LNG / LPG ARMS PROTECTION AGAINST EXCESSIVE SHIP MOVEMENT:

“A NEW APPROACH FOR ALARMS SETTING AND ESD / ERS ACTIVATION”

Bertrand LANQUETIN
Liquefied Gas Shipping Department Technical Manager
TOTALFINAELF Gas Electricity Division

ABSTRACT

The new OCIMF Design and Construction Specification for MARINE LOADING ARMS (Third
Edition – 1999) departs significantly from previous editions with regard to ESD and ERS alarms
setting. In particular it does not specify any longer figures for ESD and ERS sequences duration and
for drifting distances / speeds to be used for first stage and second stage alarms setting on loading
arms.
On the contrary, it stipulates clearly that:
“The Owner shall determine the drift velocities for each berth.
This will be based on the type and size of vessels using the jetty, wind, current, etc.”
(Chapter 2.19.4)
It stipulates also that:
“The first stage alarm will initiate the ERS ball valves closure”
(Chapter 10.2.3)

This issue of alarms setting was raised by our Company to the OCIMF loading arms Task Force as
early as 1997 after considerable work and experience have been gained on the subject in recent years
at the opportunity of various LNG / LPG terminal studies characterized by very different weather
conditions and ships dimensions (like Indonesia, Yemen, India, etc.).
As a summary we found that the traditional way of alarms setting based on the consideration of three
parameters fixed arbitrarily and taken independently from each others was not reflecting the reality
and could lead either to unnecessary arms envelope wastage or to underestimated drifting speeds and
distances. Furthermore no account has been taken until now for a catastrophic scenario like ship break-
out.
These three parameters are namely: the model of the flanging area and the “drifting” area (allowance
for ship movement at berth), the duration of the ESD and ERS sequences and the estimation of the
drifting distances and speeds.
A global approach incorporating the accurate estimation of these parameters and examining how they
interact is therefore necessary and has been successfully developed.

The present paper summarizes the work done on the subject by our Company and gives, through a
practical example resulting from a recent project of LNG jetty, our latest methodology for:
• Establishing the flanging and “drifting” areas and the setting-up of the alarm 1st step delimiting a
working envelope called “safe area”. Within this area the ship-mooring pattern remains in sound
condition and the cargo operations can take place safely. Both cases of ship reaching this alarm
with unnoticeable or with noticeable speeds are considered.
(The pre-alarm zone is located within this safe area).
• Setting-up of alarms 2nd step and arm mechanical limit beyond the limit of the safe area in a
scenario of excessive movement of the ship resulting from the mooring pattern failure.
• Alarms detection and the re-visited ESD / ERS activation philosophy accordingly.

This new approach not only supports the changes done in the Third Edition of the OCIMF
Specification, but also goes beyond the OCIMF requirements for the particular case of arms handling
liquefied gases. At the same time it eliminates any wasting of unused arms envelope through a good
adequacy between ESD and ERS sequences duration, the ship drifting distances and speeds, and the
arms design (mechanical envelop limits).

PROGRAM MENU
PART 1: TRADITIONAL METHOD FOR ARMS DESIGN AND ALARMS SETTING,
ALARMS DETECTION AND ESD/ERS ACTIVATION
This rather long introduction is necessary in the first part of the paper in order to well understand the
traditional method in use for years for arms design and alarms setting in order to identify its
insufficiencies and the reasons why improvements are necessary. This method has been used back to
1980 (Design and Construction Specification for Marine loading Arms, 1st Edition - see Ref [1]) until
now and is still continued to be used in many cases as the common practice. The method is
summarized on Figure 1 and described hereunder.

1.1 DETERMINATION OF ARM MECHANICAL ENVELOPE (ARM DESIGN)

This is done by arm vendor and is based on two sets of data:
• The flanging area: this area is approximated by a rectangular parallelepiped, the sides of which
being based on the extreme positions of the ships flanges (actual flange locations including tide
and ship condition: ballast or loaded).
• The drifting area: this area is also approximated by a rectangular parallelepiped obtained from the
flanging area by adding a longitudinal drift and a lateral drift of the ship.
There is quite a significant discrepancy on the recommended values to be used for longitudinal and
lateral drifts. Furthermore we could not find how these values have been determined in the past:
The OCIMF (1st Edition 1980 and 2nd Edition 1987 – see Ref [2]) recommends +/- 3.1 m “to be found
as a reasonable approximation of the allowable vessel surge” for sheltered areas and +/- 4.6 m for
more exposed sites. For the lateral drift a value of 3.1 m is recommended for the vessel sway-off (at
any site). It is further assumed that the “maximum sway and the maximum surge do not occur
simultaneously”.
The CEN standard EN 1474 (1st Edition 1997 – see Ref [3]) suggests a maximum drift area of 3.0 m
for both surge and sway movements of the ship.
Note: nor OCIMF or CEN specifies if the above is a quasi static drift or if it includes in addition the
movements of the ship caused by the waves.
The PIANC Criteria for Movements of Moored Ships in Harbors 1995 (see Ref [4]) has established
criteria for safe mooring of gas tankers as being 2.0 m for surge and 2.0 m for sway. These are
specified as peak to peak values for surge and zero to peak values for sway with a velocity of 0.3 m/s
in both surge and sway to be used for berth design.
Whatever the final figure for surge and sway is all here above specifications assume that the mooring
pattern is in intact condition. Although integrity of the arms is protected by excessive drift detection
and corresponding ESD (Emergency Shut-Down) and ERS (Emergency Release System) sequences
activation, the fact to base the arm mechanical envelope (or in other words the arms design) on the
assumption of an intact mooring pattern and on drifting values somehow arbitrary is a too simplistic
approach. This approach also ignores the fact that a combined longitudinal and lateral drift is a reality.
The SIGTTO Information Paper “Accident Prevention, the Use of Hoses and Arms at Marine
Terminals Handling Liquefied Gas” 1996 (see Ref [5]) states very clearly, with statistic calculations to
support that statement, that “the risk of break-out is high (1.57 for 100 port calls) – it identifies a very
real risk that should be addressed”.

1.2 SETTING UP OF ENVELOPES FOR ALARMS 1ST STEP AND 2ND STEP
It has to be first clearly mentioned that this responsibility has usually been left in the past to the arm
vendor and this is still often the case today, although the OCIMF Design and Construction
Specification for Marine loading Arms, 3rd Edition 1999 (see Ref [6]) departs from the previous
Editions and stipulates clearly that:
“The Owner shall determine the drift velocities for each berth.
This will be based on the type and size of vessels using the jetty, wind, current, etc.”
(Chapter 2.19.4)
The alarm 1st step and 2nd step envelopes have usually a spherical shape (although a rectangular shape
has been often used in the past to model the alarm 1st step envelope, the PMS – arms Position
Monitoring System - allowing both rectangular and spherical coordinates).
The traditional method for alarms setting is based on the following recommendations:
The OCIMF (1st Edition 1980 and 2nd Edition 1987, see Ref [1] and [2]) states (page 47) that “…the
signals from these sensors are relayed to a control console where alarms are actuated when

PROGRAM MENU
FLANGING
DRIFTING DRIFTING SPEEDS
AREA ESD ERS
AREA 15cm/s for ERS,
based on 30S 2s TO 30s
4.5m or 3.0m 5cm/s for ESD
ships data

ARM MECHANICAL ENVELOPE SETTING UP OF ENVELOPES FOR ALARMS

(ARM LENGTH DESIGN) 1st STEP AND 2nd STEP

DRAWBACKS:
• SUGGESTED DISTANCES (IE. DRIFTING AREA), SPEEDS (IE. DRIFTING SPEEDS) AND DURATIONS
(IE. ESD, ERS) ARE TAKEN INDEPENDENTLY, WHILE THEY ARE IN FACT CLOSELY INTERRELATED
• DO NOT REFLECT ACCURATLY OPERATING CONDITIONS, IE. WEATHER CONDITIONS AND SHIPS
SIZES FOR A GIVEN SITE
• DRIFTING SPEED PROFILE (IE. SHIP ACCELERATION) NOT PROPERLY MODELLED, IN ADDITION
SAME DRIFTING SPEED IS USED FOR LONGITUDINAL AND LATERAL SPEEDS
• SEQUENCE OF PERC ACTIVATION, LOGICS OF CONTROL (PROXIMITY SWITCHES AND/OR PMS)
NOT INCORPORATED IN THE METHOD BECAUSE THEY ARE RELEVANT TO COMPANIES POLICIES

