10.1.1 A quantitative risk assessment (QRA) on the proposed Permanent Aviation Fuel facility (PAFF) to be built at Tuen Mun Area 38, together with its operations, has been undertaken and the QRA will form the Hazard Assessment section of the EIA. A Hazard Assessment has been conducted on the proposed facility and its operations in order to meet the requirements of the EIA Study. This section contains the details of the Hazard Assessment, the analysis, its findings and recommendations.
10.2
Objectives and Scope of Work
10.2.1 The main objective of the hazard assessment is to identify the risk to life due to the storage and transport of the aviation fuel at the PAFF and propose measures to mitigate its impact. The study assessed the risk to life, due to marine transport in the vicinity of the jetty, jetty operations, tank farm storage and pipeline transfer of aviation fuel. The study considers two cases, the 2016 case and 2040 case.
10.2.2 The detailed objectives and scope of work were:
¨ identify all hazardous scenarios associated with the marine transport in the vicinity of the jetty (within 500m), jetty operations, tank farm storage and pipeline transfer of aviation fuel, which may cause fatalities;
¨ estimate the frequency of release of aviation fuel from the above operations;
¨ evaluate the consequences of such a release;
¨ quantify the risk to life in both individual and societal terms;
¨ evaluate the calculated risks against the Hazard to Life criteria established in Annex 4 of the EIAO-TM; and
¨ identify practicable cost effective risk mitigation measures should they be required.
10.3.1
Existing Arrangement of
Fuel Supply
10.3.1.1 The existing arrangement of fuel supply to the airport comprises of an aviation fuel receiving facility (AFRF) at Sha Chau. Aviation fuel is supplied to the terminal primarily in purpose built 5000 dwt tankers from Tsing Yi. Aviation fuel is then pumped through two submarine pipelines to the tank farm on the Hong Kong International Airport (HKIA).
10.3.2
Proposed Facility at Tuen
Mun Area 38 (TMA38)
10.3.2.1 The proposed PAFF at TMA38 will consist of the following facilities:
¨ a jetty with two berths, one able to accommodate 10,000 to 80,000 dwt vessels; and the other able to accommodate 10,000 to 40,000 dwt;
¨ a gross aviation fuel tankage capacity of approximately 140,000 m3 in 2005 increasing in stages to match the anticipated growth in aviation fuel uplift to an ultimate gross tankage capacity of approximately 400,000 m3
¨ the diameter of the tank will be 40m while the height of tanks vary from 23m to 32m. The volume of the largest tank will be 39,000 m3 and this is considered for the hazard assessment;
¨ pumps and associated facilities; and
¨ pipelines to transfer the fuel to the aviation fuel system on the airport.
10.3.2.2 Details of these elements of the facility are provided in Section 3 of this EIA Report. Marine transport in the vicinity of the jetty (ie within 500m) is included in this assessment. This mainly consists of vessels bringing the aviation fuel to the tank farm, vessels travelling in Urmston road channel within 500m of the jetty and vessels arriving at jetties in the vicinity of the PAFF jetty. The assumed frequency of tanker visits to the PAFF jetty per year is summarised in Table 3.1.
10.3.2.3 The Emergency Shutdown (ESD) Control Philosophy at the facility provides for the shutdown of the following:
¨ receipt of fuel from the jetty;
¨ tank farm facility; and
¨ delivery lines.
10.3.2.4 There are two ESD valves on the inlet to the tank farm (from the jetty) and two on the outlet of the tank farm (to the airport). These valves are operated via Rotork motorized actuators and are fed via the UPS (Uninterrupted Power Supply) so that they can operate during an interruption to the facility’s main power supply.
10.3.2.5 Each of the above systems has different means of initiating the system. Manual push-buttons provide the primary mode of initiation, however, other initiating devices such as the actuation of the fire alarm system, fuel tank high high level and a sudden drop in pressure in the delivery pipelines can also activate the ESD system. A leak detection system is provided for the delivery pipeline.
10.3.2.6 The fire fighting facility at the PAFF includes dedicated sea water pumps to provide fire water for tank cooling, foam injection and fire hydrants. Four pumps of total capacity of about 45,000 litres/min are provided.
10.3.2.7 The storage tanks are provided with water spray to cool the tank shell. The fire water system is designed to provide for cooling of the tank on fire as well as cooling of adjoining tanks.
10.3.2.8 In addition, foam injection facilities are provided for injecting foam in to the base of the tank (for the event of tank on fire) as well as for foam monitors.
10.3.2.9 A drencher system and a foam monitor system are also provided at the jetty.
10.3.2.10 An automatic fire alarm system is provided for the tank farm as well as the jetty.
10.3.2.11 External fire fighting resources including fireboats will be provided by the Fire Services. The Pillar Point Fire Station which has both fire trucks and foam trucks is the nearest to the PAFF facility and could be reached within the graded response time for the area of 6 minutes under normal traffic conditions.
10.3.2.12 The storage tanks will be located within a bund, which is designed to contain any spills from the tank or tank piping. The bund is designed to hold 110% of the contents of the largest tank in the bund. The bund will be provided with a drain, which will be discharged by a manually operated valve to the sea through an interceptor. Drainage from other areas onsite will be discharged through the storm water drain to the sea. The storm water drain will be provided with a block valve to contain any oil spill on site (outside of the bunded areas) if required.
10.3.2.13 The tank farm will be designed in accordance with the Code of Practice for Oil Storage Installations, 1992 issued by the Buildings Department. This Code makes reference to international codes such as API 650 for the design of tanks.
10.3.2.14 The PAFF facility will be provided with a number of security measures such as double security fencing, CCTVs within the security fence, and security guards on 24 hours duty.
10.3.3.1 The weather data used for the study is taken from the Tuen Mun weather station and is presented in Table 10.1 below.
Table 10.1 Weather Data from Tuen Mun Weather Station (1996)
|
|
10.3.4.1 To take into account the varying levels of population (on land and sea) at different times of day, the following time periods are assumed for the assessment.:
¨ For Tank Farm Events:
Weekday : 0.35
Weekend : 0.15
Night : 0.5
¨ For marine events:
Full Ferry : 0.2
Half Ferry : 0.6
Ferry 10% full : 0.2
10.3.4.2 This breakdown is based on the marine population during different times periods.
Land Population
10.3.5.1 Population data is based on estimates available for year 2016. The hazard assessment for PAFF is based on facilities that will be built by 2016 and 2040.
10.3.5.2 There is no residential development in the area around the tank farm, with the closest residential properties being about 2km away at Lung Kwu Tan. No residential development is planned in the future, with the Tuen Mun Area 38 and surrounding areas allocated for industrial land uses. The industrial population around the tank farm area is presented in Table 10.2.
10.3.5.3 The population on Lung Mum Road is based on an AADT value of 10,000 vehicles. A 1km section of the road is considered, with an average speed of 40km/hr and an average occupancy of 3.3 persons per vehicle. The calculation is as follows:
No of persons = (AADT x Length of Road x Vehicle occupancy/24 x Speed)
= 10000 x 1x 3.3/24 x 40 = 34.4.
Table 10.2 Population within 500m of
Tank Farm
|
CTS Zone |
Area (sq.m) |
Development |
2001 |
2016 |
|
163A |
179,000 |
Green Island Cement Plant (2) |
136 (1) |
136 (1) |
|
258A |
360,000 |
Shiu Wing Steel Plant (2) Proposed Airport Fuelling Facilities Proposed Special Industries Area |
300 0 0 |
300 21 992 (1) |
|
- |
- |
Access Road to Tank Farm |
2 |
2 |
|
- |
- |
Lung Mun Road |
35 |
35 |
|
TOTAL |
|
|
473 |
1,486 |
Source: Territorial Population and Employment Data
Matrices Dec. 1999.
Note:
1)
It is assumed
the number of worker population will be evenly distributed within the built up
areas inside the CTS Zone (excluding areas of greenbelt or country park).
2)
It is assumed
the worker population at existing developments will remain constant.
3)
The
population on Lung Mun Road is estimated based on an AADT (Annual Average Daily
Traffic) of 10,000 vehicles and a 1km section of the road.
10.3.5.4 The data on marine population in Urmston Road Channel was taken from the DNV (2001) report (7). The data is presented in Table 10.3.
|
Area |
Population
Density in 2011 and 2040 (people/hectare) |
||
|
Urmston Road Channel |
Ferry Nearly Full |
Ferry Half Full |
Ferry 10% Full |
|
0.15 |
0.12 |
0.10 |
|
10.4.1
Hazardous Properties of
Aviation Fuel
10.4.1.1 Aviation fuel is a mixture of aliphatic hydrocarbons with 10-16 carbon atoms and some aromatic hydrocarbons and naphthalene derivatives. It is a colourless to pale yellow oily liquid with a “fuel” odour. It is flammable and toxic if inhaled at high concentrations or ingested. According to Dangerous Goods Regulations, aviation fuel is classified as Category 5, Class 2 Dangerous Goods (those with a flash point of or exceeding 23oC but not exceeding 66oC). The UN classification is Class 3, Division 2. The physical and chemical properties of aviation fuel are summarised in Table 10.4 below.
Table 10.4 Physical
and Chemical Properties of Aviation Fuel
|
Property |
Value |
|
Molecular Weight |
156 g/mole |
|
Liquid density |
800 kg/m3 |
|
Boiling Point |
200-260oC |
|
Specific Gravity at 15oC |
0.8 (liquid) |
|
Flash Point |
38oC |
|
Flammable Limits |
0.7-5% vol |
|
Burning Rate |
0.04 mm/sec |
10.4.1.2 Aviation fuel floats on water and it is combustible when exposed to heat or flame. For the normal range of ambient temperatures in Hong Kong, the vapour pressure of aviation fuel is too low for it to form a flammable vapour cloud. Flashpoint is defined as the temperature at which the vapour pressure of the flammable substance is sufficient enough to give a concentration of vapour in the air that corresponds to the lower flammability limit.
10.4.1.3 The smoke effects of burning aviation fuel pool fires on the sea have not been modelled in this study, as the potential for interaction of hot smoke plumes with marine vessels or buildings is limited (the shipping channel is about 300m away). The smoke effects on buildings in neighbouring sites due to tank farm events are discussed in Section 10.6.2.25 to 10.6.2.28.
10.4.2
Hazards from the Tank Farm
10.4.2.1 Aviation fuel is classed as a flammable liquid and any spillage should be regarded as a potential fire risk. It has a flashpoint of >38°C(2) so a release of aviation fuel to the atmosphere may not immediately ignite given a source of ignition unless the ambient temperature is very high or the release is heated for some time, e.g. due to escalation such as fire from an adjoining tank.
10.4.2.2 The proposed tanks will be of a fixed roof design. There is a potential for light hydrocarbon vapours to build up in the headspace of the tank. This can present flammability/explosion hazards even at temperatures below the normal flashpoint. Openings are provided on the roof to minimise vapour buildup.
10.4.2.3 Overfilling is one of the main hazards from the tank farm. Level indication and independent level shutdown of inlet are provided.
10.4.2.4 Other hazards posed by the tank farm are:
¨ leak from tanks and associated piping due to corrosion, weld or material defect;
¨ fire or explosion during tank maintenance; and
¨ fire due to lightning strike.
10.4.2.5 Multiple tank failures, leaks from tanks and associated piping and overfill events have the potential to result in a bund fire. Some events, such as multiple tank failures, also have potential to result in liquid overtopping the bund and spreading to areas outside of the tank farm.
10.4.2.6 Tank fire events have potential to escalate to cause fire to adjoining tanks. This is discussed in Section 10.6.2.23 to 10.6.2.24.
10.4.2.7 Tosco Refining Corporation, Pennsylvania, USA: On 16th October 1998, at around 2:20 pm, a tank containing 700,000 gallons of jet fuel at the Tosco Refining Corporation’s Trainer, PA refinery caught fire and burned for more than four hours before being brought under control by 100 firefighters from 20 communities. Despite the magnitude of the event, there were no deaths or serious injuries.
10.4.2.8 Williams Tank Farm, Anchorage, Alaska, USA: On 6th April 2000, a fire started at the Williams Tank Farm in a tank that had reportedly been taken out of service. The fire started when something sparked and ignited 2000 gallons of Jet A1 inside the tank. The tank workers in the area were evacuated safely. After 15 fire engines responded, chemical foam was used to extinguish the blaze. The tank was apparently taken out of service, so workers could make repairs and clean the tank. No injuries were reported.
10.4.2.9 Tank Farm Fire, Denver, Colorado USA: On Sunday, November 25th 1990, a fire occurred at a flammable liquid tank farm supporting Denver’s Stapleton International Airport. Eight of the farm’s twelve storage tanks contained jet fuel-A totalling almost 4.2 million gallons. The fire, considered accidental in nature, burned for approximately 55 hours. 7 tanks were destroyed or damaged and over 1.6 million gallons of jet fuel was consumed. There were no reported fire fighter or civilian injuries as a result of this incident.
