gravity Wave Theory', Proceedings of

Seventh Conference on Coastal Engineering, 1961, pp. 184-196.
43 O.J. Andersen, E. Førland and S. Haver, Design Basis, Environmental Conditions,

Statfjord, Statoil Report no. F&U-ST 88007, Stavanger, April 25 1988.
44 R.C.T. Rainey, 'Christchruch Bay Tower Compliant Cylinder Project, Final Summary
Report and Conclusions', OTH-90-139, WS Atkins Engineering Sciences, Surrey, January,
1991.
45 DNV Classification note 30.5, 'Environmental Conditions and Environmental Loads', July
1990.
46 'Practice for the Site Specific Assessment of Jack-up Units', By Marine Technology
Division, SIPM, EDP-5, The Hague, May 1989,
47 WAJAC, Veritas Sesam Systems Report no. 82-6108, Høvik, Dec. 1984.
48 N. Pharr Smith, D.B. Lorenz, C.A. Wendenburg and J.S. Laird, 'A Study of Drag
Coefficients for Truss Legs on Self-Elevating Mobile Offshore Drilling Units', SNAME
Tansactions, Vol. 91, 1983, pp. 257-273.
49 N. Pharr Smith and C.A. Wendenburg, 'A study of Drag Coefficients', The Jack-Up
Drilling Platform, Ed. L. Boswell, City University, London, 1989.
50 Yoshiharu, Ideguchi, 'Legs Drag Coefficients of Enhanced 300IC Jack-Up Rig', TSU
Research Laboratories Technical Research Center, Nippon Kokan K.K. Report No.
822117, June 1982
51 Borgman, L.E., 'Random Hydrodynamic Forces on Objects', Annals of Mathematical

Statistics, Vol. 38., 1967, pp. 37-51.
52 Atkins Engineering Services, 'Fluid Loading on Fixed Offshore Structures', OTH 90 322,
1990.
53 Taylor, P., 'Current Blockage - Reduced Forces on Steel Platforms in Regular and Irregular
Waves with a Mean Current', Offshore Technology Conference, OTC 6519, Houston,
1991.
54 Heideman J.C. and Schaudt K.J., 'Recommended Equations for Short-term Statistics of
Wave Heights and Crest Heights', 1 April 1987
55 Moe, G. and Verley, R.L.P, 'Hydrodynamic Damping of Offshore Structures in Waves and
Currents', The Offshore Technology Conference, OTC 3798, Houston, 1980.
56 American Petroleum Institute, proposal for an update of the API-RP2A, 'Hydrodynamic

Force Guidelines for U.S. Waters', received 6. February 1992.
57 The Norwegian Petroleum Directorate, 'Regulation for structural design of loadbearing

structures intended for exploitation of petroleum resources', 1988.
58 Health and Safety Executive, 'Offshore Installations: Guidance on design, construction and
certification', London, 1990.
59 N.D.P. Barltrop, A.J. Adams, 'Dynamics of Fixed Marine Structures', Third Edition,
Butterworth Heinemann, 1991.
60 MacCamy R.S., Fuchs R.A., 'Wave Forces on Piles: A Diffraction Theory", U.S. Army
Corps of Engineers, Beach Erosion Board, Tech. Memo No 69, Washington DC, 1954.
Other project reports related and technical notes:
61 G.H.G. Lagers, 'Morison Coefficients of Jack-Up Legs', MSC report 1005, Schiedam, The
Netherlands, February 1990.
62 G.H.G. Lagers, 'Collected Morison Coefficients of Jack-Up Leg Elements', MSC report
1715, Schiedam, The Netherlands, June 1991.
63 S.Th. Schurmans et. al., DHL, 'Wave Forces on Jack-Up Legs', Delft Hydraulics
measurement report 8603-1505, Delft, 1990.
64 A. Løken, 'Review of DHL test data', Veritec report no. 91-3372, Høvik, 1991.
65 Løseth R.M., Arnesen Y., 'Check of data reduction of DHL data', DNVC report No 92-
1054, Høvik, August 1992.
Other project reports related and technical notes (continued):
66 D. Karunakaren, 'Scaling of Hydrodynamic Loads According to Computational Models',

Technical memo no. 710762, SINTEF, Trondheim, July, 1991.
67 N.P. Smith, TEES Test Data, Extract, December 6, 1991.

APPENDIX C4.A : EXAMPLE OF EQUIVALENT MODEL COMPUTATIONS
Figure C4.A.1 Model of a bay for test purposes
Table C4.A.1 Computations of equivalent model for heading 0° to be used in

site assessment for z < MWL + 2m, chord W/D = 1.13.
i αi βi cos…*) CDi Di li CDi*Di*li*cos…
1 (30) 90. 1.0 1.0 .65 5.0 3.25

Chords: 2 (30) 90. 1.0 1.0 .65 5.0 3.25
3 (90) 90. 1.0 2.124 .65 5.0 6.90
4 -30 26.7 0.25 1.0 .30 11.2 0.84

5 -30 -26.7 0.25 1.0 .30 11.2 0.84
Inclined 6 30 26.7 0.25 1.0 .30 11.2 0.84
braces 7 30 -26.7 0.25 1.0 .30 11.2 0.84
8 90 26.7 1.0 1.0 .30 11.2 3.36
9 90 -26.7 1.0 1.0 .30 11.2 3.36
span 10 -30 0. 0.125 1.0 .10 5.0 0.06

breakers 11 30 0. 0.125 1.0 .10 5.0 0.06
12 90 0. 1.0 1.0 .10 5.0 0.50
ΣCDi*Di*li*cos… = 24.10
s = 5.0 m
C De * D e = ∑ C Di D i l i / s = 4.82
D e = ( ∑ D i 2 l i / s) . ⎫⎪
= 158
C De = 4.82 / 158
. = 3.05⎬equivalent model
C Me = 2.0 ⎪
⎭
*) Geometric factor,
see Section 4.6.6 [sin2βi + cos2βi sin2αi]3/2
the PRACTICE
Table C4.A.2 Computations of equivalent model for heading 0° to be compared
with model test results, chord W/D = 1.13, model scale = 1:4.264
1 (30) 90 1.0 0.65 .152 1.178 0.1164

Chords: 2 (30) 90 1.0 0.65 .152 1.178 0.1164
3 (90) 90 1.0 2.124 .152 1.178 0.3803
4 -30 26.7 0.25 0.65 .07 2.628 0.02899

5 -30 -26.7 0.25 0.65 .07 2.628 0.02899
Inclined 6 30 26.7 0.25 0.65 .07 2.628 0.02899
braces 7 30 -26.7 0.25 0.65 .07 2.628 0.02899
8 90 26.7 1.0 0.65 .07 2.628 0.11957
9 90 -26.7 1.0 0.65 .07 2.628 0.11957
span 10 -30 0. 0.125 0.65 .025 1.173 0.00238

breakers 11 30 0. 0.125 0.65 .025 1.173 0.00238
12 90 0. 1.0 0.65 .025 1.173 0.01906
s = 1178
.
C De * D e = 0.842 C De * D e = 359
.
De = ∑ D i 2 1i / s = 0.370 ⎫⎪ De = 158
. ⎫⎪
C De = 0.842 / 0.370 = 2.277⎬ equiv. model C De = 2.277⎬ equiv. model
C Me = 2.0 ⎪ (model scale) C Me = 2.0 ⎪⎭ (full scale)
⎭
Table C4.A.3 Computations of equivalent model for heading 30° to be compared

with model test results, chord W/D = 1.13, model scale = 1:4.264
1 (60) 90 1.0 1.663 .152 1.178 0.2978

Chords: 2 (60) 90 1.0 1.663 .152 1.178 0.2978
3 (30) 90 1.0 0.65 .152 1.178 0.1164
4 0 26.7 0.091 0.65 .07 2.628 0.01088

5 0 -26.7 0.091 0.65 .07 2.628 0.01088
Inclined 6 60 26.7 0.716 0.65 .07 2.628 0.08561
braces 7 60 -26.7 0.716 0.65 .07 2.628 0.08561
8 30 26.7 0.254 0.65 .07 2.628 0.03037
9 30 -26.7 0.254 0.65 .07 2.628 0.03037
span 10 60 0. 0.650 0.65 .025 1.173 0.01239

breakers 11 60 0. 0.650 0.65 .025 1.173 0.01239
12 30 0. 0.125 0.65 .025 1.173 0.00238
s = 1178
.
C De * D e = 0.843 C De * D e = 359
.
De = ∑ D i 2 1i / s = 0.370 ⎫⎪ De = 158
. ⎫⎪
C De = 0.843 / 0.370 = 2.278⎬ equiv. model C De = 2.278⎬ equiv. model
⎭
Figure C4.A.2 Square bay with triangular chords
Table C4.A.4 Square bay with triangular chords, Equivalent model to be used in
site assessment z < MWL + 2m.
1 45 90 1.0 1.65 .71 3.4 3.983

Chords: 2 45 90 1.0 1.65 .71 3.4 3.983
3 135 90 1.0 1.79 .71 3.4 4.321
4 135 90 1.0 1.79 .71 3.4 4.321
5 0 40.2 0.2689 1.0 .32 10.6 0.912

Inclined 6 90 40.2 1.0 1.0 .32 10.6 3.392
braces 7 90 40.2 1.0 1.0 .32 10.6 3.392
8 0 40.2 0.2689 1.0 .32 10.6 0.912
Side 9 0 0. 0.091 1.0 .32 11.2 0.0

Horiz 10 90 0. 1.0 1.0 .32 11.2 3.584
11 90 0. 1.0 1.0 .32 11.2 3.584
12 0 0. 0.091 1.0 .32 11.2 0.0
span 13 45 0. 0.354 1.0 .23 11.4 2.622

breakers 14 45 0. 0.354 1.0 .23 11.4 2.622
brackets 15 90. 0. 1.0 2.0 .98 0.98 1.921

s = 3.4
C De * D e = 1163.
De = ( ∑ D i l i / s)
2
= 2.30⎫⎪
C De = 1163
. / 2.30 = 5.06 ⎬ equivalent model
C Me = 2.0 ⎪
⎭
Table C4.A.5 Square bay with triangular chords, Equivalent model to be used in
comparison with test results, model scale 1:4.256.
1 45 90 1.0 1.65 .167 0.8008 0.2207

Chords: 2 45 90 1.0 1.65 .167 0.8008 0.2207
3 135 90 1.0 1.79 .167 0.8008 0.2989
4 135 90 1.0 1.79 .167 0.8008 0.2989
5 0 40.2 0.2689 0.65 .076 2.481 0.0330

Inclined 6 90 40.2 1.0 0.65 .076 2.481 0.1226
braces 7 90 40.2 1.0 0.65 .076 2.481 0.1226
8 0 40.2 0.2689 0.65 .076 2.481 0.0330
Side 9 0 0. 0.091 0.65 .076 2.628 0.0

Horiz. 10 90 0. 1.0 0.65 .076 2.628 0.1298
11 90 0. 1.0 0.65 .076 2.628 0.1298
12 0 0. 0.091 0.65 .076 2.628 0.0
span 13 45 0. 0.354 0.65 .054 2.680 0.0333

breakers 14 45 0. 0.354 0.65 .054 2.680 0.0333
brackets 15 90. 0. 1.0 2.0 .231 0.231 0.1070

SCDi*Di*li*cos… = 1.7836
s = 0.8008
C De * D e = 2.227 C De * D e = 9.48
De = ∑ D i 2 1i / s = 0.542⎫⎪ De = 2.307⎫⎪
C De = 2.227 / 0.542 = 4109
. ⎬ equiv. model C De = 4109
. ⎬ equiv. model
⎭
APPENDIX C4.B : COMPARISON CASES TO ASSESS IMPLICATIONS OF PRACTICE
FORMULATION
Computations are performed on a 'simplified model' with no mass. The irregular and regular
wave results are computed according to the PRACTICE, Section 3 and 4. These computations
are made to asses the implications of changes made concerning drag coefficients and wave
kinematics formulations compared with previous practices.
The significant wave height is chosen as judged realistic for the two water depths investigated:
Water depth 30m:

significant wave height Hsrp = 10m
Water depth 90m:

significant wave height Hsrp = 14m
The period range specification is taken from PRACTICE, Section 3.5:
. H s ) < T < (19.5H s )

(118
. ( H s ) < TZ < 3.6 ( H s )
35
The current and current profile is often site dependent. The current is here set be constant over
the water depth, extrapolated to sea surface.
The example design for the computations is defined by:
Three legs, split tube chords,

Diameter chord (tubular) Dc = 0.7m,
Rack width W = 0.8m,
Diameter of braces Db = 0.3m,
length of braces per m. height lb = 13.44m,
leg spacing xleg = 50m
leg diameter D1 = 10m
Particulars for two existing practices and the PRACTICE
Practice I: PRACTICE:
Irregular waves:
CD d l cos CD d l cos
Tubular 1.0 *0.3 *13.44 *0.6 = 2.419 1.0 *0.3 *13.44 *0.6 = 2.419
Chord 1 2.114 *0.7 *1.0 *1.0 = 1.479 2.057 *0.7 *1.0 * 1.0 = 1.440
Chord 2,3 1.279 *0.7 *2.0 *1.0 = 1.790 1.056 *0.7 *1.0 *2.0 = 1.478
CDeDe 5.688 z < 1.5m 5.338
z > 1.5m 4.491
kinematics according to
Delta stretching Wheeler stretching
Regular waves :
Stokes' fifth Stokes' fifth (regular)
Tubular 0.7 *0.3 *13.44 *0.6 = 1.693

Chord 1 1.486 *0.7 *1.0 *1.0 = 1.040
Chord 2,3 0.896 *0.7 *2.0 *1.0 = 1.254
CDeDe = 3.989
No shielding assumed No shielding for waves

Reduction of current by
a factor
1/[1 + CDeDe/(4D1)] = 0.88
Practice II:
Regular waves:
Stokes' fifth
Tubular 0.64 *0.3 *13.44 *0.6 = 1.548

Chord 1 1.307 *0.7 *1.0 *1.0 = 0.915
Chord 2,3 0.973 *0.7 *2.0 *1.0 = 1.362
CDeDe = 3.825
Irregular waves:
Airy with constant stretching
CDeDe = 3.825 * 1.3 = 4.973

Table C4.B.1 Comparison including wave height scaling,
Water depth = 30 m, Hsrp = 10 m.
Case Environment Current Base Overturning

H/T Shear Moment
Hs/Tz
m and sec m/sec MNm
MN
Regular 0.0 5.58 143
waves 18.6/14.0
Practice I 1.0 8.03 197
H = Hmax = 1.86 Hsrp

Regular 0.0 5.42 139
waves 18.6/14.0
Practice II 1.0 7.70 189

Irregular 0.0 3.18 108
waves 10.0/11.0
Practice II 1.0 6.89 145
Hs = Hsrp
Regular 0.0 4.90 115
waves PRACTICE 16.5/14.0
0.88 6.96 157
H = Hdet = 0.86 Hmax
Irregular 0.0 5.95 122
Waves PRACTICE 11.84/11.0
Hs = [1+.5exp(-d/25)] 0.88 7.88 154

*Hsrp
Table C4.B.2 Comparison including wave height scaling,
Water depth = 90 m, Hsrp = 14.0 M.
Case Environment Current Base Overturning

H/T Shear Moment
Hs/Tz
m and sec m/sec MNm
MN
Regular 18.6/16.5 0.0 9.82 668
waves
Practice I 26.0/16.5 0.5 12.30 819

Regular 18.6/16.5 0.0 9.41 641
waves
Practice II 26.0/16.5 0.5 11.79 785

Irregular 14.0/13.0 0.0 11.22 747
waves
Practice II 14.0/13.0 0.5 13.11 859
Hs = Hsrp
Regular 23.14/16.5 0.0 8.90 578
waves PRACTICE
23.14/16.5 0.44 11.20 709
H = Hdet = 0.86 Hmax
Irregular 14.34/13.0 0.0 9.12 573
Waves PRACTICE
Hs = [1+.5exp(-d/25)] 14.34/13.0 0.44 10.80 671

*Hsrp
Comments to the results
The results for both the 30m and 90m water depth cases in Tables C4.B.1 and C4.B.2 show
improved agreement between regular and irregular wave force calculations for the PRACTICE
methodology as compared to Practice II.
The main differences between Practice II and the PRACTICE are that:
- the PRACTICE uses a reduced wave height for regular wave analysis instead of a
reduced drag coefficient.
- the PRACTICE includes a shallow water wave height correction to be applied to the
significant wave height used in irregular wave analysis.
The shallow water wave height correction term is described and justified in C3.5.1.1. The effect
of wave asymmetry in shallow water in Practice II is included only by a conservative kinematics
model above the mean water level for an irregular wave analysis. Other practices give no
consideration to shallow water effects in irregular wave analysis.
The agreement between regular and irregular wave forces is better at the 90m water depth case
than for the 30m water depth case. However, the correction term for shallow water cases is
justified as compared to the Practice II results.
APPENDIX C4.C : COMPARISON OF TEST RESULTS FOR CHORDS
Split tube chords compared in the following

rack ratio W/D
F&G 1.08
NKK 1.10
MLMC 1.13
MSC 1.18
MLMC 1.24
Figure C4.C.1 : Comparison of PRACTICE formulation with model tests, ratio

W/D = 1.08, [48]

W/D = 1.10, [50]

W/D = 1.13, [49]
Figure C4.C.4 : Comparison of PRACTICE formulation with model tests, rack

W/D = 1.18, [51]
Figure C4.C.5 : Comparison of PRACTICE formulation with model tests, rack

W/D = 1.24, [49]
C5 COMMENTARY TO CALCULATION METHODS - STRUCTURAL
ENGINEERING
C5.1 INTRODUCTION
The application of the procedures and techniques given in Section 5 is consistent with
the guidance given in the other sections of the Recommended Practice (PRACTICE).
Furthermore, it is assumed that the user of the PRACTICE is familiar with the general
philosophy and design/assessment approach specifically applicable to jack-ups. To
provide additional guidance to the analyst less familiar with these procedures and to
also ensure consistency of application by all users, this commentary has been prepared.
In general, the structural modeling for the assessment of a jack-up must achieve the
following objectives for both the static and (where applicable) dynamic responses:
• Realistic global response (i.e. displacement, base shear, overturning moments, etc.)
for the unit under the applicable environmental and functional loads.
• Represent the correct linear and non-linear characteristics of the leg, leg-hull
connection and the leg-foundation interaction.
• Sufficient detail to allow for detailed assessment of the adequacy of the leg
structure, structural/mechanical components of the jacking system and the
foundation.
C5.2 GENERAL
Prior to beginning the actual modeling of the jack-up unit the analyst should ensure that
all data necessary to perform the assessment is available. Refer first to the
accompanying "Guideline for Site Specific Assessment of Mobile Jack-up Units" and
Section 3 of the PRACTICE - Assessment Input Data for guidance on the data needed
and for the rationale as to why they are important to the assessment procedure.
Once these data are collected it would also benefit the analyst to review Section 4 -
Calculation Methods - Hydrodynamics and Wind Forces and Section 6 - Calculation
Methods - Geotechnical Engineering of the PRACTICE. This will serve not only as
confirmatory check to assure the completeness of the data being collected for the
analysis, but will also allow the analyst to evaluate the level of analysis techniques to
be used in light of the computer software available. It is important that the analyst
anticipate the complete scope of the assessment in terms of the level of quasi-static and
dynamic analyses which may be required. This will allow the analyst to optimize the
structural modeling and reduce the duplication of effort.
In the remaining sections of this Section of the Commentary the analyst is guided
through the overall structural modeling of a jack-up for the following:
(1) Quasi-static analysis procedures
(2) Assessment criteria procedures
For guidance on dynamic analysis techniques, refer to Section 7 of the PRACTICE.

This Commentary has been focused on providing a general discussion of key points and
their impact/importance on the final assessment results. Thus it is important that if
additional detail is required, the analyst refer to appropriate technical references or
contact the jack-up designer for further guidance.
C5.3 GLOBAL RESPONSE
C5.3.1 In the analysis of jack-ups the global response is developed by appropriate

combination/modeling of both the first order (linear) response and the second order
(non-linear) response. The first order response of a jack-up can be derived from
models of varying degrees of complexity as discussed in Section 5.6 of the
PRACTICE. Please refer to Figure C5.1 for the resulting response/reactions which are
derived from a first order analysis.
C5.3.2 The effects noted in Section 5.4 - 5.5 of the PRACTICE (e.g. leg inclination, P-Δ)
should be determined and combined with the forces generated by the first order analysis.
These effects have been shown to be significant (in varying degrees of importance) in the
assessment of jack-ups based on the specific circumstances of the jack-up and the
specific site conditions being considered. The actual treatment/derivation of the response
is highly dependent on the modeling complexity chosen and the computer software
analysis packages available to the analyst. The guidance given in Sections 5.4 - 5.5 of
the PRACTICE allows for this needed flexibility in addressing the full spectrum of
approaches by allowing for implicit incorporation of these second order effects by using
the appropriate non-linear modeling capabilities of the computer software or by explicit
addition of a conservative incorporation of these effects by approximate/simplified hand
calculations. Please refer to Figure C5.2 for further representation of the effects of P-Δ
and leg inclination and Figure C5.3 as to the applicable contribution of these second
order effects that are required to be incorporated with the first order response.
It should be noted that the PRACTICE recommends that P-Δ effects are included
throughout the assessment. It also recommends that the effects of initial leg inclination
are included only at the end of the analysis in the structural strength checks (by means
of an additional applied moment).
The derivation of the alternative geometric stiffness approach to P-Δ of Section 5.5.4.3
of the PRACTICE is given in Appendix C5.A.
C5.3.3 Although the "Recommended Practice" gives guidance as to simplified calculations that
can be performed for P-Δ and leg inclination, no specific guidance has been given for
the additional moment caused by hull sagging.
When a unit is installed on location the legs will normally engage the seabed with the
hull supported by its own buoyancy in a hogged condition. Subsequently, with the hull
slightly clear of the water, preload ballast will be taken on board thus preloading the
legs to achieve their final penetration. This will normally lead to an extreme hull
sagging condition. Finally the preload ballast is dumped and the hull elevated to the
required airgap for the location. In this condition the hull will be sagging under self
weight and variable loads.
As explained in Section 5.7 of the PRACTICE, the leg shear and bending moments
caused by hull sagging are very dependent on leg guide clearances, the design of the
jacking system, operational parameters and the modeling used in the analysis. A
simplified approach for a conservative quantitative assessment is to assume that 25 to 50
percent of the theoretical hull sagging moment at the lower guide is seen in practice.
This may be accounted for in a global model by reducing the distributed hull mass by 75
to 50 percent and applying the residual mass as point masses on the hull adjacent to the
connections to the legs. This procedure is not applicable when hull stresses are required.
A more thorough method is to apply self equilibrating pairs of forces/moments across the
spring connections between hull and legs:
An alternative approach is to allow a relaxation of the horizontal seabed restraint by

means of prescribed displacements.
C5.4 DISCUSSION OF THE LEG-HULL CONNECTION
In addition to the importance of understanding the global response of the jack-up, it is

important that the analyst has a full appreciation of the leg-hull interface. A
representative leg-hull connection is shown in Figure C5.4. The basic function of the
leg-hull connection is to provide for a transition of forces between the leg and hull via:
1) the upper and lower guide which transfer bending moments by a set of
horizontal forces
and
2) a jacking system and/or fixation system which transfers vertical load and
bending moment via a set of vertical forces.
Section 5.6.6 gives guidance on the detailed modeling requirements for each of the
following components:
• upper and lower guides

• jacking pinions
• fixed jacking system
• floating jacking system
• fixation system
• shock pads
• jackcase and associated bracing
The various combinations of the above components to create the leg-hull connections
on typical jack-ups now in service are given in Figure C5.5. Close attention to the leg-
hull connection should be given by the analysts to ensure a thorough understanding of
the jacking system so that proper modeling is realized. The number of key variables
which must be properly incorporated are:
• bending, shear and torsional stiffness of the leg between the upper and lower
guide
• axial, bending, shear and torsional stiffness of the jackcase stiffness of the upper
and lower guides
• amount of clearance/tolerance of the legs within the guides (see Figure C5.6)
• amount of backlash in the jacking system (see Figure C5.7)
• type of leg guide arrangement (see Figure C5.8)
• rack/pinion arrangement (opposed versus unopposed pinions) (see Figure C5.9)
When accounting for the effects of clearances (e.g. between guides and leg) in simpler
models there are several approaches available:
• For estimates of extreme behavior, use a 'secant stiffness' approach, so that the
springs used provide a realistic displacement for the load/deflection levels
expected. Thus the equivalent stiffness will include any slack behavior:
• An alternative to the above, which is less sensitive to the initial estimate of

load and deflection, is to use a pair of self equilibrating applied forces to
represent the slack so that the spring is initially under a compressive load Po,
where Po = kδo:
Note: When using the methods above any spring with true tensile loading must be
manually released.
• For natural frequency calculations, providing that the gaps do not open and
close, the use of the stiffness for closed gaps is appropriate.
C5.5 DETERMINATION OF PROPERTIES FOR EQUIVALENT MODELLING OF LEG
AND LEG-HULL CONNECTION
C5.5.1 Equivalent leg stiffness

The determination of stiffnesses for the equivalent leg model referred to in Section
5.6.4b) of the PRACTICE may be accomplished the following means:
• Hand calculations using the formulae presented in Figure 5.1 (after DNV, with
corrections). Provided that there are no significant offsets between the brace
work points these will be reasonably accurate for cases A (sideways K bracing),
C (X bracing) and D (Z bracing); case B (normal K bracing) should be used
with caution as the values of equivalent shear area and second moment of area
are dependent on the number of bays being considered. If the leg scantlings
change in different leg sections this can be accounted for by calculating the
properties for each leg section and creating the equivalent leg model
accordingly.
• The application of unit load cases to a detailed leg model in accordance with
Section 5.6.4 a) of the PRACTICE. The following load cases should be
considered, applied about the major and minor axes of the leg:
- Axial load. This is used to determine the axial area, A, of the equivalent
beam according to standard theory:
FL FL
Δ= => A =
AE EΔ
where;
Δ = axial deflection of cantilever at point of load application
F = applied axial end load
L = length of cantilever (from rigid support to point of load
application)
E = Young's modulus
- Pure moment applied either as a moment or a couple. This is used to derive the
second moment of area (I) according to standard beam theory:
ML2 ML2 ML ML
δ= => I = and θ = => I =
2EI 2Eδ EI Eθ
where;
δ = lateral end deflection of cantilever at point of load application
M = applied end moment
θ = end slope of cantilever at point of load application
It should be noted that the value of I resulting from the two equations may
differ somewhat.
- Pure shear, P, applied at the end of the leg which may be used to derive I
according to standard beam theory:
PL2 PL2
θ= => I =
2EI 2Eθ
Using either this value of I, or a value obtained from the pure moment case, the
effective shear area, As, can then be determined from:
PL3 PL 7.8 PLI
δ= + => As =
3EI As G 3EIδ − PL3
where;
G = shear modulus = E/2.6 for poissons ratio of 0.3
C5.5.2 Equivalent leg-hull connection stiffness
The determination of stiffnesses for the equivalent leg-hull connection model referred
to in Section 5.6.6 f) of the PRACTICE may be accomplished the following means:
• Hand calculations using the formulae presented in Section 7.3.5.3 of the

PRACTICE.
• The application of unit load cases to a detailed leg model in combination with a
detailed leg-hull connection model in accordance with Sections 5.6.4 a) and
5.6.6 a)-e) of the PRACTICE. Unit load cases are again applied, as described in
C5.5.1. In this instance the differences between the results from the detailed leg
model alone (see C5.5.1) and the detailed leg plus leg-hull connection model
allow the effective stiffness of the connection to be determined:
- Axial load. This is used to determine the vertical leg-hull connection

stiffness, Kvh from the axial end displacements of the detailed leg model, Δ,
and the axial end displacements of the combined leg and leg-hull connection
model, ΔC, under the action of the same loading, F:
Kvh = F/(ΔC-Δ)
- Pure moment applied either as a moment or a couple. This is used to derive

the rotational connection stiffness, Krh from either the end slopes, θ and θC,
or the end deflections, δ and δC, of the two models under the action of the
same end moment, M:
Krh = M/(θC-θ) or Krh = ML/(δC-δ)
- Pure shear which may be used to determine the horizontal leg-hull

connection stiffness, Khh, in a similar manner, accounting for the rotational
stiffness already derived. Normally the horizontal leg-hull connection
stiffness may be assumed infinite.
If the model contains nonlinearities due to the inclusion of gap elements care should be
taken to ensure that suitable levels of 'unit' loading are applied such that the derived
stiffness is applicable to the analysis to be undertaken.
C5.6 LOAD APPLICATION
Section 5.7 of the PRACTICE indicates appropriate methods for applying the various
loads to the analytical model(s). The importance of capturing the distributed nature of
the self weight and distributed loadings is emphasized.
C5.7 EVALUATION OF FORCES
This Commentary is directly applicable for the structural modeling of jack-ups for
either quasi-static or dynamic analysis. Key points which impact the final conclusions
drawn from the assessments of jack-ups have been emphasized to complement the
guidance given in the PRACTICE. A successful analysis will conclude with a set of
forces which can then be used in the final evaluation of the adequacy of the unit the
specific site, as contained in Section 8.0 - Assessment Criteria. The specific areas of
review include:
• Structural Strength Check (see Section 8.1 of the PRACTICE).