CONSEQUENCES:
• ARM LENGTH MIGHT NOT BE OPTIMIZED
• WORKING ENVELOPE MIGHT NOT BE OPTIMIZED
• DRIFTING SPEEDS MIGHT BE GREATER

Figure 1: Traditional method for arms design and alarms setting, ESD/ERS detection and activation
predetermined slew and drift limits are reached”. As said before, the owner and vendor’s
responsibilities described in Part III of these specifications did not cover the alarms setting and this
was left in most of the cases to the arms vendor.
Traditionally the figures deemed to be safe and used by the vendors as a “standard” are 0.5 m between
the alarm 2nd step and the arm mechanical limit (which supposes a drifting speed of 15 cm/s for a
typical ERS sequence duration of 5 s) and a distance of 1.5 m between the alarm 1st step and the alarm
2nd step (which supposes a drifting speed of 5 cm/s for a typical ESD sequence of 30 s).
The CEN standard EN 1474 (1st Edition 1997 - see Ref [3]) suggests also the same figures to be used,
i.e. 0.5 and 1.5 m.
Similarly to the values given for the drifting area we could not find how these values of drifting speeds
have been determined and what were the hypothesis used (weather conditions, size of the ship, intact
mooring or ship break-out, etc.). The fact that the drifting speed increases with time after break-out is
somehow taken into account (speed increasing from 5 to 15 cm/s between ESD and ERS sequences)
but nothing is said on the hypothesis retained for ship acceleration, although one can easily recognize
that the use of constant drifting speeds can be very misleading. Finally it is assumed that the
longitudinal and lateral drifting speeds are the same, which is obviously not the case and further, no
combined drifting is considered.

The deficiencies of the traditional method for arms design (mechanical envelope) and alarms settings
are not limited to the above. The major approximations made are in our opinion:
• The assumption that the drifting area and the drifting speeds are universal: it is clear that they
vary with the size of the ship, the mooring pattern composition and its integrity, the
environmental conditions and the drifting direction of the ship.
• The assumption that the longitudinal and lateral drifts are the same while it is not the case and
while the limitations with regard of arms design are different: a big lateral drift can be
accommodated at design stage by longer arms for example, while a big longitudinal drift cannot
be always accommodated, the limit being the maximum arm slewing angle which is limited to
approximately 45° (an illustration is given in PART 4 Figure 12 of the paper).
• The assumption that the ESD sequence has to be taken universally at 30 s.
• The assumption that the arms design (mechanical envelope based on the flanging area and the
drifting area), and the setting of the alarms 1st step and 2nd step (based on drifting speeds and ESD
and ERS sequences duration) are two independent things, while in practice they are closely
interrelated: for example high drifting speeds can be expected on a weather exposed loading jetty
berthing big LNG ships. This can be accommodated in several ways: longer arm design (bigger
mechanical envelope), shorter ESD and ERS sequences, or ideally by an optimum combination of
the two.
The above approximations may result in an underestimation of the drifting speeds with, in such case,
a serious risk of damaging the loading arms in case of ship break-out (in case the ESD and ERS
sequences cannot be timely completed), or at the opposite in a wastage of the arm mechanical
envelope: operating envelopes showing a drift of several meters before reaching the envelope alarm 1st
step are not uncommon. It is most likely that such drift could never be reached without a partial or
total failure of the moorings. In such scenario the activation of the ESD sequence will be probably too
late. On these examples there is an immediate benefit of triggering the ESD sequence earlier by
moving the alarm envelop closer to the working area, and hence allowing a better use of the arm
mechanical envelope space.
The above concerns were conveyed as early as in 1997 to the Task Force in charge of preparing the 3rd
Edition 1999 of the OCIMF Design and Construction Specification for Marine loading Arms (see Ref
[6]) and as a result the following new wording was used:
“The Owner shall determine the drift velocities for each berth.
This will be based on the type and size of vessels using the jetty, wind, current, etc.”
(Chapter 2.19.4)
One should realize that this apparently innocent wording invites in practice to a major change in the
approach adopted for years in the field of arms design and alarms setting and indicates clearly were
the responsibilities are laying.

PROGRAM MENU
The OCIMF does not give however any recipe on how to achieve this goal. We trust that our
experience described in this paper will be helpful and valuable for terminal owners and designers and
for arms vendors.

A last point has to be mentioned on the traditional method described above because it has been
neglected in the past or fully left to the arms vendors, who have developed the PMS (arms Position
Monitoring System). This monitoring tool can be very powerful providing the philosophy of alarm
detection is well understood and used in the appropriate way.

1.3 ALARM 1ST STEP AND ALARM 2ND STEP DETECTION AND ACTIVATION
The way we propose to achieve this is detailed later in PART 2 of the paper. We want to highlight
hereunder some of the uses, which have been developed following the appearance of the ERS (1980)
and the PMS (1986). As it can be seen, many discrepancies have been found in the various approaches
denoting that a serious rethinking of the whole ESD/ERS alarm detection and activation was
necessary. Hereunder is a rapid summary on the various uses:
• Alarm detection: the detection is done traditionally by proximity switches (inductive type),
number of which and position being clearly defined in the OCIMF specification. PMS sensors
come additionally (potentiometric type or absolute encoder type sensors). These sensors can
detect both arms flange position and speed (drifting distances and drifting speed). The position of
the PMS sensors is usually left to the arm vendor.
• Back-up proximity switches by PMS sensors: this is left either to the owner or more generally to
the arm vendor.
• Shape of alarm envelopes: it is clear that the lateral envelope (in case the ship moves away from
the jetty) has a spherical shape because proximity switches measure angular co-ordinates.
However PMS allows either to measure angular or rectangular co-ordinates. On many
installations the alarm 1st step detected by PMS has a rectangular shape (this alarm level being
considered as an extension of the arm drifting area, which has also a rectangular shape) while the
same alarm 1st step detected by proximity switches has a spherical shape (this alarm level being
considered as a protection before to reach the arm mechanical envelop, which has also a spherical
shape. To our knowledge however the alarm 2nd step detected either by proximity switches or
PMS sensors has always a spherical shape.
• Integration of ship speed to anticipate triggering of alarms: the PMS being able to measure the
drifting speed, allows to anticipate the triggering of an alarm. This facility has been used
sometimes for the early activation of the alarm 1st step, however sometimes also for the activation
of the alarm 2nd step and sometimes for both, without taking into account the laws of
hydrodynamic for a drifting ship, i.e. drifting acceleration after break-out. In other words, if the
alarm 2nd step is anticipated by the PMS, there is a real risk that an ESD sequence (alarm 1st step)
triggered for example by proximity switches, will not have the time to be completed before
activation of the ERS sequence (in case interlock is provided, there is then a real risk to reach the
arm mechanical envelope before the ERS sequence is completed.
• ESD and ERS sequences duration, time delays between ESD and ERS and by-pass flow
restriction on the hydraulic activating the ERS sequence: ESD sequence is usually 30 s,
sometimes more, rarely less (to take into account the duration of the ship manifold ESD valve
closure set at 30 s). The ERS sequence duration is from 2 to 30 s. Some mitigating measures are
often taken in order to reduce the risk of surge pressure, depending on sequences duration, time
delay between ESD and ERS sequences (see Ref [7] and [8]). Sometimes, and in order to
accelerate the ERS sequence in case a high drifting speed is detected by the PMS, a by pass flow
restriction on the PERC hydraulic (Powered Emergency Release Coupling) is automatically
activated (this device has been installed on some recent installations).
• Sequence of closing PERC valves: it seems that the majority of owners prefer to have the PERC
valves closing during the ESD sequence, with the ERS sequence therefore limited to PERC
opening. Some owners however prefer to have the ERS sequence to be entirely done when alarm
2nd step is triggered, in that case the sequence includes the PERC closing valves following by
PERC opening activation.

