DET NORSKE VERITAS Report Port toolkit risk profile LNG bunkering Port of Rotterdam, Ministry of Infrastructure & Environment, Port of Antwerp, Port of Amsterdam and Zeeland Seaport Report No./DNV Reg No.: PP035192-R2 Rev. 2, 28 August 2012
DET NORSKE VERITAS
Report
Port toolkit risk profile
LNG bunkering
Port of Rotterdam, Ministry of Infrastructure &
Environment, Port of Antwerp, Port of Amsterdam and
Zeeland Seaport
Report No./DNV Reg No.: PP035192-R2
Rev. 2, 28 August 2012
DET NORSKE VERITAS
MANAGING RISK
Port toolkit risk profile LNG bunkering
DNV Reg. No.: PP035192-R2 Revision No.: 2
Date : 2012-08-28 Page ii of iii
Table of Contents
CONCLUSIVE SUMMARY ................................................................................................... 1
1 INTRODUCTION ............................................................................................................. 3
2 LNG BUNKERING SYSTEM DEFINITION ............................................................... 5
2.1 LNG Bunkering at a glance ........................................................................................ 5
2.2 Bunker configurations ................................................................................................ 7
2.3 Location .................................................................................................................... 10
2.4 Characterization of bunker parameters .................................................................... 11
3 WHAT IS LNG? .............................................................................................................. 13
4 METHODOLGY AND SCENARIO DEFINITION .................................................... 14
4.1 Scenario’s and parameters related to safety distances for passing ships .................. 14
4.1.1 Hazard identification and Loss of containment scenarios ................................. 14
4.1.2 Selection of representative scenario .................................................................. 16
4.1.3 Calculation of effect of representative scenario ................................................ 16
4.2 Risk distances to vulnerable objects ......................................................................... 16
4.2.1 What can go wrong: Loss of containment scenarios ......................................... 17
4.2.2 How bad? Consequence Modelling ................................................................... 17
4.2.3 How often? Failure frequencies ......................................................................... 18
4.2.4 So What? Risk Assessment ............................................................................... 20
4.2.5 What do I do? Risk Management ...................................................................... 20
5 ASSESSMENTS RESULTS ........................................................................................... 21
5.1 Safety distances for passing ships ............................................................................ 21
5.1.1 Discussion .......................................................................................................... 22
5.2 Risk distances to vulnerable objects ......................................................................... 24
5.2.1 Category 1 - LNG bunkering with a small bunker vessel ................................. 24
5.2.2 Category 2 - LNG bunkering with a large bunker vessel .................................. 27
5.2.3 Category 3 - LNG bunkering with a tank truck ................................................. 30
5.2.4 Category 4 - LNG bunkering from a bunker pontoon ....................................... 32
5.2.5 Category 5 - LNG STS transfer ......................................................................... 34
5.2.6 Discussion .......................................................................................................... 36
RISK MITIGATION AND FURTHER RESEARCH ........................................................ 39
5.3 General risk mitigation measures ............................................................................. 39
5.4 Study specific ........................................................................................................... 40
DET NORSKE VERITAS
MANAGING RISK
Port toolkit risk profile LNG bunkering
DNV Reg. No.: PP035192-R2 Revision No.: 2
Date : 2012-08-28 Page iii of iii
6 CONCLUSIONS .............................................................................................................. 42
7 REFERENCES ................................................................................................................ 43
Appendix 1 Scenarios
Appendix 2 Nautical risk assessment of moored LNG bunker vessels
Appendix 3 Background parameters for risk calculation
Appendix 4 Results distances to 10-6
per year risk contour
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 1
CONCLUSIVE SUMMARY
The Port of Rotterdam, the Port of Antwerp, the Port of Amsterdam and Zeeland Seaports are
preparing for the arrival of LNG as a fuel. Large scale bunkering of LNG is novel and differs
on several aspects from the bunkering of conventional marine fuels. LNG is stored at low
temperatures and the development of a gas cloud in case of a potential release to the
atmosphere requires insight into the risks which have to be translated into procedures for safe
and practical operation during For successful incorporation of these activities into their
current safety systems (e.g. guidelines, operational procedures) and operations they DNV has
been asked to develop a “harbour toolkit safety distances LNG bunkering” to help identify (a)
the safety distances for passing ships and (b) the risk distance to vulnerable objects such as
residential housing, offices, hospitals etc. The toolkit will help get insight what safety/risk
distance should be taken into account given a specified bunker configuration and as function
of the number of bunkers. As such it can be used as a (first) screening tool for suitability of
bunker locations in the port area.
The Port of Rotterdam identified various LNG bunker activities in a port area. Those activities
can be grouped in five different categories:
1) LNG bunkering from small inland bunker vessel to small vessels
2) LNG bunkering from large bunker vessel to seagoing vessels
3) LNG bunkering from trucks to small vessels
4) LNG bunkering from bunker pontoons to small vessels
5) LNG transfer from ship to ship
The determination of the safety distances for passing ships is determined by a consequence
based methodology, where a representative risk scenario and consequence are selected. The
selection of the representative scenarios is based on a (desktop) hazard identification.
DNV found that the calculated safety distance for categories 1, 3 and 4 are in line with the
nautical safety distances that are prescribed in the existing Dutch shipping regulation BPR.
The calculated safety distances for the categories 2 and 5 are significantly larger than the
nautical safety distances that are currently prescribed in the existing Dutch shipping
regulation BPR. The relatively large safety distances that are found for those categories could
enforce additional rules for the LNG bunkering of large vessels in area with intensive nautical
traffic. It may not give restriction on LNG bunkering activities in area where lower nautical
activities take place and collision scenarios are less credible / likely.
The risk distances to vulnerable objects are calculated using a Quantitative Risk Assessment
(QRA) methodology which includes both frequency and consequences calculations of
possible Loss of Containment (LOC) scenarios (e.g. hose leakage). DNV have calculated the
risk distances to vulnerable objects towards the Dutch risk criteria for vulnerable objects, i.e.
10-6
/year.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 2
The risk distances that are found for the different categories vary from 10 to 510 meter,
depending on the category, bunker parameters, ignition method and number of bunker
activities. The risk distances for the categories 1, 2 and 5 are mainly caused by the scenario
where ship collision results in a loss of containment of the LNG cargo (bunker) tank on the
LNG bunkering vessel. It is found that lowering of the nautical activities in the bunkering area
may reduce the risk distance. For areas where (very) low nautical activities take place the risk
distance is driven by the rupture of the bunkering hose and leakage through a 25 mm hole.
It is found that the risk distance of category 3 and 4 is independent of the nautical activities
because the distance is driven by the scenario ‘rupture of the bunkering hose’ and leakage
through a 25 mm hole.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 3
1 INTRODUCTION
Liquid Natural Gas (LNG) is becoming more and more interesting as fuel for shipping. One
of the main reasons is the upcoming stringent requirements for emissions from ship engines
following the adoption of Emission Control Area’s as of 2015. A ship engine running on LNG
is emitting significant lower amounts of NOx, SOx and particulate matter as well as reduces
the output of CO2 than a ships engines running on conventional fuels. Next to that costs
saving may be achieved.
In order to use LNG as a shipping fuel there is a need to develop a LNG bunkering
infrastructure. Several bunkering configurations are possible to deliver the LNG to the vessel.
An example of a LNG bunkering configuration already used today can be seen on the DNV
LNG blog [1].
Figure 1: Screenshots bunkering film
Large scale bunkering of LNG is novel and differs on several aspects from the bunkering of
conventional marine fuels. LNG is stored at low temperatures and the development of a gas
cloud in case of a potential release to the atmosphere requires insight into the risks which
have to be translated into procedures for safe and practical operation during LNG bunkering
activities.
Different ports are preparing for the arrival of LNG as a fuel. For successful incorporation of
these activities into their current safety systems (e.g. guidelines, operational procedures), the
Port of Rotterdam, Ministry of Infrastructure & Environment, Port of Antwerp, the Port of
Amsterdam and Zeeland Seaports asked Det Norske Veritas (DNV) to determine the
following:
Safety distances for the determination of exclusion zones related to passing vessels during
LNG bunkering activities, i.e. what is a safe passing distance for other traffic related to an
ongoing LNG bunkering activity?
Risk distances to vulnerable objects, i.e. what should be the minimum distance between an
LNG bunker location and (fixed) vulnerable objects such as residential housing, offices,
hospitals etc. based on the quantified risk. The purpose is to develop a risk-based toolkit,
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 4
with which suitable locations for LNG bunkering activities can be identified in the port at
any given time.
Concerning the bunkering of LNG a significant amount of uncertainties still exist, e.g. related
to ship designs (both for bunker vessel and recipient vessel), vessel types, bunkering
equipment, process parameters etc. as there is only a limited number of LNG bunker stations
operational to date In this study considerable effort has been made to make defendable
assumptions but in case choices needed to be made the conservative option have been chosen
in an attempt not to underestimate the risks related to LNG bunkering.
And although in this study a lot of effort has been put in making the assessment as detailed
and realistic as possible, future real life bunker configurations might have different parameters
and characteristics. As such the results presented should be interpreted with care and as a
minimum real life situations should be verified in detail to see whether their parameters are in
line with the assumptions made in this study before conclusions are drawn based on this
study.
The report outline is as follows: Chapter 2 gives an overview of the LNG bunkering system
definition. This chapter include a brief overview of the history of LNG and a description of
the categorised LNG bunkering activities considered in this study. Next, a brief overview of
the (safety) characteristics of LNG is presented in Chapter 3. Chapter 4 describes the applied
methodology and scenario definition to determine the safety and risk distances. A detailed
assessment of the results is provided in Chapter 5. Chapter 6 provided general and specific
mitigation measures to reduce the risk that is found in Chapter 5. The report ends with
Chapter 7 in which the main findings are summarised and conclusions are drawn.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 5
2 LNG BUNKERING SYSTEM DEFINITION
In order to correctly assess and quantify the risks of LNG bunkering definition of the
following key aspects are important:
The configuration of bunkering: several bunkering configurations are feasible and each
configuration has specific risks;
The location of bunkering: the vicinity and number of passing vessels, presence of
ignition sources, distance to vulnerable objects and such all can have a significant
influence on the risk level;
Bunker parameters: the number of bunker operations, the volume of the LNG flow, and
the characteristics of safety equipment, e.g. time needed to close Emergency Shut Down
(ESD) valves and other also have a significant impact on the risk level.
In the paragraphs below a more detailed description of these various aspects is given. A short
description of the LNG bunker process is also presented.
2.1 LNG Bunkering at a glance
The typical main steps in a bunkering process are:
Approach of the bunker vessel
Setting up mooring arrangement
Grounding and connecting of the bunker hoses;
Inerting and purging of the filling lines;
The bunker transfer;
Stripping, purging and inerting of filling lines;
Disconnecting of grounding and bunker hoses.
The figure below is an illustration of the different elements in a LNG bunkering transfer
chain. Focus is on the different equipment and the systems involved.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 6
Figure 2: Different elements in a LNG bunkering transfer chain
LNG fuelled ships have bunkered successfully for ten years, since the first LNG fuelled ship –
a car and passenger ferry named Glutra – was put into operation in 2000. Currently, a total of
26 LNG fuelled ships are in operation, and ~30 more new-builds have been ordered. In
addition, some conversions are planned for convention fuelled ships.
The bunkering operations currently in use for LNG fuelled ships in Norway are bunkering
from trucks or from a land based tank via a fixed installation on pier or jetty. Trucks are
common due to small scale supply, long distances between LNG production sites and
bunkering sites, as well as between locations where the LNG fuelled vessels operate.
Last year a LNG bunkering JIP in the Netherlands (LESAS project) started to investigate the
possibility of LNG ship bunkering. Recently, the same initiative is launched in Singapore.
The LNG bunkering JIP in Singapore identified the (manual) bolted connection of LNG
hoses, as used in Norway at the moment, to be disputable. They prefer breakaway couplings
which can be equipped with ESD2 option for rapid disconnection of the transfer hose / from
the ship.
Next to that the loading and unloading of cryogenic substances has been around for decades.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 7
2.2 Bunker configurations
The Port of Rotterdam has identified various LNG bunker configurations in a port area and
created a virtual port to visualise those bunker configurations. This virtual port is known as
the Beanport after its creator Cees Boon. An overview of the Beanport is found in Figure 3.
Some of these bunker configurations are classed as land base, e.g. bunkering from an onshore
storage facility. These configurations are not part of the scope of this study. Instead this study
focuses on the water based configurations with one addition: the bunkering of small vessels
by truck. The water based configurations can be divided in five different groups:
1) LNG bunkering from small inland bunker vessel to small vessels
2) LNG bunkering from large bunker vessel to seagoing vessels
3) LNG bunkering from trucks to small vessels
4) LNG bunkering from bunker pontoons to small vessels
5) LNG transfer from ship to ship
Each group is discussed in more detail and represented in figure 3 below.
