PART A Assessment
for Potentially Hazardous Installations (PHIs)
14A.1
The
HATS Stage 2A, Figure 14A.1, is to collect the sewage from 8
upgraded Preliminary Treatment
Works (PTW) at the northern and south-western
parts of Hong Kong Island and to convey,
through over 20 km of deep sewage tunnel system, to the expanded Stonecutters Island Sewage Treatment Works (SCISTW) for centralized treatment. According to the EIA Study Brief No. ESB-129/2005 [1] Clause 3.4.8.2, Hazard to Life Assessment (HA) is required to be carried out since the
proposed works sites at Aberdeen and Ap Lei Chau are located within the
consultation zones of two respective Potential Hazardous Installations (PHIs),
namely the Hong Kong & China Gas Company’s Gas Holder Depot (PHI No. H4)
(hereinafter referred to as “HKCG Depot”) and the Shell LPG Transit Depot/Bulk
Domestic Supply (PHI No. H5) (hereinafter referred to as “Shell Depot”).
14A.2
The
scope of the HA is focused on addressing the following issues, together with
any other key issues identified during the course of the EIA Study.
·
Potential
hazard impacts on two proposed works sites at
·
Potential
impacts on the existing PHIs, namely the HKCG Depot (PHI No. H4) near Aberdeen
PTW and the Shell Depot (PHI No. H5) near Ap Lei Chau PTW arising from the
construction of any drop shaft, trench or underground tunnel, etc that has the
propensity to cause ground subsidence (off-site accident initiator) leading to
catastrophic failure of vessels, equipment or installations within the PHIs
(on-site accident).
14A.3
According
to the technical requirements specified in Section 3.4.8 of the EIA Study Brief
[1], the HA is carried out following the criteria for evaluating hazard to life
as stated in Annexes 4 and 22 of the TM [2] (Hong Kong Risk Guidelines).
14A.4
The
HA includes the followings:
·
assess
the risks associated with all aspects of the construction and operation of the
Project;
·
assess
the hazards associated with the storage and use of any other dangerous goods or
hazardous activities present on site;
·
ascertain
whether the overall risks posed by the activities of the Project are
acceptable, taking into account the risk guidelines set out in section 4.4 of
Chapter 12 of the Hong Kong Planning Standards & Guidelines;
·
recommend
mitigation measures where the risk is considered in the ALARP (As Low As
Reasonably Practicable) region or above, and to quantify the reduction in risk
achievable by these means; and
·
recommend
measures, including relocation of works area(s), that would prevent accidental
damage to the adjacent PHIs and their associated pipelines during construction
and operation of the Project.
·
Baseline Case – Assessment of baseline case risk
level based on current PHI operations, population and existing PTW operation
conditions. If information from previous risk assessment of the two PHIs were
available, technical parameters from previous studies would be adopted in
assessment of the baseline case; otherwise best available information will be
adopted to construct the baseline case model. Results from this baseline case
would be benchmarked with selected previous studies wherever they are
available.
·
Construction Stage Case – Assessment of risk level during
construction of the Project taking into account the peak construction workforce
level and adjacent population at that time.
·
Operational Stage Case – Assessment of risk level during
operational stage of the Project taking into account the number of operators in
PTW and adjacent population at that time.
14A.7
The
HA consists of five major steps given below:
·
Study Base: identifies details of PHIs.
Meteorological data/ignition sources and population data are collected for
consequence modelling and risk summation respectively. Topographical and
architectural features are identified for any shielding effect to population.
·
Hazard Identification: identifies hazards associated with
the PHIs and the construction and operation activities. HAZOP workshops have
been conducted in June 2007. The potential hazards identified and recorded in
the Hazard Registers will be used as an input for scenarios development for
analysis in the next stages.
·
Frequency Assessment: assess the likelihood of occurrence
of the identified hazardous scenarios by reviewing historical accident data or
using Fault Tree Analysis. Event Tree Analysis will be adopted to determine the
possible outcome from the identified hazardous events and to estimate the
frequencies.
·
Consequence Assessment: the consequences will be
established for every outcome developed from initial event by using DNV’s SAFETI Professional, Version 6.51
software to assess the impacts from gas leaks, fires, explosions, toxicity and
other process hazards.
·
Risk Assessment: evaluates the risks level, in terms
of individual risk and societal risk, associated with the identified hazardous
scenarios. The overall risk level will be compared with the criteria as
stipulated in Annex 4 of the TM to determine their acceptability. Mitigation
measures will be identified where the risk is considered in the ALARP (As Low As Reasonably Practicable) region
or above. The reduction in risk achievable by these means will then be
quantified.
·
Recommendation of Safety Measures: Upon completion of the risk
assessment, safety measures will be identified to reduce risk of accident
damage to the adjacent PHIs and their associated pipelines during construction
and operation of the Project. In addition, measures for reducing impacts to
construction workers due to accident in PHIs will also be provided.
·
Individual Risk: the maximum level of off-site
individual risk should not exceed 1 in 100,000 per year, i.e. 1 x 10-5
/ year.
·
Societal Risk: it can be presented graphically as
in Figure 14A.2. The Societal
Risk Guideline is expressed in terms of lines plotting the frequency (F) of N
or more fatalities in the population from accidents at the facility of concern.
14A.10
The
plant details including storage of flammable inventories, plant facilities and
operation details of the HKCG Depot and the Shell Depot are collected from the
corresponding PHI owners. This information is reviewed and used to identify
potential hazardous scenarios associated with the PHIs. In case that data is
not fully available, engineering judgement and/or assumptions are made. These
assumptions and judgement are clearly documented in the Appendix 14A.7 of this
study report.
14A.11
Future
development, if exists, of the PHIs is considered to ensure using the latest
data for risk calculation. In additional to physical layout of the PHIs,
operation details, such as LPG Road Tanker transfer schedules, are consolidated
for determination of hazardous scenarios.
14A.12
The
HKCG Depot adjoins the Aberdeen PTW at the east boundary of the depot.
According to the 2007 PHI Register (as at 31st January 2007) issued by the
Housing, Planning, and Lands Bureau, the Gas Holder has a storage capacity of
17.7 tonnes of Towngas, which is above the threshold limit of 15 tonnes for
classification as a PHI. The designated consultation zone (CZ) of 150m from the
gasholder is applied to the facility, as shown in Figure 14A.3.
14A.13
The
HKCG Depot consists of a gas holder and a plant building in which the governor,
the boosters and engines are installed.
14A.14
The
gas holder is served as a temporary storage to cope with network requirement
and peak shaving in winter. There are approximately 48 operations in a year for
the gas holder. In each operation, the gas holder is loaded up to approximately
80% of the total capacity for maximum 24 hours. Normal inventory level is kept
to 3 tonnes when the gas holder is not in operation. The gas pressure is
reduced from medium pressure (MP) through spill-over governor and stored in the
gas holder under low pressure (LP) via a 20m in length inlet gas pipe.
Out-flowing gas is fed to booster room from the gas holder via a 20m outlet gas
pipe and is pressurised to MP for distribution. Both inlet and outlet gas pipe
are 600mm in diameter. There is an overhead section at 5m aboveground near the
depot entrance. Layout plan of the HKCG depot is depicted in Figure 14A.4.
14A.15
Town
gas is mainly composed of hydrogen, methane, carbon dioxide and carbon
monoxide. Physical properties of the mixture are listed in Table 14A.1 and Table 14A.2.
Table 14A.1 Composition of Towngas
Material |
Composition (%
volume) |
Hydrogen |
48.1 |
Methane |
29.4 |
Carbon dioxide |
19.5 |
Carbon monoxide |
3 |
Table 14A.2 Physical Properties and Storage Conditions of Towngas
Properties |
Values |
Physical state |
Gaseous (pressurised) |
Storage pressure |
7.5 – 240 kPa (MP) 2 – 7.5 kPa (LP) |
Storage temperature |
Ambient |
Average molecular mass |
15.1 kg/kmole |
Specific heat ratio |
1.344 |
Calorific value |
17.27 MJ/m3 |
Lower Flammability Limit |
5.5% |
14A.16
The
Shell Depot (PHI No. H5) adjoins the Ap Lei Chau PTW at the south boundary.
According to information provided by the depot operator, the depot is composed
of 2 sections, namely the storage & transit section (approximately 60m away
from PTW site) for LPG products and the LPG compound for supply of LPG to
consumers at
14A.17
The
LPG compound, Figure 14A.7, adjoins the
northern boundary of the ALC PTW site. The existing facility includes 2 units
of 20-ton mounded LPG vessels and 7 units of vaporizers (5 units are hot water
type and the rest are coil-heated type). Under normal operation, the 2 LPG
vessels are filled up to no more than 85% of the total capacity (34 tons).
Daily consumption is approximately 7 tons. Minimum stock of 28 tons is
maintained for supply of 4 days in case of interrupt to delivery. Replenishment
frequency is about 6 times a week. Loading process takes 2 hours during daytime
only. Gaseous LPG is delivered to South Horizons through a 150mm gas pipeline.
The beginning section of the pipeline is above ground running along the west
boundary of the depot for about 60m. The rest is buried underground and runs
across the depot towards and along the east boundary.
14A.18
Within
the storage & transit section, there is a LPG cylinder shed at the north of
the depot, Figure 14A.6. Cylinders of capacities
ranging from 2kg to 49kg are stored onsite with average inventory of 2,000
cylinders. The depot operator also loads/unloads gas cylinders to/from LPG
cylinder wagons of local gas distributors at LPG storage platform for
consumption in Hong Kong Island South. The LPG storage platform is located at
waterfront near entrance of the depot and is for temporary handling of LPG
cylinders only. LPG cylinders are delivered to the store via ferry service.
Apart from LPG cylinders, the ro-ro ferry pier is a landing place for LPG road
tankers (9 Tons) and LPG cylinder wagons for supply of LPG in
Table 14A.3 Daily Utilisation of Ferry Service Between 1st and 27th Sept 2007
Vehicle Type |
Arrival |
Departure |
LPG Road
Tanker |
2 (loaded) |
2 (empty) |
LPG
Cylinder Wagon |
2 (loaded) |
2 (empty) |
14A.19
Population
Data, which indicate the presence and locations of people in surrounding areas,
are collected in order to evaluate the human impact of a hazardous release from
the PHIs. The manning level of the construction works and the number of
operators working in the upgraded PTW are also assessed.
14A.20
Data
from the Population Census are consolidated to determine the existing
population in vicinity of the PHIs. For the population at locations not covered
by the Population Census, a site visit was conducted to estimate the population
at various locations based on observation on-site and communication with
on-site personnel.
14A.21
Transient
population, such as vehicle traffic, are consolidated from the Annual Traffic
Census of Transport Department. The number of people travelling is derived
according to traffic flow rate and traffic mix at different time modes (e.g.
day, night, peak, weekend, weekday, etc).
Traffic mix includes passenger car, light and heavy goods vehicle and
bus.
14A.22
For
future stationary and transient population, the growth rates of the population are
estimated based on the Planning Department’s “Projection of Population
Distribution 2006-2015” for the assessment years 2009 and 2014 and data
provided by Planning Department. These growth rates are applied to the 2006
residential population consolidated to obtain the population in assessment
years.