PROGRAM MENU
• Other considerations: some other considerations must be cited here like hydraulic interlock
between ESD and ERS sequences. More recently these traditional interlock systems
recommended by OCIMF had to be modified on the so called “no spill” PERC design, where the
interlock system between ESD and ERS sequences had to become both mechanical and hydraulic
due to the fact that the upper ball valve of the PERC closes during ESD sequence while the lower
ball valve sequence closes during the ERS sequence.
It can be seen from the here above that a lot of various approaches – if not discrepancies - exist on the
particular subject of ESD/ERS alarm detection and activation.
This was brought forward also in 1997 to the OCIMF Task Force in charge of the revision of arms
design and construction specification and as a result the following changes in the new specification
must be highlighted:
- No mention is made any more on the ESD sequence duration
- The specification stipulates that:
“The first stage alarm will initiate the ERS ball valves closure”
(Chapter 10.2.3)
It is clear here again that the aim is to put more responsibility on the owner to establish its own safe
and reliable design based on the proper characteristics and operating conditions of his jetty.
Here again as well the OCIMF does not give any recipe. We have in the past years, at the opportunity
of various projects, revisited the ESD/ERS philosophy with regard to detection, activation, etc. and we
trust that our experience described in this paper will be helpful and valuable for terminal owners and
designers and for arms vendors.

PART 2: A NEW APPROACH FOR ALARMS SETTING, ALARMS DETECTION

AND ESD/ERS ACTIVATION
This revisited methodology is the result of our experience gained from various recent projects and is
now enforced in our Company. These projects were for example the Yemen LNG project
characterized by a difficult SouthWest monsoon and the Trombay LNG project characterized by
significant tidal currents. For projects like this it appeared very quickly that the traditional method for
arms design, alarm setting and detection and ESD/ERS activation was not appropriate and this has
been at the origin of our reflection on the subject and on the perfecting of the method described
hereunder. The methodology is based on:
• The laws of hydrodynamic (behavior of the ship at berth and after ship break-out) with the use of
sophisticated computer programs to model the behavior of the ship.
• The logical and the fault tree analysis of the failure events (intact or defective mooring after lines
failure).
• Considerations of the risk analysis (integrity of the arms).
• Increased level of safety and protection of the installation (anticipation of the events by initiating
proper and adequate actions).
• The concern of maximum use of the arm mechanical envelop without any wasting (in both normal
and emergency operations).

2.1 ALARMS ZONES AND ALARMS DETECTION

The revisited concept is summarized on Figure 2 to which the hereunder description refers to. To be
complete, a description of the type of computer programs we recommend to use for the evaluation of
the ships movements at berth (intact mooring conditions) and for the drifting distances and speeds
after break-out (total mooring failure) is given in PART 3 of this paper. This will be followed by a
PART 4, in which we will give a concrete example of loading arms design and alarms setting based on
a recent project where this methodology has been applied.

A. Flanging area:
This space allows for a variation of ship manifold flanges positions for the range of ships handled. The
hip is supposed to be in ideal position on the spotting line and on berthing line, i.e. with no fender
compression.

PROGRAM MENU
 SAFE AREA , ELEVATION
UNNOTICEABLE
DRIFTING SPEED
MECHANICAL ENVELOPE (FIXED)
ALARM 2ND STEP (FIXED)
ALARM 1ST STEP (FIXED)

AREASHIP
MOVEMENTS AT BERTH
ALARM 1ST STEP (VARIABLE)

DRIFTINGFOR
FLANGING

ALARM 1st STEP IS IN

AREA

ALLOWANCE SPHERICAL COORDINATES

(D1 & D2 ARE BASED ON

D1 D2 SIMULATED DRIFTING
DISTANCES; DRIFTING AREA
IS CALCULATED BY MODEL)
A A
A

B) ALLOWANCE FOR SHIP MOVEMENTS AT BERTH IS USUALLY CALLED "DRIFTING AREA"

DETECTION:
PROXIMITY SWITCHES

IMPORTANT NOTES: A) IN THE SAFE AREA IT IS SUPPOSED THAT THE MOORING PATTERN IS INTACT;
&
PMS SENSORS
DETECTION:
PROXIMITY SWITCHES
&

BEYOND IT IS SUPPOSED THAT THE MOORING PATTERN HAS A COMPLETE FAILURE

PMS SENSORS
DETECTION:
PMS SENSORS ,
FUNCTION OF SHIP DRIFTING
SPEED

SAFE AREA ,
UNNOTICEABLE
DRIFTING
SPEED
SECTION A-A
D'2
MECHANICAL ENVELOPE (FIXED)
D'1
ALARM 2ND STEP (FIXED)
ALARM 1ST STEP (FIXED)
ALARM 1ST STEP (VARIABLE)
FLANGING
SPOTTING LINE
AREA

(D'1 & D'2 ARE BASED ON

ALLOW. FOR SHIP
SIMULATED DRIFTING
MOVT
DISTANCES; DRIFTING AREA
D'1
IS CALCULATED BY MODEL)
D'2
TO BE ON CONSERVATIVE
SIDE, DRIFTING AREA HAS A
RECTANGULAR SHAPE

Figure 2: The revisited methodology - alarms zones and alarms detection

PROGRAM MENU
It is determined by tabulation of X1, X2, Y, Z1, and Z2 values for the range of ships handled (see
illustrations on Figures 3 and 4 enclosed). These values take into account the tide range and the ship
draft variations between the loaded and operational ballast conditions.
It has a polygonal shape, which we recommend not to be approximated by a rectangular
parallelepiped. The reason being that, due to the recommendations for ship manifold standardization
(see Ref [9]) and Ref [10]) and the relatively comparable geometrical characteristics of liquefied gas
ships in the medium to the big size range, very few ship flanges are in the lower part of the flanging
area, and those who are in that area have generally their flanges close to the berth because they
correspond to relatively small ships. Approximating the polygonal shape by a rectangular
parallelepiped would considerably inflate the volume of the lower part of the flanging area, knowing
that the arm length is most of the time determined by the difficulty to reach the lowest flanges due to
the high elevations of the loading platform position. To approximate the flanging area by a rectangular
parallelepiped would result in an unnecessary over-sizing of the arm length (arm mechanical
envelope) for purely geometrical reasons and without bringing more safety in a system (it is a much
better and safer approach to accept an increase the arm mechanical envelope because of high expected
drifting speeds compared to an increase of this mechanical envelope for unjustified geometrical
reasons, i.e. unduly approximation of the flanging area). The Figure 5 illustrates well the here above
point, valid in a majority (some exceptions exist) of cases.

B. “Drifting” area, setting up of alarm 1st step, pre-alarm:

It is the maximum excursion of the ship under wind, current, and wave, with an intact mooring
pattern, rounded up to a value to be given by the owner (the rounding takes into account the fact that
some ships may have smaller pre-tensions in their lines or use mooring lines with greater elasticity for
example). The “drifting” area is the result of the ship movement around its ideal position described
above, resulting from forces and movements under wind, current and waves action including quasi-
static and dynamic movements. As such the traditional wording “drifting area” is unfortunately very
misleading and a confusion can be made with the ship drifting after partial or total rupture of the
moorings (ship break-out). A simple wording like “allowance area for ship movement at berth”
would be much more appropriate and will be used in this paper.
To be on the conservative side, this drifting area for a given ship with intact mooring is assumed to
have a rectangular shape and not a circular shape (arm vendors utilize both; a circular shape implies
that the ship movement at berth is the same in all directions, which is obviously not the case. A
rectangular shape is a more accurate representation and will lead to longer arms to encompass this
rectangular shape).
The values of the lateral ship excursion (affecting the arm luffing movement) and of longitudinal
ship excursion (affecting the arm slewing movement) are based on the maximum sway and surge
motions given by the "Berthing Model Studies" (this will be analyzed in PART 3 of this paper) and are
rounded-up to a value to be given by the arm owner.
These values (lateral and longitudinal excursions) are added to the lateral Y and to the longitudinal
X1, X2 components of the flanging area to determine a polygonal volume, which as a result, is used to
set-up the alarm 1st step envelope of the loading arms. The alarm 1st step envelope has a spherical
shape (and not a rectangular shape as it has been done sometimes in the past) and is determined so as
to encompass the above polygonal volume (see section "A - A" on Figure 2).
It is assumed that this alarm can be reached for a ship drifting – ship excursion movement - with an
unnoticeable speed (for example the ship can reach an equilibrium position under the action of the
elements). In such case the alarm detection is done both by proximity switches and the PMS
sensors backing up each other as indicated on Figure 2. (Enough sensors allow detecting luffing and
slewing arms movements respectively).
Alternately, this alarm envelope can be reached with a noticeable drifting speed of the ship denoting
either an abnormal dynamical behavior of the ship or an improper line tensioning (too slack pattern) or
a partial failure of the mooring lines. In such case, and depending on the drifting speed, the alarm 1st
step is anticipated by the PMS so that the ESD sequence can be triggered earlier, see Figure 2.