10.4.2.10 The above recent incidents suggest that even relatively large fire events involving jet fuel do not normally result in deaths or serious injuries offsite.
10.4.2.11 In 1996, about 5,000 litres of aviation fuel at Kai Tak International Airport was spilt from the top of one tank during product receipt. Most of the spilt product was contained within the bunded area. A small amount of fuel, about 200 litres, overspilt onto the road due to lack of clearance between the tank and the tank farm boundary wall
10.4.2.12 The duty operator, who was working close to the tank, arrested the spill within 2 minutes, by activating the emergency shut down button, thus closing the main inlet valve of the receiving line. There was no injury. The entire spilt product was subsequently recovered and disposed as a downgraded product. The 200 litres of aviation fuel evaporated within a few hours.
10.4.2.13 The above scenario however is very unlikely within the proposed PAFF tank farm for the following reasons:
¨ an automatic monitoring and a shut down systems will be installed to prevent tank overfilling from human error;
¨ unlike Kai Tak, there is at least 10 metres clearance between the tank and the bund wall; and
¨ there is also a much greater distance between the public road and the tank farm.
10.4.2.14 Nevertheless, the potential for bund overtopping is considered in the consequence assessment section.
Historical Tank Fires
10.4.2.15 A survey has been conducted to review historical tank fires involving kerosene and jet fuel. The primary database that was used was the MHIDAS database (3), which includes incidents from primarily the US and UK. Incidents were analysed in the period going back to 1970. Incidents prior to this time were thought not relevant due to the increasing levels of safety and safety management and improvements in technical integrity. A summary of the events is provided in Table 10.5.
Table 10.5 Kerosene/Jet
Fuel Tank Fires Since 1970
|
Date |
Location |
Type of Facility |
Ignition Source |
Injuries |
|
24th Oct 1995 |
Cilacap, Java |
Not known |
Lightning |
None |
|
25th Nov 1990 |
Denver, Colorado |
Airport tank farm |
Pump motor |
None |
|
13th Oct 1981 |
Yokohama, Japan |
US navy oil depot |
Not known |
Two |
|
23rd May 1980 |
Cologne, Germany |
Hydrogenation plant |
Not known |
Two |
|
30th Nov 1971 |
Virgin Islands |
Refinery |
Not known |
Eighteen |
|
29th April 1971 |
Calcutta, India |
Not known |
Not known |
None |
|
12th Jan 1970 |
Ajjacio, Corsica |
Storage facility |
Not known |
None |
10.4.2.16 It can be seen from the table that no off-site fatalities have resulted from the tank fires studied. For three of the kerosene/jet fuel events, injuries did occur, however, two of these events were at a refinery and a process plant, which are not directly comparable with PAFF. The only event at a tank farm where injuries did occur from a jet fuel tank fire was in Yokohama, Japan on 13th October 1981. The available information suggests that a tank containing 153,000 barrels of jet fuel at a US navy oil depot caught fire but the fire was controlled after 4 hours.
10.4.2.17 In summary, historical data of tank fire incidents show that major fires in storage terminals that contain kerosene/jet fuel are rare events. The available data show that there has only been one such event in the past twenty years. Furthermore, in none of the incidents is a major bund fire mentioned. The data would also suggest that major fires in storage terminals involving kerosene/jet fuel do not cause off-site fatalities.
Ignition Sources
10.4.2.18 For all tankage, the historical data show that lightning is a relatively common ignition source (where the ignition source is known). The incidents tend to be in areas of high electricity activity, certainly relative to Hong Kong. There was one incident in 1997 in Hong Kong when a 31,000 tonne tanker was struck by lightning after completion of unloading at one of the fuel terminals in Tsing Yi Island. The fire occurred on the gas vent pipe and was extinguished by Fire Services.
10.4.2.19 Welding has resulted in a number of incidents, with hot work being carried out in the vicinity of flammable materials. Hot work permits used at the facility decrease the likelihood of such an event considerably.
10.4.2.20 Several incidents have occurred in the past where there has been a catastrophic failure of an atmosphere storage tank containing petroleum products. Following such a failure the tank contents have been released and in some instances the material has been lost outside the secondary containment due to the momentum from the initial surge.
10.4.2.21 Table 10.6 below gives details of the catastrophic tank failure incidents found on the MHIDAS database. These relate to incidents, since 1970, involving tanks containing petroleum products. Since the tank design for all petroleum products are similar, references to these incidents are also included. Where there was a spill outside the secondary containment, this is noted.
Table
10.6 Catastrophic
Tank Failures of Petroleum Products Tanks Since 1970
|
Date |
Location |
Fuel |
Failure Cause |
Spill Contained by Bund |
|
29th July 1993 |
El Segundo, CA, USA |
Fuel oil |
Not known |
No - about 2% lost |
|
11th May 1993 |
Fawley, UK |
Bunker oil |
Mechanical |
Unknown if bund used |
|
Oct 1989 |
Richmond, CA, USA |
Gasoline |
Earthquake |
Yes |
|
6th Feb 1989 |
New Haven, CT, USA |
Heating oil |
Mechanical |
Yes |
|
11th July 1988 |
Brisbane, Australia |
Gasoline |
Corrosion |
Not known |
|
2nd Jan 1988 |
Floreffe, PA, USA |
Diesel oil |
Mechanical |
No - 40 to 71% lost |
|
28th Dec 1980 |
El Dorado, KS, USA |
Solvents |
Mechanical |
Not known |
10.4.2.22 It can be seen that, in addition to the incident at Kai Tak, there have been other incidents where it is known that material overtopped a bund wall. There have also been a number of incidents involving crude oil storage tanks where material has been lost over the bund walls(4).
Other Hazards
10.4.2.23 Other hazards such as landslide, subsidence, typhoon, wind and earthquake, which could cause damage to the tanks, will be addressed in the design in accordance with the relevant engineering codes and standards adopted for the PAFF and therefore not considered further. As regards aircraft crash, the PAFF facility does not lie near to the flight path and therefore such incidents are not considered further.
10.4.3
Hazards From Submarine
Pipeline
10.4.3.1 Loss of containment could be due to various causes such as corrosion, material/weld defect but is largely dominated by marine traffic impact.
10.4.3.2 In the event of a fuel release, fuel could be ignited by a passing vessel or by the vessel that caused damage and is still present in the vicinity.
10.4.3.3 A brief description of the causes of failure of a submarine pipeline is included below.
Anchor
Drop/Drag
10.4.3.4 Anchor drop/drag is the dominant cause of failure or damage to a submarine pipeline. This occurs when a ship's anchor is set off inadvertently or due to an emergency. When an anchor is dropped, it undergoes a free fall, reaches the bottom with a known velocity and penetrates the soil and causes damage to any pipeline in its path. The type of damage that could be caused will vary depending on the size of anchor and other factors such as whether the pipeline is buried etc. Generally, it could damage the concrete coating, cause a dent or cause the pipe to tear open.
10.4.3.5 The potential for anchor drop depends on the proximity of the pipeline route to port/harbour areas, fairways and anchorage areas.
10.4.3.6 In the fairways, vessels will be on the move and if any vessel drops anchor it is more likely to collide with other passing vessels and hence the frequency of such an event is expected to be low. Also, since the pipeline will be marked on admiralty nautical charts, it is expected that passing vessels will be aware of the presence of the pipeline and therefore will not drop anchor in the vicinity.
10.4.3.7 Nevertheless, anchor drop incidents occur due to emergency conditions or due to human error. Emergency situations may include ship machinery failure, collision or poor weather (adverse wind, typhoon, fog etc.). Emergency anchoring due to poor weather conditions does not usually occur since all ocean-going vessels have advanced navigation systems on board.
10.4.3.8 Along the proposed route, the section crossing the Urmston Road Channel (it is to be noted that it is not a designated channel), there is a significant amount of River trade vessels. However, these vessels range from less than 1000 dwt to about 5000 dwt and very rarely greater than 10,000 dwt.
10.4.3.9 Anchor drag occurs when a moving vessel drops anchor and therefore the anchor gets dragged over some distance. The drag distance could be assumed as about 50m although it could be higher if the anchor is dropped at high vessel speed. If there is a submarine pipeline along the anchor drag path, anchor dragging onto the pipeline may result in localised buckling or denting of the pipeline, or over-stressing from bending if the tension on the anchor is sufficient to laterally displace the pipeline. A dragged anchor may also hook onto a pipeline during retrieval causing damage as a result of lifting the pipeline.
10.4.3.10 It is to be noted that an anchor dropped vertically will penetrate deeper than anchor drag. However, the probability of a direct hit on a pipeline from an anchor drop is generally low as compared to damage due to drag, as the impact area is much larger.
10.4.3.11 Not all types of vessels have potential to cause anchor damage. This can be estimated deterministically. Damage to a submarine pipeline lowered to 3m below the seabed will occur only when anchors penetrate 3m into the seabed. The anchor penetration depth is a function of the water depth, soil conditions and anchor characteristics. It is therefore possible to estimate the minimum anchor size that can penetrate 3m into the seabed for given soil condition and thereby estimate the proportion of vessels that have anchors exceeding the minimum size.
10.4.3.12 Anchor sizes are broadly related to vessel sizes and conform to international standards. The estimated average anchor sizes based on the commonly used US stockless anchor for typical vessel sizes are given as 4.2 tonnes for 25,000 dwt vessel, 6.8 tonnes for 50,000 dwt vessel and 11.6 tonnes for 100,000 dwt vessel.
10.4.3.13 Based on theoretical calculations on penetration depth for various anchor sizes for an assumed soil shear strength (of 12 kPa), it is found that the minimum anchor size required to penetrate to 3m depth is about 7 tonnes. This will correspond to vessel sizes of 25,000 dwt. For lower soil shear strength as in the case of top soil, anchor sizes that can penetrate to 3m may be lower.
10.4.3.14 The data gathered by the Consultants on another study(5) provided the following details on the type and size of vessels operating in Hong Kong waters:
¨ 69% of all ocean-going (OG) vessels entering Hong Kong waters are less than 10,000 dwt while 96% of all OG vessels are less than 50,000 dwt;
¨ if only fully cellular container vessels are considered, 42% are less than 10,000 dwt while 92% are less than 50,000 dwt; and
¨ all river trade cargo vessels are less than 5000 dwt.
10.4.3.15 It is found that all river trade cargo vessels, fishing vessels, construction barges etc. which are less than 10,000 dwt, even if they anchor, are unlikely to cause damage to the pipeline. Amongst ocean-going vessels, only about 15% have the potential to cause anchor damage.
10.4.3.16 In addition, the pipeline under study will also have rock armour protection further mitigating any anchor impacts.
Vessel
Sinking
10.4.3.17 An analysis of incidents of vessel sinking/grounding in Hong Kong waters for the years 1997 and 1998 showed that these incidents occurred mostly in Victoria Harbour and Ma Wan Channel. Also, the incidents are dominated by mid-stream and construction vessels while about 10% involved river trade cargo vessels. The size of these vessels varied between 100 to 300 dwt but less than 1000 dwt.
10.4.3.18 Also, the pipeline will be trenched 3m below sea bed (more than 6m below sea bed in some sections such as Urmston Road shipping channel crossing) with rock armour protection and therefore vessel sinking is not considered to pose a hazard to the pipeline.
Accidental Dropping of Containers
10.4.3.19 Freight containers may get dropped accidentally due to collision, vessel sinking or improper stowage. These containers typically weight about 10 tonnes and would not cause damage to the pipeline if it were to land on top of the pipe.
Fishing
Activity
10.4.3.20 The vessels of concern are stern trawlers with lengths up to 30m. Trawl gear operation is however, unlikely to involve penetration depths greater than 1m. In the present case, where the pipeline will be laid to 3m below the seabed, potential for damage due to fishing is not expected.
Dredging
Activities
10.4.3.21 Dredging vessels could cause damage due to dredging operations that involve cutting heads. They could also cause damage to the pipeline by anchoring.
10.4.3.22 Deep maintenance dredging activities currently occur along the pipeline route for a coal berth for CLP. Therefore, in this section, the pipeline will be lowered to around 6-7m below seabed. It is assumed that dredging operations by others will be closely monitored and controlled and therefore potential for damage due to dredging is considered to be low.
Corrosion
10.4.3.23 The proposed pipeline will be protected against external corrosion by sacrificial anodes. However, ineffective corrosion protection due to a failure or breakdown of the system could cause external corrosion resulting in general or local loss of wall thickness leading to failure.
Construction
Damage
10.4.3.24 Damage to the pipeline during construction is recognised as a potential hazard. For example, during pipelay, the pipeline will be laid by barge or bottom pulled into position to 3m below seabed followed by installation of additional protection such as sandfill and rock armour. During this transient phase, where the pipeline lies in the trench unprotected, damage due to anchoring is a threat.