14A.23
In
order to reflect temporal distribution of population, time period is divided
into 4 time modes namely daytime (07:00–19:00) and night-time (19:00–07:00) for
both weekend and weekday. In general, assumption of temporal variation in
population for different population categories is tabulated in Table 14A.4.
Table 14A.4 Temporal Changes in Population for Various Categories
Time period |
Residential Dwellings |
Shopping Centre |
Industrial/Commercial Buildings |
Weekday
(day) |
25% |
50% |
100% |
Weekday
(night) |
100% |
0% |
10% |
Weekend
(day) |
70% |
100% |
40% |
Weekend
(night) |
100% |
0% |
5% |
14A.24
These
figures have been applied to various hazard assessments such as Comprehensive
Feasibility Study for the Revised Scheme of South East Kowloon Development
(SEKD) [6] in which building types are similar to those in the
14A.25
Population
indoor is well protected by the building structure and therefore impacts to
indoor populations are relatively lower than that of the outdoor populations.
Therefore, escape and shielding factors are applied to the population data to
ensure a realistic assessment. While population is usually protected by
building or population site, indoor ratio of 95% is assumed for buildings [3].
14A.26
There
is a seafront, Kai Lung Wan, to the south of the HKCG Depot. A gradual hill
slope can be found to the north of the gas holder. Buildings sit on the same
ground level as the HKCG Depot at the east and the west direction. Shielding
effect of topographical features is not considered as significant. Populations
in the following locations adjacent to the Aberdeen PTW Site are considered in
this study.
·
The
·
Concrete
batching plant at 70m west of the PTW
·
Tin
Wan
·
Shek
Pai Wan Road
·
Wah
Kwai Estate and associated social facilities/ recreational facilities
·
·
Hing
Wai Ice & Cold Storage
·
Hing
Wai Centre
·
Open
area car park opposite to the HKCG Depot
·
14A.27
The
numbers of personnel currently present on the PTW site are up to 22 during
daytime and 6 during night-time. Due to recent advances in plant
instrumentation and control, it is assumed that the present staffing level
required to operate the PTW would be relatively lower than that in late 1970s
when the PTW was first commissioned. Furthermore, the number of operation
personnel on the PTW site will be reduced to 18 during construction phase.
After completion of construction (from 2014 onwards), the number of operation
personnel on the PTW site will remain the same as at present.
14A.28
During
construction stage, i.e. from 2009 to 2014, the upgrading works of the PTW
mainly include demolition/reconstruction of existing structure in small scale,
low-rise buildings and installation of electrical and mechanical equipment.
Yet, this would have to be conducted over a relatively long period of 3.5
years, as the activities must be done in phases to minimize interruption to PTW
operation. It is estimated that the total number of HATS construction workers
for the PTW works would be 60 in which 40 is for PTW upgrading works, 14 is for
aboveground SCS works and 6 is for underground SCS works during construction
stage. However, PTW works and SCS tunnelling works will be carried out in
separate construction phases. Thus, the maximum number of construction workers
at any time will be 40 from PTW upgrading works. And it is assumed that no
construction worker would work overnight.
14A.29
Consultation
with the current operator of the concrete plant has revealed that the number of
personnel currently present on site is about 49 (comprising 12 plant staff and
37 concrete truck drivers), corresponding to an average production of 25,000m3
of concrete per month. Previously, during the peak periods before 2004 when the
plant ran at 47,000m3 per month, there was as many as 80 staff (20
plant staff and 60 drivers) on site. To take a conservative approach, maximum
population of the concrete plant in the assessment is 80.
14A.30
The
Consultant was informed that the current lease for the concrete plant will
expire in 2008. It is unclear if the concrete plant will continue to operate
after 2008. According to circulation between Planning Department and other
relevant authorities, strong local requests for closing down the plant have
been received since 1990’s and no in-principle objection to changing the use of
the site even without a replacement site for the plant has been received from
relevant authorities, Appendix 14A.3.
Although Commission of Tourism has proposed a conceptual design of using the
site for coach parking, no further detailed information is available at this
stage. Therefore, it is assumed that the concrete plant will still be operating
during the construction phase of the PTW and onwards for conservativeness in
this assessment.
14A.31
Summary
on population data being adopted in the assessment is tabulated in Table 14A.5.
Table 14A.5 Population Data for the
Population Group |
Existing |
Construction Phase |
Operation Phase |
Wah Kwai
Estate (1) |
11,284 |
10,842 |
10,569 |
Wah Kwai Community
Centre (3) |
60 |
58 |
56 |
Wah Kwai
Vocational and Re-habilitation Centre (3) |
150 |
144 |
140 |
Wah Kwai
Shopping Centre (2) |
500 |
480 |
468 |
Wah Kwai
Bus Terminus (3) |
100 |
96 |
94 |
Wah Kwai
Open Area Car Park (3) |
8 |
8 |
8 |
Wah Kwai
Outdoor Playgrounds (at 4 locations) (3) |
80 |
80 |
80 |
Wah Kwai
Sport Courts (at 3 locations) (3) |
60 |
60 |
60 |
Wah Kwai
Sitting-out Areas (at 2 locations) (3) |
40 |
40 |
40 |
|
5,264 |
5,058 |
4,931 |
Ka Lung Court
Outdoor Playgrounds (at 1 location) (3) |
20 |
20 |
20 |
Ka Lung
Court Sport Courts (at 2 locations) (3) |
20 |
20 |
20 |
|
8 |
8 |
8 |
Dairy
Farm Ice & Cold Storage (4) |
60 |
60 |
60 |
Hing Wai
Ice & Cold Storage (3) |
60 |
60 |
60 |
Hing Wai
Centre (3) |
600 |
600 |
600 |
|
50 |
50 |
50 |
Shek Pai
Wan Playground (3) |
20 |
20 |
20 |
Shek Pai
Wan Road (3,5) |
92 |
111 |
94 |
Tin Wan |
13 |
13 |
13 |
Concrete Plant
(3) |
80 |
80 |
80 |
|
22 |
18 |
22 |
|
0 |
40 |
0 |
Sources:
(1)
HK Census/ Planning
Department
(2)
Based on data available from
The Link
(3)
Site/ Telephone survey
(4)
Estimated from Hing Wai Ice
& Cold Storage
(5)
Based on Traffic Census
–100% and 25% of
(6)
Data from DSD
(7)
Project consultants;
consider only aboveground workers at the PTW work area
14A.32
There
is a steep slope at east of the Shell Depot along
14A.33
At
the top of the slope, Ap Lei Chau No. 2 Fresh Water Service Reservoir is located
approximately 100m east of the LPG depot at 80m elevated height. Besides, it is
outside the maximum fireball diameter obtained from consequence modelling
(PHAST model) and is far beyond the LPG cloud height. Workers at the service
reservoir are protected by the topographical feature and are not further
considered in the assessment.
14A.34
Although
there are more than 30 residential buildings together with the
14A.35
Having
taken into account the topographical features, populations in the following
locations adjacent to the Ap Lei Chau PTW Site are considered in this study.
·
The
Ap Lei Chau PTW
·
South
Horizons
·
·
·
Material
store for Highway Department to the south of the PTW
·
Dah
Chong Hong (industrial building)
·
14A.36
The
numbers of personnel currently present on the PTW site are up to 4 during
daytime and 2 during night-time. The number of personnel will be reduced to 3
and 2 at daytime and night-time respectively during construction phase. After
completion of construction (from 2014 onwards), the number of operation personnel
on the PTW site will remain the same as at present.
14A.37
During
the HATS Stage 2A peak construction periods, around 50 workers with
approximately 10% at indoor and 15 (11 for aboveground work and 4 for
underground work) construction personnel related to PTW upgrading and tunnel
construction (including spoil removal truck) respectively will be present in
daytime at any one time. It is assumed that no workers will work at night-time.
14A.38
Summary
on population data being adopted in the assessment is tabulated in Table 14A.6.
Table 14A.6 Population Data for the Ap Lei Chau Project
Population
Group
|
Existing
|
Construction
Phase
(Year
2009)
|
Operation
Phase
(Year
2014)
|
South Horizons Blocks 7 to 33A
(1)
|
26,819
|
27,252
|
27,800
|
|
3,000
|
3,180
|
3,480
|
Dah Chong Hong (6)
|
750
|
795
|
870
|
Driving School (2)
|
50
|
50
|
50
|
Highway Department Material Store
(2)
|
2
|
2
|
2
|
|
16
|
16
|
16
|
Ap Lei Chau PTW – DSD Operators
(4)
|
4
|
3
|
4
|
Ap Lei Chau PTW Construction Site
(5)
|
0
|
61
|
0
|
Sources:
(1)
HK Census/ Planning
Department
(2)
Site/ Telephone survey
(3)
Based on Traffic Census –
20% of
(4)
Data from DSD
(5)
Project consultants; consider
only aboveground workers at the PTW work area
(6)
Estimation based on floor
area
14A.39
Meteorological
conditions affect the consequence of gas release in particular the wind direction,
speed and stability which influences the direction and degree of turbulence of
gas in the dispersion process. Meteorological data for year 2006 from Wong Chuk
Hang weather station of the Hong Kong Observatory are collected and adopted in
the consequence model to determine the various gas dispersion, fire and
explosion effect. 6 sets of weather class (combination of wind speed-stability
class) for both daytime and night-time are identified in accordance with [7]
and adopted in the risk assessment. Summary on meteorological data analysis is
tabulated in Table 14A.7.
Table 14A.7 Meteorological Data Analysis
Day |
||||||
Wind Direction |
Weather Class |
|||||
B 3.1 |
D 1.7 |
D 3.8 |
D 7.8 |
E 2.9 |
F 1.4 |
|
0 |
0.0160 |
0.0062 |
0.0052 |
0.0002 |
0.0027 |
0.0107 |
30 |
0.0170 |
0.0045 |
0.0090 |
0.0000 |
0.0035 |
0.0067 |
60 |
0.0292 |
0.0055 |
0.0102 |
0.0000 |
0.0062 |
0.0110 |
90 |
0.0959 |
0.0090 |
0.0400 |
0.0092 |
0.0092 |
0.0155 |
120 |
0.1815 |
0.0200 |
0.0498 |
0.0155 |
0.0110 |
0.0217 |
150 |
0.1146 |
0.0157 |
0.0085 |
0.0007 |
0.0015 |
0.0150 |
180 |
0.0152 |
0.0057 |
0.0010 |
0.0000 |
0.0007 |
0.0055 |
210 |
0.0100 |
0.0042 |
0.0000 |
0.0000 |
0.0000 |
0.0047 |
240 |
0.0355 |
0.0057 |
0.0010 |
0.0000 |
0.0000 |
0.0075 |
270 |
0.0280 |
0.0052 |
0.0010 |
0.0000 |
0.0002 |
0.0055 |
300 |
0.0212 |
0.0032 |
0.0035 |
0.0000 |
0.0015 |
0.0052 |
330 |
0.0245 |
0.0035 |
0.0082 |
0.0005 |
0.0035 |
0.0105 |
|
||||||
Night |
||||||
Wind Direction |
Weather Class |
|||||
B 1.0 |
D 1.9 |
D 4.1 |
D 7.5 |
E 3.0 |
F 1.3 |
|
0 |
0.0000 |
0.0000 |
0.0058 |
0.0003 |
0.0053 |
0.0671 |
30 |
0.0000 |
0.0003 |
0.0178 |
0.0006 |
0.0122 |
0.0395 |
60 |
0.0000 |
0.0000 |
0.0131 |
0.0003 |
0.0145 |
0.0448 |
90 |
0.0000 |
0.0000 |
0.0913 |
0.0170 |
0.0465 |
0.0737 |
120 |
0.0000 |
0.0000 |
0.0810 |
0.0145 |
0.0493 |
0.0940 |
150 |
0.0000 |
0.0000 |
0.0036 |
0.0006 |
0.0047 |
0.0604 |
180 |
0.0000 |
0.0000 |
0.0003 |
0.0000 |
0.0000 |
0.0365 |
210 |
0.0000 |
0.0000 |
0.0000 |
0.0000 |
0.0003 |
0.0351 |
240 |
0.0000 |
0.0003 |
0.0014 |
0.0000 |
0.0025 |
0.0398 |
270 |
0.0000 |
0.0000 |
0.0006 |
0.0000 |
0.0022 |
0.0323 |
300 |
0.0000 |
0.0000 |
0.0050 |
0.0003 |
0.0011 |
0.0259 |
330 |
0.0000 |
0.0000 |
0.0134 |
0.0017 |
0.0072 |
0.0359 |
14A.40
The
presence of ignition sources is a primary concern in case of inflammable gas
release. Ignition sources (other than onsite one), such as dwellings and vehicles
along carriageways, contribute to delayed ignition in VCE and flash fire. The
energy level, timing, location of ignition sources in the vicinity of the PHIs
and hence the probability of ignition of gas cloud is reviewed and assessed.