• Overturning Stability (see Section 8.2 of the PRACTICE).
• Foundation Assessment (see Section 8.3 of the PRACTICE):
- Bearing Capacity
- Sliding Resistance.
• Other Responses of Interest (see Sections 8.4, 8.5 and 8.6 of the PRACTICE
respectively):
- Horizontal Deflection
- Holding Capacity of the Jacking System
- Hull Strength (if required).
Care should be taken to ensure that the appropriate load factor is reflected in the
determination of the load set applied during the analysis. This allows for direct use of
the resulting loads in the assessment formulation.
Figure C5.1 : Responses/reactions from first order analyses
Figure C5.2 : P-Δ and leg inclination effects

Figure C5.3 : Contribution of Second Order Effects to First Order Responses
Figure C5.4 : Representative leg-hull connection

Figure C5.5 : Leg-hull connection component combinations
Figure C5.6 : Guide clearances
NB: Additional backlash may arise due to slack in gear train, clearances between
floating elevating system and shockpads etc.
Figure C5.7 : Jacking system backlash

Figure C5.8 : Types of leg guide arrangement
Figure C5.9 : Unopposed and opposed pinion arrangements

A = axial area of equivalent beam.

As = effective shear area of equivalent beam.
E = Young's modulus.
F = applied axial end load.
G = shear modulus = E/2.6 for poissons ratio of 0.3.
I = second moment of area of equivalent beam.
Khh = horizontal leg-hull connection stiffness.
Krh = rotational leg-hull connection stiffness.
Kvh = vertical leg-hull connection stiffness.
L = length of cantilever (from rigid support to point of load application).
M = applied end moment.
P = applied end shear.
δ = lateral end deflection of cantilever at point of load application for detailed leg
model alone.
δC = lateral end deflection of cantilever at point of load application for detailed leg plus
leg-hull connection model.
Δ = axial deflection of cantilever at point of load application for detailed leg model
alone.
ΔC = axial deflection of cantilever at point of load application for detailed leg plus leg-
hull connection model.
θ = end slope of cantilever at point of load application for detailed leg model alone.
θC = end slope of cantilever at point of load application for detailed leg plus leg-hull
connection model.
APPENDIX C5.A - DERIVATION OF ALTERNATIVE GEOMETRIC STIFFNESS
FORMULATION FOR P-Δ EFFECTS
C5.A.1 SUMMARY
The method described below allows a simple procedure for incorporating P-Δ effects in
a jack-up structural analysis. The advantage of this simple procedure is the ability to
include such effects without the necessity to adopt the iterative procedures required by
other methods. This method is accurate in determining the global response parameters,
including hull displacement and base overturning moment. It is also accurate in
determining the leg moment below the lower guide (usually the most critical part of the
leg). In its simplest form the procedure will conservatively predict the shear in the legs
(by roughly 10%). However leg shear is rarely a controlling factor in structural
assessments; therefore this difference is insignificant.
C5.A.2 DESCRIPTION OF THE METHOD
It is assumed that a structural computer model of a jack-up is used to determine the

effects of applied loads. Typically, a 'static' wave load is applied to a structural model,
and resulting deflections, forces and moments are determined. Note, however, that the
method can be similarly applied to dynamic analysis. (The method may not be
permitted by some software packages which 'prohibit' the use of a negative spring
stiffness).
The incorporation of P-Δ effects in the structural analysis is accomplished by including

a correction term in the global stiffness matrix of the structure. When an analysis is
performed with the correction term included, the resulting deflections, etc. will include
P-Δ effects. Note that since the global stiffness matrix is modified before the analysis,
no subsequent changes to the matrix are required (i.e. no iterations are required in the
solution).
The correction term to the global stiffness matrix is determined by a simple hand
calculation:
The correction term is: -Pg/L
where;
Pg = Effective hull gravity load. This includes hull weight and weight of the legs
above the hull.
L = The distance from the spudcan point of rotation to the hull center of gravity.
This single (negative) value is then incorporated into the global stiffness matrix of the
jack-up structural model. This can be accomplished in various ways depending on the
software in use. Typically, an orthogonal pair of horizontal translational earthed spring
elements can be attached to a node representing the hull center of gravity, and the
negative value is entered for each of the spring constants. Some software packages
allow direct matrix manipulation.
The effect of the negative stiffness is to produce an additional overturning load at the
hull. The overturning moment produced by this lateral load about the base is equal to
the overturning moment caused by the vertical load (of hull and legs above the hull)
times the deflection of the hull. Thus, the effect of the translation of the vertical load is
incorporated as a lateral force couple.
C5.A.3 BASIS FOR THE METHOD
The P-Δ effect is a consideration of the displacement of the structure under the applied
loads. In its most general form, the solution considers the displacements of each
element of the structure under loading. This is typically called a 'large displacement'
solution. In this general procedure, the deflections of the structure are used to reform
the stiffness matrix, which is then used to recalculate the displacements. While this is
analytically correct, there is a requirement to resolve the stiffness matrix several times
for each loading condition.
If the overall structural displacements are not very large, approximate solutions may be
used. Typically, approximate solutions are valid if tanθ ≈ θ, where θ is the rotation of
the structure about its base. These approximate solutions are known as 'geometric
stiffness' solutions. The classical column moment magnification or 'Euler
amplification' term is an example. The simple method presented here is another
example. A comparison of these two methods is given in Section C5.A.4 and the
derivation is presented in Section C5.A.5.
The P-Δ effect for jack-up structures is manifested as a change in lateral stiffness of the
individual legs, given a change in the axial load in each leg. For jack-ups the change in
axial load in each leg is caused by the application of the gravity loading and
environmental loading. As shown in Section C5.A.5, the net effect on the P-Δ of the
axial load changes in each leg due to the environmental loading will cancel out. Thus,
for overall structural response, only the gravity load need be considered in the
calculation of P-Δ effects. The reduced stiffness will then affect the response to
environmental loadings.
C5.A.4 VERIFICATION AGAINST 'EXACT' SOLUTION
Verification of this simple procedure was made against an 'exact' solution. In this case,
the 'exact' solution was performed using analysis software which accounts for large
displacements. In this procedure, the displaced configuration of the structure is used to
update the stiffness matrix, and iteration is used to converge on a given solution.
Verification was performed using a jack-up structural model as shown in Figure

C5.A.1. The leg chord, horizontal and diagonal members are modeled as individual
elements. The hull (in this case) is assumed to act as a rigid body. The hull to leg
connection included leg clamping devices. These were modeled along with the leg
guides. For the purpose of this verification the detailing of the hull and leg/hull
connection is not important. The spudcans were modeled as 'pinned'.
Loading of the model was accomplished as shown on Figure C5.A.2. Loadings due to
wave and wind (and dynamic inertial) were considered separately to verify the behavior
under the two separate types of loading. The loading direction was towards the bow in
both cases. For each case, a vertical load was applied at the hull center of gravity. It is
interesting to note that the vertical load is necessary for solution using 'exact' large
displacement methods, but is not needed to obtain a solution using the simple method.
A summary of the comparison results is given on Tables C5.A.1 (wave load) and
C5.A.2 (wind load). Verification with these two load cases was done separately, since
the loading occurs on different parts of the structure. The level of loading is arbitrary.
Values assumed here are greater than used in the site assessment of this particular jack-
up.
Discussion of the individual response parameters from Tables C5.A.1 and C5.A.2 is
given below.
C5.A.4.1 Global Response Parameters
The fundamental response quantities of deck displacement and overturning moment

agree to within 1%. The base shear for the simple method is not correct since it
includes the additional (fictitious) lateral force applied to the hull. The difference
between the total applied force and the base shear is the additional lateral force applied
at the hull. In theory the moment due to the vertical load (P-Δ moment) should be
replaced by a lateral force couple, i.e. lateral loads at the hull and base. Reduction of
the base shear (in global axes) by this additional lateral force at the hull will equate the
global base shear with the applied load.
C5.A.4.2 Windward and Leeward leg parameters
The values for individual leg axial load and moment at the lower guide agree to within
1%. These quantities are the most critical parameters for structural assessments.
The distribution of global base shear among the individual legs is not as accurately
matched by the simple method. For each leg, the lateral stiffness is decreased by
increasing axial load. Thus, the distribution of global base shear will depend on the
axial load present in each leg. The simple method, since it lumps the effects of all legs
into one correction term, cannot accurately predict the shear re-distribution among the
legs.
This lack of re-distribution of global base shear loading is not generally important to a
structural assessment. The amount will depend on the level and type of loading (wave
or wind). For the two cases given, 1% and 5% of the total base shear load (in global
axes) is shifted from the leeward leg to the windward legs.
When the leg base shears are not corrected, the simple method conservatively over-
predicts the shear in the legs. Since shear force is not as critical as the leg bending
moment this conservatism is not very restrictive.
If a correction is desired, the added lateral load at the hull can be subtracted in equal
fractions from the leg spudcan reactions (in global axes). Note that, for the case of the
windward leg, this will slightly under-predict the 'correct' global shear reaction.
TABLE C5.A.1 - VERIFICATION OF SIMPLE PROCEDURE FOR P-Δ EFFECT WITH
EXACT SOLUTION - WAVE LOADING CASE
Simple Exact
No P-Δ Method Solution
Global response parameters
Hull displacement (inches) 21.6 24.5 24.7
Base OTM (Kip-ft x 103) 227. 253. 251.
Base shear (kips) 711. 780. 711.
(Added lateral load at hull, kips) (69.)
Windward leg parameters

Axial force (kips) 3638. 3524. 3534.
Shear at spudcan (kips) 250. 272. 254.
(Corrected by 69/3 = 23 kips) (249.)
Moment at lower guide (Kip-ft x 103) 48. 56. 56.
Leeward leg parameters

Axial force (kips) 5477. 5706. 5685.
Shear transferred from leeward leg to 0. 8.
windward legs due to P-Δ (kips)
TABLE C5.A.2 - VERIFICATION OF SIMPLE PROCEDURE FOR P-Δ EFFECT

WITH EXACT SOLUTION - WIND LOADING CASE
Simple Exact
No P-Δ Method Solution
Global response parameters
Hull displacement (inches) 44.9 51.0 51.4
3
Base OTM (Kip-ft x 10 ) 490. 545. 541.
Base shear (kips) 1055. 1198. 1055.
(Added lateral load at hull, kips) (143.)
Windward leg parameters

Axial force (kips) 2538. 2300. 2318.
Leeward leg parameters

Axial force (kips) 7803. 8279. 8242.
Shear transferred from leeward leg to 0. 47.
windward legs due to P-Δ (kips)
Figure C5.A.1 - Analysis model
Figure C5.A.2 - Load application

C5.A.5 DERIVATION OF THE SIMPLIFIED CORRECTION TERM
Beam Deflection Moment diagram

(due to secondary bending only)
Ref.: Salmon and Johnson, 'Steel Structures', p 620.
y0 = Deflection without axial load, P

y1 = Additional lateral deflection due to axial load, P
To calculate y1, take moment of M/EI diagram between support and midspan
P(y 0 + y1 ) ⎧ 2L ⎫⎧ 2L ⎫
y1 = ⎨ ⎬⎨ ⎬
EI ⎩ π ⎭⎩ π ⎭
Centroid of area under moment diagram
Area under moment diagram
Rearranging:
4PL2
y1 = (y0 + y1) 2
π EI
using:
π 2 EI
PE = (with k = 2)
( kL) 2
P
y1 = (y0 + y1)
PE
⎧ P / PE ⎫
y1 = y0 ⎨ ⎬
⎩1 − P / PE ⎭
Total lateral deflection: ymax = y0 + y1
⎧ P / PE ⎫
ymax = y0 + y0 ⎨ ⎬
⎩1 − P / PE ⎭
⎧ 1 ⎫
= y0 ⎨ ⎬
⎩1 − P / PE ⎭
'Euler amplification' term
Determine effect on stiffness
H
Define Ko = Lateral stiffness without axial load, P
y0
3EI
=
L3
H
Define K1 = Lateral stiffness with axial load, P
y max
⎧
H P⎫
= ⎨1 − ⎬
⎩ PE ⎭
y0
⎧ ⎡ 4L2 ⎤ ⎫
= Ko ⎨1 − P ⎢ 2 ⎥ ⎬
⎩ ⎣ π EI ⎦ ⎭
⎧ 3EI ⎫⎧ 4L ⎫
2
= Ko - P ⎨ 3 ⎬⎨ 2 ⎬
⎩ L ⎭⎩ π EI ⎭
⎧ 12 P ⎫
= Ko - ⎨ 2 * ⎬
⎩π L⎭
12P
Conclusion: Effective lateral stiffness reduced by -
π 2L
P = axial load in one leg
Note: This is based on assuming a sine curve for deflection.

Repeat calculations assuming a linear deflection (rather than a sine curve):
P( y 0 + y1 ) ⎧ L ⎫⎧ 2L ⎫
y1 = ⎨ ⎬⎨ ⎬
EI ⎩ 2 ⎭⎩ 3 ⎭
Centroid of area under moment diagram
Area under moment diagram
Rearranging:
PL2
y1 = (y0 + y1)
3EI
⎧ PL2 / 3EI ⎫
y1 = y0 ⎨ ⎬
⎩1 − PL / 3EI ⎭
2
ymax = y0 + y1
ymax = y0 ⎧⎨ ⎫
1
⎬
⎩1 − PL / 3EI ⎭
2
Amplification for linear assumption
Determine effect on stiffness

H
Define Ko = Lateral stiffness without axial load, P
y0
3EI
=
L3
H
Define K1 = Lateral stiffness with axial load, P
y max
H ⎧ PL2 ⎫
= ⎨1 − ⎬
y 0 ⎩ 3EI ⎭
⎧ ⎡ PL2 ⎤ ⎫
= Ko ⎨1 − ⎢ ⎥⎬
⎩ ⎣ 3EI ⎦ ⎭
P
= Ko -
L
Conclusion: Based on linear deflection assumption, lateral stiffness is reduced by -P/L

term. This is the usual approximation of geometric stiffness.
Effect on total jack-up stiffness:
KE = Ke1 + Ke2 + Ke3 sum of individual leg stiffnesses (neglects hull rotation)
For each leg:

P = Pgravity + Penvironment
12 1 12 1
KE = [k01 - 2 (Pg1 + Pe1)] + [k02 - 2 (Pg2 + Pe2)] + …
π L π L
12 1 12 1
= (k01 + k02 + k03) - 2 (Pg1 + Pg2 + Pg3) - 2 (Pe1 + Pe2 + Pe3)
π L π L
Assume net vertical environmental load (Pe1 + Pe2 + Pe3) = 0

12 P
KE = (ΣK0i) - 2 G , where PG is total gravity load only
π L
If linear deflection assumption is made:

P
KE = (ΣK0i) - G
L
C6 COMMENTARY TO CALCULATION METHODS –
GEOTECHNICAL ENGINEERING
C6.1 INTRODUCTION
This Commentary is compiled to support Section 6 of the PRACTICE and should only
be used as a reference document in conjunction with the text of the PRACTICE.
A Glossary of Terms used for the Geotechnical Engineering Analysis is included at the
end of this Commentary.
C6.2 PREDICTION OF FOOTING PENETRATION DURING PRELOADING
C6.2.1 Analysis Method for Leg Penetration Prediction
The equations in Sections 6.2.2 and 6.2.3 may be applied for estimating the penetration
of the spudcan during preloading. In this case the backflow and contribution of the
spudcan buoyancy (due to the weight of soil replaced by the spudcan) will have a direct
effect on the penetration depth. Hence the effect of backflow and spudcan buoyancy
should be included in the calculations.
Predictions of spudcan penetration are based on a direct application of conventional

bearing capacity formulae for shallow circular flat foundations. However, the analysis
methods for shallow foundations and spudcan penetration predictions are
fundamentally different as illustrated in Figure C6.1. Conventional foundation
analyses for a circular footing at depth D firstly comprise the determination of the
ultimate bearing capacity, FV, at this depth and subsequently computing the vertical
displacement of this footing, zu, which is required to mobilize this resistance. Thus the
analyses consist of a strength analysis followed by a deformation analysis.
Figure C6.1 : Comparison of bearing capacity analytical procedures

for shallow foundations and jack-ups
In a spudcan penetration analysis the deformation at ultimate resistance (i.e. the

spudcan penetration D) is taken as an input and from this the associated soil resistance
is directly computed. The conventional analysis consist of only one step using the
same bearing capacity criteria as for the former shallow foundation analyses. The
incompatibility between these approaches is generally accounted for by the application
of empirical corrections to classical bearing capacity formulations.
These empirical corrections are generally introduced by the selection of appropriate soil
strength parameters. It is also noted that these corrections account for other significant
differences between conventional shallow foundations and spudcans, such as:
1. Spudcans are relatively smooth (steel) and (semi) conical whereas the other
footings are usually rough (concrete) and flat.
2. Spudcan foundations stress soil which, during installation, has been subjected to
large strains, whereas conventional foundations are placed on soil which has not
failed. Also for conventional foundations the soil may have improved due to
"pre-design" foundation loads causing increased strength by consolidation.
3. Spudcans are an order of magnitude larger than most conventional foundations.
For the conservative evaluation of the hole backflow the PRACTICE recommends that
the stability factors of Meyerhoff [1] are used. However, for normally consolidated
clay profiles the Britto and Kusakabe [2] curve may be more appropriate (see Figure
C6.2).
It should be noted that the expression in Section 6.2.1 is based on static hole stability.
In reality, during penetration of the spudcan the soil will probably flow along the
spudcan upwards on to the top of the spudcan. Hence, the hole stability derived from
the expression provided in Section 6.2.1 may be too optimistic.
Figure C6.2: Stability factors for cylindrical excavations in clay

C6.2.2 Penetration Analysis For Clays
For the selection of undrained shear strength, cu, it is recommended that the mean value
to a depth of half a spudcan diameter beneath the level where the maximum spudcan
diameter is in contact with the soil is used (Young [3]).
This method is applicable if the shear strength values up to one diameter below the
spudcan do not vary more than 50 percent from the average value (after Skempton [4]).
If significant cu variations occur, then the bearing capacity should be computed using a
method for layered soil conditions.
Analytical solutions are available for computing the bearing capacity of footings on
clay with increasing shear strength with depth (Davis and Booker, [5]); Salencon and
Matar, [6]; Houlsby and Wroth, [7]). These methods give bearing capacities less than
those resulting from the use of Skempton's [4] and Vesic [8] relationships. Empirical
correction factors for the Skempton [4] and the Davis and Booker [5] methods are
recommended by Endley et al [9].
However these empirical methods take no account of the spudcan equivalent cone
angle, the spudcan roughness factor or the depth of spudcan embedment of the
uppermost part of the bearing area below the soil surface.
An alternative bearing capacity factor Nc' has been developed which takes these factors,
and that of increased shear strength with depth, into account. In this case the ultimate
bearing capacity of a spudcan in clay can be expressed by:
FV = {cuoNc' + po'} A
The maximum preload, VLo, is equal to the ultimate vertical bearing capacity, FV,
taking into account the effect of backflow, Fo'A, and the effective weight of the soil
replaced by the spudcan, γ'V i.e.:
noting that the terms -F'oA + γ'V should always be considered together.
Table C6 provides values for Nc' which is an alternative dimensionless bearing capacity
factor dependent on:
(a) The equivalent cone angle of the spudcan β. For spudcans with multiple cone
angles, the equivalent cone has a base equal to the base of the largest
component and a volume equal to the total volume of the components,
(b) The roughness factor for the spudcan surface α,
(c) The depth of embedment of the uppermost part of the bearing area below the
soil surface D,
(d) The rate of increase of the undrained shear strength with depth below the
spudcan "ρ" (see Figure C6.3).
The alternative non-dimensional bearing capacity factors "Nc'" shown in Tables C6.1 to
C6.6 have been derived using a computer program which is able to calculate lower
bound (conservative) collapse loads for both axisymmetric and plane strain
foundations. Vertical bearing capacity has been computed for all combinations of the
following parameters:
Figure C6.3: Conical footing bearing capacity - problem definition and

notation
Cone angle, β = 30°, 60°, 90°, 120°, 150°, 180°.
Footing embedment depth, D/R = 0.0, 0.2, 0.5, 1.0, 2.0, 5.0.
Roughness factor, α = 0.0, 0.2, 0.4, 0.6, 0.8, 1.0

(where α = 0.0 for fully smooth and 1.0 for fully rough)
Rate of increase of clay shear strength with depth,

ρ2R/cum = 0.0, 1.0, 2.0, 3.0, 4.0, 5.0.
The effects of the depth of embedment and the rate of increase of shear strength with
depth are expressed by use of the dimensionless factors D/2R and ρ2R/cum, where cum is
the undrained strength at the soil surface (equal to cuo - Dρ assuming a linear variation
of strength with depth) and R is the radius of the spudcan.
Values of the dimensionless factor Nc' are given in Tables C6.1 to C6.6 for the range:
β = 30° to 180° ; α = 0.0 to 1.0
D/2R = 0.0 to 2.5 ; ρ2R/cum = 0.0 to 5.0
The factors are calculated assuming a linear variation of undrained strength with depth.
The best fit to the profile of undrained strength between the depth of the lowermost
point of the maximum bearing area and one radius below that point should be used in
deriving the value of ρ.
In the model a field of slip lines is formed between the footing and the horizontal free
soil surface. This type of "general shear" failure mechanism is appropriate for the
shallow footing penetrations being considered. At larger embedments, however, the
slip lines do not propagate to the surface as the "local shear" failure mechanism
becomes critical (i.e. it gives a lower bearing capacity). A transition from general to
local shear failure may be predicted at footing embedments between 6R and 8R.
The greatest embedment considered in the tables is 5R, so the general shear bearing
capacity factors are still appropriate for this depth. Therefore Table C6.6 should not be
extrapolated to higher values of D/2R, since the bearing capacity factor does not
increase significantly with further embedment. Values for D/2R = 2.5 should be used
for large embedments.
Although a footing may be fully rough (α = 1.0), full adhesion is only mobilized at the
cone surface when β ≤ 90°. For cone angles greater than 90°, only partial friction is
mobilized. In general, if the roughness factor is α, full friction is mobilized only when
β < π - sin-1α. This relationship may be derived using Mohr's Circle.
For the selection of appropriate roughness factors the results of mathematical models
(Noble Denton [10a]) suggest that the presence of a sharp secondary cone, forming the
tip of spudcan, tends to cause "rough" behavior. Rough blunt spudcans behave in a
similar manner to flat circular plates but more pointed spudcans behave as neither fully
rough nor fully smooth and have intermediate roughness factors of between α = 0.3 to
0.5. In the absence of detailed information, and as an approximation, a value of α = 0.4
may be appropriate for typical "double cone" spudcan shapes.
For further information regarding this alternative method for bearing capacity analysis
in clay reference should be made to Houlsby [11, 12, 13], Koumoto [14] and Houlsby
[15].
It is noted that footing penetration predictions are generally made using shear strength
data from simple laboratory tests such as torvane, pocket penetrometer, motorvane
and/or unconfined compression tests. Strength values from such tests are generally
lower than those of higher quality in-situ or laboratory tests, particularly if samples for
the latter tests are obtained from push/piston sampling rather than the percussion
sampling method.
It is likely that the former testing methods may yield low bearing capacity values for
very sensitive clays and/or strongly strain softening clays. Engineering judgment is
required in such cases to assess the likely footing penetration.
It is noted that in some clays, following remolding during spudcan penetration, the
shear strength may increase over a short time period. For certain clays the strength
may be regained in a matter of hours. In such cases, a crust of stronger material may
develop underneath the spudcan and this crust may then be underlain by weaker clay.
In this condition a potential punch-through situation could occur during subsequent
reloading. Several actual failures have been attributed to this type of soil behavior
(Young et al., [3]). For soils where this type of strength hardening (thixotropy) is
possible caution should be exercised as interruptions during the preloading operations
could lead to severe consequences.
For the conservative assessment of the effects of cyclic loading on clay foundations the
following vertical bearing capacity reduction factors may be applied to the capacities
calculated from static soil properties (Andersen [16]):
Leeward leg vertical bearing capacity reduction factor = 1.0
Windward leg vertical bearing capacity reduction factor = 0.8

Commentaries to Recommended Practice for Site Specific Assessment of
TABLE C6.1 TABLE C6.2
30 degrees cones FV/(Acuo) factors 60 degrees cones FV/(Acuo) factors
ρ2R D cuo ρ2R Roughness ρ2R D cuo ρ2R Roughness
--- - --- --- --- - --- ---
cum R cum cuo 0.0 0.2 0.4 0.6 0.8 1.0 cum R cum cuo 0.0 0.2 0.4 0.6 0.8 1.0
0.0 0.0 1.00 0.00 4.61 5.51 6.38 7.22 8.03 8.78 0.0 0.0 1.00 0.00 4.45 4.96 5.45 5.90 6.32 6.69
0.0 0.2 1.00 0.00 4.80 5.70 6.56 7.40 8.20 8.95 0.0 0.2 1.00 0.00 4.68 5.19 5.67 6.12 6.53 6.90
0.0 0.5 1.00 0.00 5.05 5.94 6.80 7.63 8.43 9.18 0.0 0.5 1.00 0.00 4.98 5.50 5.96 6.40 6.81 7.18
0.0 1.0 1.00 0.00 5.41 6.29 7.14 7.79 8.76 9.50 0.0 1.0 1.00 0.00 5.41 5.90 6.37 6.81 7.21 7.57
0.0 2.0 1.00 0.00 5.98 6.85 7.70 8.51 9.29 10.03 0.0 2.0 1.00 0.00 6.07 6.55 7.01 7.43 7.84 8.18
0.0 5.0 1.00 0.00 7.12 7.98 8.81 9.61 10.38 11.10 0.0 5.0 1.00 0.00 7.33 7.81 8.25 8.66 9.05 9.39
1.0 0.0 1.00 1.00 7.53 9.02 10.46 11.84 13.19 14.46 1.0 0.0 1.00 1.00 5.81 6.51 7.15 7.77 8.34 8.87
1.0 0.2 1.10 0.91 7.45 8.89 10.27 11.61 12.90 14.13 1.0 0.2 1.10 0.91 5.92 6.59 7.23 7.83 8.38 8.89
1.0 0.5 1.25 0.80 7.38 8.73 10.05 11.32 12.55 13.72 1.0 0.5 1.25 0.80 6.04 6.70 7.30 7.88 8.42 8.91
1.0 1.0 1.50 0.67 7.28 8.55 9.78 10.98 12.14 13.24 1.0 1.0 1.50 0.67 6.20 6.84 7.41 7.96 8.47 8.94
1.0 2.0 2.00 0.50 7.20 8.38 9.51 10.61 11.68 12.68 1.0 2.0 2.00 0.50 6.43 7.05 7.58 8.12 8.59 9.03
1.0 5.0 3.50 0.29 7.34 8.39 9.39 10.37 11.30 12.19 1.0 5.0 3.50 0.29 6.97 7.55 8.08 8.54 8.98 9.39
2.0 0.0 1.00 2.00 10.45 12.51 14.51 16.44 18.31 20.10 2.0 0.0 1.00 2.00 7.14 8.02 8.84 9.60 10.32 10.99
2.0 0.2 1.20 1.67 9.65 11.53 13.33 15.08 16.79 18.40 2.0 0.2 1.20 1.67 6.92 7.73 8.49 9.21 9.88 10.50
2.0 0.5 1.50 1.33 8.89 10.58 12.19 13.76 15.27 16.72 2.0 0.5 1.50 1.33 6.74 7.50 8.18 8.84 9.46 10.03
2.0 1.0 2.00 1.00 8.20 9.67 11.09 12.47 13.80 15.08 2.0 1.0 2.00 1.00 6.59 7.29 7.91 8.53 9.09 9.61
2.0 2.0 3.00 0.67 7.60 8.87 10.10 11.30 12.45 13.54 2.0 2.0 3.00 0.67 6.55 7.20 7.76 8.33 8.83 9.30
2.0 5.0 6.00 0.33 7.37 8.44 9.48 10.18 11.44 12.35 2.0 5.0 6.00 0.33 6.99 7.49 8.03 8.50 8.95 9.37
3.0 0.0 1.00 3.00 13.36 15.98 18.56 21.03 23.42 25.71 3.0 0.0 1.00 3.00 8.49 9.54 10.50 11.42 12.29 13.10
3.0 0.2 1.30 2.31 11.51 13.76 15.92 18.02 20.05 22.00 3.0 0.2 1.30 2.31 7.77 8.70 9.56 10.38 11.14 11.85
3.0 0.5 1.75 1.71 9.98 11.89 13.72 15.49 17.21 18.85 3.0 0.5 1.75 1.71 7.24 8.03 8.80 9.53 10.20 10.82
3.0 1.0 2.50 1.20 8.74 10.33 11.87 13.36 14.80 16.18 3.0 1.0 2.50 1.20 6.82 7.56 8.21 8.86 9.45 10.00
3.0 2.0 4.00 0.75 7.79 9.11 10.39 11.64 12.83 13.98 3.0 2.0 4.00 0.75 6.60 7.27 7.85 8.44 8.94 9.43
3.0 5.0 8.50 0.35 7.40 8.46 9.51 10.52 11.49 12.42 3.0 5.0 8.50 0.35 6.99 7.47 8.01 8.49 8.94 9.36
4.0 0.0 1.00 4.00 16.27 19.46 22.57 25.62 28.52 31.32 4.0 0.0 1.00 4.00 9.83 11.02 12.16 13.24 14.26 15.18
4.0 0.2 1.40 2.86 13.10 15.68 18.14 20.54 22.86 25.08 4.0 0.2 1.40 2.86 8.51 9.52 10.48 11.38 12.22 13.00
4.0 0.5 2.00 2.00 10.83 12.87 14.86 16.79 18.66 20.44 4.0 0.5 2.00 2.00 7.61 8.44 9.26 10.04 10.75 11.41
4.0 1.0 3.00 1.33 9.11 10.77 12.38 13.96 15.47 16.91 4.0 1.0 3.00 1.33 6.97 7.74 8.41 9.08 9.69 10.26
4.0 2.0 5.00 0.80 7.91 9.26 10.57 11.84 13.06 14.23 4.0 2.0 5.00 0.80 6.64 7.31 7.90 8.49 9.01 9.51
4.0 5.0 11.00 0.36 7.40 8.47 9.52 10.54 11.52 12.46 4.0 5.0 11.00 0.36 6.86 7.45 8.00 8.48 8.94 9.35
Rev 3, August 2008