Figure 3: Overview of the LNG bunkering configurations in “Beanport”
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 8
1) Bunkering of LNG from a small inland bunker vessel
Small inland bunker vessels will load LNG from large scale or intermediate LNG terminals
and transport it to bunker locations. At the bunker location the LNG is bunkered to small
(inland) vessels through a flexible hose. The size of the bunker barge is strongly depending on
the number of bunker operations and the bunker volume of the recipient vessel to be
bunkered. Bunkering of short sea vessels and short sea Ro/Ro ferries does require more LNG
bunker volume than inland vessels. The bunkering of LNG from a small inland bunker vessel
will take between 1 or 2 hours depending on the LNG demand of the bunkered vessel and the
bunkering flow rate. Bunkering flow rates vary from 80 to 500 m3/h.
The volume of LNG per cargo tank is limited for small inland bunker vessel. The ADN [1]
limit the volume of cargo tanks for standard inland vessel to 380 m3, which means that cargo
tanks of bunker vessels cannot be constructed above this threshold value. Bunker barges
which will transport more LNG must be equipped with multiple cargo tanks.
After special permission the ADN [1] allows cargo tanks with a volume up to 1000 m3. To
comparison; The Pioneer Knutsen, which is at the moment the only LNG small bunker vessel
in operation, contains two spherical stainless steel 550 m3 cargo tanks. This bunker vessel
operates at 3 bar(a) [4].
2) Bunkering of LNG with use of a large LNG bunker vessel
Large LNG bunker vessels will load LNG from large scale or intermediate LNG terminals
and transport it to the bunker locations. At the bunker location the LNG is bunkered to small
vessels which could vary from tankers to large container ships. The size and main dimensions
of small scale LNG carriers can vary significantly, depending on different market demands,
draught and other physical limitations of the ports and bunker sites to be used. Typical
cargo capacity for small scale LNG carriers may be approximately 7.000 to 21.000 m3, but
smaller and bigger vessels exist. According to the information received by the Port of
Rotterdam the large container ships require a maximum bunker volume of 21.000 m3, which
means that the entire volume of a large LNG bunker vessel is needed to bunker one container
vessel.
The LNG bunkering time of vessels in this study is limited to 7 hours. To meet the required
bunkering volume in 7 hours LNG is bunkered by 3 flexible hoses. Bunkering flow rates per
hose vary from 500 to 1000 m3/h.
3) Bunkering of LNG with use of LNG tank truck
Regional land-based distribution of LNG can be carried out by heavy duty trucks, for example
to serve nearby industries, other ports in the region and transportation within the port. LNG
trucks are also used for transporting LNG from small scale liquefactions plants to customers
who are not connected to the gas network. Examples of countries where LNG is distributed by
trucks are Norway, Sweden, Finland, Belgium, Germany, the Netherlands, Poland, Spain,
Turkey, China and Russia. LNG terminals with regional distribution of LNG by trucks are
equipped with facilities for loading and unloading of trucks. Flexible hoses are used for the
transfer of LNG between the terminal and the truck. The size of the truck is in Europe limited
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 9
by the Accord européen relatif au transport international de marchandises Dangereuses par
Route ADR to 40 m3 [3]. A normal bunkering operation from a semi-trailer takes up to two
hours including signing of documents and safety procedures. The actual pumping / transfer
time is approximately one hour.
According to the information received by the Port of Rotterdam short sea Ro/Ro vessels
require a maximum bunker volume of 200 m3, which means that multiple trucks are required
to bunker a single Ro/Ro vessel. Inland vessel require less bunkering volume. The first LNG
fuelled inland vessel, Argonon, requires 40 m3 of LNG. This inland vessel is bunkered with a
single LNG truck.
4) Bunkering of LNG with use of a fixed land based tank/installation (bunker pontoon)
Bunker pontoons and other small scale land based LNG installation will load LNG from LNG
feeder vessels. Small scale land based LNG installation in port area could be as large as
100000 m3. The volume of land based installations such as bunker pontoons, is much lower.
Here, typical volumes up to 1000 m3 can be expected. The smaller land based installations are
likely to be used for the bunkering of inland vessels, harbour tugs or fishing vessels or even
trucks. The bunkering of LNG from small scale installations to inland vessel will take
approximately 1 hour depending on the LNG demand of the bunkered vessel and the
bunkering flow rate. Bunkering flow rates vary from 30 to 80 m3/h.
5) LNG transhipment on the buoys or dolphins (ship to ship transfer)
Most of the cargo operations normally take place between the large or small bunker vessels
and intermediate LNG storage terminals. Nevertheless, it is possible to transfer LNG from a
LNG feeder to a bunker vessel. During this ship-to-ship (STS) transfer the LNG bunker
vessel will moor alongside the LNG feeder to transfer LNG. The advantage of the transfer
method is the absence of an immediately land based terminal to transfer LNG to bunker
vessels. The flow at which the LNG is transported is highly depending on the size of the
receiving bunker vessel. LNG STS transfer can be applied to both large and small bunker
vessels.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 10
2.3 Location
Although the outcome of this study will be used to assess the suitability of proposed bunker
locations, a definition of representative locations is needed in order to reach this outcome.
Together with the port of Rotterdam it was decided two simulate three situations:
1. An area with intensive nautical activity: The intensive nautical traffic area is defined as a
location where large ships with high velocity will pass the moored LNG bunker vessel.
The potential impact energy of this scenario is high due to the combination of large ships
with high velocities. In the Port of Rotterdam area, de Oude Maas could be seen as a
representative case for an intensive nautical traffic area.
The average traffic density on the Oude Maas is estimated at approx. 108,000 ships per
year. The majority of the ships are less than 140 meter and have a speed around 8 knots.
This combination of length (mass) and velocity results in a high potential impact energy
on the waterway.
2. An area with low nautical activity: The low nautical traffic area is defined as a location
where the traffic density is less than the intense nautical traffic area. At this location the
average velocity of the ships is low and large vessels are supported by tugs which limits
the probability of collision. The potential impact energy at this location is low.
Representative locations for this scenario could be found in the Caland canal area.
The average traffic density on the Caland canal is estimated at approx. 48,500 ships per
year. The majority of the ships is less than 140 meter and have a speed around 8 knots. At
this location the combination of length (mass) and velocity also results in high potential
impact energy on the waterway. However the number of ships is significantly smaller than
the Oude Maas, which significantly reduces the probability of collision.
3. An area with very low nautical activity: The very low nautical traffic area is defined as a
location where ships are going to be berthed. At this location the average velocity of the
ships is low (<5 knots) and large vessels are supported by tugs which limits the
probability of collision. The potential impact energy at this location is low. Representative
locations for this scenario could be found in a dock in the Amazonehaven area.
The average traffic density on the Amazonehaven area is estimated at approx. 4,200 ships
per year. At this location the average velocity of the ships is low and large vessels are
supported by tugs. The potential impact energy at this location is low.
All three areas will be used for the simulation of the risk associated with LNG bunker
activities. For each of the locations the width of the waterway is estimate on 300 meter. The
locations are displayed in Figure 4.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 11
Figure 4: Overview of the location of nautical traffic areas.
2.4 Characterization of bunker parameters
A final key aspect that influences the overall risk levels is the characteristics of the bunkering
process itself: e.g. the volume of the LNG flow or the number of bunker activities has a
significant impact to the overall risk levels. To account for this some assumptions had to be
made. These assumptions are described below.
Number of bunker activities
Since the number of bunkering activities will increase over time (as more gas fuelled ships
will become available) three levels of bunker activity have been defined per bunker
configuration: an upper bound, mean and lower bound level. The number of bunker activities
per level is an indication for the visualisation of the risk. They do not represent the expected
number of bunker activities in the Port of Rotterdam.
The upper bound level of bunker activities for the categories 1, 3 and 4 are 10 per day. The
mean and lower bound level for these categories equal 5 bunker activities per day and 1
bunker activity per week.
For the category 2 the upper bound level of bunker activities is 1 per day. The mean and lower
bound level for category 2 equal 1 bunker activities per 2 days and 1 bunker activity per
month.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 12
For the last category, category 5, the upper bound level of bunker activities is 3 per week. The
mean and lower bound level for category 5 equal 7 bunker activities per month and 1 bunker
activity per month.
Process characteristics
Since the actual bunkering infrastructure is not yet available there is uncertainty in what the
bunkering process parameters such as flow and pressure will be. To solve this situation two
sets of process parameters have been used based on current comparable processes as well as
known designs. Per bunkering configuration a minimal and maximal process parameter set
has been defined. The variations in parameters include flow, diameter, bunker time, number
of hoses and ignition method. Quantification of these parameters is given in appendix I and
the addendum [13]. It must be note that the risk calculation with the maximum parameter set
is based on the conservative ignition method where the flammable cloud is ignited at its
largest volume. The minimum parameter set is based on the less conservative ignition method
where the actual ignition sources like passing ships are taken into account. More information
regarding ignition sources is found in Appendix III.
Based on the above the following simulation scenarios are possible for each of the five bunker
configurations. This leads to a total of 90 simulation scenarios.
Figure 5: Overview of various simulation scenarios per bunker category
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 13
3 WHAT IS LNG?
LNG is Liquefied Natural Gas and as such is the same gas (mostly consisting of methane) that
is used for cooking in many homes. In its liquid state, LNG is not flammable, nor explosive.
When LNG is heated and becomes a gas, the gas is not explosive if it is unconfined. Natural
gas is only flammable within a narrow range of concentrations in the air (5% to 15%). Less
air does not contain enough oxygen to sustain a flame, while more air dilutes the gas too
much for it to ignite. Ignition without ignition source (auto ignition) is not possible in normal
conditions. The temperature at which auto ignition may occur is above 500°C (in an air-fuel
mixture of about 10% methane in air, the auto ignition temperature is approximately 540°C,
while the auto ignition temperature for diesel oil is in the range of 260°C to 371°C).
In the event of a spill, LNG vapours will disperse with the prevailing wind. Cold LNG vapour
will appear as a white cloud. Parts of that cloud contain flammable concentrations of gas. The
flammable concentrations of gas could be ignited, which can results in a jet fire, flash fire or
explosion (if confined). The probability of explosion could be limited by a good design of the
facility or vessel which is fuelled by LNG. This means that facilities or vessels should have an
open design where confinement is limited, so no significant overpressures can be built up
after ignition.
Localized jet or flash fires would burn with intense heat. To keep the public at a safe distance,
thermal exclusion zones are established for installation that handle LNG. When LNG is
released the liquid droplets rain out and may form a pool of LNG. The LNG pool cannot be
ignited but the flammable concentration above can. Ignition of the flammable concentration
above the pool will result in a pool fire. For pool fires also thermal exclusion zones are
established to keep the public safe.
When skin touches an extremely cold body or LNG, heat is transferred from the skin and
organs to the cold body or LNG. This will cause damage to the skin and underlying tissues.
The normal functioning of the body may be disturbed by the cooling of internal organs, which
will lead to a critical condition called hypothermia. The cooling of the brain or heart is very
dangerous. Proper procedures and the use of protective clothing and equipment to prevent any
contact with the LNG are hence imperative. Large scale exposure to LNG will cause a
fatality.
However, the extremely low temperatures are not only hazardous to people. While stainless
steel will remain ductile, carbon steel and low alloy steel will become brittle and fractures are
likely if they are exposed to such low temperatures. Standard ship steel must therefore be
protected and insulated from any possible exposure to LNG (e.g. using stainless steel drip tray
etc.).
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 14
4 METHODOLGY AND SCENARIO DEFINITION
This chapter will discuss the methodology that is used for the determination of the safety
distance to passing ships and the methodology that is used for the determination of risk
distances to vulnerable objects.
4.1 Scenario’s and parameters related to safety distances for passing ships
The determination of the safety distances for passing ships is determined by a consequence
base methodology, where a representative scenario and consequence are selected. The
methodology is illustrated in Figure 6.
Figure 6: Methodology of determination of safety distance
4.1.1 Hazard identification and Loss of containment scenarios
The selection of the representative scenarios is based on a (desktop) hazard identification
which is focussed on hazards that can results in a loss of containment of LNG. During the
hazard identification the cause, consequences and credibility of each of the hazards were
identified. It is reasonable to assume to the overfilling of the fuel tank and improper boil of
gas control do not occur if proper measures are in place. Hazards that arise from the
intermediate LNG storage and/or fuel tank are not considered within the scope of this study.
The other identified hazards that could occur are grouped in two different categories:
Coupling failure
Damage to the hose
These two categories will be discussed below in more detail:
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 15
Coupling failure
Before the bunkering operation the hose is connected to the ship’s manifold. The connection
should be established by operators which could make an operational error while connection
the hose. It is assumed that the system is tested (purged) prior to each bunkering operation,
using nitrogen as inert gas. After the bunkering of LNG is finished the LNG hose is purged to
prevent possible releases of LNG when disconnecting the hose.
Nevertheless, a leak could occur at the flange face, resulting in an initial slow release, with
little impact and growing to a wire cut across the flange face. The operator, which is present
during the loading operation, would detect the leak and will shut down the installation. The
shutdown action of the operator would isolate the leak and further release of LNG is
prevented. A conservative estimate of the wire cut diameter/hole size that could occur during
this incident would be between 5 – 10 mm.