14A.41
Two
types of ignition source are defined in the risk model. Transportation
polylines are defined for roads. The presence time of a road is calculated from
its traffic density, average vehicle speed and total length of the road
segments. Traffic data is consistent with those for population calculation.
Ignition probability is taken as 0.5. Another ignition source, due to
activities of population such as cooking and using electrical appliances, is
assigned implicitly to all population groups by SAFETI.
Hazard Events at the
Existing HKCG Depot
14A.42
The
hazardous event of the HKCG Depot is gas leakage that may lead to fire,
explosion and toxicity. Gas leakage could be initiated by external events such as
damage of gas installations by third parties. Historical data sources, previous
risk assessment report [5] and MHIDAS database have been reviewed. Generic
hazardous events associated with the operation of the HKCG Depot are listed in Table 14A.8.
Table 14A.8 Hazards Associate with Operation of HKCG Depot
Hazard event |
Potential Cause |
Spontaneous
failure |
·
Gas holder failure ·
Pipework failure ·
Flange gasket failure ·
Valve leakage failure ·
Pump failure |
Partial
failure |
·
Pipework leakage ·
Blown seal |
External
event |
·
Earthquake ·
Car crash ·
Aircrafts crash ·
Ship collision at seafront ·
Landslide ·
Severe environmental events ·
Lightning Strike ·
Dropped Object ·
Subsidence ·
External Fire ·
Collapse and Strike by object ·
Vibration |
14A.43
The
gas holder at
14A.44
Under
additional strain, the gas holder may be distorted and may not operate
properly. Subsequently, alignment of guide rollers or guiding columns could be
changed. Crown movement can be hindered by structural deformation. Those
damages may develop into other failure modes (such as crack, distortion,
dislocation, fracture and destruction). Moreover, a jammed crown could lead to
tank overpressure and blown-seal when Towngas is pumping continuously into the
gas holder. Similarly, a jammed crown could lead to tank under-pressure and
collapse of roof when Towngas is drawing continuously from the gas holder.
14A.45
Damage
to the gas holder or its structure may lead to disruption of gas supply. In the
severe case, such damage may lead to gas leakage and even structural failure.
External interferences increase shear load and weaken welding lines and joints
as well as impose additional strains to gas holder sheeting.
14A.46
14A.47
Governors
controls and regulates pressures of gas inflow from the medium pressure network
to the gas holder and is sensitive to interferences. In case of minor accident,
interferences would disturb inflow of gas towards the gas holder. In case of
overpressure, inlet pipeline would be overloaded and lead to full bore rupture
or leak at the pipeline. In the worst scenario, it triggers failure of the gas
holder leading to lifting off the floating roof or blown seal. Damage to gas
governor may lead to large scale Towngas release from distribution network side
in short period of time. However, emergency isolation of gas supply for a
particular section of pipework can effectively stop the leak.
14A.48
Booster
pressurises gas stock and feeds gas stock from gas holder back to distribution
network. Damage to booster leads to interruption of gas supplied from the gas
holder. It would result in pressure drop in the distribution network when
demand from consumers is high.
14A.49
14A.50
Failures
of gaskets and valve leak would only tend to give relatively small scale of
leakage and will not contribute to the off-site risk. The results from gasket
failure will not be considered separately but absorbed into pipework failure in
the study.
14A.51 No unauthorised vehicle is allowed and speed restriction is imposed within the HKCG Depot. Besides, safety markings and marked protective fencing are provided to the above ground pipelines near the entrance and the gas holder. Based on statistical data for fatal traffic accident involving medium and heavy goods vehicles between year 2002 and year 2006, accident rate of 6.37x10-9 per km.year is calculated. Car crash leading to failure of gas facility is estimated 5.81x10-7 per year from fault tree analysis (Appendix 14A.5).
14A.52
In
14A.53
The
distance between the nearest arrival flight path and the Depot is more than 2
miles. The distance between the Depot and
14A.54
Ships
anchoring at Kai Lung Wan are mostly fishing boats and barges for the adjacent
concrete plant. Size of ships is rather small and low in profile and speed.
Besides, seafront is protected by a seawall. Ship collision at seafront leading
to gas facilities failure does not exist.
14A.55
The
HKCG Depot is at least 40m from toe of a slope of 20-35m height Slope No.
11SW-C/F 348. The slope is not classified as dangerous hillside slope and does
not require non-routine maintenance. There is no record of severe landslide for
the slope. The slope is well maintained, fully covered by vegetation and it is
separated by Tin Wan Praya and an open area car park.
14A.56
Loss
of containment due to severe environmental event such as typhoon or tsunami
(large scale tidal wave) is not possible as the Depot is designed to withstand
wind load for local typhoon while
14A.57
Subsidence
is usually slow in movement and such movement can be observed and remedial
action can be taken in time. Besides, the Depot has been built for more than 30
years. During construction phase, the increase of probability of subsidence is
derived from the EGIG pipeline failure database [10], taking the failure
frequency for “ground movement” and assuming it will always lead to a full bore
pipe rupture. After considering the length of pipe section (40 m) and duration
of tunnelling work (4 months of out 12 months in 1 year, equivalent to time
fraction 0.33), the frequency of subsidence leading to pipework failure at
Aberdeen PTW is 3.88x10-7 per year.
14A.58 External fire means the occurrence of fire event which lead to the failure of the gas holder or other facilities. The key potential concern relates to the gas holder and pipelines being affected by the on-site diesel tank. By considering location of the diesel tank, the tank is installed underground. Fire fighting equipment is also provided. Therefore, it can be assumed that such external fire will not lead to any disastrous outcome.
14A.59
The
frequency of a lightning strike on properly protected building is extremely low
in
14A.60
Production
process of the concrete plant right next to the HKCG Depot is carried out in
enclosed environment. Moreover, there is a concrete fence wall next to
14A.61
There
is no risk of damage to the gas depot by collapse and strike of surrounding
object in the existing setup. This factor is further elaborated in sections
regarding construction activities in the Aberdeen PTW.
Hazard Events Initiated by
14A.62
A
systematic brainstorming HAZOP workshop was conducted in June 2007 for the HKCG
Depot. Participants include the representatives from HKCG, risk specialist and
subject engineers. Details of hazards identified are given in the hazard
register attached in Appendix 14A.1
of this document. All these identified hazards are evaluated for development of
the risk model.
14A.63
Referring
to Paragraph 72 of judgment from the Court of Final Appeal (FACV 28/2005): “…it
is appreciated that a QRA, in order to satisfy the exigencies of Annex 4, must
be both generic and project-specific, that the methodology searches for the
relevant scenarios in the history of projects of the same genus- and thus
identifies scenarios for the purposes of para. (i) – then quantifies risk by
reference to that history and the specific features of the instant project …”.
A review of historical incidents has been conducted to identify hazardous scenarios and
frequencies of occurrence of these scenarios.
14A.64
Impact
on the HKCG Depot due to activities of construction and operations as well as
the DG storage of the Project is assessed. The following subsections summarised the
hazardous scenarios identified from the activities of the Project during
construction and operational stage.
14A.65
Amongst
all external events in Table 14A.8,
risk of occurrence for subsidence, external fire and dropped objects would be
increased by construction activities of the PTW works. These activities are
further developed to various hazardous scenarios in the following subsections.
14A.66
Review
on construction and operation activities have been conducted prior to the HAZOP
workshops held in June 2007. These activities and the hazards identified in the
workshops are recorded in Hazard Registers as attached in Appendix 14A.1 for the
14A.67
It
has been confirmed that construction activities will be conducted within the
150m CZ of the HKCG Depot (except that the production shaft of the tunnel is
outside the zone). Activities related to the Project in vicinity of the HKCG
Depot are listed below and are indicated on Figure 14A.8.
·
Upgrading of PTW - Relevant construction works will
last for about 3.5 years and will involve demolition of existing buildings and
construction of new buildings & structures, installation of electrical and
mechanical equipment, reconstruction of a section of the existing seawall as
well as laying of sewers, drainage pipes, and other utility lines. Shallow excavation (less than 5m) will be
required for
construction of the
treatment plant building, except for manholes/pits and the grit traps where
excavation at these locations could be down to depths of about 8m below ground
level. Sheet piling and non-percussive piling works will also be undertaken.
·
Construction of Drop Shaft and
Tunnels - A sewage
tunnel “P” of finished diameter 1,500mm (approximate excavated diameter to be 3,000mm) will be
constructed underneath
·
Reconstruction of Seawall – 29m of the existing seawall will
be demolished and reconstructed to cater for PTW upgrading works and a new seawater intake connecting to
the new seawater pumping house. Another 24m of seawall will be demolished and
reconstructed only for the PTW upgrading works. Extent of the affected seawall (Total 53m)
is indicated in Figure 14A.9.
·
Temporary Storage - Tunnel construction plants,
materials and equipment will be stored at a work area opposite to the HKCG
Depot on the other side of
14A.68
The
upgraded PTW, as shown in the layout plan Figure 14A.10, will be
operated in a similar fashion to the existing one although the treatment
capacity will be increased. An administration office building, a maintenance
workshop, and a spare parts store will be established in the upgraded PTW. It
will serve as DSD’s depot for Hong Kong Island South region. A DG store will be
set up as a depot to support regional need. Lubricants, paints, gas cylinders
for gas welding and bleach will be stored in this store.
14A.69
External
interferences affecting the normal operation of the HKCG may originate from the
following,
·
Excessive
vibration
·
Ground
movement/settlement
·
Strike
by other objects from the PTW construction area
·
Fire
impingement or high thermal radiation
14A.70
In
general, potential damage initiated by these hazards is listed below.