5.0 0.0 1.00 5.00 19.18 22.94 26.61 30.20 33.63 36.92 5.0 0.0 1.00 5.00 11.17 12.52 13.83 15.06 16.20 17.26
Page 102
5.0 0.2 1.50 3.33 14.48 17.33 20.06 22.72 25.29 27.75 5.0 0.2 1.50 3.33 9.14 10.23 11.26 12.25 13.15 13.99
5.0 0.5 2.25 2.22 11.46 13.64 15.75 17.80 19.78 21.68 5.0 0.5 2.25 2.22 7.90 8.78 9.63 10.43 11.17 11.87
5.0 1.0 3.50 1.43 9.37 11.08 12.77 14.38 15.94 17.43 5.0 1.0 3.50 1.43 7.08 7.84 8.55 9.24 9.86 10.45
5.0 2.0 6.00 0.83 7.98 9.35 10.68 11.97 13.21 14.40 5.0 2.0 6.00 0.83 6.66 7.32 7.94 8.53 9.06 9.56
5.0 5.0 13.50 0.37 7.40 8.47 9.53 10.56 11.55 12.48 5.0 5.0 13.50 0.37 6.85 7.44 7.99 8.47 8.93 9.35
90 degrees cones FV/(Acuo) factors 120 degrees cones FV/(Acuo) factors
--- - --- --- --- - --- ---
0.0 0.0 1.00 0.00 4.64 5.02 5.36 5.67 5.95 6.17 0.0 0.0 1.00 0.00 4.96 5.25 5.51 5.73 5.92 6.05
0.0 0.2 1.00 0.00 4.90 5.28 5.61 5.91 6.18 6.41 0.0 0.2 1.00 0.00 5.23 5.52 5.77 5.99 6.17 6.30
0.0 0.5 1.00 0.00 5.22 5.59 5.93 6.23 6.49 6.71 0.0 0.5 1.00 0.00 5.57 5.85 6.10 6.31 6.49 6.62
0.0 1.0 1.00 0.00 5.68 6.03 6.36 6.66 6.92 7.14 0.0 1.0 1.00 0.00 6.04 6.31 6.55 6.76 6.93 7.05
0.0 2.0 1.00 0.00 6.37 6.71 7.05 7.32 7.58 7.79 0.0 2.0 1.00 0.00 6.74 7.01 7.24 7.44 7.61 7.72
0.0 5.0 1.00 0.00 7.65 8.03 8.32 8.60 8.86 9.05 0.0 5.0 1.00 0.00 8.07 8.32 8.55 8.75 8.90 8.99
1.0 0.0 1.00 1.00 5.57 6.05 6.47 6.87 7.22 7.53 1.0 0.0 1.00 1.00 5.69 6.04 6.36 6.65 6.89 7.09
1.0 0.2 1.10 0.91 5.74 6.21 6.62 7.00 7.36 7.65 1.0 0.2 1.10 0.91 5.89 6.24 6.55 6.82 7.07 7.26
1.0 0.5 1.25 0.80 5.94 6.38 6.79 7.16 7.50 7.79 1.0 0.5 1.25 0.80 6.12 6.45 6.76 7.02 7.26 7.45
1.0 1.0 1.50 0.67 6.16 6.61 6.99 7.36 7.68 7.97 1.0 1.0 1.50 0.67 6.39 6.72 7.01 7.27 7.48 7.66
1.0 2.0 2.00 0.50 6.50 6.93 7.30 7.64 7.95 8.21 1.0 2.0 2.00 0.50 6.80 7.10 7.37 7.61 7.82 7.97
1.0 5.0 3.50 0.29 7.25 7.57 7.94 8.25 8.53 8.78 1.0 5.0 3.50 0.29 7.52 7.82 8.08 8.29 8.49 8.61
2.0 0.0 1.00 2.00 6.46 7.03 7.54 8.01 8.45 8.82 2.0 0.0 1.00 2.00 6.38 6.79 7.16 7.50 7.80 8.04
2.0 0.2 1.20 1.67 6.41 6.94 7.43 7.88 8.28 8.65 2.0 0.2 1.20 1.67 6.41 6.80 7.16 7.47 7.75 7.97
2.0 0.5 1.50 1.33 6.41 6.88 7.35 7.76 8.14 8.46 2.0 0.5 1.50 1.33 6.46 6.83 7.17 7.46 7.72 7.94
2.0 1.0 2.00 1.00 6.40 6.88 7.29 7.69 8.03 8.35 2.0 1.0 2.00 1.00 6.56 6.91 7.22 7.49 7.74 7.92
2.0 2.0 3.00 0.67 6.54 6.99 7.37 7.73 8.06 8.33 2.0 2.0 3.00 0.67 6.80 7.12 7.40 7.65 7.87 8.03
2.0 5.0 6.00 0.33 7.16 7.49 7.86 8.18 8.47 8.72 2.0 5.0 6.00 0.33 7.43 7.72 7.99 8.21 8.41 8.53
3.0 0.0 1.00 3.00 7.36 8.00 8.59 9.14 9.65 10.08 3.0 0.0 1.00 3.00 7.04 7.51 7.93 8.31 8.66 8.93
3.0 0.2 1.30 2.31 6.99 7.57 8.10 8.60 9.05 9.45 3.0 0.2 1.30 2.31 6.84 7.27 7.65 8.00 8.31 8.57
3.0 0.5 1.75 1.71 6.70 7.24 7.73 8.17 8.59 8.94 3.0 0.5 1.75 1.71 6.71 7.09 7.45 7.76 8.05 8.27
3.0 1.0 2.50 1.20 6.54 7.04 7.47 7.88 8.24 8.57 3.0 1.0 2.50 1.20 6.66 7.02 7.34 7.63 7.88 8.08
3.0 2.0 4.00 0.75 6.56 7.02 7.41 7.78 8.11 8.39 3.0 2.0 4.00 0.75 6.81 7.11 7.41 7.67 7.89 8.06
3.0 5.0 8.50 0.35 7.12 7.46 7.83 8.15 8.44 8.46 3.0 5.0 8.50 0.35 7.38 7.68 7.95 8.17 8.38 8.51
4.0 0.0 1.00 4.00 8.22 8.96 9.64 10.25 10.82 11.33 4.0 0.0 1.00 4.00 7.70 8.22 8.69 9.11 9.49 9.81
4.0 0.2 1.40 2.86 7.49 8.11 8.68 9.22 9.70 10.14 4.0 0.2 1.40 2.86 7.20 7.66 8.07 8.44 8.77 9.03
4.0 0.5 2.00 2.00 6.94 7.50 8.01 8.48 8.92 9.29 4.0 0.5 2.00 2.00 6.88 7.28 7.65 7.98 8.27 8.53
4.0 1.0 3.00 1.33 6.63 7.15 7.58 8.01 8.38 8.72 4.0 1.0 3.00 1.33 6.72 7.08 7.42 7.71 7.97 8.18
4.0 2.0 5.00 0.80 6.57 7.03 7.43 7.80 8.14 8.42 4.0 2.0 5.00 0.80 6.80 7.12 7.41 7.68 7.90 8.08
4.0 5.0 11.00 0.36 7.05 7.44 7.81 8.13 8.42 8.67 4.0 5.0 11.00 0.36 7.39 7.66 7.93 8.15 8.36 8.49
Rev 3, August 2008

5.0 0.0 1.00 5.00 9.11 9.93 10.66 11.35 12.00 12.56 5.0 0.0 1.00 5.00 8.35 8.91 9.43 9.89 10.31 10.67
5.0 0.2 1.50 3.33 7.87 8.55 9.17 9.74 10.26 10.75 5.0 0.2 1.50 3.33 7.52 7.99 8.43 8.82 9.18 9.95
5.0 0.5 2.25 2.22 7.12 7.71 8.24 8.72 9.17 9.57 5.0 0.5 2.25 2.22 7.01 7.43 7.81 8.15 8.45 8.72
5.0 1.0 3.50 1.43 6.70 7.22 7.67 8.09 8.47 8.82 5.0 1.0 3.50 1.43 6.77 7.13 7.47 7.77 8.03 8.25
Page 103
5.0 2.0 6.00 0.83 6.57 7.04 7.44 7.82 8.16 8.44 5.0 2.0 6.00 0.83 6.80 7.12 7.42 7.69 7.91 8.09
5.0 5.0 13.50 0.37 7.03 7.42 7.80 8.12 8.41 8.66 5.0 5.0 13.50 0.37 7.34 7.64 7.91 8.14 8.34 8.48
150 degrees cones FV/(Acuo) factors 180 degrees cones FV(Acuo) factors
--- - --- --- --- - --- ---
0.0 0.0 1.00 0.00 5.32 5.55 5.74 5.89 6.01 6.05 0.0 0.0 1.00 0.00 5.69 5.86 5.97 6.03 6.05 6.05
0.0 0.2 1.00 0.00 5.60 5.82 6.00 6.16 6.26 6.30 0.0 0.2 1.00 0.00 5.97 6.13 6.24 6.29 6.30 6.30
0.0 0.5 1.00 0.00 5.94 6.16 6.34 6.49 6.59 6.61 0.0 0.5 1.00 0.00 6.31 6.47 6.57 6.61 6.61 6.61
0.0 1.0 1.00 0.00 6.41 6.62 6.80 6.94 7.03 7.05 0.0 1.0 1.00 0.00 6.79 6.93 7.02 7.05 7.05 7.05
0.0 2.0 1.00 0.00 7.13 7.32 7.49 7.62 7.71 7.72 0.0 2.0 1.00 0.00 7.49 7.63 7.70 7.71 7.71 7.71
0.0 5.0 1.00 0.00 8.46 8.65 8.81 8.93 8.99 8.99 0.0 5.0 1.00 0.00 8.82 8.94 8.99 8.99 8.99 8.99
1.0 0.0 1.00 1.00 5.94 6.22 6.46 6.67 6.84 6.97 1.0 0.0 1.00 1.00 6.25 6.47 6.65 6.79 6.90 6.95
1.0 0.2 1.10 0.91 6.16 6.43 6.67 6.87 7.04 7.15 1.0 0.2 1.10 0.91 6.48 6.69 6.87 7.00 7.10 7.14
1.0 0.5 1.25 0.80 6.41 6.67 6.90 7.09 7.25 7.36 1.0 0.5 1.25 0.80 6.74 6.94 7.11 7.23 7.32 7.35
1.0 1.0 1.50 0.67 6.71 6.96 7.18 7.36 7.51 7.60 1.0 1.0 1.50 0.67 7.05 7.24 7.39 7.51 7.58 7.60
1.0 2.0 2.00 0.50 7.13 7.36 7.57 7.73 7.86 7.95 1.0 2.0 2.00 0.50 7.47 7.64 7.79 7.88 7.93 7.94
1.0 5.0 3.50 0.29 7.91 8.12 8.31 8.44 8.56 8.61 1.0 5.0 3.50 0.29 8.26 8.32 8.52 8.60 8.61 8.61
2.0 0.0 1.00 2.00 6.50 6.82 7.11 7.35 7.57 7.73 2.0 0.0 1.00 2.00 6.73 6.98 7.20 7.39 7.53 7.63
2.0 0.2 1.20 1.67 6.59 6.90 7.16 7.40 7.59 7.74 2.0 0.2 1.20 1.67 6.85 7.08 7.30 7.46 7.59 7.68
2.0 0.5 1.50 1.33 6.69 6.98 7.23 7.45 7.63 7.76 2.0 0.5 1.50 1.33 6.98 7.20 7.39 7.55 7.66 7.72
2.0 1.0 2.00 1.00 6.84 7.10 7.34 7.54 7.70 7.82 2.0 1.0 2.00 1.00 7.15 7.36 7.53 7.67 7.76 7.80
2.0 2.0 3.00 0.67 7.11 7.35 7.57 7.74 7.89 7.99 2.0 2.0 3.00 0.67 7.45 7.63 7.78 7.90 7.96 7.98
2.0 5.0 6.00 0.33 7.81 8.01 8.21 8.35 8.47 8.53 2.0 5.0 6.00 0.33 8.16 8.27 8.43 8.50 8.53 8.53
3.0 0.0 1.00 3.00 7.03 7.40 7.72 7.98 8.24 8.43 3.0 0.0 1.00 3.00 7.16 7.45 7.69 7.91 8.08 8.21
3.0 0.2 1.30 2.31 6.94 7.27 7.56 7.81 8.03 8.21 3.0 0.2 1.30 2.31 7.13 7.40 7.62 7.81 7.96 8.07
3.0 0.5 1.75 1.71 6.88 7.18 7.45 7.68 7.88 8.03 3.0 0.5 1.75 1.71 7.15 7.37 7.58 7.75 7.88 7.96
3.0 1.0 2.50 1.20 6.91 7.18 7.43 7.63 7.81 7.94 3.0 1.0 2.50 1.20 7.21 7.42 7.61 7.75 7.86 7.91
3.0 2.0 4.00 0.75 7.10 7.35 7.57 7.75 7.90 8.00 3.0 2.0 4.00 0.75 7.43 7.62 7.78 7.90 7.97 7.99
3.0 5.0 8.50 0.35 7.76 7.97 8.16 8.31 8.43 8.49 3.0 5.0 8.50 0.35 8.13 8.23 8.38 8.46 8.49 8.49
4.0 0.0 1.00 4.00 7.55 7.94 8.30 8.58 8.88 9.10 4.0 0.0 1.00 4.00 7.56 7.87 8.15 8.38 8.58 8.73
4.0 0.2 1.40 2.86 7.23 7.58 7.89 8.16 8.40 8.59 4.0 0.2 1.40 2.86 7.38 7.64 7.89 8.09 8.26 8.39
4.0 0.5 2.00 2.00 7.02 7.34 7.62 7.86 8.07 8.23 4.0 0.5 2.00 2.00 7.26 7.50 7.71 7.89 8.03 8.13
4.0 1.0 3.00 1.33 6.95 7.23 7.49 7.70 7.88 8.01 4.0 1.0 3.00 1.33 7.25 7.46 7.65 7.80 7.92 7.98
4.0 2.0 5.00 0.80 7.09 7.34 7.56 7.75 7.90 8.00 4.0 2.0 5.00 0.80 7.44 7.61 7.77 7.89 7.97 8.00
4.0 5.0 11.00 0.36 7.72 7.94 8.13 8.29 8.41 8.47 4.0 5.0 11.00 0.36 8.09 8.19 8.36 8.44 8.47 8.47
Rev 3, August 2008

5.0 0.0 1.00 5.00 8.05 8.48 8.86 9.19 9.48 9.74 5.0 0.0 1.00 5.00 7.94 8.27 8.57 8.83 9.05 9.23
5.0 0.2 1.50 3.33 7.46 7.83 8.16 8.44 8.69 8.90 5.0 0.2 1.50 3.33 7.56 7.85 8.10 8.32 8.50 8.64
5.0 0.5 2.25 2.22 7.13 7.45 7.74 7.99 8.20 8.37 5.0 0.5 2.25 2.22 7.34 7.59 7.81 8.00 8.15 8.25
5.0 1.0 3.50 1.43 6.99 7.27 7.53 7.74 7.93 8.07 5.0 1.0 3.50 1.43 7.27 7.49 7.68 7.84 7.96 8.02
5.0 2.0 6.00 0.83 7.09 7.34 7.56 7.75 7.91 8.01 5.0 2.0 6.00 0.83 7.43 7.60 7.77 7.89 7.97 8.00
Page 104
5.0 5.0 13.50 0.37 7.70 7.93 8.12 8.27 8.40 8.46 5.0 5.0 13.50 0.37 8.07 8.18 8.35 8.43 8.46 8.46
C6.2.3 Penetration Analysis For Silica Sands
Various bearing capacity and shape factors are given in the literature for the analysis of
bearing capacity in silica sand. The method described in the PRACTICE text is that of
Vesic [8]. However, of particular relevance is the method proposed by Brinch Hansen
[17] which is also in common use particularly where the angle of internal friction has
been determined for plane strain conditions.
FV = A (0.5 γ' B Nγdγsγ + po' Nq sq dq)
where;
Nγ = 1.5(Nq - 1) tanφ Nq = eπtanφ tan2 (45° + φ/2)
dγ = 1 sq = 1 + B/L sinφ
sγ = (1 - 0.4B/L) dq = 1 + 2tanφ (1 - sinφ)2 D/B
and the maximum preload, VLo, is equal to the ultimate vertical bearing capacity, FV,
taking into account the effect of backflow, Fo'A, and the effective weight of the soil
replaced by the spudcan, γ'V, i.e.:
VLo = FV - Fo'A + γ'V
Empirical bearing capacity factors show reasonable agreement with model footings of
less than 2.0 metros diameter. However, significant disagreement has been observed in
centrifuge tests on larger size spudcans and for actual jack-up rig footings in the North
Sea for which the footing diameters ranged from 3.0 to 15.0 m, and laboratory triaxial φ
values for the sand indicated φ values in the range of 30° to 40°.
Observed penetrations were significantly larger than the φ values would indicate for
both the Vesic and Brinch Hansen methods. Analyses by various researchers (Graham,
[18]; James, [19]; Kimura, [20]) suggest that reduced φ values be used to account for
these "scale effects".
In view of these observations it is recommended that laboratory triaxial φ test values

should be reduced by 5° for the prediction of large diameter footing penetrations in
silica sands, i.e.
φdesign = φtriaxial - 5°
If laboratory test data are unavailable the following design φ values may be applicable:
DESIGN PARAMETERS FOR COHESIONLESS SILICEOUS SOIL*
Density Soil Description φ°des. Nγ Nq

Very Loose Sand 15 2.6 3.9
Loose Sand-Silt
Medium Silt
Loose Sand 20 5.4 6.4
Medium Sand-Silt
Dense Silt
Medium Sand 25 11 11
Dense Sand-Silt
Dense Sand 30 22 18
Very Dense Sand-Silt
Dense Gravel 35 48 33
Very Dense Sand
* See notes overleaf

* NOTE: "Sand-Silt" includes those soils with significant fractions of both sand and
silt. Strength values generally increase with increasing sand fractions. For
spudcans on sand the effects of cyclic loading may be to either increase or decrease
the pore water pressure. Positive excess pore water pressure will weaken the soil
and in severe cases may cause partial fluidization. Negative excess pore water
pressures may temporarily strengthen the soil. Approximate methods are available
for the assessment of excess pore water pressure development and associated
foundation settlement (Dean [20]).
The following qualifications apply:
1. Footing penetrations in a thick layer of clean silica sand are usually minimal with
the maximum diameter of the spudcan rarely coming into contact with the soil. It is
therefore not usual to consider the effects of soil backflow in this situation.
2. If various sand layers occur to 1.5 B below the footing depth an average value for φ'
can be selected using the graph developed by Meyerhof [22] as shown in Figure
C6.4. Considering the overall inaccuracy in the prediction of footing penetration in
sand, this refinement does not generally influence the accuracy of the prediction.
3. Emphasis should be placed on the identification and analyses of potential punch-

throughs into a clay layer (or a silt layer which may behave undrained).
Figure C6.4 : Depth of failure zone in sand

C6.2.4 Penetration Analysis For Carbonate Sands
Relatively large footing penetrations have been reported for uncemented carbonate
materials despite high laboratory friction angles (Dutt, [23]). This may be attributed to
either the high compressibility of these materials or low shear strengths due to high
voids ratio and a collapsible structure.
The leg penetration is governed by both strength and deformation characteristics of

foundation soils as noted in Commentary 6.2.1. The compressibility of carbonate sands
is relatively higher than for silica sands. Hence, greater penetrations should be
expected for carbonate sands relative to silica sands despite the similar or even higher
laboratory friction angles. This is supported by both experimental (Poulos, [24]) and
theoretical (Yeung, [25]) studies on model foundations.
The predictions of footing penetrations in carbonate sands are likely to be performed to

a lower degree of accuracy compared with those for silica sands. The conventional
method is to use the plasticity based formulation for bearing capacity of shallow
foundations in sand. However, friction angles to be used in the formulae should be
considerably smaller than laboratory values to account (in an artificial manner) for the
soil behavior.
C6.2.5 Penetration in Silts
Cyclic loads imposed in silty fine sands/silt foundations may cause liquefaction due to
the generation of excess pore water pressures. In this situation foundation settlements
would be anticipated. Conservative assessments of reduced bearing capacities and
increased settlements should be conducted as appropriate.
C6.2.6 Penetration Analysis For Layered Soils
C6.2.6.1 Squeezing of Clay
For a squeezing clay layer the resistance on the footing cannot become larger than the
resistance of the layer underneath the soft clay layer. Thus there is a limit to the
squeezing process.
C6.2.6.3 Punch-through: Dense Sand Over Soft Clay
Traditionally the bearing capacity of a footing on a thin sand layer overlying soft clay
has been computed according to the method developed by Hanna and Meyerhof [26] as
described in the PRACTICE. This method appears to provide reasonable predictions of
the ultimate resistance at the onset of punch-through, however it may overpredict the
resistance after the initiation of punch-through as the soil shear planes then approach
the vertical and the assumed modeling is then incorrect. The approximation Kstanφ ≈
3cu/Bγ' is a lower bound applicable to the onset of punch-through. The reference paper
provides more accurate data for Ks.
An alternative method is that in which a load spread is considered through the upper
sand layer, as illustrated in Figure C6.5. In this model the bearing capacity of the
foundation is assumed to be equal to the bearing capacity of the foundation projected
onto the lower layer. This method allows capacity ranges to be developed for a range
of load spread gradients. Thus for a load spreading under a slope of 1:n, the ultimate
bearing capacity of a circular foundation is given by:
FV = FV,b - W
where;
FV = ultimate bearing capacity of footing (with diameter B and area A at
depth D).
FV,b = ultimate bearing capacity of fictitious footing (with diameter
[B + 2H/n] and area (1 + 2H/nB)2A, at the top of the soft clay layer.
W = weight of soil "plug" in between footing and fictitious footing.
= [1 + 2H/(nB)]2AHγ'
Figure C6.5 : Spudcan bearing capacity analysis - sand over clay -

load spread method
This method, using n = 3, has been recommended by Young, [27] for jack-up
foundations. However, comparison with model test data (Jacobsen, [28]; Higham, [29];
Craig, [30]) suggest a range from n = 3 to 5. Conversely, actual spudcan penetration
data are available which suggest a higher spread, i.e. smaller n values, (Baglioni, [31]).
Hence it is suggested that this method would provide reasonable, but conservative,
results if a lower bound value n = 5 be used. However, it is noted that both
observations of model test data and results of numerical analyses reveal that soil
punching failure occurs along vertical surfaces. Thus, although this method can
provide reasonable quantitative estimates on leg penetration, it may not be based on a
physically correct model.
The same comment applies to the previously referenced Hanna and Meyerhof method
which is based on failure along a truncated cone surface. This is unlike the observed
vertical shear surface. However, the ultimate resistance computed when punching
shear is initiated generally gives reasonable agreement (and is acceptably conservative)
compared with actually observed data.
It is noted that this method generally provides reasonable estimates of ultimate soil
resistance at the onset of punch-through. However, significant underpredictions of soil
resistance have been reported by Baglioni, [31]. Conversely the method appears to
overpredict soil resistance after punch-through has initiated as suggested by Craig, [32].
Craig, [30] observed in centrifuge model tests that a sand plug underneath the spudcan
is pushed with relatively little lateral deformation into the underlying clay for prototype
offshore conditions. It is suggested that account be taken of the frictional resistance on
this sand plug when penetrating the clay.
C6.2.7 Summary
The various soil failure mechanisms considered in this section are illustrated in Figure
C6.6.
The current status on analytical methods for punch-through (Section C6.2.6) is that the
widely used methods, discussed above, (i.e. the load spread method and the Hanna and
Meyerhof method) show (different) discrepancies with observed data. However, both
methods allow prediction of a lower bound resistance at the initiation of punch-through.
Figure C6.6 : Foundation bearing failure modes

C6.3 FOUNDATION STABILITY ASSESSMENT
C6.3.3/4 Bearing capacity for inclined loading
Introduction
Formulae for combined vertical (FVH, FVHM), horizontal (FH, FHM) and moment (FM)
capacity on shallow foundations in uniform soils have been suggested by various
researchers. Those due to Vesic [8] and Brinch Hansen [17] are commonly used for
offshore applications and are included in API RP2A [33] and DNV [34] guidelines
respectively. It should be noted that the equations are applicable to shallow
penetrations. Any contribution of horizontal soil resistance on the embedded legs is
ignored. In case of deep penetrations the horizontal resistance may be significant,
especially when the jack-up leg comprises a single tubular or box section or when the
spudcan is provided with skirts around the can perimeter.
The following section is an overview of the recommended criteria for spudcan

foundations.
It is noted that the analytical procedures apply to a flat footing in which all
load/resistance is transmitted through its base. The influence of horizontal resistance
on vertical areas (e.g. footing cone or vertical surfaces above the base) can either be
assessed separately or ignored. Reasons for discluding the latter resistance component
are discussed below.
1. Uncertainty of the contact area between the spudcan and soil due to the shape of the
foundations or due to removal of foundation material by scour.
2. Soil strengths and stiffnesses may be significantly reduced as a result of material
remolding during unit installation.
3. The ultimate shear resistance along the foundation base is generally mobilized at a
significantly smaller displacement than that required to mobilize the passive soil
resistance.
It is noted that the combined loading problem is generally solved in a simplistic and
generally conservative manner. For layered soil conditions and/or complex footing
configurations it may be helpful to assess the ultimate foundation capacity using finite
element techniques. Such analyses are particularly relevant for the assessment of
displacements associated with combined loads (level 3 analyses).
The equations of Sections 6.3.3.1 and 6.3.3.2 may be used to calculate the bearing
capacity of the soil beneath the spudcan under inclined loading. The expressions
provide correlation between the vertical soil bearing capacity the associated horizontal
soil bearing capacity. In this case the effect of backflow and spudcan buoyancy shall
be included in the spudcan loading.
The total vertical load during storm conditions may comprise:

- dead, live and environmental loads including associated P-Δ effects
- backflow
- spudcan buoyancy (due to weight of soil displaced by the spudcan)
Uniform Clay
Clay pinned foundation (FM = 0):
A comparison of the Vesic [8] and Brinch Hansen [17] criteria for surface footings in
clay shows that the latter are slightly more conservative than the former (i.e. the Brinch
Hansen results show a reduced capacity compared with those given by the Vesic
approach). The Brinch Hansen criteria appear to provide a lower bound to finite
element analysis results and (model) test data reported by Noble Denton [10b] and by
Santa Maria [35].
The relationship between maximum vertical capacity (FVH) and horizontal soil capacity
(FH) for a circular surface footing in clay is graphically presented in Figure C6.7. The
graph has been non-dimensionalized by dividing FVH and FH by the maximum vertical
soil capacity FVmax (which occurs when FH = 0). Also shown is a curve for deep
footings where D/B≥2.5.
For spudcans founded on overconsolidated clay (OCR ≥ 4), cyclic degradation may
reduce the horizontal bearing capacity by a factor of 0.3, i.e. the horizontal bearing
capacity calculated from static soils properties should be multiplied by a reduction
factor:
Horizontal bearing capacity reduction factor = 0.7
For these materials the horizontal and vertical soil stiffnesses calculated from static soil
properties may be multiplied by factors of 1.25 and 3 to 8 as a result of cyclic effects
(Anderson [36]).
Figure C6.7 : Vertical/horizontal load envelopes for footings in clay

Clay foundation with moment fixity (FM > 0):
The FVH-FH envelope reduces in size if an overturning moment is applied in addition to

horizontal and/or vertical loading. Reference is made to Brinch Hansen [17] or to DNV
[34]. Some guidance can also be obtained from Santa Maria [35].
Uniform Sand
Pinned sand foundation (FM = 0):
Figure C6.8 shows a comparison of Vesic, [8] (=API RP2A, [33]) and Brinch Hansen,
[17] (= DNV, [34]) criteria for surface footings on sand with test results reported by
Noble Denton [10b]. These data suggest that the Vesic criteria provide a reasonable
lower bound to the test data for FH/FVH ratios less than 0.3. (It is noted that Tan [39]
reported tests results and analysis data which indicate higher soil resistances than those
due to Vesic and Brinch Hansen.)
The relations between FVH/FVmax and FH/FVmax for a circular surface footing and for a
deep footing in sand are graphically presented in Figure C6.9. The graph for surface
footings can be used to make a lower bound estimate of FVH-FH relations at any depth.
Based on the above studies it is recommended to adopt the Vesic criteria for spudcan
analysis in sands.
Sand foundation with moment fixity (FM > 0):

The FVH-FH envelope reduces in size if an overturning moment is simultaneously
applied. Reference is made to Vesic [8] and API RP2A [33] for details on the
computation procedure. Further guidance can be obtained from Tan [39].
Figure C6.8 : Foundation combined vertical/horizontal loading on sand

- comparison of design criteria and observed data
Figure C6.9 : Vertical/horizontal load envelopes for footings in sand