Hose Failure
There are various failure mechanisms for (flexible) hoses. For the LNG bunkering purpose the
following failure mechanisms where identified:
Fatigue due to high pressure or low temperature;
Ship securing/ mooring line failure;
Collision of ships;
Extreme weather conditions
External impact due to lifting activities or maintenance
The flexible hoses that are used for the bunkering of LNG are in European countries subjected
to the Pressure Equipment Directive (PED). The PED prescribes periodic inspection of
flexible hose if a certain threshold value of pressure and diameter is exceeded. European
design standard EN1472-2 [7] states that he maximum allowable working pressure in a hose
should not be less than 10 bar(g). For pressures of 10 bar(g) and higher the PED prescribes a
periodic inspection for hoses with a diameter above 2.5 inch. To ensure the technical integrity
of the hose, the periodic inspection should be performed by an independent party.
In industry it is also common to perform a visual inspection before the hose is connected. It is
likely to assume that the inspection by bunkering company and independent party will secure
the technical integrity of the bunker hose and failure or rupture of the hose is therefore not
selected as credible scenario.
During the bunker operation the receiving and bunker vessel are connected with mooring lines
to prevent drifting. For all bunkering operations the receiving vessel is fixed to the shore as
well. In case of collision the mooring lines will (partly) absorb the first impact energy.
Colliding vessels with low amount of impact energy (e.g. low mass and/or velocity) will not
have sufficient energy to rupture the mooring lines and loss of containment is not expected.
For higher impact energies the mooring lines can fail and the tensile strength of the bunker
hose could be the limiting factor for loss of containment. However coupling of flexible hose
for gas transfer are equipped with a breakaway coupling to limit the spilled volume.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 16
The traffic density on the waterways around the Port of Rotterdam is high. Therefore collision
scenarios are not negligible and collision of passing vessels could be seen as a credible
hazardous scenario.
4.1.2 Selection of representative scenario
In the above analyses two representative scenarios are identified as credible scenarios;
leakage at the flange face with a hole diameter between 5 – 10 mm and disconnection of the
breakaway coupling in case of a collision scenario.
For bunker locations where the traffic density is relatively high the safety distance could be
best determined with the disconnection of the breakaway coupling scenario. This selected
scenario would be applicable for the Rotterdam port area where the traffic density is high.
4.1.3 Calculation of effect of representative scenario
The results of the effects of representative scenarios are presented in the results chapter. Here
the maximum effect of the representative scenario is used to determine the safety distance.
4.2 Risk distances to vulnerable objects
The risk distances to vulnerable objects are calculated with using a Quantitative Risk
Assessment (QRA) methodology. The QRA methodology is a well-known and widely
accepted approach to determining risk levels associated with Loss of Containments (e.g.
spills). The modeling practice is described in the Dutch Reference Manual Risk Assessments
[4].
A QRA gives insight into the risks to human life of a certain activity by calculating the
potential effects of a variety of scenarios as well as considering the probability of occurrence
of these scenarios.
A QRA tries to answer five simple questions. Beside each question, the technical term is
listed for that activity in the risk assessment process:
What can go wrong? Hazard Identification
How bad? Consequence Modelling
How often? Frequency Estimation
So What? Risk Assessment
What do I do? Risk Management
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 17
4.2.1 What can go wrong: Loss of containment scenarios
The individual risk contours will be calculated for all the 60 individual bunkering scenarios
which are defined in Figure 5. A detailed scenario definition is provided in appendix I. Note
that for each simulation scenario separate loss of containment scenarios are defined, which are
reported in the next paragraph.
During bunkering activities loss of containment (LOC) might occur due to various reasons
(e.g. external ship collision, failure of hoses) and at several locations. As mentioned earlier,
only potential LOC scenarios are taken into account during bunkering activities (i.e. failure of
hoses or tanks on bunkering vessels due to external ship collision). LOC scenarios related to
failures of storage tanks on bunkering vessels or inland bunker pontoons/tank trucks are
outside the scope of this study and therefore not considered.
The loss of containment scenarios used for the risk calculations per bunkering operation are
specified in Table 1.
Table 1: Loss of containment scenarios
Scenario Description Hole size
(mm)
1 Hose leakage 5
2 Hose leakage 25
3 Hose rupture Full bore
4 Tank leakage (only cat 1,2,5) 250
5 Disconnection of hose due to
ship collision
Full bore
The hose failure scenarios are representative for failure of hoses taken from the ARF
document [6], which gives a suggestion for typical hole size diameters of leakages that could
be taken into account when considering loss of containment during liquefied gas transfer in
hoses or arms. The full bore rupture hole size varies for each bunkering activity/scenario setup
depending on the typical hose diameter used (appendix I). For ship collisions between passing
ships and bunkering or receiving vessels additional rupture scenarios are defined. The tank
leakage scenario is only applicable in case ship collisions between passing ships and
bunkering ship are possible. For a complete definition regarding potential loss of containment
caused by ship collisions, a reference is made to Appendix II.
4.2.2 How bad? Consequence Modelling
In parallel with the frequency analysis, consequence modeling evaluates the resulting effects
if the accidents occur, and their impact on personnel, equipment and structures, the
environment or business. In chapter 3 it is commented that natural gas is flammable in a
narrow range of concentrations (5% to 15% in air). When ignited it can results in, jet,-, pool, -
or flash fire depending on the time of ignition and place of ignition. Explosions could only
occur when ignited flammable concentrations of gas are enclosed. The consequences of the
fire are mostly dependent on the loss of containment parameters and the process conditions
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 18
during the release. The loss of containment scenarios are discussed above. The process
conditions are discussed below.
Process conditions
European design standard EN1472-2 [7] states that he maximum allowable working pressure
in a hose should not be less than 10 bar(g). According to internal sources within DNV a
typical pressure in a bunker hose is around 5-6 bar(g). For this study a generic bunkering hose
pressure of 5 bar(g) (stagnant, absolute pressure) is assumed, which is independent of type of
bunkering activity.
Pressure in the storage tank of the bunker vessel/pontoon or tank truck is estimated at 2
bar(g). The Swedish Marine Technology Forum (together with other organizations in a joint
industry project) has developed a LNG bunkering Ship to Ship procedure, which is in
principle accepted and approved by DNV [8]. The procedure states that a LNG bunker ship
may be equipped with an insulated storage tank type C for liquefied natural gas, which could
contain around 1000 m3 at 3 bar(g) and -163°C. However, internal sources within DNV state
that a typical operating pressure of LNG tanks in vessels would be closer to 2 bar(g). The
latter pressure is used for the risk calculations for any type of bunkering vessel or LNG tank
truck for that matter. A recent QRA study carried out by DNV confirms that storage pressure
of LNG in tank trucks is typically equal to 2 bar(g). Based on the above mentioned pressures,
it is reasonable to assume that the pump head is sufficient to realize the flows for each
bunkering scenario provided by the Port of Rotterdam (ranging from 30 m3/hour for category
4, minimal transfer parameters and 1500 m3/hour for category 5, maximal transfer
parameters).
A typical LNG storage and transfer temperature of -162°C is used, under the assumption that
bunkering vessels/installations have the ability to maintain the temperature constant by
handling/escaping the boil-off vapors to, for instance, a compressor and subsequently a re-
condenser for liquefaction. Tank trucks are usually equipped with double-walled tanks with
vacuum and insulation between the outer (carbon steel) and inner (aluminum) tank in order to
maintain the low temperature.
4.2.3 How often? Failure frequencies
The frequencies given in ARF are based on road or rail tanker transfer accident data and are
therefore not suitable to use for this study. The technical notes of DNV for process failure
frequencies [9], proposes a base frequency of 6.77 x 10-5
/visit for failure of loading arms
(resulting in leakages). It is the best available data and is taken from the ACDS document
from 1991 [10]. This frequency is based on liquefied gas transfers using articulated arms
(rather than transfer hoses). Furthermore, incidents reported in the ACDS document are
dominated by LPG spills rather than LNG spills. Nonetheless, the frequency is considered to
be “conservative best-estimate”, if not ‘upper bound’ for LNG transfers. The base frequency
of articulated arms is factored with the following aspects to obtain the frequency that is used
for transfer hoses:
The base frequency is per visit and must be converted into failure frequency per hour per
hose
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 19
The base frequency is based on failure of articulated arms rather than hoses. Therefore a
factor is applied to the base frequency.
A reference is made to the Addendum [13] for the assumptions made to factor the base
frequency. The frequency distribution between of respective loss of containment scenarios is
based on the Dutch guideline for risk calculations (HARI) [4], which states that 10% of all
leaks consist of rupture scenarios. The ARF document states that 10% of all small leaks are
ruptures, which is more or less in the same order of magnitude. The frequencies of the smaller
leaks (5, 25mm) are equally distributed, which is in agreement with the ARF document.
Table 2: Loss of containment scenarios and likelihood
Scenario Description Hole size
(mm)
Frequency
(1/hour/hose)
Frequency
distribution (%)
1 Hose leakage 5 1.5 x10-6
45%
2 Hose leakage 25 1.5 x 10-6
45%
3 Hose rupture Full bore 3.4 x 10-7
10%
4 Tank leakage (only cat 1,2,5)* 250 - -
5 Disconnection of hose due to ship
collision*
Full bore - -
* Failure frequency is dependent on the level of nautical activity in the bunkering area (a
reference is made to Appendix II for a specification of these frequencies)
Intervention times of operators and EMS/ESD systems in place
Measures such as the presence the operators, emergency shutdown systems (ESD/EMS)
during transfer might mitigate discharge effects by tripping pumps or closing valves in case of
a LOC. Small intervention times, which usually vary for each mitigation measure, are
essential to significantly limit the amount of material being discharged during loss of
containment.
For operators, the intervention time is taken from HARI [4] and is equal to 120 seconds. The
following conditions have to be met to ensure an intervention time of 120 seconds by an
operator is achievable:
The operator has the possibility to visual monitor the hose during the entire transfer.
The presence of an operator is ensured by either a dead man’s switch or a procedure in the
safety management system. These measures should be inspected on a regular basis.
Manual activation of an emergency shutdown during loss of containment by an operator
should be well-documented in a procedure.
The operator should be well-trained and is also familiarized with the procedures
applicable.
The emergency shutdown button should be positioned according to applicable rules and
standards, which ensures fast, manual activation independent of release direction in case
of loss of containment.
The probability of failure on demand (PFD) of an operator is 0.1 and is also taken from HARI
[4].
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 20
A fully automatic emergency shutdown system can detect leakages (gas detection, flow
measurements) and are able to trip pumps/close valves automatically. Operation intervention
is not necessary in case the EMS/ESD system is working properly. The PFD of automated
shutdown systems is equal to 0.001 (source: HARI). In the event that an automated shutdown
system fails, an operator is still able to initiate the shutdown manually. Intervention time is
assumed to be equal to 20 seconds for the 25mm hole and full bore rupture scenarios based on
data provided by the Port of Rotterdam. It must be noted that the exact duration of
intervention would be highly dependent on the design of the intervention system in place. It is
reasonable to assume that a small leak will take longer to detect automatically and as such, the
intervention time for the 5 mm hole size scenarios is set to 120 seconds.
An EMS/ESD system and operator with each a response time of 20 and 120 s, respectively, is
present at all bunkering vessels and installations (bunkering categories 1, 2, 4 and 5). Note
that ESD systems might not be available for tank trucks (cat 3). For category 3, LNG
bunkering with a LNG tank truck, only intervention of an operator is deemed possible.
All flexible bunkering hose are equipped with a safety breakaway coupling. The breakaway
coupling is a passive device that is located between the bunker hose and the receiving vessel.
For external impact scenarios, like ship collision, the breakaway coupling will disconnect the
bunker hose and immediately close the outflow area. The closure of the outflow area will be
mechanical driven and last for less than a second. It is reasonable to assume that the closure of
the breakaway coupling will be less than 5 seconds. Based on breakaway manufacturers’
information, the shut-off valves inside the breakaway coupling close immediately in the event
of sudden disconnection (i.e. less than 1 second). As such, 5 seconds reaction time can be
considered as a conservative estimate.
4.2.4 So What? Risk Assessment
Up to this point, the process has been purely technical, and is known as risk analysis. The next
stage is to introduce criteria which are yardsticks to indicate whether the risks are
“intolerable” or “negligible” or to make some other value-judgment about their significance.
This step begins to introduce non-technical issues of risk acceptability and decision making,
and the process is then known as risk assessment.
The Dutch risk criteria are implemented in the Decree External Safety Establishments 2011.
For this study the Individual risk criteria is used to assess the calculated risk related to LNG
bunkering activities. The Dutch Individual risk criteria states for vulnerable objects, a risk
limit value of 10-6
per year must not be exceeded. For objects with limited vulnerability, the
same value applies as an orientation norm and may be exceeded under certain conditions.
4.2.5 What do I do? Risk Management
In order to make the risks acceptable, risk reduction measures may be necessary. The benefits
from these measures can be evaluated by repeating the QRA with them in place, thus
introducing an iterative loop into the process. Detailed investigation of risk mitigation
measures and their impact of the risk calculation is not part of the scope of this study.
