Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

DEVELOPMENTS IN HEAVY TRANSPORT DESIGN CALCULATIONS

M.J.A. van Exsel MSc and J.B. de Jonge MSc, Dockwise, The Netherlands

SUMMARY

With continuous development in the Oil&Gas industry, there is an increase of transports of heavier and bigger
structures. As these transports tends to reach the structural limits of heavy transport vessels there is a growing demand to
perform more detailed transport analyses.

The paper addresses developments in heavy transport design calculations from a theoretical and an operational
perspective. Both view-points contribute to a more detailed analysis of the transports and increase the safety level.
Theoretical developments of design calculations are achieved by means of a so-called “Global Hull Analysis”. An
integrated finite element model of the transport vessel and cargo ensures that the stiffness interaction between both is
accurately represented. Hydrodynamic analysis –combined with the “design wave approach” – is performed to
determine critical load cases, containing hull pressures and accelerations. These load cases are input for structural
analysis to determine stresses in and support loads between cargo and heavy transport vessel.

Developments on the operational side are achieved by means of a fleet wise motion monitoring project gathering motion
and weather data. This project increases insight in operational margins during transports. Concluded is that the recorded
nowcast significant wave height from almost 1.200.000 nm of data hardly ever exceeds 5.0 [m], giving a significant
operational margin for high design sea-states, typically Hs > 7.0 [m]. For low design sea-state the margin is small. The
aim is to reach a constant operational margin by increasing the design sea-state at the lower range and reducing the
design sea-state at the higher range. This will be implemented in a new design method.

NOMENCLATURE
2. GLOBAL HULL ANALYSIS
CoG Center of Gravity
CSM Cargo Securing Manual For transports of large structures, especially
DW Dockwise where deflection of the heavy transport vessel or cargo
ETLP Extended Tension Leg Platform may be of importance, it is appropriate to perform
FEM Finite Element Method extended structural analyses with loads defined by
GHA Global Hull Analysis hydrodynamic analysis. This development in design
GHS General Hydrostatics calculations is called a “Global Hull Analysis”.
GWS Global Wave Statistics In this chapter the GHA method is described
HTV Heavy Transport Vessel including a comparison between the results obtained via
RAO Response Amplitude Operator the conventional rigid-body calculation method and the
SPOS Ship Performance Optimization results obtained with this more detailed GHA.
System
UC Unity check
WSF Wave Shear Forces
WBM Wave Bending Moment
x acc longitudinal acceleration
y acc transverse acceleration

1. INTRODUCTION

Developments in heavy transport design

conditions from theoretical as well as operational
perspective are described. Improvements are gained by a
more detailed calculation method, which is here defined
as theoretical improvement.
A weather and motion monitoring project on the
Dockwise fleet of vessels leads to operational
improvements. Interesting insights are given from the
extensive amount of data, gathered in a period of four
Figure 1: FE model of combined HTV and cargo
years. Both theoretical as operational improvements lead
to more insight and more accurate analyses in the heavy
transport engineering.

