Proceedings of OMAE2006

25th International Conference on Offshore Mechanics and Arctic Engineering

June 4-9, 2006, Hamburg, Germany

OMAE2006-92328

SQUALL: NIGHTMARE FOR DESIGNERS OF DEEPWATER WEST AFRICAN

MOORING SYSTEMS

Franck Legerstee, Bureau Veritas Michel François, Bureau Veritas

[email protected] [email protected]

Cédric Morandini, Bureau Veritas Stéphane Le-Guennec, Total

[email protected] [email protected]

ABSTRACT winds are specified, one for sustained wind conditions, the
Squalls have been present in the environmental other for squalls.
specifications for floating units in West Africa for the last
couple of years. However it appears that such phenomena tend West Africa is particularly affected by the position of the
to be the designing factor for mooring systems of deepwater Inter-Tropical Convergence Zone (ITCZ), resulting from the
FPSO’s (in spread or turret configuration) and offloading convergence of trade winds from the North-East and from the
buoys. At that stage, due to the lack of proper South-East, and which is characterised by a great activity of
modelling/characterisation, squalls tend to be represented for cumulonimbus vertical formation and by heavy rains and
design purposes by on-site recorded time series of varying squalls. Squalls are thunderstorms generated on land, which
wind velocity and associated relative headings applied from tend to form along lines separating air masses and which
any direction. This leads to rapid changes in offsets and loads usually travel westward over the sea. They trigger a sudden and
in the mooring lines induced by the transient response of the large increase of wind speeds: a squall event usually lasts less
vessel to sudden load increase generated by such squall signal. than an hour at one given location.
Through diverse exemplary simplified calculations, this paper
illustrates the influence of the consideration of squalls in the Squalls are particularly frequent before and after the rainy
design process, together with the present shortcomings in the season, i.e. from April to June and from September to October.
modelling process, either in terms of extreme conditions, or in
terms of operating conditions, knowing that such events are While satellite and hindcast data reproduce quite well
difficult to forecast. In addition the effect of tugs, and sustained wind speeds and directions, local perturbations such
associated operating limitations are also discussed. Areas as squalls are only captured by on-site measurements. Indeed,
needing further investigation are finally identified.. satellite data does not provide information about squall winds,
as these are not at all stationary events: only measurements
INTRODUCTION with a sampling acquisition period of at least 1Hz may enable
For floating structures from a to derive squall wind speeds with adequate accuracy. Worth to
mooring/risers/hydrodynamic point of view, current design mention that some gusts extreme values available for various
practice consists of considering extreme events due to projects have been calculated from in-situ measurements,
combinations of extreme environment components such as offshore Nigeria, Angola and Congo, analysed during Phase 1
wind, waves and currents. Along the coast of West Africa, of the West Africa Gust Joint Industry Project (WAG JIP). In
given the moderate environment, particular phenomena are addition, it should be noted that the standard wind speed
likely to prevail over other conditions and become governing in conversion factors applicable for sustained winds do not apply
the design process: squalls. For this reason, two sets of extreme for strong gusts because regular trade winds and less frequent,
stronger winds or squalls, are different phenomena.

1 Copyright © 2006 by ASME

Downloaded From: https://proceedings.asmedigitalcollection.asme.org on 06/28/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 All calculations are based on time domain calculations,
Squalls are highly unstationary phenomena: the very rapid performed with the ARIANE software, developed by Bureau
change of wind speed intensity is often accompanied by a rapid Veritas.
change of direction. In addition available records show that,
although squall winds are usually known to turn East (i.e. from
East), this is not always the case. Model Data
A typical West African FPSO has been considered for the
In order to illustrate the main issues raised by the present study: it consists of a 2 million-barrel FPSO, spread-
consideration of squalls in the design process, and to highlight moored by 4 bundles of 3 lines in about 1,000 m water depth.
some of the limitations of the current practice, the effects of a As the focus is essentially on wind effect, only wind force
squall on a typical FPSO spread moored in the Gulf of Guinea coefficients were necessary for the purpose of this study. These
are analysed. Then a simplified squall model is built, in order to wind coefficients were originally established through wind
assess the sensitivity to some key parameters. The details of the tunnel testing. In addition some damping has been
procedure are detailed in the next sections. implemented, based on Ref. 1.