PROGRAM MENU
 3000 3000

3000 Y 6000

200 Fender compression

Figure 3: Flanging area - tabulation of X1, X2 & Y Figure 4: Flanging area - tabulation of Z1, Z2
 Alarm 1st Step
Alarm 2nd Step
MAX.REACH ENVELOPE for 12" Arms
MAX.REACH ENVELOPE for 8" Arms

Alarm 1st Step

moves by # 3 m

MAX.REACH ENVELOPE 3m
moves by # 3 m 3m

Figure 5: Consequence on arm length of approximation of polygonal flanging area by a rectangular parallelepiped
The area within the alarm 1st step envelope is considered as a safe working area (mooring pattern in
intact condition). On the drawings “arms operating envelop” provided by the arm vendors, we ask
specifically not to indicate the so called “drifting” area anymore because it does not add any useful
information and again this wording is extremely misleading. Only the flanging area (polygonal shape)
and the alarm 1st step envelop (spherical shape) are shown in this safe working area.
In any case of detection, the alarm 1st step triggers the ESD sequence.
Finally, and within the safe area, a pre-alarm is provided, should the ship flange moves beyond 0.75
m (typical figure, can be adjusted) in any direction. This alarm is only audible and visible on the PMS
screen. It is provided solely for the operator information and might indicate that during mooring
operation the ship has deviated significantly from her spotting line or that a re-tensioning of the
mooring lines might be necessary.
We further ask in our specification for arm vendors that a re-initialization of the PMS (memory of
initial flanges coordinates) is done once the ship is all fast (and not each time one arm is finished to be
connected). This is to face a situation where a ship would have slightly moved from her spotting line
during the mooring operation.

C. Beyond the safe area:

a) setting up of alarm 2nd step:

Beyond the alarm 1st step envelope, it is assumed to be on a conservative side that the ship faces a
complete failure of her mooring pattern (catastrophic scenario called ship break-out).
It is further assumed that the ESD sequence (including PERC valves closure time) has been optimized
for weather exposed loading terminals where high drifting speeds are expected, as a compromise
between arm length (mechanical envelop) and the pressure surge calculations. Generally speaking it is
actually not recommended in weather exposed locations to use the standard ESD sequence duration of
30 s because this will allow a break-out event to develop in too great drifting speeds and distances: it
can be seen through “Drifting Model Studies” (see PART 3 of the paper) that, during the first minute
of drifting, the ship acceleration is constant, which means that the drifting speed increases linearly
with the time elapsed from the break-out event and that the drifting distance versus time follows a
quadratic curve!
The drifting distances shown on Figure 2 (lateral drift D1 and longitudinal drift D'1) at completion
of ESD sequence are given by the Drifting Model Studies. They have further to include a safety
margin due to PMS accuracy.
These drifting distances D1 and D'1 are used to set up the alarm 2nd step envelope.
The alarm 2nd step detection is done both by proximity switches and the PMS sensors backing up
each other. It has to be noted that, due to the quadratic curve of the drifting distance versus time after
break-out, it would be unsafe to anticipate also the alarm 2nd step using the PMS sensors; therefore
the alarm 2nd step, contrarily to the alarm 1st step, is fixed.
The fact that a safety margin is taken on the drifting area (see above chapter 2.1.B.: predicted ship
lateral and longitudinal excursions are rounded-up) and that the detection of alarm 1st step is
anticipated in case of noticeable drifting speed of the ship in the safe working area (see above chapter
2.1.B.), and finally that the drifting scenarios used for the setting up of the alarm 2nd step are
conservative (ship break-out), leads to the conclusion that the ESD sequence should be always
completed prior to the detection of the alarm 2nd step. However, should the ESD sequence not be
completed for any reason at the time the alarm 2nd step is detected, then an additional safety margin is
taken between the alarm 2nd step envelope and the arm mechanical envelope as explained in
hereunder chapter 3.2. allowing completing the ESD and ERS sequences sequentially prior to reach
the arm mechanical limit.
The detection of the alarm 2nd step triggers the activation of the ERS sequence (PERC opening).

PROGRAM MENU
b) arm mechanical envelope:
The definition of the arm mechanical envelope (or arm mechanical limit) and therefore of the arm
length required for the project is based on two conservative assumptions in order to build further
safety margin in the concept:
• it is assumed that the ERS sequence (i.e. PERC opening) duration is 5 s (in reality it is 2 s or
less).
• rather than using the predicted drifting distances between alarm 2nd step and arm mechanical
envelope, we use the maximum drifting speeds (lateral and longitudinal) predicted at the time of
ERS sequence completion after ship break-out, and this in order to calculate the lateral distance
D2 and the longitudinal distance D' 2 between the alarm 2nd step and the arm mechanical
envelope.
The drifting speeds (lateral speed and longitudinal speed) at the end of the ERS sequence are given in
the Drifting Speed Studies. They include on the top a safety margin due to PMS accuracy.
Finally it has also to be checked that the part of the arm mechanical envelope close to the loading
platform is behind the berthing line with compressed fenders.
The value of the fender compression to be used is given by the Berthing Model Studies and is rounded
up (i.e. we use predicted fender compression figures rather than the standard conventional figure of
200 to 300 mm used by arms vendors).
NOTES:
• It is important to note that the longitudinal and lateral drifting speeds and distances are not the
same. The constraints for loading arms design (mechanical limits) are not the same as well for
lateral and longitudinal drifts:
- The lateral drift (affecting the arm luffing movement) requirements can usually easily be taken
into account by adjusting the arm length
- For the longitudinal drift at contrary (affecting the arm slewing movement), very little can be done
due to the limit 45° slewing angle (arm length has no effect on this limit). This should be a cause
of concern in areas with strong currents for example or for ship where starting the main engine by
mistake could be a credible scenario.
Again, and for loading terminals, a solution to build more safety in the design is to consider a
reduction of the ESD sequence duration compared to the conventional value of 30 s (as said before
in PART 1, this ESD sequence duration is no more indicated in the OCIMF Marine Loading Arms
specification 3rd Edition, see Ref. [6]).
• The most important parameters for arms design and alarms setting are not independent (namely
the model of the flanging area and the area for ship movements at berth, the ESD and ERS
sequences durations, the drifting distances and speeds). Several tries have generally to be done in
order to find a good compromise in particular with regard to the arm length required. The Figure 6
shows how such iterative process can be done in order to reduce an unacceptable arm length.
When comparing with Figure 1, it can be seen how the above method departs significantly from
the traditional one.