© 2012: The Royal Institution of Naval Architects

 Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

2.1 BACKGROUND In the FE-analysis a set of element pressures and

accelerations in the center of gravity are given as input,
For standard analyses an in-house program is used representing global loads, where pressure and
to calculate the design parameters for the transports of acceleration are directly correlated because phasing
large structures on heavy transport vessel. Several between the accelerations and the wave is included.
assumptions are made to simplify the calculations: The hydrodynamic analysis is performed using
• Cargo and vessel are treated as rigid body the so-called “diffraction” method, which is incorporated
• Beam model to represent heavy transport vessel in in AQWA-Line. The AQWA-Suite of programs is
calculation of shear and still-water bending capable of determining wave shear and bending and to
moment export wave loads on the wetted surface for structural
• Cribbing pressure calculations based on rigid analysis. The structural analysis is performed using a FE
body cargo and HTV package.
For the determination of loads for FEM the
• Cribbing pressure in 4 points, i.e. FWD, AFT, PS following basic steps are made:
& SB extreme fibers of the cribbing arrangement,
• Based on the construction plans, loading condition
see Figure 2
and the weight distribution of the vessel and
cargo, a 3-D FE Model is prepared, see Figure 1.
From this FE Model, the mass, centre of gravity
and the radii of gyration are calculated. With the
mass data and the AQWA model, diffraction
calculations are carried out, determining all
action- (wave) and reaction-forces (added mass
and damping). With the overall added mass,
damping and wave forces and moments, the
hydrodynamic response (the six motion
components and their phase angles) of the vessel
in waves is determined.
• Based on these results wave shear force (WSF)
Figure 2: Cribbing pressure determination conventional and bending moment (WBM) characteristics or
method force- and moment-characteristics between Cargo
To compensate for the simplifications as and HTV - generally called splitting forces and
described above, the maximum allowable cribbing moments - are determined.
pressure from the conventional rigid-body method are set
• Based on frequency domain analysis splitting
to 20 [kg/cm^2]. This gives additional margin and makes
forces and moment design values are determined.
the program suitable for most transports of large
In addition linear combinations of splitting force
structures on a heavy transport vessel.
RAO’s are produced and processed to determine
However, with continuous development in the
cribbing pressure and stress in transverse sections.
Oil & Gas industry, the amount of transports of heavier
and bigger structures increases. As these transports tend • Based on the maximum splitting forces, the most
to reach the structural limits of the HTV, there is a severe wave spectrum and directions are defined.
growing demand to perform more detailed transport
analyses to verify the structural integrity of the HTV and • A regular wave representing equal and
the cargo. Especially with large overhanging- or lengthy comparable splitting forces is selected and used as
cargoes, which are supported over a large part of the input for AQWA-wave. This selection process is
vessels deck, stiffness interaction between cargo and generally known as the “regular design wave
HTV can have a significant impact on the design approach”.
calculations. • Load-sets are created and transferred to FEM.

2.2 GHA METHOD • Structural analysis is performed to check the

integrity of cargo and HTV and to determine
Often overall accelerations are applied to the FE cribbing pressures and deflections of the HTV.
model, sometimes with a simplified method to account
for wave load deflection of the HTV. For such a 2.3 VALIDATION
simplified method the relation between acceleration
components and wave loads is ignored and assumptions A comparison of the results obtained with the
are made. A more appropriate method is to apply conventional rigid-body method and the results obtained
hydrodynamic loads directly from 3D-diffraction with the more detailed GHA are presented in this
software to an FE model; hydrodynamic loads consist of paragraph.
pressures over the underwater hull and accelerations in
the system CoG.

© 2012: The Royal Institution of Naval Architects

Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

2.3(a) Test case description paragraph 3.1. The cribbing pressures are converted to
Two transports have been selected as test cases Unity Checks, U.C. to be able to compare them with the
to investigate the influence of the more detailed GHA on results of the GHA.
the design parameters for larger structures onboard heavy
transport vessels. Table 1: Cribbing pressures test case 1
First test case is the transport of an extended Item Support pressures ETLP (Test case 1)
tension leg platform, ETLP, stowed onboard an HTV, see PS (ex1) SB (ex2) FWD (ey1) Aft (ey2)
2 2 2 2
(kg/cm ) (kg/cm ) (kg/cm ) (kg/cm )
Figure 1. A GHA is performed for this transport due to Static 11.85 11.39 11.62 11.62
the fact that weight of the ETLP is close to the limits of Static + dynamic 20.26 19.80 16.08 16.12
the HTV carrying capacity in combination with a large
overhang of the ETLP on both sides of the vessel. It has PS (ex1) SB (ex2) FWD (ey1) Aft (ey2)
to be noted that a shaped cribbing design is used for this (kg/cm2 ) (kg/cm2) (kg/cm2) (kg/cm2)
Static 0.59 0.57 0.58 0.58
transport. With a shaped cribbing the static cribbing Static + dynamic 1.01 0.99 0.80 0.81
pressure will be equally divided over the vessel width by
adapting the cribbing height. The conventional rigid-
body method uses two rigid-bodies to determine the Cribbing pressures calculated using the GHA
cribbing pressure for which the results shall correspond are presented in Figure 4 & 5. Figure 4 presents the
to a shaped cribbing design. Furthermore, the HTV will actual calculated cribbing pressure and Figure 5 presents
be equipped with four sponsoons to increase the support these pressures converted to U.C.’s based on an
area of the ETLP. allowable cribbing pressure limit of 32 [kg/cm2].
Second selected test case is the transport of a
60.0
semi-submersible longitudinally stowed onboard an Static + Dynamic
HTV, see Figure 3. Due to the longitudinal stowage, Static
50.0
there is a significant contact area in longitudinal direction
Cribbing pressure [kg/cm2]

of the vessel. A GHA has been performed to investigate 40.0

the impact of stiffness interaction between cargo and
HTV. Strength of HTV is not an issue. 30.0

