This section presents the
findings of the hazard assessment undertaken for the Project. The hazard assessment includes an evaluation
of the risk to off-site population of the Project due to the transport and use
of natural gas, and also an evaluation of the risk during construction and
operation phases of the Project due to the transport, storage, and use of other
non-fuel dangerous goods.
The key legislation and
guidelines that are considered relevant to the development of the Project are
as follows:
·
EIA Study Brief (ESB-286/2015), Section
3.4.4 and Appendix B;
·
Environmental Impact
Assessment Ordinance (EIAO), Cap. 499;
·
Technical Memorandum on EIA
Process (EIAO-TM);
·
Hong Kong Planning
Standards and Guidelines (HKPSG), Chapter 12;
·
Gas Safety Ordinance, Cap. 51; and
·
Dangerous Goods Ordinance, Cap. 295.
Section 2 of Annex 4 of EIAO-TM specifies the individual risk
guidelines and societal risk guidelines.
Individual risk is the
predicted increase in the chance of fatality per year to a hypothetical
individual who remains 100% of the time at a given stationary point. The individual risk guidelines require that
the maximum level of off-site individual risk associated with a hazardous
installation should not exceed 1 in 100,000 per year, i.e. 1 × 10-5
per year.
Societal risk expresses the
risks to the population in the vicinity of the hazardous installations. The societal risk guidelines for acceptable
risk levels are presented graphically in Figure 5.1. It is expressed in terms of lines plotting
the frequency (F) of N or more deaths in the population from incidents at the
installation. Two FN risk lines are used
in the Hong Kong Government Risk Guidelines (HKRG) to demark “Acceptable” or
“Unacceptable” regions. In order to
avoid major disasters resulting in more than 1,000 deaths, there is a vertical
cut-off line at the 1,000 fatality level extending down to a frequency of 1 in
a billion years. The intermediate region
indicates the acceptability of societal risk is borderline and should be
reduced to a level which is “As Low As Reasonably
Practicable” (ALARP). It seeks to ensure
that all practicable and cost-effective mitigation measures which can reduce
risks will be considered.
The objective of this
hazard to human life assessment is to assess the risk to life of the off-site
population from the Project during construction and operation phases. The detailed objectives of the assessment
are:
·
To
identify all credible hazardous scenarios associated with transport and use of
natural gas, and transport, storage and use of other non-fuel gas dangerous
goods, with a view to determining a set of relevant scenarios to be included in
a quantitative risk assessment (QRA);
·
To
execute a QRA of the set of relevant hazardous scenarios, expressing population
risks in both individual and societal terms;
·
To
compare individual risk and societal risk with the criteria for evaluating
hazard to life stipulated in Section 2 of Annex 4 of EIAO-TM; and
·
To
identify and assess practicable and cost-effective risk mitigation measures, if
required.
The methodology adopted in
this QRA study is consistent with the approved studies, including both safety
case and EIA studies, listed in the following:
·
DNV,
Safety Case Report for Black Point Gas
Supply Project, Report No. PP019678, Revision No. 2 ([1]) ;
·
ERM,
EIA for Black Point Gas Supply Project,
Revision 3 ([2]);
·
ERM,
EIA for Liquefied Natural Gas (LNG)
Receiving Terminal and Associated Facilities ([3]);
·
ERM,
EIA for 1,800 MW Gas-fired Power Station
at Lamma Extension ([4]);
·
ERM/
BV, EIA for Desalination Plant at Tseung
Kwan O – Feasibility Study ([5]);
·
ERM,
EIA for In-situ Reprovisioning
of Sha Tin Water Treatment Works – South Works – Designs and Construction –
Hazard to Life Assessment ([6]);
·
ERM,
Lamma Power Station Extension: QRA for
Gas Facilities ([7]);
·
ERM,
EIA for South Island Line (East) – Hazard
to Life Assessment ([8]);
and
·
ERM,
EIA for Development of a Biodiesel Plant
at Tseung Kwan O Industrial Estate ([9]).
The elements of this QRA
study are depicted in Figure 5.2, and each of the
elements is described as below.
Relevant data on the
Project and existing facilities such as their layout drawings, design basis,
etc. are provided by CLP. Weather data
and population data in the vicinity of BPPS are collected and reviewed.
Section
5.5 presents detailed information about
the surrounding land, road traffic and marine vessel populations, project time
frame, proposed assessment years and meteorological data. Section
5.6 describes the Project facilities
while Section 5.7 describes the existing BPPS
facilities.
A Hazard Identification
(HAZID) Study including a HAZID Workshop has been conducted to identify all
potential hazards, including both generic and site specific for the
Project. A review of literature and
incident/ accident database was also conducted.
These form the basis for identifying all potential hazardous scenarios,
which were included in this QRA study.
Section
5.8 presents the details of the HAZID.
The frequencies, or the
likelihood, of the various hazardous scenarios resulting from a natural gas or
non-fuel gas dangerous goods release scenario have been derived from historical
database and, where necessary, are modified to take into account local factors.
Section
5.9 presents the details of the frequency
analysis.
Flammable hydrocarbon and
hydrogen, and toxic gas releases were modelled using the PHAST consequence
modelling package developed by DNV.
Fragment range analysis for
a carbon dioxide (CO2) boiling liquid expanding vapour explosion
(BLEVE) was conducted based on the similar approach adopted in the approved EIA
study ([10]).
Section
5.10 presents the details of the
consequence analysis.
The consequence and
frequency data together with weather and population data were subsequently
combined using ERM’s proprietary software, RiskplotTM,
which is in line with the previous QRA studies (1) ([11]) ([12]), ([13]) approved by
EPD, and EMSD respectively. The software
has been independently validated by United Kingdom Health and Safety Executive
(UK HSE) to produce the required risk calculations.
The results from the
cumulative risk assessment were compared with the risk criteria in Section 2 of
Annex 4 of EIAO-TM.
Section
5.11 presents the results for the
cumulative risk assessment including both individual risk and societal risk.
The BPPS area is generally
remote with very low population density in the vicinity. All surrounding population, including land,
road traffic and marine vessel population was considered in the analysis, for
the following proposed assessment years:
·
Case
0: 2016 as the baseline condition year;
·
Case
1a: 2019 as the expected peak impact (in HtLA
context) construction year of the 1st CCGT unit;
·
Case
1b: 2020 as the expected year of operation of the 1st CCGT unit;
·
Case
2a: 2034 as the assumed peak impact (in HtLA context)
construction year of the 2nd CCGT unit; and
·
Case
2b: 2035 as the assumed year of operation of the 2nd CCGT unit (for
worst-case assessment purpose).
The
current plan for implementation of the 1st additional CCGT unit is
from 2016 to 2019; the implementation of the 2nd additional CCGT
unit is expected to be after 2019.
From
the Hazard to Life Assessment (HtLA) point of view,
there is an expectation of increase in surrounding population in the vicinity
of the BPPS; therefore the resulting F-N curve will be on the conservative side
should the assessment year for each scenario be further away from 2016.
Therefore,
the assessment year of 2019 is considered as the expected peak impact
construction year of the 1st CCGT unit from the HtLA
point of view; the societal risk impact is expected to be the most
conservative among the construction period of the 1st
CCGT unit.
The
operation year of the 2nd additional CCGT unit is assumed as 2035,
15 years from the operation year of the 1st additional CCGT unit,
2020. This is considered as a reasonably
worst-case scenario for the purpose of the HtLA. In the same manner as the assumption for the
1st CCGT unit, the assessment year of 2034 is
considered as the expected peak impact construction year of the 2nd
CCGT unit from the HtLA point of view.
The safety management
measures described in Section 5.6 shall be implemented as requirements
to protect construction workers of the Project from the risks associated with
the existing BPPS facilities. With the
safety management measures in place, the potential risks to construction
workers are expected to be low. The
construction workers are not considered as off-site population and are not
taken into account in this QRA study.
The assumptions of
population estimation are summarised in Annex 5A.
Based on a review of aerial
maps, there is no land based population (building development) in the vicinity
of BPPS. The nearest industrial
facilities in Lung Kwu Sheung
Tan are about 1.4 km away and the nearest proposed residential development in
Lung Kwu Tan reclamation area is about 800 m from the
site boundary of BPPS. With reference to
the detailed consequence analysis, all potential hazardous consequences are not
able to reach any land based population (building development) in the vicinity
to BPPS. Therefore, land based
population (building development) will not affect the societal risk levels of
the Project and therefore not considered in this QRA study.
Access to BPPS is via Lung Kwu Tan Road and this is the only road providing access to
villages and industrial sites in this western region of the New Territories.
The population estimation
for Lung Kwu Tan Road is based on 2014 Annual Traffic
Census, which is the latest available road traffic data. The Annual Average Daily Traffic (AADT) value
is 4,170 vehicles per day for station number 5481 from Lung Fai Street to Tsang
Kok. With an
assumed average speed of 50 km hr-1 and an average of three (3)
persons per vehicle, the number of persons on the road was estimated as:
No. of persons = (4,170 × Vehicle Occupancy / 24 / Vehicle
Speed)
= 4,170 × 3 / 24 / 50
= 10.4 persons
km-1
As a conservative approach,
it was assumed the annual traffic growth at Lung Kwu Tan
Road as one (1) % from 2014 to the proposed operation phase year of 2nd
CCGT unit in 2035.
The traffic flow of Yung
Long Road was assumed as 10% of day-time traffic flow from Lung Kwu Tan Road, and 10% of day-time traffic flow was assumed
during the night-time in this QRA study.
Black Point is situated
near Deep Bay. The marine traffic in the
vicinity includes passenger ferries, container ships and river trade vessels
going to Guangzhou and other Pearl River ports.
Small fishing vessels and leisure crafts also contribute to the marine
traffic in the Black Point region.
Marine Vessel Population
The marine vessel
population used in this QRA study are as given in Table
5.1.
The figures are based on BMT’s Marine Impact Assessment report ([14]) except those
for fast ferries. The maximum population
of fast ferries is assumed to be 450, based on the maximum capacity of the
largest ferry operating in the area.
However, the average load factors for fast ferries to Macau and Pearl
River ports are 62% and 47% respectively ([15]). Hence, a distribution in ferry population was
assumed as indicated in Table
5.1.
This distribution gives an overall load factor of about 58% which is
conservative and covers any future increase in vessel population. There is an additional category in the
traffic volume data called “Others”.
These are assumed to be small vessels with a population of 5.
Table 5.1 Marine Vessel Population
|
Type of
Marine Vessel |
Average Population per Vessel |
% of Trips |
|
|
Ocean-Going Vessel |
21 |
|
|
|
Rivertrade Coastal Vessel |
5 |
|
|
|
Fast Ferries |
450 |
(largest ferries with max
population) |
3.75 |
|
|
350 |
(typical ferry with max population) |
3.75 |
|
|
280 |
(typical ferry at 80% capacity) |
22.50 |
|
|
175 |
(typical ferry at 50% capacity) |
52.50 |
|
|
105 |
(typical ferry at 30% capacity) |
12.50 |
|
|
35 |
(typical ferry at 10% capacity) |
5.00 |
|
Tug and Tow |
5 |
|
|
|
Others |
5 |
|
|
Marine Vessel Protection
Factors
The
population on marine vessels is assumed to be provided with some protection
from the vessel structure. The degree of
protection offered depends on factors such as:
·
Size
of vessel;
·
Construction
material and likelihood of secondary fires;
·
Speed
of vessel and hence its exposure time to the flammable cloud;
·
The
proportion of passengers likely to be on deck or in the interior of the vessel;
and
·
The
ability of gas to penetrate into the interior of the vessel and achieve a flammable
mixture.
Small
vessels such as fishing boats provide little protection but larger vessels such
as ocean-going vessels provide greater protection. Fast ferries are air conditioned and have a
limited rate of air exchange with the outside.
Based on these considerations, the fatality probabilities assumed for
each type of marine vessel are as given in Table
5.2.
|
Marine
Vessel Type |
Population |
Fatality Probability |
Population at Risk |
|
Ocean-Going Vessel |
21 |
0.1 |
2 |
|
Rivertrade Coastal Vessel |
5 |
0.3 |
2 |
|
Fast Ferries |
450 |
0.3 |
135 |
|
|
350 |
0.3 |
105 |
|
|
280 |
0.3 |
84 |
|
|
175 |
0.3 |
53 |
|
|
105 |
0.3 |
32 |
|
|
35 |
0.3 |
11 |
|
Tug and Tow |
5 |
0.9 |
5 |
|
Others |
5 |
0.9 |
5 |
Methodology
In this QRA study, the
marine traffic population in the vicinity of BPPS has been considered as both
point receptors and average density values.
The population of all marine vessels is treated as an area average density
except for fast ferries which are treated as point receptors.
The marine area around
Black Point was divided into 12.67 km2 grid cells, each grid being
approximately 3.6 km × 3.6 km. The
transit time for a marine vessel to traverse a grid is calculated based on the
travel distance divided by the marine vessel’s average speed. The average speed ([16]) and transit time for
different vessel types are presented in Table
5.3.
Table 5.3 Average Speed and Transit Time of Different Marine Vessel Type
|
Marine
Vessel Type |
Assumed Speed (m s-1) |
Transit Time (min) |
|
Ocean-Going Vessel |
6.0 |
9.9 |
|
Rivertrade Coastal Vessel |
6.0 |
9.9 |
|
Fast Ferries |
15.0 |
4.0 |
|
Tug and Tow |
2.5 |
23.7 |
|
Others |
6.0 |
9.9 |
The number of marine
vessels traversing each grid daily was provided by the marine study (1). These are given in Table
5.4, where the grid cell reference numbers
are defined according to Figure 5.3. The marine study was based on 2003 data,
extrapolated to years 2011 and 2021, which were used to estimate the marine
traffic population at years 2016, 2019, 2020, 2034 and 2035.
The number of marine
vessels present within each grid cell at any instant in time is then calculated
from:
Number of vessels = No. of
vessels per day × Grid length / 86,400 / Speed
(Equation 1)
This was calculated for
each type of marine vessel, for each grid and for years 2011 and 2021. The values obtained represent the number of
marine vessels present within a grid cell at any instant in time. Values of less than one are interpreted as
the probability of a vessel being present.
Table 5.4 Number of Marine Vessels per Day
|
Grid No. |
Average Number of Marine Vessel per Day |
|
||||||||||
|
2011 |
2021 |
|
||||||||||
|
OG |
RT |
TT |
FF(*) |
OTH |
OG |
RT |
TT |
FF(*) |
OTH |
|||
|
1 |
19 |
788 |
368 |
44 |
567 |
23 |
863 |
403 |
52 |
621 |
||
|
2 |
0 |
0 |
21 |
0 |
84 |
0 |
0 |
23 |
0 |
92 |
||
|
3 |
19 |
557 |
263 |
77 |
294 |
23 |
610 |
288 |
91 |
322 |
||
|
4 |
0 |
368 |
168 |
11 |
294 |
0 |
403 |
184 |
13 |
322 |
||
OG:
Ocean-Going Vessel
RT:
Rivertrade Costal Vessel
TT:
Tug & Tow Vessels
FF:
Fast Ferries
OTH:
Others
(*):
Fast ferries are treated separately
Average Density Approach
The average marine
population for each grid was calculated by combining the number of marine vessels
in each grid as per Equation 1 with
the population at risk for each marine vessel Table
5.2.
The population for the estimated marine populations for the assessment
years is summarised in Table
5.5.
This grid population is assumed to apply to all time periods.
Table 5.5 Estimated Marine Populations for the Assessment Years
|
Year |
2016 |
2019 |
2020 |
2034 |
2035 |
|
Grid No. 1 |
93.8 |
96.5 |
97.3 |
111.5 |
115.1 |
|
Grid No. 2 |
5.4 |
5.6 |
5.6 |
6.4 |
6.6 |
|
Grid No. 3 |
63.2 |
65.1 |
65.7 |
75.4 |
77.9 |
|
Grid No. 4 |
43.0 |
44.2 |
44.6 |
50.9 |
52.4 |
It is noted however that fast
ferries are excluded since ferries are treated separately in the analysis (see
below).
When simulating a possible
release scenario, the impact area is calculated from dispersion modelling. In general, only a fraction of the grid area
is affected and hence the number of fatalities within a grid is calculated
from:
Number
of Fatality = Grid Population × Impact Area / Grid Area
(Equation 2)
Point Receptor Approach
The average density
approach, described above, effectively dilutes the population over the area of
the grid. Given that ferries have a much
higher population than other classes of vessel, combined with a relatively low
presence factor due to their higher speed, the average density approach would
not adequately highlight the impact of fast ferries on the FN curves. Fast ferries are therefore treated
differently in this QRA study.
In reality, if a fast ferry
is affected by an accident scenario, the whole ferry will likely be
affected. The likelihood that the ferry
is affected, however, depends on the size of the hazard area and the density of
ferry vessels. To model this, the
population is treated as a concentrated point receptor, i.e. the entire
population of the ferry is assumed to remain focused at the ferry
location. The ferry density is calculated
the same way as described above (Equation
1), giving the number of ferries per grid at any instant in time, or
equivalent a “presence factor”. A hazard
scenario, however, will not affect a whole grid, but some fraction determined
by the area ratio of the hazard footprint area and the grid area. The presence factor, corrected by this area
ratio is then used to modify the frequency of the hazard scenario.
Probability that ferry is
affected = Presence Factor × Impact Area / Grid Area
(Equation 3)
The fast ferry population
distribution adopted was described in Table
5.1.
Information from the main ferry operators suggested that 25% of ferry
trips take place at night time (between 7 pm and 7 am), while 75% occur during
daytime. Day and night ferries are
therefore assessed separately in the analysis.
The distribution assumed is given in Table
5.6.
Table 5.6 Fast Ferry Population Distribution for Day and Night Time Period
|
Population |
Population at Risk |
% of Day Trip |
% of Night Trip |
% of All Trips (=0.75 × day+0.25 × night) |
|
450 |
135 |
5 |
- |
3.75 |
|
350 |
105 |
5 |
- |
3.75 |
|
280 |
84 |
30 |
- |
22.50 |
|
175 |
53 |
60 |
30 |
52.50 |
|
105 |
32 |
- |
50 |
12.50 |
|
35 |
11 |
- |
20 |
5.00 |
The
ferry presence factor (Equation 1) and
probability that a ferry is affected by a release scenario (Equation 2) are calculated for each
ferry occupancy category and each time period.
Stationary Marine
Population
Other stationary marine
vessel population such as that for the Urmston Road Anchorage area are more
than 1,000 m from BPPS, with reference to the detailed consequence analysis,
all potential hazardous consequences cannot reach other stationary marine
vessel population. Therefore, no other
stationary marine population is taken into account in this QRA study.
The indicative project
programme is that the construction of the Project would be implemented in
stages commencing from the second half of 2016, with commercial operation of
the first unit anticipated by the end of 2019.
The second unit would be constructed and operated after 2019.
Five assessment years were
proposed to assess the risks from the Project and associated facilities, and
the existing BPPS facilities within the site boundary for this QRA study, and
are listed below:
·
2016
as the baseline condition year;
·
2019
as the expected peak impact (in HtLA context)
construction year of the 1st CCGT unit;
·
2020
as the expected year of operation of the 1st CCGT unit;
·
2034
as the assumed peak impact (in HtLA context)
construction year of the 2nd CCGT unit; and
·
2035
as the assumed year of operation of the 2nd CCGT unit.
Data on local meteorological
conditions such as wind speed, wind direction, atmospheric stability class,
temperature, and relative humidity were obtained from the Hong Kong
Observatory.
The annual average
temperature and relative humidity have been taken to be 23.3 °C and 78% respectively
according to the 1981-2010 Normals for Hong Kong ([17]).
The location of weather
station in the vicinity of BPPS is Sha Chau weather station. Data from the Sha Chau weather station were
adopted for this QRA study as this weather station is closest to BPPS and also
the most relevant based on the topography.
The meteorological data used in this QRA study are based on the data
recorded by the Sha Chau weather station over a five-year period from 2010 to
2014.
The raw data from the Hong
Kong Observatory are a series of readings taken hourly over the five-year
period. These data were rationalised
into four combinations of wind speed and atmospheric stability class, denoted
as 2.5B, 3.0D, 7.0D and 2.0F where 2.0F for example refers to a wind speed of
2.0 m s-1 and atmospheric stability class F. The data were then further sorted in twelve
(12) wind directions. This sorting of
meteorological data was performed for the two time periods, day time and night
time respectively.