Clay & Sand Foundations with moment fixity (FM > 0)
Section 6.3.4.6 describes three methods by which fixity may be included in the
analysis. The intermediate method, using linear fixity, is an approach that was not
previously encompassed. It is now included as the more detailed methods are not
readily available to most analysts. It must however be noted that the linear rotational
stiffness must be selected with care to ensure that wave force cancellation effects do
not drive the resulting DAF’s. Refer to C7.4.
Elastic Spring Stiffnesses - Sand and Clay
The elastic stiffness factors are calculated assuming full contact of the spudcan with the
seabed. If the vertical load is insufficient to maintain full contact as the moment
increases then reduced stiffnesses should be used. The stiffness factors are derived for
a homogeneous, linear, isotropic soil. Choice of the appropriate shear modulus should
take into account the expected stress level and strain amplitude. In general, the shear
modulus decreases with increasing strain amplitude.
Selection of Shear Modulus, G, in clay
The value of the initial, small-strain shear modulus for clay should be based on the
value of the shear strength (cu) measured at the depth z = D + 0.15 B where B is the
diameter of the spudcan and D is the depth below mudline of the lowest point on the
spudcan at which this diameter is attained. Where the clay is significantly layered the
average strength within the range z = D to z = D + 0.3B should be used. Except in
areas with carbonate clays or clayey silts the shear modulus should be calculated as
[ref. 53]:
600
G = cu with G < CuIrNC and subject to the limitations given below.
OCR 0.25
Where;
OCR = The overconsolidation ratio;
IrNC = The Rigidity Index for Normally Consolidated clay..
For extreme loading conditions and in the absence of other data IrNC shall be
conservatively limited to 400 (Noble Denton, [53]) based on overconsolidated
clay sites with Plasticity Indices (Ip) of up to 40%; the data for normally
consolidated clay published by Andersen in Figure 10.2 of [55], reproduced as
Figure C6.10 below, supports the use of IrNC of 400 up to about Ip = 60% if, as
suggested by Andersen in correspondence, the low points at Ip of around 50%
are given less weight as they fall outside the main trend. Due consideration
should be given to the possibility of determining site-specific shear moduli for
normally consolidated and slightly overconsolidated clays and/or where the
Plasticity Indices exceed 60%.
IrNC = 600 for clays with low OCR and Ip less than about 40% is supported by
field data for jack-up response in the Gulf of Mexico (Templeton [54]).
It should be noted that IrNC appears to be fundamentally inversely proportional to
the Plasticity Index (Andersen [55], Figure 10.2, reproduced as Figure C6.10
below). For sites with Plasticity Indices in excess of 60%, and not covered by
field data, the analyst should account for the inverse relationship when
determining G.
In some cases higher ratios of IrNC may be used. The data published by Andersen
([55], Figure 10.2, reproduced as Figure C6.10 below) would support use of
values as high as 1000 or even 2500, particularly for Plasticity Indices less than
20%.
These recommendations are intended for use in both site assessments for extreme
loading and applications involving small strain beneath the spudcan. In the calculation
of fixity for extreme loading, the moment stiffness based on the small strain G values
will be degraded using the stiffness reduction formulae given in 6.3.4 of the Practice.
In the case of small strain applications, such as in structural fatigue analysis, the
stiffness reductions do not apply and it may be appropriate to adopt upper-bound
values.
3000
1.0 < OCR <1.5
OCR > 1.5. OCR indicated
at data point
1.9
3
2000
Gmax / suDSS
40
2.8
2.8
1000
2
4
10
25
40
0
0 20 40 60 80 100
Plasticity Index, Ip (%)
Figure C6.10 : Normalised initial shear modulus as a function of Plasticity Index, Ip, for 11
different clays. Figure 10.2 from Anderson [55]
Note: On the vertical axis Gmax, the initial shear modulus from shear waves generated by bender elements in
direct simple shear (DSS) tests is normalised against SuDSS, the undrained montonic shear strength from
the DSS tests. Whilst these parameters may differ from those determined by other means, due to rate
effects, etc., the differences are expected to be sufficiently small that the data and trends remain
applicable.
Note: If it is assumed that Poisson's ratio for clay is 0.5 then,

GB 3
K3 =
1.5
and, ignoring other terms and factors:
πB 2
V Lo = N c cu
4
where VLo is the effective seabed vertical reaction under preload.
This simplifies to:
⎛ 4 × 600 ⎞ 1
K 3 = VLo B⎜⎜ ⎟⎟
⎝ 1.5πN c ⎠ OCR
0.25
So that the rotational stiffness is directly proportional to the diameter, directly

proportional to the preload, and depends weakly on the OCR. The bracketed term is
almost a constant factor in the region of about 80. Since full embedment will usually
apply, neither preload nor diameter will vary very much for any one unit. In fact the
OCR is the only factor that alters the stiffness significantly.
Selection of Shear Modulus in Sand

For sands the initial, small-strain shear modulus should be computed from:
G p a = j (Vswl Ap a )
0.5
where
⎛ D ⎞
j = 230⎜ 0.9 + R ⎟ (Dimensionless stiffness factor)
⎝ 500 ⎠
p a = Atmospheric pressure
DR = Relative Density (percent)
Vswl = Seabed vertical reaction under still water conditions.
Note: The above gives

0.5
G ⎛ D ⎞⎛ 4V ⎞
= 230⎜ 0.9 + R ⎟⎜⎜ 2swl ⎟⎟
pa ⎝ 500 ⎠⎝ πB pa ⎠
Combining this with
GB 3
K3 =
3(1 − ν )
and
πB 2 γ ′B
V Lo = Nγ
4 2
(for a partially embedded foundation in sand, and also approximately true for the fully
embedded case if one ignores the N q term and depth and shape factors) gives:
⎛ D ⎞
2 × 230⎜ 0.9 + R ⎟ pa0.5Vswl
0. 5
K3 = ⎝ 500 ⎠
B2
3(1 − ν )π 0.5
and
8VL 0
B3 = .
πγ ′N γ
Two cases emerge. If there is (rarely) full embedment then the rotational stiffness is
proportional to the square of the diameter and the square root of the load - since for any
particular unit not much can be done about either, this results in almost constant
rotational stiffness in the embedded case.
In the partially embedded case we substitute for the diameter and get:
⎛ D ⎞
2 × 230⎜ 0.9 + R ⎟ p a0.5Vswl
1.17
0.67
⎛ V Lo ⎞ ⎛ 8 ⎞
0.67
⎝ 500 ⎠ ⎜ ⎟
1 1
K3 = ⎜ ⎟ ⎜ ⎟
3(1 − ν )π 0. 5
⎝ Vswl ⎠ ⎝ π ⎠ γ ′
0.67
N γ0.67
This shows that the stiffness depends on the vertical load (increasing slightly more than
linearly) and reduces with increasing bearing capacity factor. Note that N γ increases
much more rapidly than D R as relative density increases. The rather surprising effect
of density is due to the reduced penetration and hence reduced effective diameter. In
the limit an infinitely strong soil would result in point contact, and no rotational
stiffness! None of the other factors in the above equation vary much.
Therefore weaker soils (NC rather than OC clays, loose rather than dense sands) in
each case result in, paradoxically, higher rotational stiffnesses.
Effect of Embedment of the Spudcan on the Elastic Spring Stiffness

A study of the effect of embedment of flat plate and conical type footings has been
performed by Bell [42] to demonstrate the effect of penetration depth on the
translational and rotational spring stiffnesses. In order to take the embedment into
account the spring stiffness derived from the elastic solutions may be multiplied by the
depth factor Kd. The results of the study are summarized in the tables below for
Poisson’s ratios of 0.0, 0.2, 0.4 and 0.5. In the tables Kd1, Kd2 and Kd3 represent the
depth factors for the vertical spring stiffness, horizontal spring stiffness and the
rotational spring stiffness, respectively. Case 1 represents an open hole above the
spudcan, case 2 a back-filled hole.
Stiffness factors for ν=0.0

Kd1 Kd2 Kd3
D/R Case 1 Case 2 Case 1 Case 2 Case 1 Case 2
0.5 1.15 1.21 1.33 1.49 1.28 1.64
1.0 1.28 1.41 1.44 1.71 1.43 2.05
2.0 1.42 1.70 1.51 1.92 1.51 2.31
4.0 1.59 2.00 1.61 2.06 1.57 2.41

Kd1 Kd2 Kd3
0.5 1.11 1.18 1.32 1.47 1.23 1.54
1.0 1.21 1.34 1.42 1.67 1.37 1.90
2.0 1.34 1.59 1.48 1.85 1.44 2.15
4.0 1.49 1.85 1.58 1.98 1.51 2.25

Kd1 Kd2 Kd3
0.5 1.08 1.14 1.31 1.45 1.18 1.43
1.0 1.16 1.27 1.41 1.64 1.31 1.76
2.0 1.27 1.48 1.48 1.80 1.39 2.01
4.0 1.41 1.72 1.57 1.92 1.47 2.13

Kd1 Kd2 Kd3
0.5 1.07 1.10 1.32 1.44 1.18 1.39
1.0 1.15 1.23 1.44 1.62 1.31 1.71
2.0 1.25 1.44 1.51 1.78 1.40 1.99
4.0 1.40 1.69 1.59 1.91 1.51 2.16
Avoidance of numerical problems
An analysis in which the spudcan is considered as pinned to the seabed is equivalent to the
assumption that K1 and K2 are infinite and K3 is zero. In any analysis taking into account
foundation fixity (K3 non-zero) it is recommended that the effects of K1 and K2 are
considered, as they will tend to decrease the sway stiffness, and hence increase the natural
period and second order effects. As a first approximation elastic springs are recommended.
In some instances the vertical deformations resulting from the inclusion of such springs may
be large, and could compromise the numerical accuracy of the solution. A possible method
of reducing the absolute value of the deflections is given below.
Figure C6.11: Vertical load-displacement curves for leeward and windward legs
With reference to Figure C6.11, consider the windward and leeward legs as follows:
Leeward leg Windward leg

Vmax = VD + VE Vmin = VD - VE
where;
VD = Vertical reaction due to dead load
VE = Vertical reaction due to environmental load (and any change from the
variable load level used in computing VD)
The deflection due to environmental load alone Δd can be derived as:

V − VD Vmax VD − Vmin Vmin
Δd = max = Δd = =
Kv K *v Kv K *v
The modified vertical spring stiffness K*v is then:

Vmax Vmin
K*v = Kv K*v = Kv
Vmax − VD VD − Vmin
VD + VE VD − VE
K*v = Kv K*v = Kv
VE VE
where;
Kv = K1 from the PRACTICE
= 2GD/(1-ν)
For further information regarding foundation stiffness evaluation reference should be

made to Bell, [44].
C6.4 OTHER ASPECTS OF JACK-UP UNIT FOUNDATION PERFORMANCE
C6.4.1 Leaning Instability

A lower bound estimate of the leaning stability can be performed using the theory of
Hambly [45]. However, it should be recognized that such estimates have proven to be
generally conservative due to the omission of beneficial effects such as spudcan fixity
and lateral soil resistance on the legs.
The potential for jack-up unit leaning instability may largely be discounted if
appropriate installation procedures are adopted.
C6.4.2 Footprint Considerations

Installing a jack-up with its spudcans near or adjacent to existing footprints, or zones of
weaker material (naturally infilled spudcan footprints) may induce soil failure.
Mathematical models are available for the evaluation of ground stability in such
situations and, in particular, finite element techniques are becoming more widely used.
It is not possible to advise on a minimum acceptable distance between the proposed

spudcan locations and existing footprints as this will depend on several parameters.
These parameters include the soil conditions, the depth and configuration of the
footprints, the degree of soil backfill during and after spudcan removal, the elapsed
time since the last installation, the spudcan geometry and foundation loading.
As a general guideline it is usually acceptable for a spudcan to be installed at a

minimum distance (from the edge of the bearing area to the edge of the footprint) of
one diameter measured at the spudcan bearing area. However, in soft clay conditions,
with consequentially deep footing penetrations, the situation may be complicated by the
fact that the footprints may have larger diameters than the spudcans.
Also in dense sand or stiff clay conditions, where shallow footprints are unlikely to
influence the integrity of the spudcan foundations, the above guideline may be
conservative.
C6.4.3 Scour
The seabed is susceptible to scour when the shear stresses induced by fluid flow exceed
a certain value and/or when turbulent intensity of the flow is sufficiently large to lift
individual grains and entrain these in the flow. Both wave action and currents can
induce scour although in deep water, the effect of wave action on seabed scour is
negligible.
The following parameters are important for the assessment of scour potential:
a) Seabed material - size, shape, density and cohesion

b) Flow conditions - current velocity, wave-induced oscillatory velocities
interaction of waves and currents
c) Shape, size and penetration of jack-up footing.
Methods are available to determine whether significant scour is likely under waves and
currents. These generally proceed by considering the velocities near the seabed and by
calculation of the shear stresses. Guidance is given with regard to the assessment of
scour potential in the US NCEL [46] Marine Geotechnical Engineering Handbook.
C6.4.4 Seafloor instability
Seafloor instability may be caused by a number of mechanisms and where the potential
for unstable ground conditions is recognized it is recommended that expert local advise
is obtained.
In areas where liquefaction is known to be possible its potential must be assessed.

Liquefaction, or cyclic mobility, occurs when the cyclic stresses within the soils cause a
progressive build up of pore pressure. The pore pressure within the profile may build
up to a stage where it becomes equal to the initial average vertical effective stress.
Foundation failure may result depending on the location and extent of pore pressure
developed in the soil. The rate and degree of pore pressure build up will depend on
three factors:
a) The loading characteristics; that is, the amplitude, period and durations of the
different cyclic loading components
b) The cyclic characteristics of the soil deposits
c) The drainage and compressibility of the strata comprising the soil profile.
The cyclic loads may be induced by environmental or mechanical means, or by the

oscillatory ground accelerations imposed during earthquake conditions.
If appropriate soil conditions prevail, the potential for cyclic mobility should be
considered for a wide variety of load cases. Of particular interest is the windward
footing during storm conditions, where reduced vertical load and increased horizontal
load may theoretically induce lateral sliding or bearing failure.
C6.4.6 Spudcan - pile interaction
Where it is recognized that jack-up footings may adversely effect the piles of an
adjacent structure it will be necessary to assess the implications. Procedures such as
that proposed by Siciliano [47] may be used for deeply embedded footings in clay.
Otherwise, if adequate soil data is available, mathematical modeling techniques, such
as finite element modeling, could be used to assess the significance of the spudcan-pile
interaction.
au = Adhesion.
A = Spudcan effective bearing area.
B = Effective spudcan diameter at uppermost part of bearing area in contact
with the soil (for rectangular footing B = width).
B' = Increased effective spudcan diameter - load spread method.
cu = Undrained cohesive shear strength.
cul = Undrained cohesive shear strength at spudcan tip.
cum = Undrained cohesive shear strength at mudline.
cuo = Undrained cohesive shear strength at max bearing area.
d = Critical depth of failure below spudcan in sand.
dq = Bearing capacity factor = 1 + 2tanφ(1- sinφ)2 D/B.
dγ = Bearing capacity factor = 1.
D = Distance from mudline to spudcan maximum bearing area.
DR = Relative Density (percent).
e = Voids ratio.
e(e) = Voids ratio factor.
f(eL) = Voids ratio factor for loose sand.
f(eD) = Voids ratio factor for dense sand.
FH = Horizontal foundation capacity (envelope).
FM = Foundation moment capacity (envelope).
Fo' = Effective overburden pressure due to backfill at depth of the uppermost part
of the bearing area.
FV = Vertical foundation capacity.
FV,b = Vertical bearing capacity of fictitious footing on the surface of the lower
(bottom) clay layer with no backfill.
FVH = Vertical foundation capacity when horizontal load is present.
FVmax = Maximum vertical soil resistance (occurs when FH = 0).
G = Shear modulus.
GLoose = Shear modulus for loose sand.
GDense = Shear modulus for dense sand.
H = Distance from spudcan maximum bearing area to weak strata below.
Ir = Coefficient relating undrained shear strength to shear modulus.
I = Height of soil column above spudcan.
j = Dimensionless stiffness factor for sand.
ks = Coefficient of punching shear.
kv = Vertical foundation stiffness (= K1).
K1,K2,K3 = Vertical, horizontal and rotational stiffness.
Kd1 = Stiffness factor on vertical stiffness to account for embedment.
Kd2 = Stiffness factor on horizontal stiffness to account for embedment.
Kd3 = Stiffness factor on rotational stiffness to account for embedment.
K*v = Modified vertical foundation stiffness.
L = Foundation length, for circular foundation L = B.
n = Inverse slope of load spreading (slope of spread = 1:n).

N = Stability factor.
NC = Bearing capacity factor (taken as 5.14).
NC' = Alternative bearing capacity factor for normally consolidated clays.
Nq = Bearing capacity factor = eπtanφtan2(45 + φ/2).
Nγ = Bearing capacity factor
= 2(Nq + 1) tanφ for Vesic analysis [8]
= 1.5(Nq - 1) tanφ for Brinch Hansen analysis [17].
OCR = Over consolidation ratio.
pa = Atmospheric pressure.
po' = Effective overburden pressure at spudcan base level (i.e. depth of maximum
bearing area).
QV = Factored vertical leg reaction.
QH = Factored horizontal leg reaction.
R = B/2.
sq = Bearing capacity shape factor = 1 + (B/L)tanφ.
sγ = Bearing capacity shape factor = (1 - 0.4B/L).
(= 0.6 for circular footing under pure vertical load).
T = Thickness of weak clay layer underneath spudcan.
V = Embedded spudcan volume.
VD = Vertical reaction due to dead load.
VE = Vertical reaction due to environmental load (and any change from the
variable load level used in computing VD).
VLo = Maximum vertical foundation load during preloading.
Vmax = Maximum footing reaction on leeward leg.
Vmin = Minimum footing reaction on windward leg.
Vswl = Seabed vertical reaction under still water conditions.
W = Weight of soil plug (load spread method) = [1 + 2H/(nB)]2AHγ'.
z = Vertical foundation settlement for conventional bearing capacity analysis.
zu = Vertical displacement required to mobilize capacity FV.
α = Roughness factor = au/cu.

β = The equivalent cone angle of the spudcan.
Δd = Vertical deflection due to environmental load.
γ' = Submerged unit weight of soil.
φ = Angle of internal friction for sand - degrees.
φ' = Angle of internal friction for sand - degrees, dependent on d/B ratio.
ν = Poisson's ratio.
ρ = Rate of increase of cohesive shear strength with depth.
1 Meyerhof G.G. (1972), "Stability of Slurry Trench Cuts in Saturated Clay", Proceedings
of the Speciality Conference on Performance of Earth and Earth Supported Structures,
ASCE, pp. 1451-1466.
2 Britto A.M., Kusakabe, Osanu (1983) "Stability of Axisymmetric Excavations in

Clays", Journal of Geotechnical Eng., Vol 109, No. 5.
3 Young A.G., Remmes B.D., Meyer B.J., (1984) "Foundation Performance of Offshore
Jack-Up Drilling Rigs" Journal of Geotechnical Engineering, Vol. 110, No. 7, pp. 841-
859.
4 Skempton A.W. (1951), "The Bearing Capacity of Clays", Building Research Congress.
5 Davis E.H., Booker J.R., (1973), "The Effect of Increasing Strength with Depth on the
Bearing Capacity of Clays", Geotechnique Vol. 23, No. 4, pp. 551-563.
6 Salencon J., Matar M., (1982), "Capacite portante des Foundations superficielles
circulaires", Journal de Mecanique theorique et applique, Vol. 1, No. 2, pp. 237-267.
7 Houlsby G.T., Wroth C.P., (1983), "Calculation of Stresses on Shallow Penetrometers

and Footings", Proc. IUTAM Symp. on Seabed Mechanics, Newcastle, pp. 107-112.
8 Vesic A.S., (1975), "Bearing Capacity of Shallow Foundations", Foundation

Engineering Handbook (H.F. Winterkorn and H.Y. Fang, eds.), 121-147, Van Nostrand.
9 Endley, S.N., Rapoport, V., Thompson, P.J., and Baglioni, V.P. (1981), "Prediction of
Jack-up rig Footing Penetration", OTC, Houston, OTC 4144.
10a Noble Denton & Associates (1987), "Foundation Fixity of Jack-up Units, Joint Industry
Study", Volumes I, II.
10b Noble Denton & Associates (1988), "Foundation Fixity of Jack-up Units, Joint Industry
Study, Extra work", Volume III.
11 Houlsby G.T., Wroth C.P., (1982), "Determination of undrained strengths by cone

penetration tests", Proceedings of the Second European Symposium on Penetration
Testing / Amsterdam.
12 Houlsby G.T., Wroth C.P., (1982), "Direct Solution of Plasticity Problems in Soils by
the Method of Characteristics", Proceedings of the Fourth International Conference on
Numerical Methods in Geomechanics, Edmonton, Canada.
13 Houlsby G.T., (1982), "Theoretical Analysis of the Fall Cone Test" Geotechnique 32,
No. 2, 111-118.
REFERENCES FOR SECTION C6 (continued)
14 Koumoto T., Kaku K. (1982), "Three-Dimensional Analysis of Static Cone Penetration"
Proceedings of the Second European Symposium on Penetration Testing, Amsterdam.
15 Houlsby G.T. (1991), "Bearing Capacity Factors for Conical Footings on Clay -
Comments on Derivation of Factors", presented to Jack-Up Working Group
Foundations Sub-Committee, London.
16 Andersen K.H. (1988), "A Review of Soft Clay under Static and Cyclic Loading",
Invited lecture, International Conference on Engineering Problems of Regional Soils,
Being, China.
17 Brinch Hansen J., (1970) "A Revised and Extended Formula for Bearing Capacity",
Bulletin No. 28, Danish Geotechnical Inst., Copenhagen.
18 Graham J., Stuart J.G. (1971), "Scale and Boundary Effects in Foundation Analysis",
Journal of the Soil Mechanics and Foundation Division, ASCE, Vol. 97, No. SM11,
November, pp. 1533-1548.
19 James R.G., Tanaka H. (1984), "An Investigation of the Bearing Capacity of footings
under Eccentric and Inclined Loading in Sand in a Geotechnical Centrifuge", Proc.
Symp. Recent Advances in Geotechnical Centrifuge Modelling, University of
California, Davis, pp. 88-115.
20 Kimura T., Kusakabe O., and Saitoh K. (1985), "Geotechnical Model Tests of Bearing
Capacity Problems in Centrifuge", Geotechnique, Vol. 35, No. 1, pp. 33-45.
21 Dean E.T.R. (1991), "Some Potential Approximate Methods for the Preliminary
Estimation of Excess Pore Water Pressures and Settlement-Time Curves for Submerged
Foundations subjected to Time Dependent Loading", Cambridge University
Engineering Department, CUED/D-Soils/TR240.
22 Meyerhof G.G. (1984), "An Investigation of the Bearing Capacity of Shallow Footings
on Dry Sand", Proceedings 2nd ICSMFE, Rotterdam.
23 Dutt R.N., Ingram W.R. (1988), "Bearing Capacity of Jack-up Footings in Carbonate
granular Sediments", Proceedings of the International Conference on Calcareous
Sediments, Perth, pp. 291-296.
24 Poulos H.G., Chua E.W. (1985), "Bearing Capacity of Foundations on Calcareous
Sand", Proceedings 11th ICSMFE, San Francisco, Vol. 3, pp. 1619- 1622.
25 Yeung S.K., Carter J.P. (1989), "An Assessment of the Bearing Capacity of Calcareous
and Silica Sands", International Journal for Numerical and Analytical Methods in
Geomechanics, Vol. 13, pp. 19-26.
26 Hanna A.M., Meyerhof G.G. (1980), "Design Charts for Ultimate Bearing Capacity of
Foundations on Sand Overlying Soft Clay", Canadian Geotechnical Journal Vol. 17.
27 Young A.G., Focht J.A. Jr. (1981), "Subsurface Hazards Affect Mobile Jack-up Rig
Operations", Sounding, McClelland Engineers Inc., Houston, Texas, Vol. 3, No. 2, pp.
4-9.
28 Jacobsen M., Christensen K.V., Sorensen C.S. (1977), "Gennemlokning of Tynde
Sandlag (Penetration of Thin Sand Layers)", Vag-och Vattenbyggaren 8-9, Sweden.
29 Higham M.D. (1984), "Models of Jack-up Rig Foundations:", M.Sc. Thesis, University
of Manchester.
30 Craig W.H., Chua K. (1990), "Deep Penetration of Spudcan Foundations on Sand and
Clay", Geotechnique, Vol. 40, No. 4, pp. 541-556.
31 Baglioni V.P., Chow G.S. Endley S.N. (1982) "Jack-up Rig Foundation Stability in
Straified Soil Profiles", Proceedings, 14th OTC, Houston, OTC 4408.
32 Craig W.H., Higham M.D. (1985), "The Applications of Centrifugal Modelling to the
Design of Jack-up rig Foundations". Proceedings Offshore Site Investigation
Conference, Vol. 3. ISBN 0-86010-668-3.
33 API RP2A (1989), "API Recommended Practice for Planning, Designing, and
Constructing Fixed Offshore Platforms", API Recommended Practice 2A (RP2A) 18th
Edition, Washington.
34 Det Norske Veritas (1977), "Rules for the Design and Inspection of Offshore Structures,
Appendix F, Foundations", H`vik, Reprint with corrections (1980).
35 Santa Maria P.E.L. de (1988), "Behavior of Footings for Offshore Structures under
Combined Loads", Ph.D. Thesis, Oxford University.
36 Andersen K.H. (1992), "Cyclic effects on Bearing Capacity and Stiffness for a Jack-up
Platform on Clay", NGI Oslo report 913012-1, Rev 1.
37 Brekke J.N., Murff J.D., Lamb W.C. (1989) "Calibration of Jackup Leg Foundation
Model Using Full-Scale Structural Measurements", Proceedings Offshore Technology
Conference, Houston, (OTC 6127).
38 Matlock H. (1970), "Correlations for Design of Laterally Loaded Piles in Soft Clay",
39 Tan F.S.C. (1990), "Centrifuge and Theoretical Modelling of Conical Footings in
Sand", Ph.D. Thesis, Cambridge University.
40 Wroth et al. (1979), "A Review of the Engineering Properties of Soils with Particular
Reference to the Shear Modulus", Cambridge University Engineering Department.
Report No 1523/84./SM049/84.
41 Dean et al. (1992a), "A New Procedure for Assessing Fixity of Spudcans on Sand",
Andrew N Schofield and Associates Ltd., Cambridge, for Joint Industry Jack-Up
Committee.
42 Dean et al. (1992b), "A New Procedure for Assessing Fixity of Spudcans on Sand -
Further Notes", Andrew N Schofield and Associates Ltd., Cambridge, for Joint Industry
Jack-Up Committee.
43 Hardin B.O., and Drnevich V.P. (1972), "Shear Modulus and Damping in Soils: Design
Equations and Curves", J. Soil Mech. Foundation Division, ASCE Vol 98, SM7, 667-
692
44 Bell R.W. (1991), "The Analysis of Offshore Foundations Subjected to Combined
Loading", MSc. Thesis presented to the University of Oxford.
45 Hambly E.C. (1985), "Punch-through Instability of Jack-up on Seabed", Journal of
Geotechn. Eng., ASCE, Vol. 111, No. 4.
46 US NCEL (1985), "Handbook for Marine Geotechnical Engineering". Deep Ocean
Technology, Naval Civil Engineering Laboratory, Port Hueneme, CA 93043.
47 Siciliano R.J., Hamilton J.M., Murff J.D. (1990), "Effect of Jackup Spudcans on Piles",
48 Dean, et al. (1995), "Centrifuge Modelling of 3-Leg Jackups with Non-Skirted and
Skirted Spuds on Partially Drained Sand", Proceedings Offshore Technology
Conference, Houston, (OTC 7839).
49 Wong P.C. and Murff J.D. (1994), "Dynamic Analysis of Jack-up Rigs Using Advanced
Foundation Models", Proceedings OMAE, Houston, paper 94-1315
50 Svano and Tjelta (1993), "Skirted Spudcans - Extending Operational Depth and
Improving Performance", 4th City University Jack-up Platform Conference, London.
51 Baerheim (1993), "Structural Effects of Foundation Fixity on a Large Jack-up", 4th City
University Jack-up Platform Conference, London.
52 Van Langen and Hospers (1993), "Theoretical Model for Determining Rotational
Behavior of Spudcans", Proceedings Offshore Technology Conference, Houston, (OTC
7302).
53 Noble Denton Europe & Oxford University (2005), "The Calibration of SNAME
Spudcan Footing Equations with Field Data", Report No L19073/NDE/mjrh, Rev 4,
dated 21st November 2005.
54 Templeton J.S., Lewis D.R., Brekke J.N. (2003), "Spud Can Fixity in Clay, First
Findings of a 2003 IADC Study", 9th City University Jack-up Platform Conference,
London.
55 Andersen K.H. (2004), "Cyclic clay data for foundation design of structures subjected
to wave loading", Invited lecture, International Conf. on CyclicBehaviour of Soils and
Liquefaction Phenomena, CBS04, Bochum, Germany. Proc. p. 371 – 387.
56 Svanø G. (1996), "Foundation Fixity Study for Jack-Up Unit", SINTEF report STF22
F96660, August 1996.
57. Murff, J. D., M. D. Prins, E. T. R. Dean, R. G. James, A. N. Schofield (1992), "Jack-Up
Rig Foundation Modeling", Proceedings, Offshore Technology Conference, Houston,
(OTC 6807).
58. van Langen, H., P. C. Wong, E. T. R. Dean (1997), “Formulation and Validation of a
Theoretical Model for Jack-Up Foundation Load-Displacement Assessment”,
Proceedings, 6th International Conference on the Jack-Up Platform – Design,
Construction and Operation, London.
59. Wong, P. C., J. C. Chao, J. D. Murff, E. T. R. Dean, R. G. James, A. N. Schofield, Y.
Tsukamoto (1993), “Jack-Up Rig Foundation Modeling II”, Proceedings, Offshore
Technology Conference, Houston, (OTC 7303).
C.7 COMMENTARY TO CALCULATION METHODS –