20.0

10.0

0.0
-30.0 -20.0 -10.0 0.0 10.0 20.0 30.0
Distance from HTV CL
2
Figure 4: Cribbing pressures [kg/cm ] test case 1

1.60
Figure 3: Combined FE model of a Heavy transport Static + Dynamic
Static 1.40
vessel and a semi-submersible longitudinally stowed
U.C. = 1 1.20

2.3(b) Validation criteria 1.00

The following design criteria are compared: 0.80

• Cribbing pressures 0.60
• Stillwater longitudinal strength of heavy
0.40
transport vessel
• Deflections of heavy transport vessel 0.20

0.00
2.3 (c) Cribbing pressures validation -30.0 -20.0 -10.0 0.0 10.0 20.0 30.0
The limit for soft wooden cribbing pressure Distance from HTV CL
calculation based on the conventional rigid-body method Figure 5: Cribbing pressures test case 1 converted to
is set at 20 [kg/cm2]. This limit contains a margin to U.C. [-]
compensate for the simplifications in the calculation
method. Since fewer assumptions are made in the GHA, After comparing the cribbing pressure results of
the allowable limits for this analysis are increased to 32 the first test case for both methods the following is noted:
[kg/cm2] for soft wooden cribbing.
Cribbing pressures calculated for test case 1 • The influence of the weight distribution and the
using the rigid-body method are presented in Table 1. stiffness of the cargo and HTV results in
Table 1 contains the actual calculated cribbing pressures significant discrepancies between the calculated
for both the static as well as the static + dynamic cribbing pressures of both methods.
situation based on the method as briefly described in

© 2012: The Royal Institution of Naval Architects

 Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

• As result of the shaped cribbing as used in the 1.20

FEM analysis the static cribbing pressure is

1.00
found to be uniform and similar for both
methods.
• For the static + dynamic load case the difference 0.80

between the two methods is found to be

0.60
significant. Especially the calculated cribbing
pressures below the columns are much higher
0.40
for the GHA in comparison to the ones
calculated by the conventional rigid-body
0.20 Static
method. Dynamic
U.C. = 1
0.00
The benefit of the more detailed GHA with 0 20 40 60 80 100
respect to cribbing pressure calculation is proven to be Semi-submersible PS Pontoon AFT - FWD [m]
essential for this test case. Due to the non-uniform load Figure 7: Cribbing pressures test case 2 converted to UC.
distribution of the cargo over the width of the vessel,
cribbing pressures below the columns are exceeding the Based on the cribbing pressure results for test
allowable limits showing the need of additional cribbing case 2, the following is noted:
wood. This could not have been obtained from the
calculated cribbing pressures for the conventional rigid- • Static cribbing pressures are not similar since
body method. In addition it shows that shaped cribbing is weight distribution along pontoon has
required to guarantee a uniform load distribution in the significant influence on GHA results
static case. • A peak in the cribbing pressures is calculated by
Table 2 contains the cribbing pressures and the GHA method. This high cribbing pressure
unity check for test case 2 calculated using the occurs in a cribbing beam which is located
conventional rigid-body method. between a transverse bulkhead in cargo and
HTV.
Table 2: Cribbing pressures test case 2 • Neglecting the peak pressure calculated by the
Item Support pressures Semi (Test case 2) GHA, the maximum U.C.’s for both calculation
PS (ex1) SB (ex2) FWD (ey1) Aft (ey2)
2
(kg/cm ) (kg/cm )
2 2
(kg/cm )
2
(kg/cm )
methods are in the same range.
Static 10.24 10.60 10.81 10.04
Static + dynamic 15.81 16.17 15.40 14.61 Based on these two test cases, the value of
having a more detailed GHA method on top of the
PS (ex1) SB (ex2) FWD (ey1) Aft (ey2) current existing rigid-body method is proven for cribbing
(kg/cm2 ) (kg/cm2) (kg/cm2) (kg/cm2)
Static 0.51 0.53 0.54 0.50 pressure determination. Especially for large overhanging
Static + dynamic 0.79 0.81 0.77 0.73 cargoes, or cargoes with a non-uniform load distribution,
the calculated cribbing pressures with the GHA method
Figure 6 & 7 present the cribbing pressures can significantly differ from the ones determined by the
calculated using the GHA method. First Figure shows the conventional rigid-body method. It must be noted that,
actual cribbing pressure while Figure 7 shows the due to the large margins which are taken into account in
cribbing pressure converted to a unity check. the rigid-body method, design criteria calculated by the
rigid-body method with respect to cribbing pressures will
Semi-submersible PS Pontoon AFT -> FWD [m] be sufficient in most cases.
40