PROGRAM MENU
CHANGE ESTABLISH MINIMUM
FLANGING AREA +
ARM POSSIBLE ESD, ERS
LENGTH SEQUENCES DURATIONS
ESD,ERS SEQUENCE DURATION
ALLOWANCE AREA FOR
taking into account loading
SHIP MOVEMENT AT BERTH resp. 30s ? 2 to 30s?
or discharging port,
(UNACCURATELY CALLED
piping design (surge),
"DRIFTING AREA")
ship valve closing time

WORKING AREA WITH INTACT MOORING

ESTABLISH DRIFTING SPEEDS
CONDITIONS (INCLUDES SHIP
AND DRIFTING DISTANCES BASED
MOVEMENTS IN OPEN SEAS,
ON ACTUALSHIPS CHARACTERISTICS,
MAINLY HEAVE, ROLL, SWAY, SURGE)
ACTUAL WEATHER DATA
AND VARIOUS DRIFTING SCENARIOS
this area has a polygonal contour.
At this stage atypical ships which
versus time starting from t = 0s
might unnecessary increase
(date of break-out event)
the flanging area can be identified

SETTING-UP ALARM 1ST STEP SETTING UP OF ENVELOPE 2ND STEP

ENVELOPE AND ARM MECHANICAL LIMIT
(DELIMITS THE SAFE AREA) (ARM LENGTH)

CHECK IF ARM if YES, OK

LENGTH REMAINS if NO: 1st step: check ships with restrictions in order to possibly reduce
WITHIN the flanging area
REASONABLE 2nd step: check if the ESD (and ERS) sequences duration can be
FIXED DATA LIMITS reduced this will reduce the distance required between
alarm 1st step and arm mechanical envelopes
3rd step (after 1st and/or 2nd step): change arm length

Figure 6: Recommended approach for combining arms design and alarms setting in case iterations are necessary
2.2 ESD AND ERS SEQUENCES AND ACTIVATION FROM PMS SENSORS AND ALARM
PROXIMITY SWITCHES

a) ESD and ERS sequences

The sequences are summarized on Figure 7. The alarm 1st step detection triggers the ESD sequence
where, among other things, pumps stop and ESD valves close; The sequence duration has to be
optimized particularly for loading terminals where it has no reason to be set arbitrarily to 30 s.
According to the OCIMF Marine Loading Arms specification 3rd Edition the ESD sequence must
include the closing of the ball valves of the PERC (Powered Emergency Release Coupling also called
“dry” break coupling). The exception to this is for the so-called “no spill PERC” where the upper ball
valve closes first due to the mechanical interlock between upper and lower ball valves.
Once the ESD sequence is completed and the alarm 2nd step is detected, then the ERS sequence is
trigger and consists simply in the PERC opening. More and more this ERS sequence is followed (and
we recommend to have this facility on all liquefied gas terminals) by a hydraulically controlled raise
and retract clear sequence of the arms.
It has to be noted here that modern arms for LNG are of DCMA type (Double Counterweight Marine
Arms, usually DCMA-S, i.e. supported) while RCMA type arms (Rotating Counterweight Marine
Arms) are generally used for other liquefied gases and LPG in particular because they are smaller in
diameter and do not require sophisticated DCMA technology.
This difference of technologies has to be highlighted when we go through arms balancing and
emergency raise and retract sequence, knowing that just before PERC opens arms are in free wheel
condition; therefore understanding and analysis of the natural movement of the arm at time of
disconnection is important:
• DCMA arms are easier to balance and due to the fact that the outboard arm is balanced by an
independent counterweight, the natural movement of the outer arm after disconnection (at flange
or at PERC) is to retract simultaneously upwards and backwards, hence avoiding the risk for the
outer arm to hit ship manifold at the time of the disconnection.
• RCMA arms are different due to the single counterweight. The apex angle (vertical angle between
inner arm and outer arm) remains constant after arm disconnection (at outer flange or at PERC)
with the natural tendency for the outer arm to hit the ship manifold at the time of the
disconnection. Therefore the raise and retract clear mechanism has to be design more carefully for
these arms and the absence of such mechanism is not recommended.

b) ESD and ERS sequences activation from PMS sensors and proximity switches on arms

The logic resulting from the methodology described in above chapter 2.1 is given in Figure 8 showing
the PMS CPU (Central Processing Unit), the PLC (Programmable Logic Controller) and the electronic
control panel. It illustrates also how the PMS sensors and the proximity switches back-up each other.
There has been some discussion on the reliability of PMS sensor in comparison with the reliability of
proximity switches, going sometimes up to the decision not to install the PMS and rely only on
proximity switches. We are in the opinion that this is a wrong debate because it goes nowhere to
compare a piece of a system with another piece of a system.
What is important is the way the whole system is conceived from sensors to a multiplexer through
signal cables, then connection to PMS board through ILS circuit, then connection to the PLC and
finally display of the information on the electric control panel, and what is provided for PMS failure
detection. The system of failure detection is called the PMS auto-check and consists of continuous
interrogation of values given by the sensors (three sensors per arm), which are compared each other. If
the values are not coherent, the wrong sensor information is refused by the micro-processor and an
error message is given to the operator (additional information in order to localize the defect is also
provided such as default of a PMS card, default on multiplexer, default of sensor on arm, etc.).
There is therefore no reason to think that a failure will not be detected.
Should a failure being detected on a particular sensor, the alarm 1st step and the alarm 2nd step will still
be detected by the proximity switch and the ESD or ERS triggered because the alarm detection by
PMS sensors or by potentiometric sensors back-up each other in our concept.

PROGRAM MENU
 ESD 30 SEC? ERS < 5 SEC

• STOP SHIP PUMPS

• CLOSE SHIP ESD VALVE
• CLOSE ARM MOV VALVE
• OPEN ARM VENT VALVE
• ETC.

CLOSE UPPER
T.D.
PERC VALVE IN • CLOSE LOWER PERC VALVE (b)
5S (a) & FOLLOWED BY PERC OPENING,
TOTAL SEQUENCE IN < 5 SEC

• THEN AUTOMATIC ARMS

RETRACTION CLEAR FROM
T.D. = TIME DELAY TO BE ADJUSTED DEPENDING ON SURGE
DRIFTING SHIP
PRESSURE CALCULATION

• ARMS LOCKING

NOTE (a): THIS IS FOR THE ENHANCED "NO SPILL" PERC.

SHOULD CONVENTIONAL PERC TO BE SELECTED, THEN BOTH PERC VALVES CLOSE
SIMULTANEOUSLY IN THAT SEQUENCE.

NOTE (b): THIS IS FOR THE ENHANCED "NO SPILL" PERC ONLY. SHOULD CONVENTIONAL
PERC TO BE SELECTED, THEN PERC OPENING SEQUENCE CAN BE REDUCED < 2S.

Figure 7: Sequences of ESD and ERS activation

 * * * PERC UNLOCKED *
INBOARD OUTBOARD SLEWING

POTENTIOMETRIC
SENSORS ON ARM
* * PROXIMITY SWITCHES
ALARM 1 ALARM 2
ON ARM

INTRINSICALLY SAFE
TRANSMITTER

PMS IN OPERATION
PMS IN OPERATION

PMS SENSORS OK *
PMS SENSORS FAILURE *
PMS
2ND ALARM *

ELECTRIC
1ST ALARM *
CONTROL PANEL
(OUTPUT SIGNALS) AND AND
AND PROXIMITY SWITCHES
FAILURE *

OR OR
PERC UNLOCKED *

AND AND PLC

PRE-ALARM
PRE-ALARM

ESD ERS
* PER ARM

Figure 8: ESD and ERS sequences activation from PMS and alarms proximity switches
PART 3: COMPUTER PROGRAMS USED TO MODELIZE SHIP BEHAVIOR AT
BERTH AND SHIP DRIFTING AFTER BREAK-OUT
The use of computer programs (often completed by physical model tests) is necessary in order to set-
up properly and accurately the alarms 1st step and 2nd step on loading arms.
More generally these models allow to understand as well the arm swivels solicitations (with regard to
angular amplitudes, period and angular speeds) for the ship at berth in a weather exposed area and are
a helpful support for swivel design.