The principal characteristics of the FPSO are given in

WIND MODELING Table 1 and Figure 1.
For such location in West Africa, two sets of extreme
winds are usually given, a set for sustained wind conditions and The 4x3 spread mooring system consists of the following
another set for squalls. segments, from anchor to fairlead:
• 300 m of R3 studless chain,
It should be noted that these extreme winds have not to be • 1,500 m of sheathed spiral strand wire rope,
associated to extreme waves, because the phenomena are not • 200 m of R4 studless chain (adjusted according to
correlated, particularly for squalls. Furthermore, the time pretension)
during which they blow over the sea and their relatively limited
extent is not sufficient to significantly modify the existing sea The mooring pattern is illustrated in Figure 2.
state.
As the natural period of the moored FPSO is expected to
In the design process, 100-year return period squall have a significant influence on the dynamic response to a
conditions will be associated with ‘background’ underlying sea squall, two sets of pretensions have been considered: 100 t and
states, based on the 95 percent non-exceedance probability 300 t leading respectively to natural periods in surge and sway
values. More details on associated sea states are given in Ref. 1 of 400 s and 250 s.
and Ref. 3 and are currently the subject of on-going joint
studies, based on both measurements and theoretical
approaches.
Squall time record and averaging periods:
For sustained winds, usual design practice consists of Time records depend from the acquisition equipment and
using a mean wind speed in combination with a wind spectrum data processing. Most of the time, averaged values will be
to reproduce the wind speed variations, the intensity by given and stored to a specific time step. A typical wind squall
direction being defined by the metocean data. recording time series is given in Figure 3.

In terms of analysis procedure, squalls are better accounted This time series is applied to the numerical model defined
for in the time domain, and their impact needs to be assessed above, and the dynamic response in sway is analysed (cf Figure
for any direction, with the directional extreme gust speeds, if 4).
available.
From Figure 4, firstly a transient response to the rise time
of the squall signal is observed, followed by more or less
damped oscillations at the motion natural period. Remarkably,
ANALYSIS the high frequency oscillations of the input signal are
In order to investigate the potential impact of squalls on completely filtered by the FPSO.
the design of deepwater mooring systems, a typical measured
wind squall time series has been considered. Then an idealised This strongly suggests that the high frequency part of the
simplified model has been used for further numerical analyses. wind signal has absolutely no effect on the overall response of
The frequency content of the squall time records is briefly the FPSO, and can be safely removed from the signal. This can
discussed, considering also the filtering brought by the FPSO. be further demonstrated by using running average of the signal

2 Copyright © 2006 by ASME

Downloaded From: https://proceedings.asmedigitalcollection.asme.org on 06/28/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 over different lapses of time and then analysing the FPSO
response.

Figure 5 shows that there is no noticeable difference Spatial distribution

between the response to the original signal and the signal
averaged over 1 minute. However, for averaging periods longer It is current conservative industry practice to assume that
than 1 minute, the response becomes clearly underestimated. the same wind speed is simultaneously applied to the whole
This can be linked to the increase of the rise time by half of the volume of the FPSO.
averaging period, due to the averaging process. The rise time Under sustained winds, a static wind speed vt is assumed,
therefore appears to be the key factor for the dynamic response averaged over a period t function of the dimensions of the
to a squall. object to which it is applied (e.g. 3s for local action, 1 minute
for larger structures, cf Ref. 2). For a floater, where the low
Thus, the time series of the 1-minute wind speed can be in frequency actions of the wind are significant, like for
a first approach characterised by a rise time (time from instant deepwater systems, a spectrum is used. However, the
of minimum recorded wind speed just before the peak and components with period less than 1 minute in the resulting
instant when the peak is reached) and a decay time. In the signal will be filtered by the FPSO, as illustrated in the
numerical models used in next section, the rise has been previous section for squalls. Their spatial coherence is therefore
modelled as a linear function of time, whereas the decay of no interest as regards the overall response.
section has been modelled by an exponential function,
characterised for instance by the half-decay time, ie the time to For squalls, there is currently no available data concerning
decrease from the peak value to half of this value (cf Figure 6). the characteristics of the wind field and the spatial coherence. It
is worth pointing out that wind speed measurements are
generally acquired using one single anemometer.
Parametric Study Appropriate measurements are necessary to resolve the
current uncertainties as regards the wind field characterisation.
The response is that of a single degree of freedom However, as of today, there is no other alternative for the
oscillator and therefore depends essentially of the natural designer but to apply the same conservative assumptions.
period of the system, and the damping level.
Two natural periods have been considered, corresponding The potential consequences of a different assumption on
to the two sets of pretensions mentioned above, with the same the spatial distribution are shown in Figure 8 to Figure 10.
level of damping, corresponding to a real situation. Two types of runs have been performed:
• The first one applying instantaneously a quartering
From this series of runs, the following conclusions can be wind to the FPSO
drawn: • The second one consisting of a beam wind profile,
1. The maximum response is generally higher than the travelling along the FPSO axis as shown in Figure 8.
static response. The dynamic amplification depends of
the rise and decay times, and of the natural period Depending on the propagation speed of this front, a
(Figure 7). This confirms the results that can be obtained dynamic response greater than that generated by a
from theoretical approaches, published in the literature. unidirectional quartering squall can be observed.
2. The dynamic amplification factor is maximum for a rise That means that even a beam squall unidirectional, but
period to natural period ratio in the range of 0.2 – 0.3. propagating longitudinally may generate an unexpected
For the two systems considered in this paper, this dynamic response of an FPSO, having similar effects with that
corresponds to rise times ranging from 50 to 120 generated by an oblique environment or by a wind speed
seconds, matching what can be observed from actual reversal as it is often observed during squalls.
recordings. Beyond this, dynamic effects generated by a
unidirectional squall can be considered negligible This in turn may explain some discrepancies observed
(Figure 7a). when trying to correlate on-site simultaneous measurements of
3. The decay time has also a notable effect: it can be seen environment and motions of a floating unit with numerical
from Figure 7b that for the same rise time to natural models.
period ratio, a longer decay time will lead to higher
dynamic amplification.
Vertical profile
The observations made above are to be taken with caution:
the direction changes, not considered in this study, will Wind force coefficients are usually derived from wind
introduce additional terms in the response. tunnel tests, using a controlled flow, in terms of intensity,