35 2.3 (d) Longitudinal strength validation

Generally for each loading Condition the still
30 water bending moment and shear force is checked to
Cribbing pressure [kg/cm2]

ensure that the strength requirements of the Classification

25
Societies, i.e. allowable vertical shear and longitudinal
20
bending moment are met.
The shear force curve is calculated by
15 combining the weight and buoyancy distribution over the
HTV’s length. Division by the sectional shear area
10
results in the shear stresses distributed over the length of
5
the HTV. Integration of the shear force curve over the
Dyna mic
Static length of the HTV results in the still-water bending
Average sta tic cribbing p ressure
0 moment distribution. Minimum and maximum bending
0 20 40 60 80 100
moment occur at the section with zero shear load.
Figure 6: Cribbing pressures [kg/cm2] test case 2

© 2012: The Royal Institution of Naval Architects

Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

Maximum allowable shear force and bending moment in

conventional rigid-body method are defined as a
percentage of the total shear and bending. The method
reserves a part of the capacity which is required for wave
shear and bending contribution.
Figure 8 shows the still-water bending moment
and shear force curve for test case 1 determined by the
conventional rigid-body method. In this graph also the
bending moment and shear force limits are presented.
Shear force is close to its limits just in front and aft of
both pontoons of the ETLP. Figure 10: Bending stress in main deck of HTV for the
Figure 9 shows the shear stress distribution for static condition
the static condition over the different longitudinal
members calculated by the GHA. Comparing the shear
stress distribution with the shear force strength graph of
Figure 8 it can be concluded that the trend in longitudinal
direction of the HTV is similar. The shear force in the FE
model is also determined just aft the aft pontoon of the
ETLP (looking from stern) being 6847mt. This means
that a difference of about 2% occurs between the two
methods. It is concluded that in this case stiffness
interaction does not have a significant influence on the
calculated maximum shear force. Benefit of the FE
model is that it shows the distribution over the different Figure 11: Bending stress in HTV bottom hull for the
longitudinal members where the rigid-body cannot make static condition
a distinction between those members.
Bending stress in main deck and bottom of the Figures 12 to 15 present the same graphs for test
HTV are presented in Figures 10 and 11. The trend in case 2 as reported above for test case 1. These figures
longitudinal direction is similar to the bending curve as prove that this case was not selected due to HTV strength
presented in Figure 8. but to investigate stiffness interaction between cargo and
HTV and cribbing pressures. Stresses, as well as shear
force and bending moment are fluctuating around zero
not coming close to its limits.

Figure 8: Still-water bending moment and shear curve

Figure 12: Still-water bending moment and shear curve

Figure 9: Shear stress in longitudinal HTV members for Figure 13: Shear stress in longitudinal HTV members for
the static condition the static condition

© 2012: The Royal Institution of Naval Architects

 Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

Based on above mentioned assumptions, a static

analysis is performed. In 12 steps the wave top is moved
along the vessel length to calculate the deflection at
certain phases. The deflection is calculated assuming the
following:

• Vessel is a beam with a certain cross-section

• Static analysis, no accelerations

From the integrated FE model used for the

GHA, also the deflection of the vessel can be determined.
In contradiction to the GHS method, the GHA uses:

Figure 14: Bending stress in main deck of HTV for the • Realistic wave length determined by the spectral
static condition analysis
• Dynamic analysis including accelerations
• Transverse/torsional deflections can be
determined using the GHA results

Figures 16 & 17 present a comparison between the HTV

vertical deflections for both methods.