However chapter 0 of this rapport gives a summation of risk mitigating measures that could
be used to lower the risk related to LNG bunkering.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 21
5 ASSESSMENTS RESULTS
This chapter will give an indication of the safety distance for passing ships as well as the risk
distance to vulnerable objects. Both distances are given for the categories defined in chapter
2. The calculation of the distances is based on the scenarios defined in chapter 4. The
background information, e.g. weather data, ignition sources and general risk parameters, that
are used for the risk calculation are enclosed in Appendix III.
5.1 Safety distances for passing ships
This section determines the indicative safety distance for all bunkering scenarios. Thereafter
the determined safety distances are compared with the safety distances that are currently in
place in the Rotterdam port area. The section is closed with a sensitivity analysis of two key
parameters.
As stated earlier the safety distance can best be determined based on the disconnection of the
breakaway coupling scenario. The released volume and corresponding consequence is
strongly depending on the flow rate and closure time of the breakaway coupling. For the
determination of the safety distance the maximum bunker parameters are used. For the closure
time of the breakaway coupling a value of 5 seconds is considered.
The safety distance is based on the maximum effect of the selected scenario. The maximal
effect for the disconnection of the breakaway coupling is a flash fire. A flash fire could occur
when the released flammable cloud is ignited. This ignition could occur till the Lower
Flammable Limit (LFL) concentration. The weather type wherefore the flash fire contour is
calculated is stability class D and a wind speed of 5 m/s.
The safety distance is determined as the LFL distance corresponding to the released volume
of a disconnected breakaway coupling. The determined safety distances are summarised in
Table 3.
Table 3: Safety distances for the different bunker categories
Bunker category Safety distance [m] (based on LFL)
1 LNG bunkering with a small bunker vessel 61
2 LNG bunkering with a large bunker vessel 218*
3 LNG bunkering with a tank truck 49
4 LNG bunkering from a bunker pontoon 45
5 LNG STS transfer 235*
*the calculated safety distance for category 2 is based on the simultaneous disconnection of three hoses. For category 5 the safety distance is based on the simultaneous disconnection of two hoses.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 22
5.1.1 Discussion
This section will compare the calculated safety distances for passing ships with the current
nautical safety distances that are applicable in the Port of Rotterdam area. Hereafter the
sensitivity of the closure time of breakaway coupling is discussed. The discussion section will
be closed with a discussion about the applicability of the calculated safety distances for
“very” low nautical risk areas. The term “very” low nautical risk areas must be seen in the
context of the traffic density in the Port of Rotterdam. It may be possible that “very” low
nautical risk areas in the Port of Rotterdam are seen as normal traffic densities in smaller
ports.
Current nautical safety distances
Currently, the Rotterdam Port Management Bye-Laws (version: June 2011) [11] state that
open ignition sources, like flames or areas where the temperature equal to or higher than the
minimum ignition temperature of the substance in the cargo tank of the ship, are prohibited
within a distance of 25 metres of the ship, with some exception cases. However, it is also
suggested that this distance may have to be extended for ship of a specialized nature such as
gas tankers. Furthermore, article 4.8 of the Port Bye-Laws states that activities related to the
operation of the ship or objects on the ship have to be performed at least 25 metres away from
dangerous substances or combustible material.
Existing shipping regulations BPR [12] enforce a minimum passing distance of 50 metres
between ships carrying specific explosive substances and other ships, unless ships are passing
each other in opposite directions.
A Swedish bunkering procedure [8] states that bunkering areas on both ships (bunkering
vessel and receiving vessel) should be EX-classified and restricted area during bunkering. The
size of the EX-zone shall be according to class rules for gas-dangerous space and 10 m
horizontally on each side of the receiving ship bunker station plus the whole shipside
vertically.
The calculated safety distance for categories 1, 3 and 4 are in line with the nautical safety
distances that are prescribed in the existing Dutch shipping regulation BPR and the Belgian
shipping regulations. However, the safety distances are roughly a factor two higher than the
distances in the Rotterdam Port Management Bye-Laws. The calculated safety distances for
the categories 2 and 5 are significantly larger than the nautical safety distances that are
prescribed in the existing Dutch inland shipping regulation and the distances in the Rotterdam
Port Management Bye-Laws. The relatively large safety distances that are found for those
categories restrict the LNG bunkering of large vessels in area with intensive nautical traffic. It
may not give restriction on LNG bunkering activities in area where lower nautical activities
take place and collision scenario are less credible. The sensitivity of the nautical activities is
investigated in the following section.
Sensitivity to Nautical activities
For “very” low nautical risk area the collision of passing ships into the bunker operation may
not be a credible scenario. In this case it would be better to selected the leakage at the flange
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 23
face with a hole diameter between 5 – 10 mm. The driving force of this scenario is the
pressure during the bunker operation. In section 4.2.2 a pressure of 5 bar(g) is considered
during the bunkering activity. For area where the nautical risk is not significant the
determined safety distance for each category is 20 meter. From the assessment of the nautical
activities is found that the safety distances could theoretically be reduced to 20 meter in case
collision scenarios are not significant. However the safety distances cannot be less than the
safety distances that is prescribed in the Dutch shipping regulations.
Sensitivity of breakaway coupling closure time
The closure of the outflow area of the breakaway coupling will be mechanically driven and
will be accomplished in less than a second. However for the determination of the safety
distances the closure time is conservative estimated on 5 seconds. The effect of the closure
time of the breakaway coupling on the safety distance is investigated for the category 1 and 2.
The results of this investigation are shown in Figure 7, where the left figure represents
category 1 and the right figure category 2. The results of the sensitivity analysis show that the
closure time of the breakaway coupling do not have a major effect within the investigated
time range. The conservative time of 5 seconds to close to breakaway coupling does not have
a significant influence on the safety distance, which means that the determined safety distance
cannot be significantly reduced by quicker closure of the breakaway coupling.
Figure 7: Safety distances for different closure time of the breakaway coupling; category
1 (left), category 2 (right)
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 24
5.2 Risk distances to vulnerable objects
The distances to the 10-6
/year individual risk (IR) contours for all bunkering scenarios defined
in paragraph 4.2.1 are visually displayed in the figures in this section. The distance to the 10-
6/year individual risk (IR) contours for the three levels of bunker activities per year (upper
bound, mean and lower bound) are visualised with a dot. For practical reasons the dots are
interconnected with a fitted line. The line between the calculated results does not give the
exact distance to the 10-6
/year risk contour and should be used with care. More accurate
results could be obtained in case more bunker activities per category are calculated. However
this is not part of the scope of this project.
It is important to mention that the results are based on the maximum and minimum parameter
set as discussed earlier in section 2.4. The maximum parameter set is a summation of
conservative assumptions (e.g. maximum flow, maximum diameter, maximum bunker time,
maximum number of hoses and the most conservative ignition methodology). This means that
the maximum parameter set gives an upper bound of expected risk level. The minimum
parameter set is based on more average parameters and does not necessary represent the
minimum / lower bound risk level. A detailed overview of all maximum and minimum
parameters per bunker category is found in appendix I.
5.2.1 Category 1 - LNG bunkering with a small bunker vessel
The distances to the 10-6
risk level for the different simulation scenarios from category 1 are
given in Figure 8.
From Figure 8 can be observed that for the maximum parameter set the distance to the 10-6
/year risk level differs for the intense and the less dense nautical traffic areas. The intense
nautical traffic area does results in a higher distance to the 10-6
/year risk level. In the lower
range of the nautical activities, there is no significant difference in distance. This means that
the differences in distance to the 10-6
/year risk level for the low and very low nautical traffic
area negligible. For the same figure could be seen that for minimum parameter set there is a
difference in distance to the 10-6
/year risk level for each of the different nautical traffic area.
In this parameter set a reduction of the nautical activities decreases the distances to the 10-
6/year risk level with 50 -60 %.
The individual risk contours for maximum parameters are shown in Figure 9 where the orange
line represents the 10-6
/year risk contour. The purple and light blue lines in Figure 9 represent
the 10-5
/year and 10-4
/year risk contours. It is seen that there is only a small difference in
distance between the 10-6
/year and 10-5
/year risk contour. The individual risk contours for
minimal parameters are shown in Figure 10, where the same colours represent the different
risk levels.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 25
Figure 8: LNG bunkering toolkit for category 1
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 26
Figure 9: IR contour for 5 bunker activities per day in cat 1, maximal parameters in intensive nautical risk traffic areas (left), low
nautical traffic areas (middle) and very low nautical traffic areas (left). The orange line represents the 10-6
/year risk contour.
Figure 10: IR contour for 5 bunker activities per day in cat 1, minimal parameters in intensive nautical risk traffic (left), low nautical
traffic (middle) and very low nautical traffic (left). The orange line represents the 10-6/year risk contour.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 27
5.2.2 Category 2 - LNG bunkering with a large bunker vessel
The distances to the 10-6
/year risk level for the different nautical scenarios from category 2 are
shown in Figure 11. From the figure can be observed that for the maximum parameter set the
10-6
risk level do not significantly changes over the nautical scenarios. From this observation
it can be concluded that the nautical activities does not have a significantly contribution the
individual risk level of 10-6
/year. However for the minimum parameter set, changes in
distance could be observed. In this parameter set a reduction of the nautical activities
decreases the distances to the 10-6
/year risk level with 37%.
In the intense nautical traffic scenario with minimum parameters the individual risk level of
10-6
/year is dominated by the collision scenario where a 250 mm hole is formed in hull of the
large bunker vessel. The remaining risk is not related to collision scenarios and caused by full
bore rupture of the bunkering hose. For low nautical traffic areas the collision scenarios are
less contributing to the individual risk level of 10-6
/year, but still have a significant
contribution. For the very low nautical traffic areas the collision scenarios are negligible.
The individual risk contours for maximum parameters are shown in Figure 12 where the
orange line represents the 10-6
/year risk contour. The purple and light blue lines in Figure 12
represent the 10-5
/year and 10-4
/year risk contours. It is seen that there is only a small
difference in distance between the 10-6
/year and 10-5
/year risk contour. The individual risk
contours for minimal parameters are shown in Figure 13, where the same colours represent
the different risk levels.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 28
Figure 11: LNG bunkering toolkit for category 2
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 29
Figure 12: IR contour for 1 bunker activities per 2 days in category 2, maximal parameters in intensive nautical traffic areas (left), low
nautical traffic areas (middle) and very low nautical traffic areas (left). The orange line represents the 10-6
/year risk contour.
Figure 13: IR contour for 1 bunker activities per 2 days in cat 2, minimal parameters in intensive nautical traffic areas (left), low
nautical traffic areas (middle) and very low nautical traffic areas (left). The orange line represents the 10-6/year risk contour.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 30
5.2.3 Category 3 - LNG bunkering with a tank truck
The distances to the 10-6
risk level for the different simulation scenarios from category 3 are
given in Figure 15
The difference in the distances to the 10-6
/year risk contour for intensive, low and very low
nautical traffic areas that could be observed from Figure 15 is negligible. In this category
there is no scenario where a possible LOC from the bunker vessel could occur. Bunkering
operation is performed from a truck. The collision scenarios in this category could only results
in rupture of the bunker hose. The LOC scenarios from the truck are not considered within the
scope of this study.
For most of the scenarios the 10-6
/year risk level is mainly caused by a combination of rupture
and leakages through a 25 mm hole in the bunker hose. The risk caused by the rupture of the
bunker hose during collision is negligible to the risk caused by rupture and leakage of the
bunker hose during bunkering. With other words, the nautical activities in the surrounding of
the bunker location do not have a significant influence on the 10-6
risk contour.
Figure 14 shows the individual risk contours for maximum and minimum parameters at an
intensive nautical traffic area.
Figure 14: : IR contour for 1826 bunker activities in cat 3, intensive nautical traffic area
, maximal parameters in (left), minimum parameters intensive nautical traffic area
(right)
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 31
Figure 15: LNG bunkering toolkit for category 3
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 32
5.2.4 Category 4 - LNG bunkering from a bunker pontoon
The distances to the 10-6
risk level for the different simulation scenarios from category 4 are
given in Figure 16.
The risk drivers of category 4 do show a huge similarity to category 3. For most of the
scenarios the 10-6
/year risk level is mainly caused by a combination of rupture and leakages
through a 25 mm hole in the bunker hose. The LOC scenarios from equipment on the bunker
pontoon are not considered within the scope of this study. It is expected that a separate risk
assessment is performed for the permit application of the bunker pontoon and its equipment.
Therefore collision scenarios that resulted in LOCs of the storage tank of the bunker pontoon
are not considered.
The risk caused by the rupture of the bunker hose during collision is negligible to the risk
caused by rupture and leakage of the bunker hose during bunkering. This results in a limited
difference between the distances to the 10-6
/year risk contour for the different nautical traffic
areas. With other words, the nautical activities in the surrounding of the bunker location do
not have a significant influence on the 10-6
risk contour.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 33
Figure 16: LNG bunkering toolkit for category 4
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 34
5.2.5 Category 5 - LNG STS transfer
The distances to the 10-6
risk level for the different simulation scenarios from category 5 are
given in Figure 17.