3 Copyright © 2006 by ASME

Downloaded From: https://proceedings.asmedigitalcollection.asme.org on 06/28/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 direction, turbulence intensity and vertical profile. This is well In light of the above, a line failure might occur during a
established for ocean wind but, given the different nature of squall event, and possibly during an offloading operation.
physical phenomena intervening in the generation of squall Although the probability of such combination of events appears
winds, the question whether a standard exponential profile to be lower than that for usual design conditions, it has not
(Ref. 2) is applicable to squall is legitimate. As there is no or been quantified yet and is not addressed in the present versions
little data concerning vertical distribution of wind speed during of mooring codes. There is a trend in the industry to consider
squalls, the effect of a different vertical profile cannot be such events in the design loop, as transient effects, thus
quantified. It is expected that at least isolated structures such as allowing lower safety factors.
flares might be affected even though the overall response is
kept unchanged. Further work is also needed in this area, in order to keep
consistency between probability levels and applied safety
This gap may be bridged by on-site measurements, using factors.
the adequate vertical distribution over the exposed areas of an
FPSO in order to update the present state-of-the-art. Wind load
coefficients derived through the procedure mentioned above
may no longer be valid in case of squalls.
CONCLUSIONS

The present paper shows the influence of the consideration

of events such as squalls in the design process of floating units
OFFLOADING OPERATIONS in West Africa.
The uncertainties linked to this phenomenon have also
Limiting operating conditions been highlighted.
Further research is needed to understand the physical
Offloading operations, either from an FPSO or an phenomenon associated to squall events, and should lead to
offloading buoy, are likely to be affected by squall events, since better description for design purpose, and better forecasting
these cannot be forecasted with sufficient time to allow proper from the operating point of view.
disconnection. Typical time for disconnection ranges from 45 Active participation of operators of West African units
minutes to 1 hour (from decision to effective disconnection), needs to be pursued, through dedicated instrumentation,
and cannot be reduced without compromising flushing of the sharing of collected data and statistical processing. Such work
floating hoses. This is about the order of magnitude for squall will give a proper ground to guidance for design, when the
forecast. Furthermore, as forecasts are mainly based on visual major uncertainties listed in this paper have been understood
observations, events occurring at night are difficult to forecast, and satisfactorily resolved.
unless proper radar detection is in place.
Consequently, design procedure for offloading operations
should take into account squall events, in general rescaled to 1-
year return period values (typical maximum offloading
conditions for West Africa). ACKNOWLEDGEMENTS
Beyond wind speed sudden increase, changes in wind
direction will have a significant influence on the behaviour of The authors wish to thank Bureau Veritas and Total for
the shuttle tanker, and dynamic effects will be prevailing, not permission to publish this paper. The views expressed are those
only in terms of tensions in the mooring lines or in the hawser of the authors, and do not necessarily reflect those of Bureau
but also in terms of relative heading. Veritas and Total.
Tug assistance is generally provided during offloading
operations; however, as tugs are experiencing the same squalls
as the shuttle tankers, they have only limited control on the
bollard pull, and tend to act so as to maintain the tension in the
tug line to acceptable levels, either through repositioning or REFERENCES
line pay-out. As a result, tug action cannot be considered
effective during squall event.
Ref. 1: Bureau Veritas NI 493 Classification of Mooring
Systems for Permanent Offshore Units
Line failure during squall event
Ref. 2: ISO FDIS 19901-1, Petroleum and natural gas
industries — Specific requirements for offshore