Figure 15: Bending stress in HTV bottom hull for the

static condition

Based on test 1 it is concluded that stiffness

interaction has a negligible influence on the maximum Figure 16: HTV deflection comparison between GHA &
shear force and bending moment since the values for GHS for test case 1
both methods are equal.
Benefit of the GHA method is that the shear
distribution over the longitudinal members in transverse
direction of the HTV is calculated while for the rigid-
body method it is assumed to be equally distributed.
Furthermore with the GHA method actual dynamic
situations are assessed while the conventional rigid-body
method takes a percentage of the total capacity into
account for the dynamic behaviour of the HTV
independent of actual wave loads.

2.3 (e) HTV deflection validations

For certain projects, vessel deflections are of
importance for cargo and/or seafastening. Furthermore it
can be used to determine whether uplift will occur. Since
head- or following waves will result in the largest vessel
deflections, this wave heading is considered in both Figure 17: HTV deflection comparison between GHA &
methods. GHS for test case 2
To calculate vessel deflections, the conventional
GHS method uses the following assumptions: The same trend in deflections is recognized for
• Wave length = vessel length (lpp) both test cases; i.e. deflections calculated by the
• Wave height = max wave height = 1.86 * Hsig conventional GHS method are more severe than for the
GHA method.

© 2012: The Royal Institution of Naval Architects

Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

Reason why these deflections are more severe the most severe area. Given the season and the transit
has mainly to do with the assumptions taken for the GHS time through the area, the probability of exceedance of
method. Especially the wave length has significant the individual “3-hour stationary storm” design wave
impact on the deflections. Because it is not known for the height is calculated. To account for calm periods the
rigid-body method which wave length is most severe, a calculation is corrected for wave smaller than 4.0 [m].
worst case wave length is assumed. Reference is made to Dockwise guidelines and criteria
Another reason why the deflections are lower [3] for detailed description about the design wave height
for the GHA method has to do with stiffness interaction. determination.
Especially for test case 2, stiffness of the cargo has a Wave height relaxation in non-head sea
significant contribution to the overall stiffness of the conditions may be considered for vessels with redundant
system, which mainly visible in the sagging condition. propulsion systems. According to the Noble Denton
Due to the stiffness of the cargo, deflections in sagging Guidelines for marine operations [6] a vessel with a
condition are significantly lower. redundant propulsion system is defined as having, as a
It is concluded that the deflections calculated minimum:
with the GHS method are more severe than the ones
calculated with the GHA method due to: • 2 or more independent main engines
• Applied wave • 2 or more independent fuel supplies
• Stiffness interaction • 2 or more independent power transmission
systems
2.4 CONCLUSION & RECOMMENDATIONS • 2 or more independent switchboards
• 2 or more independent steering systems, or an
Overall it is concluded that the more detailed alternative means of operation of a single
GHA method to determine the design criteria is required steering system (but excluding emergency
in the following cases: steering systems that cannot be operated from
• Critical transports in the conventional rigid- the bridge
body method calculations • The ability to maintain any desired heading in
• Large overhanging cargo all condition up to and including the design
• Expectation that stiffness of cargo has storm, taking into account the wind age of the
significant impact on the design criteria like cargo.
cribbing load The maximum reduction of design wave height is given
in Table 3.
It is recommended to compare the results of
more transports with each other to find out whether a Table 3: directional wave height relaxation Noble Denton
trend could be obtained from the results. In future it Dir wrt bow [deg] % of design wave height
could increase safety and save resources when general 30 / 0 / 330 100
knowledge from the GHA could be applied to all 45 / 315 90
transports. 60 / 300 80
75 / 285 70
3. OPERATIONAL IMPROVEMENTS 90 / 270 60
105 / 255 70
An overview is given on the methods used within 120 / 240 80
Dockwise to generate the design wave height for heavy
135 / 225 90
marine transports. Operational insights are given and the
150 / 180 / 210 100
increased amount of data, compared to [2], is analysed,
shown and reviewed. In addition a prognosis is given
3.1 (b) Short trip scenario
about future developments to come to a design wave
In a very limited number of cases, where vessels
height based on the measured weather data.
are not redundant or where accelerations due to
bow/stern waves result in unacceptable forces to the
3.1 DESIGN WAVE HEIGHT DETERMINATION
cargo, a short trip scenario could be used to reduce the
design wave height. This method is only to be used if
This paragraph gives an overview of three
enough safe havens exist along the planned route. In
different methods to derive the design wave height for
addition extensive safety precautions are to be taken
heavy transport voyages.
when a short trip scenario is included in the design.
Weather forecast need to be obtained from 2 different
3.1 (a) Standard design method
providers and it is advisable to provide dedicated route
As a standard design method the scatter
forecast by a specialized company.
diagrams from the Global Wave Statistics are used at
Dockwise. From these available wave statistics the short
term design sea state is derived where the significant
wave height has a probability of exceedance of 5% for