The risk that is observed for the maximum parameter set seems to be independent of the
nautical traffic area. For the minimum parameter set there is a significant difference between
the three nautical areas. In general it can be concluded that the risk drivers of category 5 do
show a huge similarity to risk driver in category 1 and 2. For most of the scenarios the 10-
6/year risk level is mainly caused by the collision scenario where a 250 mm hole is formed in
hull of the large bunker vessel.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 35
Figure 17: LNG bunkering toolkit for category 5
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 36
5.2.6 Discussion
From the previous sections it could be concluded that for the maximum parameter set the
distance to the 10-6
risk level do not significantly changes over the nautical scenarios. The
bunkering of LNG with small bunker vessel, category 1, is an exception where the intense
nautical traffic area differs from the low and very low nautical traffic areas. Despite the small
difference in distance between the low and very low nautical traffic areas, the 10-6
risk level
of the low nautical risk area is dominated by the collision scenario where a 250 mm hole is
formed in hull of the small bunker vessel.
For the maximum parameter set of category 1 and 2 it is found that the difference in distance
between the 10-6
and 10-5
risk level is small. A detailed analysis of the risk drives reveals that
the 10-5
risk level is dominated by the hose rupture scenario where the ESD system works
probably. This result clarifies why the distances of the low and very low nautical traffic areas
do not significantly differ despite that they do not share the same risk driver.
The maximum parameter set is based on the free field method which means that flammable
clouds ignite when the lower flammable limit is reached. Table 4 shows the effect distances
(Lower Flammable Limit (LFL)) of the different LOC scenarios from category 1. The largest
effect distances are found for the collision scenario where a 250 mm hole is formed in hull of
the small bunker vessel.
Table 4: Effect distances LOC scenarios category 1 (likelihood is not taken into account)
Scenario Effect distance [m]
F 1.5m/s D 5m/s
Full bore rupture ESD works 416 179
Full bore rupture due to collision 226 136
25 mm hole ESD works 93 67
250 mm hole in bunker vessel 595 205
The minimal parameter set is based on specific ignition sources that are present in the
surrounding of the bunkering activity. In most of the cases the cloud is ignited before it
reaches is LFL distance. Figure 18 shows the risk distribution for 5 bunker activities per day
in category 1. The figure shows that for the minimal parameters the 10-6
/year risk level for
intense and low nautical traffic areas is dominated by the collision scenario where a 250 mm
hole is formed in hull of the small bunker vessel. In the very low nautical traffic area the 10-
6/year risk level is dominated by the scenario where the hose is ruptures and the ESD system
works properly. The same trends can be seen for bunker activities in category 2.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 37
Figure 18: Risk distribution of 5 bunker activities per day in cat 1 for three different
risk levels; left, intense nautical traffic; middle low nautical traffic, right; very low
nautical traffic
For the bunker categories 3 and 4 is observed that the distance to the 10-6
risk level do not
significantly changes over the nautical scenarios. For the maximum parameter set of category
3 and 4 it is found that the distance to the 10-6
risk level is mainly caused by the hose rupture
scenarios. An increase of the number of bunker activities shifts the risk driver to scenarios
that are less likely to occur. For instance; for the lower bound activities the 10-6
/year risk level
is mainly caused by the hose rupture scenario where the ESD system works probably, while
for the higher bound activities the 10-6
/year risk level is mainly caused by the hose rupture
scenario where the ESD system fails to work.
The maximum parameter set is based on the free field method which means that flammable
clouds ignite when the lower flammable limit is reached. Table 5 shows the effect distances
(Lower Flammable Limit (LFL)) of the different LOC scenarios from category 3. The largest
effect distances are found in case of rupture of the bunker hose.
Table 5: Effect distances LOC scenarios category 3 (likelihood not taken into account)
Scenario Effect distance [m]
F 1.5m/s D 5m/s
Full bore rupture ESD works 92 63
Full bore rupture due to collision 59 49
25 mm hole ESD works 87 61
The minimal parameter set is based on specific ignition sources that are present in the
surrounding of the bunkering activity. In most of the cases the cloud is ignited before it
reaches is LFL distance. The left side of Figure 19 shows the risk distribution for 1 bunker
activities per week in category 3. The right side shows the risk distribution for 10 bunker
activities per day.
The figure shows that for the minimal parameters and 1 bunker activity per week the 10-6
/year
risk level is dominated by leakage through a 25 mm hole. For the lower risk levels, (10-8
/year)
the dominant scenarios are shifted to the hose rupture scenarios. The figure also shows that
for an increase in bunker activities (10 bunker activities per day) the dominant scenarios that
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 38
contribute to the 10-6
/year risk level are shifted to the hose rupture scenarios. The same trends
can be seen for bunker activities in category 4.
Figure 19: Risk distribution in cat 3 for two different bounds; left, lower bound (1
bunkeractivity a week); right upper bound (10 bunkeractivities a day)
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 39
RISK MITIGATION AND FURTHER RESEARCH
This chapter gives an overview of risk mitigation measures that can improve the safety
performance of the LNG bunker operations. Two set of measures are given:
General mitigation measures that will ensure safe bunker operations,
More study specific mitigation measures and topics for further research will be given to
ensure safe LNG bunkering in the future.
5.3 General risk mitigation measures
Risk levels could be reduced by two different sets of mitigation measures; mitigated measures
that reduce the consequence of the loss of containment and mitigated measures that reduce the
likelihood that loss of containment can occur.
Mitigated measures should be focused on the prevention of loss of containment (LOC).
Prevention of LOC scenario is often accomplished by technical and procedural measures. For
bunkering of LNG the technical measures that could reduce the risk are:
The technical integrity of the bunkering hose is secured by inspection. It is recommended
that the bunker hose is visually inspected before each bunkering operation;
Purging of bunker hose with for instance nitrogen before bunkering operations. Purging of
transfer hose in common practice in industry for transferring large amounts of LPG.
Leakages and coupling errors could be noticed when purging operations are performed;
Specify rules sets for the distance between the hull of the bunker vessel and the LNG
cargo tank for inland vessels. From the risk analysis of bunkering with inland vessels is
concluded that the LOC of the cargo tank caused by collision is the main contributor the
10-6
risk level.
Additional collision protection of hull on LNG bunker vessels
More procedural measures that could reduce the risk are:
Training of bunker operators. The safety aspects of bunkering of LNG can be compared
with the bunkering of convention diesel. Bunkering personnel must be aware of the risks
associated with LNG operations;
Another procedural measure to prevent LOC scenarios is the bunker procedure which
should be followed during the preparation and bunkering operation itself. At this moment
most of the ports do not have a bunker procedure for bunkering of LNG. Before bunkering
activities can be realised a bunkering procedure must be in place;
Nevertheless, loss of containment could occur. If LOC occurs the consequences of the release
must be kept as minimal as possible. Technical measures are often in place to minimise the
release volume or/and consequence of the release:
The release volume can be minimised by a proper ESD system that quickly detects the
leak and shutdown the pumps and valves to prevent further outflow. The detection time of
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 40
leakage is often depending on the amount of gas detectors and or/sensitivity of the excess
flow valve. For LNG bunkering operations it could be recommended to install a ESD1
and ESD2 system where:
- ESD 1 stop transfer pumps and compressors, hence providing a quick and safe
means of stopping the transfer and isolating bunker vessel and recipient vessel
systems
- ESD 2 gives an additional level of protection by providing for a rapid
disconnection of the transfer hose / loading arms from the ship. ESD 2 could, for
example, be used if there is a fire on one of the ships (bunker vessel or recipient
vessel).
The domino effect to the bunker vessel and receiving ship can be minimised by equipping
with a water spray or curtain. The water curtain sprays the affected area with water to
prevent the deck steel from cracking. Another advantage is that the water curtain will
dilute the released LNG cloud and lower the concentration of gas in the air. This measure
reduces the distances at which the cloud is flammable.
5.4 Study specific
The risk drivers for the different bunker categories are identified in chapter 5. Roughly two
different risk drivers are identified:
Category 1, 2 and 5; ship collision results in a loss of containment of the LNG cargo
(bunker) tank.
Category 3 and 4; rupture of the bunkering hose and leakage through a 25 mm hole
For each of the risk drivers suggestions for mitigating measures or suggestions for further
research will be given below:
Nautical risk
The individual risk level of 10-6
/year is for the category 1 and 2 activities mainly caused by
the scenario where ship collision results in a loss of containment of the LNG cargo (bunker)
tank. This conclusion is application for both intensive and low nautical traffic areas.
The nautical risk calculations are based on a model that predicts the collision frequency. The
predicted collision frequency is used as a starting point for the assessment of loss of
containment frequency. The model predicts for the intensive nautical traffic area a collision
frequency around 1.4 x 10-2
per year. For the low nautical traffic area the collision frequency
is estimated at 3.8 x 10-3
per year. For a more detailed representation of the nautical risk these
collision frequencies must be compared with the actual collision frequency in the Port of
Rotterdam area.
The calculated loss of containment frequency from LNG cargo tanks is strongly influenced by
the layout of the bunker vessel. The key design parameters are the strength of the hull and the
distance between the cargo tank and the hull of the bunker vessel. At this moment no inland
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 41
bunker vessels do exist and estimation of the detailed technical layout of the vessel is
difficult. For estimation of a more accurate loss of containment frequency, detailed
investigation between the impact energy and the hole formation in cargo tanks of inland
bunker vessels is necessary. The conclusion of this detailed investigation could lower the
nautical risk significantly.
Another option is to prescribe a minimum distance between the LNG cargo tank and the hull
of the bunker vessel. For seagoing vessel this prescription is already in place: a minimum
distance between the cargo tank and the hull is prescribed in the class rules of the
classification societies. As most of the bunker vessels will be inland vessels class rules do not
apply.
A fourth option could be to lower the nautical risk by procedural measures. Although speed
limitation of passing vessels and other procedural measures are not practical, they do lower
the nautical risk significantly.
Bunkering risk
The individual risk level of 10-6
/year for the category 3 and 4 activities is mainly caused by
rupture of the bunkering hose and leakage through a 25 mm hole. Risk reduction measures
must be applied to the bunker activity itself since collision risk is negligible. The measures
should be focused on limiting the amount of outflow or limiting the frequency at which the
loss of containment can occur.
The frequency at which rupture of the bunkering hose by external impact occur can for
example be lowered by mitigation measures such as the application of a safety net above the
bunker operation to mitigate the risk of falling objects (e.g. container couplings) or by
limiting the number of maintenance and/or lifting activities in the surrounding of the bunker
activity.
The consequences of a release could be lowered by quick closure of valves and pumps in case
of a loss of containment. Quick detection of the loss of containment is a key parameter in
reduction of the possible consequences. For most of the categories in this study is assumed
that an EMS/ESD system does have a response time of 20. For category 3, LNG bunkering with a
LNG tank truck, it is assumed that only intervention measures of an operator are possible. In this
case a response time of 120 seconds is taken into account. This response time could be
significantly lowered if an automatic system is in place.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 42
6 CONCLUSIONS
The Port of Rotterdam has identified various LNG bunker activities in a port area. Those
activities can be grouped in five different categories:
1) LNG bunkering from bunker barges to small vessels
2) LNG bunkering from small scale LNG carriers to seagoing vessels
3) LNG bunkering from trucks to small vessels
4) LNG bunkering from bunker pontoons to small vessels
5) LNG transfer from ship to ship
DNV has calculated the indicative safety distances for the determination of exclusion zones
related to passing vessels during LNG bunkering activities. It is found that the calculated
safety distance for categories 1, 3 and 4 are line with the nautical safety distances that are
prescribed in the existing Dutch shipping regulation BPR. The calculated safety distances for
the categories 2 and 5 are significantly larger than the nautical safety distances that are
prescribed. The relatively large safety distances that are found for those categories could
enforce additional rules to the LNG bunkering of large vessels in area with intensive nautical
traffic. It may not give restriction on LNG bunkering activities in area where lower nautical
activities take place and collision scenario are less credible.
DNV calculates the risk distances to vulnerable objects for the Dutch risk criteria of 10-6
/year.
The risk distances that are found for the different categories vary from 10 to 510 meter,
depending on the category, bunker parameters, ignition method and number of bunker
activities. The risk distance for the categories 1, 2 and 5 is mainly caused by the scenario
where ship collision results in a loss of containment of the LNG cargo (bunker) tank. It is
found that lowering of the nautical activities in the bunkering area may reduce the risk
distance. For areas were (very) low nautical activities take place the risk distance is driven by
the rupture of the bunkering hose and leakage through a 25 mm hole. It is found that the risk
distance of category 3 and 4 is independent of the nautical activities because the distance is
driven by the rupture of the bunkering hose and leakage through a 25 mm hole.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 43
7 REFERENCES
[1] DNV LNG Blog; http://blogs.dnv.com/lng/2011/10/lng-bunkering-operation-
caught-on-tape/
[2] European Argeement concerning the International Carriage of Dangerous Goods by
Inland Waterways (ADN), United Nations Economic Commission for Europe
(UNECE) and the Central Commission for the Navigation of the Rhine (CCNR),
Geneva ,2008
[3] Accord européen relatif au transport international de marchandises Dangereuses par
Route (ADR), United Nations Economic Commission for Europe (UNECE),
Geneva, 1968
[4] Pioneer Knutsen, website Vuyk Engineering Rotterdam B.V.