10.4.3.25 There are a number of procedural measures that can be adopted such as deployment of mooring buoys or patrol boats to warn passing ships and thereby prevent potential incidents. However, if damage were to occur, this will be revealed during hydrotest and pigging and accordingly rectified. The only consequence is the costs for repair and therefore this is not considered further in the risk assessment.
Natural
Hazards
10.4.3.26 Natural hazards such as subsidence, earthquake and typhoon, environmental loads (currents and waves) etc. may cause varying degrees of damage to the pipeline. It is expected that the pipeline will be designed to suitable standards taking into account prevailing local conditions.
10.4.4
Hazards Due to Marine Transport
10.4.4.1 There are many possible types of accident, which may occur to a vessel during transit and result in a spillage, or leakage of cargo. In this study only marine transport in the vicinity of the jetty ie within 500m is considered. The main causes of loss of containment are discussed below.
Collisions
10.4.4.2 Collision is defined as a contact between the tanker and another vessel underway, drifting, on tow or otherwise untethered. This event is largely related to the level of marine traffic in the channel.
Grounding
10.4.4.3 Grounding is defined as a ship coming into unintended contact with a seabed or shore.
Fires/Explosions
10.4.4.4 In this study, only fires or explosion within the vessel itself and associated with the vessel's own system are considered. Fires caused by collision/grounding (leading to spill and subsequent fire) are dealt with separately.
10.4.4.5 Most fires on the vessels will have no significant risk other than to the crew or fire fighters. The only scenarios, which pose risk to third parties, are:
¨ large spills of aviation fuel, that ignite on sea surface and spread to boats and other marine traffic in the vicinity (covered under grounding/collisions); and
¨ explosions of ship tanks containing fuel vapour causing blast and debris.
10.4.5.1 In this section the hazards associated with jetty operations is described. The sequence of events leading to a loss of containment is briefly discussed.
Cargo
Unloading Operations
10.4.5.2 The size of tankers delivering cargoes to each jetty will range from 10,000 dwt to 80,000 dwt.
10.4.5.3 At the jetty, the following can lead to a loss of containment.
Loading arm rupture
10.4.5.4 The loading arm could be incorrectly connected; the purge valve left open during delivery or liquid could still be in the line when the purge valve is opened. These events could lead to a relatively small leak of aviation fuel. The loading arm could rupture due to a variety of reasons, including corrosion, material defect, construction defect and excessive movement. The rupture could result in a large amount of aviation fuel release, particularly if loading is not stopped immediately.
Pipeline rupture or leak
10.4.5.5 The submarine pipeline from the loading arm to the tank farm could rupture due to material defect, corrosion or impact. However, as compared to the subsea pipeline from the tank farm to the airport this section will be less exposed to impact from marine vessels.
Riser Rupture
10.4.5.6 At the jetty there is a potential for the riser to rupture in the event of a vessel impacting the jetty or a passing vessel striking a berthed vessel. However, the riser is built into the jetty structure and the maximum spill quantity will be limited to the inventory in the pipeline to the tank farm from the jetty, which is about 400m long.
Striking
10.4.5.7 Striking involves a drifting vessel (which probably lost control while in the channel) impacting the aviation fuel tanker while it is berthed.
Impact
10.4.5.8 Impact is defined as a vessel running into a dock wall or a jetty. This event depends upon the number of floating objects to be encountered and the space available to take avoiding action. The effect on the vessel will depend on the size of the obstruction, in the context of this study it is considered that the strength of the impact is very unlikely to result in rupture. It is assumed that all the tankers arriving at the jetty will be of double hull construction and therefore should contain the fuel to a certain extent following an impact.
10.4.6
Checklist of Representative
Events
10.4.6.1 The following checklist of representative events has been selected from the discussion on hazards above. This is presented in Table 10.7.
Table 10.7 Checklist of Representative Events
|
Source |
Cause |
Size
|
Consequence |
|
Marine Transport |
Collision |
Small Leak |
Sea Surface Pool Fire |
|
|
|
Large Leak |
Sea Surface Pool Fire |
|
|
|
Rupture |
Sea Surface Pool Fire |
|
|
Grounding |
Small Leak |
Sea Surface Pool Fire |
|
|
|
Large Leak |
Sea Surface Pool Fire |
|
|
|
Rupture |
Sea Surface Pool Fire |
|
|
Fire/Explosion |
|
Blast Overpressure/Fragments |
|
Marine Jetty |
Impact |
Small Leak |
Sea Surface Pool Fire |
|
|
|
Large Leak |
Sea Surface Pool Fire |
|
|
|
Rupture |
Sea Surface Pool Fire |
|
|
Striking |
Small Leak |
Sea Surface Pool Fire |
|
|
|
Large Leak |
Sea Surface Pool Fire |
|
|
|
Rupture |
Sea Surface Pool Fire |
|
|
Loading Arm |
Leak |
Sea Surface Pool Fire |
|
|
|
Rupture |
Sea Surface Pool Fire |
|
Submarine Pipeline |
Corrosion/Defect/ Third Party |
Small Leak |
Sea Surface Pool Fire |
|
|
|
Medium Leak |
Sea Surface Pool Fire |
|
|
|
Rupture |
Sea Surface Pool Fire |
|
Tank Farm |
All Causes |
Tank Top |
Tank Top Fire |
|
|
|
Catastrophic Rupture |
Bund Fire/Pool Fire |
10.5.1.1 This section of the report includes an assessment of the frequency of release of aviation fuel. The following releases are considered:
¨ tank farm releases;
¨ twin submarine pipeline;
¨ marine transport events; and
¨ marine jetty operations.
10.5.1.2 The frequencies derived in this section correspond to initiating event frequencies and the probabilities of various outcomes following the initiating event.
10.5.1.3 The approach to frequency analysis is based on the application of historical data worldwide for similar systems modified suitably to reflect local factors. Previous studies(7) carried out for the proposed facility have also been reviewed to ensure consistency of approach and data.
10.5.2
Frequency of Tank Farm Releases
Atmospheric
Storage Tanks Fires
10.5.2.1 OPITSC, 1990: The best available data on fire frequencies in atmospheric storage tanks are taken from the OPITSC study, which is based on the following sources. An analysis of the tank fire frequencies from these sources is summarised in Table 10.8:
¨ the API Risk Analysis Task Force (Document No 1, May 1977). Source data was for the period between 1969 and 1977;
¨ USA Refinery and Petrochemical data from a confidential operating source covering the period between 1965 to 1975;
¨ Saval-Kronenburg bv (a manufacturer of fire extinguishers) database for the period 1981-1984 covering world-wide experience with tanks installed with their extinguishing systems.
¨ Scottish North Sea Terminals for 641 tank-years; and
¨ Singapore data since 1945.
Table 10.8 Analysis
of Tank Fire Frequency Data
|
Source |
Tank Type |
No of Incidents |
No of tank years |
Failure Frequency (per tank-year) |
|
API
(1969-1977) |
Cone Roof |
270 |
100,000 |
2.7 x 10-3 |
|
|
Floating Roof |
63 |
30,000 |
2.1 x 10-3 |
|
Lees (US refineries 1982-85) |
Cone Roof |
29 |
100,000 |
2.9 x 10-4 |
|
|
Floating Roof |
25 |
30,000 |
8.3 x 10-4 |
|
Saval-Kronenburg (1981-1984) |
Floating Roof |
1 |
673 |
1.5 x 10-3 |
|
US
Refinery & Petrochemical Data (1965-1975) |
Floating Roof |
10 |
3,883 |
2.6 x 10-3 |
|
Scottish
North Sea Oil Terminals |
Floating Roof |
1 |
461 |
2.2 x 10-3 |
|
Singapore
(1945-1989) |
Cone Roof |
2 |
11,125 |
1.8 x 10-4 |
|
|
Floating Roof |
12 |
2,151 |
5.6 x 10-3 |
10.5.2.2 The US refineries and Singapore data are more relevant for cone roof tanks. These data also account for advances in storage tank safety and more accurately reflects the case under study. The frequency for fixed cone roof tanks is generally 1.8 to 2.9 x 10-4 per tank year. A value of 3.0 x 10-4 per tank year is adopted.
10.5.2.3 Fixed cone roof tanks are generally used for flammable (flashpoint between 23-66oC) liquids and combustible liquids (>60oC) as is the case for aviation fuel, which is flammable with a flashpoint of 38oC. Therefore, the fire frequency of cone roof tanks, as applied to the proposed tank farm, is considered appropriate. However, this value is regarded as conservative as it is based on average worldwide conditions.
10.5.2.4 Therefore for 12 tanks the total tank fire frequency is 3.6 x 10-3 per year (2040 case). For 2016 case, there will only be 6 tanks in the tank farm and therefore the total tank fire frequency will be 1.8 x 10-3 per year.
Atmospheric
Storage Tank Failures
10.5.2.5 Davies et al.(8) cite the following reasons for catastrophic releases from storage vessels after inspection of incidents recorded on the MHIDAS database (Major Hazardous Incidents Database):
¨ brittle failure of primary containment, sometimes caused by rapid changes in ambient temperature;
¨ failure of tank seams due to fire impingement;
¨ failure of the tank during the initial filling process;
¨ boilover of tank contents (applicable mainly for crude oil and fuels containing water); and
¨ acts of vandalism or sabotage.
10.5.2.6 Of the above, brittle failure due to rapid changes in ambient temperature is not expected in Hong Kong. Rapid changes in temperature of contents are also not expected, as aviation fuel received will be at ambient temperature. Failure of the tank due to overfilling could occur but an independent high level shutdown system is provided to prevent overfilling. Boilover of tank contents is not relevant in the case of aviation fuel.
10.5.2.7 Other causes of tank failure are weld or material defect, corrosion, settlement, and fire due to ignition of tank vent by lightning. Stringent quality control measures will be adopted during procurement of plate material and during construction and therefore failures due to weld or material defect are not expected at the PAFF facility. Corrosion will be monitored during the operational phase and therefore failure due to corrosion is not expected at the PAFF facility. Settlement of tank foundation will be monitored during construction as well as during initial operations. Although the site is located on reclaimed land, the reclamation was carried out some years ago and therefore general settlement of the site is not expected.
10.5.2.8 The tank vapour space could be in the flammable range due to vent opening to atmosphere and therefore ignition of tank vent due to lightning could result in a tank fire and subsequent failure at the roof to shell connection (API 650 tanks are provided with a weak roof to shell connection which will fail preferentially to any other joint). Such failures could also occur in the event of fire impingement to relieve excess vapours. It may be assumed that in the event of roof failure, the top most plate of the shell connecting to the roof may also fail resulting in spill onto bund. Each plate is about 3m high which is about 10% of the tank height. The catastrophic failure of the tank is therefore assumed to result in a release of 10% of tank contents (ie about 3,900 m3) on to the bund. Aircraft crash is the only conceivable incident that can result in more than 10% of tank contents but this can be discounted, as the proposed site does not lie near to the flight path.
10.5.2.9 A number of studies have been carried out to estimate the catastrophic failure frequency of storage tanks. The following Table 10.9 gives the failure rates derived for failures of storage tanks in literature.
Table 10.9 Failure
Frequencies of Atmospheric Storage Tanks
|
Source |
Type of Failure |
Failure Frequency (per tank-year) |
|
Batstone
& Tomi (9) |
Catastrophic
rupture |
3 x 10-5 |
|
COVO Study (10) |
Serious leakage (50 mm hole) |
1 x 10-4 |
|
|
Catastrophic rupture |
6 x 10-6 |
|
Taylor (11) |
Large Leak |
8.8 x 10-4 |
|
1 x 10-5 |
||
|
E&P Forum (12) |
Major Release |
6.9 x 10-6 |
|
Catastrophic Rupture |
2 x 10-7 |
|
|
Christiansen & Eilbert (14) |
4 x 10-6 |
|
|
|
Catastrophic Rupture (tanks > 10,000 barrels) |
9.2 x 10-6 |
10.5.2.10 In this study, a frequency of 6.6 x 10-6 per tank-year is adopted for catastrophic tank rupture. This frequency is generally similar to the COVO study(10) and to that given by the Hartford Steam Boiler paper for a ‘major release’.
Bund Fire
10.5.2.11 Davies (8) et al suggests a bund fire frequency in a common bund of 1.2 x 10-5 per year for a flammable liquid. This frequency includes all causes such as tank failure, overfilling, pipework failure etc. Based on this value the bund fire frequencies for 2016 case (1 bund) and 2040 case (2 bunds) are derived as 1.2 x 10-5 per year and 2.4 x 10-5 per year.
10.5.2.12 A summary of the tank event frequencies is given in Table 10.10. Further discussion on the outcomes following these initiating events is given in the consequence assessment section.