The fraction of occurrence
for each combination of wind direction, speed and atmospheric stability for
each time period is summarised in Table
5.7.
Wind directions, such as
90°, refer to the direction of the prevailing wind. For example, 90° refers to an easterly wind,
0° is northerly, 180° is southerly and 270° is westerly.
Table 5.7 Data from Sha Chau Weather Station (2010 – 2014)
|
|
Day |
Night |
|||||||
|
Wind speed (m s-1) |
2.5 |
3.0 |
7.0 |
2.0 |
2.5 |
3.0 |
7.0 |
2.0 |
|
|
Atmospheric stability |
B |
D |
D |
F |
B |
D |
D |
F |
|
|
Wind direction |
|
|
|
|
|
|
|
|
|
|
0° |
0.076 |
0.006 |
0.135 |
0.002 |
0.000 |
0.006 |
0.120 |
0.010 |
|
|
30° |
0.011 |
0.004 |
0.058 |
0.002 |
0.000 |
0.006 |
0.098 |
0.009 |
|
|
60° |
0.008 |
0.005 |
0.010 |
0.002 |
0.000 |
0.006 |
0.029 |
0.010 |
|
|
90° |
0.045 |
0.010 |
0.053 |
0.005 |
0.000 |
0.014 |
0.124 |
0.028 |
|
|
120° |
0.073 |
0.007 |
0.146 |
0.004 |
0.000 |
0.008 |
0.247 |
0.023 |
|
|
150° |
0.016 |
0.003 |
0.027 |
0.002 |
0.000 |
0.003 |
0.045 |
0.011 |
|
|
180° |
0.032 |
0.003 |
0.035 |
0.002 |
0.000 |
0.002 |
0.048 |
0.011 |
|
|
210° |
0.079 |
0.006 |
0.066 |
0.004 |
0.000 |
0.003 |
0.093 |
0.014 |
|
|
240° |
0.000 |
0.000 |
0.000 |
0.000 |
0.000 |
0.000 |
0.000 |
0.001 |
|
|
270° |
0.000 |
0.000 |
0.000 |
0.000 |
0.000 |
0.000 |
0.000 |
0.000 |
|
|
300° |
0.003 |
0.000 |
0.000 |
0.000 |
0.000 |
0.000 |
0.000 |
0.001 |
|
|
330° |
0.030 |
0.003 |
0.025 |
0.002 |
0.000 |
0.003 |
0.022 |
0.005 |
|
The Pasquill-Gifford
atmosphere stability classes range from A through F.
A: Turbulent
B: Very unstable
C: Unstable
D: Neutral
E: Stable
F: Very stable
Wind speed and solar radiation
interact to determine the level of atmospheric stability, which in turn
suppresses or enhances the vertical element of turbulent motion. The latter is a function of the vertical
temperature profile in the atmosphere; the greater the rate of decrease in
temperature with height, the greater the level of turbulence.
Class A represents
extremely unstable conditions, which typically occur under conditions of strong
daytime insolation. Class D is neutral
and neither enhances nor suppresses atmospheric turbulence. Class F on the other hand represents stable
conditions, which typically arise on clear nights with little wind.
The Project comprises the
following key new facilities:
·
Turbine
hall building, housing the following major equipment;
·Combined cycle gas turbine;
·Generator;
·Steam turbine;
·Condenser; and
·Associated facilities for
compressed air, oil systems, service water.
·
Electrical
annex (connected to the turbine hall building);
·Control room building; and
·Equipment building.
·
Two
further buildings will be added outside of the turbine hall building and
electrical annex; and
·Water treatment plant
building; and
·Electro-chlorination plant
building.
·
The
following equipment and installations will be required to support the
additional CCGT units but not located in buildings.
·Heat Recovery Steam
Generator (HRSG);
·Generator step-up
transformers;
·Main and auxiliary
transformers;
·Exhaust stack;
·Fuel gas compressors/ fuel
gas conditioning and metering stations;
·Air intake;
·Auxiliary boiler;
·Electrical and electronic
control compartment;
·Effluent handling plants;
·Fuel oil filters/ transfer
pumps; and
·Water treatment/ chemical
sampling plant.
The new proposed units will
utilise a tie-in to the existing natural gas pipelines which serve CCGT units
1-8. Natural gas will be used as the
primary fuel in combustion turbines which drive generators to produce
electricity. The hot exhaust gas from
the gas turbine combustion process is sent through an HRSG in order to extract
the remaining energy out of the combustion exhaust and create steam. The energy stored in the steam is then
extracted via a steam turbine which is also driving the same generator to
produce additional power output.
Fuel gas (natural gas) is
piped from the existing fuel gas headers to new fuel gas conditioning and
metering stations located adjacent to each HRSG. From there the fuel gas pressure and quality
are controlled to meet the required gas turbine parameters set by the gas
turbine supplier. New work as part of
the Project includes tie-ins with two existing fuel gas headers. It is anticipated that there will be new fuel
gas compressors to raise the gas pressure for efficient combustion, as well as
a metering station and filters.
The new gas turbines are
also capable of burning fuel oil (distillate oil) from the existing fuel oil
storage facility as a secondary fuel in the case that natural gas is not
available. New work as part of the Project
includes new fuel oil filters, transfer pumps and piping up to the gas turbines
from the existing fuel oil storage tanks.
The new generators will
utilise hydrogen as a cooling medium during operation. The new units will tap into the existing bulk
hydrogen supply and there will not be any hot tapping operations during the
installation of new CCGT units. When the
generators are not producing electricity, CO2 will be used to purge
them of leftover hydrogen to eliminate the potential explosive atmosphere
during maintenance or outages. The
frequency of purging is expected to be less than once a year. New work as part of the Project includes
piping from the existing bulk hydrogen supply up to the generator hydrogen
control systems.
Cooling water extracted
from the seawater cooling process will be run through an electrochlorination
system in order to produce sodium hypochlorite (NaOCl)
which is used to treat the seawater before it passes through the cooling water
system. This process is continuous and
does not rely on any stored chlorine gas or hypochlorite brought from
off-site. New work as part of this
system will be new electrochlorination modules, dosing pumps, and piping as
required to inject NaOCl into the cooling water
system.
Urea is transported to BPPS
and ammonia is generated on-site by a urea-to-ammonia system to avoid the
hazards of ammonia transportation, transfer and storage. The generated ammonia is fed to the selective
catalytic reaction system to control the level of nitrogen oxide and nitrogen
dioxide to be released to atmosphere.
The layout of the Project
facilities is depicted at Figure 3.2.
To minimise the potential
risks to construction workers on-site during the construction phase of the
Project, construction safety plan will be developed before commencement of
construction and the following key safety management measures will be
implemented.
·
All
construction workers shall comply with CLP’s safety policy and requirements;
·
Method
statements and risk assessments shall be prepared and safety control measures
shall be in place before commencement of work;
·
All
work procedures shall be complied with the operating plant procedures or
guidelines and regulatory requirements;
·
Work
permit system, on-site pre-work risk assessment and emergency response
procedure shall be in place before commencement of work;
·
All
construction workers shall equip with appropriate personal protective equipment
(PPE) when working at the Project Site;
·
Safety
training and briefings shall be provided to all construction workers;
·
All
construction workers shall be under close site supervision; and
·
Regular
site safety inspections shall be conducted during the construction phase of the
Project.
With the implementation of
the above safety management measures, the potential risks to construction
workers on site are expected to be insignificant. The construction workers are not considered
as off-site population and therefore are not taken into account in this QRA
study.
The existing BPPS
facilities include following facilities, and the respective detailed
description for each facility is presented in the following sections.
·
Gas
receiving station (GRS) for submarine pipeline from Yacheng
Field 13-1 & 13-4 (Yacheng);
·
GRS
for submarine pipeline from Dachan Launching Station
in Shenzhen;
·
Fuel
gas facilities including 400 m pipeline from GRS to the existing eight (8)
CCGTs and the associated fuel gas systems;
·
Interconnecting
pipeline between BPPS and Castle Peak Power Station (CPPS);
·
Hydrogen
facilities;
·
CO2
facilities;
·
Distillate
oil facilities; and
·
NaOCl facilities.
The inlet emergency
shutdown valve (ESDV) defines the transition between the submarine pipeline and
the GRS. The inlet shutdown valve (SDV)
is a primary shutdown in the event of a problem within GRS or gas facilities at
BPPS. The valve closes within five (5)
seconds from activation following an emergency shutdown (ESD) or pressure
shutdown (PSD) condition.
A pig receiver receives
maintenance and inspection pigs from the Yacheng
pipeline. A pig receiver opening
interlock (keyed valve) system is included to minimise the potential for
opening the receiver while under pressure.
Redundant pig signals verify pig arrival.
Motorised valves on the
receiver piping will close in the event of fire detection to complete
isolation. It is unusual and undesirable
for liquids to enter the pipeline and pigging to clean the line is not
routine. The pig receiver is designed
for normal brush pigs, as well as instrumented smart pigs. Special connections were used during pipeline
hydro-testing, dewatering and commissioning.
A slug catcher intercepts
liquids from the pipeline for holding, handling and disposal, although in fact
no liquids are expected. If any do
occur, they will be collected in a slop receiver and disposed of by tank truck. The liquids would be condensate
hydrocarbons. It is assumed that no
liquids occur under normal operation for this QRA study. The slug catcher is small at 200 bbl. Liquids recovered in the slug catcher, if
any, will be disposed of by tank truck.
Two parallel inlet gas
filter separators remove traces of liquid mist and solid particles in the
incoming gas. There are vertical
cylinder type separators. Minimal
quantities of collected condensate are anticipated. These separators use replaceable high
efficiency coalescing cartridges for removal of 99% of particles over three (3)
microns in size. Mist carryover and
solids collected by the brush pig are captured.
Filter flow is from the inside of the cartridge to outside with a liquid
dump to the slop system. The same
cartridges are used in the outlet filter/ separators.
Seven (7) water bath gas
heaters warm the natural gas to counteract the temperature drop which occurs on
pressure reduction (Joule-Thompson cooling), and maintain the sale gas at least
28 °C above hydrocarbon dew point. The
heaters are horizontal cylinders of 4 m diameter and 13 m in length. Five (5) heaters are required for full
capacity. Gas flow through individual
heaters is controlled by open/ close inlet block valves. As the sale gas demand increases, inlet block
valves to the heaters are opened in sequence.
Two (2) parallel pressure
reduction stations are available to adjust the delivery pressure to 35 - 41 barg. Each station
comprises two (2) 350 mm trains in parallel for peak flow delivery and one (1)
100 mm train for low flow. Each station
provides 100% capacity, in order to ensure high reliability, and both are
normally kept pressurised to act as back-up in case of valve failure in the
operating train. Each train consists of
a primary pressure control valve, followed by a 14” monitoring (tracking) valve
and a 20” SDV. Individual 18” SDVs are
installed to prevent overpressure by their fast closure times of under one (1)
second, instead of a single 28” valve with slower closure times, in each
train. In addition, low flow
self-contained pressure regulators provide the water bath with stand-by fuel.
Two (2) parallel outlet gas
filter separators remove any liquid formed in the pressure let-down. No liquids should form if the heating system
is operating properly. There are
3-micron thick coalescing filter elements which are identical to those in the
inlet filter/ separators.
Metering and sales
facilities consist of six (6) parallel meter runs. Five (5) are required for maximum flow and
one (1) is spare. All meter streams are
normally kept pressurised. The meter
runs are opened in sequence as the gas flow is increased. Flow is totalled on two (2) independent
metering computers with redundant transmitters.
Flow signals are transmitted to the control room at BPPS. A composite sampler gathers samples for gas
chromatography testing in an independent laboratory.
The outlet SDV defines the
transition between the GRS and the downstream BPPS facilities.
A pig launcher for the
interconnecting pipeline to CPPS was taken for this QRA study as part of the
GRS facilities.
Both open and closed drain
systems are installed in the GRS. The
open drain collects waste water from limited areas under the process equipment
and drains it to the open drain sump.
Waste is transferred to BPPS open drain system by the way of lift pumps
and is primarily for rain water discharge.
The closed drain system collects high pressure hydrocarbons wastes and
drains to the closed drain sump. Liquids
are pumped from the sump to the slop receiver by pump for storage and disposal
by way of tank truck. Vapours are vented
to atmosphere through a flame arrestor.
An elevated vent stack
safely disposes of process blowdown gas, pressure safety valve releases and blowdown gas from the pipeline. The stack is 180 feet high and 28” in
diameter. A CO2 snuffing
system is installed to extinguish fires in cases where the vent stack is struck
by lightning, static electricity or other causes. The system is designed for full pipeline gas
flow during the three (3) seconds closing time of the SDVs. The height limits radiation from an
accidental ignition to a person on the ground to 3,000 BTU per hour-square foot
to under five (5) seconds. As
aforementioned, the existing vent stack will provide a common vent for all two
GRSs to allow depressurisation of equipment.
Gas is received via the Dachan Pipeline and the first major piece of equipment in
the station is an ESDV, which can be closed by means of the station ESD system
in the event of an emergency, isolating the station from the source of natural
gas.
Downstream of the ESDVs is
the pig receiver. This enables the
running of cleaning and inspection pigs in the pipeline. Following the pig receiver are the inlet
filter units, metering runs, heaters and pressure letdown
section where the pressure is reduced to about 40 barg. The natural gas is then sent out to
distribution headers to supply all CCGT units at BPPS. The headers from two GRSs are combined at the
mixing station.
The existing vent stack
provides a common vent for both GRSs to allow depressurisation of equipment in
an emergency.
The maximum gas flow rate
at the manifold is 507,000 m3 hr-1,
while the nominal gas flow to each existing gas turbine is 63,000 m3
hr-1.
Typical operating
conditions are approximately 38 barg and 41 °C.
According to operating
experience over the past years, the variations in these are slight. The designed variation in delivery pressure
is 34 – 41 barg.
The gas conditions remain broadly constant throughout the fuel gas
system. The equipment is designed for 50
barg.
The interconnecting
pipeline is part of the gas transmission system linking the GRS at BPPS and the
Intermediate Pressure Reduction Station (IPRS) at Castle Peak Power Station
(CPPS). The pipeline runs mainly
underground along the Yung Long Road, Lung Kwu Tan
Road, Lung Fai Street and through the CPPS site. It starts at the outlet ESDV from the BPPS
facilities, and ends at the inlet ESDV to the IPRS.
Only the section of the
interconnecting gas pipeline within the site boundary of BPPS was included in
this QRA study for the purpose of cumulative risk assessment.
The hydrogen facilities are
used to unload hydrogen tube trailers/ pallets charged up from an external
supplier. The charged tube trailers/
pallets provide a storage and supply of hydrogen to the generator gas control system
where it is used as a cooling medium in the generator cooling circuit via the
hydrogen distribution system.
Hydrogen is delivered in
tube trailer/ pallets at a maximum pressure of 200 barg
and connected to the station supply manifold using a flexible hose. Replacement of fully-filled hydrogen tube
trailer/ pallet is required before hydrogen is used up.
The current delivery
frequency of hydrogen road trailers to BPPS by an external supplier is about
once per two weeks, depending on the consumption rate of hydrogen at BPPS. It is expected that an additional 20%
hydrogen delivery is required for each of the proposed CCGT units.
The hydrogen trailer bay is
designed and constructed in line with NFPA 50A.
It is an open structure enclosed by a security fence and lockable
gates. Concrete kerbs and guard stop are
provided to prevent the trailer damaging any part of the installation during
trailer reversing or unloading.
Each tube trailer consists of
trailer chassis, container frame, hydrogen cylinders and interconnection
pipework. The chassis is designed in
compliance with the relevant transportation regulations in Hong Kong. The trailer will have the relevant licences
governing its use before it is put into service.
Each chassis has rigid side
guards and rear guards to protect against traffic collision, and is also
equipped with an anti-towaway system.
The trailers have a
container frame, which will accommodate multiple cylinders occupying a total of
11.1 m3 water capacity or 2,115 m3
gas capacity at 200 barg storage pressure.
The hydrogen distribution
system is designed to supply hydrogen from hydrogen trailer(s) and distribute
it to the generator gas control system of each CCGT unit, where it is used as
the rotor cooling medium.
The system supplies the
generator gas control systems with hydrogen for filling of the generators prior
to returning them to services after a generator overhaul/ shutdown which
required purging the generator of hydrogen with CO2/ air.
The system also provides
hydrogen to the generator control system during normal operation at 4 barg to make up for leakage from the generator units.
The system operates at a
normal working pressure of 6.9 barg.
For off-site transport of
hydrogen via trailers, only the road section from the junction of Nim Wan Road/ Yung Long Road to BPPS, approximately 1 km,
was included in this QRA study for the purpose of cumulative risk
assessment. This is selected as Yung
Long Road is dedicated for access to BPPS, and 1 km is an indicative distance
comparable to other approved EIAs where the Consultation Zone boundary of the
Potentially Hazardous Installation was taken into consideration.
The CO2 storage
and distribution system is used to purge the generator of each CCGT unit during
hydrogen filling and emptying operations, and for continuous purging of the
generator gas dryer control cabinets. In
addition, the system also provides CO2 for purging the fuel gas
distribution pipework.
The current delivery
frequency of CO2 road tankers to BPPS by an external supplier is
about once per month, depending on the consumption rate of CO2 at
BPPS. It is expected that an additional
20% CO2 delivery is required for each of the proposed CCGT units.
The CO2 storage
equipment is located outdoors in a compound with easy access to the storage
tank to enable filling from a road tanker.
The equipment is skid mounted for ease of installation.
The CO2 is
stored in a storage tank as a liquid.
The storage tank is provided with 2 × 100% duty refrigeration units.
One (1) of the two (2) main
vaporisation units is selected for duty by the operators at the storage system
local control panel. When large
quantities of CO2 are required for purging, the station operators
command the duty vaporiser to start from the local control panel or from the
distributed control system (DCS). This
command will cause the vaporiser to be powered up and the inlet isolation valve
to open admitting liquid CO2 to the vaporiser. When the CO2 is no longer required
the station operators command the vaporiser to shutdown
from the local control panel or from the DCS.
For off-site transport of
carbon dioxide via road tankers, only the road section from the junction of Nim Wan Road/ Yung Long Road to BPPS, approximately 1 km,
was included in this QRA study for the purpose of cumulative risk
assessment. This is selected as Yung
Long Road is dedicated for access to BPPS, and 1 km is an indicative distance
comparable to other approved EIAs where the Consultation Zone boundary of the
Potentially Hazardous Installation was taken into consideration.
BPPS is a dual fuel
combined cycle power plant, and the primary fuel is natural gas while the
secondary fuel is distillate oil.
The current distillate oil
bunkering and auxiliaries at the Oil Unloading and Heavy Load Berths have a
capacity to facilitate the operation of double berthing.
The distillate oil
unloading system receives distillate oil unloading from oil barge. There are two (2) marine unloading arms
associated with the distillate oil unloading system,
each marine unloading arm is sized to deliver at a flow rate of 1,000 m3
hr-1. The first one is
located at the Oil Unloading Berth while the second one is located at the Heavy
Load Berth.
The existing distillate oil
unloading system, distillate oil bunkering and auxiliaries at oil unloading
facilities are operated only during the distillate oil unloading operation with
frequency of about once every two years.
For each unloading
operation, about 500 to 1,000 tonnes distillate oil is unloaded at either the
Oil Unloading Berth or Heavy Load Berth to distillate oil storage tanks though
fuel oil bunkering line and fuel oil strainer.
The unloaded distillate oil is to replenish the use of distillate oil at
BPPS. Bund wall and oil spill kits are
provided to minimise the risk due to accidental leakage of distillate oil.
There are three (3)
distillate oil storage tanks in the tank farm nearby the Oil Unloading and
Heavy Load Berths. Each of them has a
22,000 m3 capacity which is sufficient for approximately 100 hours
continuous firing of the existing eight (8) CCGT units. This tank farm is surrounded by a bund wall
which can retain any oil leaks within the tank farm area.
The distillate oil transfer
system downstream of the storage tanks consists of pumping equipment and three
(3) forwarding streams.
There are two (2) identical
CCGT forwarding streams, each supplies distillate oil
to one half of BPPS (four (4) CCGT unit liquid fuel modules).
The third (3rd)
stream supplies the essential services diesel generators service tanks and the
auxiliary boiler day tank.
The piping and valve
arrangement around the tanks allows the streams to draw distillate oil from any
one of tanks.
The distillate oil storage
tanks and auxiliary boiler day tank bunds are provided with instruments to
detect the presence of liquid in the bunds.