3.1 BERTHING MODELS

An interesting paper on the subject is given in Ref. [4]. A lot of work has been done in the past by
several working groups on good mooring practices, on design of fenders, on wind and current drag
coefficients, establishing very useful guidelines. The use of these documents is satisfactory in most of
the circumstances in order to mitigate the risk of accident on a ship at berth.
However these documents are not well suited for the design of jetties where more sophisticated tools
are required, which include the non linearity of the mooring lines (elasticity curve), the non linearity of
the fenders (fender characteristic curve) and factors controlling the moored response to wave action
like the added inertia and the dumping. Such models should be also able to predict the dynamic
response of the ship at berth under wind gusty phenomena and analyze the effect of long swells.
It is also essential to identify at a early stage possible high pick tension values, which may result from
coupling of main oscillation modes and apply corrective actions like use of tail ropes for example in
order to modify one of the natural frequencies of the system.
The programs we usually use are ARIANE developed by Bureau Veritas and ELF of France and
widely accepted worldwide or DIODORE developed by PRINCIPIA of France or AQUAPLUS
developed by SIREHNA of France. Several similar programs exist, developed by various maritime
institutes in different countries such as (MARIN, DHI, CETENA, etc.). It is not possible to mention all
of them here. They are well known by the engineering companies.
Usually these programs make a two-stage analysis by assuming a summation of the cinematic effects
resulting from the two components of the real ship movement:
• One is dedicated to the low frequency response to weather-induced stress (swell, wind, current),
traditionally known as “drift” effects, including the permanent deformations and thrusts of time-
average slow sway oscillations.
• The other is dedicated to the response to the frequency of oscillations imposed by the swell, which
constitutes a relatively high frequency exciter of the six degrees of freedom around their mean or
slowly varying state.
The results of the above programs in various sets of environment conditions of wind (including gusty
phenomena), irregular waves and current allow to determine the maximum lateral and longitudinal
ship excursions at ship manifold (sum of static plus dynamic effects, combining linear and angular
components of ship movements) to be used for setting the alarm 1st step as explained in above chapter
2.1 B.
An example of berthing model results is given in Table 1. Simulations were done using the DIODORE
model from PRINCIPIA (France) on a recent project of a LNG jetty located at the edge of a natural
channel affected by tidal currents, winds and attenuated waves arriving at location after diffraction and
refraction in the channel. The jetty orientation is 200° (ship-moored bow out), more or less parallel to
the prevailing tidal flow and ebb currents. The same project will be taken as an example for illustrating
the results of the drifting model hereunder and the example of arms design and alarms setting given in
PART 4.
We have chosen this recent example in relatively sheltered conditions to give a general illustration of
the method on a representative project with regard to ship size, weather data, etc. Obviously a project
in open sea would give much greater arms envelopes and more “spectacular” results with regard to
drifting speeds.
The max lateral and longitudinal motions are considered, i.e. the sum of the quasi-static and dynamic
components. The small values obtained by the model reflect the particular characteristics of the site:
almost no transversal components due to the berth approximately parallel to the tidal current directions
and very attenuated waves. On more weather exposed sites these figures would be significantly higher,
remembering that the forces on the ship are proportional to the squared values of the wind and current.

PROGRAM MENU
 DESCRIPTION OF MAIN DIMENSIONNING TESTS IN DYNAMIC CONDITIONS

CURRENT WIND WAVES

CASE MODE LOADING CONDITION SPEED (m/s) DIRECTION SPEED (m/s) DIRECTION Hs (m) Tp (s) DIRECTION
CONDITION TO °N FROM °N FROM °N
7 Dynamic Full Ebb 1.1 240 20 0 0.7 7 180
8 Dynamic Full Ebb 1.1 240 20 0 0.7 14 180
9 Dynamic Full Ebb 1.1 240 20 110 0.7 14 180
10 Dynamic Full Ebb 1.1 240 16 90 0.7 14 180
11 Dynamic Ballast Ebb 1.1 240 16 90 0.7 14 180

TABLE 1: RESULTS WITH REGARD TO MANIFOLD MOVEMENTS

X MOTIONS (m) Y MOTIONS (m) Z MOTIONS (m) DIMENSIO-

NING CASE
CASE MEAN MAX Rms MEAN MAX Rms MEAN MAX Rms
7 1.14 1.32 0.06 -0.03 0.12 0.05 0.00 0.09 0.03
8 1.14 1.33 0.06 -0.03 0.12 0.05 0.00 0.11 0.03 YES (X)
9 1.12 1.31 0.06 -0.04 0.12 0.05 0.00 0.11 0.03
10 1.12 1.31 0.06 -0.04 0.12 0.05 0.00 0.11 0.03
11 1.09 1.24 0.06 -0.03 0.20 0.04 0.00 0.11 0.03 YES (Y)

NOTE: THE PROGRAM GIVES DIRECTLY THE X, Y, Z MOVEMENTS AT MANIFOLD FLANGE (AND SPEEDS/ACCELERATIONS IF REQUIRED)
TOGETHER WITH THE THREE LINEAR AND THREE ANGULAR MOVEMENTS AT CENTER OF GRAVITY OF THE SHIP
3.2 DRIFTING MODELS
Until now we have used real time navigational programs in order to model the drifting of the ship in the
first minute after ship break-out under various environment conditions of wind, irregular waves and
current allow to determine the drifting distances, speeds and ship rate of turn after break-out. Lateral,
longitudinal and combined drifts can be modeled.
We use usually TRAJNAV model developed by SOGREAH/STNMTE of France or other models
developed by various maritime institutes in various countries.
A very interesting feature of these programs is that a scenario like engine start can be added to the various
drifting scenarios. Also interesting is the fact that drifting distances and speeds can be expressed both in
rectangular or angular coordinates. We ask for both as illustrated in Figure 9, knowing that the PMS who
will detect the ship flange position can use both types of coordinates. The same Figure 9 gives an example
of ship drifting after break-out during five minutes (one plot every 30 s). In practice we are looking only at
what happens during the first minute for arms alarms set-up.
In order to illustrate the drifting laws, we have chosen an example of pure lateral drift simulated with the
real time navigational model SIMON from CETENA (Italy). The LNG jetty studied is part of the studies
carried-out on a recent project. The jetty is located at the edge of a natural channel affected by tidal
currents and has an orientation 200° of (ship moored bow out), more or less parallel to the prevailing tidal
flow and ebb currents.
The test conditions were:
• Wind 20 m/s from 290°
• Current 0.25 m/s to 110° (reversing tide condition)
• No waves, no engine, no rudder.
The results of the simulation are given on Figure 10.
These conditions where the dimensioning case for the lateral drift in this particular project.
In the first minute of drifting after break-out, the drifting speed is approximately proportional to the wind
speed squared and current speed squared. The curve drifting speed versus time after break-out is linear,
which means that the drifting acceleration “B” is constant and that the drifting distance is a quadratic
function of the time elapsed after ship break-out “t” (after two or three minutes after break-out the drifting
speed stabilizes and do not increase anymore).
In the example given on Figure 10:
• Drifting speed: V = A + B.t , with A = 0.25000 cm/s (should be theoretically 0) and B = 0.56429
cm/s2 (A and B are obtained by linear regression)
• Drifting distance: D = 1/2 B.t 2
Similar simulations are done for longitudinal drift and combined lateral/longitudinal drifts. Results are
summarized in Table 2 attached (negative signs for drifting distances and speeds are due to convention
direction of axes).
.
The following important observations can be made:
• Average drifting speeds:
a) Lateral drift (TEST 3): 8.50 cm/s for the ESD sequence and 18.50 cm/s for the ESD sequence.
b) Longitudinal drift without starting engine (TEST 2): 6.50 cm/s for the ESD sequence and 14.50 cm/s
for the ERS sequence.
c) Combined lateral and longitudinal drifts (TEST 1): 8.00 cm/s for the ESD sequence and 17.50 cm/s
for the ERS sequence.
Notes:
• longitudinal drift with starting the engine and ebb current (TEST 4): 16.50 cm/s for the ESD sequence
and 37.00 cm/s for the ERS sequence.
• average drifting speeds are given for indicative purpose only because, if used as such in the design,
they would be too misleading.