·
damage
of Gas Holder
·
damage
of Pipeline inside the HKCG Depot
·
damage
of other gas installation, such as booster, pump and governor inside the HKCG
Depot
14A.71
Hazards
with potential impact to the adjacent HKCG Depot have been identified during
the aforementioned HAZOP workshop. These hazards and the potential outcomes due
to construction works are categorised and tabulated in Table 14A.9. Detailed evaluation of the potential impacts is
followed.
Table 14A.9 Hazards with Potential Impact to HKCG Depot
Hazardous Scenario |
Damage Outcomes |
Construction Activities |
Potential Cause |
Hazard Log Ref. |
Ground
settlement inside Gas Holder Depot |
Ground
Collapse damage Gas Holder |
Excavation near Gas Holder
- PTW - excavation for Tunnel P - Construction of seawater pump house |
Unexpected drawing down of groundwater table |
1,
11 |
Ground
Collapse damage Gas Holder and/or Booster/Governor |
Tunnelling - Drop shaft construction (inside Aberdeen
PTW) by mechanical boring - Tunnel Q construction by horizontal
directional drilling (HDD) - Tunnel P construction by “drill and
blast” |
Unexpected drawing down of groundwater table
during tunnel/shaft construction |
8,
9, 12 |
|
Structural
Damage of Gas Holder and Pipework |
Gas
Holder foundation and water seal damage; |
Mobilising and usage of PTW construction or
tunnelling equipment. e.g. Crane operation
accidentally hitting gas holder |
Loss
of Stability or Mechanical |
5,
10 |
Tunnel Q construction by horizontal directional
drilling (HDD) |
Construction
of Tunnel Q strayed from design alignment |
9 |
||
PTW
Upgrading Piling Work and Demolition of existing
PTW structures; demolition and reconstruction of seawall |
Excessive
ground vibration |
2,
3, 7 |
||
Gas
Holder Wall and Guiding Frame Damage |
Production
shaft construction |
Fly
rocks due to drill and blast construction method |
13 |
|
Governor,
Booster and gas pipework Damage |
Tunnel P construction by “drill and blast” |
Ground vibration and air overpressure due to
blasting |
12 |
|
Demolition of existing PTW structures |
Excessive vibration |
3 |
||
Fire/
Explosion Hazard affect Gas Depot |
Gas
Holder damage due to missile of explosion |
Installation of electrical and mechanical
equipment |
Ignition of flammable material (DGs or
Construction material) due to Hot Works. Explosion due to fire escalation
inside construction site |
4/6 |
Gas
Holder damage due to missile of explosion |
Temporary storage of construction plant,
equipment, and materials in temporary works area opposite the Gas Holder/ PTW
site |
Ignition of flammable material (DGs or Construction
material) due to Hot Works. Explosion due to fire escalation inside
construction site |
14,16 |
14A.72
Drill
and blast method is to be adopted only in rock head level for the construction of
the production shaft which is located more than 150m away from the Gas Holder.
The soil depth at the production shaft would be in the range of 4 to 5 m. Two
boreholes will be carried out at the production shaft to obtain Site
Investigation (SI) information for detailed design purpose. During blasting
operation, protection against sideways projection of flyrock will be provided
by the shaft walls, and protection against vertical ejection through the top of
the shaft will be provided by a steel roof. A steel roof over the shaft is
required to ensure there is no risk of flyrock ejecting form the shaft. The
arrangement would be specific to the chosen method of working, and details
would be provided in the Contractors method statement for blasting and to be approved
by the Supervising Officer of the design and build contract and Mines Division
of CEDD. Therefore the risk of flyrock is minimized with proper measures.
Moreover, the throw distance of a flyrock using Terrock's Flyrock model [12] is
predicted to be less than 120m which is below the 200m separation distance
between the production shaft and the Gas Holder. Given standard precautionary
measures implemented by contractor, no serious concern to ground level is
envisaged regarding flyrocks. Therefore, no damage of gas holder will be caused
by flyrocks in blasting activities. Flyrock damage to gas holder is not further
assessed.
14A.73
Tunnel
P is about 2.5km long located within volcanic rocks. It has been characterised that volcanic
rocks in the area is relatively abrasive and hard, which is not favourable to
Tunnel Boring Machine (TBM) operation. In fact, some areas along Tunnel P are
associated with geological features such as
14A.74
Emulsion
based cartridge explosive will be deployed in construction of Tunnel P because
of the small charge-weight requirement. Delivery of detonator and explosive
will not be on the same truck. Besides, the emulsion cannot be detonated until
a gassing agent has been introduced. Explosive transport is via
14A.75
A
blasting assessment report for the project has been prepared under separate
cover. Calculation has been carried out for estimating the amount explosive to
be used within the CZ for construction of Tunnel P, Figure 14A.12.
The calculation is based on the assumption of 5mm/s PPV at the foundation of
the gas holder. This value is much lower than the proposed allowable limit for
gas pipelines, 25mm/s, stated in communications with HKCG, Appendix 14A.4. In the calculation,
maximum charge-weight per delay 0.79kg is estimated to be applied to the
closest section to the HKCG Depot at where the true distance is about 47.8m.
Although higher charge-weight per delay for construction the rest of Tunnel P
will be used and will pass the nearest section, emulsion and gassing agent will
not be mixed until the rock surface has been reached. Explosive will not be
accidentally detonated when it is transported inside the Tunnel P. Thus, the
HKCG Depot will not be subject to excessive vibration due to use of explosive.
14A.76
As
there will be 2 blasts per day maximum, explosive can be delivered on demand
and no overnight storage of explosive onsite is required.
14A.77
After
the information on the use of explosives in this project has been reviewed,
hazard to the HKCG Depot due to use and handling of explosive within the
project area is not an issue.
14A.78
Tunnel
Q will start from ground surface on
14A.79
Figure 14A.13 shows the interface diagram for
drop shaft construction and PTW upgrading work. In the diagram, work area for
HDD works is right next to the east boundary of the HKCG Depot. Mobile crane
will be used for assemble of construction equipment and drilling rig at ground
level within the HDD work area. Crane outreach height for assemble of drilling
rig is estimated between 8m and 12m. Besides, there is minimum 5m separation
distance between the work area west boundary and the gas holder. It should
provide sufficient buffer for HDD works.
14A.80
For
construction of PTW administration building and workshop, a tower crane may be
used. It is feasible that the tower crane is erected near the drop shaft
location. While the administration building would be 2 to 3 storey in height,
the height of the tower crane would not cause any threat to the gas holder with
the maximum crane height of 20m.
14A.81
The
erection, operation and dismantle of tower crane will follow all regulations,
requirements and code of practice for safe use of tower crane by Labour
Department. In addition, specific risk assessment will be conducted before the
erection and operation of the crane to ensure safety and stability of crane
operations. Control measures can be taken to ensure no damage to the gas holder
even in collapse accident of mobile crane. A crane may topple or collapse
towards the load carrying end. Such accidents can be avoided with a well
planned site layout and engineering mean. Control measures within the HDD work
area may impose the following restrictions,
·
use
of tall mobile crane.
·
lifting
height of a crane.
·
angular
movement and location of crane.
14A.82
Potential
ground settlement adjacent to the HKCG due to the construction of Tunnel P has
been estimated by the project geotechnical engineers as shown in Figure 14A.14. The
magnitude of potential settlement is around 8mm.
14A.83
The
gasholder is a piled structured which is founded on Grade III rock which will
not be affected by this magnitude of settlement. For pipeline, usually a higher
value of 25mm can be used for settlement analysis. Nevertheless, monitoring
would be carried out during tunnel construction to closely monitor any ground
movement. Instrumentation such as settlement marker can monitor the ground
settlement. Piezometer can be used to monitor the ground water table which may eventually
cause ground settlement.
14A.84
Alert
and action limits will be set such that construction works will be stopped when
vibration or ground settlement exceed the corresponding action limit.
Construction works will proceed further only if sources of settlement/vibration
have been identified and rectification of the problem has been completed.
14A.85
In
addition to hazards identified in the HAZOP workshop, control and monitoring
measures were recommended in the workshop and these measures are recorded in
the Hazard Register for future implementation. Example of these measures
include establishment of effective communication channel between HKCG operators
and PTW operators in case of emergency evacuation initiated from either side,
regular meeting with HKCG and closer and more frequent supervision in early
construction period from both HKCG and construction team representatives.
14A.86
Based
on the aforementioned control measures and monitoring procedures, any ground
settlement, soil movement and vibration during construction can be well
controlled and kept below allowable limits. Thus, there is no damage to the gas
holder and the associated facilities due to construction of the project works
14A.87
During
the HAZOP workshop, it was confirmed that the operation of upgraded PTW will be
similar to that of the current practice. There will be no adverse impact on the
HKCG Depot during operational stage.
14A.88
It
was also advised that a DG store would be set up within the PTW to support
regional need of DSD operations. Besides the DG store, limited quantities of
chemicals for general operation and maintenance of the PTW would be handled and
stored in the maintenance workshop. Estimated quantities and types of chemical
to be handled and stored in the upgraded Aberdeen PTW are given in Table 14A.10. Safety of the DG store is
evaluated by quantifying hazard ranges for various chemicals in the following
sections.
Table 14A.10 Estimated Quantities for Chemicals Handled in the
Upgraded
Substance |
Quantity |
Grease |
25
x 18 Litre |
Paint |
50
x 3.78 Litre |
Gasoline |
11
x 1.8 Litre |
Diesel |
4 x
18 Litre |
Thinner |
70
x 0.95 Litre |
Oxygen
gas |
2
cylinders |
Acetylene |
2
cylinders |
Note:
According to the list of dangerous goods tabulated in Dangerous Goods (General)
Regulations, Chapter 295B, Laws of Hong Kong (DG(G)R)), the on-site storage of
these DGs require a DG Licence.
14A.90
Acetylene
and oxygen cylinders will be used for welding purpose. Acetylene can form
explosive mixture in air. Hazard range of acetylene is assessed through consequence
modelling. In case of cylinder rupture, fireball of 11m radius is predicted. In
continuous release, jet flame lengths are 13m and 24m for 5mm and 10mm leaks
respectively. Thus, hazard will not go beyond the project site.
14A.91
Grease
and paint are chemically stable with high boiling points 310oC and
100oC respectively. They are both flammable but grease is not
volatile. Fire and extremely high temperature should be avoided. Fire is the major
hazard for these 2 chemicals. Apart from fire, paint in closed container may
cause explosion. Although both chemicals emit smoke and fumes in fire, they do
not have off-site hazard.
14A.92
Based
on this preliminary assessment, hazard due to onsite DGs storage can be
contained within the project site. The DG storage will not pose risk or cause
damage to the HKCG Depot.
Hazard Events at the Existing Shell Depot
14A.93
Historical
data sources, previous risk assessment reports [5][6] and MHIDAS database, etc,
have been reviewed. Generic Hazard events associated with the Shell Depot are
tabulated in Table 14A.11 below.