[5] Reference Manual Bevi Risk Assessments version 3.2, National Institute of Public
Health and the Environment (RIVM), Bilthoven, 2009.
[6] ARF document T14 rev 1; Process equipment failure frequencies for transfer
equipment , Det Norske Veritas, Høvik, 1999.
[7] EN1472-2: Installation and equipment for liquefied natural gas – Design and testing
of marine transfer systems – Part 2: Design and testing of transfer hoses, European
Committee for standardization (CEN), Brussels, 2008
[8] LNG ship to ship bunkering procedure, Swedish Marine Technology forum | Linde
Cryo AB | FKAB Marine Design | Det Norske Veritas AS | LNG GOT | White
Smoke AB , Sweden,
[9] DNV process failure frequencies, standardized offshore leak frequencies, technical
note 14, D.23 loading arms, rev0, Det Norske Veritas, Høvik, 2011
[10] ACDS; Major Hazard Aspects of the Transport of Dangerous Substances, Advisory
Committee on Dangerous Substances, Health & Safety Commission, HMSO Major
hazard aspects of the transport of dangerous substances, 1991.
[11] Rotterdam Port Management Bye-Laws (version juni 2011), ‘Part A – Bulk Liquids
– General – Physical checks-ups, section 37: naked light regulations are observed,
Port of Rotterdam, Rotterdam, 2011
[12] Inland shipping police regulations (‘Binnenvaartpolitieregelement’), article 6.18,
subsection 2, Ministry of infrastructure and environment, the Hague, 2012.
[13] Addendum; Input assumptions for risk calculations bunkering study Port of
Rotterdam, Det Norske Veritas, Rotterdam, 1999.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 44
APPENDIX I SCENARIOS
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 45
A high-level overview of all calculated scenarios is given in Figure 5, section 2.4. This
appendix provides more detailed specifications on the applied LNG transfer parameters used
for each scenario. Some transfer parameters vary per bunkering category and are divided into
two extremes: minimal and maximal. Only the transfer parameters that vary per scenario are
given in Table 6 (e.g. hose diameter, pump rate, number of hoses used per transfer, bunkering
duration per transfer). For a complete overview of all used transfer parameters a reference is
made to the addendum [13].
Table 6: Detailed bunkering scenario definition – transfer parameters
Scena
rio
Nautical
activities
Transfer
parameters
Ignition
method
Hose
diameter
(inch)
Single
pump
rate
(m3/hou
r)
Number
of hoses
used
Duration
per
transfer
(hours)
Category 1 – LNG bunkering with a small inland LNG bunker vessel
1.1 Intensive Maximal Free field 5 500 1 2
1.2 Intensive Maximal Free field 5 500 1 2
1.3 Intensive Maximal Free field 5 500 1 2
1.4 Intensive Minimal Ignition
sources
3 80 1 1
1.5 Intensive Minimal Ignition
sources
3 80 1 1
1.6 Intensive Minimal Ignition
sources
3 80 1 1
1.7 Low Maximal Free field 5 500 1 2
1.8 Low Maximal Free field 5 500 1 2
1.9 Low Maximal Free field 5 500 1 2
1.10 Low Minimal Ignition
sources
3 80 1 1
1.11 Low Minimal Ignition
sources
3 80 1 1
1.12 Low Minimal Ignition
sources
3 80 1 1
1.13 Very Low Maximal Free field 5 500 1 2
1.14 Very Low Maximal Free field 5 500 1 2
1.15 Very Low Maximal Free field 5 500 1 2
1.16 Very Low Minimal Ignition
sources
3 80 1 1
1.17 Very Low Minimal Ignition
sources
3 80 1 1
1.18 Very Low Minimal Ignition
sources
3 80 1 1
Category 2 – LNG bunkering with a large LNG bunker vessel
2.1 Intensive Maximal Free field 7 1000 3 7
2.2 Intensive Maximal Free field 7 1000 3 7
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 46
Scena
rio
Nautical
activities
Transfer
parameters
Ignition
method
Hose
diameter
(inch)
Single
pump
rate
(m3/hou
r)
Number
of hoses
used
Duration
per
transfer
(hours)
2.3 Intensive Maximal Free field 7 1000 3 7
2.4 Intensive Minimal Ignition
sources 5
500 3 6.7
2.5 Intensive Minimal Ignition
sources 5
500 3 6.7
2.6 Intensive Minimal Ignition
sources 5
500 3 6.7
2.7 Low Maximal Free field 7 1000 3 7
2.8 Low Maximal Free field 7 1000 3 7
2.9 Low Maximal Free field 7 1000 3 7
2.10 Low Minimal Ignition
sources 5
500 3 6.7
2.11 Low Minimal Ignition
sources 5
500 3 6.7
2.12 Low Minimal Ignition
sources 5
500 3 6.7
2.13 Very Low Maximal Free field 7 1000 3 7
2.14 Very Low Maximal Free field 7 1000 3 7
2.15 Very Low Maximal Free field 7 1000 3 7
2.16 Very Low Minimal Ignition
sources 5
500 3 6.7
2.17 Very Low Minimal Ignition
sources 5
500 3 6.7
2.18 Very Low Minimal Ignition
sources 5
500 3 6.7
Category 3 – LNG bunkering with a LNG tank truck
3.1 Intensive Maximal Free field 3 80 1 2.5
3.2 Intensive Maximal Free field 3 80 1 2.5
3.3 Intensive Maximal Free field 3 80 1 2.5
3.4 Intensive Minimal
Ignition
sources
3 40 1 1
3.5 Intensive Minimal
Ignition
sources
3 40 1 1
3.6 Intensive Minimal
Ignition
sources
3 40 1 1
3.7 Low Maximal Free field 3 80 1 2.5
3.8 Low Maximal Free field 3 80 1 2.5
3.9 Low Maximal Free field 3 80 1 2.5
3.10 Low Minimal
Ignition
sources
3 40 1 1
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 47
Scena
rio
Nautical
activities
Transfer
parameters
Ignition
method
Hose
diameter
(inch)
Single
pump
rate
(m3/hou
r)
Number
of hoses
used
Duration
per
transfer
(hours)
3.11 Low Minimal
Ignition
sources
3 40 1 1
3.12 Low Minimal
Ignition
sources
3 40 1 1
3.13 Very Low Maximal Free field 3 80 1 2.5
3.14 Very Low Maximal Free field 3 80 1 2.5
3.15 Very Low Maximal Free field 3 80 1 2.5
3.16 Very Low Minimal
Ignition
sources
3 40 1 1
3.17 Very Low Minimal
Ignition
sources
3 40 1 1
3.18 Very Low Minimal
Ignition
sources
3 40 1 1
Category 4 – LNG bunkering from a bunkerpontoon
4.1 Intensive Maximal Free field 3 80 1 1
4.2 Intensive Maximal Free field 3 80 1 1
4.3 Intensive Maximal Free field 3 80 1 1
4.4 Intensive Minimal Ignition
sources
3 30 1 1
4.5 Intensive Minimal Ignition
sources
3 30 1 1
4.6 Intensive Minimal Ignition
sources
3 30 1 1
4.7 Low Maximal Free field 3 80 1 1
4.8 Low Maximal Free field 3 80 1 1
4.9 Low Maximal Free field 3 80 1 1
4.10 Low Minimal Ignition
sources
3 30 1 1
4.11 Low Minimal Ignition
sources
3 30 1 1
4.12 Low Minimal Ignition
sources
3 30 1 1
4.13 Very Low Maximal Free field 3 80 1 1
4.14 Very Low Maximal Free field 3 80 1 1
4.15 Very Low Maximal Free field 3 80 1 1
4.16 Very Low
Minimal Ignition
sources
3 30 1 1
4.17 Very Low
Minimal Ignition
sources
3 30 1 1
4.18 Very Low Minimal Ignition 3 30 1 1
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 48
Scena
rio
Nautical
activities
Transfer
parameters
Ignition
method
Hose
diameter
(inch)
Single
pump
rate
(m3/hou
r)
Number
of hoses
used
Duration
per
transfer
(hours)
sources
Category 5 – Ship to ship LNG transfer
5.1 Intensive Maximal Free field 8 1500 2 7
5.2 Intensive Maximal Free field 8 1500 2 7
5.3 Intensive Maximal Free field 8 1500 2 7
5.4 Intensive Minimal Ignition
sources
5 1500 1 2.7
5.5 Intensive Minimal Ignition
sources
5 1500 1 2.7
5.6 Intensive Minimal Ignition
sources
5 1500 1 2.7
5.7 Low Maximal Free field 8 1500 2 7
5.8 Low Maximal Free field 8 1500 2 7
5.9 Low Maximal Free field 8 1500 2 7
5.10 Low Minimal Ignition
sources
5 1500 1 2.7
5.11 Low Minimal Ignition
sources
5 1500 1 2.7
5.12 Low Minimal Ignition
sources
5 1500 1 2.7
5.13 Very Low Maximal Free field 8 1500 2 7
5.14 Very Low Maximal Free field 8 1500 2 7
5.15 Very Low Maximal Free field 8 1500 2 7
5.16 Very Low
Minimal Ignition
sources
5 1500 1 2.7
5.17 Very Low
Minimal Ignition
sources
5 1500 1 2.7
5.18 Very Low
Minimal Ignition
sources
5 1500 1 2.7
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 49
APPENDIX II NAUTICAL RISK ASSESMENT OF MOORED LNG BUNKERVESSEL
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 50
NAUTICAL RISK ASSESMENT OF THE MOORED LNG
BUNKERVESSEL
When the small inland or large LNG bunker vessel is moored and bunkering activities are
ongoing, various accidental scenarios can be identified. For the scenario that a passing vessel
collides with the docked LNG bunker vessel a detailed and quantitative assessment is
performed using a DNV Energy Model. Bunkering activities do always take place in port
areas where ship velocities are still relatively small. To enable a LNG bunker infrastructure
LNG bunkering along waterways could be a possibility as well. Along the (inland) waterways
velocities could be high.
To assess the nautical risk an intensive, low and very low nautical traffic area is considered.
All three scenarios are based on estimations of maritime traffic on three representative
scenarios. This appendix will determine the loss of containment frequency of three identified
nautical traffic areas (intensive, low and very low nautical traffic). Figure 20 shows the three
different nautical traffic areas.
Figure 20: Nautical traffic areas
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 51
Methodology
The risk analysis of the moored LNG bunker vessel is performed using DNV`s methodology
for probabilistic hole assessment in moored LNG Carrier cargo tanks due to ship collisions.
The methodology has been based on a DNV research program where the damage extents have
been estimated for different ship sizes as a function of striking angles and bow types. This
DNV energy model approach is developed to analyse only the risk of a moored LNG Carrier,
since this activity is expected to be most relevant with regards to potential external risk
exposure on land.
The methodology is divided into a frequency and a collision part. In the frequency assessment
“average maximum impact energy” is estimated with corresponding frequency. In the
collision assessment experience data from previous studies is used to transform the “average
maximum impact energy” to probability distribution for size impact damages.
This applied probabilistic approach for collision risk assessment is visualized in the figure
below.
Figure 21: Probabilistic approach for collision risk assessment
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 52
Frequency Assessment
Estimation of maximum impact energy
In the frequency assessment the vessel traffic passing a bunkering LNG vessel is dived into
six classes of vessels, given by the properties of the passing vessels. These six classes are:
Bulb bow vessels with length below 120 meters
Raked bow vessels with length below 120 meters
Bulb bow vessels with length between 120 and 180 meters
Raked bow vessels with length between 120 and 180 meters
Bulb bow vessels with length above 180 meters
Raked bow vessels with length above 180 meters
For each of the classes the vessels are grouped according to their type and displacement. The
impact energy is estimated for each group and then summarised to weighted “average impact”
energy for the vessel class, which is used as input in the collision assessment.
Probability of collision
The methodology assumes there are two dominant failure modes leading to collision:
Steering gear failure
Black out
The probabilities for steering gear failure have been found from DNV`s internal database
RiskNet.
From this reference the following basis figures are applied:
Steering gear failure : 8.3 x10-7
per nautical mile
Black-out : 4.8 x10-6
per nautical mile
The probability for one of these failure modes leading to an actual impact with an LNG
Carrier dock at a berth is assumed to be a function of the geometric probability of hitting the
docked carrier and the time available to implement mitigating actions.
Geometric probability of hitting a passing carrier
The geometric probabilities are a function of the length of the potentially stroked LNG
Carrier, the distance to passing shipping lanes and physical obstacles such as breakwaters or
shallows.
Time to implement mitigating actions
It is assumed that the probability of having time to implement mitigating action has a
“Weibull” distribution. This means that the probability for implementing actions is very low
up to a given time, where the probability increases sharply, to a time where very high
probability that preventive action is implemented successfully. The typical actions that are
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 53
implemented to mitigate striking incidents are assessed to be emergency anchoring, running
engine full astern and re-powering of the vessel.