PROGRAM MENU
 N
Y

DRIFT PATH Ψ(COURSE)

θ OF SHIP
N

200° <X>
1 plot
START POSITION
every 60
seconds

Figure 9: Definition of ship drift path parameters and example of drifting simulation
 30
SPEED Y
25
DRIFTING SPEED (cm/s)

20

15
TEST 3 (LATERAL DRIFT)

10

0
0 10 20 30 40 50

TIME AFTER BREAK-OUT (s)

800
DISTANCE Y
700
DRIFTING DISTANCE (cm)

600

500

400

300

200

100

0
0 10 20 30 40 50
TIME AFTER BREAK-OUT (s)

Figure 10: example of simulation of lateral drift

PROGRAM MENU
 TABLE 2 : DETERMINATION OF THE DIMENSIONING CASES FOR LONGITUDINAL AND LATERAL DRIFTS

TESTS DRIFTING DIRECTION AFTER 30 s AFTER 35 s DIMENSIONING CASE

COMPONENT (completion of ESD (completion of ERS
sequence) sequence)
(no waves for all tests) DRIFTING DRIFTING DRIFTING DRIFTING
SPEED DISTANCE SPEED DISTANCE
(cm/s) (m) (cm/s) (m)
TEST 1: COMBINED: X -10 -1.52 -11 -2.06
wind 20 m/s from 230° LONGITUDINAL Y 13 1.88 15 -2.58
current 0.9 m/s to 40° PLUS LATERAL RHO 16 2.42 19 3.30 YES-COMBINED DRIFT
no engine, no rudder THETA 129 129
TEST 2: LONGITUDINAL X 13 1.75 16 2.41
wind 20 m/s from 350° Y
current 1.1 m/s to 250° RHO
no engine, no rudder THETA 0 0
TEST 3: LATERAL X
wind 20 m/s from 290° Y 17 2.59 20 3.51 YES-LATERAL DRIFT
current 0.25 m/s to 110° RHO
no engine, no rudder THETA 90 90
TEST 4: LONGITUDINAL X 33 3.93 41 5.78 YES-LONGITUDINAL DRIFT
wind NO Y
current 1.1 m/s to 200° RHO
engine full ahead harbor THETA 0 0

NOTE: IN CASE START ENGINE DOES NOT CONSTITUTE A REALISTIC SCENARIO, THEN THE TEST 2 BECOMES THE
DIMENSIONING CASE FOR THE LONGITUDINAL DRIFT
The above values (except for the scenario where the engine is started) are close to the values 5 cm/s and
15 cm/s traditionally used, although they are slightly above. However the drifting speeds would be
significantly higher for jetties exposed to abeam current and waves as shown for example in our Yemen
LNG project where higher drifting speeds have been taken into account in our design (in above example
the jetty is oriented as parallel as possible to the current and there is no wave; the only transverse current
occurs in reversing tide condition where the transverse current is weak).
• Although it is not shown above, it is important to note that the maximum longitudinal drift speed
component is usually obtained for combined drifting speeds (for example quarterly conditions) and
not for pure longitudinal environmental conditions.
• It is evidence, but it is good to remind that longitudinal and lateral drift speeds are not the same and
the setting-up of the alarms envelopes should take that into account.
• The drifting speed of the ship is increasing linearly after ship break-out, it is therefore in our opinion
not advisable to have the PMS anticipating the alarm 2nd step (as it is done for the alarm 1st step)
otherwise there is a serious risk not to have sufficient time to complete the ESD sequence before the
alarm 2nd step is detected.
• In the above particular example the TEST 4 (combination of start engine and ebb current) leads to too
high longitudinal speeds and the arm swiveling mechanical limit (around 45° angle) is reached before
the ESD and ERS sequences are completed (see PART 4 Figure 12 for more details). The LNG jetty
study presented here belongs to a receiving terminal where it is not possible to reduce the ESD
sequence duration below 30 s in order to avoid to generate pressures surges on board the ship (where
ESD valves close on 30 s). Therefore this scenario has been excluded from the design of the jetty and
measures to be taken so that this scenario will never happen.

Finally one may ask if the break-out scenario is not too severe and if a scenario of partial failure could be
more appropriate for the purpose of arms design (mechanical envelope) and alarms setting. The question
has been addressed and the problem is how to model accurately a progressive mooring line pattern failure.
It is most likely that there is no law of mooring lines failure propagation and two similar mooring
arrangements may lead to different results of failure propagation, even if the first failure of a mooring line
occurs in the same group of lines (for example a group of aft breast lines). In such a case the ship would
weather vane around the forward group of breast lines and an angular acceleration would result (compared
to a linear acceleration in case of ship break-out).

However the question is far to be purely academic and there is a need to understand better the
phenomenon and the association of circumstances leading to a first mooring lines and the failure
propagation. We have been informed recently that the powerful software DIODORE from Principia of
France could perform the job. We would like to take the opportunity of present paper to arouse some
enthusiasm for such a study. The problem is not specific to liquefied gases tankers, however large ships
should be addressed in priority in such study. Such a study could be sponsored jointly by S.I.G.T.T.O. and
O.C.I.M.F.

PROGRAM MENU
PART 4: EXAMPLE OF LOADING ARM DESIGN AND ALARMS SETTING BASED
ON RECENT PROJECT STUDIES
For continuity we will use the same LNG jetty project study as the one mentioned in above chapters 3.1.
and 3.2. (berthing and drifting models). Figures 11 and 12 illustrate the present PART 4 of this paper.

4.1 FLANGING AREA

It is shown on Figure 11 as a polygonal shape body obtained for a range of ships selected for this project

4.2 ALLOWANCE AREA FOR SHIP MOVEMENT AT BERTH (INTACT MOORING

PATTERN)
The maximum excursions (X motion and Y motion of the ship at berth under various sets of environment
conditions) are given in chapter 3.1 (berthing model study, Table 1):
• Lateral Y = 0.20 m rounded up to 1.00 m to be on the safe side
• Longitudinal X = 1.33 m rounded up to +/- 2.00 m to be on the safe side
We have chosen a rather important rounding here in order to be in the safe side, knowing that the arm
length would not be a concern on this project (standard arm length of 60 feet expected). This is also to
take into consideration possible greater movements due for example to lower pre-tensions in the mooring
lines or greater elasticity in case synthetic lines or tail ropes are used. On Figure 11 this area for ship
movement – lateral and longitudinal - has been exceptionally indicated (we do not give it any more on the
arms operating envelopes because it adds no significant information to the jetty supervisor or to the arm
operator. Among other things it can be seen on the top view that the shape of the so-called “drifting” area
is rectangular and not circular for conservative reasons in setting-up the alarm 1st step.
It has been kept for this paper in order to show the setting-up of the alarm 1st step envelop in case of
unnoticeable drifting speed. It can be seen on the side view of Figure 11 that the arm length will be fixed
like on many liquefied gas jetties by the smaller ships. Therefore the spherical 1st step alarm envelop is
set-up so that to be tangential to the longitudinal corners of the allowance area for ship movement at berth
at the lowest elevation of EL +12.850 m (see also cross section ‘A-A’ on Figure 2).
This alarm 1st step delimits the safe working area, i.e. a ship at berth under various weather conditions
with an intact mooring pattern. Once this alarm is reached the ESD sequence is triggered immediately.
NOTE: the fender compression used for this project is 425 mm resulting from the calculations done with
the berthing model (it departs from the value 200 to 300 mm traditionally used).

4.3 SETTING-UP OF ALARM 2ND STEP

As said above the arm length here is not a concern and more over the jetty project given in this example is
a receiving terminal, therefore a 30 s figure has been chosen for the ESD sequence duration. The design is
done keeping in mind that beyond the alarm 1st step a total failure of the mooring system has occurred as
the most catastrophic scenario.
In such conditions the drifting distances and speeds after ship break-out are given in chapter 3.2 (drifting
model study – Table 2).
The drifting distances after 30s are:
• Lateral drift (TEST 3) D1 = 2.59 m + 0.20 m for PMS accuracy = 2.79 m rounded-up to 2.80 m
• Longitudinal drift:
a) in case engine is not started (TEST 2) D2 = 1.75 m + 0.20 m for PMS accuracy = 1.95 m
b) in case engine is started (TEST 4) D2 = 3.93 m + 0.20 m for PMS accuracy = 4.15 m
• a combined drift (TEST 1) has been also analyzed in order to make sure that it will not be
dimensioning for alarm 2nd step setting, which is the case.
The PMS accuracy is somehow arbitrary taken and is added to be on the conservative side. It is assumed
to be 0.06° for arm opening angle (or apex angle), that translates into 1 to 2 cm/s for speed detection and
0.20 m for distance.
The above values (scenario start engine being excluded) are shown on Figure 11 and are used to set-up the
alarm 2nd step. Once this alarm is reached the ERS sequence is triggered immediately.