Table 10.10 Summary
of Tank Event Frequencies
|
Year |
Tank Fire Frequency (per year) |
Catastrophic Tank Rupture Frequency
(per year) |
Bund Fire Frequency (per year) |
|
2016 |
1.8 x 10-3 |
3.96 x 10-5 |
1.2 x 10-5 |
|
2040 |
3.6 x 10-3 |
7.92 x 10-5 |
2.4 x 10-5 |
10.5.3.1 The most comprehensive failure database for submarine pipelines is described in the report published by UK Health and Safety Executive titled 'PARLOC 96’ (16), which covers incidents until year 1995. This provides information on pipeline failure causes and frequency of such failures. The information in this database is based on data obtained from regulatory authorities in the UK, Norway, the Netherlands, Denmark and Germany, Operators in the UK, Dutch and Danish sectors and published sources.
10.5.3.2 An extensive review of the database has been conducted in a previous study (5) and the failure frequency has been derived for a submarine pipeline considering only those failures relevant to the pipeline under consideration.
10.5.3.3 The failure frequency has been derived separately for mid-line and pipelines within platform safety zone (500m). The higher failure rate in the safety zone (an order of magnitude higher than mid-line) is due to the effect of increased ship/barge movements in the vicinity and the potential for anchor damage as a result.
10.5.3.4 For those sections along pipeline route, where there is a high level of marine traffic (consisting of large vessels > 10,000 dwt), the potential for anchor/impact incidents is expected to be higher and therefore the failure rate corresponding to the safety zone is considered appropriate. For other sections, where there is low marine traffic (consisting of large vessels), the failure rate derived for mid-line is applied. For moderate level of marine traffic, an intermediate value is considered.
10.5.3.5 The generic failure frequency values are given in Table 10.11.
Table 10.11 Generic
Failure Frequency
|
Cause |
Failure
Rate (per km per year) Based on Level of Marine Activity |
||
|
|
High |
Moderate |
Low |
|
Anchor/ Impact |
5 x 10-4 |
8.4 x 10-5 (a) |
2.8 x 10-5 |
|
|
|
|
|
|
Corrosion/ Others |
1.6 x 10-5 |
1.6 x 10-5 |
1.6 x 10-5 |
|
Total |
5.2 x 10-4 (1 in 1,923 km-yrs) |
1 x 10-4 (1 in 10,000 km-yrs) |
4.4 x 10-5 (1 in 22,727 km-yrs) |
Note: (a)
Value assumed 3 times the value for mid-line, i.e. 3 times 'low' value.
10.5.3.6 The submarine pipeline from the jetty to the airport passes through Urmston Road and North Lantau Channel, both of which have ‘low’ marine traffic in terms of ocean-going vessels relative to other shipping channels in Hong Kong such as East Lamma Channel. The DNV 2000 Report (7), assumes a value of 3.4 x 10-5 per km per year as the leak frequency. For consistency the same value will be assumed. However, this value does not assume any protection. The PAFF pipeline will be lowered to 3m below seabed and protected by backfill and rock armour. The design basis for rock armour protection is 20 tonnes anchors. In a previous study (5), ERM has estimated that this anchor size of 20 tonnes corresponds to vessel sizes much greater than 100,000 dwt. Also, 99% of ocean-going vessels are less than 100,000 dwt. Therefore; it is considered appropriate to reduce the frequency of pipeline failure by 99% to 3.4 x 10-7 per km per year.
10.5.3.7 Three leak sizes are assumed. The sizes and their proportions are given in Table 10.12.
Table 10.12 Leak
Size Distribution and Failure Frequencies
|
Size |
Hole
Size (mm) |
Proportion
of Leaks |
Frequency
(per km- year) |
|
Small Leak |
20 |
57% |
3.88 x 10-7 |
|
Medium Leak |
50 |
15% |
1.02 x 10-7 |
|
Rupture |
500 |
28% |
1.90 x 10-7 |
10.5.3.8 An ignition probability of 0.01 is assumed for the subsea release of the fuel igniting on the sea surface. This is the value derived for minor liquid releases by Cox (27). Given the emulsification of fuel with water following a submarine release, this ignition probability (which is lower than ignition of a surface spill of fuel) is considered appropriate. Also, it is assumed that only rupture of the pipeline will result in sufficient fuel to reach the sea surface and ignite. Details of the behaviour of submarine release of aviation fuel are given in the consequence section.
10.5.3.9 Therefore, the resultant scenario frequency of the pool fire on the sea surface following rupture of the pipeline is calculated as 1.9 x 10-7 x 0.01 = 1.9 x 10-9 per km per year. For 4.8km of the pipeline, the frequency per year is calculated as 9.12 x 10-9 per year.
10.5.4
Frequency of Marine Events
10.5.4.1 Three main hazard events have been discussed in the section on Hazard Identification. These are:
¨ collision;
¨ grounding; and
¨ fire/explosion
Collision
10.5.4.2 The frequency of collisions will depend upon the number of ship visits and encounters.
10.5.4.3 The collision frequency is estimated as 3.5 x 10-5 per encounter. The encounter frequency for the 2011 and 2040 is given as 0.69 per km (DNV (2000) (7) report). The interaction distance is 0.5km as only marine transport within 500m of the jetty is considered in this study. Therefore the collision frequency is given as 3.5 x 10-5 x 0.69 x 0.5 = 1.2 x 10-5 per visit. The frequencies of collisions have been allocated based on visits of different tanker sizes. These values are presented in Table 10.13 and Table 10.14 for the years 2016 and 2040 respectively.
Table
10.13 Frequency
of Collisions (2016 Case)
|
Tanker Size (dwt) |
Base Frequency per visit |
No of visits/year |
Frequency per year |
|
20,000 |
1.2
x 10-5 |
60 |
7.2 x 10-4 |
|
45,000 |
1.2
x 10-5 |
62 |
7.44 x 10-4 |
|
60,000 |
1.2
x 10-5 |
21 |
2.52 x 10-4 |
Table 10.14 Frequency
of Collisions (2040 Case)
|
Tanker Size (dwt) |
Base Frequency per visit |
No of visits/year |
Frequency per year |
|
30,000 |
1.2
x 10-5 |
70 |
8.4 x 10-4 |
|
45,000 |
1.2
x 10-5 |
80 |
9.6 x 10-4 |
|
80,000 |
1.2
x 10-5 |
38 |
4.56 x 10-4 |
Grounding
10.5.4.4 The frequency of grounding is influenced by the following factors:
¨ distance travelled by the ship in restricted water;
¨ width of navigable water;
¨ nature of shoreline and seabed (weather smooth or complex);
¨ likelihood of poor visibility in the area;
¨ reliability of machinery on the ships;
¨ density of marine traffic in the area; and
¨ availability of VTS guidance etc.
10.5.4.5 The frequency of grounding is expressed on a per km basis. This value (7) adopted is 4.3 x 10-6 per km travelled.
10.5.4.6 The distance travelled by vessels within the vicinity of the jetty is 0.5km. Therefore the frequency of grounding per year is calculated to be 4.3 x 10-6 x 0.5 x 143 = 3.1 x 10-4 per year (2016 case). For the 2040 case, the frequency of grounding per year is 4.04 x 10-4 per year.
10.5.4.7 The allocation of the total grounding frequency for different tanker sizes is based on the number of visits. This is presented in Table 10.15 and Table 10.16.
Table 10.15 Grounding
Frequency for Different Size Tankers (2016 case)
|
Tanker
Size (dwt) |
Base
Frequency (per km) |
Interaction
Distance (km) |
No
of visits/year |
Frequency
per year |
|
20,000 |
4.3 x 10-6 |
0.5 |
60 |
1.29 x 10-4 |
|
45,000 |
4.3 x 10-6 |
0.5 |
62 |
1.33 x 10-4 |
|
60,000 |
4.3 x 10-6 |
0.5 |
21 |
4.52 x 10-5 |
Table 10.16 Grounding
Frequency for Different Size Tankers (2040 case)
|
Tanker
Size (dwt) |
Base
Frequency (per km) |
Interaction
Distance (km) |
No
of visits/year |
Frequency
per year |
|
30,000 |
4.3 x 10-6 |
0.5 |
70 |
1.51 x 10-4 |
|
45,000 |
4.3 x 10-6 |
0.5 |
80 |
1.72 x 10-4 |
|
80,000 |
4.3 x 10-6 |
0.5 |
38 |
8.17 x 10-5 |
Fire/Explosion
10.5.4.8 Explosion/fire frequencies in the channel within 500m from the jetty depends mainly on the following factors:
¨ time spent by ship in harbour;
¨ cargo carried;
¨ operations permitted (such as tank cleaning);
¨ standard of operational safety on ship;
¨ design of tanks and equipment on ships; and
¨ probability of lightning storms (for ignition).
10.5.4.9 The frequency of explosion due to fire on board has been taken as 1.2 x 10-8 per km. This value is consistent with that adopted in the DNV 2000 report. The distance travelled by each tanker is only 500m and therefore this frequency is 1.2 x 10-8 x 0.5 x number of visits per year. The allocation of the total fire/explosion frequency for different tanker sizes is based on the number of visits. This is presented in Table 10.17 and Table 10.18.
Table 10.17 Fire
and Explosion Frequencies (2016 Case)
|
Tanker
Size (dwt) |
Base
Frequency (per km) |
Interaction
Distance (km) |
No
of visits/year |
Frequency
per year |
|
20,000 |
1.2 x 10-8 |
0.5 |
60 |
3.6 x 10-7 |
|
45,000 |
1.2 x 10-8 |
0.5 |
62 |
3.7 x 10-7 |
|
60,000 |
1.2 x 10-8 |
0.5 |
21 |
1.26 x 10-7 |
Table 10.18 Fire
and Explosion Frequencies (2016 Case)
|
Tanker
Size (dwt) |
Base
Frequency (per km) |
Interaction
Distance (km) |
No
of visits/year |
Frequency
per year |
|
30,000 |
1.2 x 10-8 |
0.5 |
70 |
4.2 x 10-7 |
|
45,000 |
1.2 x 10-8 |
0.5 |
80 |
4.8 x 10-7 |
|
80,000 |
1.2 x 10-8 |
0.5 |
38 |
2.28 x 10-7 |
10.5.5
Frequency of Jetty Events
Impact
10.5.5.1 The berthing impact frequency has been taken from the Caltex Safety Case (17). This value is assumed to be 7.4 x 10-5 per visit. Based on the number of visits to the jetty the failure frequencies due to impact are presented in Table 10.19 and Table 10.20.
Table 10.19 Frequency
of Impacts (2016 Case)
|
Tanker Size (dwt) |
Base Frequency (per visit) |
No of visits/year |
Frequency per year |
|
20,000 |
7.40 x 10-5 |
60 |
4.44 x 10-3 |
|
45,000 |
7.40 x 10-5 |
62 |
4.59 x 10-3 |
|
60,000 |
7.40 x 10-5 |
21 |
1.55 x 10-3 |
Table 10.20 Frequency
of Impacts (2040 Case)
|
Tanker Size (dwt) |
Base Frequency (per visit) |
No of visits/year |
Frequency per year |
|
30,000 |
7.40 x 10-5 |
70 |
5.18 x 10-3 |
|
45,000 |
7.40 x 10-5 |
80 |
5.92 x 10-3 |
|
80,000 |
7.40 x 10-5 |
38 |
2.81 x 10-3 |
Striking
10.5.5.2 The frequency of strikings will depend upon the number of ship visits.
10.5.5.3 The striking frequency is estimated as 8 x 10-6 per movement (struck when berthed). The frequencies of strikings have been allocated based on visits of different tanker sizes. These values are presented in Table 10.21 and Table 10.22 for the years 2016 and 2040 respectively.
Table 10.21 Frequency of Strikings (2016 Case)
|
Tanker Size (dwt) |
Base Frequency (per visit) |
No of visits/year |
Frequency per year |
|
20,000 |
8
x 10-6 |
60 |
4.8 x 10-4 |
|
45,000 |
8
x 10-6 |
62 |
4.96 x 10-4 |
|
60,000 |
8
x 10-6 |
21 |
1.68 x 10-4 |
Table 10.22 Frequency
of Strikings (2040 Case)
|
Tanker Size (dwt) |
Base Frequency(per visit) |
No of visits/year |
Frequency per year |
|
30,000 |
8
x 10-6 |
70 |
5.6 x 10-4 |
|
45,000 |
8
x 10-6 |
80 |
6.4 x 10-4 |
|
80,000 |
8
x 10-6 |
38 |
3.04 x 10-4 |
Loading
Arm Failure
10.5.5.4 The frequency of loading hoses failure is derived as 9 x 10-8 per hour of operation (EMSD studies). The frequency of loading arm ruptures have been considered to occur at an order of magnitude lower than loading hoses ie 9 x 10-9 per hour.