DETERMINATION OF RESPONSES
C7.1 INTRODUCTION
The main objective of this Section is to provide documentation of the considerations

applied to the recommendations given in the Recommended Practice (PRACTICE)
concerning the determination of extreme responses. The PRACTICE recommends that
extreme response determination should always follow a procedure which always
considers the potential dynamic magnification of the jack-up's behavior.
C7.2 QUASI-STATIC EXTREME RESPONSE WITH INERTIAL LOADSET
Section 7.2 of the PRACTICE recommends that quasi-static responses are normally
determined by stepping the design wave through the structure. The extreme response is
obtained by combining the quasi-static wave-current loading with wind loads, dead
loads, etc., and an inertial loadset to represent the loading due to dynamic response.
This method approximates the random nature of wave excitation and implicitly
assumes that the extreme response is uniquely related to the occurrence of the extreme
wave.
As an alternative to the deterministic quasi-static design wave analysis a probabilistic

quasi-static random wave analysis may be used. This procedure is identical to a
random dynamic analysis procedure with the dynamic effects suppressed. Where
software permits, suppression may be achieved by setting all jack-up vibrational
masses to zero; see Section 7.3 of the PRACTICE. The significant wave height used
for the random wave analysis should include a water depth correction as shown on
Section 3.5.1 of the PRACTICE.
C7.3 CONSIDERATIONS AFFECTING THE DYNAMIC RESPONSE
The following considerations are relevant:

a) The highest natural period of the jack-up, in absolute terms, in relation to the
environmental excitation periods.
For the present purpose the highest natural period (Tn) is used as the single
indicator representing the properties of the structural system for the given
application.
The environmental excitation is due to sea waves and may contain energy at periods
in the range 2-3 to some 20 seconds. The energy content in the wave spectrum at
the lower end of this period range is controlled by wave saturation and is
independent of the geographical location. Experience has shown, and theoretical
calculations support, that the dynamic magnification may be neglected if the highest
natural period does not exceed 2.5 seconds.
The extent of the upper end of the period range, and its energy content, varies
significantly with the geographic location, an important indicator being the peak
period (Tp) of the wave spectrum of the extreme sea state. If Tn > 0.5 Tp the
dynamic behavior becomes increasingly significant and complex, which always
demands an appropriate dynamic analysis. Tn may be determined in accordance
with Section 7.3.5 of the PRACTICE. The derivation of the equations is given in
Appendix C7.A. The derivation of Tn and Tp are therefore critical to the dynamic
analysis. Considerations for the calculation of Tn and Tp are given in Section C7.4.
b) The magnitude of the dynamic magnification, relative to the quasi-static response.
Engineering calculations are subject to inherent (in)accuracy as a result of

uncertainties/inaccuracies in calculation methods and input data. Therefore, if the
magnitude of the dynamic contribution remains smaller than some 5%, it may be
considered to be covered by the overall margins included in the engineering
assessment criteria in Section 8 of the PRACTICE, without explicitly quantifying
and including the dynamic effect. For this evaluation, the magnitude of the
dynamic contribution may be estimated by the Dynamic Amplification Factor
(DAF) of an idealized single degree-of-freedom system in accordance with Section
7.3.6.1 of the PRACTICE. If the DAF < 1.05, dynamic magnification may be
neglected.
c) As the dynamic responses to periodic and random excitations can be significantly

different, the random nature should (where possible) be retained in the modeling of
the excitation. The simplified method described in Section 7.3.6.1 implicitly
assumes periodic excitation.
C7.3.7 Free surface corrections for frequency domain spectral wave load analysis
When using frequency domain spectral techniques, the wave forces are evaluated using
linear kinematics up to the mean water level only. Thus the force in the wave crests
may be underestimated. The underestimation is perhaps further compounded by the
drag force linearization. To account primarily for the effects if inundation, but also
partially to correct any errors in the drag linearization, empirical factors (FSE's) may be
derived to adjust the wave induced force and overturning moment obtained from a
frequency domain spectral analysis. By way of an example, the maximum wave force
and overturning moment on a pile group (representing jack-up legs) and accounting for
free surface effects and drag force non-linearity, have been compared with the wave
load on the same pie group when ignoring the sea surface variability and using
linearized drag force. This yields separate FSE's for shear and overturning moment
which may be used to make an initial correction for the above effects. Such factors are
dependent on the kinematic stretching algorithm assumed. Using Wheeler stretching
for a drag-dominated jack-up of typical size gives:
π⎧ H⎫
FSES = ⎨1 + ⎬
2 ⎩ 2d ⎭
π⎧
2
H⎫
FSET = ⎨1 + ⎬
2 ⎩ 2d ⎭
where;
FSES = the base shear correction factor
FSET = the overturning moment correction factor
d = water depth
H = the most probable maximum wave height (Hmax or Hmpm)
Similar expressions may be derived for different wave stretching algorithms.

C7.3.7 For a jack-up subjected to wave and current loading these factors should be used to
scale the MPM responses prior to combination with the mean (wind) response to form
MPME.
C7.4 SELECTION OF APPROPRIATE EXCITATION PERIOD
Sections 7.3.3, 7.3.6, 7.3.7 and 7.4 of the PRACTICE require that an appropriate
selection of excitation period is made. In making the selection consideration should be
given to the position of the natural period(s) in relation to the cancellation and
reinforcement points in the global wave loading of the jack-up which is important for
the magnitude of any dynamic wave magnification. Cancellations and reinforcements
in global loading are due to spatial separation of the wave load attracting legs and may
be different for different wave directions. The global loading may be characterized by
the total horizontal wave loading or overturning moment; cancellation and
reinforcement of points for these may appear at slightly different wave periods.
Figure C7.1 presents the periods at which first and second cancellations and
reinforcements occur in the total wave loading. It is valid for the main wave directions
of 3 and 4 -legged jack-ups in water depths exceeding 30m.
The calculation of natural period(s) is subject to uncertainty as a result of uncertainty in

the parameters affecting the natural period. In order to avoid the possibility of under-
conservative dynamic amplification factors, it is important to investigate the
relationship between the jack-up natural period and the cancellation and reinforcement
points in the transfer functions relating wave height to base shear and overturning
moment [1]. The range of possible natural period(s) should be bracketed and
compared
with the relevant cancellation/reinforcement points in the global wave loading. The
natural period(s) used in the dynamic analysis should be selected such that a realistic
but conservative value of the dynamic response is obtained for the particular
application envisaged, avoiding maximum dynamic amplification to coincide with
minimum environmental loading. Figure C7.1 may be used for a first evaluation of the
position of the calculated natural period(s), but it is recommended that the definitive
selection of natural period(s) be based on the shape of the global horizontal wave
loading (base shear) and overturning moment transfer functions calculated for the
actual application under consideration.
When the natural period occurs at a cancellation point in the transfer functions, the
mass or stiffness should be adjusted in a logical manner to move the natural period
away from the cancellation point.
If the analysis is for pinned footings with maximum hull mass, then the adjustment
should be made by reducing the hull mass (within the normal range) and/or by
introducing a degree of rotational fixity at the seabed.
If the analysis is for a case with footing moment fixity, then the adjustment would
most logically be made by varying the degree of rotational fixity at the seabed.
To minimise cancellation effects, it is suggested that the dynamic analysis may be
carried out for a single wave heading along an axis which is neither parallel nor
normal to a leg line. Thus, for a 3-legged unit with equilateral leg positions and a
single bow leg, suitable analysis headings would be with the environment approaching
from approximately 15o or 45o off the bow. The dynamic amplification factors
(DAF’s) should be determined for one, or both, of these headings, with suitably
adjusted natural period. The DAF’s (or more conservative DAF’s) may then be
applied to the final quasi-static analysis for all headings and hull weight cases with,
when applicable, non-linear fixity iteration according to Section 6.3.4.1 of the
Practice.
C7.5 METHODS FOR DIRECT DETERMINATION OF THE DYNAMIC RESPONSES
Section 7.3.7 of the PRACTICE outlines some of the considerations which are relevant
to direct methods for determining the dynamic responses. Appendix C7.B provides
further details of appropriate methodologies, together with flowcharts indicating their
implementation. An overview of the applicability of various modeling combinations is
given in Table C7.1 (overleaf).
For applications incorporating linearised foundation fixity in the dynamic analysis, the
methodology of Appendix C7.B.2.1 is recommended using the time domain approach
per Appendix C7.B.1.2.
For applications incorporating non-linear foundation fixity in the dynamic analysis,
the methodology needs to be selected with care, to ensure stable results. When the
analysis is used solely to determine DAF’s it is probable that a time domain simulation
of appropriate duration using the approach per Appendix C7.B.1.2 will be sufficient
with the extremes determined using the methodology of Appendix C7.B.2.1.
However, if the analysis is to be used to directly determine the extremes of other
responses, then the methodology of Appendix C7.B.2.3 is recommended using the
time domain approach per Appendix C7.B.1.2. This is because the results from the
non-linear fixity analysis (where non-linear foundation response occurs) are dependent
upon the load history experienced by the foundations. Consequently, the analysis
should be carried out for a number of load histories in order to determine a reliable
extreme value.
Model of the Environmental

Analysis Structural Model Excitation (always nonlinear)
Level Random
Periodic Frequency Time
Domain Domain
Linear A B
Simple SDOF
full Nonlinear Unsuitable Unsuitable
results not
available Linear Generally C
directly MDOF not
Nonlinear recommended E D
Complex - Linear Generally C

full MDOF not
results Nonlinear recommended E D
available
Notes:
A - Combines a simplistic model of the structural system with a simplistic model of

the excitation.
B - Is a refinement of case A. It remains simplistic to execute and lends itself to

both frequency and time domain methods. In the latter method the main
nonlinearities in the excitation can be retained and therefore non-gaussian
effects in the random response can be accounted for. Same limitations as for
case A.
C - Is a simplification of case D, if linear modeling of the structural system is a

sufficiently accurate representation.
D - Is the more complete and accurate representation of reality, but also the most
complex. This is a necessary combination for a detailed evaluation of the
dynamic behavior of a jack-up. Both random time and frequency domain
methods can be used; the latter requires some approximation in the form of
appropriate linearization of nonlinear terms and, therefore, the former are the
most suitable.
E - Incompatible combination.
Table C7.1 : Recommended combinations of the structural system and

environmental excitation models for a dynamic analysis
Figure C7.1 : Periods for wave force cancellation and reinforcement as

a function of leg spacing
APPENDIX C7.A - DERIVATION OF JACK-UP STIFFNESS EQUATION
To determine the natural period of a jack-up, the effective lateral stiffness seen by a horizontal
load acting at the level of the jack-up hull is required. To determine this stiffness the following
effects which cause hull lateral deflections are considered:
1. bending of the legs, leg-soil and leg-hull rotational stiffness.
2. shear deformation of the legs.
3. axial deformation of the legs.
4. hull bending deformation.
5. horizontal soil and leg-hull connection stiffness.
6. vertical soil and leg-hull connection stiffness.
7. second order P-Δ or Euler amplification.
Effects 4, 5 and 6 may readily be considered by means of modifications to terms in the stiffness
equation that can be derived for effects 1, 2 and 3. Taking each effect in turn:
1. Bending of the legs, leg-soil and leg-hull rotational stiffness.
Consider one leg as shown in the Figure:
F = shear transmitted from the hull

E = Young's modulus
ν = Poisson's ratio
I = second moment of area of leg
As = effective leg shear area
L = length considered
Krh = leg-hull connection rotational
stiffness
Krs = leg-soil connection rotational
stiffness
Mh = moment on leg-hull spring
Ms = moment on leg-soil spring
The bending equation may be written for any section z-z as:
Mzz = F.z - Ms
substituting the general equation of flexure:
∂2 x
EI 2 = − M zz = M s − F. z
∂z
hence:
∂x z2
EI = −M s . z − F + A
∂z 2
2 3
z z
EIx = M s . − F. + A. z + B
2 6
∂x M s
Apply the boundary condition: z = 0, x = 0, =
∂z K rs
EIM s
Hence: B = 0 and A =
K rs
The deflection at any point is then given by:
M s . z 2 F. z 3 M s . z. EI
EIx = − + (1)
2 6 K rs
∂x M h
To determine Ms, apply the boundary condition: z = L, Mh = F.L - Ms, =
∂z K rh
also from (1):
∂x M s . z F. z 2 M s
= − +
∂z EI 2EI K rs
Thus when z = L:
∂x M s . L F. L2 m s M h F. L − M s
= − + = =
∂z EI 2EI K rs K rh K rh
rearranging:
⎧ F. L F. L2 ⎫
⎨ + ⎬
⎩ K rh 2EI ⎭
Ms = (2)
⎧ 1 1 L⎫
⎨ + + ⎬
⎩ K rs K rh EI ⎭
The deflection xLB at x = L, due to bending is (from (1)):

M . L2 M s . L F. L3
xLB = s + −
2EI K rs 6EI
Rearranging and substituting from (2), the effective bending stiffness, KB = F/xLB, at z = L is
obtained thus:
⎧⎧ L L2 ⎫⎧ L2 L ⎫ ⎫
⎪⎨ + ⎬⎨ + ⎬ 3 ⎪
⎪ ⎩ K rh 2EI ⎭⎩ 2EI K rs ⎭ L ⎪
x LB = F⎨ − ⎬
⎪ ⎧ 1 1 L⎫ 6EI ⎪
⎨ + + ⎬
⎪ ⎩ K K EI ⎭ ⎪
⎩ rs rh ⎭
−1
⎧⎧ L L2 ⎫⎧ L2 L ⎫ ⎫
⎪⎨ + ⎬⎨ + ⎬ 3 ⎪
⎪ ⎩ K rh 2EI ⎭⎩ 2EI K rs ⎭ L ⎪
KB = ⎨ − ⎬
⎪ ⎧ 1 1 L⎫ 6EI ⎪
⎨ + + ⎬
⎪ ⎩ K rs K rh EI ⎭ ⎪
⎩ ⎭
After rearrangement and manipulation:
3EI / L3
KB = (3)
⎧ 3L 3( EI) 2 ⎫
⎪ 4 − LK K ⎪
⎪ ⎪
1− ⎨ rs rh
⎬
⎪ EI EI ⎪
+ L+
⎪⎩ K rs K rh ⎪⎭
2. Shear deformation of the legs.
Considering the shear force at any section zz is constant, the deflection xzzS due to shear is:
xzzS = F.z/(As.G)
but:
G = E/{2(1 + ν)} and, for steel, ν = 0.3
hence:
xzzS = 2.6F.z/(As.E) (4)
and the shear stiffness, KS, when x = L is:
F A .E
KS = = s (5)
x LS 2.6L
3. Axial deformation of the legs.
A) Consider a 3-leg jack-up, and assume that the legs are placed at the vertices of an equilateral
triangle. The shear applied to the hull is 3F, i.e. F acting on each leg.
Case 1 Case 2
3.F.L - 3.Mu - R.Y = 0 3.F.L - 3.Mu - R.X = 0

thus: thus:
3( F. L − M s ) 3( F. L − M s )
R = R =
Y X
applying Hook's law:
3( F. L − M s )L 3( F. L − M s )L
δaxial = δaxial =
A. E.Y A. E. X
The resulting hull rotation is:

θhull = 3.δaxial/(2.Y) θhull = 2.δaxial/X
9( F. L − M s )L 6( F. L − M s )L
= =
2A. E.Y 2 A. E. X 2
and the horizontal hull deflection is:
Δhorz = θhull.L Δhorz = θhull.L
9( F. L − M s )L2
6( F. L − M s )L2
= =
2A. E. Y 2 A. E. X 2
Y 3
If X = Y/cos30 =
2
9( F. L − M s )L2
Δhorz =
2A. E. Y 2
i.e. assuming an equilateral hull, the two loading directions yield the same horizontal
displacement at the hull:
9( F. L − M s )L2
Δhorz = (6)
2A. E. Y 2
B) Consider an N-leg jack-up, where N = 4, and assume that the legs are placed in two parallel
rows. The shear applied to the hull is NF, i.e. F acting on each leg.
Applying similar methods as above:

4( F. L − M s )L2
Δhorz = (7)
A. E. Y 2
where Y is the distance between the windward and leeward leg rows.
Comparing equations (6) and (7), it can be seen that (6) is a factor, Fg, of:
Fg = (9/2)/4 = 1.125
larger than (7).
The effective horizontal stiffness due to axial deformation, KA, rearranging (7), including Fg and
substituting for Ms from (2) is:
F
KA =
Δ horz . Fg
A. E.Y 2 / 4Fg L2
=
⎧ L L2 ⎫
⎨ + ⎬
⎩ K rh 2EI ⎭
L−
⎧ 1 1 L⎫
⎨ + + ⎬
⎩ K rs K rh EI ⎭
A. E. Y 2 / 4Fg L3
= (8)
⎧ EI L ⎫
⎨ + ⎬
⎩ K rs 2 ⎭
⎧ EI EI ⎫
⎨ + L+ ⎬
⎩ K rs K rh ⎭
4. Hull bending deformation.
Assume that the hull can be represented by equivalent beams joining the legs, of typical bending
stiffness IH:
If it is assumed that the hull deflects in double-curvature under the influence of the moments
transmitted by the leg-hull connection springs, and that the rotational deflections at the two sides
are equal (the side with higher stiffness has two legs acting on it) we can write, for one half of
the beam:
M.(Y / 2)
θ =
E. I H
Hence the hull rotational stiffness Khull, = M/θ = 2E.IH/Y
If this stiffness is considered as acting in series with the leg-hull connection spring Krh, the
modified stiffness is:
⎛ 1 1 ⎞
Krh' = 1 / ⎜ + ⎟
⎝ K rh K hull ⎠
Rearranging, and substituting for Khull gives:
Y. K rh
Krh' = Krh/(1 + )
2E. I H
Hence the modification factor Fr, to be applied to the leg-hull connection stiffness, Krh, to
account for hull flexibility is:
1
Fr = (9)
⎧ Y. K rh ⎫
⎨1 + ⎬
⎩ 2E. I H ⎭
5. Horizontal soil leg-hull connection stiffness.
The horizontal soil and leg-hull connection stiffnesses, Khs and Khh, may be considered to act in
series with the lateral stiffness due to leg shear deformation (ASG/L). The combined stiffnesses
is then:
⎛ L 1 1 ⎞
KS' = 1 / ⎜ + + ⎟
⎝ A S G K hs K hh ⎠
rearranging, gives:
A G A G
KS' = (ASG/L)/(1 + S + S )
LK hs LK hh
AS.E AS.E
= (ASG/L)(1 + + )
2.6L. K hs 2.6L. K hh
If it is considered that the modified leg deformation stiffness Ks' is linked to the unmodified
value by a factor, Fh:
1
Fh = (10)
⎧ A s .E A s .E ⎫
⎨1 + + ⎬
⎩ 2.6L. K hs 2.6L. K hh ⎭
6. Vertical soil and leg-hull connection stiffness.
The vertical soil and leg-hull connection stiffnesses, Kvs and Kvh, may be considered to act in
series with the axial stiffness due to leg axial deformation (AE/L). The combined stiffnesses is
then:
⎛ L 1 1 ⎞
KA' = 1 / ⎜ + + ⎟
⎝ AE K vs K vh ⎠
rearranging:
AE AE
KA' = (AE/L)/(1 + + )
LK vs LK vh
If it is considered that the modified leg deformation stiffness KA' is linked to the unmodified
value by a factor, Fv:
1
Fv = (11)
⎧ AE AE ⎫
⎨1 + + ⎬
⎩ L. K vs L. K vh ⎭
7. Second order P-Δ or Euler amplification.
The deflection will (approximately) be amplified by a factor (1 - [P/PE]) due to second order
effects. The Euler load, PE, may be derived as follows, accounting for the soil and leg-hull
connection rotational springs:
P = axial load in leg

E = Young's modulus
I = second moment of area of leg
L = length considered
Krh = leg-hull connection rotational
stiffness
Krs = leg-soil connection rotational
stiffness
Mh = moment on leg-hull spring
Ms = moment on leg-soil spring
xh = hull deflection
The bending equation may be written for any section z-z as:
MZZ = P.x - MS
substituting the general equation of flexure:
∂2x
EI 2 = -Mzz = Ms - P.x
∂z
hence:
∂ 2 x P. x M s
+ =
∂z 2 E. I E. I
let μ2 = P/E.I
hence:
∂2x Ms
2 + μ (x − )=0
2
(12)
∂z P
The solution to (12) is:
M
x = A.Cosμz + B.Sinμz + s (13)
P
differentiating (13):
∂x
= -μA.Sinμz + μB.Cosμz (14)
∂z
When z = 0, x = 0 and hence, from (13), A = -Ms/P
∂x M s
When z = 0, = and hence, from (14) B = Ms/(μ.Krs)
∂z K rs
− Ms Ms M
x= . Cosμz + .Sinμz + s
Thus: P μK rs P (15)
∂x μM s M
= .Sinμz + s . Cosμz
and: ∂z P K rs (16)
Apply boundary conditions at leg-hull interface to derive the equation yielding the Euler load:
∂x M h
When z = L, = (17)
∂z K rh
and x = xh (18)
also Mh = P.xh - Ms (19)
From (16) and (17):

M h μM s M
= .SinμL + s . CosμL (20)
K rh P K rs
From (15) and (18):
M Ms M
xh = − s . CosμL + .SinμL + s (21)
P μK rs P
Substituting (21) into (19) gives:
P. M s
Mh = − M s . CosμL + .SinμL
μK rs
Mh P
hence: = .SinμL − CosμL (22)
M s μK rs
Rearranging (20) gives:
M h μK rh K
= .SinμL + rh . CosμL (23)
Ms P K rs
Equating the (22) and (23):
⎧ P μK rh ⎫ ⎧ K rh ⎫
SinμL⎨ − ⎬ = CosμL⎨1 + ⎬
⎩ μ. K rs P ⎭ ⎩ K rs ⎭
⎧ K rh ⎫
⎨1 + ⎬
⎩ K rs ⎭
or: TanμL =
⎧ P μK rh ⎫
⎨ − ⎬
⎩ μ. K rs P ⎭
μK . P + μK rh . P
= 2 rs 2
P − μ K rs . K rh
By definition P = μ2EI so:
( K rs + K rh )μEI
TanμL = (24)
(μEI) 2 − ( K rs . K rh )
Notes:
1. When Krs = 0, and Krh = ∞, (24) reduces to TanμL = ∞
i.e. μL = π/2, 3π/2, 5π/2, …
The smallest finite value satisfying (24) is π/2, thus μL = π/2 and μ2 = P/(EI) hence:
PE = π2EI / (4L2)
2. When Krs = ∞, and Krh = ∞, (24) reduces to TanμL = 0
i.e. μL = 0, π, 2π, 3π, …
Rejecting the first value (μL = 0) as this give PE = 0, the smallest value satisfying (24) is
μL = π
hence:
PE = π2EI / L2
3. For finite values of Krs and Krh the Euler load may be determined using a graphical solution.
For example:
Krs = 2.65 x 1010 Nm/rad
Krh = 5.30 x 1010 Nm/rad
E = 2.10 x 1011 N/m2
I = 7.45 m4
L = 100m
From equation (24) the LHS = TanμL = Tan100μ (Note μ is in radians/m)

( K rs + K rh )μEI
the RHS =
(μEI) 2 − ( K rs . K rh )
124.4μ
=
2448μ 2 − 14045
.
Plotting these as shown in Figure C7.A.1 the smallest non-zero value in the example is
μ1 = 0.018248. Thus the Euler crippling load is:
PE = (0.018248)2EI
or, in the more general form:
PE = 0.337389π2EI / L2
Figure C7.A.1 Graphical solution of equation (24)

COMBINING EFFECTS 1 TO 7 ABOVE:
For the leg under consideration, all the effects can be combined by considering the components
as springs in series, thus Ke, the effective spring stiffness for one leg is deduced from:
1 1 1 1
= + +
K e K B KS K A
where the stiffness terms KB, KS and KA are derived in (3), (5) and (8). Rearranging and
including the Euler amplification effect:
⎡ P⎤
⎢1 − ⎥
⎣ PE ⎦
Ke =
1 1 1
+ +
K B KS K A
⎡ P⎤
⎢1 − ⎥
⎣ PE ⎦
=
⎧ 3L 3( EI ) ⎫ 2
⎧ EI L ⎫
⎪ 4 − LK K ⎪ ⎪⎪ K + 2 ⎪⎪
⎪ rh ⎪
1− ⎨ rs
⎬ ⎨ EI EI ⎬⎪
rs
⎪ EI + L + EI ⎪ ⎪ + L+
⎪⎩ K rs K rh ⎪⎭ 2.6L ⎪⎩ K rs K rh ⎪⎭
+ +
3EI / L3 As.E A. E. Y 2 / 4Fg L3
3EI ⎡ P⎤
3 ⎢1 − ⎥
L ⎣ PE ⎦
=
⎧ 3L 3( EI) 2 ⎫ ⎧ 3( EI) 2 3EI ⎫
⎪ 4 − LK K +
⎪
⎪
⎪ 7.8I 4Fg L ⎪⎪ L3 K rs 2L2
3 ⎪
⎪
1− ⎨ rs rh
⎬+ 2 + ⎨ ⎬
⎪ EI EI ⎪ A . L AEY 2 ⎪ EI EI ⎪
+ L+ s
+ L+
⎪⎩ K rs K rh ⎪⎭ ⎪⎩ K rs K rh ⎪⎭
3EI ⎡ P⎤
3 ⎢1 − ⎥
L ⎣ PE ⎦
=
⎧ 3L 4Fg L3 ⎧ 3( EI ) 2 3EI ⎫ 3( EI ) 2 ⎫
⎪ − 2 ⎨ 3 + ⎬− ⎪
⎪ 4 AEY ⎩ K rs L 2L2 ⎭ LK rs K rh ⎪ 7.8I
1− ⎨ ⎬+ 2
⎪ EI EI ⎪ As.L
+ L+
⎪ K rs K rh ⎪
⎩ ⎭
3EI ⎡ P⎤
3 ⎢1 − ⎥
L ⎣ PE ⎦
=
⎧ 3L 12Fg I ⎧ EI L ⎫ 3( EI) 2 ⎫
⎪ − 2 ⎨ + ⎬− ⎪
⎪ 4 AY ⎩ K rs 2 ⎭ LK rs K rh ⎪ 7.8I
1− ⎨ ⎬+ 2
⎪ EI EI ⎪ A s .L
+ L+
⎪ K rs K rh ⎪
⎩ ⎭
If the correction terms to Krh, As and A which are Fr, Fh, and Fv as defined in (9), (10) and (11)
respectively are included:
3EI ⎡ P⎤
3 ⎢1 − ⎥
L ⎣ PE ⎦
Ke =
⎧ 3L 12Fg I ⎧ EI L ⎫ 3( EI) 2 ⎫
⎪ − ⎨ + ⎬ − ⎪
⎪ 4 Fv AY ⎩ K rs 2 ⎭ Fr LK rs K rh ⎪
2
7.8I
1− ⎨ ⎬ + 2
⎪ EI EI ⎪ Fh . A s . L
+ L+
⎪ K rs Fr K rh ⎪
⎩ ⎭
If the foundation is effectively pinned, and Krs = 0, the equation can be simplified as follows
(multiply top and bottom of central term in denominator by Krs, and then set Krs = 0):
3EI ⎡ P⎤
3 ⎢1 − ⎥
L ⎣ PE ⎦
Ke =
12Fg I 3EI 7.8I
1+ 2 + +
Fv AY Fr LK rh Fh . A s . L2
If the foundation and leg-hull connection are effectively encastré, and Krs = Krh = ∞, the equation
can be simplified as follows (note that the Fr term to incorporate hull stiffness has vanished, as
its definition relies on a finite value of Krh; if an alternative definition were applied, its effect
could be retained).
12EI ⎡ P⎤
3 ⎢1 − ⎥
L ⎣ PE ⎦
Ke =
24Fg I 312I .
1+ 2 +
Fv AY Fh . A s . L2
In the absence of any of the terms for effects other than bending (i.e. setting A and As to
infinity), this further reduces to 12EI/L3, which is as expected for a beam, encastré at each end,
with one end free to slide.
APPENDIX C7.B - DETAILS OF APPROPRIATE DYNAMIC ANALYSIS METHODS
Section 7.3.6.1 of the PRACTICE provides a simple approach to determining the

dynamic response, based on the SDOF approximation. Sections 7.3.6.2 and 7.3.7 of the
PRACTICE outline more complex approaches and some specific recommendations are
included in Tables 7.1, 7.2 and 7.3.
It should be noted that the basic analysis may be carried out in either the frequency or
time domain and that there are then a number of approaches for determining the
required most probable maximum (MPM) response which is defined in Table 7.3 of the
PRACTICE as the mode value (or highest point on the PDF with a 63% chance of
exceedance). This corresponds to a 1/1000 probability level in a 3-hour storm.
The recommended, more complex, analysis methods are described below, together with
appropriate methods of determining the MPME. These may be summarized as follows:
C7.B.1 Analysis Methods
C7.B.1.1 Frequency Domain Methods:

- Use entire RAO (DAF) from simplified SDOF model.
- Use RAO from multi-degree-of-freedom (MDOF) model, with appropriate
linearization.
Frequency domain methods require the linearization of the wave-current drag loading.
It is recommended that the statistical (or least squares) linearization procedure
formulated by Borgman is adopted [L.E. Borgman, 'Ocean Wave Simulation for
Engineering Design', Civil Engineering in the Oceans, ASCE conference, San
Francisco, September 1967]; other forms of linearization may not adequately handle the
current velocity, wave induced particle velocity and the structures velocity (if a relative
velocity formulation is used). Table 7.2 of the PRACTICE makes some additional
recommendations regarding the generation of the random seastate.
C7.B.1.2 Time Domain methods:

- Use simplified SDOF model.
- Use MDOF model.
Time domain simulations require a suitable generation of the random seastate, that the
validity of the generated seastate is checked, and that the time-step for the solution of
the equations of motion is sufficiently small. It is also necessary to ensure that the
duration of the simulation(s) is sufficient for the method being used to determine the
MPME. Specific recommendations are given in Tables 7.2 and 7.3 of the PRACTICE.
C7.B.2 Methods for Determining the MPME
It should be noted that the simpler modeling approaches will not lead directly to the
MPME of all quantities of interest. For example, SDOF based models will provide
directly only the MPME hull displacement; simpler multi-degree-of-freedom models
may provide the MPME of total leg loads, but will not lead directly to loads in
individual members of a truss-leg.
As a means of circumventing this difficulty the analysis may be used solely to
determine the inertial loadset which represents the contribution of dynamics over and
above the quasi-static response (see Figure C7.B.1). The inertial loadset is then applied
to a structural model of appropriate complexity together with all the quasi-static loads
(due to wind, wave/current, weight, etc.) and the required responses determined. The
simplest inertial loadset uses a single point load at deck level. The magnitude of this
force is calculated to match the inertial overturning moment effects as shown on the
right hand side of figure C7.B.1 (blocks 18, 23, 24, and 25). It is possible to refine this
loadset to match both base shear and overturning moment inertial effects by simply
determining the magnitude of the loadset to match the inertial base shear and then
applying this loadset (single point load) at an elevation such that the inertial
overturning moment is matched.
However, the use of a distributed inertial loadset is considered more representative and
will, in turn, result in a more accurate description of the component dynamic
amplification effects as well as the amplification of global responses. The distribution
of the inertial loadset is based on the fundamental sway modes and the mass
distribution and is determined so that both the global base shear and overturning
moment responses are matched. Figure C7.B.1 (on the left hand side) outlines how a
distributed loadset (2-dimensional response) is determined based on the first two
fundamental bending modes (in the same direction) and the mass distribution.
An alternative to the inertial loadset approach is to use transfer functions to link known
responses with other required responses (for example to determine leg member loads
from total leg loads). The derivation of such transfer functions requires the use of
appropriately detailed models. Where non-linearities are significant the transfer
functions are not linear (and cannot be linearized) and may vary, for example, as a
function of the level of leg load(s).
The following methods are recommended for determining the MPME:
- Use of drag-inertia parameter (or equivalent) determined from mean and

standard deviation of a frequency or time-domain analysis.
- Fit Weibull distribution to results of a number of time-domain simulations to
determine responses at required probability level and average the results.
- Fit Gumbel distribution to histogram of peak responses from a number of time-
domain simulations to determine responses at required probability level.
- Apply Winterstein's Hermite polynomial method to the results of time domain
simulation(s).
Further details of the approaches are given below:

C7.B.2.1 Use of drag-inertia parameter (or equivalent) determined from mean and standard
deviation of a frequency or time-domain analysis
This procedure relies on the identification of the two components of the total dynamic
response, i.e. the quasi-static and the 'inertial' parts. The 'inertial' part is the
amplification of the quasi-static part due to dynamic effects, and should not be
confused with inertial wave loading. The procedure requires the determination of the
basic statistical parameters of the mean, μ, and the standard deviation (excluding the
mean), σ, of the required response variable(s). In general the root-mean-square, RMS,
≠ σ, unless μ = 0. The notation MPMR is used to refer to the most probable value of
the response minus the mean response, R(t) - μR, for a given storm duration. When the
mean is included the MPM value is referred to as the most probable maximum extreme
of R(t) and denoted by MPMER.
The response quantity of interest is indicated by the general notation R; this can be any
quantity which is related to the random wave excitation (e.g. base shear BS,
overturning moment OTM, etc.) Where necessary to distinguish between different
forms of response a second subscript is used as follows: 's' for (quasi-)static, 'i' for
inertial and 'd' for total dynamic (quasi-static plus inertial) response.
The procedure for estimating the extreme response is shown on Figure C7.B.4, and
requires the means and standard deviations of the (overall) dynamic and quasi-static
response, and the standard deviation of the 'inertial' response. These can be determined
from time domain simulations (Figure C7.B.2) or frequency domain analyses (Figure
C7.B.3). Figures C7.B.5 or C7.B.6 form an input to Figure C7.B.4. These Figures are
based on [SIPM EPD/51/52 'Dynamic Analysis and Estimation of Extreme Responses
for Jack-Ups', August 1991].
C7.B.2.2 Fit Weibull distribution to results of a number of time-domain; simulations to
determine responses at required probability level and; average the results.
This procedure requires a suitable length time domain simulation record for each
quantity of interest. The input seastate record should be checked for 'Gaussianity'.
Guidance is given in Tables 7.2 and 7.3 of the PRACTICE. The procedure requires the
following steps.
Step 1
The signal record is first analyzed to calculate the mean, μR, as:
n
∑ R(t
i −1
i )
μR =
n
where
R(ti) = time history of signal
ti = time points
n = number of useable time points in simulation
(discounting the run-in)
Step 2
The individual point-in-time maxima are next extracted according to the following
criteria:
A maximum occurs at ti if:

R(ti-1) < R(ti) and R(ti+1) ≤ R(ti)
Suppose Nmax maxima are found in the extraction.
Step 3
From the Nmax maxima, the mean of the signal, μR, is subtracted and the maxima R(max,i)
are ranked into 20 blocks having mid-points in ascending order. The blocks all have
the same width and the upper bound of block 20 is taken as being 1.01 x the largest
value, the lower bound of the first block being zero. A distribution of maxima
observations is then found, using for each block the Gumbel plotting position in order
to obtain the best possible description of the distribution for large values of R. If each
block has ni maxima, the cumulative probability Fi to be plotted against the mid point
for block i is then given by:
j= i −1 j= i
Fi = [(1 + ∑ n )∑ n ]
j= 0
j
j= 0
j
0.5
/ ( N max + 1)
where n0 = 0.
C7.B.2.2 Step 4.a
A Weibull distribution is fitted against the cumulative distribution of the maxima as
defined under Step 3 (see Steps 4.b to 4.d). The 3-parameter Weibull cumulative
distribution function is defined as:
⎡ ⎧ R − γ ⎫β ⎤
F(R;α,β,γ) = 1 - exp. ⎢− ⎨ ⎬ ⎥
⎣ ⎩ α ⎭ ⎦
where;
F(R;α,β,γ) = probability of non-exceedance
α = scale parameter
β = slope parameter
γ = threshold parameter
and α,β,(R-γ) > 0.0
Step 4.b
Only data points R(max,i), corresponding to a probability of non-exceedance greater than
a threshold value of 0.2 are used to fit the Weibull distribution, i.e. only the points:
⎡ ⎧ N max − i + 1⎫⎤
⎢ R (max, i ) ⎨ ⎬⎥ for i>0.2 x Nmax
⎣ ⎩ N max ⎭⎦
Notice that R(max,i) are in ascending order.
Step 4.c
For each of these points, the deviations between the Weibull distribution and the values
R(max,i) (transformed to Weibull scales) are calculated as:
δi = ln[-ln{1-F(R(max,i),α,β,γ)}] - β[ln(r(max,i)-γ) - ln(α)]
Step 4.d
The parameters α,β,γ are now estimated by a non-linear least square technique, i.e.
N max
∑δ i
i = 0.2 N max
2
is minimized
The procedure may be based on a Levenberg-Marquardt algorithm, using the

parameters of a 2-parameter Weibull distribution (found by the maximum likelihood
method) as initial estimates.
Step 5
The MPM value RMPM is found as the value of R for which:
1
F(R,α,β,γ) = 1 -
⎧ 3 hours ⎫
⎨N max . ⎬
⎩ simulation duration ⎭
Step 6
The total extreme MPM value, RMPME is found as:
RMPME = μR + RMPM
where μR = the mean value of R established in Step 1 RMPM = the MPM value
(excluding the mean) established in Step 5.
Step 7
The procedure is repeated for each required response parameter.
C7.B.2.3 Fit Gumbel distribution to histogram of peak responses from a number of time-domain
simulations to determine responses at required probability level.
The basic assumption of this method is that the 3-hour extreme values follow a Gumbel
distribution:
⎡ ⎧ x − ψ ⎫⎤
F3h(x) = exp ⎢− exp ⎨− ⎬
⎣ ⎩ κ ⎭⎥⎦
where;
F3h(x) = the probability that the 3-hour maximum will not exceed value x.
ψ = location parameter
κ = scale parameter
The following steps are followed for each required response parameter:
Step 1
Extract maximum (and minimum) value for each of 10 3-hour response signal records.
Step 2
A Gumbel distribution is fitted through these 10 maxima/minima. This is done using
the maximum likelihood method, yielding ψ and κ.
Step 3
The Most Probable Maximum Extreme is found according to:
⎧⎪ ⎫⎪
MPME = ψ - κ ln ⎨− ln{F 3h ( MPME )}⎬
⎪⎩ ⎪⎭
with;
F3h (MPME) = 0.37
The 0.37 lower quantile is used because the extreme of recurrence of once in 3 hours
will have a probability of exceedance of 0.63 (= 1 - 0.37). In this case it can be seen
that:
MPME = ψ
Step 4
The procedure of Step 3 is similarly applied for minima.
C7.B.2.4 Apply Winterstein's Hermite polynomial method to the results of time domain
simulation(s).
For Gaussian processes, analytical results exist for the determination of the MPM
values (e.g. MPM wave height = 1.86 x significant wave height). For general non-
inear, non-gaussian, finite band-width processes, approximate methods are required to
generate the probability density function of the process. The method proposed by
Winterstein [Winterstein S.R., 'Non-Linear Vibration Models for Extremes and
Fatigue', Journal of Engineering Mechanics, Vol. 114, No 10, 1988] fits a Hermite
polynomial of gaussian processes to transform the non-linear, non-gaussian process
into a mathematically tractable probability density function. This has been further
refined by Jensen [Jensen, J.J. 'Dynamic Amplification of Offshore Steel Platform
Responses due to Non-Gaussian Wave Loads', The Danish Center for Applied
Mathematics and Mechanics Report No 425, May 1991, Submitted to Journal of
Structural Engineering, ASCE] for processes with large kurtosis.
This procedure requires a suitable length time domain simulation record for each
quantity of interest. The input seastate record should be checked for 'Gaussianity'.
Guidance is given in Tables 7.1 and 7.2 of the PRACTICE. The calculation procedure
to determine the maximum of a time series, R(t), in duration T is as follows:
Step 1
Calculate the following quantities of the time series for the parameter under
consideration:
μ = mean of the process
σ = standard deviation
α3 = skewness
α4 = kurtosis
Step 2
Hence construct a standardised response process, z = (R - μ)/σ. Using this standardised
process, calculate the number of zero-upcrossings, N. In lieu of an actual cycle count
from the simulated time series, N = 1000 may be assumed for a 3-hour simulation.
Step 3
Compute the following quantities from the characteristics of the response parameters
identified earlier:
h3 = α 3 / [4 + 2 {1 + 15
. ( α 4 − 3)}]
h4 = [ {1 + 15. (α 4 ]
− 3)} − 1 / 18
K = [1 + 2h 3
2
+ 6h 4 2 ]
− 12
Step 3
It is necessary to seek a more accurate result by determining the solution of the
following equations for C1, C2 and C3:
σ2 = C12 + 6C1C3 + 2C22 + 15C32

σ3α3 = C2(6C12 + 8C22 + 72C1C3 + 270C32)
σ4α4 = 60C24 + 3C14 + 10395C34 + 60C12C22 + 4500C22C32 + 630C12C32 +
936C1C22C3 + 3780C1C33 + 60C13C3
using as initial guesses:
C1 = σK(1-3h4)
C2 = σKh3
C3 = σKh4
with σ, K, h3 and h4 from above. Following the solution for C1, C2 and C3, the values
for K, h3 and h4 are computed as follows:
K = (C1 + 3C3)/σ
h3 = C2/(σK)
h4 = C3/(σK)
Step 4
The most probable value, U, of the transformed process is computed by the following
equation:
⎛ 3 hours ⎞
U = 2 log e ⎜⎜ N ⋅ ⎟
⎝ simulation time (in hours ) ⎟⎠
Where U is a Gaussian process of zero mean, unit variance.
Step 5
The most probable maximum, transformed back to the standardised variable, z, is then
given by:
zMPM = K[U + h3(U2 - 1) + h4(U3 - 3U)]
Step 6
Finally, the most probable maximum extreme in the period T, for the response under
consideration, can be computed from the following equation:
RMPME = μ + σzMPM
Notes with Figure C7.B.1 - Part 1
General
The figure shows two possible paths. The path on the left through blocks 19
to 22 matches the dynamic base shear and the dynamic overturning moment,
by making up the difference between the dynamic and static base shears by a
distributed inertial force. This distributed inertial force is established by an
appropriate combination of structural mode shapes and lumped masses. The
basis for the calculation is that the base shear and overturning moment
inertial effects are simultaneously matched and combined in phase with the
quasi-static loads such that the levels of total global response are maximized.
By contrast, the path on the right chooses to match the dynamic overturning
moment by an inertial force in the form of a point load at deck level. This is
a very reasonable approximation of the inertial loadset, for cases where the
mass of the hull is much larger than the masses of the legs and the mode
participation factor (the relative horizontal displacement of the vibrating
jack-up) is also largest at the deck elevation. The inertial point load thus
determined is again not equal to the difference in dynamic and static base
shears; generally it overmatches the dynamic base shear. In this case the
remaining excess force is not compensated for (as was possible for the path
on the left) and must be accepted as an element of some conservatism.
Re blocks 17 and 18: The input to these blocks is obtained from Figure C7.B.4
(blocks 14, 15 and 16). Note that DAF3T will be greater than
DAF3S. This is in agreement with experience and supported by
theory.
Re block 19: An outline calculation of the distribution of F1i over height is
given in Figure C7.B.1 (Part 2).
Re block 23: The force F2i follows directly from the increase in OTM and the
height h at which F2i is applied above the effective hinge or
fixation points of the legs Therefore this does not require
knowledge of the mass distribution and mode shape.
Re block 24: The excess F3i in representing the dynamic base shear is
calculated as general verification. If F3i is found to be relatively
large compared to the dynamic base shear it is recommended
to follow the path on the left instead of the path on the right. A
criterion for this should be set by the user; as a suggestion the
Rev 3, August 2008

excess should not be greater than up to 5% of the dynamic
base shear.
.
Page 154
Figure C7.B.1, Part 1 - Procedure for determining inertial loadset
Notes with Figure C7.B.1 - Part 2

Procedure for Determining (Distributed) Inertial Loadset.
The extreme inertial load will generally be three dimensional in nature. It should be noted that vertical dynamic response effects are
not normally significant for storm conditions. It is assumed that the response of the Jack-up under combination of wave/current and
inertial loading will be in-line with the applied wave and current actions. Hence a 2-D response is considered along one of the global
structural axes.
The first two bending sway modes (i.e. global modes rather than local leg bending modes) acting along the selected axis are
combined to form a pair of simultaneous equations which match both inertial base shear and overturning moment. Base shear is
given by the product of mass and assumed acceleration profile, and overturning moment by the product of mass, assumed lateral
acceleration profile and lever arm above footing level i.e.
M1i = α φ1 MZ + β φ2 MZ
F1i = α φ1 M + β φ2 M (1)
where M1i is the global inertial overturning moment (zero mean)
F1i is the global inertial base shear (zero mean)
φ1 is the first global sway/bending mode shape
φ2 is the second global sway/bending mode shape
M is a matrix of structural masses
Z is a vector of point elevations above footing level
α and β are scalars
Global inertial responses are calculated from the global response DAFs generated by the dynamic analyses, combined with the
design wave and current load i.e.
F1i = (DAF3S-1)mpmeSs
M1i = (DAF3T-1)mpmeTs (2)
where DAF3Ts is the global overturning moment DAF (using mpme responses)
DAF3Ss is the global base shear DAF (using mpme responses)
M1i is the maximum design wave and current overturning moment
F1i is the maximum design wave and current base shear
mpmeSs is the most probable maximum extreme static shear
mpmeTs is the most probable maximum extreme static overturning moment
The simultaneous equations (1) are solved for scalar multipliers α and β, which are used to calculate the inertial load set i.e.
Fin = α φ1 M + β φ2 M (3)
In its current format, Fin is a distributed load vector consisting of horizontal forces applied to each point mass in the structure.
Equations (1) to (3) can be readily adapted such that the inertial load is fully three dimensional in nature, by using the first and
second global (3-D) sway modes along both horizontal axes, and extending equation set (1) to 4 components.
Jack-up structures exhibit several unique properties which allow the use of a simplified inertial load set calculation procedure. For
the majority of units, approximately 80% of the total system mass (including added fluid mass) effectively acts at the hull COG. In
addition, the mass and stiffness distribution results in the ratio of the first and second bending/sway mode periods for each principal
direction being in excess of 5. This leads to the resonant component of response being largely confined to the fundamental modes
in each direction (sway and surge), with a potential contribution from the first torsional mode (yaw). On this basis, and assuming
torsion can be ignored, equation set (1) can be reduced:
M1i = αδHMHZH
F1i = αδHMH (4)
where δH is the first mode shape ordinate at the hull COG
MH is the point mass acting at the hull COG
ZH is the elevation of the hull COG above footing level
We can clearly relate the second of equations (4) with the inertial load set given in Section 7.3.6.1 of the Practice.
Figure C7.B.1, Part 2 - Procedure for determining (distributed) inertial loadset

Specific notes with Figure C7.B.2
General
The procedure for estimating the extreme response due to hydrodynamic loading shown in Figure C7.B.4 requires knowledge of the
mean and the standard deviation of the quasi-static and dynamic responses, and the standard deviation of the "inertial" response. A
time domain procedure may be used to determine these.
Re blocks 4, 5, 6: The mean of the "inertial" response is not used in the procedure. In most cases the mean of the static
response will be (approximately) equal to the mean of the dynamic response. Therefore, the mean of the
"inertial" response will be (approximately) zero. This may serve as a check on the simulations performed.
However, under certain conditions the means may truly be different. this can most clearly be seen when
relative velocities (i.e. the wave induced water particle velocity minus the structure's velocity) are used to
perform the dynamic simulation.
Figure C7.B.2 Time domain procedure for determining mean and standard deviation

General
The procedure for estimating the extreme response due to hydrodynamic loading shown in Figure C7.B.4 requires knowledge of the
mean and the standard deviation of the quasi-static and the dynamic responses, and the standard deviation of the "inertial"
response. A frequency domain procedure may be used to determine these. In order to reflect the interactions between the current
velocity, the absolute wave induced water particle velocity and the structure's velocity (if a relative velocity formulation is adopted)
and to linearize the associated drag loading adequately it is necessary to adopt a statistical or least squares linearization procedure
as first formulated by Borgman (see Ref. below). Other forms of linearization in frequency domain analysis cannot handle these
interactions.
For the least square linearization procedure, there only is a mean response in case of a non-zero current. The magnitude of the
mean depends on the value of the current velocity and on the standard deviation of the wave induced (horizontal) water particle
velocity, both taken at the same elevation z, and subsequently integrated over the full water depth. The wave induced water particle
velocity may be the absolute or the relative velocity, depending on which of these is more appropriate for the case considered.
The transfer functions HRs(ω) and HRd(ω) between the response and the water surface elevation are similarly dependent on both the
wave induced (horizontal) absolute or relative velocities and the current velocities at various elevations.
The means mRs and mRd and the transfer functions HRs(ω) and HRd(ω), are therefore a function of the sea state and the current sued
in the environmental definition.
Re block 3: The transfer function representing the difference between the dynamic and the quasi-static response is
only notionally associated with "mass inertial" forces (not to be confused with inertial wave loading). The
difference may additionally be due to damping forces and any effect causing (frequency dependent) phase
differences between HRd(ω) and HRs(ω). (e.g. associated with multi degree of freedom system responses).
Re blocks 4, 5, 6: The spectral analyses operate on the transfer functions HRx(ω), which by definition represent the time
varying part of the response minus the mean, i.e. Rx(t)-μRx.
A similar note on the mean values of the various responses as given with Figure C7.B.2 should be made
here. The mean value of the "inertial" response cannot be determined in a frequency domain analysis and
is not required either. However, the fact remains that in most cases the mean of the static response will be
(approximately) equal to the mean of the dynamic response. This should again serve as a useful check o
the analyses performed.
From the above general note it can be seen that both means will only be non-zero if there is a current
present. When relative velocities are used in the analysis of the dynamic problem the interaction between
the current and the relative velocity may be different for the dynamic and the static case, resulting in
realistically different mean values.
Reference:
L.E. Borgman
"Ocean wave simulation for engineering design"
Civil Engineering in the Oceans, ASCE conference, San Francisco, September 1967
Figure C7.B.3 Frequency domain procedure for determining mean and standard deviation
Figure C7.B.4 Procedure for estimating the extreme response


Re block 7: The correlation coefficient ρR is theoretically a value between -1 and +1. For virtually all
applications to offshore structures problems it is expected that:
δ2Rd > δ2Rs + δ2Ri so that 0 < ρR < 1.
For rR = 0, zero correlation, the quasi-static and "inertial" responses do not influence one
another and will be well separated in the frequency domain. This is generally only to be
expected for relatively low natural periods which fall in the (very) high frequency tail of the wave
spectrum and where HRs(ω) is also very low. Under these circumstances the variance (or
mean square) of the full dynamic response is equal to the sum of the variances of the quasi-
static response and the "inertial" response:
δ2Rd = δ2Rs + δ2Ri
Geometrically, this means a direct addition of non-overlapping areas of the two parts of the
response spectrum.
For rR = 1, full correlation, the quasi-static and "inertial" responses are fully dependent on one
another. The two parts of the response spectrum overlap strongly and will not really be
distinguishable. This will increasingly be the case for high natural periods, considerably closer
to the peak of the wave spectrum and therefore associated with a region of significant wave
energy, and where HRs(ω) is also having appreciable values. Under these circumstances the
standard deviation (instead of the variance) of the full dynamic response is equal to the sum of
the standard deviations of the quasi-static response and the "inertial" response:
δRd = δRs + δRi
Re blocks 8, 12, 16: Several definitions of the dynamic amplification factor DAF are in use. The purest and most
meaningful definition is believed to be DAF1, the ratio of the standard deviations of the
dynamic and static responses (block 8), i.e. after eliminating the means which are not affected
by dynamic magnification. If the static and the dynamic processes are both gaussian, or to an
equal degree non-gaussian, then DAF2 = DAF1; however, this will not be the case in general.
The ratio DAF3 of the most probably maximum extremes, including the means, is a practical
overall measure of the increase in response due to dynamics.
Re block 9: The mpm-factor for an arbitrary non-gaussian response is not known. As an engineering
postulate it is assumed that this is equal to the mpm-factor for Morison type wave loading on a
cylindrical element of unit length. The factor for a nominal number of 1000 peaks
(corresponding approximately with a 3 hr storm duration) then varies between the extremes of
3.7 (for inertial wave loading only and hence a gaussian process) and 8.0 (for drag waver
loading only and consequently a strongly non-gaussian process). It can be determined on the
basis of a drag-inertia parameter or the kurtosis of the response, as shown in Figure C7.B.5
with its associated notes.
Note that the factor 8.0 is different from the previously used factor of 8.6. This is due to the
fact that in this report the most probably maximum is consistently used as a predictor for the
maximum of a random process. The previous factor 8.6 referred to the expected maximum
instead of the most probably maximum.
Re block 10: The mpm-factor for the "inertial" part of the response is associated with the dynamic behavior
and predominantly of a purely narrow banded resonant nature. Experience has shown (and
theory supports this) that such lightly damped dynamic processes tend towards gaussianity so
that a mpm-factor of 3.7 is a reasonable and confident assumption for engineering purposes.
Re block 11: The relationship between the mpm-values is entirely analogous with the relationship between
the standard deviations from which the correlation coefficient is determined. However, while it
is theoretically proven equation for the standard deviations, it is an engineering postulate for
the mpm-values.
Re blocks 14, 15: It should be recalled that the procedure depicted in Figures C7.B.2 to C7.B.6 is aimed at
estimating the extreme short-term response due to hydrodynamic loading only (see General
Note 1). Therefore, the effect of wind should be excluded from the most probably maximum
extreme static and dynamic responses in block 14 and 15, respectively. Wind is assumed to
produce a static load and a static response, and not to influence the dynamic behavior. To
determine the ultimate response the mean response due to wind should be determined
separately and added to mpmeRs and mpmeRd.
Notes to Figure C7.B.4

Figure C7.B.5 Procedure for determining the mpm-factor of the static response

Re blocks 9a and 9b: These standard deviations may be obtained from separate time domain simulations in the same manner
as shown in Figure C7.B.2 (blocks 1 and 4) or, alternatively, from separate frequency domain analyses as
shown in Figure C7.B.3 (blocks 1 and 4). The mean values are not required and neither is it necessary to
subtract dynamic and static response time series or transfer functions, respectively.
Re block 9c: The drag-inertia parameter is defined as the ratio of the magnitude of the drag force to the magnitude of
the inertia force due to waves. All relationships given below are valid for the case of zero current, which is
used as the basis for the whole procedure in view of the engineering approximations involved. For an
element of a circular cylinder of diameter D and unit length, subjected to a periodic wave, the drag-inertia
parameter then becomes:
1 1
ρ Cd D v ) / ( Cm ρ π D a)
2 2
K = (
2 4
2
2C d v
= (1)
πC m Da
Where v and a are the velocity and the acceleration normal to the element, respectively. As both v and a
depend on the wave parameters (wave height, wave period, waterdepth) and the elevation at which the
element is located, it is obvious that K is also a function of depth, waterdepth, wave height and wave
period. therefore, the theoretical definition of K is only meaningful for Morison wave loading per unit length
of the element.
The definition of K can be generalized to random instead of periodic wave conditions by replacing the
deterministic normal velocity v by the standard deviation of the random normal velocity σv and replacing
the deterministic normal acceleration a by the standard deviation of the random normal acceleration σa.
Equation (1) then becomes:
2
2C d σ v
K = (2)
πC m Dσ a
Using a statistical or least squares linearization procedure in the frequency domain, as developed by
Borgman (see notes with Figure C7.B.3), it can be shown that for the wave force on an element of a single
member the standard deviations of the two parts of the wave force are as follows:
σR(Cm = 0) = 8 / x. 1/2 ρ Cd D.σv
2
σR(Cd = 0) = ρ Cm.1/4 π D .σa

2
These relationships can be used to determine σv and σa, which can then be substituted into equation 2 to
2
result in:
π σ R ( C m = 0)
K = . (3)
8 σ R ( C d = 0)
With R being the wave force per unit length in a random sea.
Equation (3) may subsequently be generalized to apply to any other local or global response R selected for
interest. It will be clear that such a generalization is purely an engineering postulate and not founded on
theoretical reasoning. It is an attempt to incorporate the important but unknown non-gaussian effects on
the maximum response through the assumed similarity with the wave loading process for which the non-
gaussian statistics are known.
Yet another way to determine the drag-inertia parameter K for a generalized response R is by using the
kurtosis of R. The kurtosis is defined through the expected values of the second and fourth order
moments of the time simulations of R, i.e.:
κ
4 2 2
= E {R } / [E {R }] (4)
For Morison wave loading per unit length of member the relationship between K and the kurtosis k is (see
Ref. 2 below):
4 2
105 K + 18 K + 3
κ = 2 2 (5a)
( 3K + 1)
or in the inverse form:
Notes to Figure C7.B.5

1/ 2
⎡ ⎧ 26(κ − 3) ⎫ ⎤
1/ 2
⎢ (κ − 3) + ⎨ ⎬ ⎥
⎢ ⎩ 3 ⎭ ⎥
K = (5b)
⎢ ( 35 − 3κ ) ⎥
⎢ ⎥
⎣ ⎦
While K varies between 0 (inertia loading only) and infinity (drag loading only) κ ranges from 3 to 35/3. It
may now be assumed that the same relationship holds for an arbitrary response variable R. Therefore, if
the kurtosis of R is know the corresponding drag-inertia parameter K can be determined. If this is done,
separate time domain simulations for the standard deviations in blocks 9a and 9b are not required but the
route through block 93 cannot be followed. One enters the diagram in block 9c and must read CRs from
Figure C7.B.6 as per block 9d.
Both the kurtosis and the drag-inertia parameter may be subject to appreciable statistical variability and
their determination may require time domain simulations of substantial length; see Ref. 2 below.
Re blocks 9d and 9e: Figure C7.B.6 (referred to in block 9d) is equivalent to the figure that was derived by Brouwers and
Verbeek and presented in Ref. 1 below as well as in Figure A1 of the SIPM - Practice (EP 89-0550).
However, this latter figure presented the ratio of the expected value of the extreme to the standard
deviation for a 1000 peaks, rather than the mpm-factor CR which is the ratio of the most probable maximum
value of the response to the standard deviation, which is used in this report. Therefore, Figure C7.B.6 has
been recalculated in accordance with Ref. 3 and now truly presents the mpm-factor CR. It should be noted
that the figure is valid for a narrow band process, the corresponding ratios for a broad band process being
somewhat smaller. Therefore, CR is a slightly conservative estimate for the mpm-factor. This is in
accordance with the general principles underlying a simplified engineering method and is well within the
accuracy of the overall procedure.
An alternative and practical method to estimate K is to apply the engineering assumption for estimating the
most probably maximum value of the dynamic response, as used in block 11 of Figure C7.B.4, to separate
responses due to hydrodynamic drag loading only and inertia loading only, replacing RS from block 9 and
Ri from block 10, respectively. These two hydrodynamic loading components are fully uncorrelated and so
are the responses caused by them; hence the correlation coefficient r = 0. Further, the mpm-factor for a
totally drag dominated Morison force is 8.0 and for a totally inertia dominated Morison force it is 3.7. With
these substitutions the equation in block 11 of Figure C7.B.4 becomes:
{8.0 σR (Cm = 0)} + {3.7 σR (Cd = 0)}
2 2 2
mpmR =
For zero correlation the standard deviation of the overall response is obtained from the equation:
σR {σR (Cm = 0)} + {σR (Cd = 0)}
2 2 2
=
(see note with block 7 of Figure C7.B.4).
These are the equations presented in block 9e. The comments made with regard to conservatism
included in the route through block 9d remain equally valid here.
Its determination in block 9c could therefore, strictly speaking be avoided. The input of KRs into block 93 of
Figure C7.B.5 is symbolic, representing the implicit use through σRs (Cm = 0) and σRs (Cd = 0), resulting
directly from blocks 9a and 9b. In practical applications it is recommended that both routes through block
9d and 93 are followed as a check on the calculations.
Reference 1:
J.J.H. Brouwers and P.H.J. Verbeek
"Expected fatigue damage and expected extreme response for Morison-type wave loading"
Applied Ocean Research, Vol. 5, No. 3, 1983, pp. 129-133
Reference 2:
P.M. Hagemeijer
"Estimation of drag/inertia parameters using time-domain simulations and the prediction of the extreme response"
Applied Ocean Research, Vol. 12, No. 3, 1990, pp. 134-140.
Reference 3:
J.J.M. Baar
Extreme values of Morison-type processes"
Report EP 90-33365, October 1990.
To be published shortly in Applied Ocean Research
Notes to Figure C7.B.5 (cont.)