© 2012: The Royal Institution of Naval Architects

 Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

3.1 (c) SafeTrans advance. Please note that the actual wave height could
The MARIN program SafeTrans is an integrated have been larger when the forecast underestimated the
tool to design marine heavy lift transports in a safe and weather.
efficient manner using state of the art analysis methods, Figure 18 also shows that with increasing design
databases, and hydrodynamics. This DNV approved wave height the operational margin increases. While the
software includes the master’s decisions to avoid adverse lowest operational margin occurs at 3.0-5.0 [m]
weather and uses Monte Carlo simulations based on a significant design wave height, the δnowcast_Design for
hind cast/fore cast database. By using the knowledge and higher design sea states reduces to values smaller than
experience of the master in the design the wave height 60%. This trend is traced back to the fact that master’s
can be reduced. In addition an estimate is given on the decisions about routing and bad weather avoidance
amount of weather delays. Reference is made to Aalbers increase the operational margin at larger design waves.
et al [5]. Generally speaking it is shown that design waves larger
than 5.0 [m] are never encountered for the analysed
3.2 OPERATIONAL INSIGHTS voyages in this database.

The theoretical methods to determine the design

wave height and accelerations are evaluated by using full
scale weather forecast and motion measurements.
Reference is made to [2] for a detailed description on the
Octopus motion monitoring system as installed on all
Dockwise vessels. The latest results are presented from
the Octopus system and insights are given with respect to
using nowcast weather data, influence of the angle of loll
and overhang of cargo.

3.2 (a) Nowcast versus Design wave height

An update is given about the nowcast wave
height registrations from the Dockwise vessels. As
recommended in [2] the significance of this research has
increased from 55 to 140 voyages and from 530.000 nm Figure 18: nowcast significant wave height vs. Design
to 1.178.292 nm, see Figure 23. It is important to wave height per Area according to GWS
investigate if the presented conclusions are still valid.
The logged GPS position is used to determine 3.2 (b) Measured versus design accelerations
which Areas according the GWS are crossed during each In order to exclude any uncertainty of weather
transport. The nowcast significant wave height is data and the conversion from weather data to motion
compared to the design wave height for each area of a data, the focus is transferred to measured accelerations.
voyage. From the nowcast wind and swell waves a total The measured acceleration is compared to a single design
wave height per time step is determined by taking the acceleration value, which is determined with Shipmo for
quadratic sum of both waves. Results of the wave height each voyage. Reference is made to Dockwise Guidelines
comparison are shown in Figure 18. The following and Criteria [4] for a detailed description on ship motion
formula is used to determine the non-dimensional calculations by Shipmo. The following formula is used to
δnowcast_Design. determine the non-dimensional δMeasured_Design to be able to
compare multiple transports.

_ =
_ =
Figure 18 shows that the design wave height
was almost never exceeded according to the recorded The number of occurrences versus the value of
nowcast wave height from SPOS. At 4 points the this δ gives important information on the actual
Nowcast wave height was slightly exceeded. Looking to operational margin during heavy transports. The total
the maximum recorded nowcast significant wave height number of 3-hour occurrences, as plotted in Figures 19 to
of 5.0 [m] our masters have been able to avoid larger 22 is 31980, which is around 4000 sailing days of
forecasted waves. This is an interesting conclusion different Dockwise vessels.
because it shows that our masters avoid regions where Figures 19 and 20 are presented to focus on
the forecast wave height is larger than 4.5-5.0 [m]. This measured longitudinal, x and transverse, y accelerations
graph validates the DW/Anglo Eastern policy about in the cargo CoG compared to design accelerations.
avoiding heavy weather for HTV’s. In addition there is a These figures show that the operational margin increases
time factor which avoids vessels ending up in excessive with increasing design acceleration. This conclusion
waves, in the order of 9 to 10 [m], which typically need confirms the conclusion from the Nowcast wave data.
some time to build up and hence are better predictable in