Cooling water extracted
from the seawater cooling process will run through an electrochlorination
system in order to produce NaOCl which is used to
treat the seawater before it passes through the cooling water system. This process is continuous and does not rely
on any stored chlorine gas or hypochlorite brought from off-site.
Chlorinated seawater,
together with the entrained hydrogen gas, which is a byproduct
of the electrolysis reaction, flows to one of the two hydrogen disentrainment tanks.
Due to the instability of seawater/ NaOCl, the
residence time in the disentrainment tank is
restricted to approximately one (1) hour.
At the inlet to these open topped tanks, the NaOCl
solution weirs over into the tank, causing separation of the entrained hydrogen
gas, which discharges to atmosphere by natural ventilation.
The limited hydrogen byproduct is generated from the electrochlorination reaction, it is dispersed through the open tank top without
additional pressure. Currently,
lightning protection system is provided for the NaOCl
solution storage tanks from which hydrogen is vented to atmosphere. As there is no heat source in the vicinity of
the NaOCl solution storage tanks, there is no chance
that the small amount of hydrogen can be ignited and any risk impact will only
be localised. Considering the points
above, NaOCl facilities in BPPS have not been
modelled in this QRA study.
The safety philosophy is to
ensure containment of hazardous materials used on-site by incorporating various
safety features, maintenance and operation philosophies to minimise any
potential escalation from an accident.
The safety equipment
implemented at BPPS to mitigate against the effect of
any accident can be summarised in the following categories:
·
Emergency
shutdown facilities;
·
Blowdown
facilities;
·
Overpressure
protection facilities;
·
Leak
detection facilities; and
·
Fire
detection and protection facilities.
Emergency
Shutdown Facilities
The inlet ESDV provides
isolation from the submarine pipeline in the event of an operating problem in
the GRS or CLP facilities, including both BPPS and CPPS.
Another ESDV is located
downstream of the pig launcher for the interconnecting pipeline to CPPS. It is closed by a push-button from the GRS
control room.
Process SDVs, located
immediately downstream of the pressure control valves, provide protection of
the downstream facilities from overpressure in the event of pressure control
valve failure.
Unit SDVs are located
upstream and downstream of the main units in the GRS, including on the outlet
line to both BPPS and CPPS. They are
closed by a push-button from the GRS control room.
Plant ESD is initiated
automatically if high or low pressure is detected by two (2) out of three (3)
pressure sensors downstream of the pressure reduction stations. It may also be initiated manually from the
GRS control room or externally on the plant by means of ESD push-buttons.
ESD automatically activates
the inlet ESDV, the interconnecting pipeline ESDV, the process SDVs and the
unit SDVs.
Isolation of the pig
receiver is by motorised valves, which close in the event of fire detection.
Blowdown Facilities
In the event of
overpressure which is not contained by the process SDVs, the system is designed
to vent automatically from a point downstream of the pressure reduction
station.
In the event of an ESD,
each unit on the plant may be blown down to the vent. This is initiated manually from the GRS
control room. The performance of the
blowdown system is assumed to be a reduction of the pressure in each unit to 7 barg within fifteen (15) minutes after initiation.
Blowdown, vent and pressure
relief valve (PRV) released gas is disposed at a vent stack. A CO2 snuffing system is installed
to extinguish the vent if the gas is ignited by lightning or other causes.
Overpressure Protection
Facilities
Overpressure protection on
the downstream piping is provided by:
·
Primary
pressure control valve;
·
Secondary
pressure control valve;
·
The
fast-closing process SDVs;
·
Vent
to atmosphere; and
·
PRVs.
Leak Detection
IR point/ open path flammable
gas detectors are installed at GRS area to provide flammable gas leak detection
capability.
Fire Detection and
Protection Facilities
A 300 mm fire water ring
main for the site is supplied from BPPS.
There are seven (7) fire water monitors, eight (8) street fire hydrants
and twenty-two (22) hose reels on-site.
There are also UV/ IR flame detectors, manual fire alarm system with
audio alarm and nine (9) kg portable powder fire extinguishers.
For the satellite
instrument enclosure (SIE) Room, there are automatic and manual fire alarm and
protection system including very early smoke detection & alarm (VESDA)
system, conventional smoke detectors, heat detectors and manual call points as
well as total gas flooding system (NOVEC 1230).
Portable 5 kg CO2 and 9 litre foam fire extinguishers are
also provided.
Emergency Shutdown
Facilities
Emergency isolation of the
supply pipeline can be achieved by means of:
·
Outlet
SDV on the GRS; and
·
ESDVs
located downstream of the eight (8) filter separators.
Isolation valves are also
provided on the inlet and the eight (8) outlets of the manifold, but these are
not ESDVs and could only be closed on independent actuation by the operator.
The ESDVs will be closed
automatically on fire detection.
Blowdown Facilities
It is necessary that
natural gas can be safely vented at a high level from any section of the system
prior to maintenance on that section. It
is also necessary that all sections can be adequately purged and that a positive
check of the effectiveness of the purging can be performed.
The complete fuel gas
supply (FGS) system is provided with vent and purge points to enable the system
to be purged with a CO2 supply for maintenance. All vent points are suitably earthed to
prevent any danger of electrostatic sparks.
The whole piping system is
electrically trace heated and thermally insulated. This maintains the stagnant fuel gas at the
correct temperature during shutdown ready for restart of the gas turbine.
Any section of the system
on which work is to be carried out must be physically isolated from any section
that is live with natural gas by a double block and bleed system, or
equivalent.
The removal of blanks and
opening of normally closed tested/ purge valves is controlled by written procedures.
Each gas turbine is
contained in an acoustically-lined steel enclosure. The enclosure also serves as containment for
the turbine ventilation and fire protection systems and protects the turbine
from external impact.
The fuel gas module is also
covered by an enclosure. The ventilation
system is designed such that under permanent operations a negative pressure of
5 mm WG is maintained with the enclosure.
Overpressure Protection
Facilities
Overpressure protection on
the downstream piping is provided by:
·
The
primary pressure control valve;
·
The
secondary pressure control valve;
·
The
fast-closing process SDVs;
·
Vent
to atmosphere; and
·
PRVs.
Leak Detection Facilities
The gas leak detector is normally
installed along the flammable gas path above the potential flammable gas leak
sources or inside the exhaust vent ducts to provide flammable gas leak
detection capability.
Fire Detection and
Protection Facilities
The gas turbine modules are
fitted with CO2 fire extinguishing systems.
Fire detectors are provided
in different zones within the gas turbine modules. On alarm by two (2) out of three (3) fire
detector loops in a zone, the turbine and its ventilation are tripped and CO2
is released automatically. Alarms are
indicated on the fire panel in the central control room. The fire protection system can also be
activated from manual call points beside the gas turbine enclosures.
Emergency Shutdown Facility
Hydrogen supply to the
generator gas control system will be suspended in case of emergency by the
following provision:
·
ESDVs;
·
Pipe
rupture safety valves; and
·
Manual
SDVs.
Operation of the ESDV will
be initiated automatically on receiving signals from the following scenarios:
·
Hydrogen
leakage detection in trailer bay;
·
Fire
detection in trailer bay;
·
Manifold
pressure low;
·
Hydrogen
purity low;
·
Oxygen
concentration high in hydrogen gas supply; and
·
Manual
triggering.
Each hydrogen manifold is
equipped with two pipe rupture safety valves arranged in parallel. The two valves will be shutdown automatically
when the supply gas exceeds the design flow rates of 250 Nm3 hr-1
(for generator charging operation) and 10 Nm3 hr-1 (for
normal generator make-up) respectively.
This is to prevent uncontrolled continuous escape of hydrogen in case of
sudden pipe rupture in the distribution manifolds.
In case of malfunction of
the ESDVs during an emergency, hydrogen supply can be isolated by the manually operated
SDV on the manifolds located outside the trailer bay at the boundary fence.
Leak Detection
Facility
Hydrogen detectors are
installed strategically inside the hydrogen trailer bay compound to detect
signs of hydrogen leakage. On detection
of hydrogen with a concentration of 25% lower flammable limit (LFL), flashing
and audible alarm will be initiated at entrance to the area being
monitored. The alarms will also trigger
the ESDV and fire protection water spray system.
Low pressure alarms at
hydrogen distribution system also provide leak detection capability.
Fire Detection and
Protection Facilities
Fire is the primary risk
during the distillate oil unloading process.
Special precautions, therefore, have to be taken to avoid this from
happening. Smoking or naked light are
forbidden in the Oil Unloading and Heavy Load Berths and fire-fighting
equipment should be ready and in good condition. Proper connection of the marine unloading arm
to the oil barge and constant watch during unloading is required.
Distillate oil off-loading
hose connections are ready for the compressed service air. Floodlighting for 24-hour operation and
closed circuit television (CCTV) system are ready to provide clear vision. It is for monitoring the bunkering/
fire-fighting operation at the locations around each marine unloading arm in
front of its pair of foam monitors, and the foam monitor house respectively.
Hazards associated with the
Project facilities have been identified based on a detailed review of known
incident records worldwide and experience gained from operations at similar
facilities. In addition, a systematic
HAZID process including a HAZID Workshop has been conducted to identify any
local or site specific factors.
The major hazards arising
from the Project facilities are mainly related to the loss of containment of
flammable hydrocarbons and hydrogen, and the release of toxic gas such as CO2.
This section includes the
following subsections:
·
Review
of hazardous materials;
·
Review
of the potential major accidental events;
·
Review
of industry incidents relevant to identified hazardous materials;
·
Review
of potential initiating events leading to major accidental events;
·
HAZID
Workshop for the Project facilities; and
·
Development
of hazardous sections.
Natural gas is supplied
from Yacheng Field and Dachan
Launching Station in Shenzhen, and it is a primary fuel in combustion turbines which
drive generators.
Natural gas is composed of
primary methane gas with other fossil fuels such as ethane, propane, butane and
pentane, etc. Natural gas is extremely
flammable when mixed with appropriate concentration of air or oxygen in the
presence of an ignition source.
The compositions of natural
gas supplied from Yacheng pipeline is summarised at Table
5.8. The average gas molecular weight is 19.84 and its
hydrogen sulphide content is less than 50 ppm by volume.
The composition of natural
gas supplied from Dachan Pipeline is given at Table
5.9
and the
total hydrogen sulphide content in the natural gas is less than 20 mg m-3
(about 15 ppm).
Not only is the maximum
surface emissive power of pure methane higher, but also are the consequence
distances for both flash fire and jet fire hazardous scenarios associated with
pure methane larger than that of natural gas.
Therefore, pure methane has been conservatively selected as a
representative material for natural gas in the consequence modelling conducted
by PHAST.
The major hazards arising
from loss of containment of natural gas may lead to hazardous scenarios,
including jet fire, flash fire, fireball, and vapour cloud explosion (VCE).
Table 5.8 The Composition of Natural Gas Supplied from Yacheng
Field
|
Composition |
Molar Percentage (%) |
|
Methane |
84.683 |
|
Ethane |
2.706 |
|
Propane |
0.935 |
|
Butane |
0.458 |
|
Pentane |
0.151 |
|
C6 - C13 |
0.108 |
|
Nitrogen |
0.796 |
|
CO2 |
10.163 |
Table 5.9 The Composition of Natural Gas Supplied from Dachan
Launching Station
|
Composition |
Maximum Molar Percentage (%) |
Minimum Molar Percentage (%) |
|
Methane |
N.A. |
85.0 |
|
Ethane |
10.0 |
N.A. |
|
Propane |
4.0 |
N.A. |
|
Butane |
2.0 |
N.A. |
|
Pentanes + |
0.5 |
N.A. |
|
Nitrogen |
3.0 |
N.A. |
|
CO2 |
3.0 |
N.A. |
The details of each of
dangerous goods related to the Project are analysed as below and summarised at Table
5.10.
Hydrogen
CAS number of hydrogen is
1333-74-0, and hydrogen is a colourless and odourless gas at ambient
temperature and pressure. It has a
boiling point of 253 °C at 1 bar, a critical temperature of -240 °C and a
critical pressure of 13 bara.
Hydrogen is extremely
flammable in oxygen and air, and has the widest range of flammable
concentrations in air among all common gaseous hydrocarbons. This range is between the lower limit of 4%
to an upper limit of 75% by volume.
Because of this wide range, a given volume of hydrogen release will
present a large flammable volume, thus increasing the probability of ignition.
Also, when diluted with
inert gas, hydrogen can still burn with only 5% oxygen. It can be ignited by low energy sources;
hence it is easily ignited by static electricity.
The major hazards arising
from loss of containment of hydrogen may lead to hazardous scenarios, including
jet fire, flash fire, fireball, and VCE.
Distillate Oil
Distillate oil is an
ultra-low sulphur diesel and it is a clear oily liquid. Distillate oil has a relatively high flash
point (> 66 °C), which is above ambient temperature, and a high boiling
point. Thus, evaporation from a liquid
pool is expected to be minimal.
The major hazards arising
from distillate oil are therefore considered to be pool fire, or even flash
fire.
Carbon Dioxide
CAS number of CO2
is 124-38-9, and CO2 is a colourless and odourless gas at ambient
temperature and pressure. In high concentration,
a sharp smell may become apparent. It
has a melting point of -56.6 °C at 1 bar, a critical temperature of 30 °C and a
critical pressure of 74 bara.
When liquid CO2
under pressure is released to atmosphere, the discharge consists of gaseous and
solid CO2 only. When solid CO2
is in direct contact with skin will cause acute cold damage, “cold burn”.
High concentrations of CO2
may cause asphyxiation. Symptoms may
include loss of mobility/ consciousness, victim may not be aware of
asphyxiation. Low concentrations of CO2
can also cause increased respiration and headache.
As per DNV’s CO2RISKMAN 2013 ([18]), a CO2
BLEVE occurs when a very rapid depressurisation of a pressurised liquid of CO2
due to a vessel rupture creates a superheated liquid phase that suddenly
vaporises in an explosive manner. This
may give a transient overpressure peak inside the remaining vessel that further
bursts the vessel creating a shockwave and projectiles.
The major hazards arising
from CO2 are therefore considered to be toxicity effect, and
projectile effect from a CO2 BLEVE, which is explained in Section 5.8.3.
Sodium Hypochlorite
CAS number of NaOCl is 7681-52-9, and NaOCl
solution is a corrosive liquid with the appearance of colourless to yellowish,
and with a chlorine-like odour. NaOCl is not flammable, but it can decompose and release
corrosive chorine gas when it is in contact with acids. NaOCl is produced
by an electrochlorination system at BPPS.
This process is continuous and does not rely on any stored chlorine gas
or hypochlorite brought from off-site.
The major hazard associated
with NaOCl is the possibility of giving off toxic
chlorine gas under the decomposition of the solution. However, once generated on-site, NaOCl is consumed immediately for treatment of seawater
before it passes through the cooling water system. Hence, an accidental mixing of NaOCl with incompatible chemicals such as acids leading to
a toxic chlorine gas release is not expected to occur, and it is not foreseen
to have risk impact from NaOCl solution generated
on-site on off-site population.
Therefore, NaOCl is not taken into account in
this QRA study.
Ammonia
CAS number of ammonia is
7664-41-7, and ammonia is a corrosive alkaline solution and can react with body
tissues. In case of an accidental
exposure, it can severely damage tissues of the mucous membranes and upper
respiratory tract. Symptoms include
burning sensation, coughing, wheezing, laryngitis,
shortness of breath, headaches and vomiting.
In case of inhalation of high concentration ammonia, it may cause
fatality due to spasm inflammation and edema of the
larynx and bronchi, chemical pneumonitis and pulmonary edema.
Nevertheless, ammonia is
barely able to cause any flammable incidents due to its high auto-ignition
temperature at 651 °C and the limited flammable range, from 16 to 25%, by
volume of air.
The major
hazard related to ammonia is its toxicity characteristic, instead of
flammability; however, once generated on-site, ammonia is consumed immediately
to treat the exhaust gas from the Project through selective catalytic reduction
system. Hence, it is not foreseen to
have risk impact from ammonia generated on-site on off-site population. Nevertheless; ammonia detector and monitor
system will be in place as safeguard measures, therefore, ammonia is not taken
into account in this QRA study.
Urea
CAS number of urea is
57-13-6, but urea is not classified as dangerous goods based on “Fire Protection Notice No.4, Dangerous Goods
General” by Fire Services Department ([19]);
therefore, urea is not taken into account in this QRA study.
The existing dangerous
goods delivery routes, depicted in Figure 5.4, are through the Central
Road, which is adjacent to the Project Site.
During the construction
phases of the Project, alternative routes, either through the Workshop Road or
the Transformer Road, can be used for dangerous goods delivery. Alternative routes for dangerous goods
delivery are depicted in Figure 5.5.
The alternative dangerous
goods delivery routes are kept separated from the Project Site to minimise the
impacts from the construction activities within the Project Site to the
dangerous goods during the delivery, and vice versa, during the construction
phases of the Project.
Table 5.10 Dangerous Goods Related associated with the Project
|
Chemical |
Dangerous
Goods Classification* |
Exempt
Quantity |
Location |
Storage
Quantity |
Maximum
Storage Quantity |
Storage Temperature (°C) |
Storage Pressure (barg) |
|
Hydrogen |
Category 2 |
1 Cylinder |
Hydrogen trailer bay |
2 x 64 Cylinders |
128 Cylinders# |
25 |
200.0 |
|
Distillate oil |
Category 5 |
20 L |
Distillate oil tank #1 |
2.2 x 107 L (22,000 m3) |
2.2 x 107 L (22,000 m3) |
25 |
0.0 |
|
Distillate oil |
Category 5 |
20 L |
Distillate oil tank #2 |
2.2 x 107 L (22,000 m3) |
2.2 x 107 L (22,000 m3) |
25 |
0.0 |
|
Distillate oil |
Category 5 |
20 L |
Distillate oil tank #3 |
2.2 x 107 L (22,000 m3) |
2.2 x 107 L (22,000 m3) |
25 |
0.0 |
|
CO2 |
Category 2 |
1 Cylinder |
Station liquid CO2
storage tank |
6 tonnes (tank capacity) |
6 tonnes (tank capacity) |
-18 |
20.7 |
|
NaOCl |
Category 4 |
250 L |
On-site generation at
electrochlorination plant |
- |
- |
- |
- |
|
Ammonia |
Category 2 |
1 Cylinder |
On-site generation at the Project
Site |
- |
- |
- |
- |
Note
*: The dangerous goods
category is classified based on “Fire
Protection Notice No.4, Dangerous Goods General” by Fire Services
Department ([20]).
#: The maximum cylinder
capacity is 26,784 L.
Leakage or rupture
scenarios of process equipment, pipeline or pipework handling flammable gas
such as natural gas or hydrogen can result in a flammable gas cloud, which may be
ignited if it encounters an ignition source while its concentration lies within
the flammable range. In some cases,
static discharge may also cause immediate ignition of flammable gas release.
Distillate oil is heavier
than air and can also form a flammable mixture with air, an ignition of
distillate oil spillage from storage tanks or associated pipework can result in
pool fire or even flash fire depending on circumstances.
In addition, CO2
facilities contain a large amount of pure CO2; leakage or rupture
scenarios from these facilities could result in a toxicity effect,
or even projectile effect from a CO2 BLEVE.
Hazardous scenarios
evaluated in this QRA study include:
·
Jet
fire;
·
Flash
fire;
·
Pool
fire;
·
Fireball;
·
Vapour
cloud explosion;
·
Boiling
liquid expanding vapour explosion; and
·
Toxicity
effect.
The characteristic of the
hazardous scenarios are described separately below:
A jet fire results from an
ignited release of pressurised flammable gas such as natural gas or
hydrogen. The momentum of the release
carries the flammable material forward in a jet, entraining air to give a
flammable mixture, an ignition of which results in a jet flame.
Following the release of
flammable liquid or gas, if there is no immediate ignition, the vapour will
disperse and be diluted as a result of air entrainment. The dispersing vapour cloud may subsequently
come into contact with an ignition source and burn rapidly with a sudden flash.
If the source of flammable
material which created the cloud is still present, then the fire may flash back
to the source giving a jet fire or pool fire.
Direct contact with the burning vapour of a flash fire may cause
fatalities but the short duration of the flash fire means that thermal
radiation effect is not significant outside the flammable gas cloud and thus no
fatalities are expected outside of the flash fire envelope.