The equation for the curve is Ref. 3, Specific notes with Fig. C7.B.5
3.72 / A C <B ( K < 0135
. )
CR = if R
( 6.91 + D ) / C C R < B ( K < 0135
. )
Where A, B, C and D are functions of k as follows:
A = 3K2 + 1
⎡ ⎤
B = 1 / ⎢( 2K ) (3K 2 + 1) ⎥
⎣ ⎦
C = ⎡ (3K 2 + 1) ⎤ / ( 2 K )
⎢⎣ ⎥⎦
2
D = 1/(8K )
Figure C7.B.6 Ratio CR of most probable maximum to standard deviation as a

function of drag-inertia parameter K for N = 1000 peaks
Figure C7.B.7 Comparison between the normalized spectra Sη(ω), Sφ(ω) and SPM(ω)
1. Parker G.J. (1997), ‘Calibration of an SDOF-Based Dynamic Analysis Procedure Including

Non-Linear Foundation Fixity’, Sixth International Conference The Jack-Up Platform,
Design Construction and Operation, City University, London.
C8 COMMENTARIES TO ACCEPTANCE CRITERIA
C8.0 BACKGROUND TO PARTIAL LOAD FACTORS
C8.0.1 General
Reliability analysis was used in the derivation of load effect factors which are
conservatively presented in the PRACTICE as load factors. All factors associated
with strength or resistance have been derived either from consensus (e.g. weight in
overturning) within the JUWG or from other relevant codes (e.g. AISC LRFD). The
philosophy used in the derivation of the load effect factors is discussed below. Further
references on the technique etc. are given in [1], [2], [3] and [4].
C8.0.2 Fundamental Question
When a jack-up is offered for operation at a marginal location, a number of issues such
as overturning stability, soil capacity and leg strength are addressed to ascertain the
fitness for purpose of the jack-up. In all these assessments, it is necessary to establish
an acceptable safety margin (or safety factor) between load and resistance. The
question is, how do we establish, quantitatively, the safety factor required for the
performance assessment?
C8.0.3 Solution
Loads and resistances are not uniquely defined due to physical, statistical and
methodological uncertainties. Acceptance of this fundamental principle has led to the
understanding that the use of safety factors merely assists in maintaining a level of
safety. Furthermore, the true goal of assessment is to achieve as consistent a level of
safety as possible when the safety factors are just satisfied. This demonstrates the need
to perform reliability analysis which would provide a framework to link the safety
factors to the safety levels. The various key stages of the analysis are described below:
C8.0.3.1 Probabilistic Description of Input
The code calibration project was performed in two stages. In Stage 1, sensitivity
studies were performed to identify the key parameters which influence the response of
a jack-up [5], [6]. These showed that significant wave height (Hs), peak period (Tp),
drag diameter (CDD), tidal current (VT) and the permissible interaction ratio for
structural elements were important items. The reliability studies, [2], showed that the
significance of Tp, CDD and VT was not critical and therefore the variability in these
parameters was ignored in Stage 2 studies, [3]. However, it was clear that the largest
value of the responses varied between different realizations of the same seastate and it
was therefore necessary to account for this variability. Thus in the final stages of the
study, Hs, the variability in the largest value in a storm and the permissible interaction
ratio were considered as variables. The following table summarizes the variables and
associated probability distributions used in Stage 2 studies [3]:
Parameter Distribution
Significant Waveheight, Hs Gumbel
Largest value in a storm Poisson
Permissible Interaction Ratio Log-normal
C8.0.3.2 Limit States
Three limit states, namely, overturning, preload and leg strength were considered.
These limit states are recognized to possess some degree of reserve of safety from
actual failure. As the PRACTICE is focused upon component failures, and target
safety levels were determined from average safety levels of exemplary rigs (see section
C8.0.3.5 below), reaching any of the selected limit states does not indicate "true"
failure. This is not significant to the code calibration.
C8.0.3.3 Response Model
A major objective of the JUWG has been to develop as comprehensive an analysis

method as possible which reflected the behavior of jack-ups in the elevated condition.
The developed method was applied to derive the response parameters required for the
reliability analysis. It must be noted that time domain analysis was performed in order
to properly capture all the non-linearities in jack-up loading and response. The
following table summarizes the key facts associated with the simulations and analysis
methodology:
• 0.5s time step used in simulations

• 3 hour simulations used for MPM
• 1 hour simulations with correction for response surfaces
• Hs increased by factor [1 + 0.5e(-d/30)]*
• 75% of max. variable load used in reliability analysis
• P-Δ effects included
• Hull Sag and Flexibility included
• Pinned Foundations
• Initial leg inclination included in leg strength evaluation .
• Relative velocity Morison's equation used
• 4% critical damping (plus damping from above)
• Jonswap spectrum used
• Winterstein [7], Juncher Jensen [8] used for MPM .
* This is changed to [1 + 0.5e(-d/25)] in the current draft of the PRACTICE.
The following response model, which linked Hs to the safety index (or probability) of
exceeding the given limit state in that Hs was derived from three simulations with
different but large (near 50-year return) Hs:
β = A + B.Hs + C.Hs2
Once this link was established, then using the probability distribution of the annual
extremes of significant wave height, the probability of limit state exceedance in any
one year (or the annual safety index) was computed.
C8.0.3.4 Safety Index vs. Safety Factor
By repeating the procedures described in C8.0.3.3 for different values of the resisting
quantities such as righting moment, permissible interaction ratio and preload a
rig/location specific link between safety factor and safety index was generated (e.g. see
Figure C8.0.1). This then permitted the evaluation of the required safety factor for a
specified level of safety.
C8.0.3.5 Reference or Target Safety Level
No absolute target or reference safety level was imposed. Instead, the four rig/location
combinations which were, on average, considered to be close to the limit were analyzed
using reliability techniques and the target safety level obtained by averaging the
individual safety levels achieved by each rig. This process is described graphically in
Figure C8.0.1. The process was repeated for each of the limit states considered.
It is not possible to directly compare safety levels achieved by the exemplary rigs with
safety levels of other offshore structures. However, the safety levels achieved are
broadly comparable.
Figure 8.0.1 : Link between safety factor and safety index (β)
C8.0.3.6 Derivation/Calibration of Safety Factors
As noted in section C8.0.3.5, four exemplary rig/location combinations were analyzed

using reliability techniques. The reliability analysis was further extended to tropical
cyclone locations by simply modifying the statistical distribution of the environmental
parameters to reflect the greater variability. The spread in the safety index vs. safety
factor curves was shown to be due to the varying levels of dynamicity of the four
different rig/location combinations and the environmental variability between North
Sea and tropical cyclone locations.
The response quantities were split into quasi-static and dynamic components in order to
investigate the potential for reducing the spread in safety levels across rigs and
locations. The objective of the optimizing function was to minimize the squared
difference between the achieved safety index and the target safety index. This
approach did show a reduction in spread of safety levels with the use of partial factors,
however, as the initial safety index spread itself was small, the decision was taken to
adopt a single load factor which minimized the spread whilst maintaining the target
safety level as discussed in section C8.0.3.5.
C8.1 STRUCTURAL STRENGTH CHECK
C8.1.1 Introduction
Code Basis
Currently, the most widely used codes for structural strength assessment for the
offshore industry are based on "working stress design". Examples of commonly used
codes are AISC ASD [9] and API RP2A [10]. Due to a number of inadequacies of
these codes, there has been a move by their authors to replace them with the "Load and
resistance factor design (LRFD)" approach. Although AISC and API allow use of both
the "working stress" and LRFD codes in parallel, it is their intention to phase out the
"working stress" methods in the future. To follow the trend in the industry, it was
decided that the structural strength assessment code to be used in the PRACTICE
should also be based on LRFD. Both the AISC LRFD [11] and API LRFD [12] codes
were reviewed as bases for jack-up structural assessment.
It was decided that the AISC LRFD should be used as the basis for the PRACTICE, for
the following reasons:
i) API LRFD covers tubular members thoroughly but refers the user to AISC LRFD
for non-tubular members. Since non-tubular members are commonly encountered
in jack-ups, AISC LRFD offers the greater applicability.
ii) Although a limit state code, the equations used in API LRFD are expressed in terms
of stresses and not loads. This would cause difficulties in the integration of AISC
LRFD with API LRFD for use in assessment computer programs for the non-
tubular cases (see (i)).
iii) A parametric study of the two codes for tubular members produced results for the
AISC LRFD equations (including the η exponent discussed below) similar to those
for the API LRFD code.
For analysis of a structure using the LRFD approach, it is necessary to define the
structure more comprehensively. Certain characteristics which occurred in jack-up
structures which could be fitted in with the AISC "working stress" codes had to be dealt
with specifically for AISC LRFD. Whether the treatment of these characteristics was
correct in the AISC "working stress" code was doubtful and hence the use of LRFD has
not created additional problems but has highlighted inadequacies of the previously
accepted codes. Particular points of concern include local buckling limit states, hybrid
beam-columns and biaxial bending.
For the PRACTICE it was desirable to produce a code simple to use, unambiguous and
as close as possible to AISC LRFD to avoid the need for validation. Except for the
sections on shell members, the code uses the same equations as given in AISC LRFD
apart from two areas of extension, relating to beam-column biaxial bending and hybrid
beam-columns, discussed below. For instance, the same ranges of D/t ratios have been
used as specified in AISC to define the different ranges of limit state. These ranges
may not be instantly recognizable since they been given in terms of R/t ratios in many
places. This is considered more appropriate for jack-up members which may contain
partial tubular sections for which a section radius is a more meaningful quantity than a
diameter.
Most sections are drawn from the numerous chapters and appendices in AISC LRFD
and are placed in a logical sequence. In some cases it has been necessary to interpret
which equations are applicable especially in the area of local buckling for beam-
columns.
AISC LRFD does not cover very high D/t ratio tubes and tubes with stiffeners, so
reference has been made to relevant sections of a different code. The "DNV Rules for
Classification - Fixed Offshore Installations" [13] was selected as the most suitable,
since this is a limit state code. This document refers the user to "DNV Classification
Notes - Note 30.1" [14] for obtaining member resistances or strengths. Some guidance
on the use of these notes is given in section C8.1.5.
Hybrid beam-columns
Hybrid beam-columns are quite common in jack-ups such as chords with high yield
stress racks welded to lower yield stress plate constructions. The treatment of hybrid
beams of this nature is not adequately covered in AISC LRFD and hence it has been
necessary to state rules dictating the method for establishing the axial and bending
strengths of such beam-columns. The methods described are based upon engineering
understanding of the problem and have been made as straight forward as possible.
Limitations
The first limitation, (a) has been stated as a warning that the code given in the
PRACTICE must be restricted to the type of geometries as described in section 8.1.4.
The intention has been to cover all of the geometries likely to be encountered in jack-
ups although there may be some exceptional cases. If this is so, the user must refer to
AISC LRFD [11]. One notable exception could be 'I' type sections which are
sometimes used in jack house frames. Since AISC LRFD is oriented towards the
assessment of 'I' type sections, it is reasonably straight forward to use. It is
recommended that the equations given in Appendix H are used for 'I' sections since
these should give less conservative results than the general equations in Chapter H.
The limitation (b) is stated in AISC LRFD [11] for the reasons that experimental
validation has not been carried out for steels with higher yield stresses than 100 k.s.i.
The equations may not be valid for higher yield stresses although there does not appear
to be any theoretical reason for this to be the case. However, if higher yield stress
steels are to be assessed using the practice, it will be necessary to validate the equations
for whatever yield stress is used in the design. Currently steels with yield stresses
greater that 100 k.s.i. are not generally encountered in jack-ups with the exception of
the mechanical components in the holding system which are to be treated under other
assessment criteria.
C8.1.2 Definitions
Structural Members and Components
The definitions of members and components have been stated so that the appropriate
analysis method can be used for a particular type of geometry. The emphasis of the
assessment is on structural members, for which loads and properties must be known.
Components are assessed only when the section is identified as being prone to local or
lateral torsional buckling. This type of specification is in keeping with conventional
jack-up analysis procedures in which chord scantlings are modeled as single beams; the
modeling of the individual plates not being necessary.
C8.1.3 Factored Loads
The load factors used in AISC are inappropriate for jack-up analysis since these have
been derived for land based buildings. The derivation of load factors specific to jack-
ups is discussed in Section C8.0.
C8.1.4 Assessment of Members - excluding stiffened and high D/t ratio tubulars
C8.1.4.1 General interaction equations
The treatment of biaxial bending in AISC LRFD tends to be conservative for beam-
columns laterally supported at both ends. This is most apparent when assessing a
tubular member. The bending strength of the tube must be the same in all directions,
but this is not reflected in the AISC LRFD equations. The linear addition of x- and y-
axis bending moment terms in effect reduces the nominal bending strengths in all cases
of biaxial bending. For example, a tubular member subject to bending in a plane at 45°
to the x-axis has in the AISC LRFD code a nominal strength of 71% of that for uniaxial
bending in the x- or y- planes.
The problem is not confined to tubulars, as most sections would have a reduction in
nominal strength on account of this linear addition. Only for `I' sections does AISC
LRFD allow a more liberal formulation. In the AISC ASD formulation, such a problem
does not arise, as the stress points are considered explicitly.
Since optimal design is required for jack-ups, and since previous ASD-based design did
not suffer from this problem, it was considered necessary to remove the conservatism
for the PRACTICE. The general interaction equations have therefore been modified
from the AISC LRFD equations.
In deriving a suitable form the problem for the tubular was considered first. Clearly,
since the tubular has equal bending strength in all directions, the correct actual bending
moment should be the vectorial sum of the x- and y-axis bending moments. Expressed
as a unity check equation for bending only:
2
⎧ M ux ⎫ ⎧⎪ M uy ⎫⎪
2
⎨ ⎬ +⎨ ⎬ ≤ 10
.
⎩ φ b M nx ⎭ ⎪⎩ φ b M ny ⎪⎭
and with the addition of axial load (for Pu/φaPn > 0.2)
8 ⎛⎜ ⎧ M ux ⎫ ⎪⎧ M uy ⎪⎫ ⎞⎟
2 2 2
Pu
+ ⎨ ⎬ +⎨ ⎬ ≤ 10
.
φ a Pn 9 ⎜⎝ ⎩ φ b M nx ⎭ ⎩⎪ φ b M ny ⎭⎪ ⎟⎠
Since most jack-up chords are closed sections with high torsional stiffnesses similar to
tubulars, the logical step was to formulate a similar equation which had the ability to
account for sections not exhibiting circular symmetry. This was carried out by using a
generalized exponent η to form the two equations given in the PRACTICE. One of the
equations is given below as an example (for Pu/φaPn > 0.2). This resembles the
formulation in the AISC LRFD for I sections, although the exponent η has a different
determination procedure.
8 ⎛⎜ ⎧ M ux ⎫ ⎧⎪ M uy ⎫⎪ ⎞ η
η η
Pu
+ ⎨ ⎬ +⎨ ⎬ ⎟ ≤ 10
.
φ a Pn 9 ⎜⎝ ⎩ φ b M nx ⎭ ⎪⎩ φ b M ny ⎪⎭ ⎟⎠
With η = 1.0, the equations revert to the standard AISC LRFD equations, and hence a
conservative assessment can be made. However, if the limit is required with more
accuracy, then it is necessary to determine the value for η (discussed later).
If the nominal bending strengths Mux and Muy are the same and η = 2.0, then this
would imply that the section has equal bending strength in all directions. A value to η
= ∞ implies that the bending capacities in the x- and y-axes are independent of each
other. Favorable interaction between, for example, the -Mx and +My moments acting
on triangular chords with a single rack cannot be reproduced by the above equation.
In such cases recourse to the section-specific interaction surface is recommended (see
Section C8.1.4.7).
C8.1.4.2 Nominal Axial Strength
C8.1.4.3
Whereas the nominal axial strength of tension members of one material are fully
catered for in the AISC LRFD code, some interpretation was required hybrid beam-
columns.
The basic measure of tensile strength is 0.9Fyi, but in certain cases this value may be
unacceptably close to the ultimate strength. Therefore the provision is introduced that
the factored strength is the lesser of 0.9Fyi and 0.75Fui. This ensures that an acceptable
margin is applied to each component as illustrated in Figures C8.1.1 and C8.1.2.
Figure C8.1.1 : ultimate strength Figure C8.1.2 : yield strength

much bigger than yield strength close to ultimate strength
For hybrid members a nominal strength is required that takes into account the
properties of each component. If there is no likelihood of fracture of any one
component then an addition of the nominal strengths of each component is appropriate
for the member (Figure C8.1.3.).
Figure C8.1.3 : Stress/strain curves for two component member for which
addition of nominal strengths is permissible
However it is conceivable that fracture of one component may take place at a strain
level below that at which another component is loaded to its nominal strength
(Figure C8.1.4.).
Figure C8.1.4 : Stress/strain curves for two component member in which

one component fractures before the other is loaded to its nominal
strength.
For such an eventuality it is stipulated that the strength of the whole section is that for
the weakest component, applied across the whole section, so that:
Pn = FminΣAi
This formulation is suitable when component materials are similar. When material
properties differ widely from component to component then the formulation may be
over conservative, and a rational analysis may be preferred. An example follows.
Example
Consider a member in tension. The section consists of a rectangular portion of steel 1

sandwiched between two rectangular portions of steel 2. The areas occupied by steels 1
and 2 are both half the total section area A. By symmetry the section is balanced.
Figure C8.1.5 : Stress/strain curves for

components of example member
Consider the portion of steel 1 as component 1 and the portions of steel 2 as component
2. The stress/strain plots of the materials (Figure C8.1.5.) show that the strain level for
component 1 to reach its nominal strength is well below that for fracture of 2. The
ductility of 2 means that the component can support a stress of just over Fn2 for strains
up to those at which component 1 reaches its nominal capacity. Therefore, a less
conservative nominal strength for the member is:
Pn = Fn1A/2 + Fn2A/2
Because the section is balanced, plastic deformation of 2 does not induce any extra
loads or moments on the member. Were the section not balanced, then this would not
be true. It is essential that such aspects are considered in a rational analysis of strength.
C8.1.4.4 Effective Applied Moment

C8.1.4.5
The PRACTICE allows for structural analyses of a range of levels of sophistication. In
some, it may be necessary to manipulate the calculated moment to produce a more
"true" value for application in the PRACTICE. This leads to the use of the effective
applied moment for members with compressive axial loads. Adjustments for tensile
axial loads are not considered significant.
It has been noted that the P-Δ effect produces an extra moment on a leg under hull
sway, and that this moment should be included in the structural analysis. The similar,
local P-Δ effect on the individual members of truss leg must also be included, directly
in the structural analysis, or by use of the B term. If for a truss, the structural analysis
includes the local P-Δ effect then no manipulation is required. If the local P-Δ is not
included, for example through a linear elastic analysis of the leg segment, then the
amplifier B is required. For many non-truss leg cases there is no local P-Δ effect, such
as for a jack-up with large diameter tubular legs.
The use of the single B term differs from that in the AISC LRFD code. There, the first
order moment is separated into two parts: a moment assuming no lateral deflection of
the frame Mnt, and a moment attributed only to lateral deflection Mlt. Then the
effective applied moment is the sum of B1Mnt and B2Mlt, where B1 is similar to B in the
PRACTICE and B2 is a second coefficient. It is important to note that both these
moments are first order, and do not include P-Δ. The use of B1 and B2 is to simulate
the P-Δ effects at local and global level respectively.
Therefore, in the PRACTICE the calculated applied moment is not the same as the Mu
in the AISC LRFD code. The use of
Mue = B Mu
performs the necessary step of adding the local P-Δ moment to the calculated moment
which already includes global P-Δ.
Note that in a plastic analysis, yielding can take place within the members and bending
moments can hence be redistributed. The types of analysis to be used for the structural
assessment of jack-ups are to be elastic analyses where yielding does not take place, so
this aspect is not covered. For reference, AISC LRFD states their code is only valid for
plastic analysis if material yield stresses do not exceed 65 k.s.i.
C8.1.4.6 Nominal Bending Strength
The calculations of nominal bending strength for compact and noncompact sections
require knowledge of the plastic moment capacity of the section. For a section
composed of uniform material this is given by the lesser of:
Mp = FyZ
and
Mp = 5/6FuZ
where Z is the plastic section modulus. For hybrid sections there is more than one set
of material properties to consider. Standard techniques are recommended for
evaluation of Mp and an example is provided below.
Example
Consider the simplified problem of a square rack section (component 1) of properties:

Fy1 = 700 MN/m2
Fu1 = 828 MN/m2 ; Fy' = 828 x 5/6 = 690 MN/m2
connected to a solid square chord section (component 2) of properties:
Fy2 = 345 MN/m2
Fu2 = 485 MN/m2 ; Fy' = 485 x 5/6 = 404 MN/m2
as shown in Figure C8.1.6 below. The nominal strengths are the lesser of the Fyi and
5/6Fui, namely:
component 1 strength = 690 MN/m2
component 2 strength = 345 MN/m2
Dimensions are as marked.
Figure C8.1.6 Example Figure C8.1.7 Fully plastic

hybrid chord section stress distribution
On the assumption that the strain for component 1 to be loaded to its nominal strength
is not sufficient to lead to fracture of component 2, the plastic stress distribution for
pure bending is as shown in Figure C8.1.7. The Plastic Neutral Axis is a distance zo
from the back face of the chord component, such that:
345 x 0.3 x zo = 345 x 0.3 x (0.3-zo) + 690 x 0.1 x 0.1
i.e.
zo = 0.183 m
The section plastic moment is then:
Mp = 345 x 0.3 x 0.183 x (0.183/2)
+ 345 x 0.3 x 0.117 x (0.117/2)
+ 690 x 0.1 x 0.100 x (0.117 + 0.100/2) = 3.59 MNm
C8.1.4.7 Determination of η
Determination of the correct value of η is carried out by calculation of the nominal

strength of the member about axes other than the x- and y-axes. This can be done in
the normal manner based on the effective plastic section modulus with reductions for
local buckling if applicable. Although a beam will not necessarily bend in the same
plane as the applied moment when the bending plane is at an angle to the orthogonal
axes, it is not expected that the capacity will be greatly affected.
Once the nominal bending strength has been calculated for a few angles between the x-
and y-axes, the value for η can be calculated using the graphical procedure given in the
practice, or by an iterative procedure. A successful iterative procedure was found to be
by the use of the coupled equations, setting a = M'uex/Mnx and b = M'uey/Mny:
1n(1 − b ηi )
ηi+1 =
1na
with the accelerating step:
ηi+2 = 0.5(ηi+1 + ηi)
and the initial value η = 1.5.
The three angles which were chosen, 30°, 45° and 60° give a good spread over the 90°
range. It is not the intention to fit a curve through all the values from the three angles
but merely find the lowest value to η. This may still make the equation conservative
although considerably less so than for η = 1.0.
Plastic Interaction Curve Approach

Alternatively, interaction equations and curves for generic families of chords are
presented in Figures C8.1.8 - C8.1.11. The offset distance between the elastic
centroid (used in the structural analysis) and the 'center of squash', together with
other geometric data for the members of each family of chord is presented in
Tables C8.1.1 to C8.1.4.
C8.1.5 Assessment of Members - other geometries
For high D/t ratio tubulars, reference is made to the DNV Rules for Fixed Offshore
Installations, as these are based on a suitable LRFD format. Care must be taken to
adapt the usage factors in the Rules to the correct resistance factor format.
The buckling strength of shells is best described in terms of buckling stress. For this
reason, the stress to cause buckling in the shell must be determined and compared with
the stresses caused by the factored loads. Since the analysis model usually gives the
overall member loads, it is necessary to calculate the stresses in the shell. It may be
possible, with caution, to allow the analysis model to also calculate membrane stresses.
The detailed stress formulations in the DNV Class note 30.1 are amenable to some
simplification. For example, the pressure loading terms may usually be omitted, since
high D/t tubulars in jack-up legs are generally flooded.
For beam-column interaction, the effects from global axial buckling are added to the
effects of local buckling due to flexural bending. Global buckling effects in bending
such as lateral-torsional buckling only occur in sections in which the stiffness out of
plane is less than the stiffness in the plane of bending. Thus tubular and rectangular
sections, hollow or otherwise, in which the depth is less than or equal to the width do
not suffer lateral torsional buckling.
Strength Interaction Equations
1/ 2
⎧⎪⎛ M ⎞ 2 ⎛ M ⎞ 2 ⎫⎪
⎨⎜⎜ ⎟⎟ + ⎜⎜ y ⎟⎟ ⎬ ≤ 100
x
.
⎪⎩⎝ M' px ⎠ ⎝ M' py ⎠ ⎪⎭
0.7
⎧⎪ ⎛ π. P ⎞ ⎫⎪
For (P/Py) ≤ 0.6: M'px = Mpx ⎨cos⎜⎜ ⎟⎟ ⎬
⎪⎩ ⎝ 2. Py ⎠ ⎪⎭
11
.
⎧⎪ ⎛ π. P ⎞ ⎫⎪
M'py = Mpy ⎨cos⎜⎜ ⎟⎟ ⎬
⎪⎩ ⎝ 2. Py ⎠ ⎪⎭
⎛ P⎞
For (P/Py) > 0.6: M'px = 1.71Mpx ⎜⎜1− ⎟⎟
⎝ Py ⎠
⎛ P⎞
M'py = 1.39Mpy ⎜⎜1− ⎟⎟
⎝ Py ⎠
Figure C8.1.8 Interaction equations/curves for tubular chords

with double central racks.
Chord Dimensions - Tubular Chord with Central Double Racks
All dimensions are in millimeters, Yield Stresses are in MPa

<--- Yield --->
Stress
Design L1 t1 L2 t2 D t3 Fy1 Fy2 Fy3 Bay Ht
BMC JU-300-CAN (Zapata Scotian) 991 127 0 0 914 44 690 0 690 5532
48
CFEM T2001 (Hitachi Redesign) 960 18 121 140 960 52 690 690 690 4500 Btm 3 bays
34 4100 Top 3 bays
26 4050 Middle bays
34
42
CFEM T2005 650 20 108 140 800 28 700 685 650 or 5050
30 700*
31
32
33
35
36
38
40 700 685 700
44
34 700 685 650
38
42
* Note: Early CFEM T2005 designs use 650 MPa steel for tube, later designs use 700 MPa steel.
…continued
Table C8.1.1 Data for tubular chords with double central racks
<--- Yield --->

Stress
Design L1 t1 L2 t2 D t3 Fy1 Fy2 Fy3 Bay Ht
CFEM T2600 650 20 120 140 800 33 700 700 700 6000
35
38
40
41
43
45
47
49
50
51
52
55
56
57
58
MODEC 200 450 15 102 127 559 27 490 690 490 5486
27
MODEC 300 450 25 102 127 559 34 490 690 490 5486
28 34
15 40
20 40
27 40
60 40
115 40
MODEC 400 (Trident 9) 690 20 102 127 800 30 490 690 490 6200
20 35
35 35
Hitachi K1025/31/32 900 18 100 127 900 32 690 690 690 5160
18 36
18 50
20 40
20 42
Hitachi K1026 (Neddrill 4) 950 18 100 127 950 32 690 690 690 4360
18 36
20 42
Hitachi K1056/7 1000 28 130 178 1000 47 690 730 690 4600
30 50
30 52
30 60
30 64
60 60
60 64 4000
ETA Robray 300 (Asia Class) 627 10 127 127 762 22 690 690 690 5486
11
13
14
16
17
19
25
32
ETA Europe Class 627 38 140 140 762 22 690 690 690 5486
51
64
76
89
102
114
127
Table C8.1.1 (Continued) Data for tubular chords with double central racks
1/ 2
⎧⎪⎛ M ⎞ 2 ⎛ M ⎞ 2 ⎫⎪
⎨⎜⎜ ⎟⎟ + ⎜⎜ y ⎟⎟ ⎬ ≤ 100
x
.
⎪⎩⎝ M' px ⎠ ⎝ M' py ⎠ ⎪⎭
⎛ ⎛P ⎞ ⎞
2.25
where; M'px = Mpx ⎜⎜1 − ⎜ ⎟ ⎟