© 2012: The Royal Institution of Naval Architects

Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

Another interesting conclusion is that the

operational margin for x acceleration is smaller than the
operational margin for y acceleration. This conclusion
confirms that during heavy weather the master turns his
vessel into bow waves. The determination of the design
pitch motion and y acceleration could also influence this
conclusion, because of usage of the strip theory with
limited accuracy for longitudinal motions i.e. surge and
pitch.
It should also be kept in mind that x acceleration
is generally much smaller than y acceleration. Design
accelerations are used to determine seafastening forces
between vessel and cargo. As x accelerations are
generally small and cribbing friction is significant, the
seafastening force is often replaced by a minimum of 5%
of the cargo weight, which increases the operational Figure 20: Y acceleration Measured vs. Design
margin. This additional safety margin is not displayed in
these graphs because measured and design accelerations
are compared directly and thus show the operational
margin.
Figure 20 shows also that two measured
acceleration points for y acceleration were almost 100%
of the design. Please note that for that particular voyage
wave height relaxation was used in order to keep the
acceleration within limits for cargo strength. The
seafastening forces were calculated with non reduced
wave heights in order to maintain equal safety standards
for all transports. The conclusion is that this practice
reduces the operational margin as shown in this Figure.
In addition one point exceeds the design acceleration
with 50%, as a result of large overhang and low stability.
Reference is made to paragraph 3.2 (d) for review of this Figure 21: X acc in Cargo CoG occurrences vs
voyage. δMeasured_Design
Several conclusions are drawn from Figure 21
and 22, which display for x and y acceleration the
δMeasured_Design number of 3-hour occurrences. For x and y
acceleration the average δMeasured_Design is around 0.1375
and 0.125 respectively. Next to this it can be seen that
only a small number of occurrences is larger than 50% of
the design value. For x and y direction the probability of
3-hour occurrences larger than 50% is 2.84% and 2.15%
respectively.

Figure 22: Y acc in Cargo CoG occurrences vs

δMeasured_Design

3.2 (c) Wind influences on accelerations

For transports with a large cargo windage area
like jack-up rigs, the wind speed could have effect on the
measured transverse acceleration. In the transport design
analysis the wind contribution is split in two parts, a
static wind force and a force due to the angle of loll. It is
to be noted that the forces from the wind are included in
Figure 19: X acceleration Measured vs. Design the seafastening load, but the transverse acceleration