A pool fire occurs when a
flammable liquid, such as distillate oil, is split onto the ground and
ignited. The pool formed from the
release initially spread due to the gravitational and surface tension forces
acting on it. As the pool spreads, it
absorbs thermal energy from its surrounding causing evaporation from the pool
surface, resulting in a flash fire if this vapour encounters an ignition
source.
Immediate ignition of
flammable gas release from a full bore rupture of natural gas manifold, or a
catastrophic rupture of a hydrogen road trailer/ a hydrogen trailer at hydrogen
trailer bay, may give rise to a fireball.
Persons caught in the open area within the fireball diameter are assumed
to be 100% fatality.
Flash fire is the most
likely outcome upon an ignition of a dispersing vapour cloud from a flammable
hydrocarbon or hydrogen release. However,
if the vapour cloud accumulates in an area with significant congestion from
surrounding equipment, buildings/ structures, damaging overpressure from a VCE
may result upon an ignition.
Flammable gas cloud
concentration is also one of the important factors to determine whether an
explosion can occur. An explosion with the maximum loading is expected if the
flammable material is sufficiently mixed with air at its stoichiometric
concentration.
A CO2 BLEVE due
to a catastrophic rupture of a CO2 road tanker or a CO2
storage tank was taken into account in this QRA study.
As per DNV’s CO2RISKMAN 2013 ([21]), a CO2
BLEVE occurs when a very rapid depressurisation of a pressurised liquid of CO2
due to a vessel rupture creates a superheated liquid phase that suddenly
vaporises in an explosive manner. This
may give a transient overpressure peak inside the remaining vessel that further
bursts the vessel creating a shockwave and projectiles.
Toxicity effect for a CO2
release is due to a scenario of depleting oxygen in air following a large
amount release of CO2. The
ability of the toxic gas cloud to disperse below dangerous concentrations
strongly depends on wind speed and stability.
The toxic gas cloud will mix and disperse faster in neutral and unstable
wind conditions. Conversely in stable
wind conditions, the toxic gas cloud will remain dense and travel a further
distance.
To investigate further the
possible hazards from the Project facilities, a worldwide review of past
industry incidents at similar facilities was conducted based on following
incident/ accident databases:
MHIDAS database (MHIDAS is
a Major Hazard Incident Data Service developed by the Safety and Reliability
Directorate of the UK Atomic Energy Authority.
MHIDAS contains incidents from over 95 countries particularly the UK,
U.S., Canada, Germany, France and India.
The database allows access to many other
important sources of accident data, such as the Loss Prevention Bulletin, and
is continuously updated) ([22]);
eMARS, which contains reports of
chemical accidents and near misses provided to the Major Accident and Hazards
Bureau (MAHB) of the European Commission’s Join Research Centre from EU, OCED
and UNECE countries ([23]);
ARIA, operated by the
French Ministry of Ecology, Sustainable Development and Energy, lists the
accidental events, which have or could have damaged health or public safety
nature or the environment ([24]);
EGIG 1970 to 2013, which is
a European gas pipeline incident database since 1970, and a valuable and
reliable source of information that is used to establish causes of failure and
failure frequencies in the gas transmission pipeline system ([25]);
·
Energy
Institute;
·
Global
Congress on Process Safety 2013; and
·
US
Gas Pipeline Incident Database, 1984 to 2013 ([26]).
The review of past
incidents/ accidents for similar facilities are summarised in Annex 5B.
The potential hazardous
incidents arising from the Project and the existing facilities are mainly associated
with a loss of containment of flammable hydrocarbons and hydrogen, or a toxic
gas release. An initiating event is
defined as a significant accidental release of hazardous materials beyond its
normal barriers. This could occur as a
result of failure of a major piece of process equipment, pipeline, or
pipework. The various failure modes of
different facilities that could result in a loss of containment are discussed
below.
Natural gas facilities
include additional CCGT units and associated pipework from the Project
facilities, the existing CCGT units, GRSs, and associated natural gas manifolds
and pipework. The major causes for a
loss of containment from natural gas facilities are:
·
Corrosion,
both internal and external;
·
Third
party interference due to work in adjoining areas, etc.;
·
Material
defect;
·
Construction
defect;
·
Operator
errors during operation and maintenance;
·
Defect
caused by pressure cycling; and
·
External
effects.
Hydrogen facilities include
hydrogen road trailers, hydrogen trailers at hydrogen trailer bay, associated
hydrogen manifold and pipework, and generator gas control systems at each of
both additional and existing CCGT units.
The major causes for a loss of containment from hydrogen facilities are:
·
Corrosion,
both internal and external;
·
Material
defect;
·
Construction
defect;
·
Operator
errors during operation and maintenance;
·
Traffic
accidents; and
·
External
effects.
CO2 facilities
include CO2 road tankers, a CO2 storage tank, vaporiser,
and associated CO2 manifold and pipework to CCGT units. The major causes for a loss of containment
from CO2 facilities are:
·
Overfilling;
·
Corrosion,
both internal and external;
·
Material
defect;
·
Construction
defect;
·
Operator
errors during operation and maintenance;
·
Traffic
accidents; and
·
External
effects.
Distillate oil facilities
include a distillate oil unloading system, three distillate oil storage tanks,
an auxiliary boiler day tank, associated pumps and pipework. The major causes for a loss of containment
from distillate oil facilities are:
·
Overfilling;
·
Corrosion,
both internal and external;
·
Material
defect;
·
Construction
defect;
·
Operator
errors during operation and maintenance; and
·
External
effects.
This section outlines these
hazards that are outside the control of BPPS, but could still pose a risk on
the Project and the existing BPPS facilities.
Such hazards are termed as “external hazards” because they are independent
of the process operations on-site but still can lead to major hazard
accidents. These external hazard events
include the following:
·
Earthquake;
·
Subsidence;
·
Lightning;
·
Hill
fire;
·
Storm
surge and flooding;
·
Tsunami;
·
Aircraft
crash; and
·
Helicopter
crash.
Since these events are
independent of the operations, protection against them needs to be incorporated
in the system design. Design criteria
for the protection of hazardous facilities against such events are well
established, and are generally dependent on the geographical locations of the
facilities. Additional criteria could
also be imposed (for instance development of hazardous facilities is subject to
stricter planning regulations in areas of high seismic activity).
Each of these external hazard
events is further analysed in the following section.
Earthquake
An earthquake has the
potential to cause damage to process equipment and pipework. Damage to pipework could be due to ground
movement/ vibration, with guillotine failure of pipework.
Studies by the Geotechnical
Engineering Office ([27])
and Civil Engineering Development Department ([28]) conducted in
the last decades indicate that Hong Kong Special Administrative Region is a
region of low seismicity.
The seismicity in the
vicinity of Hong Kong is considered similar to that of the areas of Central
Europe and the Eastern areas of the U.S. ([29]). As Hong Kong is a region of low seismicity,
an earthquake is an unlikely event. The
generic failure frequencies adopted in this QRA study are based on historical
incidents that include earthquakes in their cause of failure. Since Hong Kong is not at disproportionate
risk from earthquakes compared to other similar facilities worldwide, it is
deemed appropriate to use these generic frequencies without adjustment. There is no need to address earthquakes
separately as they should be already included in the generic failure rates in Section 5.9.
Subsidence
For
subsidence which would result in failure of process equipment or pipework, the
ground movement must be relatively sudden and severe. Normal subsidence events occur gradually over
a period of months and thus appropriate mitigation action can be taken to
prevent failures. In the worst cases,
the BPPS plant could be shutdown and the relevant
equipment isolated and depressurised.
The Project and the existing facilities are built on coastal land with
solid foundation. No undue risk from
subsidence is therefore foreseen and failures due to this should deemed to be included in generic failure frequencies in Section 5.9.
Lightning
Lightning strikes
have led to a number of major accidents worldwide. For example, a contributory cause towards the
major fire at the Texaco refinery in the UK in 1994 was thought to be an
initial lightning strike on process pipework.
The Project
and the existing facilities will be protected with lightning conductors to
safety earth direct lightning strikes.
The grounding will be inspected regularly. The potential for a lightning strike to hit
the Project and the existing facilities and cause a release event is therefore
deemed to be unlikely. Failures due to
lightning strikes should be covered by generic frequencies in Section 5.9.
Hill Fire
Hill fires
are relatively common in Hong Kong, and could potentially occur near the BPPS
site. Recent statistics for these fires
in Hong Kong country parks have been reviewed.
Although the BPPS site is not actually located in a country park, some
of the surrounding terrain and vegetation is similar to those typically found
in country parks. According to
Agriculture, Fisheries and Conservation Department (AFCD) statistics, the
average number of hill fires is 32.4 per year during the five years 2008-2012
(range: 16 to 49). The area affected by
fire each year is available from AFCD annual reports for 2009-2013 at the
following table. These are compared to
the total area of country parks in Hong Kong of 44,004 – 44,239 Ha.
Averaging the data for the
5-year period suggests that 1% of vegetation areas are affected by fire each
year, or equivalently, the frequency of a hill fire affected a specific site is
0.01 per year.
Table 5.11 Hill Fire Data for Hong Kong
|
Year |
Area Affected (Ha) |
Country Park Area (Ha) |
% of Total Country Park Affected |
|
2012 |
79 |
44,239 |
0.18 |
|
2011 |
27 |
44,239 |
0.06 |
|
2010 |
897 |
44,239 |
2.03 |
|
2009 |
275 |
44,004 |
0.62 |
|
2008 |
501 |
44,004 |
1.14 |
BPPS site is protected from
external fires (e.g. hill fire) by a firebreak line in BPPS Landscape Plan and
regular inspection to ensure the firebreak line is properly maintained. Also, BPPS and the route between BPPS and
CPPS are patrolled by station security, who will report to Fire Services
Department and BPPS/ CPPS central control room immediately in case of hill
fires.
Also, there is
no incident record of any hill fire in the vicinity of the BPPS site. Taking into account the low frequency of hill
fire near BPPS and safety management measures are implemented in placed in
BPPS, hill fire leading to impact at BPPS site is considered unlikely and thus
is not taken into account in this QRA study.
Storm Surges and Flooding
If the
pipework becomes submerged under water, it is possible for buoyancy forces to
lift the pipework/storage tanks, causing damage and possible a loss of
containment.
Flooding
from heavy rainfall is not possible due to the coastal location of BPPS
including the Project Site. The slopes
of the natural terrain will channel water to the sea. The primary hazard from typhoons is the storm
surge. Winds, and to a lesser extent
pressure, cause a rise in sea level in coastal areas. In general, storm surges are limited to
several metres.
The Project
and existing BPPS facilities, located at +6 mPD above
sea level, are therefore protected against any risk from storm surges, waves
and other causes of flooding.
As a result,
storm surges and flooding are not taken into account in this QRA study.
Tsunami
Similar to storm surges,
the main hazard from tsunamis is the rise in sea level and possible floatation
of process equipment and pipework. The
highest rise in sea level ever recorded in Hong Kong due to a tsunami was 0.3 m
([30]), and occurred
as a result of the 1960 earthquake in Chile, the largest earthquake ever
recorded in history at magnitude 9.5 on the Richter scale. The Project and the existing BPPS facilities
are approximately at +6 mPD. The effect of a tsunami on BPPS including the
Project Site is therefore considered negligible.
The reason for the low
impact of tsunamis on Hong Kong may be explained by the extended continental
shelf in the South China Sea which effectively dissipates the energy of a
tsunami. Also, the
presence of the Philippine Islands and Taiwan acts as an effective barrier
against seismic activity in the Pacific ([31]). Secondary waves that pass through the Luzon
Strait diffract and lose energy as they traverse the South China Sea.
Seismic
activity with the South China Sea area may also produce tsunamis. Earthquakes on the western coast of Luzon in
the Philippines have produced localised tsunamis but there is no record of any
observable effects in Hong Kong. Also,
with the shelter effects offered by Lantau Island and Hong Kong Island, tsunami
comes from South East of Hong Kong cannot affect the BPPS.
As a result,
tsunami is not taken into account in this QRA study.
Summary of natural hazards
The Project
Site and design of the existing BPPS facilities are such that there will be no
special risks from natural hazards.
Natural hazards are therefore not treated separately in the analysis but
should be included in the generic failure frequencies in Section 5.9.
Aircraft Crash
The BPPS
site does not lie within the flight path of Chek Lap Kok (depicted at Figure 5.6), being about 10 km from
the nearest runway.
The frequency of aircraft
crash was estimated using the methodology of the HSE ([32]). The model takes into account specific factors
such as the target area of the proposed hazard site and its longitudinal (x) and perpendicular (y) distances from the runway threshold (Figure
5.7). The crash frequency per
unit ground area (per km2) is calculated as:
(Equation 4)
where:
N is the number of runway movements per year
R is the probability of an accident per movement (landing or
take-off)
F(x,y) gives the spatial distribution of crashes and is
given by:
Landings
(Equation 5)
for
Take-off
(Equation 6)
for
Equations
5 and Equations 6 are valid only for the
specified range of x values. If x lies
outside this range, the impact probability is zero.
NTSB data ([33]) for fatal
accidents in the U.S. involving scheduled airline flights during the period
1986-2005 are given in Table
5.12.
The 10-year moving average suggests a downward trend with recent years
showing a rate of about 2´10-7 per
flight. However, only 13.5% of accidents
are associated with the approach to landing, 15.8% are associated with take-off
and 4.2% are related to the climb phase of the flight ([34]). The accident frequency for the approach to
landings hence becomes 2.7×10-8 per flight and for take-off/climb
4.0´10-8 per flight. The number of flights at Chek
Lap Kok in 2014 was about 390,000. The number of flights at Chek
Lap Kok in 2035 (the operation phase of 2nd
CCGT unit) was estimated as 685,000 (assumed as linear growth of a period from
1999 to 2014), which is in the same order of magnitude with the estimated
number of flights as 620,000 in 2032 from the approved EIA study for Expansion of Hong Kong International Airport
into a Three-Runway System ([35]).
Table 5.12 U.S. Scheduled Airline Accident Rate
|
Year |
Accident rate per
1,000,000 flights for accidents involving fatalities |
10-year moving
average accident rate per 1,000,000 flights |
|
1986 |
0.14 |
- |
|
1987 |
0.41 |
- |
|
1988 |
0.27 |
- |
|
1989 |
1.10 |
- |
|
1990 |
0.77 |
- |
|
1991 |
0.53 |
- |
|
1992 |
0.53 |
- |
|
1993 |
0.13 |
- |
|
1994 |
0.51 |
- |
|
1995 |
0.12 |
0.451 |
|
1996 |
0.38 |
0.475 |
|
1997 |
0.30 |
0.464 |
|
1998 |
0.09 |
0.446 |
|
1999 |
0.18 |
0.354 |
|
2000 |
0.18 |
0.295 |
|
2001 |
0.19 |
0.261 |
|
2002 |
0.00 |
0.208 |
|
2003 |
0.20 |
0.215 |
|
2004 |
0.09 |
0.173 |
|
2005 |
0.27 |
0.188 |
Considering landings on runway
25L for example, the values for x and
y to the Project Site according to Figure
5.6 are 1.4 km and 11.7 km respectively. Applying Equation
5 gives FL =
5.2×10-5 km-2.
Substituting this into Equation 4
gives:
g(x,y)
= NRF(x,y) = 685,000 / 8 × 2.7 × 10-8 ×
5.2 × 10-5 = 1.2 × 10-7 per km2-year
The number of plane
movements has been divided by eight (8) to take into account the 8 flight
routes in the approved EIA study for Expansion
of Hong Kong International Airport into a Three-Runway System ([36]). This effectively assumes that each runway is
used equally and the wind blows in each direction with equal probability.
The target area is
estimated at 40,000 m2 (0.04 km2). This gives a frequency for crashes into the
Project Site associated with landings on runway 25L as 4.8×10-9 per
year. Repeating the calculation for
landings and take-offs from all runways gives the results shown in Table
5.13.
Table 5.13 Aircraft Crash Frequency onto the Project Site
|
Runway |
Landing (per year) |
Take-off (per year) |
|
07R |
- |
2.4×10-11 |
|
07C |
- |
1.3×10-10 |
|
07L |
- |
No
take off at 07L* |
|
25L |
4.8×10-9 |
- |
|
25C |
No
landing at 25C* |
- |
|
25R |
6.8×10-9 |
- |
|
Total |
1.2×10-8 |
1.6×10-10 |
*:
The preferred operation mode is referred to the approved EIA study (1).
The combined frequency of
all take-off and landing crashes onto the Project Site from activities on all
runways is 1.2×10-8 per year, which has been taken into account in
this QRA study.
Helicopter
Crash
The BPPS site is provided
with a helicopter landing pad although the frequency of use is expected to be
low with perhaps one landing/take-off per week.
The approach, landing and take-off stages of an aircraft flight are
associated with the highest risk and therefore the possible impact of
helicopter crashes on the Project facilities were assessed.
The frequency of helicopter
crash was estimated using the methodology of the HSE ([37]). The model takes into account specific factors
such as the target area of the proposed hazard site. The crash frequency per unit ground area (per
km2) is calculated as:
CHEL = NRHF
(per km2-year)
(Equation 7)
where:
N is annual number of helicopter
movements
RH is helicopter landing/ take-off
reliability
(2.4×10-6 crashes per movement)
F is 29.6 (0 – 100 m) or 0.74 (100 – 200
m)
Considering one flight per
week using the helipad and within 100 m from the helipad, the values for N and F are 52 and 29.6 respectively.
Applying Equation 7 gives CHEL = 3.69x10-3
per km2-year.
Only accidents involving
fatalities were therefore taken into account in this QRA study. 4% of incidents resulted in one or more
fatalities and so the frequency of uncontrolled crashes was calculated as
3.69×10-3 ×0.04 = 1.48×10-4 per km2-year. Similar approach was applied to the impact
range between 100 m and 200 m from the helipad.
The target areas within 100
m and between 100 m and 200 m from the helipad were identified and estimated
based on plot plan to calculate the helicopter crash frequency.
Regarding passing
helicopters, there are no helicopter flight paths in the vicinity of BPPS. The possibility of a passing helicopter
crashing into the Project facilities is therefore much smaller than the generic
failure frequencies used in this QRA study.
Helicopter crashes are therefore not considered separately but are
deemed to be included in the generic failure frequencies in Section 5.9.
A HAZID Workshop for the
Project facilities was conducted at BPPS on 8 September 2015 to identify the
potential hazards from the Project facilities based on HAZID Workshop team’s
specialist opinion, past incidents/ accidents, lessons learned and checklists.
A systematic approach was
adopted, whereby the Project facilities were divided into a number of
“subsystems” based on the layout and the hazards/ keywords from the checklist
applied to each subsystem as relevant.
The key findings from HAZID
Workshop were summarised at Table
5.14.
Table 5.14 HAZID Workshop Worksheets for the Project Facilities
|
System: 1. CCGT Units |
||||||||||||
|
Subsystem: 1. Process
hazards |
||||||||||||
|
Hazards/ Keywords |
Description/ Causes |
Consequences |
Safeguards |
Recommendations |
||||||||
|
|
|
|
|
||||||||
|
|
|
||||||||||
|
|
|||||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
|
|
||||||||||
|
|
|
||||||||||
|
|
|
|
|
||||||||
|
|
|
||||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
|
|
||||||||||
|
|
|
||||||||||
|
|
|
|
|
||||||||
|
|
|||||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
System: 1. CCGT Units |
||||||||||||
|
Subsystem: 2. Natural
hazards |
||||||||||||
|
Hazards/ Keywords |
Description/ Causes |
Consequences |
Safeguards |
Recommendations |
||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
|
|||||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
||||||||||||
|
System: 1. CCGT Units |
||||||||||||
|
Subsystem: 3. External
hazards |
||||||||||||
|
Hazards/ Keywords |
Description/ Causes |
Consequences |
Safeguards |
Recommendations |
||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
System: 1. CCGT Units |
|
||||||||||||
|
Subsystem: 4. Material
hazards |
|
||||||||||||
|
Hazards/ Keywords |
Description/ Causes |
Consequences |
Safeguards |
Recommendations |
|||||||||
|
|
|
|
|
|||||||||
|
|||||||||||||
|
|
|
|
|
|||||||||
|
System: 1. CCGT Units |
||||||||||||
|
Subsystem: 5. Loss of
utilities |
||||||||||||
|
Hazards/ Keywords |
Description/ Causes |
Consequences |
Safeguards |
Recommendations |
||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
System: 1. CCGT Units |
||||||||||||
|
Subsystem: 6. Layout
hazards |
||||||||||||
|
Hazards/ Keywords |
Description/ Causes |
Consequences |
Safeguards |
Recommendations |
||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
System: 1. CCGT Units |
||||||||||||
|
Subsystem: 7. Interface
with existing facility |
||||||||||||
|
Hazards/ Keywords |
Description/ Causes |
Consequences |
Safeguards |
Recommendations |
||||||||
|
|
|
|
|
||||||||
|
System: 1. CCGT Units |
||||||||||||
|
Subsystem: 8.