Table 14A.11 Hazards Associate with Operation of Shell Depot
Hazard event |
Potential Cause |
Spontaneous failure |
·
Storage vessel / cylinder ·
LPG Road Tanker failures (both transit and
LPG vessel refilling) ·
Pipework failures ·
Vaporizer failures ·
Flexible hose failures ·
Flange gasket failures ·
Valve leak failures ·
Outlet pipeline failure |
Loading from LPG Road Tanker to vessel |
·
Vessel filling hose misconnection ·
Vessel filling hose disconnection error ·
Disconnection with valve open ·
LPG Road Tanker drive away ·
LPG Road Tanker impact onto LPG facilities ·
LPG Road Tanker collision during unloading ·
Loading pipework over pressurisation ·
Storage vessel overfilling |
External event |
·
Earthquake ·
Aircrafts crash ·
Car crash ·
Natural terrain landslide ·
Severe environmental events ·
Lightning strike ·
Dropped object ·
Subsidence ·
External fire ·
Collapse and strike by object ·
Vibration |
14A.94 Storage vessel failure can be cold catastrophic leading to instantaneous release or cold partial failure resulting in continuous release of LPG to the atmosphere. The generic failure rate of 1.8x10-7 per vessel year [3] has been adopted for cold spontaneous catastrophic failure. For partial failure, a generic value of 5.0x10-6 per vessel year [3] has been adopted. The vessel is assumed stress relieved and 100% radiograph tested.
14A.95
LPG
road tankers in
14A.96 Cylinders of small size are stored in stack. Apart from cylinder shell, failure can also be initiated from cylinder valve. Cylinders may have cold catastrophic leading to instantaneous release or cold partial failure resulting in continuous release. Since LPG cylinders are subject to full inspection and re-qualification every 5 years and undergone leak test at filling plant, failure rates for LPG road tanker [3] are also applied to LPG cylinder for similarity in non-stationary nature. The catastrophic and partial failure rates of a LPG cylinder for the study are 1.0x10-6 [7] and 2.6x10-6 [14] per cylinder year respectively.
14A.97
A
side wall of LPG Storage Shed up to ceiling height and solid fence wall of 3m
high is erected along the site boundary next to Lee Nam Road and adjoining
boundary with South Horizons to prevent LPG cylinder fragments from hitting
outdoor population (no direct line of sight to LPG cylinders at ground level
along Lee Nam Road and at South Horizons). Besides, the LPG Storage Shed is
fitted with a metal sheet roof which can confine travelling distance of debris.
Although façades of some dwellings at South Horizons have direct line of sight to
LPG cylinders, indoor population is protected by building structures. Figure 14A.15 illustrates architectural features
which provide protection to ground population from projectiles. Since there is
no direct line of sight from ground level population, there is no direct impact
to ground level population and is not further considered in the
assessment.
14A.98
Cylinder
fragments flying through a window façade in a BLEVE accident which may lead to
fatality of indoor population at South Horizons. Frequency of flying fragment
in BLEVE is estimated to be 4.36x10-9 per year. Calculation is based
on the frequency of leak failure of a cylinder 2.6x10-6 per cylinder
year, probability of immediate ignition 0.005, vertical and horizontal view
angles as shown in Figure 14A.15. Fraction of window pane along a
vertical building façade and horizontal building façade has been considered.
From findings of the recent TNO report [13], BLEVE occurs after a time varying
from 5 minutes to 25 minutes when it is placed on a fire. In accordance with
PHAST modelling results, leak size <5mm can lead to continuous release over
5 minutes. Such leak size is equivalent to damage caused by impact of forklift
truck in moving of cylinders during operation hours. This leads to further
assumption that 50% of leak failure can cause BLEVE. While cylinders are stored
in stacks, 50% of fragments are considered being contained by stacks of
cylinders. In a site survey, 1664 cylinders of various sizes were found in the
transit depot. Since the transit depot is utilised to handle both fully loaded
and empty cylinders, 50% of cylinders are considered fully loaded and can cause
BLEVE. Event frequency for flying fragment and number of fatalities are
considered low in comparison with other events. This event is not further
assessed in the study.
14A.99
According
to a recent review of compressed gas safety by Cadwallader [8], a runaway or
almost runaway cylinder involves opening/removal of valve or regulator.
Operation of the transit depot involves movement of LPG pellets of cylinders
using forklift trucks. Due to human error, cylinder valve may be broken or
cylinder body may be punctured by a forklift truck. However, downward or
sideward force is generated and leads to spinning of the cylinder. While the
transit depot is fenced off with wire mesh and solid wall, a runaway cylinder
would not cause offsite fatality.
14A.100
14A.101 LPG road tankers carry a flexible hose for LPG unloading to LPG storage vessels. A generic guillotine failure rate of 9x10-8 per hour [3] is adopted. An Excess Flow Valve (EFV) is fitted immediately upstream of the hose. Similar to pipework failure, the effect of partial failure of the hose is small but is also considered in the assessment with failure rate of 3.3 times guillotine failure rate applied.
14A.102 Failures of gaskets and valve leak would only tend to give relatively small scale of leakage and will not contribute to the off-site risk. The results from gasket failure will not be considered separately but absorbed into pipework failure in the study.
14A.103 This study only considers misconnection errors which results in hose coming completely apart, giving a full-bore release. Small leaks will be rectified instantaneously by the truck driver or his assistant, and hence are not considered. A failure rate of 3x10-5 per operation [3] resulting from human error which leads to misconnection is adopted.
14A.104 This is initiated by human error which requires a complete disregard of normal operating procedure as well as the failure to preventing it from happening. A failure rate of 2x10-6 per operation [3] for an operator to disconnect a hose during loading operation has been used in the study.
14A.105 The event is significant only if the release is fed from the storage vessel and when the vessel is over-pressurised with failure of the associated safety valves (i.e. a Non-return Valve in the present case), as well as driver’s failure to shut down the Manual Valve 0.5 per operation [3] together with the hose disconnection error.
14A.106
A
drive-away error could be resulted from repositioning of truck during delivery
or inadvertent drive-away before completion of replenishment. The outcome of
this failure matches those of hose misconnection. A number of measures such as
the use of wheel chocks, interlocks on shutters and parking brake have been
implemented in
14A.107 LPG Road Tanker impact refers to the LPG installation being hit by a LPG road tanker causing damage to the installation, pipework, the road tanker itself, the LPG road tanker fittings or the hose connection pipework. Failure rate of 1.5x10-4 per operation [3] for “tanker impact during unloading” is adopted.
14A.108 A dedicated loading bay for parking of the LPG road tanker is provided and a speed bump at entrance of the LPG compound is installed. Frequency of 1x10-8 per operation [3] for vehicle impact into an unloading LPG road tanker is adopted and fed into fault tree analysis.
14A.109 It is possible that a LPG road tanker driver makes an error while unloading from the road tanker to the storage vessel. Over pressurisation of the liquid filling line would be resulted should the operator forget to open all relevant valves on the pipe/hose. However, over-pressurisation protection system of the LPG road tanker should fail and other failure to isolate leak system (e.g. possibility of the leak being isolating using manual valves) does not work. The concerned scenario will have a much lower probability to happen than the “misconnection” error event (which will lead to a similar outcome) and the misconnection error has already accounted for this factor.
14A.110
The
on-site practice in unloading LPG to the storage vessel is that the vessel will
only be filled to up to 85% of the maximum capacity, through monitoring with a
level gauge during loading operation. The reserved capacity is able to contain
a full load of replenishment. It is also an offence in
·
failure
of the truck pump overpressure protection system;
·
failure
of pressure relief valve (PRV) on storage tank;
·
failure
of driver and his assistant to detect problem and to take effective mitigation
action.
14A.111 In equipment failure, it is possible for staff to rectify the problem before any hazard event occurs. Two staff is responsible for the unloading process (i.e. the driver and one site staff). As the staff should have undergone training programme for the job, the probability that the problem cannot be rectified before hazard event occurs can be assumed to be lower than 0.5. In this assessment, a probability of 0.5 is assumed for human error [3].
14A.112
Similar
to spontaneous failure of pipework, failure of this supply line can be
guillotine or partial failures. LPG transforms from liquid state to gaseous
state through heating of vaporisers. Vaporiser failure would cause release of
liquid state LPG in the worst case. There are 2 types of vaporizer in use with
production rates of 500kg/hr and 1000kg/hr.
14A.113
Similar
to spontaneous failure of pipework, failure of this supply line can be
guillotine or partial failures. However, LPG is in gaseous state after it has
passed vaporisers.
14A.114
A
LPG road tanker/cylinder wagon in transport is subject to the same failure
hazards as it is stationary. LPG road tankers and cylinders wagons may experience
spontaneous failure as well as collision impact or subsequent fire impact in a
severe traffic accident. Although speed restriction is imposed within the
depot, serious accidents are considered to be possible within the depot for
conservativeness. By reference to historical incidents and SEKD [6] study, the
following hazards can lead to gas release from LPG road tankers,
·
BLEVE/
Fireball - engulfed in fire with fire protection system failure
·
Cold
catastrophic failure – spontaneous failure in transport, LPG road tanker
rupture in traffic accident or thermal expansion of an overfilled LPG road
tanker with PRV failure
·
Partial
failure – LPG road tanker shell leak, access hatch leak and un-isolated
pipework leak lead to liquid release. PRV spontaneous failure and PRV leak in
traffic accident lead to gaseous release
14A.115
Hazards
lead to gas release from LPG cylinder wagon including,
·
Cold
catastrophic failure – spontaneous failure in transport or cylinders rupture in
traffic accident
14A.116
All
LPG road tankers/cylinder wagons waiting for aboard ferry at the Depot. Their
arrival time is not far apart. Moreover, fully loaded LPG road tankers arriving
by ferry leave the transit depot as soon as they disembark. Cylinder wagons
arriving by ferry unload to the LPG cylinder shed. An average wait time of
20-30 minutes per LPG road tanker/ cylinder wagons is considered reasonable and
conservative. Since parked LPG road tankers/ cylinder wagons have their engine
turned off, fire hazard to the queuing LPG road tankers/ cylinder wagons due to
traffic accident fire are considered not possible. Except multiple BLEVE
hazard, other hazards relevant to movement of road tankers/ cylinder wagons
inside the Depot are also applicable to queuing vehicles.
14A.117
No
unauthorised vehicle is allowed and speed restriction is imposed within the
Shell Depot. Besides, the depot is fenced off from
14A.118
In
14A.119
The
distance between the nearest arrival flight path and the Depot is more than 2
miles. The distance between the Depot and
14A.120 Soil/rock slopes Slope No. 15NW-A/C2 and 15NW-A/C3 are located at 15m from east boundary of the Shell Depot and are separated by Lee Nam Road. The slopes are not classified as dangerous hillside slope and do not require non-routine maintenance. Surface protection was applied with 50% and 100% shotcrete for the 2 slopes.
14A.121
Loss
of containment due to severe environmental event such as typhoon or tsunami
(large scale tidal wave) is not possible as the LPG vessel is mounded. However,
typhoon may tumble LPG cylinder stacks at the LPG cylinder shed. Damage to a
cylinder is possible when it falls from stack. While
14A.122 Subsidence is usually slow in movement and such movement can be observed and remedial action can be taken in time. Besides, the Depot has been operating over 15 years. The probability of hazardous event due to subsidence is therefore assumed zero.