10.5.5.5 There are 143 visits to the jetty per year (2016 case) and 188 visits for 2040 case and each transfer operation is assumed to take 20 hours on an average. Therefore, the rupture frequency of the loading arms is calculated as
Frequency of Rupture: 9 x 10-9 x 143 x 20 = 2.57 x 10-5 per year (2016 case)
Frequency of Rupture: 9 x 10-9 x 188 x 20 = 3.38 x 10-5 per year (2040 case)
Submarine
Pipeline Failure Rate
10.5.5.6 The submarine pipeline from the loading arm to the tank farm will be treated in exactly the same manner as the submarine pipeline from the tank farm to the airport. This has already been discussed in Section 10.5.3.
10.5.6
Spill Size Distribution
10.5.6.1 The following spill size distribution (Table 10.23) is assumed in this study.
Table 10.23 Spill Size Distribution for
Various Failures
|
Hazard
Source |
Cause |
Size
of Leak |
Probability
of Leak |
|
Tank Farm |
All Causes |
Rupture |
1.0 |
|
Submarine Pipeline |
All Causes |
Small (20mm) |
0.57 |
|
|
|
Medium (50mm) |
0.15 |
|
|
|
Rupture (500mm) |
0.28 |
|
Marine Transport |
Collisions |
Small (0.3% of dwt) |
0.2 |
|
|
|
Medium (1% of dwt) |
0.2 |
|
|
|
Rupture (7% of dwt) |
0.6 |
|
|
Grounding |
Small (0.3% of dwt) |
0.2 |
|
|
|
Medium (1% of dwt) |
0.2 |
|
|
|
Rupture (7% of dwt) |
0.6 |
|
|
Fire/Explosion |
N/A |
N/A |
|
|
|
|
|
|
Marine Jetty |
Impact |
Small (0.3% of dwt) |
0.2 |
|
|
|
Medium (1% of dwt) |
0.2 |
|
|
|
Rupture (7% of dwt) |
0.6 |
|
|
Strikings |
Small (0.3% of dwt) |
0.2 |
|
|
|
Medium (1% of dwt) |
0.2 |
|
|
|
Rupture (7% of dwt) |
0.6 |
|
|
Loading Arm |
Rupture |
1.0 |
|
|
Submarine Pipeline |
Small (20mm) |
0.57 |
|
|
|
Medium (50mm) |
0.15 |
|
|
|
Rupture (500mm) |
0.28 |
10.5.7.1 The spill probabilities used in this assessment (taken from the DNV 2000 report) are given in Table 10.24.
Table
10.24 Spill
Probabilities
|
Cause |
Double
Hull Tanker |
|
|
|
<
20,000 dwt |
>
20,000 dwt |
|
Collisions |
0.015 |
0.0075 |
|
Striking
|
0.015 |
0.0075 |
|
Grounding
|
0.03 |
0.015 |
|
Impact
|
0.015 |
0.0075 |
10.5.8.1 The ignition probabilities assumed for the various release scenarios is given in Table 10.25. The ignition probabilities for releases on the sea are consistent with the DNV Report (2001). For releases on land (tank farm releases), the ignition probability has been taken from a recent study conducted in the UK on the Mere Tank Farm (kerosene).
Table
10.25 Ignition
Probabilities
|
Hazard
Source |
Cause |
Leak
Size |
Ignition
Probability |
|
Tank Farm* |
All Causes |
Rupture (Bund Fire) |
0.1 |
|
|
|
Rupture (Pool fire from bund
overtopping) |
0.1 |
|
Submarine Pipeline |
All |
Rupture |
0.01 |
|
Marine Transport |
Grounding/Collisions |
Small |
0.02 |
|
|
|
Large |
0.03 |
|
|
|
Rupture |
0.05 |
|
|
|
|
|
|
Marine Jetty |
Loading Arm Failure |
Rupture |
0.05 |
|
|
Striking/Impact |
Small |
0.02 |
|
|
|
Large |
0.03 |
|
|
|
Rupture |
0.05 |
Note:
See discussion in Section 10.6.2.16 on the applicability of the ignition
probability for tank farm events to the PAFF site.
10.5.9
Frequencies of Outcome
Scenarios
10.5.9.1 The frequencies of outcome scenarios are presented in Table 10.26 (scenarios initiating from collisions, striking, grounding and explosion). These values are for the 2016 case. The frequency per year values presented in the tables has been calculated from base frequencies described earlier.
Table 10.26 Outcome Scenarios for
Collision/Striking/Grounding/Impact and Explosion (2016)
10.5.9.2 For the 2040 case, the frequencies of outcome scenarios are presented in Table 10.27 (scenarios initiating from collisions, striking, grounding, impact and explosion).
Table 10.27 Outcome Scenarios for
Collision/Striking/Grounding/Impact and Explosion (2040)
10.5.9.3 The pool fire frequency following a rupture of the loading arm is estimated as 1.28 x 10-6 per year (2016 Case) and 1.69 x 10-6 per year (2040 Case).
10.6.1.1 For this study, the thermal radiation level used as impact criteria is 37.5 kW/m2.
10.6.1.2 In pool fires, this level is reached close to the edge of the pool (i.e. 2 to 3m from the edge of small pools and much shorter distances from large pools – this is due to wind drag effects which affects small pools greater than larger pools). Thus the lethal area is approximately the same as the pool area. A lethal area of radius r+3 is considered where r is the radius of the pool fire.
10.6.2
Tank Farm Event
Consequences
10.6.2.1 The main consequences arising from spill in the tank farm are:
¨ tank fire;
¨ bund fire; and
¨ overtopping of a bund following catastrophic tank failure or escalation from bund fire.
Tank
Fires
10.6.2.2 Tank fires are essentially treated as pool fires, the difference being that the liquid surface and hence the base of the flame is elevated and is surrounded by a metal wall. Hence, at a given distance, the thermal radiation will tend to be lower than for a pool fire at ground level.
10.6.2.3 The effect distance for tank fires will be limited to the tank radius (20m) plus 3m (ie 23m from the centre of the tank), which is confined to the site boundary. Tank fires therefore do not pose any offsite risk and is hence not considered further in the assessment, although escalation of tank fires is discussed in the later paragraphs.
Bund
Fires
10.6.2.4 Bund fires may occur following the catastrophic release of liquid from a tank, overfilling or piping failure and subsequent ignition of the material. In this case, full surface bund fires are considered where the flammable material covers the surface of the whole bund. Smaller bund fires may occur due to small leaks into the bund, but can be controlled quickly.
10.6.2.5 The effect distance for a bund fire will be limited to 3m from the bund wall, which is confined to the site boundary. Escalation of bund fires is discussed in the later paragraphs. .
Fires
Following Overtopping of Bund
10.6.2.6 It has been hypothesised that material may escape over the bund wall following the initial surge of the released material (following catastrophic tank failure). A rough estimate given by Trbojevic and Slater (20) is that 30% of liquid would escape over a square dike following catastrophic failure of a 30 m diameter tank with liquid height 10.5 m and bund height of 1.5m under the assumption that the integrity of the dike is preserved. Wilkinson’s (4) work suggests that 50% of the material would be lost over the bund wall, and cites different cases where the bund may not be able to contain the liquid. The only case that is relevant here is the momentum of surging fluid sufficient to overtop the bund. Wilkinson cites some examples where material has overtopped the bund wall, but there are also examples in MHIDAS where the bund has retained all of the liquid following a catastrophic failure.
10.6.2.7 Analytical model and experiments carried out by Greenspan et al (28,29) provide a correlation for the fraction of the original liquid volume, which spills over as a function of the ratio h/H, where h is the height of the bund and H is the original height of the liquid in the tank with the bund r radius. This correlation provides a value of up to 50% for bund overtopping.
10.6.2.8 The UK HSE has recently published a report on a series of experiments involving tank ruptures to determine the spreading of liquid (30). The experiments were based on a scale of 1:20, representing a 70m diameter storage tank. Water was released through a slot at the base. The size of the slot was such that the tank emptied in about thirty seconds. It was found that if a large release occurs around the base of the tank, fluid is capable of overtopping the bund walls, even if the bunded area has the capacity to hold the fluid. The percentage overtopping reported in the experiments vary from 1% (case 1) to 11% (case 2) depending on the fill height and bund configuration (case 1: 4.43m distance from tank center to bund wall, bund height of 200mm and fill height of 1.45m; case 2: 8.89m distance from tank center to bund wall, bund height of 50mm and fill height of 1.75m). The lower figures in the HSE study appear to result from a larger bund radius, which may be applicable for single tank bunds.
10.6.2.9 The bund (which will hold 6 tanks) at the PAFF facility will be designed to provide a total volume of 110% of the largest tank capacity. The drain from the bund is normally kept closed and opened only during rainfall. In the event of a spill within bund or traces of oily water, it is routed to the oily water interceptor. For the 2040 case, the bunds will be interconnected to provide containment for at least 3 tank volumes.
10.6.2.10 A cross section of the tank, the bund wall and the boundary fence is shown in Figure 10.1. The tank height is 32m and its diameter is 40m. The distance from the nearest tank shell to bund wall is 10m. The height of the proposed bund wall is 4.6m with respect to the bund floor. The site roads around the bund wall (which form the general site area) is raised to about 2.6m with respect to the bund floor, ie the bund wall is not free standing but will act as a retaining wall. Therefore potential for failure of the bund wall due to momentum surge will be limited. Furthermore, a security wall (of breeze block type) of 2m high from road level is provided at 8m from the bund wall, which will act as a secondary containment in the event of overtopping of the bund. The roads around the tank bund will be provided with storm water drains, which will collect any liquid overtopping the bund. There is further separation beyond the security wall to the boundary fence, which will be planted with trees for minimising visual impact from the site. A drainage ditch with a sloping catchment will be provided in the 4m strip between the security wall and the security fence to trap any liquid splashed over the security wall and the gate. This ditch will be designed to handle 39m3 of liquid over a 100m length and connected to the stormwater drains, which discharge to the sea. Also, the security gate will be provided with a 1m ramp as well as a leak tight seal at the bottom of the gate up to the first hinge to contain any spill within the site.
10.6.2.11 As discussed in section 10.5.2.8, the failure mode assumed for the PAFF facility is a rupture of the top plate of the tank following a fire, resulting in spillage of 10% of tank contents (ie 3,900 m3). This would fall vertically into the bund and form a pool inside the bund of depth 0.5 to 1.2m depending on the distance from the tank to the bund wall and splashing outside the bund. The bund wall is about 4.6m high and therefore the bund will contain bulk of the spillage as per the design intention. However, vertical momentum from the release of 3,900 m3 at a height of 23 to 32m could result in splashing of some liquid over the bund wall (of about 10% of the liquid spilt) similar to that from a waterfall.
10.6.2.12 The splashed liquid (390m3) would approximately spread over a 100m length of the site inner road (of width 8m). The depth of this pool is estimated as about 0.5m which will be contained by the 2m high security wall and the 1m ramp provided up to the security gate.
10.6.2.13 Most of the fluid would enter the stormwater drain provided on site for the PAFF. The storm water drain consists of a 750mm diameter pipe and is designed to discharge more than 1000 m3/hr. Therefore most of the liquid (assumed as 90% of liquid splashed over the bund, ie 350m3) will be drained to the sea through the storm water drains provided for the inner road.
10.6.2.14 Some portion of the liquid splashed over the bund (about 10%, ie 39m3) on to the inner road may further splash over the security wall and the gate over a length of about 100m along the security wall. A drainage ditch with a sloping catchment is provided in the 4m strip between the security wall and the security fence to trap any liquid splashed over the security wall and the gate. This ditch will be designed to handle 39m3 of liquid over a 100m length and connected by storm water drains to discharge to the sea.
10.6.2.15 Based on the above, it can be seen that the structural integrity of the bund, security wall and the security gate will not be compromised since the quantity splashing over the bund is limited to 390m3(due to the water fall effect rather than a tidal wave). Furthermore, this splashed liquid will drain into the sea through the storm water drains on the site inner road and through the gulley along the security wall. Due to these specific drain arrangements on site, the liquid spill will be contained within the site boundary and not extend off-site as to affect public outside the facility.
10.6.2.16 Ignition of the spill could result in a pool fire. Ignition of spill could occur within the bund, on the site inner road, in the area between the security wall and the security fence or on the sea where the liquid splashed over the bund will be discharged. There are no ignition sources within the bund except for hot work which may be carried out during maintenance and lightning. Hot work will be governed by stringent procedures, which will include immediate termination of hot work in the event of any spill. Lightning is more likely to affect tank top than ground level releases. Therefore the likelihood of ignition within the bund is expected to be very low. Ignition of spill on the site inner road may occur due to transfer pumps, pipeline pigging operations, emergency or maintenance vehicle movements and street lighting (although hazardous area classification will be adopted for electrical equipment and fittings). Based on these considerations, the ignition probability of 0.1 (given in Table 10.25) is perhaps more appropriate for ignition of the splashing liquid on the site inner road rather than for the spill within the bund. An ignition probability of 0.1 for spill within the bund is highly pessimistic. Ignition of 39m3 liquid splashed over the security wall/gate is unlikely because the quantity of spill is very low, will permeate rapidly into the gravel in the ditch between the security fences, drain way to the sea and there are no potential ignition sources between the security wall and the security fence. This corridor is provided for security purposes only and therefore will not be accessed normally.