Construction phase |
||||||||||||
|
Hazards/ Keywords |
Description/ Causes |
Consequences |
Safeguards |
Recommendations |
||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
||||||||||||
|
||||||||||||
|
|
|
|
|
||||||||
|
||||||||||||
Construction activities of
the Project, which may cause potential external hazards on the existing BPPS facilities
and discussed during the HAZID session, are summarised as follows:
·
Movement
of large equipment/ construction vehicles within BPPS to the Project Site;
·
Dropped
object from crane operation;
·
General
construction hazards such as hot work, drilling, etc.; and
·
Tie-in
works onto existing BPPS facilities.
Each of the construction
activities is discussed below:
One of the hazards
identified from the Project construction activities to existing BPPS facilities
is the movement of large equipment/ construction vehicles due to Project works
leading to potential damage of BPPS equipment nearby the Project Site.
Based on the layout plan,
the Project Site (for each phase) is well segregated from existing BPPS equipment. Also, during the construction phases, a
dedicated access route will be used for construction vehicles. Therefore, the chance of vehicle crash impact
on the existing CCGT units is negligible during construction phases.
As for other existing
equipment (e.g. distillate oil pumping house for construction phase of first
new unit, first new CCGT unit for construction phase of second new unit), the
chances of construction vehicle crash impact can be minimised by ensuring speed
limit enforcement being specified in the contractor’s method statement to limit
speed. Speed checks will be conducted to ensure enforcement of speed
limits. A procedure will be developed to
ensure adequate site access control.
Also, lifting plans, with detailed risk assessment, will be prepared and
endorsed for heavy lift of large equipment.
Moreover, the construction
vehicle impact can be mitigated by providing vehicle crash barrier, which is
designed for the specific limited speed between the construction site and the
distillate oil facilities during 1st CCGT unit construction
phase. A crash barrier will also be
provided between the construction site and the operating 1st CCGT
unit facilities during 2nd CCGT unit construction phase.
With the implementation of the
above mitigation measures in place during the construction phases, the
construction vehicle crash hazards can be minimised and managed. The failure due to construction vehicles
should be covered by the generic failure frequencies in Section 5.9.
However, as a conservative
approach, external construction vehicle impact on the distillate oil pumping
house facilities during the construction phase of 1st CCGT unit, and
on the operating 1st CCGT unit facilities during the construction
phase of 2nd CCGT unit, are taken into account in this QRA
study. The worst case scenario approach
for construction vehicle impact was considered, by conservatively assuming one
hundred (100) vehicle movements per day and a running distance of one (1)
kilometre for each vehicle movement within the Project Site.
The fault tree analysis of
the construction vehicles impact for Construction of 1st CCGT unit
and 2nd CCGT unit is summarised in Annex 5C.
Any lifting operation near
or over live equipment may cause damage to the existing BPPS facilities during
the construction phases.
However, the lifting
equipment operation procedure will be specified in the contractor’s method
statement to ensure that any lifting operation near or over live equipment
should be strictly minimized. If such
lifting operation cannot be avoided, lifting activities will be assessed,
controlled, and closely supervised by contractors and operation staff. Also, adequate protection covers will be
provided on the existing BPPS facilities in case the operation of lifting
equipment has a potential to impact live equipment at the BPPS. Process isolation will be achieved in case
that live equipment protection becomes impractical.
A Job Safety
Analysis will be conducted for construction activities of the Project during
the construction phases, to identify and analyse hazards associated with the
construction activities. Also, risk from
the Project and existing facilities will be minimised through the contractor’s
method statement, work permit system, strict supervision and adequate
protection covers on live equipment. The
potential for a dropped object to cause damage on the Project and the existing
facilities and cause a release event is therefore considered included in the
generic leak frequency in Section 5.9.
General construction
activities such as hot works within the Project Site will increase the number
of ignition sources.
The ignition sources
potentially present at the Project Site during the construction phases can be
managed through a hazardous area classification study and updated hazardous
area maps before the start of the construction activities to control ignition sources
during construction phases. Also, work
permit system for hot work activities within the Project Site will be specified
in the contractor’s method statement.
However, as a conservative
approach, the ignition sources potentially present at the Project Site during
the construction phases are taken into account in this QRA study and all
delayed events are considered to be ignited when the consequence of the cloud
can reach the Project Site (in modelling terms, the delayed ignition events
from scenarios when the wind is blowing from northeast, are conservatively
modelled with 100% ignition probability as a worst case scenario).
During the tie-in works
(connection of pipework), the potential hazard is the same as an increased
number of ignition sources on the Project Site, which has been discussed in the
paragraph above. During the
commissioning works (start-up of new CCGT unit), the potential hazards are the
same as those associated with any start-up operations, whose frequency of
occurrence has been included in the generic failure frequencies suggested in Section 5.9.
To follow industrial best
practices for construction works, the procedure for tie-in works onto existing
BPPS equipment network should be reviewed by authorized personnel as
appropriate, and then strictly followed during construction phases via a work
permit system.
The Project and the
existing BPPS facilities were divided into numbers of hazardous sections for
the detailed analysis in this QRA study, based on following consideration:
·
Isolation
valves:
Sections are delimited by ESDVs/ SDVs where applicable. These ESDVs/ SDVs will ensure that the
inventory released from a potential leak is limited to one section; and
·
Process
conditions, such as operating temperature, operating pressure, etc.
The hazardous section lists
for the Project and the existing BPPS facilities are summarised in Table
5.15 and Table
5.16.
Table 5.15 Hazardous Section List for the Project Facilities
|
Code |
Plant
Section Description |
Potential
Hazardous Scenarios |
No. of
items |
Pipe Diameter (mm) |
Length of Section |
Operating Temperature (°C) |
Operating Pressure (barg) |
|
FG_01 |
Manifold
extending from Yacheng fuel gas manifold to
blending station |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
120 |
41 |
38.0 |
|
FG_02 |
Manifold
extending from Dachan fuel gas manifold to blending
station |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
120 |
41 |
38.0 |
|
FG_03 |
Manifold
from blending station for 1st CCGT Unit |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
10 |
41 |
38.0 |
|
FG_04 |
Piping
from manifold to 1st CCGT unit |
Jet fire, flash fire, VCE |
1 |
200 |
120 |
41 |
38.0 |
|
FG_05 |
1st
CCGT unit gas filter separator |
Jet fire, flash fire, VCE |
1 |
200 |
100 |
41 |
38.0 |
|
FG_06 |
1st
CCGT unit fuel gas module |
Jet fire, flash fire, VCE |
1 |
200 |
5 |
41 |
38.0 |
|
FG_07 |
Manifold
for 2nd CCGT |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
95 |
41 |
38.0 |
|
FG_08 |
Piping
from manifold to 2nd CCGT unit |
Jet fire, flash fire, VCE |
1 |
200 |
120 |
41 |
38.0 |
|
FG_09 |
2nd
CCGT unit gas filter separator |
Jet fire, flash fire, VCE |
1 |
200 |
100 |
41 |
38.0 |
|
FG_10 |
2nd
CCGT unit fuel gas module |
Jet fire, flash fire, VCE |
1 |
200 |
5 |
41 |
38.0 |
|
HY_01 |
Additional
hydrogen road trailer for 1st CCGT unit (on-site and off-site
transport) |
Jet
fire, flash fire, fireball, VCE |
1 |
- |
- |
25 |
200.0 |
|
HY_02 |
Hydrogen
manifold for 1st CCGT |
Jet
fire, flash fire, VCE |
2 |
25 |
295 |
25 |
6.9 |
|
HY_03 |
Piping
from hydrogen manifold to 1st CCGT unit |
Jet
fire, flash fire, VCE |
1 |
25 |
115 |
25 |
6.9 |
|
HY_04 |
Additional
hydrogen road trailer for 2nd CCGT unit (on-site and off-site
transport) |
Jet
fire, flash fire, fireball, VCE |
1 |
- |
- |
25 |
200.0 |
|
HY_05 |
Hydrogen
manifold for 2nd CCGT |
Jet
fire, flash fire, VCE |
2 |
25 |
95 |
25 |
6.9 |
|
HY_06 |
Piping
from hydrogen manifold to 2nd CCGT unit |
Jet
fire, flash fire, VCE |
1 |
25 |
115 |
25 |
6.9 |
|
DO_01 |
Distillate
oil manifold for 1st CCGT |
Pool
fire, flash fire |
1 |
350 |
45 |
25 |
3.5 |
|
DO_02 |
Piping
from distillate oil manifold to 1st CCGT unit |
Pool
fire, flash fire |
1 |
200 |
125 |
25 |
3.5 |
|
DO_03 |
Distillate
oil manifold for 2nd CCGT |
Pool
fire, flash fire |
1 |
350 |
95 |
25 |
3.5 |
|
DO_04 |
Piping
from distillate oil manifold to 2nd CCGT unit |
Pool
fire, flash fire |
1 |
200 |
125 |
25 |
3.5 |
|
CD_01 |
Additional
CO2 road tanker to 1st CCGT unit (on-site and off-site
transport) |
CO2
BLEVE, toxicity effect |
1 |
- |
- |
-18 |
20.7 |
|
CD_02 |
CO2
manifold for 1st CCGT |
Toxicity
effect |
1 |
25 |
300 |
30 |
10.0 |
|
CD_03 |
Piping
from CO2 manifold to 1st CCGT unit |
Toxicity
effect |
1 |
25 |
125 |
30 |
10.0 |
|
CD_04 |
Additional
CO2 road tanker to 2nd CCGT unit (on-site and off-site
transport) |
CO2
BLEVE, toxicity effect |
1 |
- |
- |
-18 |
20.7 |
|
CD_05 |
CO2
manifold for 2nd CCGT |
Toxicity
effect |
1 |
25 |
95 |
30 |
10.0 |
|
CD_06 |
Piping
from CO2 manifold to 2nd CCGT unit |
Toxicity
effect |
1 |
25 |
125 |
30 |
10.0 |
Table 5.16 Hazardous Section List for the Existing BPPS Facilities
|
Code |
Plant
Section Description |
Potential
Hazardous Scenarios |
No. of items |
Pipe Diameter (mm) |
Length of Section |
Operating Temperature (°C) |
Operating Pressure (barg) |
|
GRS_01 |
Above
ground piping from shore end to pig receiver of Yacheng
GRS |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
160.0 |
21.0 |
150.0 |
|
GRS_02 |
Piping
from receiver to slug catcher of Yacheng GRS |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
35.0 |
21.0 |
150.0 |
|
GRS_03 |
Piping
from slug catcher to inlet gas filter separators of Yacheng
GRS |
Jet fire, flash fire, fireball, VCE |
2 |
700 |
40.0 |
21.0 |
150.0 |
|
GRS_04 |
Piping
from inlet gas filter separator to gas heater of Yacheng
GRS |
Jet fire, flash fire, fireball, VCE |
7 |
700 |
105.0 |
21.0 |
150.0 |
|
GRS_05 |
Piping
from gas heaters to pressure reduction station, including PRS of Yacheng GRS |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
95.0 |
21.0 |
150.0 |
|
GRS_06 |
Piping
from pressure reduction station to outlet gas filter separator of Yacheng GRS |
Jet fire, flash fire, fireball, VCE |
2 |
700 |
20.0 |
21.0 |
150.0 |
|
GRS_07 |
Piping
from outlet gas filter separator to manifold, including sales gas meter unit
of Yacheng GRS |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
45.0 |
21.0 |
38.0 |
|
GRS_08 |
Pig
receiver of Yacheng GRS |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
8.7 |
21.0 |
150.0 |
|
GRS_11 |
Above
ground piping from shore end to pig receiver of Dachan
GRS |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
70.0 |
19.7 |
54.8 |
|
GRS_12 |
Piping
from receiver to gas filter of Dachan GRS |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
95.0 |
19.7 |
54.8 |
|
GRS_13 |
Filter
& inlet/outlet piping of Dachan GRS |
Jet fire, flash fire, fireball, VCE |
3 |
400 |
15.0 |
19.7 |
54.8 |
|
GRS_14 |
Piping
from filter to metering station of Dachan GRS |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
45.0 |
19.7 |
54.8 |
|
GRS_15 |
Piping
from metering station to heaters, including metering runs of Dachan GRS |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
80.0 |
19.7 |
54.8 |
|
GRS_16 |
Heater
Piping of Dachan GRS |
Jet fire, flash fire, fireball, VCE |
7 |
350 |
40.0 |
19.7 |
54.8 |
|
GRS_17 |
Piping
from heater to PRS, including PRS of Dachan GRS |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
35.0 |
19.7 |
54.8 |
|
GRS_18 |
Piping
from PRS to manifold, including HIPPS of Dachan GRS |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
265.0 |
41.0 |
38.0 |
|
GRS_19 |
Pig
receiver of Dachan GRS |
Jet fire, flash fire, fireball, VCE |
1 |
800 |
8.7 |
19.7 |
54.8 |
|
FG_101 |
Yacheng
manifold |
Jet fire, flash fire, fireball, VCE |
1 |
700 |
200.0 |
41.0 |
38.0 |
|
FG_102 |
Dachan
manifold |
Jet fire, flash fire, fireball, VCE |
1 |
600 |
200.0 |
41.0 |
38.0 |
|
FG_103 |
Piping
from manifold to Unit 1 |
Jet fire, flash fire, VCE |
1 |
200 |
100.0 |
41.0 |
38.0 |
|
FG_104 |
Unit
1 gas filter separator |
Jet fire, flash fire, VCE |
1 |
200 |
60.0 |
41.0 |
38.0 |
|
FG_105 |
Unit
1 fuel gas module |
Jet fire, flash fire, VCE |
1 |
200 |
5.0 |
41.0 |
38.0 |
|
FG_106 |
Piping
from manifold to Unit 2 |
Jet fire, flash fire, VCE |
1 |
200 |
100.0 |
41.0 |
38.0 |
|
FG_107 |
Unit
2 gas filter separator |
Jet fire, flash fire, VCE |
1 |
200 |
60.0 |
41.0 |
38.0 |
|
FG_108 |
Unit
2 fuel gas module |
Jet fire, flash fire, VCE |
1 |
200 |
5.0 |
41.0 |
38.0 |
|
FG_109 |
Piping
from manifold to Unit 3 |
Jet fire, flash fire, VCE |
1 |
200 |
100.0 |
41.0 |
38.0 |
|
FG_110 |
Unit
3 gas filter separator |
Jet fire, flash fire, VCE |
1 |
200 |
70.0 |
41.0 |
38.0 |
|
FG_111 |
Unit
3 fuel gas module |
Jet fire, flash fire, VCE |
1 |
200 |
5.0 |
41.0 |
38.0 |
|
FG_112 |
Piping
from manifold to Unit 4 |
Jet fire, flash fire, VCE |
1 |
200 |
100.0 |
41.0 |
38.0 |
|
FG_113 |
Unit
4 gas filter separator |
Jet fire, flash fire, VCE |
1 |
200 |
70.0 |
41.0 |
38.0 |
|
FG_114 |
Unit
4 fuel gas module |
Jet fire, flash fire, VCE |
1 |
200 |
5.0 |
41.0 |
38.0 |
|
FG_115 |
Piping
from manifold to Unit 5 |
Jet fire, flash fire, VCE |
1 |
200 |
100.0 |
41.0 |
38.0 |
|
FG_116 |
Unit
5 gas filter separator |
Jet fire, flash fire, VCE |
1 |
200 |
60.0 |
41.0 |
38.0 |
|
FG_117 |
Unit
5 fuel gas module |
Jet fire, flash fire, VCE |
1 |
200 |
5.0 |
41.0 |
38.0 |
|
FG_118 |
Piping
from manifold to Unit 6 |
Jet fire, flash fire, VCE |
1 |
200 |
100.0 |
41.0 |
38.0 |
|
FG_119 |
Unit
6 gas filter separator |
Jet fire, flash fire, VCE |
1 |
200 |
60.0 |
41.0 |
38.0 |
|
FG_120 |
Unit
6 fuel gas module |
Jet fire, flash fire, VCE |
1 |
200 |
5.0 |
41.0 |
38.0 |
|
FG_121 |
Piping
from manifold to Unit 7 |
Jet fire, flash fire, VCE |
1 |
200 |
100.0 |
41.0 |
38.0 |
|
FG_122 |
Unit
7 gas filter separator |
Jet fire, flash fire, VCE |
1 |
200 |
60.0 |
41.0 |
38.0 |
|
FG_123 |
Unit
7 fuel gas module |
Jet fire, flash fire, VCE |
1 |
200 |
5.0 |
41.0 |
38.0 |
|
FG_124 |
Piping
from manifold to Unit 8 |
Jet fire, flash fire, VCE |
1 |
200 |
100.0 |
41.0 |
38.0 |
|
FG_125 |
Unit
8 gas filter separator |
Jet fire, flash fire, VCE |
1 |
200 |
60.0 |
41.0 |
38.0 |
|
FG_126 |
Unit
8 fuel gas module |
Jet fire, flash fire, VCE |
1 |
200 |
5.0 |
41.0 |
38.0 |
|
FG_127 |
Interconnecting
gas pipeline from BPPS to CPPS |
Jet fire, flash fire, fireball, VCE |
1 |
600 |
950.0 |
41.0 |
40.0 |
|
HY_11 |
Hydrogen
road trailer for existing gas facilities (on-site and off-site transport) |
Jet fire, flash fire, fireball, VCE |
1 |
- |
- |
25.0 |
200.0 |
|
HY_12 |
Hydrogen
trailer at hydrogen trailer bay |
Jet fire, flash fire, fireball, VCE |
2 |
- |
- |
25.0 |
200.0 |
|
HY_13 |
Trailer
to high pressure piping |
Jet fire, flash fire, VCE |
2 |
25 |
20.0 |
25.0 |
200.0 |
|
HY_14 |
Trailer
to low pressure piping |
Jet fire, flash fire, VCE |
2 |
25 |
20.0 |
25.0 |
6.9 |
|
HY_15 |
Hydrogen
manifold |
Jet fire, flash fire, VCE |
2 |
25 |
496.0 |
25.0 |
6.9 |
|
HY_16 |
Piping
from main supply to Unit 1 |
Jet fire, flash fire, VCE |
1 |
25 |
32.0 |
25.0 |
6.9 |
|
HY_17 |
Piping
from main supply to Unit 2 |
Jet fire, flash fire, VCE |
1 |
25 |
31.0 |
25.0 |
6.9 |
|
HY_18 |
Piping
from main supply to Unit 3 |
Jet fire, flash fire, VCE |
1 |
25 |
43.0 |
25.0 |
6.9 |
|
HY_19 |
Piping
from main supply to Unit 4 |
Jet fire, flash fire, VCE |
1 |
25 |
35.0 |
25.0 |
6.9 |
|
HY_20 |
Piping
from main supply to Unit 5 |
Jet fire, flash fire, VCE |
1 |
25 |
36.0 |
25.0 |
6.9 |
|
HY_21 |
Piping
from main supply to Unit 6 |
Jet fire, flash fire, VCE |
1 |
25 |
36.0 |
25.0 |
6.9 |
|
HY_22 |
Piping
from main supply to Unit 7 |
Jet fire, flash fire, VCE |
1 |
25 |
44.0 |
25.0 |
6.9 |
|
HY_23 |
Piping
from main supply to Unit 8 |
Jet fire, flash fire, VCE |
1 |
25 |
36.0 |
25.0 |
6.9 |
|
DO_11 |
Distillate
oil tank #1 |
Pool
fire, flash fire |
1 |
- |
- |
25.0 |
Atmospheric
pressure |
|
DO_12 |
Distillate
oil tank #2 |
Pool
fire, flash fire |
1 |
- |
- |
25.0 |
Atmospheric
pressure |
|
DO_13 |
Distillate
oil tank #3 |
Pool
fire, flash fire |
1 |
- |
- |
25.0 |
Atmospheric
pressure |
|
DO_14 |
Distillate
oil pumping house |
Pool
fire, flash fire |
8 |
350 |
35.0 |
25.0 |
3.5 |
|
DO_15 |
Distillate
oil manifold section 1 |
Pool
fire, flash fire |
1 |
450 |
10.0 |
25.0 |
3.5 |
|
DO_16 |
Distillate
oil manifold section 2 |
Pool
fire, flash fire |
1 |
350 |
25.0 |
25.0 |
3.5 |
|
DO_17 |
Distillate
oil manifold section 3 |
Pool
fire, flash fire |
1 |
250 |
435.0 |
25.0 |
3.5 |
|
DO_18 |
Piping
from distillate oil manifold to Unit 1 |
Pool
fire, flash fire |
1 |
200 |
75.0 |
25.0 |
3.5 |
|
DO_19 |
Piping
from distillate oil manifold to Unit 2 |
Pool
fire, flash fire |
1 |
200 |
75.0 |
25.0 |
3.5 |
|
DO_20 |
Piping
from distillate oil manifold to Unit 3 |
Pool
fire, flash fire |
1 |
200 |
75.0 |
25.0 |
3.5 |
|
DO_21 |
Piping
from distillate oil manifold to Unit 4 |
Pool
fire, flash fire |
1 |
200 |
75.0 |
25.0 |
3.5 |
|
DO_22 |
Piping
from distillate oil manifold to Unit 5 |
Pool
fire, flash fire |
1 |
200 |
75.0 |
25.0 |
3.5 |
|
DO_23 |
Piping
from distillate oil manifold to Unit 6 |
Pool
fire, flash fire |
1 |
200 |
75.0 |
25.0 |
3.5 |
|
DO_24 |
Piping
from distillate oil manifold to Unit 7 |
Pool
fire, flash fire |
1 |
200 |
75.0 |
25.0 |
3.5 |
|
DO_25 |
Piping
from distillate oil manifold to Unit 8 |
Pool
fire, flash fire |
1 |
200 |
75.0 |
25.0 |
3.5 |
|
DO_26 |
Unloading
arm at oil unloading berth/ heavy load berth |
Pool
fire, flash fire |
1 |
200 |
20.0 |
25.0 |
3.5 |
|
DO_27 |
Fuel
oil bunkering line from unloading arm to distillate oil tanks through fuel
oil strainer |
Pool
fire, flash fire |
1 |
300 |
100.0 |
25.0 |
3.5 |
|
CD_11 |
CO2
road tanker to the existing facilities (on-site and off-site transport) |
CO2
BLEVE, toxicity effect |
1 |
- |
- |
-18.0 |
20.7 |
|
CD_12 |
CO2
storage tank |
CO2
BLEVE, toxicity effect |
1 |
- |
- |
-18.0 |
20.7 |
|
CD_13 |
CO2
manifold section 1 |
Toxicity
effect |
1 |
25 |
20.0 |
30.0 |
10.0 |
|
CD_14 |
CO2
manifold section 2 |
Toxicity
effect |
1 |
25 |
275.0 |
30.0 |
10.0 |
|
CD_15 |
Piping
from CO2 manifold to Unit 1 |
Toxicity
effect |
1 |
25 |
30.0 |
30.0 |
10.0 |
|
CD_16 |
Piping
from CO2 manifold to Unit 2 |
Toxicity
effect |
1 |
25 |
30.0 |
30.0 |
10.0 |
|
CD_17 |
Piping
from CO2 manifold to Unit 3 |
Toxicity
effect |
1 |
25 |
45.0 |
30.0 |
10.0 |
|
CD_18 |
Piping
from CO2 manifold to Unit 4 |
Toxicity
effect |
1 |
25 |
35.0 |
30.0 |
10.0 |
|
CD_19 |
Piping
from CO2 manifold to Unit 5 |
Toxicity
effect |
1 |
25 |
35.0 |
30.0 |
10.0 |
|
CD_20 |
Piping
from CO2 manifold to Unit 6 |
Toxicity
effect |
1 |
25 |
35.0 |
30.0 |
10.0 |
|
CD_21 |
Piping
from CO2 manifold to Unit 7 |
Toxicity
effect |
1 |
25 |
45.0 |
30.0 |
10.0 |
|
CD_22 |
Piping
from CO2 manifold to Unit 8 |
Toxicity
effect |
1 |
25 |
35.0 |
30.0 |
10.0 |
A number of references and the
approved EIA studies ([38])
([39]) ([40]) have been
reviewed for the frequency analysis. The
likelihood or the frequency of outcome events resulting in a flammable or toxic
gas release from process equipment or pipework, and the subsequent potential
hazardous scenarios such as fire, explosion or toxic gas dispersion scenarios
was estimated in this QRA study.