14A.123 External fire means the occurrence of fire event which lead to the failure of the gas related facilities. The key potential concern relates to the LPG compound being affected by the fire of LPG road tanker and LPG transit depot being affected by fire in collision accident.
14A.124
In
14A.125
In
severe collision accident, fire-resisting shields on LPG road tankers/cylinder
wagons can be damaged and cannot isolate a fire effectively. Fire hazard on
movement of DG vehicles are considered within the transit depot.
14A.126
The
frequency of lightning strike on a properly protected building is extremely low
in
14A.127
There
is no industrial activity close to the Shell Depot. Moreover, there is a
concrete fence wall next to
14A.128
There
is no risk of damage to the gas depot by collapse and strike of surrounding
object in the existing setup. This factor is further elaborated in sections
regarding construction activities in the Ap Lei Chau PTW.
Hazard Events Initiated by Ap Lei Chau PTW and Safety Measures
14A.130
The
hazardous event of the Shell Depot is LPG (Gas/Liquid) release that may lead to
fire and explosion. LPG leakage could be initiated by external events such as
damage of gas installations by construction activities in PTW site.
14A.132
The
Ap Lei Chau PTW is located in the 500m-consultation zone of the Shell Depot as
indicated in Figure 14A.6. The impact generated from tunnel construction
and the upgrading of the PTW and hazards associated with the Shell Depot have
been identified and presented in the following sections.
14A.133
Among
all external events in Table 14A.11,
only subsidence, external fire, collapse and strike by object and vibration are
potential hazards generated from the construction activities of the PTW. These
activities are further developed to various hazardous scenarios in the
following subsections.
14A.134
Review
on construction and operation activities have been conducted prior to the HAZOP
workshops held in June 2007. These activities and the hazards identified in the
workshops are recorded in Hazard Registers as attached in Appendix 14A.2. These hazards and the relative information are used
as the basis for further development of the risk model.
14A.135
Construction
activities in the vicinity of the Ap Lei Chau PTW are listed below. A general
layout of the PTW site is presented in Figure 14A.16.
·
Upgrading of PTW. Relevant construction works will
last for about 3.5 years and will involve demolition of existing buildings and
construction of new buildings & structures, installation of electrical and
mechanical equipment, laying of sewers, drainage pipes, and other utility lines, as well
as non-percussive
piling and ground excavations to depths up to 11m below ground level for the
Transfer Pumping Station(in which sheet piling would be adopted for temporary support during
excavation); and
·
Construction of Drop Shaft and
Tunnel “Q”.
Construction of a tunnel connecting to the Aberdeen PTW by HDD and a permanent
drop shaft adopting mechanical methods and no explosive will be involved.
14A.136
The
primary function of PTW will remain the same as the existing one. The layout
plan for the upgraded PTW is shown in Figure 14A.17. Sewage will
continue to enter the PTW via the existing sewer system. It will be pumped
through the various treatment units in the PTW before it is discharged into the
deep tunnel system.
14A.137
The
distance between the ALC PTW north boundary and the entrance of Shell Depot is
more than 100m. The LPG cylinder shed is even further away from the north
boundary at 140m. Since work area
for the ALC PTW is relatively far away from the LPG cylinder shed and the
transit depot, construction works of the PTW does not have direct impact on the
LPG cylinder store and transit tankers/cylinder wagons. However, hazards caused
by toppling/dropping of gas cylinders from loading platform which is
initialised by ground settlement, are further elaborated in the assessment.
14A.138
External
interferences affecting the normal operation of the Shell Depot may originate
from the following,
·
Excessive
vibration
·
Ground
movement/settlement. Differential ground displacement is the most concerned in
ground settlement as it causes shear stress to components
·
Strike
by other objects from the PTW construction area
·
Fire
impingement or high thermal radiation
14A.139
In
general, potential damages initiated by these hazards are listed below.
·
Liquid
release due to damage of LPG storage vessel/ LPG cylinder
·
Liquid
release due to damage of LPG road tanker
·
Gas/Liquid
release due to damage of pipeline
·
Gas/Liquid
release due to damage of other gas installation e.g. vaporizers
14A.140
Owing
to the close proximity, it is more likely that the LPG compound and its
facility are affected by the ALC PTW site and its activities. Hazards, which
are caused by construction works of the PTW, with potential impact to the
adjacent Shell Depot have been categorized and tabulated in Table 14A.12.
Table 14A.12 Hazards with Potential Impact to Shell Depot
Hazardous Scenario |
Damage Outcomes |
Construction Activities |
Potential Cause |
Hazard Log Ref |
Ground
settlement inside LPG Compound/ cylinder store |
Ground
Collapse underneath LPG Tank, underneath vaporizers rooms and/or underneath |
Excavation
at PTW site about 11m deep |
Unexpected
drawing down of groundwater table. |
1 |
Ground
Collapse underneath LPG Tank, underneath vaporizers rooms and/or underneath |
Tunnelling - Drop shaft construction (inside Ap Lei Chau PTW) by mechanical boring - Tunnel Q construction by horizontal directional drilling (HDD) |
Unexpected drawing down of groundwater table. |
7,
8 |
|
Structural
damage to gas Installation |
LPG
Tank, vaporizers, LPG Road Tanker and gas pipework damage |
Mobilising and usage of construction equipment (e.g. backhoe, bulldozer, dump truck, site vehicle etc) |
Loss
of Stability or Mechanical |
5 |
Tunnel
Q construction by horizontal directional drilling (HDD) |
Construction
of Tunnel Q strayed from design
alignment to cause unexpected damage to the nearby
LPG facility |
8 |
||
Piling
Works |
Excessive
ground vibration leading to structural damage of LPG facility. |
2 |
||
Demolition
of existing PTW structures (and substructures) |
Excessive
vibrations due to inappropriate method of demolition |
3,
9 |
||
Fire/Explosion
Hazard to Gas Depot |
LPG
Tank, LPG Road Tanker and outdoor pipework damage due to missile of explosion
at PTW site |
Installation
of electrical and mechanical equipment |
Fire
escalation from construction site |
4,
6 |
LPG
Tank, LPG Road Tanker and outdoor pipework damage due to missile of explosion
at PTW site (Fire Escalation) |
Temporary
storage of construction plant, equipment, and materials in temporary works
area (during Construction Stage) |
Ignition of flammable material (DGs or
Construction material) due to Hot Works. |
10 |
14A.141
Regarding
vibration due to construction activities, mitigation measures, such as using of
non-percussive piling methods and vibration monitoring, will be implemented to
ensure that the velocity and amplitude of vibration will not damage to the Shell
Depot. Therefore, this hazard is not assessed further in this study. Details of
the monitoring and control measures are stated at end of this section.
14A.142
For
construction of Tunnel Q using HDD, pilot hole will be drilled under close
supervision to avoid deviation from design alignment during tunnel
construction. Periodic monitoring and checking of Tunnel Q construction every
25 meters will be implemented. The Shell Depot would not be damaged due to
misalignment of tunnel construction underneath the depot.
14A.143
There
is neither materials store nor nake flame at or near the adjoining boundary.
Besides, there are a 10m buffer (access road within the PTW site) and a 3m high
fence wall separating the PTW site and the LPG compound. Thus, fire at the PTW
site would not affect the LPG compound.
14A.144
In
the interface diagram for drop shaft construction and PTW upgrading work as
shown in Figure 14A.18, the drop
shaft and temporary work site for tunnelling work is at least 60m away from the
south boundary of the Shell Depot. Mobile crane will be used for assembling of
construction equipment and drilling rig at ground level within the work site.
Crane outreach height for assembling of drilling rig is estimated between 8m
and 12m. Use of crane within the work site does not impose any risk of direct
mechanical damage to the Shell Depot.
14A.145
For
construction of the new PTW building, a mobile crane or tower crane may be
erected. There are 2 possible locations for erection of the tower crane and are
indicated in Figure 14A.19. The new PTW
building is approximately 11m from ground level while the rest of structures
such as shelter for grit trap and switch room vary from 4m to 12m. Crane
Location 1 is adjacent to the new building and approximately 10m from the Shell
Depot boundary. Use of crane at such close proximity to the LPG compound would
impound risk of damage to the associated facilities in case of mechanical
failure. Crane Location 2 is viable from safety point of view while it can
provide more than 10m buffer distance from the end boom. Outreach of cranes is
illustrated on Figure 14A.19.
14A.146
While
a mobile crane may topple or collapse towards the load carrying end, damage to
the LPG compound due to such accidents can be avoided by controlling
orientation, swing angle, lifting height of the crane as well as always facing
crane jibs away from or parallel to the adjoining site boundary with the LPG
compound. With this fail-safe arrangement, a collapsed crane will not fall on
the Shell Depot even sitting at Crane Location 1.
14A.147
Climbing
and dismantling of a tower crane is the most common cause of fatal accident
involving particular hazards relating to the carrying of unbalanced load. In
such accidents, crane arms may run across the PTW site and hit the LPG
compound. However, the damage can be avoidable by turning crane jibs parallel
to the adjoining boundary or away from the depot when carrying out this
process.
14A.148
Load
slipping and lifting accessory failing cause load falling which may damage LPG
compound. Safety zone will be set out such that a loading arm is also kept away
from the depot boundary. Protective fencing consisting of mesh will also be
used to catch projectiles or debris of fallen load.
14A.149
Due
to the relatively large separation distance and nature of use, direct impact of
PTW construction activities on the transit and LPG cylinder depot does not
exist. However, ground settlement may lead to damage to LPG cylinder shed and
LPG cylinder platform. Collapsed structures subsequently cause rupture of
cylinders because of collapse of cylinder stacks and collision between
cylinders. However, ground settlement will be monitored and controlled within
allowable limits. Significant ground settlement will be avoided. Preliminary
monitoring procedures are described in later sections.
14A.150
Strategy
for ground settlement and vibration monitoring is similar to the one for the
Aberdeen PTW. Preliminary monitoring locations and plan are described in later
sections. Alert and action limits will be set such that construction works will
be stopped when vibration or ground settlement exceed the corresponding action
limit. Construction works will proceed only after sources of
settlement/vibration have been identified and rectification of the problem has
been completed.
14A.151
Based
on the aforementioned control measures and monitoring procedures, amplitude of
ground settlement, soil movement and vibration can be well controlled and kept
below allowable limits. Thus, construction of the ALC PTW will not cause damage
to the Shell Deport.
14A.152
During
the HAZOP workshop, it is confirmed by PTW operator that the operation of upgraded
PTW will be similar to that of the current practice. Small amount of DGs
similar to those current used for maintaining the existing PTW would be stored
on site during future operations. Operation of the upgraded ALC PTW would not
have any adverse impact on the Shell Depot.
14A.154
Frequencies
for rupture failure 5x10-6 per year for process vessel [7], leak
failure for 200mm equivalent hole-size 4x10-5 per year [6] and blown
seal failure for 1m equivalent hole-size 4x10-5 per year [6] are
adopted. Based on the design of the
14A.155
In
accordance with [7], full bore rupture and leak failure frequencies for
gas/liquid pipework are 1x10-7 per m.year and 5x10-7 per
m.year respectively. After these values are localised for pipe length,
frequencies of failure for full bore rupture and leakage are 4x10-6
per year and 2x10-5 per year respectively by assumption total
pipeline length of 40m.