4 Copyright © 2006 by ASME

Downloaded From: https://proceedings.asmedigitalcollection.asme.org on 06/28/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 structures — Part 1: Metocean design and operating
considerations.

Ref. 3: François, M et. al, Directional metocean criteria for

Wind speed
mooring and structural design of floating offshore
structures, ISOPE 2004

Length between perpendiculars LPP 300 m 0 2000 4000 6000 8000 10000 12000
Breadth B 60 m time (s)
Depth 33 m
Draft (Loaded) 24 m
Figure 3: Typical wind squall recording time series
Draft (Ballast) 14 m
Displacement (Loaded) 428k t
Displacement (Ballast) 248k t
Table 1: FPSO Characteristics Response to constant wind speed

YG (m)
AERODYNAMIC DRAG FORCES PER (VELOCITY UNITY)
Wind Moment Coefficient
Wind Force Coefficient

2000 4000 6000 8000 10000 12000

Time (s)
0 45 90 135 180

Figure 4: FPSO dynamic response in sway

Incidence (deg)

Fx (kN*s²/m²) Fy (kN*s²/m²) Fpsi/Ref.(kN*s²/m²)

Wind speed

Figure 1: FPSO wind load coefficients

0 2000 4000 6000 8000 10000 12000

time (s)

3s wind speed 1 min wind speed

3-min wind speed 10-min wind speed
YG (m)

Figure 2: FPSO mooring system 3400 3450 3500 3550 3600 3650 3700 3750 3800
Time (s)

YG (m) 3s YG (m) 1min YG (m) 3min YG (m) 10min

Figure 5a&b: Effect of different averaging periods of wind

velocity on FPSO response

5 Copyright © 2006 by ASME

Downloaded From: https://proceedings.asmedigitalcollection.asme.org on 06/28/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 propagation direction of the front direction of the peak wind speed

Vmax
t0 tk
t1
Wind speed

Wind speed
Vmax/2

FPSO
Time -900 -750 -600 -450 -300 -150 0 150 300

Rise time Half decay FPSO longitudinal axis

time

Figure 8: Definition of a forward travelling front

Figure 6: Theoretical model

Lateral Wind Force on FPSO

Response increases
when
0 500 1000 1500 2000
rise time Tr decreases

Static response

kN
XG(m)

0 250 500 750 1000 1250 1500

Time (s)

Fy (kN) (Travelling front 4 m/s) Fy (kN) (Quartering squall)

Time (s)

Vv=20m/s Tr50 T1/d2 500 Vm20 Tr100 T1/d2 500 Vm20 Tr150 T1/d2 500 Vm20
Wind Moment on FPSO
Tr200 T1/d2 500 Vm20 Tr250 T1/d2 500 Vm20 Tr300 T1/d2 500 Vm20

Response increases
with decay time
0 500 1000 1500 2000
kN.m

Static response
XG(m)

Time (s)

0 250 500 750 1000 1250 1500

Mz (kN.m) (Travelling front 4 m/s) Mz (kN) (Quartering squall)

Figure 9: Loads generated on an FPSO by a quartering

squall vs. a travelling front
Time(s)

Vv=20m/s Tr100 T1/d2 360 Vm20 Tr100 T1/d2 500 Vm20 Tr100 T1/d2 720 Vm20

FPSO Response - Tension in the most loaded line

Figure 7a&b: FPSO Response to different theoretical

Static tension (quartering)
squalls Static tension (beam)
Traveling speed 1 m/s
Traveling speed 2 m/s
Tension in the most loaded line

Traveling speed 3 m/s

Traveling speed 4 m/s
Quartering squall
Beam squall

0 100 200 300 400 500 600 700 800 900 1000
Time (s)

Figure 10: FPSO response to a forward travelling front

6 Copyright © 2006 by ASME

Downloaded From: https://proceedings.asmedigitalcollection.asme.org on 06/28/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use