Failure frequencies for
various hazard sources/ equipment were adopted from historical databases in
line with the approved EIA studies, summarised in Table
5.17.
In addition, fault tree analysis was applied to analyse the failure
frequencies associated with hydrogen road trailers and CO2 road
tankers, to take into account the combination of spontaneous failures from
trailer/ tanker and external road traffic accident failures. Release scenarios including a range of hole sizes from a small leak to a full bore rupture or a
catastrophic rupture were assessed in this QRA study.
Table 5.17 Release Event Frequencies
|
Equipment |
Release
Scenario |
Release
Phase |
Release
Frequency |
Unit |
Reference |
|
Pipe size 600 mm to 750 mm |
i) 10 & 25 mm hole |
Liquid/ Gas |
1.00E-07 |
per meter-year |
Hawksley |
|
|
ii) 50 & 100 mm hole |
Liquid/ Gas |
7.00E-08 |
per meter-year |
Hawksley |
|
|
iii) Full bore rupture |
Liquid/ Gas |
3.00E-08 |
per meter-year |
Hawksley |
|
Pipe size 150 mm to 500 mm |
i) 10 & 25 mm hole |
Liquid/ Gas |
3.00E-07 |
per meter-year |
Hawksley |
|
|
ii) 50 & 100 mm hole |
Liquid/ Gas |
1.00E-07 |
per meter-year |
Hawksley |
|
|
iii) Full bore rupture |
Liquid/ Gas |
5.00E-08 |
per meter-year |
Hawksley |
|
Storage Tank (Single-containment tank) |
i) 10 mm hole |
Liquid |
1.00E-04 |
per tank-year |
Purple Book |
|
ii) Catastrophic |
Liquid |
5.00E-06 |
per tank-year |
Purple Book |
|
|
Storage Tank (Double-containment tank) |
i) Catastrophic |
Liquid |
1.25E-08 |
per tank-year |
Purple Book |
|
|
ii) Partial failure* |
Liquid |
1.25E-08 |
per tank-year |
Purple Book |
|
Pressure Vessel |
i) 10 mm hole |
Liquid/ Gas |
1.00E-05 |
per vessel-year |
Purple Book |
|
|
ii) Catastrophic |
Liquid/ Gas |
5.00E-07 |
per vessel-year |
Purple Book |
|
Cylinder |
i) Catastrophic |
Gas |
1.00E-06 |
per cylinder-year |
Purple Book |
|
|
ii) Partial failure* |
Gas |
2.60E-06 |
per cylinder-year |
Journal of hazardous material ([41]) |
|
Road Tanker |
i) Catastrophic |
Liquid/ Gas |
5.00E-07 |
per road tanker-year |
Purple Book |
|
|
ii) Leak# |
Liquid/ Gas |
5.00E-07 |
per road tanker-year |
Purple Book |
|
Pumps |
i) Leak |
Liquid |
1.00E-04 |
per year |
COVO Study |
|
|
ii) Full bore rupture |
Liquid |
1.00E-05 |
per year |
COVO Study |
|
Frequency of spontaneous truck fire |
- |
- |
4.00E-09 |
per truck-km |
Refer
to PHI QRA ([42]) |
|
Truck rollover frequency |
- |
- |
1.90E-07 |
per truck-km |
Refer to PHI QRA |
|
Truck impact frequency |
- |
- |
4.00E-07 |
per truck-km |
Refer to PHI QRA |
|
Conditional probability of vessel rupture in traffic accident |
- |
- |
4.25E-03 |
- |
DNV
QRA Report ([43]) |
|
Conditional probability of large leak on vessel in traffic
accident |
- |
- |
4.00E-03 |
- |
DNV QRA Report |
|
Conditional probability of small leak (all sizes including pipe,
valve etc.) in traffic accident |
- |
- |
1.50E-01 |
|
DNV QRA Report |
|
Driver fails to put out the fire with vehicle fire extinguisher |
- |
- |
5.00E-01 |
per demand |
Fire services fail to prevent
BLEVE ([44]) |
|
Unloading Arm |
i) Leak^ |
Liquid/ Gas |
4.05E-03 |
per year |
COVO Study |
|
|
ii) Full bore rupture |
Liquid/ Gas |
4.05E-05 |
per year |
COVO Study |
|
Truck accident frequency (for construction vehicle impact) |
- |
- |
5.90E-07 |
per truck-km |
Refer to PHI QRA |
|
|
|
|
|
|
|
Note:
*: 10 mm and 25 mm hole
releases are assumed as the representative hole size releases for the partial
failure of the storage tank with double containment tank design and cylinder.
#:
25 mm hole release is assumed as a representative hole
size release for the continuous release from a hole the size of the largest
connection.
^: 10 mm, 25 mm, 50 mm and
100 mm hole releases are assumed as the representative hole size releases for
the leak hole size of the unloading arm.
The failure frequencies for
natural gas facilities were selected from Hawksley ([45]),
who presents his own derived data for both above and below ground pipework in
1984. This historical database was also
suggested in the approved EIA studies ([46])
([47]) for natural
gas facilities.
The failure database
adopted in the Safety Case Report for Black Point Gas Supply Project (Safety
Case Report) is the generic leak frequencies developed by DNV based on the
Hydrocarbon Release Database (HCRD) from UK HSE. The database is based on approximately 4,000
recorded leaks recorded between October 1992 and March 2010 from the UK sector
of the North Sea which has been collected by the UK HSE. In general, the database for offshore
facilities gives higher leak frequencies than most of the onshore sources of
data.
The natural gas facilities
at BPPS mainly consist of both above and underground pipework; therefore, it is
more applicable to adopt failure frequencies suggested by Hawksley (1) in this QRA study.
Moreover, the failure rates
from Hawksley (1) are also compared with
that suggested by UK HSE (4), which was established by
the Hazardous Installations Directorate (HID) C15 of the UK HSE in 2012. The UK HSE failure rates ([48])are intended for use on Land
Use Planning cases and have been in use for several years. After review, it is found that the failure
rates for pipework from Hawksley (1) and UK HSE (4) are in the same order of
magnitude.
The majority of the failure
frequencies were selected from the data suggested by the approved EIA studies,
including, Purple Book ([49]), COVO Study ([50]), as these
international databases were developed for risk assessments for the use,
handling, transport and storage of dangerous substances.
These databases are based on failure data collected from a variety
of onshore facilities including refineries, LPG storage, etc., and other
dangerous substances installations such as chlorine and ammonia storage tanks
with professional judgment applied as appropriate.
The following subsections
provide the discussion on the frequency of other failure events such as
overtopping and bund overflowing.
The possibility of bund
overtopping is considered for tank catastrophic rupture scenarios. The proportion of tank contents that overtop
the bund depends on the geometry, specifically the hydrostatic head of liquid, and
the height of the bund wall and the distance of the bund from the tank.
For the distillate oil
storage site, the likelihood of bund overtopping was therefore estimated based
on the orientation of a release in relation to the nearest bund wall. Directions are considered in quadrants (see Figure
5.8). The probability of a
release being directed towards the nearest bund wall is taken to be 0.25, while
the probability of a leak being directed towards the inner regions of the
impoundment area is taken as 0.75. For
tanks located in the corner of the bund area, the probability of overtopping is
taken as 0.5.
It would, in theory, be
possible to release the contents of more than one tank into a bund at the same
time. A tank fire leading to a
subsequent escalation to other tanks is calculated below:
The frequency of a bund
fire from all causes, including an escalation, can be taken as 1.2×10-5
([51]). The risk at the BPPS site is further reduced
because:
·
Bunds
are equipped with fire detection and protection measures such as CCTV connected
to foam monitor house and a manually operated foam system;
·
Bunds
can be drained to remove flammable/combustible material. This will limit the duration of a bund fire;
and
·
Tank
failure below the liquid level is normally not to be expected during an
external fire, because of the cooling effect from the liquid.
The maximum duration of a
bund fire incident resulting from a tank catastrophic rupture can be estimated
as follows. The worst case of distillate
oil tank rupturing and spilling its entire contents of distillate oil into the
bund with surface area of 9,147 m2 has been considered; the
distillate oil then is ignited and burns without any mitigation measures.
With reference to PHAST ([52]), the value of
burning rate for fuel oil (distillate oil) is 0.035 kg per m2-s. The maximum volume of distillate oil in the
bund is A m3,
corresponding to B kg. Thus the burning time tB can be estimated
as:
tB
= ‘B’ / (0.035 × 9,147) = 1.69 × 107 / (0.035 ×9,147) =52,800
seconds = 14.7 hours
Thus, a bund fire incident
can be expected to last only D (14.7 hours).
The frequency of two (2) out of three (3) tanks within the bund randomly
failing within a day may be estimated from:
3 × 10-4 × 2 ×
10-4 × D / 8,760 × 0.02 = 2.0 × 10-12 per year
where:
10-4 per year is
the leak frequency of a tank
0.02 is the ignition
probability
The resulting frequency of
an overflowing pool fire was estimated as 2.0×10-12 per year, which is much lower than the risk
criteria in Section 2 of Annex 4 of EIAO-TM,
therefore it is not taken into account in this QRA study.
With reference to Purple
Book ([53]), the effect of
blocking valve system is determined by various factors, such as the position of
gas detection monitors and the distribution thereof over the various wind
directions, the direction limit of the detection system, the system reaction
time and the intervention time of an operator.
The probability of failure on demand for an automatic blocking system is
0.001 per demand while that for a hand-operated blocking system is 0.01 per
demand.
However, only GRS and
hydrogen trailer bay areas are protected by an automatic shutdown system, and
the other natural gas facilities and hydrogen distribution system are protected
by a hand-operated shutdown system.
As a conservative approach,
the probability of failure on demand for all detection and shutdown system was
assumed to be 1 in this QRA study.
The ignition probability
depends not only on the presence of ignition sources, but also the release rate
and release duration. Larger releases
are more likely to be ignited than smaller releases. Similarly releases that continue for a longer
duration have a higher probability of ignition than short duration releases.
Table
5.18
summarises
the ignition probabilities adopted in this QRA study for natural gas, which is
consistent with the approved EIA studies ([54])
([55]). The total ignition probability is 0.07 for
small leaks (considered to be 10 mm and 25 mm leaks) and 0.32 for large leaks
and ruptures. These ignition
probabilities are consistent with the model of Cox, Lees and Ang ([56]).
The ignition probabilities
are distributed between immediate ignition and delayed ignition. Delayed ignition is further divided between
delayed ignition 1 and delayed ignition 2 to take into account that a
dispersing gas cloud may be ignited at different points during its
dispersion. Delayed ignition 1 results
in a flash fire and takes into account the possibility that an ignition could
occur within the plant area due to the presence of ignition sources
on-site. Delayed ignition 2 gives a
flash fire after the gas cloud has expanded to its maximum (steady state)
extent.
If both delayed ignition 1
and 2 do not occur, the gas cloud disperses with no
hazardous effect.
Table 5.18 Ignition Probability for Natural Gas
|
|
Immediate Ignition |
Delayed Ignition 1 |
Delayed Ignition 2 |
Delayed Ignition Probability |
Total Ignition Probability |
|
Small leak |
0.02 |
0.045 |
0.005 |
0.05 |
0.07 |
|
Large leak/ rupture |
0.10 |
0.200 |
0.020 |
0.22 |
0.32 |
For isolation failure
scenarios, the delayed ignition probabilities given in Table
5.18
are
doubled. The longer duration and larger
inventory release from a non-isolated release is assumed to make it more likely
that an ignition takes place.
Table
5.19 summarises the ignition probabilities
adopted in this QRA study for distillate oil.
The total ignition probability is 0.03 for small leaks (considered to be
10 mm and 25 mm leaks) and 0.08 for large leaks and ruptures. These ignition probabilities are consistent
with the model of Cox, Lees and Ang (3).
The ignition probabilities
are distributed between immediate ignition and delayed ignition. If a delayed ignition does not occur, the gas
cloud disperses with no hazardous effect.
Table 5.19 Ignition Probability for Distillate Oil
|
|
Immediate Ignition |
Delayed Ignition Probability |
Total Ignition Probability |
|
Small leak |
0.01 |
0.02 |
0.03 |
|
Large leak/ rupture |
0.02 |
0.06 |
0.08 |
For isolation failure scenarios,
the delayed ignition probabilities given in Table
5.19
are
doubled. The longer duration and larger
inventory release from a non-isolated release is assumed to make it more likely
that an ignition takes place.
Hydrogen has a higher
probability of ignition; ignition energy required for hydrogen is one-tenth of
natural gas. Therefore, the immediate
ignition probability of hydrogen can be higher than the ignition probability of
natural gas by a factor of 5.
The approach is consistent
with Cadwallader and Herring ([57]) who give
ignition probabilities for hydrogen fuel vehicles as 0.5 for small spills and
0.9 for large spills. Hydrogen is much
lighter than air and therefore, if not immediately ignited, it will tend to
quickly disperse vertically (as opposed to dense gases). Therefore, the hydrogen cloud is unlikely to
find any ignition source. It should be
nevertheless noted that the risk of static ignition will persist. This risk is however only relevant for a few
minutes as the hydrogen will quickly disperse.
The probability of delayed ignition is therefore conservatively taken as
the same as the probability of immediate ignition.
The ignition probabilities
selected for hydrogen in this QRA study are presented in Table
5.20.
Table 5.20 Ignition Probability for Hydrogen
|
Ignition
Probability |
Small Leak |
Large Leak |
|
Immediate Ignition |
0.5 |
0.9 |
|
Delayed Ignition |
0.5 |
0.9 |
The explosion probability
given an ignition adopted in this QRA study was taken from Cox, Lees and Ang
model ([58]), as shown in Table
5.21.
VCE occurs upon a delayed ignition from a gas release at a congested
area. Since a liquid release is
contained in a potential explosion site (PES), it is conservative to assume an
unignited liquid release vapourises to produce a
flammable vapour cloud, subsequently ignited to produce an explosion.
Table 5.21 Probability of Explosion
|
Leak Size
(Release Rate) |
Explosion Probability |
|
Minor (< 1 kg s-1) |
0.04 |
|
Major (1 – 50 kg s-1) |
0.12 |
|
Massive (> 50 kg s-1) |
0.30 |
An initially small release
may escalate into a larger, more serious event if a jet fire impinges on neighbouring
equipment for an extended time (more than about ten (10) minutes). It is taken into account the modelling for
the isolation fail branch of the event trees for natural gas facilities. If neighbouring piping is within range of the
flame zone of a jet fire, an escalation probability of 1/6 is taken to
conservatively estimate the directional probability and chance of
impingement. Escalation is assumed to
cause a full bore rupture of the affected pipeline/ piping only.
An event tree analysis was
performed to model the development of each event from an initial release to a
final hazardous scenario. The event tree
analysis was considered whether there is immediate ignition, delayed ignition
or no ignition. The possible hazardous
scenarios include jet fire, flash fire, pool fire, toxicity effect, fireball
and VCE.
The event tree analysis for
natural gas, hydrogen, distillate oil and CO2 facilities is depicted
at Figure
5.9, Figure 5.10, Figure 5.11 and Figure
5.12 respectively.
A summary of hazardous
scenarios frequencies considered in this QRA study are summarised in Annex 5C.
This section summarises the
approaches to model the major hazardous scenarios from the continuous and
catastrophic releases considered in this QRA study. Consequence analysis comprises the following
items:
·
Source
term modelling, which involves determining the release rate variation with time
and thermodynamic properties of the released fluids;
·
Physical
effects modelling, which involves estimating the effect zone of the various
hazardous scenarios; and
·
Consequence
end-point criteria, which involves assessing of the impact of hazardous
scenarios on the exposed population during construction and operation phases of
the Project.
Source term modelling was
carried out to determine the maximum (e.g. initial) release rate that may be
expected should a loss of containment occur.
The release rate and
release variation with time were estimated using PHAST. The release rate models used in PHAST are
based on standard orifice type calculations, where the hole
size and process conditions are used to determine the discharge rate. The release rate also depends on the physical
properties of the material being released at the leak location, as well as the hole diameter.