In the methodology it is assumed that the probability of implementing mitigating actions starts
to rise sharply around an average one minute. This is represented by a “Weibull” distributing
with mean value of 60 seconds and a standard deviation of 31 seconds.
The time available to implement mitigating actions are a function of the vessels speed and the
distance from the shipping lanes and the potentially stroked vessel.
Output from the frequency assessment is a probability for striking impact in to the LNG
bunker vessel per vessel category with corresponding “average maximum impact energy”.
Collision Assessment The relationship between the striking angles, the ship speeds and the absorbed deformation
energy of the colliding ships is determined by,
In the above formula the effect of striking location against the LNG vessel, x2, is taken into
account. A mean value of x2 = 0.225*LLNG is assumed.
For each of the selected striking ship sizes of L = 90 m, L = 140 m and L = 230 m the
absorbed deformation energy has been calculated for a series of impact cases where the
apparent striking angles and the ship speeds have been varied.
The speed distribution of impacting vessels is adjusted relatively to the impact speed, based
on the speed distribution found from the HARDER study, for impact at the time of collision.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 54
Hence, the energy is distributed, with the “average max impact energy” from the frequency
assessment as the maximum impact energy and the downwards. Further, effect of the impact
angle is included, where impact angels of 0 to 22,5 degrees and 167 to 180 degrees are
assumed only to give glancing impacts, with no potential for cargo containment penetration.
The remaining impact angels are grouped and represented in the structural model with impact
angels of 45 degrees and 90 degrees.
Results An estimation of vessel movements in the intensive, low and very low nautical traffic area
was provided by the Port of Rotterdam. The DNV energy model however also needs
additional input which is summed in the nautical part of appendix I.
The collision frequency per category is calculated in the frequency assessment. For the
intensive nautical traffic area it is found that the total collision frequency is around 1.4 x 10-2
per year. The total collision frequency for the low and very low nautical risk area is around
3.8 x 10-3
per year and 6.8 x 10-5
per year. Details about the segmentation of collision
frequencies per category are given in Table 7 till Table 9.
Table 7: Collision frequency of different types of vessels at intensive nautical traffic location
Vessel type Collision
frequency
[/year]
Average
maximum impact
energy [MJ]
Bulb bow vessels with length below 120 meters 8.0 x 10-3
59
Raked bow vessels with length below 120 meters 5.3 x 10-3
59
Bulb bow vessels with length between 120 - 180
meters
7.2 x 10-3
567
Raked bow vessels with length between 120 - 180
meters
4.8 x 10-3
567
Bulb bow vessels with length above 180 meters 1
1
Raked bow vessels with length above 180 meters 1
1
Table 8: Collision frequency of different types of vessels at low nautical traffic location
Vessel type Collision
frequency
[/year]
Average
maximum impact
energy [MJ]
Bulb bow vessels with length below 120 meters 2.3 x 10-3
41
Raked bow vessels with length below 120 meters 1.5 x 10-3
41
Bulb bow vessels with length between 120 - 180
meters
6.9 x 10-6
91
Raked bow vessels with length between 120 - 180
meters
4.6 x 10-6
91
Bulb bow vessels with length above 180 meters 3.4 x 10-6
227
Raked bow vessels with length above 180 meters 2.3 x 10-6
227
1 On this representative waterway no vessel longer than 180 meters are recorded
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 55
Table 9: Collision frequency of different types of vessels at very low nautical traffic location
Vessel type Collision
frequency
[/year]
Average
maximum impact
energy [MJ]
Bulb bow vessels with length below 120 meters 4.0 x 10-5
23
Raked bow vessels with length below 120 meters 2.7 x 10-5
23
Bulb bow vessels with length between 120 - 180
meters
6.9 x 10-7
91
Raked bow vessels with length between 120 - 180
meters
4.6 x 10-7
91
Bulb bow vessels with length above 180 meters 5.2 x 10-7
227
Raked bow vessels with length above 180 meters 3.4 x 10-7
227
It must be noted that the collision frequencies are highly theoretically and may not represent
the actual collision frequency in the Port of Rotterdam area. The collision frequency also
represents all theoretical collision and not only the collisions which will results in a loss of
Containment of LNG. The probability of Loss of containment is made in the collision
assessment.
The frequency assessment is followed up by the collision assessment where the probabilities
of certain indentation depths into a moored LNG Bunker vessel are calculated. Graphical
representations of indentation depths for the three nautical traffic areas are given in Figure 22.
More details regarding the indentation depths for the nautical traffic areas are found in Table
10 till Table 12.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 56
Figure 22: Probability of indentation for intensive,
low and very low nautical traffic areas
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 57
Table 10: Probability of indentation for intensive nautical traffic areas
Representative
length [m]
Bow
type
Type
ID
Indentation depth [m]
0.5-1.0 1.0-1.5 1.5-2.0 2.0-3.0 3.0-4.0 4.0-5.0
90 Bulb 1.1 9.2x10-4
2.3x10-4
1.5x10-4
1.5x10-4
3.8x10-5
3.8x10-5
90 Raked 1.2 7.4x10-4
1.1x10-4
5.3x10-5
1.3x10-4
2.7x10-5
0
150 Bulb 2.1 7.2x10-5
0 1.4x10-5
0 1.4x10-5
0
150 Raked 2.2 4.8x10-5
0 9.5x10-6
0 9.5x10-6
4.8x10-6
240 Bulb 3.1 2
2 2 2 2 2
240 Raked 3.2 2
2 2 2 2 2
Total 1.8x10-3
3.4x10-4
2.3x10-4
2.9x10-4
8.9x10-5
4.3x10-5
Table 11: Probability of indentation for low nautical traffic areas
Representative
length [m]
Bow
type
Type
ID
Indentation depth [m]
0.5-1.0 1.0-1.5 1.5-2.0 2.0-3.0 3.0-4.0 4.0-5.0
90 Bulb 1.1 2.7x10-4
9.1x10-5
4.6x10-5
3.4x10-5
1.1x10-5
0
90 Raked 1.2 2.3x10-4
3.0x10-5
3.0x10-5
1.5x10-5
0 0
150 Bulb 2.1 1.4x10-7
6.9x10-8
6.9x10-8
6.9x10-8
1.4x10-7
3.4x10-8
150 Raked 2.2 5.5x10-7
1.4x10-7
4.6x10-8
9.2x10-8
4.6x10-8
4.6x10-8
240 Bulb 3.1 6.9x10-8
6.9x10-8
0 6.9x10-8
0 3.4x10-8
240 Raked 3.2 4.6x10-8
4.6x10-8
2.3x10-8
2.3x10-8
4.6x10-8
4.6x10-8
Total 5.0x10-4
1.2x10-4
7.6x10-5
5.0x10-5
1.2x10-5
1.6x10-7
Table 12: Probability of indentation for very low nautical traffic areas
Representative
length [m]
Bow
type
Type
ID
Indentation depth [m]
0.5-1.0 1.0-1.5 1.5-2.0 2.0-3.0 3.0-4.0 4.0-5.0
90 Bulb 1.1 5.6x10-6
1.6x10-6
6.0x10-7
3.4x10-5
1.1x10-5
0
90 Raked 1.2 4.3x10-6
8.0x10-7
2.7x10-7
1.5x10-5
0 0
150 Bulb 2.1 1.4x10-8
6.9x10-9
6.9x10-9
6.9x10-8
1.4x10-7
3.4x10-8
150 Raked 2.2 5.5x10-8
1.4x10-8
4.6x10-9
9.2x10-8
4.6x10-8
4.6x10-8
240 Bulb 3.1 1.0x10-8
1.0x10-8
0 6.9x10-8
0 3.4x10-8
240 Raked 3.2 6.9x10-9
6.9x10-9
3.4x10-9
2.3x10-8
4.6x10-8
4.6x10-8
Total 9.9x10-6
2.4x10-6
8.8x10-7
2.3x10-7
2.5x10-8
2.0x10-8
For inland LNG barges an indentation depth of 2 meter or more is expected to cause a 250
mm LNG leak. In the case of small scale LNG vessels an indentation depth of 3 meter or
more is expected to cause a 250 mm LNG leak.
2 On this representative waterway no vessel longer than 180 meters are recorded
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 58
Table 13 gives a summary of the LOC frequencies for the leakage of the cargo tank in case of
collision. It must be noticed that all frequencies in Table 13 are based on the assumption that
the bunker vessel and/or receiving vessel are continuously present (100% of the time) at the
bunker location. For the actual LOC frequency per category and parameter set (maximum or
minimum) the frequency of Table 13 should be multiplied with the presence factor of the
bunker vessel and/or receiving vessel.
Table 13: Summary of the LOC frequencies of the bunker cargo tank
Category LOC frequency [/year]
of intensive nautical
traffic area
LOC frequency [/year]
of low nautical traffic
area
LOC frequency
[/year] of very low
nautical traffic area
Cat 1 2.9 x 10-4
5.0 x 10-5
2.3 x 10-7
Cat 2 8.9 x 10-5
1.2 x 10-5
2.5 x 10-8
Cat 5 Max parameters 8.9 x 10-5
Min parameters 2.9 x 10-4
Max parameters 1.2 x 10-5
Min parameters 5.0 x 10-5
Min parameters 2.5 x
10-8
Min parameters 2.3 x
10-7
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 59
NAUTICAL RISK ASSESSMENT OF THE LOADING OPERATION
In the previous chapter the collision of ships to the moored bunker vessel resulted in a loss of
containment of the cargo tanks of the bunker vessel. This chapter will focus on the impact that
the collision can have on the bunker operation. The impact of a colliding ship can results in
rupture of the bunker hose depending on the impact energy of the colliding vessel.
Methodology
The methodology is divided into a frequency and a collision part. The frequency of the ship
collision on the bunker operation is based on the same collision frequency model that is used
for the risk analysis of the moored bunker vessel. The collision frequency in both case is
same, the difference is found in the loss of containment frequency. In the collision assessment
the collision frequency is transformed to a LOC frequency of the bunker operation.
Results
The risk analysis of bunker operation is based on the overall collision frequency and does not
make any distinction between the different shipping classes. The total collision frequency for
the intensive nautical traffic area is around 1.4 x 10-2
per year. For the low and very low
nautical traffic area the total collision frequency is reduced to around 3.8 x 10-3
and 6.8 x 10-5
per year. During the bunker operation the receiving and bunker vessel are connected with
mooring lines to prevent drift away. For all bunkering operations the receiving vessel is
fixed/moored to the shore as well. In case of collision the mooring lines will absorb the first
impact energy. Colliding vessels with low amount of impact energy (e.g. low mass and/or
velocity) will not have sufficient energy to rupture the mooring lines and loss of containment
could not occur. For higher impact energies the mooring lines could fail and the tensile
strength of the bunker hose is the limited factor for loss of containment.
It is common practice in the industry to equip bunker hoses for gas purpose with a breakaway
coupling. For this bunker hose configuration the weak point of the confirmation is shifted to
the breakaway coupling because the tensile strength of the bunker hose will exceed the
strength at which the breakaway coupling will be activated.
In case the colliding vessel contains enough impact energy to release the breakaway coupling
the coupling will close in less than a second because of the mechanical closing system. The
quick closure of the breakaway coupling limits the release volume. The probability that the
colliding vessel contains enough impact energy to release the breakaway coupling is estimate
on 0.1. This expert judgment is based on the average energy on the waterway, strength of the
mooring line and set pressure of the breakaway coupling.
Table 14 gives a summary of the LOC frequencies for the rupture of bunker hoses due to
collision. It must be noticed that all frequencies in Table 14 are based on the assumption that
the bunker vessel and/or receiving vessel are continuously present (100% of the time) at the
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 60
bunker location. For the actual LOC frequency per category and parameter set (maximum or
minimum) the frequency of Table 14 should be multiplied with the presence factor of the
bunker vessel and/or receiving vessel.
Table 14: Summary of the LOC frequencies for the rupture of bunker hoses due to collision
Category LOC frequency
[/year] of intensive
nautical traffic area
LOC frequency
[/year] of low
nautical traffic area
LOC frequency
[/year] of very low
nautical traffic area
Cat 1,2,3,4 and 5 1.5 x 10-3
3.8 x 10-4
6.8 x 10-6
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 61
APPENDIX III BACKGROUND
PARAMETERS FOR RISK
CALCULATION
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 62
MODEL PARAMETERS
Together with the scenarios defined in the chapter 4, model parameters represent the basis
input of the model. For many of the general model parameters default values in Phast Risk 6.7
are taken. However, for some parameters a deviation of the default values was needed. The
most important and influential (with respect on the output) general model parameters used to
calculate the individual risk are described below. Specific model parameters, such as weather
data and ignition sources, are also discussed in this chapter. A reference is made to addendum
[13] for a complete overview of all model parameters, which includes a justification of the
chosen value/method.
General parameters
The following general parameters are discussed below:
General risk/model parameters
Ignition parameters (for immediate and delayed explosions)
Explosion parameters
General risk/model parameters
In the event that all effect mitigating measures fail (e.g. ESD, operator), the maximum release
duration of each loss of containment scenario is limited by 1800s (based on the Dutch
guideline risk calculations). Note that this limit also applies for the failure scenario of a LNG
storage tank on a bunkering vessel caused by external collision where effective mitigation is
not deemed possible.