⎟
⎜P ⎟
⎝ ⎝ y ⎠ ⎠
⎛ ⎛P ⎞ ⎞⎟
1.85
M'py = Mpy ⎜⎜1 − ⎜ ⎟

⎜P ⎟ ⎟
⎝ ⎝ y ⎠ ⎠
Figure C8.1.9 Interaction equations/curves for split tubular chords with opposed central racks
(doubly symmetrical)
Chord Dimensions - Split Tubulars with Double Central Racks

Yield
Stress
Design L1 t1 D t2 t3 L4 t4 Y1 H1 H2 Fy1 Fy2 Bay Ht
F & G L780 (Lower bays) 400 152 381 25 25 0 0 0 191 165 621 690 3658
F & G L780 (Upper bays) 400 127 381 25 25 0 0 0 191 191 621 450 3658
F & G L780 m2 (Lower bays) 400 152 381 32 32 0 0 0 191 165 621 690 3658
F & G L780 m2 (Upper bays) 400 127 381 32 32 0 0 0 191 191 621 517 3658
F & G L780 m5 (Monitor) 401 178 381 81 57 0 0 51 178 178 690 690 4267
F & G L780 m5 (Monarch) 401 178 381 81 51 0 0 51 178 178 690 690 4267
F & G L780 m6 611 178 584 83 38 0 0 95 292 292 690 690 5486
MSC CJ62 (Lower bays) 650 210 600 65 48 0 0 75 270 270 690 690 6927
MSC CJ62 (Upper bays) 650 210 600 55 40 0 0 75 270 270 690 690 6927
MSC CJ50 (1) 550 210 520 25 25 0 0 0 260 260 690 690 5608
MSC CJ50 (2) 550 210 520 25 35 0 0 0 260 260 690 690 5608
Technip TPG 500 (1) 722 160 680 75 61 0 0 20 340 340 690 540 6000
Technip TPG 500 (2) 722 160 680 75 37 0 0 55 340 340 690 540 6000
Technip TPG 500 (3) 722 160 680 62 37 0 0 36 340 340 690 540 6000
Technip TPG 500 (4) 722 160 680 58 37 0 0 30 340 340 690 540 6000
Technip TPG 500 (5) 722 160 680 50 37 0 0 19 340 340 690 540 6000
Technip TPG 500 (6) 722 160 680 50 37 510 30 19 340 340 690 540 6000
Table C8.1.2 Data for split tubular chords with double central racks

1/ ξ
⎧⎪⎛ M ⎞ ξ ⎛ M ⎞ ξ ⎫⎪
⎨⎜⎜ ⎟⎟ + ⎜⎜ y ⎟⎟ ⎬ ≤ 100
x
.
⎪⎩⎝ px ⎠
M ' ⎝ M' py ⎠ ⎪⎭
⎧⎪ ⎛ P ⎞ 1.45 ⎫⎪
where; M'py = Mpy ⎨1 − ⎜⎜ ⎟⎟ ⎬
⎪⎩ ⎝ Py ⎠ ⎪⎭
2 3
⎛ P⎞ ⎛ P⎞ ⎛ P⎞
When Mx ≥ 0: . + 2.7⎜⎜ ⎟⎟ + 2.8⎜⎜ ⎟⎟ − 5.6⎜⎜ ⎟⎟
ξ = 18
⎝ Py ⎠ ⎝ Py ⎠ ⎝ Py ⎠
1/112
.
⎧⎪ ⎛ P ⎞ 112
.
⎫⎪
and M'px = Mpx ⎨1 − ⎜⎜ ⎟⎟ ⎬
⎪⎩ ⎝ Py ⎠ ⎪⎭
When Mz < 0: ξ = 1.8
and for (P/Py) ≤ 0.25: M'px = -Mpx

⎧⎪ ⎛ P 1⎞ ⎫⎪
1.45
for (P/Py) > 0.25: M'px = -Mpx ⎨1 − ⎜⎜ − ⎟⎟ ⎬

⎪⎩ ⎝ 0.75Py 3⎠ ⎪⎭
Figure C8.1.10 Interaction equations/curves for tubular chords with offset double racks.
Tubular Chords with Offset Double Racks
<--- Yield --->

Stress
Design D t1 L1 L2 t2 t3 Fy1 Fy2 Fy3 Bay Ht Yena Ycos E
Levingston 011-C 914 29 305 906 127 0 483 621 0 4826 84 100 16
33 75 90 15
Levingston 111 1016 32 305 1047 127 0 690 690 0 4877 73 73 0
35 68 68 0
Mitsui JC-300 (Key Hawaii) 1016 32 305 1046 127 0 690 690 0 5650 78 78 0
34 0
35 66 66 0
Mitsui 1-off (Key Bermuda) 1016 29 305 1046 127 0 690 690 0 4672 0 Most of leg
1016 29 305 1046 127 0 690 690 0 5050 0 Towage Section
32 5050 73 73 0 "
36 5050 66 66 0 "
Hitachi Drill-Hope 762 30 190 882 127 0 690 690 0 5500 57 57 0
32 55 55 0
Hitachi C-150 (Ile Du Levant) 762 30 190 890 130 0 690 690 0 5500 60 60 0
Hitachi K1040/44/45 900 30 300 882 127 0 690 690 0 4800 77 77 0 Btm 2 bays
30 5090 77 77 0 Rest of leg
35 5090
42 5090 60 60 0
Hitachi K1060 (Sagar Lakshmi) 900 30 300 854 127 13 690 690 690 5260 84 84 0
31 0
32 0
34 77 77 0
Robco 350-C 876 29 292 881 127 0 690 690 0 5461 83 83 0 Btm 3 bays
876 38 68 68 0 "
864 29 89 89 0 Rest of Leg
864 32 82 82 0 "
Table C8.1.3 Data for tubular chords with offset double racks

1/ ξ
⎧⎪⎛ M / M − K ⎞ ξ ⎛ M ⎞ ξ ⎫⎪
⎨⎜⎜ ⎟⎟ + ⎜⎜ y ⎟⎟ ⎬ ≤ 100
x px
.
⎪⎩⎝ M ' px / M px − K ⎠ ⎝ M ' py ⎠ ⎪⎭
2 3
⎛ P⎞ ⎛ P⎞ ⎛ P⎞
where; K = −0.8⎜⎜ ⎟⎟ + 0.4⎜⎜ ⎟⎟ + 0.4⎜⎜ ⎟⎟
⎝ Py ⎠ ⎝ Py ⎠ ⎝ Py ⎠
⎧⎪ ⎛ P ⎞ 2.1 ⎫⎪
and M ' py = M py ⎨1 − ⎜⎜ ⎟⎟ ⎬
⎪⎩ ⎝ Py ⎠ ⎪⎭
⎧⎪ ⎛ P ⎞ 1.45 ⎫⎪
When (Mx/Mpx) ≥ K: M ' px = M px ⎨1 − ⎜⎜ ⎟⎟ ⎬
⎪⎩ ⎝ Py ⎠ ⎪⎭
and ξ = 1.45
1/1.04
⎧⎪ ⎛ P ⎞ 1.04 ⎫⎪
When (Mx/Mpx) < K: M ' px = − M px ⎨1 − ⎜⎜ ⎟⎟ ⎬
⎪⎩ ⎝ Py ⎠ ⎪⎭
2
⎛ P⎞ ⎛ P⎞
. + 2.35⎜⎜ ⎟⎟ + 4.7⎜⎜ ⎟⎟
and ξ = 145
⎝ Py ⎠ ⎝ Py ⎠
Figure C8.1.11 Interaction equations/curves for triangular chords

with single racks
chords with single racks
Table C8.1.4 Data for triangular
Chord Dimensions - Triangular Chords with Single Rack

<-------- Yield Stress -------->
Design L1 t1 L2 t2 L3 t3 L4 t4 L5 t5 t6 X1 Y1 Y2 Y3 Fy1 Fy2 Fy3 Fy4 Fy5 Bay Ht Yena Ycos E
MarLet Standard (3/4" side plates) 711 51 466 19 213 127 0 0 0 0 0 236 457 0 0 483 483 587 0 0 3408 259 279 20
MarLet Standard (7/8" side plates) 711 51 466 22 213 127 0 0 0 0 0 236 457 0 0 483 483 587 0 0 3408 259 279 20
MarLet Standard (1" side plates) 711 51 466 25 213 127 0 0 0 0 0 236 457 0 0 483 483 587 0 0 3408 260 279 19
MarLet Standard (1.5" side plates) 711 51 466 38 213 127 0 0 0 0 0 236 457 0 0 483 483 587 0 0 3408 262 279 17
MarLet Standard + side stiffeners 711 51 466 19 213 127 0 0 127 25 0 236 457 0 211 483 483 587 0 483 3408 260 279 19
MarLet Std 116 (1"x4" rack stiffeners) 711 51 466 19 213 127 102 25 0 0 0 236 457 524 0 483 483 587 483 0 3408 278 296 18
MarLet 116 North Sea (1"x4"+1"x12" stifnrs) 711 51 466 19 213 127 102 25 305 25 0 236 457 524 118 483 483 587 483 483 3408 276 291 15
MarLet 116 (1"x4"+1.5"x12" stifnrs) 711 51 466 19 213 127 102 25 305 38 0 236 457 524 124 483 483 587 483 483 3408 277 291 14
MarLet 116 Juneau (2"x4"+1"x12" stifnrs) 711 51 466 19 213 127 102 51 305 25 0 236 457 524 118 483 483 587 483 483 3408 290 304 14
MarLet Gorilla (150-88) 813 76 573 57 222 140 0 0 0 0 0 248 600 0 0 483 483 620 0 0 5113 302 323 21
MarLet Super 300 813 76 607 38 222 140 0 0 0 0 0 268 600 0 0 483 483 620 0 0 5113 298 323 25
813 76 607 38 222 140 0 0 305 51 0 268 600 0 296 483 483 620 0 483 5113 327 346 19
MarLet 300 Slant 711 64 441 38 213 127 0 0 0 0 0 218 457 0 0 414 414 414 0 0 2556 245 245 0
LeTourneau 150 (3/4" side pl) 711 51 466 19 213 127 0 0 0 0 0 236 457 0 0 414 414 620 0 0 2556 259 302 43
LeTourneau 150 (1.125" side pl) 711 51 466 29 213 127 0 0 0 0 0 236 457 0 0 414 414 620 0 0 2556 259 303 44
LeTourneau 150 (1.5" side pl) 711 51 466 38 213 127 0 0 0 0 0 236 457 0 0 414 414 620 0 0 2556 262 298 26
LeTourneau 46,47 559 44 432 13 178 89 0 0 0 0 0 166 432 0 0 ? ? ? 0 0 3408 224 224 0
LeTourneau 4,9 559 51 565 13 197 102 0 0 0 0 0 178 533 0 0 ? ? ? 0 0 3430 286 286 0
Mitsubishi MD-T76J 750 50 574 25 225 125 0 0 0 0 0 226 575 0 0 687 687 687 0 0 3456 315 315 0
Gusto 1-off: (Maersk Endeavour) 800 60 592 30 283 127 0 0 0 0 0 359 443 0 0 620 620 620 0 0 4800 284 284 0
800 90 534 40 283 127 0 0 0 0 0 331 443 0 0 620 620 620 0 0 4800 255 255 0
Rev 3, August 2008

800 110 488 50 283 127 0 0 0 0 0 307 443 0 0 620 620 620 0 0 4800 244 244 0
Gusto 1-off: (Maersk Explorer) 800 76 535 38 279 127 0 0 0 0 0 337 453 0 0 690 690 690 0 0 4539 268 268 0
800 64 549 38 279 127 0 0 0 0 0 344 453 0 0 690 690 690 0 0 4539 282 282 0
800 51 562 32 279 127 0 0 0 0 0 351 453 0 0 690 690 690 0 0 4539 297 297 0
800 51 562 29 279 127 0 0 0 0 0 351 453 0 0 690 690 690 0 0 4539 297 297 0
BMC 1-off design (Trident 7) 711 38 356 19 279 127 0 0 0 0 25 264 204 0 0 ? ? ? 0 0 3353 197 197 0
Page 184
C8.3 FOUNDATION ASSESSMENT
The intention of the foundation capacity checks of steps 1 and 2 (Sections 8.3.1 and
8.3.2/8.3.3 of the PRACTICE) is to safeguard against foundation failure. Foundation
failure will, in most cases, manifest itself through excessive spudcan vertical and/or
horizontal displacements which may cause local or global instability of the jack-up.
Local instability occurs when a leg becomes overstressed, with global instability as a
consequential effect. Global instability may occur through overturning which will then
cause leg overstress. The key to preventing either type of failure mode is to safeguard
against excessive spudcan displacements.
Since it is difficult to accurately compute the displacements (as proposed in Step 3,

Section 8.3.4 of the PRACTICE, the checks of steps 1 and 2 are performed by
comparing the bearing capacity with the extreme combinations of load, including
applicable partial factors. In steps 1 and 2a the loads are computed assuming pinned
footings. In step 2b the check allows for fixity.
Selection of the Resistance Factors φ

The resistance factor φ is intended to cover uncertainties in the estimation of the
bearing capacity. In foundation engineering it is common to adopt φ in the range 1/1.25
- 1/1.30 (0.80 - 0.77) when the capacity is determined on the basis of available soil data
and analytical predictions. For a jack-up there is however additional information
available in terms of the preload applied at installation, which generally justifies the use
of a higher resistance factor.
During preloading the spudcan foundation experiences loading similar, but not identical
to the conditions of the leeward leg during the extreme event. Taking this information
into account, there is greater certainty in the upper part of the bearing capacity curve
applicable to bearing failure than in the lower part applicable to sliding failure.
However, some uncertainty still remains for the foundation capacity in bearing
applicable to the leeward leg determined from the preload value. This uncertainty is
due to factors such as:
- effect of cyclic loading
- effects of consolidation and creep
- loading rate effects
There is at present insufficient information available to fully quantify the likely
distribution in the actual foundation capacity curve. In the study performed by NGI
[17] it is concluded that cyclic degradation effects on clay are significantly larger for
the leeward leg than for the windward leg. It is also noted that the case of a spudcan
which has not penetrated sufficiently to mobilize the maximum available bearing area
should be differentiated from the case where the maximum bearing area is utilized.
This is because in the former case a small additional penetration will lead to a increase
in capacity as a result of the increase in bearing area.
On the basis of the above arguments the following resistance factors are proposed:
Step 1a - preload, vertical load alone:

φ = 0.9 See note overleaf
Note: Section 8.3.1.4 of the PRACTICE requires that the vertical and horizontal load
check of step 2a is made when the horizontal leg reaction at the leeward leg
exceeds prescribed limits, depending on the penetration and soil. This is
because the simplistic check in Step 1a is based on the proven ultimate vertical
bearing capacity during preloading and it is therefore assumed that the extreme
footing load is the same as the maximum footing load during preloading. This
implies that the horizontal loading on the spudcan under extreme conditions is
small and it is therefore appropriate to limit the combined horizontal and
vertical loading to the values permitted under Step 2. In the selection of the
limits for Step 1a two penetration cases can be distinguished:
- full embedment to maximum bearing area in foundation layer,
- partial embedment in the foundation layer.
For full spudcan embedment in sand the lateral soil resistance at a vertical load
of 0.9VLo is approximately 0.03VLo. Additional penetration may increase the
soil resistance, but to increase the horizontal resistance to 0.1VLo the additional
penetration will be in the order of 10% of the spudcan diameter and outside
tolerable limits.
In the case of partial penetration of the spudcan in sand (i.e., full bearing area
not mobilized), any additional penetration will result in a significant increase of
bearing capacity due to the rapid increase in the bearing area. An increase in
embedded area of approximately 10% will increase the horizontal capacity to
0.1VLo.
In clayey soils the requirement of QH < 0.1VLo is met if the ratio of the spudcan
laterally projected area to bearing area, AS/A is in the order of 0.3.
Step 1b - sliding, vertical and horizontal load vector:
φHfc = 0.8 (effective stress - sand/drained)
= 0.64 (total stress - clay/undrained)
Step 2a - bearing, vertical and horizontal load vector:
φVH = 0.9 (maximum bearing area not mobilized)
= 0.85 (maximum bearing area mobilized)
Step 2b - vertical, horizontal and moment load vector:
φVHM = φVH from step 2a for leeward legs
= φHfc from step 1b for windward legs
Selection of safety factors against punch-through
Where the potential for punch-through foundation failure is recognized, detailed
consideration regarding foundation integrity will be required. Methods have been
proposed for punch-through installation procedures and acceptability criteria but are
omitted from this document as they remain ambiguous, (Rapaport [18], Senner [19]).
Some jack-up designs are more able to tolerate rapid leg penetration than others and if
the magnitude of the potential leg plunge is acceptable then installation could be
possible even though punch-through is predicted during preloading. Significant
investigation will be required in such circumstances and it is recommended that each
potential punch-through situation is assessed on its own merit both at preloading and,
should the potential for punch-through remain after installation, for the elevated
operational and survival conditions.
C8.7 STRUCTURE CONDITION ASSESSMENT
C8.7.1 Introduction
Where an on-site structural inspection is required Section C8.7.2 provides guidance as

to how this may be carried out. Section C8.7.3 provides guidance on monitoring the
structural condition during an assignment.
C8.7.2 Scope of Condition Assessment
The aim of a condition assessment is to verify the condition of the structural

components and details that are essential for the ultimate strength of the jack-up in the
elevated condition and to confirm that the condition of the jack-up structure is in line
with the assumptions made in the site specific analysis. Structural details to be
assessed can be identified based on general considerations (see [20]), or on the results
of site specific calculations and/or fatigue considerations.
Condition assessment may comprise four steps with increasing involvement as

described below. It is necessary that step 1 should always be completed and steps 2, 3
and 4 should be carried out as necessary to validate the condition of the jack-up.
Step 1 - Review of existing records

The review should initially cover the existing certificates and operating/inspection
records. These should provide details of any incidents (indications of damage and
subsequent repairs).
Step 2 - Visual inspection

A general visual inspection is carried out to confirm the rig is in a good state of repair
and well maintained. The inspection should focus on the critical structural components
and will include checks for missing members, mechanical damage, corrosion, etc.
Normally this will only cover above water areas and should include a selection of the
fatigue sensitive areas listed in Section 7.4.4.
The visual inspection may be carried out by the same team visiting the unit for a pre-
contract inspection (safety, drilling, etc.). However it is required that qualified
personnel should be part of that team.
Step 3 - Close Visual inspection

Close visual inspection of key structural areas may be required if the condition can not
be validated based on the review of existing records (step 1) and the general visual
inspection (step 2). Hence the close visual inspection should focus on specific areas or
details that may be in doubt.
If fatigue is a consideration, it is recommended that, following steps 1 and 2, a selection

of the fatigue sensitive areas should be subjected to close visual inspection. These
close visual (weld) inspections should cover a number of locations from each of the
groups of fatigue sensitive areas identified in Section 7.4.4. This should possibly be
followed with an MPI inspection as outlined below.
C8.7.2 Step 4 - Detailed inspection
Detailed inspection may be required in special cases and in case of inconclusive

findings after completion of the three steps above. The detailed inspection is seen as a
continuation of the visual inspection and hence is aimed at confirming that the
condition of some specific details are sound for the intended operation. The major
concern is the ultimate strength of the unit. Local damage with no significance for the
planned mode of operation need not be assessed, but should be recorded for inclusion
in future maintenance work.
The scope of the inspection may include some NDT (e.g. MPI) and these inspections
must be carried out by qualified personnel. The NDT will normally cover areas of
specific concern. It may also be necessary to provide some spot checks of fatigue
sensitive areas. In this case it is recommended that MPI checks should, as a minimum,
be made at a selection of areas from each of the groups of fatigue sensitive locations
identified in Section 7.4.4.
If defects are found it may be necessary to expand the scope of the inspection so that
the full extent of the damage can be assessed.
C8.7.3 Condition Monitoring
The condition of the jack-up structure should be monitored during the assignment.
This is to ensure the continuation of the overall structural integrity during the
operations. The requirements for condition monitoring may be based on the approach
outlined in steps 1 through 4 above. The operating and maintenance records kept by
the owner are the primary source of input for the independent condition monitoring and
it should be possible to validate the condition of the unit by reviewing these records at
any time during the operation. It is not expected that, for normal operations, the scope
of the condition monitoring should extend beyond step 1.
It is noted that records should be kept for future reference.

a = M'uex/Mnx used in determination of η.

A = constant in quadratic expression for β.
A = cross sectional area.
A = effective spudcan bearing area based on cross-section taken at uppermost part of
bearing area in contact with the soil.
Ai = cross sectional area of component i.
As = laterally projected embedded area of spudcan.
b = M'uey/Mny used in determination of η.
B = constant in quadratic expression for β.
B = moment amplification factor.
B1 = moment amplification factor applicable to Mnt.
B2 = moment amplification factor applicable to Mlt.
C = constant in quadratic expression for β.
CDD = product of drag coefficient and associated diameter.
d = water depth (m).
D = diameter of tubular member.
Fmin = strength of weakest component.
Fn1 = nominal strength of component 1.
Fn2 = nominal strength of component 2.
Fui = material ultimate strength for component i.
Fy = material yield strength.
F'y = effective material yield strength = 5Fu/6.
Fyi = material yield strength for component i.
F1 = factor used in computer axial capacity for plastic interaction.
F2 = factor used in computer moment capacity for plastic interaction.
Hs = significant wave height.
K = parameter in chord strength interaction relationship.
L1,L2,etc. = length of chord component 1, 2, etc.
Mlt = moment attributed only to lateral deflection.
Mnt = moment excluding lateral deflection.
Mnx = nominal bending strength about member x-axis.
Mny = nominal bending strength about member y-axis.
Mp = plastic moment capacity.
Mpx = plastic moment capacity about member x-axis.
Mpy = plastic moment capacity about member y-axis.
M'px = effective allowable x-axis bending capacity used in strength interaction
equations.
M'py = effective allowable y-axis ending capacity used in strength interaction
equations.
Mu = applied bending moment.
Mue = effective applied bending moment.
Mux, Muex = effective applied bending moment about member x-axis.
Muy, Muey = effective applied bending moment about member y-axis.
Mx = applied bending moment about member x-axis.
My = applied bending moment about member y-axis.
P,Pu = applied axial load.
Pn = nominal axial strength.
Py = axial yield strength.
t = wall thickness of tubular member.

t1,t2,etc. = thickness of chord component 1, 2, etc.
Tp = peak period associated with Hs.
VT = tidal current velocity.
zo = distance between plastic neutral axis and back face of chord.
Z = plastic section modulus.
β = safety index.
ξ = exponent in chord strength interaction relationship.
η = exponent in bending interaction relationship.
φ = resistance factor.
φa = resistance factor for axial load.
φb = resistance factor for bending.
φHfc = foundation resistance factor - sliding.
φVH = foundation resistance factor - bearing under the action of vertical and horizontal
loads.
φVHM = foundation resistance factor - bearing under the action of vertical, horizontal
and moment loads.
1 Ahilan R.V., Baker M.J., Hoyle M.J.R. and Robinson N.J., "Reliability Based
Development of Jack-up Assessment Criteria". Presented at the Tenth Structures Congress
(ASCE), San Antonio, Texas, April 13-15, 1992.
2 Noble Denton Consultancy Services Limited. "Jack-up Assessment Criteria", Stage 1
Report, L15709/NDCS/RVA (Rev. 2) London, dated 28th February 1992.
3 Noble Denton Consultancy Services Limited. "Jack-up Assessment Criteria", Stage 2
Report, L16268/NDCS/RVA (Rev. 1), London, dated 8th January 1993.
4 Ahilan R.V., Baker M.J. and Snell R.O., "Development of Jack-up Assessment Criteria
using Probabilistic Methods". OTC7305, Houston, Texas, 1993.
5 Noble Denton Consultancy Services Limited. "Jack-up Assessment Criteria - Interim
Scope of Work Items 1.1 to 1.4 on Reliability Analysis". Report No. L15323/NDCS/RVA
(Rev. 2) London, dated 4th March 1991.
6 Noble Denton Consultancy Services Limited. "Jack-up Assessment Criteria - Statement of
Variable Selection". Report No. L15670/NDCS/RVA (Rev. 0), London, dated 12th
September 1991.
7 Winterstein, S. "Nonlinear Vibration Models for Extremes and Fatigue", J. Engineering
Mechanics, ASCE, Vol. 114, October 1988.
8 Juncher Jensen J., "Dynamic Amplification of Offshore Steel Platform Responses due to
Non-Gaussian Wave Loads", Danish Center for Applied Mathematics and Mechanics
Report No. 425, May 1991.
9 Manual of Steel Construction - Allowable Stress Design - Ninth Edition, AISC, 1989.
10 Recommended Practice for Planning, Designing and Constructing Fixed Offshore
Platforms - API RP2A, Eighteenth Edition, 1 Sept 1989.
11 Load and Resistance Factor Design Specification for Structural Steel Buildings, AISC, 1
Sept 1986.
12 Draft Recommended Practice for Planning, Designing and Constructing Fixed Offshore
Platforms - Load and Resistance Factor Design API RP2A - LRFD First Edition, 1 Sept
1989.
13 Rules for Classification - Fixed Offshore Installations, Det Norske Veritas H`vik, July
1991.
14 Buckling strength analysis of Mobile Offshore Units - Classification Notes- Note 30.1,
H`vik, October 1987.
15 Dyer A.P., "Plastic Strength Interaction Equations for Jack-Up Chords", MSc Thesis, Dept
of Mechanical Engineering, Univ. of Sheffield, Nov. 1992.
16 Duan L., Chen W.-F., "A Yield Surface Equation for Doubly Symmetrical Sections",
Engineering Structures, Vol 12, pp. 114-119, April 1990.
17 Norwegian Geotechnical Institute, "Cyclic Effects on Bearing Capacity and Stiffness for
Jack-Up Platforms on Clay', Report 913012-1, May 1992.
18 Rapaport V., Alford J., (1987) "Pre-loading of Independent Leg Units at Locations with
Difficult Seabed Conditions." Conference title : Recent developments in jack-up platforms
- design, construction and operation. The City University, London.
19 Senner D.W.F., (1992) "Analysis of Long Term Jack-up Rig Foundation Performance."
Offshore Site Investigation and Foundation Behavior. SUT International Conference,
London.
20 Sliggers P.G.F., "SIPM Practice for Site Specific Structural Fitness for Purpose Assessment
of Jack-Up Rigs", Paper 21979, SPE/IADC Conference, Amsterdam, 11-14th March 1991.

Sname (2008) PDF

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Sname (2008) PDF

Загружено:

Авторское право:

Доступные форматы

T echnical

This Bulletin contains two documents relating to the site-specific assessment of

The second document is T&R 5-5A and is referred to as a "Recommended Practice".

The Society of Naval Architects

Technical & Research Bulletin 5-5A

This volume contains the following documents:

• T&R 5-5 – Guideline for Site Specific Assessment of Mobile Jack-Up

FIRST EDITION – MAY 1994

THE SOCIETY OF NAVAL ARCHITECTS AND MARINE ENGINEERS

WILLIAM T. BENNETT, CHAIRMAN

Reviewed and Approved

William P. Stewart, Chairman

William T. Bennett Peter G. Noble

The Working Group

Friede & Goldman W.T. Bennett (Chair)

© Copyright 2008 by The Society of Naval Architects and Marine Engineers

Correspondence should be addressed to:

Fax: (281) 877-6796

2 DATA TO BE ASSEMBLED FOR EACH SITE LOCATION

2.1 Rig Data 3

3.1 Mode of Operation 6

4.1 Loading Cases 7

GUIDELINE FOR SITE SPECIFIC ASSESSMENT

1.2 Reference Document

1.2.1 The accompanying document entitled "RECOMMENDED PRACTICE FOR

1.3 Applicability and Limitations

1.4 Typical Approach to Site Assessment

2 DATA TO BE ASSEMBLED FOR EACH SITE LOCATION

2.1 Rig Data

- Pertinent drawings and specifications,

- Details of any relevant modifications.

2.1.2 When modelling the jack-up structure, it is necessary to consider the

2.2 Site Data

2.3 Environmental Data

2.3.3 If the jack-up deployment is to be of limited duration, applicable seasonal data

2.4 Geotechnical Data

3.1 Mode of Operation

3.3 Leg Length Reserve

4.1 Loading Cases

a) Dead loads, plus

b) Applicable live loads, plus

a) Displacement dependent loads, plus

4.2 Wave and Current Forces

4.3 Wind Forces

4.4 Reaction Point and Foundation Fixity

4.5 Storm Approach Angle

4.6 Centre of Gravity

4.7 Displacement Dependent Loads

4.8 Dynamic Effects

b) The magnitude of this vibration is primarily influenced by the amount of

4.9 Other Loads

- Ice loading, and

- Blast over-pressure due to an explosion on an adjacent structure.

It is seldom necessary to consider such loads in combination with the maximum

The key resistance checks include:

- Foundations and preload, and

5.2 Overturning Stability

5.3 Foundation and Preload

5.4 Structural Integrity

Any strength limitations with respect to maximum or minimum penetrations

5.5 Adjacent Structures

FIRST EDITION – MAY 1994

Rev Issue Date Details