10.6.2.17 Based on the above considerations, the only conceivable scenario is pool fire on the site inner road. The 350m3 liquid splash on the site inner road is modelled as a pool fire over a length of 100m. However this will run down the drain to the sea and may become extinguished or re-ignite. The effect distance of R+3 will be limited to within the boundary fence and therefore no off-site fatality due to liquid overtopping the bund would result. However onsite personnel or firemen could potentially be affected by this fire but they will be trained using pre-set emergency procedures to respond to such incidents.
10.6.2.18 If the liquid overtopping the bund does not find an ignition source onsite, it will drain to the sea. The liquid (total of 390m3) draining to the sea through the storm water drain on the site inner road and the drain gulley along the security wall is modelled as a continuous release on sea over a duration of 20 minutes. The stormwater drain discharge will typically be released 1m below the sea surface and any fuel would have to move to the surface and then be ignited in order for a pool fire to occur on sea. In the event of a pool fire on sea, the effect distance is estimated as about 42m using the approach described in Section 10.6.3.3. Based on a maximum marine population density of 0.15 persons per hectare (see Table 10.3), this event gives 0.085 fatalities. The approach adopted for representing fatalities less than one on the FN curve is to multiply the value 0.085 with the scenario outcome frequency to derive the equivalent frequency for 1 fatality. It may be noted that this is the only outcome scenario (for the event overtopping of the bund), which has an impact offsite (although on the sea).
Bund Fire Escalation
10.6.2.19
The impact of bund
fire following tank failure or tank/piping leak and the potential for
escalation is discussed in the following paragraphs.
10.6.2.20
The Fire Services
can reach the sitequickly, as the Pillar Point Fire Station is located close to
the site. The graded response time of 6 minutes for the area under normal
traffic conditions should be achievable.
The Fire Services will take immediate measures to extinguish the fire
using foam. Foam pourers and foam tanks are also provided on site. Also, in the
event of any tank fire or bund fire, the on-site and off-site emergency plan
will be activated which will include evacuation of people in the neighbouring
sites as well as mobilisation of additional resources including foam within
Hong Kong.
10.6.2.21
As seen from Table 10.2, about 1,500
people would need to be evacuated. It is assumed that the Police would reach
the site within 10 minutes and take immediate measures to cordon-off the access
roads and evacuate the neighbouring sites. The Marine Police will also be
informed at the same time as the Marine Rescue and Fire Fighting services. They
may take about 10 minutes to reach the site. The marine unit of the Fire
Services will augment the fire fighting capability on site. Also, the Marine
Police will cordon-off the sea lanes adjoining the site. Therefore within one
hour, all the neighbouring areas will have been evacuated and the sea lanes
adjoining the site will be cordoned-off and a full scale fire fighting
operation will be underway.
10.6.2.22
In the case of
bund fires involving kerosene/aviation fuel (ie large diameter pool fires
involving heavy hydrocarbons), the clear flame height is estimated as few
metres only (31). Furthermore the surface emissive power for
kerosene fires is estimated as 10 to 20 kW/m2, which is much lower
than smaller pool fires involving lighter hydrocarbons (about 100 to 150 KW/m2)
or jet fires (about 300kW/m2). Therefore only the liquid filled
portion of the tanks will be exposed to flame engulfment. This would result in an increase in the
liquid temperature of the tank contents and corresponding increase in the evaporation
rate. . The tank wall thickness will be greater at the bottom than at
the top due to the hydrostatic load as well as the wind load. Furthermore, API
650(32) tanks are designed to withstand fire engulfment and more
importantly to relieve vapours in the event of fire by failing along the roof
to shell connection. Therefore failure due to flame engulfment is expected only
at the top and this may lead to a tank fire due to ignition by smoke particles.
Assuming a burning rate of 0.04mm/s (see Section 10.6.3.2), the time taken to
burn a liquid height of 10cm is about 40 minutes. Therefore the tank on fire
will continue to burn for several hours (for nearly empty tank) or days (for a
full tank) but failure of the tank shell would not occur.
10.6.2.23
A review of
incidents show that tank failures following bund fires have occurred mostly in
the case of crude oil tanks due to boilover. The incident review also shows no fatalities off-site
and very few cases of fatalities involving firemen although in all major tank
farm incidents, several tanks have been damaged or burnt and the fire has taken
several hours and even days to extinguish completely.
10.6.2.24
Based on the above
considerations, it can be concluded that aviation fuel tanks exposed to pool
fires (in the bund) following an aviation fuel spill would not fail, instead
burn from the tank top for several hours or days. The effect distance for tank
fire and bund fire events are confined within the site boundary. Therefore the
potential for off-site fatality due to bund fire escalation does not occur.
Tank-to-Tank Escalation
10.6.2.25
A tank fire has
the potential to impact an adjacent tank, which may result in an adjacent tank
fire. However, the storage tanks
have a cooling system installed and the adjacent tanks (in the sector opposite
the tank on fire) will be cooled in the event of a tank fire.
10.6.2.26 Previous studies (21) on tank fire modelling and escalation have been carried out and some general observations regarding the escalation potential for non-volatile fuels (such as aviation fuel) have been made. These studies suggest that a relatively long response time is available for tank escalation. The radiation level on the adjacent tanks is approximately 13-15kW/m2 given the tank separation of 15m and may not be sufficient to threaten the structural integrity of the adjacent tanks when cooled by water. The studies also suggest that escalation potential is greater for smaller diameter tanks (around 10m) and less in the case of larger diameter tanks, as will be the case for PAFF. Based on this discussion, it is concluded that the potential of escalation of a tank fire to an adjacent tank is minimal and is not considered further in the assessment. Emergency responders would however, need to be aware of this potential (ie tank to tank fire escalation) and which therefore should be suitably addressed in the onsite and offsite emergency plan for the PAFF facility.
Smoke Effects
10.6.2.27 The combustion products of aviation fuel include carbon dioxide, nitrogen oxides and sulphur oxides. Incomplete combustion will generate thick black smoke and hazardous gases including carbon monoxide. In the case of fire involving heavier hydrocarbons such as aviation fuel and in case of large diameter tank/bund fires, smoke production is very high
10.6.2.28 The occupants of any high-rise buildings in the vicinity of the tank farm could be exposed to smoke effects following a tank or a bund fire at the tank farm. The occupants could be incapacitated due to the combined effects of CO2 (causing hyperventilation) and CO (toxic narcosis).
10.6.2.29 The composition of smoke plume of heavy hydrocarbons is estimated as about 11.8% CO2 and 800ppm of CO. At 800 ppm, the time required for incapacitation is about 48 seconds and at 300ppm, the time required is 20 minutes. These times are estimated for persons caught within the smoke plume. For persons away from the fire, the effects will be limited due to the smoke plume rise.
10.6.2.30 There are no high-rise buildings currently in the vicinity of the tank farm. It is assumed that any future buildings immediately opposite the site boundary will not be high rise to avoid the impact of any smoke ingress into buildings. The on-site and off-site evacuation plans should consider the potential for hot smoke ingress into surrounding buildings.
Summary
10.6.2.31 Table 10.28 gives the frequency and consequence of events at the tank farm that can cause fatalities offsite.
Table 10.28 Frequency
and Consequence for Tank Events
|
Year |
Scenario
Frequency (per year) |
Effect
Distance (m) |
Probability
of Death |
Scenario
Description |
|
2016 |
3.96 x 10-7 |
42 |
1 |
Pool Fire on Sea from bund overtopping
following catastrophic tank failure |
|
2040 |
7.92 x 10-7 |
42 |
1 |
Pool Fire on Sea from bund overtopping
following catastrophic tank failure |
10.6.3
Consequences of Marine
Transport Events
Spill
Sizes
10.6.3.1 Spill size assessment is based on the DNV 2000 Study (7). The analysis is based on probability distributions of damage length, penetration and location in 296 actual collisions combined with theoretical calculations of cargo outflow. A review has been carried out in the above study and a simplified representation of spill size for the collection and storage barges has been adopted. The resulting spill sizes modelled are given below and the spill quantities are given in Table 10.29.
¨ Small Leak - 0.3% of dwt
¨ Large Leak - 1% of dwt
¨ Rupture - 7% of dwt
Table
10.29 Spill Quantities
from Marine Transport Releases
|
Tanker
Size (dwt) |
Size
of Leak |
Spilt
Quantity (tonnes) |
|
20,000 |
Small |
60 |
|
|
Large |
200 |
|
|
Rupture |
1400 |
|
30,000 |
Small |
90 |
|
|
Large |
300 |
|
|
Rupture |
2100 |
|
45,000 |
Small |
135 |
|
|
Large |
450 |
|
|
Rupture |
3150 |
|
60,000 |
Small |
180 |
|
|
Large |
600 |
|
|
Rupture |
4200 |
|
80,000 |
Small |
240 |
|
|
Large |
800 |
|
|
Rupture |
5600 |
Pool Fire Size
10.6.3.2 When the pool of aviation fuel is ignited, the fire spreads rapidly across the full extent of the pool and proceeds to burn through its thickness at its burning velocity. For aviation fuel this velocity is 0.04mm/s (0.053kg/m2s). The fire this has the same extent as the pool itself.
10.6.3.3 For an unconfined continuous release (as on sea), the pool grows until equilibrium is reached where burning at the surface balances out the release rate. The pool diameter is then given by:
where: D = pool diameter (m)
Q = release rate (kg/s)
b = burning rate (kg/m2s)
10.6.3.4 For an instantaneous release, the pool grows steadily until it reaches a minimum thickness of approximately 10mm, but once ignited its thickness is reduced by the fire at a typical rate of 0.04mm/s, until it is burned out, usually within a few minutes. The pool diameter may be expressed in terms of the average thickness as:
where: D = pool diameter (m)
M = release mass (tonnes)
t = average pool thickness (m
r = density (tonnes/m3)
10.6.3.5 The pool fires are modelled as unconfined pool fires on the sea surface. The modelling assumptions are given in Table 10.30.
Table
10.30 Pool Fire
Modelling Assumptions
|
Size
of Release |
20,000
dwt |
30,000
dwt |
45,000
dwt |
60,000
dwt |
80,000
dwt |
|
Small Leak |
50kg/s for 20 minutes continuous
release |
75kg/s for 20 minutes continuous
release |
112.5kg/s for 20 minutes continuous
release |
150kg/s for 20 minutes continuous
release |
200 kg/s for 20 minutes continuous
release |
|
Large Leak |
167kg/s for 20 minutes continuous
release |
250kg/s for 20 minutes continuous
release |
375kg/s for 20 minutes continuous
release |
500kg/s for 20 minutes continuous
release |
667kg/s for 20 minutes continuous
release |
|
Rupture |
1400 tonnes instantaneous |
2100 tonnes instantaneous |
3150 tonnes instantaneous |
4200 tonnes instantaneous |
5600 tonnes instantaneous |
Pool
Fire Impact Criteria
10.6.3.6 The thermal impact criterion for releases on sea is assumed as 37.5 kW/m2, which is consistent with DNV (2001) Study. The effect distance (r+3) where r is the radius for pool fires for the various release cases listed above are presented in Table 10.31.
Table
10.31 Effect Distances
for Sea Surface Pool Fires
|
Size
of Release |
Effect
Distance (m) |
|
||||
|
20,000
dwt |
30,000
dwt |
45,000
dwt |
60,000
dwt |
80,000
dwt |
Probability
of Death |
|
|
Small Leak |
20.3 |
24.2 |
29.0 |
33 |
37.6 |
1 |
|
Large Leak |
34.6 |
41.8 |
50.46 |
57.8 |
66.3 |
1 |
|
Rupture |
239 |
292 |
357 |
411.8 |
475 |
1 |
Marine Tanker Explosions
10.6.3.7 The explosion scenarios on marine tankers normally consist of the combustion of a mixture of air and hydrocarbon inside a nominally empty tank. The scenarios followed in the study are taken from the DNV 2000(7) Report.
10.6.3.8 The consequence distances for two explosion scenarios have been adopted for the sake of consistency. These are given in Table 10.32. These values have been adopted for all tanker sizes.
Table 10.32 Effect
Distances for Explosion Scenarios
|
Scenario |
Effect
Distance (m) |
Probability
of Death |
|
Explosion – Repairable Damage |
50 |
0.3 |
|
Explosion – Fragments |
500 |
3 x 10-5 |
10.6.4.1 The main consequences of the jetty events result from:
¨ striking of vessel when berthed;
¨ impact of incoming vessel against the dock wall; and
¨ rupture of the loading arm.
10.6.4.2 The consequences of striking and impact of vessels are exactly the same as that of collisions and groundings and are hence not discussed further.