For a guillotine failure of
a pipe or a 100 mm hole size leak scenario, the release rate obtained from the
discharge modelling is found to be more than the normal process flow rate or
maximum pump outflow possible. For such
a situation, the release rate was usually limited to 30% above the maximum pump
throughput (end-of-curve) for a leak from the pipe downstream of a pump. To achieve this, the modelled pressure (i.e.
system pressure at initial state) were reduced iteratively so that the desired
release rate is achieved for the given process conditions.
With reference to Purple
Book ([59]), the closing time of an
automatic blocking system is two (2) minutes while that of a hand-operated
blocking system is thirty (30) minutes.
The closing time of two (2)
minutes for an automatic blocking system has been verified by CLP based on
their solid operation experience and is deemed applicable to the existing BPPS
facilities.
Detection and shutdown
system may however fail due to some reasons, also as per Purple Book (1), the release duration is
limited to a maximum of thirty (30) minutes.
The release duration of thirty (30) minutes was conservatively assumed
and adopted in this QRA study.
The orientation of a
release can have some effects on the hazard footprint calculated by PHAST. The models take into account the momentum of
the release, air entrainment, vaporisation rate and liquid rainout
fraction.
For a horizontal,
non-impinging release, momentum effects tend to dominate for most releases
giving a jet fire as the most serious outcome.
If a release is vertically upwards, the hazard footprint will be
significantly less compared to a horizontal release. Also, if a release impinges on the ground or
other obstacles, the momentum of the release and air entrainment is reduced,
thereby reducing the hazard footprint but also increasing the liquid rainout
fraction. In this scenario, a pool fire
becomes more likely.
Therefore, for all pool
fire scenarios, the release orientation was set to “downward impinging release” in order to obtain the worst case
consequence pool fire; while “horizontal non-impinging” was selected for
modelling other fire effects such as jet fire and flash fire, and toxicity
effect, as a conservative approach.
The physical effects
modelling by PHAST assessed the effects zones for the following hazardous
scenarios in this QRA study:
·
Jet
fire;
·
Flash
fire;
·
Pool
fire;
·
Fireball;
·
Vapour
cloud explosion; and
·
Toxicity
effect.
Each of these events is
further analysed in the following section.
A jet fire results from an
ignited release of the pressurised flammable gas. The momentum of the release carries the
flammable materials forward in form of a long plume entraining air to give a
flammable mixture. Combustion in a jet
fire occurs in the form of a strong turbulent diffusion flame that is strongly
influenced by the momentum of the release.
A
jet fire was modelled for a pressurised flammable gas release. The default jet fire correlation model in PHAST was selected, and the release
orientation was set as a horizontal non-impinging release.
If there is no immediate
ignition, the flammable gas such as natural gas and hydrogen may disperse before
subsequently encountering an ignition source giving a jet fire or pool
fire. The vapour cloud will then burn
with a flash back to the source of the leak.
A flash fire is assumed to be fatal to anyone caught within the flash
fire envelope, although the short duration of a flash fire means that radiation
effects are negligible. The fatality
probability is therefore zero for persons outside the flash fire envelope.
Dispersion modelling was
conducted by PHAST to calculate the extent of the flammable vapour cloud. This takes into account both the direct
vaporisation from the release, and also the vapour formed from evaporating
pools. The extent of the flash fire was
assumed to be the dispersion distance to LFL in this QRA study.
In case of an early
ignition of a liquid pool, an early pool fire will be formed and the maximum
pool diameter can be obtained by matching the burning rate with the release
rate. Under such a condition, the size
of the pool fire will not increase further and will be steady. In case of a delay ignition, the maximum pool
radius is reached when the pool thickness at the centre of the pool reaches the
maximum thickness.
Immediate ignition of
release caused by a catastrophic rupture of process equipment or a full bore
rupture of a pipeline may give rise to a fireball. The consequence analysis for a fireball
scenario was conducted by Roberts (HSE) method ([60]) in PHAST as
the calculation method.
The flammable mass for
fireball modelling was conservatively estimated by the initial flow rate
continuing for ten (10) seconds even though the initial release rate is
decreasing rapidly in case of a full bore rupture scenario of a pipeline. This approach is consistent with the approved
study, Safety Case Report ([61]).
Persons caught in the open
area within the fireball diameter have been assumed to be 100% fatality.
Explosions may only occur
in areas of high congestion, or high confinement. Ignition in the open may only result in a
flash fire or an unconfined VCE yielding relatively a lower damaging
overpressure.
When a large amount of
flammable gas is rapidly released, a vapour cloud forms and disperses in the
surrounding air. The release can occur
from natural gas manifold, GRS, etc. If
this cloud is ignited before the cloud is diluted below its LFL, a VCE or flash
fire will occur. The main consequence of
a VCE is damage to buildings/ structures due to an overpressure while the main
consequence of a flash fire is a direct flame contact. The resulting outcome, either a flash fire or
a VCE depends on a number of parameters.
Pietersen and Huerta (1985) (1) has summarised some key features
of 80 flash fires and AIChE/CCPS (2000) ([62]) provides an excellent summary
of vapour cloud behaviour. They describe
four features which must be present in order for a VCE to occur. First, the release material must be flammable. Second, a cloud of sufficient size must form
prior to an ignition, with ignition delays of 1 to 5 minutes considered the
most probable for generating VCEs.
Lenoir and Davenport (1992) (1) analysed historical data
on ignition delays, and found delay times from six (6) seconds to as long as
sixty (60) minutes. Third, a sufficient
amount of the cloud must be within the flammable range. Fourth, sufficient confinement or turbulent
mixing of a portion of the vapour cloud must be present.
The blast effects produced
depend on whether a deflagration or detonation results, with a deflagration
being, by far, the most likely. A
transition from deflagration to detonation is unlikely in the open air. The ability for an explosion to result in a
detonation is also dependent on the energy of the ignition source, with larger
ignition sources increasing the likelihood of a direct detonation.
In order to calculate the
distances to given overpressures, the Baker-Strehlow-Tang
(BST) model ([63]),
which is a congestion based model, was adopted in this QRA study. The volume of flammable material in congested
areas was estimated as well as the flame expansion characteristics, and then
the BST model predicts the overpressures at a given distance. The BST model predicts the blast levels based
on:
·
Mass
of flammable material involved in an explosion (determined based on dispersion
modelling by PHAST);
·
Reactivity
of the flammable material (high, medium, or low)
·
Degree
of freedom for the flame expansion (1D, 2D, 2.5D or 3D); and
·
Congestion
level of a PES (high, medium, low).
To apply the BST model, the
BPPS plant was divided into various PESs based on the equipment layout. A leak from each hazardous section of the
Project and the existing BPPS facilities is then assumed to cause an explosion
in the nearest PES.
To calculate the
overpressures from a VCE scenario, dispersion modelling was performed by PHAST
to obtain the mass of flammable gas within the flammability limits. Then, the volume of the flammable vapour
cloud from PHAST was compared with the volume of each of the PESs. If the vapour cloud is larger than the PES
volume, the mass of material involved in the explosion was assumed to be
limited by the PES volume. A
stoichiometric mixture was assumed throughout the PES volume in this case. For each explosion at each of PESs, the
explosion centre was taken as the centre of the PES area.
Similar to thermal
radiation levels, overpressure levels, corresponding to specific fatality
levels, were taken from the data published by Purple Book ([64]) for indoor/
outdoor population. The various
overpressure levels considered in this QRA study are presented in Section 5.10.3.
Figure 5.13 depicts the identified ten
(10) PESs based on a review of the BPPS plot plan, and Table
5.22 lists the input parameters, such as
level of congestion, reactivity of material, etc., to the BST model performed
by PHAST.
Table 5.22 Identified PESs at BPPS
|
Tag |
Facility
Description |
Reactivity
of Material |
Degree of
Freedom for Flame Expansion |
Level of
Congestion |
Length (m) |
Width (m) |
Height (m) |
Estimated PES Volume (m3) |
|
PES 1 |
Fuel supply pipework under HRSG of
Unit 1 |
Low |
2D |
Medium |
47.4 |
29.9 |
3 |
4,252 |
|
PES 2 |
Fuel supply pipework under HRSG of
Unit 2 |
Low |
2D |
Medium |
47.4 |
29.9 |
3 |
4,252 |
|
PES 3 |
Fuel supply pipework under HRSG of Unit
3 |
Low |
2D |
Medium |
47.4 |
29.9 |
3 |
4,252 |
|
PES 4 |
Fuel supply pipework under HRSG of
Unit 4 |
Low |
2D |
Medium |
47.4 |
29.9 |
3 |
4,252 |
|
PES 5 |
Fuel supply pipework under HRSG of
Unit 5 |
Low |
2D |
Medium |
47.4 |
29.9 |
3 |
4,252 |
|
PES 6 |
Fuel supply pipework under HRSG of
Unit 6 |
Low |
2D |
Medium |
47.4 |
29.9 |
3 |
4,252 |
|
PES 7 |
Fuel supply pipework under HRSG of
Unit 7 |
Low |
2D |
Medium |
47.4 |
29.9 |
3 |
4,252 |
|
PES 8 |
Fuel supply pipework under HRSG of
Unit 8 |
Low |
2D |
Medium |
47.4 |
29.9 |
3 |
4,252 |
|
PES 9 |
Fuel supply pipework under HRSG of 1st
CCGT Unit |
Low |
2D |
Medium |
54.8 |
17.5 |
3 |
2,877 |
|
PES 10 |
Fuel supply pipework under HRSG of 2nd
CCGT Unit |
Low |
2D |
Medium |
54.8 |
17.5 |
3 |
2,877 |
Dispersion modelling was
employed by PHAST to calculate the extent of the toxic gas cloud. This takes into account both the direct vapourisation from the release, and also the vapour formed
from evaporating pools. In this QRA
study, the extent of the toxic gas cloud was based on the dispersion distance
to the various thresholds, corresponding to the different CO2
toxicity levels described in the Section 5.10.3.
As discussed in the
approved EIA study ([65]),
a catastrophic rupture of a CO2 storage tank or a CO2 road
tanker, due to either a spontaneous failure or a CO2 BLEVE, may
produce fragments that can cause fatalities hundreds of metres away.
Fragment Range
The fragment range
estimation approach is consistent with the approved EIA study ([66]). The initial velocity of the fragment must be
determined before estimating the range of the fragment. The fragment range can be estimated by
solving equations for trajectory motions, allowing for the effects of drag.
Several models have been
proposed to estimate the initial velocity of a fragment. The Brode and Baum
models described by Lees (1996) ([67])
and the Centre for Chemical Process Safety (CCPS, 1999) ([68]) were proposed
in this QRA study to estimate the fragment range associated with a CO2
BLEVE.
Brode ([69]) assumed that a
portion of the total internal energy of a storage vessel is translated into
kinetic energy of the fragment:
where:
Ek is a kinetic energy (J)
M is the total mass of the empty cylinder (kg)
k is
a fraction of internal energy converted to fragment kinetic energy
p1 is an absolute pressure in cylinder (Pa)
po is an ambient pressure (Pa)
V is the volume of the vessel (m3)
γ is a ratio of specific heats of the gas
The weakness of the Brode model ([70])
is that it does not take into account the fragment size.
Baum ([71]) developed
empirical correlations for the initial velocity for an end cap breaking away
from a cylindrical vessel, a cylindrical vessel breaking into two (2) parts,
and disintegration of a spherical or cylindrical vessel into multiple
fragments. As an example, the model for
end cap breaking into two (2) parts is presented in the equations below. The model for a cylindrical vessel breaking
into two (2) parts and disintegration of vessel into multiple fragments are
similar.
The correlation for the
initial velocity:
where:
F is a parameter for calculation of initial velocity
A is an area of the detached portion of the cylinder wall (m2)
r is
a radius of the vessel (m)
Mf is a mass of fragment (kg)
ao is a speed of sound (m s-1)
T is a temperature of the gas inside cylinder at failure (K)
R is an ideal gas constant (J kmol-1 K-1)
m is
a molecular mass of cylinder contents (kg kmol-1)
However, the mass of
fragment from a CO2 BLEVE is not available from literature,
the range of projectile is not possible to estimate based on the above
equations.
Nevertheless, another study
by Holden and Reeves ([72])
reviewed twenty seven (27) events involving cylindrical vessels ([73]). The events are classified into two (2)
groups; vessels above 90 m3 and vessels below 90 m3. It was reported that 80% of the fragments
travelled less than 200 m and the maximum range was identified to be 500 m for
the vessels below 90 m3. The
study also reported that fragment ranges are much less than the maximum range
from theoretical estimates.
Tasneem Abbasi
et al. ([74]) reported two
(2) events about fragment from cylinder vessels. The most relevant datum is for a cylindrical
vessel containing 35,000 kg of liquid CO2, which gives a maximum
range of fragments of less than 300 m.
With reference to the above
incident databases, the maximum fragment range for a CO2 road tanker
or a CO2 storage tank at BPPS is conservatively assumed to be 500 m.
Shielding Factors
No shielding factor is
conservatively applied to the road traffic population on Nim
Wan Road and Yung Long Road, and the surrounding marine vessel population,
which is in line with the approved EIA study ([75]).
Target Area
If a fragment strikes a
critical part of the body such as head or torso, a fatality may result. Assuming a cylindrical shaped torso, about 1
m long and 12” diameter (the chest size for the average Chinese male is 35”
circumference, Alvanon, 2008) (4), gives a target area of
0.3 m2.
The probability of a road
traffic vehicle or a marine vessel being hit is estimated by taking into
account the exposure factors and the fraction of fragment that have sufficient
range to reach the road traffic vehicle or marine vessel.
The probability of a person
at a road traffic vehicle or a marine vessel being struck by a fragment is
therefore calculated based on:
·
The
probability that the road traffic vehicle or marine vessel will be hit, based
on the fraction of fragment that have sufficient range to reach the road
traffic vehicle or marine vessel;
·
A
road traffic vehicle target area of 8 m2, the typical size of
private vehicle;
·
A
marine vessel target area of 900 m2, the typical size of Rivertrade Coastal Vessel; and
·
A
person target area of 0.3 m2.
The number of road traffic
vehicles within 500 m radius from a CO2 road tanker at the junction
of Nim Wan Road and Yung Long Road to BPPS is
estimated as 34 in 2035, while the number of marine vessels within 500 m radius
from a CO2 storage tank at BPPS is estimated as 42 in 2035.
After taking into account
these factors, the probability of hitting a road traffic vehicle or a marine vessel
to cause one fatality was calculated as 7.8×10-13, which is much lower than 1×10-9 per year, the lowest
frequency considered as per risk criteria in Section 2 of Annex 4 of EIAO-TM.
Therefore, it is not taken into account in this QRA study.
The estimation of the
fatality/ injury caused by a physical effect such as thermal radiation requires
the use of probit equations, which describe the
probability of fatality as a function of some physical effects. The probit equation
takes the general form:
Y = a + b ln V
where:
Y is the probit
a,
b are constants determined from experiments
V is a measure of the physical effect such as thermal dose
The probit
is an alternative way of expressing the probability of fatality and is derived
from a statistical transformation of the probability of fatality.
The following probit equation ([76])
is used to determine impacts of thermal radiation from a jet fire, pool fire or
fireball to persons unprotected by clothing.
Y = -36.38 + 2.56 ln (t I 4/3)
where:
Y is the probit
I is the radiant thermal flux (W m-2)
t is
duration of exposure (s)
This equation gives the
thermal radiation levels presented in Table
5.23, assuming a 20-second exposure
time. For areas lying between any two
radiation flux contours, the equivalent fatality level
is estimated as follows:
·
For
areas beyond the 50% fatality contour, the equivalent fatality is calculated
using a 2/3 weighting towards the lower contour. For example, the equivalent fatality between
the 1% and 50% contours is calculated as 2/3 × 1 + 1/3 × 50 = 17%; and
·
For
areas within the 50% contour, the equivalent fatality is calculated with a 2/3
weighting towards the upper contour. For
example, the equivalent fatality between the 90% and 50% contours is calculated
as 2/3 × 90 + 1/3 × 50 = 77%.
The
different approach above and below the 50% fatality contour is due to the
sigmoid shape of the probit function.
Table 5.23 Levels of Harm for 20-second Exposure Time to Heat Fluxes
|
Incident Thermal Flux
(kWm-2) |
Fatality Probability for
20-second Exposure Time |
Equivalent Fatality
Probability for Area between Radiation Flux Contours |
|
|
9.8 |
1% |
} } } |
17% |
|
19.5 |
50% |
77% |
|
|
28.3 35.5 |
90% 99.9% |
97% |
|
With regard to a flash fire,
the criterion chosen is that a 100% fatality is assumed for any person outdoors
within the flash fire envelope. The
extent of the flash fire is assumed to be the dispersion to its LFL.
For an explosion, a
relatively high overpressure is necessary to lead to significant fatalities for
persons outdoors. Persons indoor have a
high harm probability due to the risk of building collapse and flying debris
such as breaking windows. Table
5.24 presents the explosion overpressure
levels suggested by Purple Book ([77]).
Table 5.24 Effect of Overpressure - Purple Book
|
Explosion
Overpressure (barg) |
Fraction of People Dying |
|
|
|
Indoor |
Outdoor |
|
> 0.3 |
1.000 |
1 |
|
> 0.1 to 0.3 |
0.025 |
0 |
Fatality rates due to a
toxic exposure are determined by the probit function,
as follows:
Y = a + b × ln (Cn
× t)
where:
Y is the probit corresponding
to the probability of death
a,
b, n are the constants describing
the toxicity of a substance
C is the concentration of the toxic material (ppm)
t is
duration of exposure (minutes)
As can be seen from the probit equation, the fatality due to a toxic exposure
depends on the exposed duration and the toxicity of the materials. Similar to thermal radiation, toxic cloud
contours corresponding to 90%, 50% and 1% fatality was obtained from PHAST. With reference to Purple Book (1), the exposure time was
taken as thirty (30) minutes in this QRA study.
If the scenario release duration is less than thirty (30) minutes, the
actual concentrations corresponding to the certain fatality levels are
calculated based on the release duration instead of thirty (30) minutes.
Persons indoors are
expected to be offered some protection from the ingress of a toxic cloud in
buildings, the fatality probability for indoor persons is therefore assumed to
be one tenth of the outdoor fatality probability.
Toxicity of CO2 is reviewed and values for
the constants a, b and n is referred from available literature of the “HSE” ([78]). Based on the database, a is taken as -90.8, b is taken as 1.01 and n is taken as 8.
The concentrations
corresponding to thirty (30) minutes exposure for 90%, 50% and 1% fatality for CO2 are summarised in Table
5.25.
Table 5.25 CO2 Toxic Concentration at various
Fatality Levels
|
Fatality Level |
Concentration (ppm) |
|
90% |
107,713 |
|
50% |
91,933 |
|
1% |
68,902 |
The effect zones for the
hazardous scenarios considered presents in the format in Figure 5.14.
·
d: maximum downwind distance;
·
c: maximum crosswind width;
·
s: offset distance (distance between source and upwind end of effects
zone). It is noted that a negative
offset distance indicates that the upwind end of the effects zone is located
upwind of the source, as would occur for thermal radiation and overpressure
contours, for example; and
·
m: distance between release source and location of maximum
crosswind width.
All
consequence results are summarised in Annex 5D.
The cumulative risks were
calculated by summing the risks from the Project and the existing BPPS
facilities, as follows. The details of
cumulative risk assessment to consider the hazardous facilities at each of the
assessment years are summarised in Table
5.26.
·
Transport
and use of natural gas; and
·
Natural
gas facilities.
·
Transport,
storage and use of other non-fuel gas dangerous goods.
·
Hydrogen
facilities;
·
CO2 facilities; and
·
Distillate
oil facilities.