© 2012: The Royal Institution of Naval Architects

 Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

from the angle of loll is not included in the design 3.3 OPERATIONAL BASED DESIGN METHOD
acceleration as presented in the CSM. However the
measured accelerations include both the effects of wave A research project is started to investigate how
and wind induced motions. the data can contribute to a more realistic and safer
From the analysed voyages some examples exist design wave height method. The aim is to reach a
where the angle of loll was 4-7 degrees. In one case the constant operational margin by increasing the design sea-
g*sin(φ) term from the angle of loll was larger than the state at the lower range and reducing the design sea-state
transverse acceleration from the wave induced motions. at the higher range. Marine warranty surveyors are also
To include this effect the transverse design accelerations interested in the design of this new method and have
are corrected for the wind angle of loll for three voyages. offered to contribute to the project.
At the moment of writing tools are available to
3.2 (d) effect of cargo overhang on accelerations produce GWS Area and Season specific wave scatter
In the heavy transport it is not unusual to sail with diagrams from the SPOS nowcast data, see Figure 24.
buoyant overhang of cargo in the order of 20 [m]. The black numbers are produced from the Octopus
Generally this has little or no effect on the ship motions. Nowcast data and the red numbers are generated by
However it is found from a couple of transports that the GWS observations. The figure shows again lower
effect on ship motions could be large. The ideal recipe significant wave height. In addition the mean Tz from the
that leads to small transverse design accelerations and diagram is also lower. An interesting question is how the
large chance of exceeding these is the following Octopus wave scatter diagrams could be used as input for
combination: a new design wave height method.
• large buoyant overhang Preliminary calculations are performed to
• low GM generate a design wave height based on equal probability
• large roll period > 35 [s], outside wave spectrum of exceedance compared to the standard design method
• low freeboard by using a Weibull fit. The maximum nowcast wave
• large angle of loll height is registered to be 5.5 [m], which causes that this
Due to the small stability, a small exciting force is method cannot be used directly for standard design wave
enough to result in a roll angle of several degrees. heights of 7-8 [m] or larger.
Depending on the actual loading condition and Other questions that arise are how the wind and
configuration waves of a couple of meters could cause swell waves should be combined in one scatter diagram.
the vessel to roll and let the overhang touch the water. Large differences are present between wave periods and
This suddenly gives an increase in stability, which results wave directions, which can have a large effect on the
in unexpected ship motions that were not taken into vessel motions. Would it be possible to generate separate
account in the design process. One transport out of the wind and swell scatter diagrams as input for ship motion
140 show that transverse design accelerations are analysis and combine the results afterwards?
exceeded with 50%, while seastate did not reached the
design value. 3.4 CONCLUSIONS & RECOMMENDATIONS
It is not easy to include these effects in the
frequency domain ship motion analysis due to their non- With almost 1.200.000 nm of ship motion and
linear character. Internal research was performed to use a weather data the significance of the research project has
3D diffraction program, but this did not yet lead to an increased and the conclusion are still standing and valid.
accurate method. Additional damping and excitation In addition it shows clear justification to design a
terms cause difficulties in calculating accurate motion operational based method to determine a more realistic
analysis without performing model tests. design wave height at the higher sea states.
Model tests with overhang of a large semi It is proposed to keep working on collecting
submersible show also that when waves hit the bottom of data from the fleet to further increase the statistics and
the cargo and if there is significant submergence the significance.
transverse accelerations increase. This increase could be Operational insights are given on the effects of
up to 35% compared to ship motions obtained from the wind contribution and cargo overhang. Both topics can
linear theory. have significant impact on the ship motions in particular
It is important to note that for the above circumstances.
described topic the transports were still safe because the Further investigation is scheduled to design an
minimum seafastening capacity is defined to be 5-10% of operational based method for the design wave height
the (large) cargo weight by the Marine warranty determination where masters decision to avoid adverse
surveyors. In addition the steel to wood friction weather are included. This could lead to a more realistic
coefficient is generally set to 0.2 while model tests show and safer heavy transport design.
that 0.4 or even 0.5 is a more realistic value.

© 2012: The Royal Institution of Naval Architects

Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

4. CONCLUSIONS & RECOMMENDATIONS 4. DOCKWISE GC.ENG.302 “Motion Response

Analysis”, Guidelines and Criteria, November
Both theoretical and operational improvements 2008.
lead to more insight in heavy marine transportation. It is 5. A.B. Aalbers, “A new software system for safer
valuable to improve the calculation methods to increase rig moves”, Marin, Safetrans technical notes.
knowledge and insight about transporting larger and 6. GL Noble Denton, “Guidelines for Marine
heavier structures. It is recommended to keep working on operations 0030/ND”, Sect 7.2.10, 31 March
improvements and to compare results from different 2010.
voyages to see if trends exist and if standards and quick
calculations method could be improved. 6. AUTHORS BIOGRAPHY

5. REFERENCES Martijn van Exsel holds the position of structural

engineer at Dockwise. He is involved in all kind of
1. O.A.J. Peters and L.J.M. Adegeest, ‘Motion strength related topics like vessel/cargo strength
Monitoring and Decision Support during Heavy verifications and design of supports/grillages and all
Transport’, OMAE2010- 21143, 2010. other non-custom requirements.
2. J.B. de Jonge and O.A.J. Peters, ‘Operational
margin from weather and motion database for Jan de Jonge works as a marine engineer at Dockwise
heavy transport vessels’, OMAE2011-49811, and is involved in all kind of hydromechanics topics like
2011. ship motion behaviour, mooring, environment and model
3. DOCKWISE GC.ENG101 “Routing and Design testing. In addition daily work involves leading a fleet
Environmental Conditions”, Guidelines and wise motion monitoring project, processing and
Criteria, November 2008. analysing of the data.

Figure 23 sailed routes of DW heavy transport vessels.

© 2012: The Royal Institution of Naval Architects

 Marine Heavy Transport & Lift III, 24–25 October 2012, London, UK

Figure 24 nowcast scatter diagram recorded along the sailed route, black is Octopus, red is GWS.

© 2012: The Royal Institution of Naval Architects