Fractions between day and night are set to 0.44 for day and 0.56 for night and are specifically
applicable for the Netherlands [5]. This also implies that bunkering activities are equally
distributed over a 24 hour period. The Port of Rotterdam assumes/expects no difference in
amount of bunkering activities between day and night.
Ignition and explosion parameters
Two types of ignition events can be distinguished in case of release of a flammable material:
immediate ignition and delayed ignition. Immediate ignition can lead to flammable effects
such as jet fires and pool fires in the event of a continuous discharge. Delayed ignition could
result in residual pool fires, flash fires and explosions.
Explosions
Related to explosion there are two possibilities:
When a vapor cloud enters an area of congestion, a confined explosion is possible that
could lead to potentially high overpressures depending on the level of confinement.
If there is no confinement, a potential unconfined vapor explosion could occur, which
is essentially a flash fire with accompanying explosion that usually generates relative
low overpressures.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 63
In case of continuous LNG releases during bunkering activities or LNG ship to ship transfers,
it is reasonable to assume that unconfined vapor explosions could occur in the event of
delayed ignition due to the low level of confinement in the immediate area. As such, vapor
explosions are modeled with the Multi-Energy (ME) explosion model (rather than the TNT
equivalent default explosion model, which is unrealistic and over-conservative) setting the
unconfined explosion strength to 1 (lowest ME strength curve).
Immediate ignition
Materials have different probabilities of immediate ignition depending on the flash point and
reactivity. Class 0 materials, materials having a flash point under 0 ºC and a boiling point
under 35 ºC, are divided in two categories: average to high reactivity and low reactivity. LNG
is a class 0 material with low reactivity. The immediate ignition probabilities for LNG that are
used are taken from the Dutch guideline on risk calculations and vary per type of
installation/transport asset and are highly dependent on the initial continuous discharge rate or
amount of material releases in case of instantaneous release. Note, instantaneous release
scenarios (e.g. a BLEVE domino scenario of tank truck) are not considered in this study due
to scope limitations. Immediate ignition probabilities for LNG bunkering/transfer activities
are based on those of applicable for stationary installations. This approach is in agreement
with the Dutch guidelines. Immediate ignition probabilities associated with bunkering tank
failure scenarios resulting from ship collisions are based on the probabilities applicable for
ships in the event of a continuous discharge.
An overview of all used immediate ignition probabilities in the model is given below:
Table 15: Overview of used immediate ignition probabilities for class 0 materials with
low reactivity (e.g. methane) (adopted and modified from the Dutch guideline risk
calculations)
Installation/transport asset Discharge for continuous
releases
Immediate ignition
probability
Stationary
installations/transfers >100 kg/s
0.02
Stationary
installations/transfers 10-100 kg/s
0.04
Stationary
installations/transfers >100 kg/s
0.09
Ship – gas tanker >180 m3
0.7
Delayed ignition
The delayed ignition of gas clouds is modeled using the following two approaches:
Free field method: the cloud ignites when the maximum ground footprint area to the
LFL fraction to finish is reached.
No free field (ignition sources): cloud ignition is dependent on the presence of ignition
sources in the area and their probability of causing an ignition. A specification of all
ignition sources considered in this study is given later in this appendix.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 64
Both the free field and no free field method are used in the risk calculations for the
‘maximum’ and ‘minimum’ bunkering parameters scenarios, respectively (see also Figure
23). The advantage of the free-field method is that the individual risk results are location
independent (i.e. independent of potential ignition sources present at a particular location).
However, this method generally results in higher calculated individual risk and can therefore
be considered as conservative. Note, the free field method of delayed ignition is the default
method in Safeti-NL, the mandatory QRA modeling package for Dutch QRA’s.
Weather data
The consequences of releases of flammable and toxic materials into the atmosphere are
strongly dependent upon the rate at which the released material is diluted and dispersed to
safe concentrations. The rate of dispersion is dependent on the meteorological conditions
prevailing at the time of release, particularly the wind speed and the degree of turbulence in
the atmosphere. The wind direction is also of importance as it determines the direction in
which the cloud of material will travel.
Meteorological parameters (e.g. wind speeds and directions, environmental temperatures,
atmospheric pressure, humidity) are based on Dutch weather statistics measured during a time
period of 29 years by the KNMI. The used wind data are specifically applicable for the
Rotterdam region.
Ignition sources In order to address delayed ignition without using the earlier mentioned free-field method, the
model requires information about the distribution of ignition sources in the vicinity the
bunkering activity. Several different types of ignition source could be present, such as ships,
other transport vehicles and people.
The following two sources of ignition are considered in the model (also graphically visualized
in Figure 23):
Ships passing a bunkering/ LNG ship to ship transfer location where, concurrently, a
specific bunkering activity is taking place.
The LNG receiving vessel, which is not entirely EX-zoned (only the area around the
LNG fuel tank and bunkering point / station) and could therefore be a potential
ignition source.
Ignition sources not considered in the model:
The bunkering ship is assumed to be EX-zoned, meaning that all deck equipment is
explosion-proof. As such, no ignition can occur at the bunkering ship and is not
considered to be a potential ignition source.
Electrical installations, industrial equipment, people or transport vehicles in the
immediate vicinity. The assumption is that an (inland) area in the immediate vicinity
of the activity will be EX-zoned.
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 65
In the event of a ship collision, the colliding ship strikes the bunkering vessel or a
receiving ship. It is common practice in nautical QRA’s not to include the colliding
ship as additional ignition source in case the collision results in loss of containment of
flammable materials. The immediate ignition probability with respect to the latter loss
of containment scenarios on bunkering vessels (i.e. LNG tank failure) is higher in the
event of a ship collision, which accounts to some extend that the presence of a
colliding ship is essentially an additional ignition source.
Figure 23: Overview of ignition sources considered (location Oude Maas)
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 66
APPENDIX IV RESULTS -
DISTANCES TO 10-6
/YEAR
CONTOUR
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 67
Table 16: Distance to 10-6
/year contour for each bunkering scenario
Scenari
o
Transfer
parameters
Delayed
ignition
Transfer
activities
(/year)
Nautical
activity
Distance to 10-6
/year
contour (m)
Category 1 – LNG bunkering with a small inland LNG bunker vessel
1.1 Maximal Free field 52 Intensive 273
1.2 Maximal Free field 1826 Intensive 386
1.3 Maximal Free field 3652 Intensive 510
1.4 Minimal Ignition
sources
52 Intensive 70
1.5 Minimal Ignition
sources
1826 Intensive 174
1.6 Minimal Ignition
sources
3652 Intensive 185
1.7 Maximal Free field 52 Low 215
1.8 Maximal Free field 1826 Low 345
1.9 Maximal Free field 3652 Low 355
1.10 Minimal Ignition
sources
52 Low 28
1.11 Minimal Ignition
sources
1826 Low 146
1.12 Minimal Ignition
sources
3652 Low 171
1.13 Maximal Free field 52 Very low 210
1.14 Maximal Free field 1826 Very low 342
1.15 Maximal Free field 3652 Very low 350
1.16 Minimal Ignition
sources
52 Very low 28
1.17 Minimal Ignition
sources
1826 Very low 88
1.18 Minimal Ignition
sources
3652 Very low 95
Category 2 – LNG bunkering with a large LNG bunker vessel
2.1 Maximal Free field 12 Intensive 210
2.2 Maximal Free field 183 Intensive 345
2.3 Maximal Free field 365 Intensive 358
2.4 Minimal Ignition
sources
12 Intensive 111
2.5 Minimal Ignition
sources
183 Intensive 215
2.6 Minimal Ignition
sources
365 Intensive 237
2.7 Maximal Free field 12 Low 202
2.8 Maximal Free field 183 Low 342
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 68
Scenari
o
Transfer
parameters
Delayed
ignition
Transfer
activities
(/year)
Nautical
activity
Distance to 10-6
/year
contour (m)
2.9 Maximal Free field 365 Low 352
2.10 Minimal Ignition
sources
12 Low 105
2.11 Minimal Ignition
sources
183 Low 170
2.12 Minimal Ignition
sources
365 Low 192
2.13 Maximal Free field 12 Very low 430
2.14 Maximal Free field 183 Very low 480
2.15 Maximal Free field 365 Very low 480
2.16 Minimal Ignition
sources
12 Very low 85
2.17 Minimal Ignition
sources
183 Very low 135
2.18 Minimal Ignition
sources
365 Very low 150
Category 3 – LNG bunkering with a LNG tank truck
3.1 Maximal Free field 52 Intensive 155
3.2 Maximal Free field 1826 Intensive 255
3.3 Maximal Free field 3652 Intensive 302
3.4 Minimal
Ignition
sources
52 Intensive 80
3.5 Minimal
Ignition
sources
1826 Intensive 144
3.6 Minimal
Ignition
sources
3652 Intensive 151
3.7 Maximal Free field 52 Low 155
3.8 Maximal Free field 1826 Low 255
3.9 Maximal Free field 3652 Low 302
3.10 Minimal
Ignition
sources
52 Low 85
3.11 Minimal
Ignition
sources
1826 Low 146
3.12 Minimal
Ignition
sources
3652 Low 152
3.13 Maximal Free field 52 Very low 155
3.14 Maximal Free field 1826 Very low 255
3.15 Maximal Free field 3652 Very low 302
3.16 Minimal
Ignition
sources
52 Very low 85
3.17 Minimal
Ignition
sources
1826 Very low 146
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 69
Scenari
o
Transfer
parameters
Delayed
ignition
Transfer
activities
(/year)
Nautical
activity
Distance to 10-6
/year
contour (m)
3.18 Minimal
Ignition
sources
3652 Very low 152
Category 4 – LNG bunkering from a bunkerpontoon
4.1 Maximal Free field 52 Intensive 116
4.2 Maximal Free field 1826 Intensive 136
4.3 Maximal Free field 3652 Intensive 138
4.4 Minimal Ignition
sources
52 Intensive 8
4.5 Minimal Ignition
sources
1826 Intensive 85
4.6 Minimal Ignition
sources
3652 Intensive 94
4.7 Maximal Free field 52 Low 116
4.8 Maximal Free field 1826 Low 136
4.9 Maximal Free field 3652 Low 138
4.10 Minimal Ignition
sources
52 Low 8
4.11 Minimal Ignition
sources
1826 Low 82
4.12 Minimal Ignition
sources
3652 Low 91
4.13 Maximal Free field 52 Very low 116
4.14 Maximal Free field 1826 Very low 136
4.15 Maximal Free field 3652 Very low 138
4.16 Minimal Ignition
sources
52 Very low 8
4.17 Minimal Ignition
sources
1826 Very low 82
4.18 Minimal Ignition
sources
3652 Very low 91
Category 5 – Ship to ship LNG transfer
5.1 Maximal Free field 12 Intensive 270
5.2 Maximal Free field 84 Intensive 580
5.3 Maximal Free field 156 Intensive 600
5.4 Minimal Ignition
sources
12 Intensive 115
5.5 Minimal Ignition
sources
84 Intensive 250
5.6 Minimal Ignition
sources
156 Intensive 270
5.7 Maximal Free field 12 Low 232
5.8 Maximal Free field 84 Low 560
DET NORSKE VERITAS
MANAGING RISK
Port toolkit safety distances LNG bunkering
DNV Reg. No.: PP035192-R2
Revision No.: 2 Date : 2012-08-28 Page 70
Scenari
o
Transfer
parameters
Delayed
ignition
Transfer
activities
(/year)
Nautical
activity
Distance to 10-6
/year
contour (m)
5.9 Maximal Free field 156 Low 590
5.10 Minimal Ignition
sources
12 Low 68
5.11 Minimal Ignition
sources
84 Low 152
5.12 Minimal Ignition
sources
156 Low 175
5.13 Maximal Free field 12 Low 230
5.14 Maximal Free field 84 Low 560
5.15 Maximal Free field 156 Low 580
5.16 Minimal Ignition
sources
12 Low 68
5.17 Minimal Ignition
sources
84 Low 145
5.18 Minimal Ignition
sources
156 Low 165
Det Norske Veritas:
DNV is a global provider of knowledge for managing risk. Today, safe and responsible business conduct is both a license to operate and a competitive advantage. Our core competence is to identify, assess, and advise on risk management, and so turn risks into rewards for our customers. From our leading position in certification, classification, verification, and training, we develop and apply standards and best practices. This helps our customers to safely and responsibly improve their business performance. Our technology expertise, industry knowledge, and risk management approach, has been used to successfully manage numerous high-profile projects around the world. DNV is an independent organisation with dedicated risk professionals in more than 100 countries. Our purpose is to safeguard life, property and the environment. DNV serves a range of industries, with a special focus on the maritime and energy sectors. Since 1864, DNV has balanced the needs of business and society based on our independence and integrity. Today, we have a global presence with a network of 300 offices in 100 countries, with headquarters in Oslo, Norway.
Global impact for a safe and sustainable future:
Learn more on www.dnv.com