PROGRAM MENU
MAX ENVELOPE
fixed

ESD envelope
(ALARM 1st step)
fixed

1100 2800

ERS envelope
(ALARM 2nd step)
fixed

MAX ENVELOPE
fixed
12650 level
1950 900
2000
drift
900 1950 2000
drift

ESD envelope
(ALARM 1st step)
(12650 level)
fixed

ERS envelope
1000
(ALARM 2nd step)
drift
(12650 level) fixed

Figure 11: Operating envelope four arms 16”x60’ DCMA “S” satisfying TESTS 1, 2, 3

PROGRAM MENU
4.3 ARM MECHANICAL LIMIT (OUTREACH ENVELOPE)
It is assumed that the ERS sequence duration is 5 s to be on the conservative side (usually it is near 2 s
knowing that the PERC valves have to close during the ESD sequence, which come before the ERS
sequence).
Because this is the most critical phase (we reach high drifting speeds at that stage, the sequence is very
short and we are close to the arms mechanical limit) we use the maximum predicted ship drifting speed at
the time the arm flange reaches the mechanical limit in order to determine the distance between the alarm
2nd step envelop and the max outreach envelop. This builds more safety in the system rather than using
directly the distances calculated by the drifting model.
The drifting distances in 5 s beyond the alarm 2nd step envelop are therefore as follows:
• Lateral drift (TEST 3) D’1 = (0.20 + 0.02)cm/s*5s = 1.10 m, the 0.02 cm/s being for PMS accuracy
• Longitudinal drift:
a) in case engine is not started (TEST 2) D’2 = (0.16 + 0.02)*5s = 0.90 m
b) in case engine is started (TEST 4) D’2 = (0.41 + 0.02)*5s = 2.15 m
• a combined drift (TEST 1) has been also analyzed in order to make sure that it will not be
dimensioning for alarm 2nd step setting, which is the case.
The above values (scenario start engine being excluded) are shown on Figure 11 and are used to determine
the arm mechanical limit (max outreach envelop) required for the project, which in turn translates into arm
length.

IMPORTANT NOTES:
• Although the drifting speeds in this particular example are low because of the rather mild environment
conditions, the drifting distances are significantly greater than the ones commonly used to set-up
alarms, i.e. 1.50m between alarm 1st step and alarm 2nd step and 0.50 m between alarm 2nd step and
arm mechanical limit. The distances established for that particular project are respectively 2.80 m and
1.10 m for lateral drift and 1.93 m and 0.90 m for longitudinal drift.
• The above example has been chosen also purposely to draw the attention on the fact that longitudinal
and lateral drifts (and ship movements as well) are far to be the same and that the design constraints of
loading arms are far to be the same as well if we consider lateral or longitudinal drifts and movements.
This is well illustrated here by the scenario start ship engine (full ahead harbor) combined with ebb
current (1.1 m/s to 200°) pushing the ship in the same direction. We recognize this is a little academic,
but this is to make the point. This scenario is shown as impossible on Figure 12. While in this
particular project no big constraint on arm design is due to lateral drift (and this can be corrected
easily by extending the arm length), no action is however possible on longitudinal drift due to the arm
slewing mechanical limit, which is at 45° slewing angle (the main reason being the interference of
counterweights and arms each other). This scenario has therefore to be excluded and precautions
(operating procedures, etc.) to be taken so that such event will never occur.

OVERALL CONCLUSION OF THE PAPER AND ACKNOWLEDGEMENTS

Although the subject is not a new one, very few papers or literature exist on the subject treated in
detail here, which means that lasting habits have been taken and responsibilities have been
transferred to the arms manufacturers through passing years. At the same time the technology has
developed in parallel without revisiting the whole philosophy of alarms setting and alarms
detection.
It may be good to remember here a few dates in order to illustrate that point:

PROGRAM MENU
 IMPOSSIBLE AREA
Test 4

45°
Maxi

Figure 12: Verification that TEST 4 (start engine) cannot be satisfied

• The field experience with a 12” LNG arm dates from 1963
• The field experience with a 16” LNG arm dates from 1972
• The field experience with a 20” LNG arm dates from 1975
• The first PERC has been installed in 1980
• The first PMS has been installed in 1986
• The OCIMF Editions of arms design and specifications are successively from 1980, 1987 and
the long awaited 1999.

It took us three years and three important LNG projects in very different weather conditions to
develop the above synthesis and methodology, which we believe, encompasses a smooth, logic,
simple and safe approach of the problem. We hope therefore that this paper will be helpful to all
of us involved in liquefied gas jetties: engineering companies, owners and operators, arms
manufacturers, and societies like S.I.G.T.T.O. and O.C.I.M.F., whose main concern is to promote
safe operations.
We would like to conclude this paper with warm thanks to those who helped and gave their
support during these three years: Yemen LNG, Pertamina / Pt. Badak and Indigas for the three
recent projects mentioned above, FMC Europe for their consideration and interest to our
approach and methodology, and finally S.I.G.T.T.O. and O.C.I.M.F. for having listened and
understood our concerns (we raised the subject to S.I.G.T.T.O. in May 1997 and made a
presentation during the September 1997 GPC meeting; we also raised the subject to O.C.I.M.F. in
April 1997 for the consideration of the loading arms task force preparing the third Edition of the
loading arms design and construction specification, and the subject was discussed at length
during the Port and Terminal Committee in September-October 1997).

GLOSSARY OF TERMS USED IN THE PAPER

PMS: arm Position Monitoring System
ESD: Emergency Shut-Down
ERS: Emergency Release System
PERC: Powered Emergency Release Coupling
S.I.G.T.T.O.: Society of International Gas Tanker and Terminal Operators
O.C.I.M.F.: Oil Companies International Marine Forum
CEN: Comite Europeen de Normalisation
PIANC: Permanent International Association of Navigational Congresses

PROGRAM MENU
REFERENCES

Ref [1] - Design and Construction Specification for Marine Loading Arms – OCIMF 1st Edition 1980
Ref [2] - Design and Construction Specification for Marine Loading Arms – OCIMF 2nd Edition 1987
Ref [3] - Installation and Equipment for Liquefied Natural Gas – Design and Testing of
Loading/Unloading Arms – CEN Specification EN 1474 1st Edition 1997
Ref [4] - Criteria for Movements of Moored Ships in Harbors, a Practical Guide – PIANC Supplement to
Bulletin no. 88 (PTC II/1995) Report of Working Group no. 24
Ref [5] - Accident Prevention: The Use of Hoses & Hard Arms at Marine Terminals Handling Liquefied
Gases – SIGTTO Information Paper no. 4 2nd Edition 1996 (for SIGTO Members only)
Ref [6] - Design and Construction Specification for Marine Loading Arms – OCIMF 3rd Edition 1999
Ref [7] - Guidelines on the Alleviation of Excessive Surge Pressures on ESD – SIGTTO 1987
Ref [8] - Recommendations and Guidelines for Linked Ship/shore Emergency Shut-down of Liquefied
Gas Cargo Transfer – SIGTTO 1987
Ref [9] - Recommendations for Manifolds for Refrigerated Liquefied Natural Gas Carriers –
OCIMF/SIGTTO 2nd Edition 1994
Ref [10] - Recommendations for Manifolds for Refrigerated Liquefied Gas Carriers for Cargoes from 0°C
to Minus 104°C – OCIMF 1987
Ref [11] - Mooring Equipment Guidelines – OCIMF 2nd Edition 1996
Ref [12] - Prediction of Wind Loads on Large Liquefied Gas Carriers – OCIMF/SIGTTO 1985
Ref [13] - Prediction of Wind and Current Loads on VLCCs – OCIMF 2nd Edition 1994

FIGURES

Figure 1: traditional method for arms design and alarms setting, ESD/ERS detection and activation
Figure 2: the revisited methodology – alarms zones and alarms detection
Figure 3: flanging area – tabulation of X1, X2, and Y
Figure 4: flanging area – tabulation of Z1, Z2
Figure 5: consequence on arm length of approximation of polygonal flanging area by a rectangular
parallelepiped
Figure 6: recommended approach for combining arms design and alarms setting in case iterations are
necessary
Figure 7: sequences of ESD and ERS activation
Figure 8: ESD and ERS sequences activation from PMS and alarms proximity switches
Figure 9: definition of ship drift path parameter and example of drifting simulation
Figure 10: example of simulation of lateral drift
Figure 11: operating envelop four arms 16”x60’ DCMA “S” satisfying TESTS 1, 2, 3
Figure 12: verification that TEST 4 (start engine) cannot be satisfied

TABLES

Table 1: results with regard to manifold movements

Table 2: determination of the dimensioning cases for longitudinal and lateral drifts

PROGRAM MENU