10.6.4.3 The maximum pumping rate through the loading arm is 3500 m3/hr. Should a rupture occur, it is estimated that a maximum of 20 minutes will be taken to isolate the leak. This is extremely pessimistic as typically isolation times are much lower. This gives a spill quantity of approximately 934 tonnes. All the fuel will spill on to the sea and following ignition will result in a pool fire much in the same fashion as the cases for marine transport.
10.6.4.4 The effect distance of a spill of 934 tonnes has been modelled based on a continuous release of 778 kg/s for 20 minutes. The radius of the pool is calculated to be 68.37m. Therefore the effect distance is 68.37 + 3 = 71.37m (r+3).
10.6.4.5 Due to the presence of sea wall at about 200m from the jetty, the actual effect distance for spills on sea (with effect distance greater than 200m) will not in the form of circular pools, rather spread along the sea wall in irregular shapes. A simplistic semi circular pool of equivalent diameter has been assumed to model the effect of spreading due to sea wall.
10.6.4.6 For marine events that result in less than one fatality, the approach adopted for representing such events on the FN curve is to consider this value as a fatality probability and therefore multiplied with the outcome scenario frequency to derive the equivalent outcome frequency for 1 fatality. The PLL calculations are not affected by this approach.
10.6.5
Submarine Releases of
Aviation Fuel
Release
Rate and Inventory
10.6.5.1 The submarine pipeline is operating at a pressure of approximately 15 barg and has a water table of 25m above it. In the event of a rupture, it is assumed that a response time of 3 minutes will be required to affect a shutdown. Hence, following an initial release, the pressure will quickly fall and consequently the release rate from the pipeline will drop. Upon achieving shutdown, the relaxation volume of the fuel in the pipe (which is approximately 1-2% of the pipe inventory) will be released. The water head (approximately 2 bar), sandfill and rock armour is likely to prevent any further release from the pipeline although the spill may continue at a very low level due to diffusion.
10.6.5.2 Therefore, following an initial high release (about 430kg/s), the release rate will drop quickly and the residual inventory of the pipe will be released against a 2 barg water head.
Subsurface
Plume
10.6.5.3 Various models simulating the behaviour of a subsea release of oil (or other petroleum products) have been proposed (22). However, these models are difficult to verify against actual field data. Experimental studies have also been conducted to study the behaviour of underwater plumes.
10.6.5.4 The twin submarine pipeline system will be laid in a trench to at least 3m below the seabed and covered with rock armour protection. In order to provide further protection against future proposed CLP coal vessel access in the Urmston Road, the pipeline will be at a depth of up to 6-7m in this location. A release of aviation fuel from the submarine pipeline will initially be driven by momentum close to the release point. At some distance, the plume is expected to be driven by buoyancy of the fuel droplets within the plume. Thus, the plume will contain the seawater entrained into the plume as well as the fuel droplets. Due to the stratification (the vertical variation of temperature and salinity) of the seawater, the entrained seawater is likely to be trapped below the warmer and less salty water masses closer to the surface. When the velocity of the vertical motion of plume drops below the velocity of the fuel droplets, the droplets will tend to leave the subsurface plume. From that point onwards, the plume is expected to consist of oil droplets rising through the water column on an individual basis.
10.6.5.5 Due to the depth of the pipeline and rock armour protection, the initial momentum of the fuel release will be diminished. The fuel is likely to seep/percolate through the sand backfill/rock armour and will lose all its momentum in the process. Thereafter, the fuel will be rise under its own buoyancy. As described above, the entrainment of water into the fuel droplets will create a water fuel emulsion, which will eventually reach the sea surface. However, it is expected that due to the weathering and tidal motions of the sea, by the time the fuel reaches the sea surface, it will not remain as one large pool. Rather, the fuel would have broken up into a number of small pools. The thickness of these pools is also likely to be very small (<< 10mm).
10.6.5.6 Due to the entrainment of water in the fuel droplets, the amount of fuel that can vaporise to form a flammable mixture just above the pool will also be very limited.
Ignition
of Surface Pool
10.6.5.7 A significant amount of work has been carried out on the amount of heat required to ignite an oil spill on sea surface (23) (24). While most of these studies are concerned with in situ burning as an effective means for removal of an oil slick to reduce negative environmental impact, they, nevertheless, provide valuable insight into the amount of heat required to ignite an oil spill.
10.6.5.8 Most accidental and deliberate burns of spilled oil on the sea surface suffer from the effect of wind and waves. Volatile components tend to evaporate rapidly with time (weathering), and mixing tends to form oil-water emulsions making the oil difficult to ignite. Emulsion burning was shown to be very sensitive to external heat flux and it was found to be impossible to burn an emulsion below a certain threshold. Consequently, alteration of the physical or chemical properties of the oil can require additional energy for ignition. Several studies have attempted to characterize weathering and emulsion of typical of oil-spill scenarios (25). Also, the ignition of an oil spill must be followed by flame spread, which would eventually result in mass burning (sea surface pool fire). Many studies have also shown that ignition is not always followed by spread (26) and therefore is not sufficient to guarantee a pool fire.
10.6.5.9 All the studies have been carried out on crude oils whose flash point ranges from 190C to 250C whereas the flash point for aviation fuel is 380C. As the sea surface temperature is around 200C, the likelihood of ignition is low except in the case where it is externally heated. Based on the argument presented above, the likelihood of the aviation fuel igniting following a release from the submarine pipeline is extremely low.
10.6.5.10 However, as a conservative assumption an ignition probability of 0.01 is assumed for a pool of aviation fuel on the sea surface following pipeline rupture. As discussed above, the initial release rate of 430kg/s will drop quickly to equal to the pumping rate of the fuel, which is 3500m3/hr (2800 tonnes/hr). Assuming an isolation time of 30 minutes, as a conservative assumption, this will result in a release of 1400 tonnes of fuel (Actual isolation times would be much lower due to the provision of leak detection system). This scenario is modelled as equivalent to a rupture of 20,000 dwt tanker, which releases the same quantity of fuel. Table 10.33 gives the details of the outcome scenario of release event from submarine pipeline.
Table 10.33 Frequency
and Consequence for Submarine Pipeline Event
|
Frequency
per year of Pool Fire following Pipeline Rupture |
Effect
Distance (m) |
Probability
of Death |
|
9.12 x 10-9 |
239 |
1 |
10.7.1.1 Risk Summation involves combining the estimates of the consequences of an event with the event probabilities to give an estimate of the resulting frequency of varying levels of harm.
10.7.1.2 Risk Summation has been carried out using ERM’s in-house software called RISKPLOTTM, which is amongst the most advanced currently available
10.7.1.3 RISKPLOTTM takes into account all of the above information and models each event under each combination of conditions (weather, location, etc.). It calculates the number of fatalities from each event with a given probability of occurrence. The number of fatalities is based upon the proportion of each population area overlapped by the hazard effect, taking into account protection factors. Unlike the grid system, which assumes all occupants of a given grid sector suffer the same degree of impact, RISKPLOTTM assumes only that proportion of the population with a polygonal area equivalent to that covered by the hazard effect are affected, with an accuracy level to the nearest meter. Hence, the modelling exercise is as realistic as reasonably possible.
10.7.1.4 RISKPLOTTM is used to sum risks from the events at the tank farm. For the marine events (jetty and marine transport), an Excel spreadsheet is used to calculate the risks.
Societal
Risk
10.7.2.1 Societal risk is defined as risk to a group of people due to all hazards arising from a hazardous installation or activity. The simplest measure of societal risk is the Rate of Death or Potential Loss of Life (PLL), which is the predicted equivalent fatality per year.
10.7.2.2 The frequency (f) and fatalities (N) associated with each outcome event are derived. Based on this, the Potential Loss of Life is calculated as follows:
PLL = f1N1
+ f2N2 + f3N3 + ……. + fnNn
10.7.2.3 Societal risk is also expressed in the form of an F-N curve, which represents the cumulative frequency (F) of all event outcomes leading to N or more fatalities. This representation of societal risk highlights the potential for accidents involving large numbers of fatalities.
Individual
Risk
10.7.2.4 Individual risk may be defined as the frequency of fatality per individual per year due to the realisation of specified hazards. Individual risk from a pipeline is represented in terms of risk transects which provides risk estimates at different locations from the pipeline for any cross-section of the pipeline. The risk transects provide risks to a hypothetical individual present at a location 100% of time.
10.7.3
Risk Results and Discussion
Potential
Loss of Life
10.7.3.1 The Potential Loss of Life (PLL) values are given in Tables 10.34 and 10.35.
Table
10.34 PLL Values (2016
Case)
|
Case |
PLL
value per year |
Equivalent
Fatality |
|
Tank
Farm |
3.36 x 10-8 |
1 fatality in 27,961,904 years |
|
Submarine
Pipeline |
2.12 x 10-8 |
1 fatality in 47,169,811 years |
|
Marine
Transport & Jetty |
1.59 x 10-5 |
1 fatality in 62,893 years |
|
Total |
1.59 x 10-5 |
1 fatality in 62,893 years |
Table 10.35 PLL
Values (2040 Case)
|
Case |
PLL
value per year |
Equivalent
Fatality |
|
Tank
Farm |
6.73 x 10-8 |
1 fatality in 14,880,952 years |
|
Submarine
Pipeline |
2.12 x 10-8 |
1 fatality in 47,169,811 years |
|
Marine
Transport & Jetty |
2.18 x 10-5 |
1 fatality in 45,871 years |
|
Total |
2.18 x 10-5 |
1 fatality in 45,871 years |
FN
Curves
10.7.3.2 The FN curves for the events at the tank farm, submarine pipeline, marine jetty and marine transport are presented in Figure 10.2 (2016 case) and Figure 10.3 (2040 case). The FN Curve for the tank farm and the marine events (marine transport and marine jetty events) lie in the acceptable region for the 2016 case. The FN curve for tank farm events for the 2040 case also lies in the acceptable region. However, the FN curve for marine and jetty events for the 2040 case lies partially in the ALARP region and therefore cost effective risk mitigation measures must be identified to reduce the risk to As Low As Reasonably Practicable (refer to section 10.7.5 and 10.8). The FN curve for submarine pipeline lies entirely in the acceptable region. A summary of all risk calculations is presented in the Annex.
Individual
Risk
10.7.3.3 The maximum individual risk for the tank farm, submarine pipeline and marine events are given in Tables 10.36 and 10.37 for 2016 and 2040 respectively.
Table
10.36 Individual Risk
Results (2016 Case)
|
Case |
Maximum
Individual Risk per year |
|
Tank
Farm |
2.56 x 10-6 (onsite) |
|
Submarine
Pipeline |
4.56 x 10-8 |
|
Marine
Transport & Jetty |
7.19 x 10-6 |
Table 10.37 Individual
Risk Results (2040 Case)
|
Case |
Maximum
Individual Risk per year |
|
Tank
Farm |
4.36 x 10-6 (onsite) |
|
Submarine
Pipeline |
4.56 x 10-8 |
|
Marine
Transport & Jetty |
7.24 x 10-6 |
10.7.3.4 It can be seen that the maximum individual risk from the tank farm, submarine pipeline and marine transport and jetty is less than 1 x 10-5 per year and is therefore acceptable as per the criteria set out in Annex 4 of the EIAO-TM. The individual risk contours for the tank farm and marine transport/ jetty events (for 2016 and 2040 case) are shown in Figures 10.4 and 10.5. The individual risk contour for the submarine pipeline events is shown in Figure 10.6.
10.7.4
Maximum Justifiable Spend
10.7.4.1 In order to identify cost effective risk mitigation measures, the maximum justifiable spend is calculated. This is done as follows:
Maximum Justifiable Spend = PLL per year x Value of Life x Aversion Factor x Operating Life of the Facility
10.7.4.2 The Value of Life is assumed to be HK$ 33 million, which is the value used in similar studies in Hong Kong. The aversion factor is taken to be 1 as the PLL value is close to the acceptable region. An operating life of 36 years is assumed (2004-2040)
Maximum Spend = 2.18 x 10-5 x 33 x 106 x 1 x 36 = HK$25,898
10.7.4.3 Any additional safety design features beyond those already considered in the facility is likely to cost more than the maximum justifiable spend. Therefore recommendations are made on the basis of best practice.
10.7.5.1 The following recommendations are made in line with best practices:
¨ The marine jetty event is dominated by impact, ie caused by the approaching vessel striking the jetty resulting in spill and fire. A number of measures are already proposed in the design- fenders designed for impact loads, use of tugs, use of pilots aboard every vessel, restriction on maximum velocity for approach etc Further measures to minimise the risks from impact events should be examined. This includes the use of a berthing aid system. Under this system, two radar sensors located on the jetty would provide continuous information (ships position relative to the jetty, speed of ship and angle of ship related to berthing line) about the ships. Such advanced berthing aid systems are known to reduce the likelihood of berthing impact incidents.