14A.156
Since
the gas depot neither locates underneath existing flight path nor adjacent to
an unstable slope, no adjustment to the failure frequencies is made for these 2
factors. Similarly, no adjustment to failure frequencies for gas pipeline and
gas holder is made as gas depot site is not particularly subject to attack from
various environmental factors such as large scale tidal wave which have been
discussed in the “external events” section. The aforementioned failure rates
together with other external events such as car crash and subsidence are
adopted in the Fault Tree Analysis (Appendix 14A.5) to develop the hazardous
outcome frequencies for the HKCG Depot are tabulated in Table 14A.13.
Table 14A.13 Frequencies
of Initial
Outcome |
|
|
Existing/Operation |
Construction |
|
Gas Holder Rupture |
5.00x10-6 |
5.00x10-6 |
Gas Holder Leakage (200mm diameter) |
4.07x10-5 |
4.07x10-5 |
Blown Seal Failure (1000mm diameter) |
4.07x10-5 |
4.07x10-5 |
Pipeline Rupture |
4.16x10-6 |
4.55x10-6 |
Pipeline Leakage (50mm diameter) |
2.06x10-5 |
2.06x10-5 |
14A.157 For Ap Lei Chau Site, equipment failure rates as documented in Appendix 3 of the QRA Methodology Paper [3] have been used for Fault Tree Analysis (Appendix 14A.5) to develop the hazardous outcome frequencies for the LPG compound, which is presented in Table 14A.14.
Table 14A.14 Frequencies
of Initial
Outcome |
|
Cold Catastrophic |
6.72E-07 |
Cold Catastrophic |
1.48E-07 |
Cold Partial |
1.17E-05 |
Cold Partial |
3.62E-07 |
|
5.14E-07 |
|
7.69E-07 |
|
4.00E-06 |
|
1.30E-05 |
|
3.64E-08 |
|
3.82E-05 |
|
3.92E-05 |
|
4.00E-06 |
|
1.30E-05 |
|
2.00E-03 |
|
5.20E-03 |
14A.158
For
transit operation of the depot, failure frequencies are calculated by referring
SEKD Study [6], number of vehicles movement in the depot and length of the
internal access road (0.2km).
Table 14A.15 Frequencies
of
Event |
Likelihood (per vehicle km) |
Release mass (kg) |
Hole size (mm) |
No. of vehicles per day |
No. of vehicles per year |
Frequency |
LPG road tanker |
||||||
BLEVE |
2.70E-12 |
9000 |
n/a |
2 |
730 |
3.94E-10 |
cold
rupture |
2.60E-09 |
9000 |
n/a |
2 |
730 |
3.80E-07 |
large
leak - liquid |
1.80E-08 |
9000 |
50 |
2 |
730 |
2.63E-06 |
large leak (vapour) |
2.10E-09 |
9000 |
50 |
2 |
730 |
3.07E-07 |
medium
leak (liquid) |
6.80E-09 |
9000 |
25 |
2 |
730 |
9.93E-07 |
LPG cylinder wagon |
||||||
BLEVE |
1.30E-09 |
50 |
n/a |
2 |
730 |
1.90E-07 |
Rupture |
2.80E-08 |
50 |
n/a |
2 |
730 |
4.09E-06 |
Notes: frequency =
likelihood per vehicle.km x no. of vehicles per year x length of access road
14A.159
Taking
into account time waiting for abroad or disembarkation, it is assumed each LPG
road tanker/cylinder wagons staying in the transit depot for 25 minutes. This
time factor is adopted to derive frequencies of failure events associating with
DG vehicles at the transit depot. Failure frequencies are calculated by
referring SEKD study [6] for events likelihood, number of vehicles movement in
the depot and time at the depot. BLEVE is not considered as LPG road
tankers/cylinder wagons do not engaged in fire hazard while they are parked.
Table 14A.16 Frequencies
of
|
Likelihood (per vehicle year) |
Release mass (kg) |
hole size (mm) |
No. of Vehicles |
No. of Vehicles per year |
Freq. |
LPG road tanker |
||||||
cold
rupture |
4.00E-08 |
9000 |
n/a |
2 |
730 |
1.39E-09 |
Large leak (liquid) |
3.60E-08 |
9000 |
50 |
2 |
730 |
1.25E-09 |
Large
leak (vapour) |
3.60E-08 |
9000 |
50 |
2 |
730 |
1.25E-09 |
Medium
leak (liquid) |
3.60E-08 |
9000 |
25 |
2 |
730 |
1.25E-09 |
LPG cylinder wagon |
||||||
Rupture |
6.80E-06 |
50 |
n/a |
2 |
730 |
2.36E-07 |
Notes:
frequency = likelihood per vehicle year x no. of vehicles per year x wait-time
in fraction of a year
14A.161
To
determine the outcomes frequencies, the base frequencies for initial events are
inputted to event trees together with the probabilities of damage of the gas
installation as well as the probabilities of gas ignition. Factors, such as type of release, operation of excess flow valve,
leakage isolation, presence of ignition source, timing of ignition, geographic
location, environmental conditions and meteorological conditions are taken into
account.
14A.162
SAFETI’s
built-in event trees are applied for calculating frequency of hazardous
outcomes.
14A.163
Upon
completion of frequency estimation for the hazardous outcomes, the consequence assessment
then estimates impact of each outcome in the area of concern. In SAFETI, built-in event tree is
used for determination of consequence by considering type of release, presence
of ignition sources, timing of ignition, geographic location, environmental and
meteorological conditions.
14A.164
For
the Ap Lei Chau project site, it is assumed that all LPG released will be
flashed to vapour so that flash fire can be formed instead of pool fire for delayed
ignition. Similarly, flash fire can be formed for release of town gas. This
assumption leads to the worst scenario as pool fire is usually a localised
hazard event while extent of a flash fire is larger than a pool fire.
14A.165
The
consequence assessment consists of two major parts, they are:
·
Source
term modelling – to determine the amount and rate of gas release; and
·
Effect
modelling – to determine the toxic, flammable and explosion effect to offsite
population.
14A.167
For
pipelines connecting to a gas distribution network, continuous release without
shutdown mechanism is assumed. For pipelines connecting to a storage tank,
release duration is based on time to empty the whole content.
14A.168
LPG
model is a mixture of butane and propane in 7:3 ratios. For the LPG Compound, 85% of the total
storage capacity is used by taking into account the day-to-day practice of the
operator. Since there are 2 vessels in operation, release quantity in catastrophic
failure of the vessel is based on maximum storage quantity (17 tons) for a
single vessel. In continuous release, 34 tons release quantity (equivalent to
total storage capacity) is used. 9 tons release mass (full capacity of a LPG
road tanker) is used for LPG road tanker related events.
14A.169
Since
LPG cylinders store keeps cylinder from 2 kg to 49 kg, 49 kg is assumed for
release quantity. This amount is equivalent to single cylinder failure of 49-kg
cylinder.
14A.170
While
the transit facility serves LPG road tankers and cylinder wagons, all road
tankers are assumed having capacity of 9 tons and are fully loaded. Release
quantity for LPG cylinder in wagons is taken as 49kg from a 49-kg cylinder, as
same as those for LPG cylinders store.
14A.171
PHAST
consequence model of the SAFETI 6.51 by DNV is used for calculation of
hazardous area under various consequences. The following section briefly
describes mathematical models applied to various fire and dense gas dispersion
in the consequence model.
14A.172
The
UDM model without rainout effect is used for the dispersion of towngas/LPG for
non-immediate ignition scenarios to obtain more conservative results. The model
takes into account various transition phases, from dense cloud dispersion to
buoyant passive gas dispersion, in both instantaneous and continuous releases.
Besides, toxic effect is evaluated using the UDM dispersion model when the
cloud reaches population sites for release of Towngas without ignition.
14A.173
Upon
release of flammable gas, a number of possible outcomes may occur depending on
whether the gas is ignited immediately or ignited after a period of time. The
dispersion characteristics are influenced by meteorological conditions and
material properties, such as density, of the released gas.
14A.174
Fire
scenarios of different kinds may be developed in the presence of ignition
source in the proximity of gas release. If no ignition source exists, the gas
cloud may disperse downwind and be diluted to the concentration below its Lower
Flammable Limit (LFL). In this case, the gas would become too lean to ignite
and have no harmful effect.
14A.175
A
BLEVE is a sudden rupture due to fire impingement of a vessel containing
liquefied flammable gas under pressure, which results in a fireball as the
flashing liquid ignited. BLEVE Blast model in SAFETI is applied.
14A.176
For
immediate ignition of an instantaneous gas release, a fireball will be formed.
Fireball is more likely for immediate ignition of instantaneous release from
LPG vessels/tankers due to cold catastrophic failure although it is possible
for late explosion. Instantaneous ignition of a certain mass of fuel (flammable
gas/LPG) results in explosion and fire of hemispherical shape. Heat is evolved
by radiation. The principal hazard of fireball arises from thermal
radiation. Due to its intensity,
its effects are not significantly influenced by weather, wind direction or
source of ignition. Sizes, height, shape, duration, heat flux and radiation
will be determined in the consequence analysis.
14A.177
A
jet fire is typically resulted from ignition of gas/liquid discharging from a
pressurised containment. Major concerns regarding jet fire are jet flame and
the heat radiation effect generated from the jet flame. Thermal effect of the
jet fire on adjacent population is quantified in the consequence model.
14A.178
A
flash fire is the consequence of combustion of gas cloud resulting from delayed
ignition. The flammable gas cloud can be ignited at its edge and cause a flash
fire of the cloud within the LFL and Upper Flammable Limit (UFL) boundaries. In
case of continuous release, fire is flashed back to the release source and
leads to jet fire. Major hazards from flash fire are thermal radiation and
direct flame contact. Since the flash combustion of a gas cloud normally lasts
for a short duration, the thermal radiation effect on people near a flash fire
is limited. Humans who are encompassed outdoors by the flash fire will be
fatally injured. A fatality rate of
1 is assumed.
14A.179
A
vapour cloud explosion can occur when a flammable vapour is ignited in a
confined or partially confined situation. Although VCE is unlikely for both
project sites where vessels installed aboveground, the risk model SAFETI has
accounted for the VCE hazard according to probabilities for delayed ignition in
consequence modelling. The program models the delayed ignition effect by
considering the flammable cloud area and location of ignition sources at each
time step.
14A.180
To
determine the fatality rate, the following Probit equations will be used to
determine lethal doses for various hazard scenarios.
Thermal radiation [8]
Pr = -36.38 + 2.56 ln (Q1.33 x t) where
Q is the thermal radiation intensity in W/m2.
Toxic gas dispersion for Towngas [4]
Pr = -50.95 + 3.7 ln (C x t) where
C is concentration in ppm.
14A.181
From
the consequence modeling results generated by PHAST model, hazardous distances
of various fire scenarios for
14A.182
As
a result of consequence modelling using PHAST, the representative fireball size
can be obtained from the gas holder rupture event with radius around 80m and
lift-off height around 160m. Such fireball covers a small section of
14A.183
The
extent of flame length can be as far as 52m for jet fire hazard which can be
found in pipeline rupture event. Flammable cloud can disperse downstream 50m in
pipeline rupture. Although flammable cloud can reach up to downstream 222m
(weighed average for all weather classes) in gas holder rupture event, it would
be ignited by ignition sources along
14A.184
Toxic
effect of Towngas is not significant and does not have offsite risk.
14A.185
From the consequence modelling results generated by PHAST
model, hazardous distances of various fire scenarios for Ap Lei Chau have been
determined. However, street lamps and lighting
system within the Shell Depot would ignite vapour cloud within the depot in a
release event and the vapour cloud would not disperse to the predicted
distances. Details of hazardous
distances for these scenarios are given as follows.
14A.186
Maximum
effect distance for fireball hazard occurring within the LPG compound is 58m
radius corresponding to road tanker failure with a lift-off height up to 116m.
Since the 17 tonne of LPG in each vessel are mounded within compartments with
sand, possibility of a fireball from these vessels is negligible and hence its
effect is not evaluated. Jet fires (flame length) can go up to 30m in LPG
vessel leak and supply line/vaporiser failure. Both fireballs and jet flames
cover only a section of
14A.187
For
the transit facility, maximum fireball size is obtained from LPG road tanker
BLEVE with radius up to 65m and lift-off height of 125m. Jet fire flame length
can be up to 55m in LPG road tanker leak. Flammable cloud in LPG road tanker
rupture can disperse downwind up to 150m without immediate ignition.
14A.188
For
LPG cylinder store, cylinder rupture (49kg release mass) can produce 11m in
radius fireball with lift-off height up to 25m. The fireball does not impose
any offsite risk. In case of leakage, the flammable cloud has downwind distance
less than 30m without immediate ignition.
·
Individual
risk contours; and
·
Societal
risk present in FN Curves.
14A.190
Risk
estimates from the assessment are then compared with the acceptability criteria
stipulated in Annexes 4 and 22 of the TM for evaluating hazard to life. Risk ranking exercise is also
carried out to select priorities for mitigation actions, if any.
14A.191
Individual
risk (IR) contours of construction phase (Year 2009 scenario) for
14A.192 Societal risks in terms of FN curves for existing, construction phase and operation phase scenarios are plotted on the same graph Figure 14A.22 for ease of comparison. Societal risk for all scenarios is within the acceptable region of the risk guidelines.
14A.193
For
construction phase, frequency of accidents increases slightly in comparison
with the existing risk level and the increase in fatalities is due to
construction workers at the PTW worksite.
14A.194
Societal
risk in operation phase almost overlaps with the one for existing case. No
addition risk is generated by the operation of the upgraded PTW.
14A.195 Societal risks for LPG Transit Depot and LPG Compound in terms of FN curves are shown on the Figure 14A.23A and Figure 14A.23B respectively. For ease of comparison, results for existing, construction phase and operation phase scenarios are plotted on the same graph. The existing risk level for LPG Compound and LPG Transit Depot falls into the acceptable region.
14A.196
For
construction phase, societal risk for the LPG Compound increases but still
keeps in the acceptable region. On the other hand, societal risk for the LPG
Transit Depot is similar to the existing scenario and it is in the acceptable
region. This indicates the project work site is not significantly affected by
the LPG Transit Depot.
14A.197
Societal
risk in operation phase is very similar to the existing scenario. No addition
risk is generated by the operation of the upgraded PTW. The societal risk for
the Ap Lei Chau Shell Depot, as a whole, is at acceptable level during
construction and operational phase.
Mitigation and Monitoring Measures
14A.198
The
main strategy for groundwater and settlement control will focus on limiting
groundwater inflows to acceptable levels. Groundwater inflow into the tunnels
can be limited by pre-grouting ahead of the tunnel face, post-grouting of the
tunnel or sealing the tunnel with a relatively impermeable lining at a short
distance behind the working face, thereby limiting the duration and total
volume of groundwater inflows into each section of tunnel. Pre-grouting ahead
of the tunnel face is considered to be the most effective method of controlling
groundwater inflow where linings are not installed close to the working face.
Post-grouting would be used where the pre-grouting efforts are found to be
inadequate.
14A.199
Ground
vibration would be minimized using non-percussive pre-bored H-Piles for piling
works. Ground vibration control can be achieved through close monitoring of
soil movement. Monitoring of vibration resulted from construction works to
ensure the velocity and amplitude of vibration as recommended by relevant
government authorities will not be exceeded. Monitoring plan will include
setting up monitoring points at fixed locations adjacent to more sensitive
structures which have lower vibration or air overpressure tolerance. The
monitoring will be carried out at ground level with sensors mounted on small
concrete bases embedding into undisturbed ground. Each sensor will be able to
record both ground vibration and air overpressure. For the Tunnel P, monitoring
will be carried out using a moving array of sensor locations as the tunnel is
advanced. These will be largely sited above or close to the tunnel alignment
and mounted on small concrete bases embedding into undisturbed ground. For some
particularly sensitive structure within the HKCG, sensor will be located at the
structure, farther from the blast, to ensure the vibration limit at the
structure is not exceeded.
14A.200
Other
mitigation measures include,
·
Follow
Standard and Guidelines given in Gas Production & Supply: Code of Practice
for Avoiding Danger from Gas Pipes
·
Close
supervision is recommended during peak construction period
·
The
safeguards for this potential impact will be further confirmed in detailed
design stage
14A.201
Damage
to PHIs due crane operation is avoidable through the following design factors,
·
Location
and orientation of crane installation
·
Appropriate
lifting height
·
Restriction
on swing angle
·
Special
attention should be paid for climbing and dismantling of a tower crane.
·
No
overloading
·
Set
out safety zone
14A.202
Effect
distances for different DGs have been evaluated. As long as buffer distance of
25m can be provided, use and storage of DGs do not have adverse impact on the
PHIs.
14A.203
Ground
settlement is to be monitored by a series of geotechnical instrumentations. A
series of ground settlement markers and piezometers are to be installed in the
vicinity of the proposed tunnel and near the HKCG Depot are presented in Figure 14A.24. Monitoring
of groundwater and settlement will form an important part of the strategy to
confirm the effectiveness of the mitigation measures and to adjust the inflow
limits if necessary. In order to cater for unforeseen problems encountered
during the construction stage, additional monitoring and instrumentation system
will be installed in the vicinity of the HKCG Depot at a later stage by the
Contractor before the commencement of construction works. Typical allowable
limits for well maintained buildings and historic buildings are 25mm and 5mm
respectively. Allowable limit of 13mm is proposed (Appendix 14A.4) and the actual
allowable limit is subject to further requirements of CEDD and relevant
authorities. Gas
facilities should be able to withstand ground settlement (including differential
settlement) within the allowable limit.
14A.204
Monitoring
plan will include setting up monitoring points at fixed locations adjacent to
more sensitive structures which have lower vibration tolerance. The monitoring
will be carried out at ground level with sensors mounted on small concrete
bases embedding into undisturbed ground. Each sensor will be able to record
ground vibration. For some particularly sensitive structure within the Shell
Depot such as LPG compound, sensor will be located at the structure, farther
from the vibration source, to ensure the vibration limit at the structure is
not exceeded. Allowable limit of 5mm/s PPV was proposed
to HKCG (Appendix 14A.4). The same
criteria will be also applied to the Shell Depot.
14A.205
It
was advised, during the HAZOP meeting, that no gas leakage detection alarm is
installed in the Shell Depot. It is recommended to install gas
detection/alarming system to provide warning to Shell operator and Construction
Site staff in case of gas leakage. This provides an early warning for operation
staff and construction staff in case of gas release. Specific emergency
procedures should be developed, together with Shell Depot Operation Team, and
documented in an emergency plan to cater for gas leakage scenarios before the
construction stage.
14A.206
The
following measures are recommended as “Best Practice” for DSD to implement
during construction stage for both
b)
The
emergency procedures should specify means of providing a rapid and direct
warning (e.g. Siren and Flashing Light) to construction workers in the event of
gas release in the gas facilities.
c)
The
construction site officer of DSD should establish a communication channel with
the gas operation personnel during construction stage. In case of any incidents
in the gas facilities that need site evacuation, operation staff of gas
facilities should advise the DSD’s site officer to evacuate the construction
workers.
d)
Induction
Training should be provided to any staff before working at the both work site.
e)
Periodic
drills should be coordinated and conducted to ensure all construction staffs
are familiar with the evacuation procedures. Upon completion of the drills, a
review on every step taken should be conducted to identify area of improvement.
14A.207
For
both project sites, potential damages to HKCG Depot and Shell Depot can be
avoided with implementation of safety measures and close monitoring procedures.
Operation of these PHIs will not be disturbed and gas supply will not be
disrupted provided that vibration and ground settlement caused by construction
works can be controlled within allowable limits. Noticeable change in
individual risk during construction phase is not observed while the 1x10-5
per year maximum level of off-site individual risk stipulated in Annex 4 of the
EIAO TM can be satisfied for both HKCG Gas Holder and Shell Depot.
14A.208
Societal
risk for both PHIs does not increase significantly due to construction works of
HATS project and are within the acceptable level during construction and
operation phases. Although the LPG Compound can be maintained to acceptable
level in construction phase of the Ap Lei Chau PTW, installation of gas
detection system at the LPG Compound together with ground vibration monitoring
can bring the risk level further down as an early warning to safeguard
construction workers at the project site.
14A.209
Hazard
to life assessment has been carried out in accordance with Annexes 4 and 22 of
the TM for the proposed work sites at
[1] Environmental
Impact Assessment Study Brief (No. ESB-129/2005)
[2] “Technical
Memorandum on Environmental Impact Assessment Process”, Hong Kong EPD, issued
under Section 16 of the Environmental Impact Assessment Ordinance.
[3] Reeves,
A.B., Minah, F.C. and Chow, V.H.K., ‘Quantitative Risk Assessment Methodology
for LPG Installations’, Conference on Risk & Safety Management in the Gas
Industry, EMSD & HKIE, Hong Kong. (1997)
[4] AIChemE
1989, Guidelines for Chemical Process Quantitative Risk Analysis
[5] EIA
Report for Liquefied Natural Gas (LNG) Receiving Terminal and Associated
Facilities
[6] Comprehensive
Feasibility Study for the Revised Scheme of South
[7] Committee
for the Prevention of Disasters, Guidelines for Quantitative Risk Assessment
“Purple Book”, CPR18E, 2005
[8] L. C. Cadwallader, Compressed
Gas Safety For Experimental Fusion Facilities, 2004.
[9] The
Calculation of Aircraft Crash Risk in the
[10] European Gas Pipeline Incident Data Group, “6th
EGIG Report 1970-2004 Gas Pipeline Incidents”, December 2005.
[11] Approved EIA Report “Proposed Headquarters and Bus Maintenance
Depot in Chai Wan” (EIA-060/2001).
[12] Adrian J. Moore and Alan B. Richards, Golden Pike Cut-back
Flyrock Control and Calibration of A Predictive Model, Terrock Consulting
Engineers.
[13] “Tanks – Reduction of the Risk of a BLEVE”, TNO Report, 2006.
[14]