¨ The stormwater drainage system for the PAFF site should be provided with a valve to contain bund overspills on site when required. The operational procedures for storm water drainage should be prepared in the case of any spill or fire incident at the tank farm.
¨ The access road to the site should be limited to providing access to the site and not used as general parking space by public vehicles. Access to this road may therefore be limited to personnel entering the PAFF site.
¨ The onsite and offsite Emergency Plans for PAFF should be developed and tested on a regular basis. Offsite emergency plans including evacuation plans and communication arrangements should be developed in conjunction with the Fire Services Department (FSD), Police, Marine Department and other agencies. Offsite emergency plans for the neighbouring sites will be prepared in order to have an effective evacuation within a short period of time. There will be submitted by the project proponent during detail design of the facility.
¨
The off-site emergency plan should
include procedures for the Police including the Marine Police, including
cordoning-off the access roads, evacuating the
neighbouring sites, and cordoning-off the sea lanes adjoining the site.
¨
The onsite and
off-site emergency plans
should consider tank to tank fire escalation, bund fire escalation and smoke
effects from fires in developing suitable emergency response measures.
¨
The operating procedures for unloading
fuel from tankers at the jetty and for tank farm operations should include
procedures in the event of thunderstorm warning, typhoon and lightning. Onsite emergency
procedures should include actions to be taken in the event of ignition of vents
due to lightning.
¨ Since the tank farm will be constructed in phases, suitable measures should be adopted for ignition control, restricting access to operating areas and during tie-in with operating facilities. The following control measures will be adopted:
- The tank farm will be designed as per relevant local and international codes and standards.
- An independent high level trip will be provided to actuate shutdown of unloading operations in the event of high level in the tank.
- The proposed bund configuration along with site layout and boundary wall arrangements is shown in Figure 10.1. The main features of this arrangement as described below will be adopted in the design. The distance from the nearest tank shell to bund wall will be atleast 10m. The height of the proposed bund wall will be 4.6m with respect to the bund floor. The site roads around the bund wall (which form the general site area) will be raised to about 2.6m with respect to the bund floor. Furthermore, a security wall (of breeze block type) of 2m high from road level will be provided at 8m from the bund wall, which will act as a secondary containment in the event of overtopping of the bund. The roads around the tank bund will be provided with storm water drains, which will collect any liquid overtopping the bund. Also, the security gate will be provided with a 1m ramp as well as a leak tight seal at the bottom of the gate up to the first hinge to contain any spill within the site.
- A drainage ditch with a sloping catchment will be provided in the 4m strip between the security wall and the security fence to trap any liquid splashed over the security wall and the gate. This ditch will be designed to handle 39m3 of liquid over a 100m length and connected to the stormwater drains, which discharge to the sea.
-
- Any changes to the above site parameters during detailed design that will adversely affect the consequence modelling will be submitted to EPD for approval.
- The tanks will be protected by water spray for cooling of the tank shell on fire as well as adjoining tanks. Provision will also be made for injecting foam to the base of the tank.
- Firewater pumps on site will be provided to provide firewater for cooling as well for foam injection. Fire hydrants and foam monitors will also be provided on site.
- Water curtain will be provided at the jetty to protect loading arms and the jetty head. Foam monitors will also be provided at the jetty head.
- Leak detection system will be provided for the delivery pipeline.
- Shutdown system will be provided for the tank farm facility to actuate shutdown of unloading operations as well as delivery.
- Stringent quality control measures will be adopted during construction including material procurement, welding and testing of tanks.
- Maintenance procedures during operations will include monitoring of corrosion and settlement.
10.8.1 The frequency of impacts (7.4 x 10-5 per visit) has been derived based on world-wide historical data. Based on the recommendations made above for reducing berthing impacts and the adoption of berthing aid systems, it is estimated that the frequency of berthing impacts will be lower by 50% i.e. 3.7 x 10-5 per visit. Based on this frequency, the revised risk results are calculated. The mitigated FN Curve for the 2040 and 2016 case (for marine transport, jetty events and total FN Curve) is presented in Figure 10.7 and 10.8 respectively. The revised total FN curve also lies completely in the acceptable region. The mitigated PLL value for the 2016 case is calculated as 1.0 x 10-5 per year and for the 2040 case is 1.37 x 10-5 per year.
10.9.1 A quantitative risk assessment has been carried out on the proposed PAFF at Tuen Mun Area 38. The assessment covered hazards from:
¨ tank farm;
¨ submarine pipeline;
¨ marine transport (within 500m of the jetty); and
¨ marine jetty.
10.9.2 The main hazards associated with the tank farm are tank fires, bund fires and unconfined pool fires following bund overtopping. The hazard distance from tank fires and bund fires do not extend beyond the site boundary. Hazard distances for fires following bund overtopping (following catastrophic rupture of tanks) extend offsite onto the sea due to the discharge of splashed liquid through the stormwater drain.
10.9.3 The maximum individual risk on land from events on the tank farm is calculated as 2.56 x 10-6 (2016 case) per year and 4.36 x 10-6 (2040 case) per year. The maximum individual risk contour lies within the site boundary and is zero offsite. For events at the tank farm that have potential consequences on the sea (due to discharge through the storm water drains and subsequent ignition), the maximum individual risk is calculated as 3.96 x 10-7 per year (2016 case) and 7.92 x 10-7 per year (2040 case), which is also lower than 1 x 10-5 and therefore acceptable when compared with the criteria set out in Annex 4 of the EIAO-TM. The individual and societal risk curve for tank farm events for both 2016 and 2040 cases are acceptable according to the criteria set out in Annex 4 of the EIAO-TM.
10.9.4 Risks from the submarine pipeline are found to be extremely low. This is mainly because of the behaviour of aviation fuel under water. Also, the submarine pipeline will be buried to around 3m below seabed along most of the route (in certain parts it will be buried to 6-7m) and covered with sandfill and rock armour protection. A subsea release of the fuel will be initially be driven by momentum and then buoyancy and will slowly travel upward to the sea surface. However, the fuel will emulsify with the seawater thereby increasing it flashpoint further and consequently making it extremely difficulty to ignite. The FN curve for submarine pipeline events lie entirely in the acceptable region. The PLL value for submarine pipeline is calculated as 2.12 x 10-8 per year.
10.9.5 The main hazards associated with marine transport and jetty operations are collision, strikings, groundings, impact and rupture of the loading arm. All these events could result in a spill on the sea surface. Sea surface fires have been modelled and the risks estimated. The maximum individual risk is below 1 x 10-5 per year and is therefore acceptable. The PLL value is 1.59 x 10-5 per year (2016 case) and 2.18 x 10-5 per year (2040 case). The societal risk curve lies completely in the acceptable region for 2016 case. For 2040 case, the FN Curve for the marine and jetty events lies partially in the ALARP region. Maximum justifiable spend was calculated as HK$ 25,898, which is not sufficient to effect any design or operational changes. Nevertheless, recommendations based on best practices have been made. Based on these recommendations, the mitigated risk results were calculated and these were found to lie entirely in the acceptable region of the Hong Kong Risk Guidelines. The mitigated PLL value for the 2016 case is calculated as 1.0 x 10-5 per year and for the 2040 case is 1.37 x 10-5 per year.
10.9.6 The total FN curve (tank farm, submarine pipeline and marine events) for the 2016 case and 2040 case (mitigated case) lie completely in the acceptable region.
10.9.7 Based on the analysis presented in this section, including the recommendations, it can be concluded that the offsite individual and societal risks posed by the activities at the PAFF tank farm and the associated marine environment are acceptable according to the criteria set out in Annex 4 of the EIAO-TM.
10.10.1 The hazard assessment has predicted that the risks from the operation of the PAFF and any associated hazards to life are acceptable based upon the recommendations summarised above which will be integrated into the design and operation of the facility. Therefore, no adverse residual impacts are predicted.
10.11
Environmental Monitoring and Audit
10.11.1 The risks from the operation of the PAFF have been shown to be acceptable based upon the integration of various detection, control and containment measures into the design of the facility. Based upon this, while construction and operational phase EM&A is not recommended over and above the regular programme of inspections that will be specified in the response plan, design phase audit of the spill response plan to ensure it includes the necessary elements and of the design of the pipelines, tanks and jetty to ensure key spill detection and control equipment are included is recommended. Further details are provided in Section 15 of this report and in the EM&A Manual.
1) Fax from BMT Asia Pacific, 24 January 2002
2)
BP Oil UK Limited, Safety Data Sheet,
BP Kerosene, Premium and Standard Grades.
3)
http://www.hse.gov.uk/infoserv/mhidas.htm
4)
Wilkinson, A, “Bund Overtopping - the
consequences following catastrophic failure of large volume liquid storage
vessels”, AEA Technology, SRD/HSE R530, p.56, 1991
5) Hong Kong Electric, QRA of Proposed Subsea Gas Pipeline and associated Launcher/Receiving Stations, ERM, 2000
6) ACDS, Major Hazard Aspects of Transport of Dangerous Substances, HSC, UK, 1991
7) Airport Authority Hong Kong. QRA of Aviation Fuel Transport to Hong Kong International Airport, DNV, 2000
8)
Davies, T., Harding, A.B., McKay,
I.P., Robinson, R.G.J., Wilkinson, A., “Bund Effectiveness in Preventing
Escalation of Tank Farm Fires”, Trans IChemE, Vol 74, Part B, May 1996.
9) Batstone, R.J., Tomi, D.T., “Hazard Analysis in Planning Industrial Developments”, Loss Prevention, 13, 7, 1980.
10) Rijnmond Public Authority, “A Risk Analysis of Six Potentially Hazardous Industrial Objects in the Rijnmond Area - A Pilot Study”, COVO, D. Reidel Publishing Co., Dordrecht, 1982.
11)
Taylor, J.R., “Risk Analysis for
Process Plant, Pipelines and Transport”, London: Spon, 1994.
12)
E&P Forum, “E&P Forum QRA
Datasheet Directory”, 15 October, 1996.
13)
Prokop J, The Ashland Tank Collapse,
Hydrocarbon Processing, 67(5), 105, May 1988
14)
Christensen, R.A. and Eilbert, R.F.,
“Aboveground Storage Tank Survey”, EL RN-623, Entropy Limited, Lincoln, MA,
1989.
15)
Lees, F.P., “Loss Prevention in the
Process Industries”, 2nd Edition, Butterworth-Heinemann, Oxford, UK, 1996,
p.A14/11.
16)
PARLOC 96,The Update of Loss of
Containment Data for Offshore Pipelines, 1996
17) ERM Hong Kong, Caltex Safety Case, 1996
18) CIRRUS, Consequence Modelling Software Package, Version 6.1, BP Research and Engineering Centre, 2000.
19)
A.D. Little Inc., “Experiments
Involving Pool and Vapour Fires from Spills of LNG on Water”, Report
C6-D-55-79, Cambridge,
MA, 1979
20)
Trbojevic, V.M., Slater,
D.H., “Tank Failure Modes and Their Consequences”, Plant/Operations Progress, 8,
84, 1989.
21) Purdy et al, Tank Fire Escalation – Modelling and Mitigation, Proceedings of 7th International Symposium on Loss Prevention and Safety Promotion in Process Industries, Italy 4-8 May 1992.
22) Rye, H and Brandvik, PJ, Verification of Subsurface Oil Spill Models, Proceedings of International Oil Spill Conference, 1997 pp 551-557.
23) Putorti, A.D, Evans, D and Tennyson, E.J, Ignition of Weathered and Emulsified Oils, Proceedings of 17th Arctic and Marine Oil Spill Program (AMOP) Technical Seminar, Volume 1, pp 657-667
24) Neil Wu, Gilles Kolb and Jose Torero, Piloted Ignition of a Slick of Oil on a Water Sublayer: The Effect of Weathering, 27th Symposium on Combustion/The Combustion Institute, 1998, pp 2783-2790
25) Bobra, M. "A Study of the Evaporation of Petroleum Oils", Publication EE-135, Environmental Canada, Ottawa, Ontario, K1A 0H3, 1992
26) Kasiwagi, T et al, Fourth International Microgravity Combustion Workshop, NASA Lewis Research Centre, 1997, pp 411-416
27) A W Cox, F P Lees and M L Ang (1990), Classification of Hazardous Locations, Institution of Chemical Engineers.
28) Greenspan H P and Young R E (1978) J. Fluid Mechanics, 87 : 179
29) Greenspan H P and Johansson A V (1981) Studies in Applied Mathematics, 64 : 211
30)
HSE (2002) A series of experiments to study
the spreading of liquid pools with different bund arrangements, CRR 405/2002.
31) HSE (1996) Development of Pool Fire Thermal Radiation Model, HSE CRR No. 96/1996
32) API 2000, Venting Atmospheric and Low Pressure Storage Tanks: Refrigerated and Non-Refrigerated, 5th Edition, 1998