Table 5.26 Details of Cumulative Risk Assessment
|
Potential
Risks |
2016 |
2019 |
2020 |
2034 |
2035 |
|
Existing
Natural Gas Facilities |
Yes |
Yes |
Yes |
Yes |
Yes |
|
Existing
Hydrogen Facilities, On-site and Off-site Transport of Hydrogen |
Yes |
Yes |
Yes |
Yes |
Yes |
|
Existing
CO2 Facilities, On-site and Off-site Transport of CO2 |
Yes |
Yes |
Yes |
Yes |
Yes |
|
Existing
Distillate Oil Facilities |
Yes |
Yes |
Yes |
Yes |
Yes |
|
Construction
activities for 1st CCGT Unit leading to potential impact on nearby
BPPS equipment |
|
Yes |
|
|
|
|
1st
additional CCGT unit |
|
|
Yes |
Yes |
Yes |
|
Additional
Hydrogen Distribution, On-site and Off-site Transport of Hydrogen for 1st
CCGT unit |
|
|
Yes |
Yes |
Yes |
|
Additional
CO2 Distribution, On-site and Off-site Transport of CO2
for 1st CCGT unit |
|
|
Yes |
Yes |
Yes |
|
Additional
Distillate Oil Distribution for 1st CCGT unit |
|
|
Yes |
Yes |
Yes |
|
Construction
activities for 2nd CCGT Unit leading to potential impact on nearby
BPPS equipment |
|
|
|
Yes |
|
|
2nd
additional CCGT unit |
|
|
|
|
Yes |
|
Additional
Hydrogen Distribution, On-site and Off-site Transport of Hydrogen for 2nd
CCGT unit |
|
|
|
|
Yes |
|
Additional
CO2 Distribution, On-site and Off-site Transport of CO2
for 2nd CCGT unit |
|
|
|
|
Yes |
|
Additional
Distillate Oil Distribution for 2nd CCGT unit |
|
|
|
|
Yes |
The cumulative individual
risk contours from 1E-05 to 1E-09 per year of the following assessment years
are depicted at Figure 5.15 to Figure
5.19, and individual risk contours show that there is no off-site risk
larger than 1E-05 per year and the all individual risk contour of 1E-09 per
year does not reach any land population in the vicinity of BPPS.
·
Case
0: 2016 as the baseline condition year (Figure 5.15);
·
Case
1a: 2019 as the expected peak impact (in HtLA
context) construction year of the 1st CCGT unit (Figure
5.16);
·
Case
1b: 2020 as the expected year of operation of the 1st CCGT unit (Figure 5.17);
·
Case
2a: 2034 as the assumed peak impact (in HtLA context)
construction year of the 2nd CCGT unit (Figure 5.18); and
·
Case
2b: 2035 as the assumed year of operation of the 2nd CCGT unit (Figure
5.19).
It is noted that the
individual risk contours differ between Case 0 and Case 1a, and between Case 1b
and Case 2a. The additional risk is due
to the consideration of potential hazards due to Project construction
activities (for each construction phase).
At the operation phases of
1st and 2nd additional CCGT units, the individual risk
contours from 1E-06 to 1E-09 per year in the vicinity of the Project Site have
been extended in south-westerly direction due to additional CCGT equipment
considered (for each operation phase).
All individual risk results
do not demonstrate off-site individual risk contour at 1E-05 per year, and are
acceptable according to individual risk guideline as stipulated in Section 2 of
Annex 4 of EIAO-TM.
The potential loss of life
(PLL) values for the natural gas and other non-fuel gas dangerous goods
facilities at each assessment year are presented in Table
5.27.
It can be observed that the major risk contributor is from natural gas
facilities and the risks are the highest for the worst scenario (operation
phase of 2nd CCGT unit in 2035) because of two additional CCGT units
in operation and surrounding population growth (both road traffic and marine
vessel population).
The detail breakdown of the
main risk contributor to PLL for each of the assessment year is summarised in Annex 5E.
|
Type of
Facilities |
2016 |
2019 |
2020 |
2034 |
2035 |
|
Natural Gas |
3.69E-06 |
3.92E-06 |
4.01E-06 |
4.66E-06 |
4.96E-06 |
|
Non-Fuel Gas Dangerous Goods |
5.53E-08 |
1.56E-07 |
6.53E-08 |
6.71E-08 |
7.73E-08 |
|
Total |
3.74E-06 |
4.08E-06 |
4.08E-06 |
4.73E-06 |
5.03E-06 |
The cumulative risk
presented in FN curves for following assessment years are depicted at Figure
5.20, and all FN curves are within the “Acceptable” region.
·
Case
0: 2016 as the baseline condition year;
·
Case
1a: 2019 as the expected peak impact (in HtLA
context) construction year of the 1st CCGT unit;
·
Case
1b: 2020 as the expected year of operation of the 1st CCGT unit;
·
Case
2a: 2034 as the assumed peak impact (in HtLA context)
construction year of the 2nd CCGT unit; and
·
Case
2b: 2035 as the assumed year of operation of the 2nd CCGT unit.
The breakdown for FN curves
by types of dangerous goods for each of the assessment year is also presented
at Figure
5.21 to Figure 5.25.
From the hazard to life
assessment, the risk from the baseline case is largely contributed by natural
gas release scenarios onto marine population.
It is noted that the risk in terms of fatality frequency is increasing
along the assessment years mainly due to the operation of two additional CCGT
units (additional gas-filled process equipment/ pipework). The maximum number of fatality is just
slightly increased over the years because the surrounding population growth
within the hazardous impact zone is not significant.
Therefore, it can be
concluded that the cumulative risk level of the Project is within the
“Acceptable” region and in compliance with the societal risk criteria in
Section 2 of Annex 4 of EIAO-TM.
This section assesses the
level of uncertainty in the estimation of individual risk and societal risk by
carrying out sensitivity analysis studies on the key parameters and modelling
assumptions, in association with the hazard to life assessment. The purpose of this is to gauge the level of
confidence in the hazard assessment results.
Detection and shutdown
systems have been provided at BPPS for both natural gas and hydrogen gas
facilities; however, the automatic shutdown systems are installed at GRS and
hydrogen trailer bay areas only. As a conservative approach, it was assumed
the detection and shutdown system is not available. As such, the closing time (release duration)
of thirty (30) minutes was used to estimate the inventories released from the
isolation failure scenarios for consequence modelling in this QRA study.
Natural gas containing
methane, ethane, propane, etc. was conservatively assumed as pure methane as a
representative material in this QRA study.
The maximum surface emissive power of pure methane is higher than that
of natural gas for consequence modelling.
Due to that flammable characteristic, the consequence distances for both
flash fire and jet fire hazardous scenarios associated with pure methane are
therefore more conservative.
The flammable mass for
fireball modelling was conservatively estimated by the initial flow rate
continuing for ten (10) seconds even though the initial release rate is
decreasing rapidly in case of the full bore rupture scenario of pipeline. This approach is consistent with the approved
study, Safety Case Report ([79]).
The maximum storage
capacity for non-fuel dangerous goods including hydrogen, CO2 and distillate oil was
considered as the reasonably worst case for consequence modelling in this QRA
study.
The failure frequencies for
natural gas facilities were selected from Hawksley ([80]),
who presents his own derived data for both above and below ground pipework in
1984. This historical database was also
suggested in the approved EIA studies ([81])
([82]) for natural
gas facilities.
The failure database
adopted in the Safety Case Report ([83])
is the generic leak frequencies developed by DNV based on the Hydrocarbon Release
Database (HCRD) from UK HSE. The
database is based on approximately 4,000 recorded leaks recorded between
October 1992 and March 2010 from the UK sector of the North Sea which has been
collected by the UK HSE. In general, the
database for offshore facilities gives higher leak frequencies than most of the
onshore sources of data.
The natural gas facilities
at BPPS mainly consist of both above and underground pipework; therefore, it is
more applicable to adopt failure frequencies suggested by Hawksley ([84]) in this QRA study.
Moreover, the failure rates
from Hawksley (4) are also compared with
that suggested by UK HSE ([85]), which was established by the
Hazardous Installations Directorate (HID) C15 of the UK HSE in 2012. The UK HSE failure rates (5) are intended for use on
Land Use Planning cases and have been in use for several years. After review, it is found that the failure
rates for pipework from Hawksley (4) and UK HSE (5) are in the same order of
magnitude.
For other dangerous good
facilities, the majority of the failure frequencies were selected from the data
also suggested by the approved EIA studies, including Purple Book ([86]) and COVO Study
([87]), as these
international databases were developed for risk assessments for the use,
handling, transport, and storage of dangerous substances.
These databases are based on
failure data collected from a variety of onshore facilities including
refineries, LPG storage, etc., and other dangerous substances installations
such as chlorine and ammonia storage tanks with professional judgment applied
as appropriate.
Based on the discussions
above, it can be concluded that the uncertainty related to failure frequencies
of equipment/ facilities from BPPS is deemed insignificant.
It is noted that the
ignition sources for underground environment is limited; however, it has been
conservatively assumed the ignition probability from Cox, Lees and Ang ([88]) to be applied
on both above and underground environment, particularly for fireball scenario
associated with high fatality number.
Moreover, the total
ignition probability for a large leak of hydrogen has been conservatively
assumed as 0.99 based on the study from Cadwallader
and Herring ([89]). The large leak was assumed as the release
rate is larger than 1 kg s-1 only.
The ignition estimation
approaches for both natural gas and hydrogen gas adopted in this QRA study are
probably on conservative side.
The number of flights at Chek Lap Kok in 2035 (the
operation phase of 2nd CCGT unit) was estimated as 685,000 (assumed
as linear growth of a period from 1999 to 2014), which is in the same order of
magnitude with the estimated number of flights as 620,000 in 2032 from the
approved EIA study for Expansion of Hong
Kong International Airport into a Three-Runway System ([90]). As a conservative approach, the estimated
number of flights in 2035 was also taken into account for all the proposed
assessment years in this QRA study.
The official operation mode
for a three-runway system has not been officially confirmed yet, however, the
preferred operation mode based on the latest available information from the
approved EIA study (3) has been considered in
this QRA study to analyse the aircraft crash hazard on the Project Site. The failure frequency associated with
aircraft crash hazard was estimated as 1.2×10-8 per year, which has
already been taken into account in this QRA study.
In conclusion, after taking
into account the uncertainty in the hazard to life assessment by adopting
historical database with similar hazardous facilities characteristics,
conservative assumptions/ parameters, latest available information and worst
case scenario, this give confidence that the actual risk levels associated with
hazardous facilities at BPPS during both construction phase and operation phase
of the project will not exceed the cumulative risks estimated in this QRA study
and thus in compliance with the risk criteria in Section 2 of Annex 4 of EIAO-TM.
The following mitigation
measures should be considered to further manage and minimise the external
hazards from constructions activities risk during construction phases of the
Project. With the implementation of the
following mitigation measures, the failure frequencies associated with the
construction activities should be well covered by the general failure
frequencies summarized in Section 5.9.
·
Ensure
speed limit enforcement is specified in the contractor’s method statement to
limit the speed of construction vehicles on-site;
·
Conduct
speed checks to ensure enforcement of speed limits and to ensure adequate site
access control;
·
Provide
escort for hydrogen and CO2 delivery vehicle drivers to ensure the
right access route is used during the construction phases of the Project;
·
A
lifting plan, with detailed risk assessment, should be prepared and endorsed
for heavy lifting of large equipment;
·
Vehicle
crash barrier, designed for the specific speed limit at the BPPS, should be
provided between the construction site and the distillate oil storage
facilities during 1st CCGT unit construction phase. Also, a vehicle crash barrier is to be
provided between the construction site and the 1st CCGT unit during
2nd CCGT unit construction phase;
·
Any
lifting operation near or over live equipment should be strictly
minimised. If such operation cannot be
avoided, lifting activities should be assessed, controlled and supervised. Adequate protection covers should also be
provided on the existing BPPS facilities in case the operation of lifting
equipment has a potential to impact live equipment at BPPS. Process isolation should be achieved in case
that live equipment protection becomes impractical;
·
The
hydrogen road trailer and carbon dioxide road tanker delivery should follow
alternative route, which is further from the construction site, during crane
operation and movement of construction vehicles in the vicinity;
·
Ensure
that a hazardous area classification study is conducted and hazardous area maps
are updated before the start of the construction activities to ensure ignition
sources are controlled during both construction and operation phases;
·
Ensure
work permit system for hot work activities within the Project Site is specified
in the contractor’s method statement to minimise/ control ignition sources
during construction phase;
·
Ensure
effective communication system/ protocol is in place between the construction
contractors and operation staff;
·
Ensure
the Project Construction Emergency Response Plan is integrated with the
Emergency Response Plan for the BPPS during construction phases. The plan should address stop work
instructions to be promptly communicated to all construction workers performing
hot works in case a confirmed flammable gas (natural gas and hydrogen)
detection at the BPPS;
·
Ensure
that construction activities do not impede the functions of fire and gas
detection system, fire protection system, muster areas, fire-fighting vehicle
access and escape routes; and
·
Ensure a Job Safety
Analysis is conducted for construction activities of the Project during the
construction phases, to identify and analyse hazards associated with the
construction activities (e.g. lifting operations by cranes) onto the existing
plant facilities and operations.
Potential risks of the construction activities shall be assessed, and
risk precautionary measures shall be implemented in Contractor’s works
procedures.
A hazard to life assessment
has been conducted to evaluate the risk level associated with the transport and
use of natural gas, as well as the transport, storage and use of other
dangerous goods (hydrogen, CO2 and distillate oil)
defined in Dangerous Goods Ordinance
(Cap. 295) but not covered by the Gas Safety Ordinance (Cap. 51) during the construction phase
and operation phase of the project. The
cumulative risk assessment of the Project has also been assessed in combination
with risks associated with natural gas and other dangerous goods facilities at
BPPS.
The assessment methodology
and assumptions were based on the approved EIA and safety case studies having
similar issues, for example: Black Point
Gas Supply Project ([91]),
Safety Case Report for Black Point Gas
Supply Project ([92]),
Liquefied Natural Gas (LNG) Receiving
Terminal and Associated Facilities ([93]),
Desalination Plant at Tseung Kwan O –
Feasibility Study ([94]),
In-situ Reprovisioning
of Sha Tin Water Treatment Works – South Works – Designs and Construction ([95]), South Island Line (East) – Hazard to Life
Assessment ([96]). The list of major assumptions refers to Annex 5A.
Five (5) cases were considered
in this study, which are:
·
2016
as the baseline condition year;
·
2019
as the expected peak impact (in HtLA context)
construction year of the 1st CCGT unit;
·
2020
as the expected year of operation of the 1st CCGT unit;
·
2034
as the assumed peak impact (in HtLA context)
construction year of the 2nd CCGT unit; and
·
2035
as the assumed year of operation of the 2nd CCGT unit (for
worst-case assessment purpose).
In all cases, the
individual risk is in compliance with the risk criteria in Section 2 of Annex 4
of EIAO-TM and the societal risk lies
in the acceptable region.
Therefore, the construction
and operation phases of the Project are acceptable in terms of both individual
risk and societal risk as stipulated in Section 2 of Annex 4 of EIAO-TM.
Safety management measures
shall be implemented during the construction phase of the Project to ensure
that the risks associated with the transport and use of natural gas, as well as
the transport, storage and use of other dangerous goods at BPPS are in
compliance with the risk criteria in Section 2 of Annex 4 of EIAO-TM and stays in the “Acceptable”
region.
Since both the individual
and societal risks posed by the Project and the existing BPPS facilities to
off-site population meet the risk criteria in Section 2 of Annex 4 of EIAO-TM, no further mitigation measures
are required.
([1])
DNV, Safety Case Report for Black Point Gas Supply Project, Report No.:
PP019678, Revision No.2, August 2013.
([3])
ERM, EIA for Liquefied Natural Gas (LNG) Receiving Terminal and Associated
Facilities (Register No.: AEIAR-106/2007), December 2006.
([4])
ERM, EIA for 1,800 MW Gas-fired Power Station at Lamma Extension (Register No.:
AEIAR-010/1999), February 1999.
([5])
ERM/ BV, EIA for Desalination Plant at Tseung Kwan O – Feasibility Study
(Register No.: AEIAR-192/2015), November 2015.
([6])
ERM, EIA for In-situ Reprovisioning of Sha Tin Water
Treatment Works – South Works – Hazard to Life Assessment (Register No.:
AEIAR-187/2015), January 2015.
([8])
ERM, EIA for South Island Line (East) – Hazard to Life Assessment (Register
No.: AEIAR-155/2010), Report No.: 248137/51/F, August 2010.
([9])
ERM, EIA for Development of a Biodiesel Plant at Tseung Kwan O Industrial
Estate (Register No.: AEIAR-131/2009), February 2009.
([10])
ERM, EIA for South Island Line (East) – Hazard to Life Assessment (Register
No.: AEIAR-155/2010), Report No.: 248137/51/F, August 2010.
([12])
ERM, EIA for Liquefied Natural Gas (LNG) Receiving Terminal and Associated
Facilities (Register No.: AEIAR-106/2007), December 2006.
([14])
BMT Asia Pacific Ltd, Marine Impact Assessment for Black Point & Soko Islands LNG Receiving Terminal & Associated Facilities,
Pipeline Issues, Working Paper #3, Issue 6, May 2006.
([15])
Passenger Arrivals/Departures and Passenger Load Factors at Cross-Boundary
Ferry Terminals, January to December 2014, Marine Department, Hong Kong SAR.
([18])
DNV’s CO2RISKMAN, “Guidance on CCS CO2 Safety and Environment Major Accident
Hazard Risk Management”, DNV Report No. 13IJLJW-2, 30th January 2013, Rev. 1.
([21])
DNV’s CO2RISKMAN, “Guidance on CCS CO2 Safety and Environment Major Accident
Hazard Risk Management”, DNV Report No. 13IJLJW-2, 30th January 2013, Rev. 1.
([24])
ARIA (Analysis, Research and Information on Accidents) Database, http://www.aria.developpement-durable.gouv.fr/about-us/the-aria-database/?lang=en.
([26])
PRC International, America Gas Association, Analysis of DOT Reportable
Incidents for Gas Transmission and Gathering Pipelines, January 1984 to
December 2013.
([27])
GEO Publication No. 1/2012, Review of Earthquake Data for the Hong Kong Region,
Geotechnical Engineering Office, Civil Engineering and Development Department,
The Government of the Hong Kong Special Administrative Region, September 2012.
([28])
GCO, Review of Earthquake Data for the Hong Kong Region, GCO Publication No.
1/91, Civil Engineering Services Department, Hong Kong Government, 1991.
([29])
Scott, D.N., Pappin, J.W., Kwok, M.K.Y., Seismic
Design of Buildings in Hong Kong, Hong Kong Institution of Engineers,
Transactions, Vol. 1, No. 2, p.37-50, 1994.
([31])
Lee, B. Y., Report of Hong Kong in the International Tsunami Seminar in the
Western Pacific Region, International Tsunami Seminar in the Western Pacific
Region, Tokyo, Japan, 7-12 March 1988.
([34])
Annual Review of Aircraft Accident Data: U.S. General Aviation, Calendar Year
2001, National Transport Safety Board.
([35])
Mott MacDonald, EIA for Expansion of Hong Kong International Airport into a
Three-Runway System, (Register No.: AEIAR-185/2014), November 2014.
([36])
Mott MacDonald, EIA for Expansion of Hong Kong International Airport into a
Three-Runway System, (Register No.: AEIAR-185/2014), November 2014.
([39])
ERM, EIA for Liquefied Natural Gas (LNG) Receiving Terminal and Associated
Facilities (Register No.: AEIAR-106/2007), December 2006.
([40])
ERM/ BV, EIA for Desalination Plant at Tseung Kwan O – Feasibility Study
(Register No.: AEIAR-192/2015), November 2015.
([41])
Ames S. and Crowhurst D, Domestic explosion hazards
from small LPG containers, Journal of hazardous materials, 1988.
([42])
ERM, EIA for In-situ Reprovisioning of Sha Tin Water Treatment
Works – South Works – Hazard to Life Assessment (Register No.: AEIAR-187/2015),
January 2015.
([43])
DNV Technica, Project Summary Project for the Risk
Assessment of the Transport of Non Fuel Dangerous Goods in Hong Kong, 1996.
([44])
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.
([45])
Hawksley, J.L., Some Social, Technical and Economic Aspects of the Risks of
Large Plants, CHEMRAWN III, 1984.
([47])
ERM, EIA for Liquefied Natural Gas (LNG) Receiving Terminal and Associated
Facilities (Register No.: AEIAR-106/2007), December 2006.
([50])
Rijnmond Public Authority, A Risk Analysis of Six
Potentially Hazardous Industrial Objects in the Rijnmond
Area – A Pilot Study, COVO, D. Reidel Publishing Co.,
Dordrecht,1982.
([51])
Bund effectiveness in preventing escalation of tank farm fires, T Davies, A B
Harding, I P McKay, R G J Robinson, A Wilkinson, Process Safety and Environmental
protection, 74, Part B, 88-93, May 1996. (Note: the authors of this paper are
listed alpha
([55])
ERM, EIA for Liquefied Natural Gas (LNG) Receiving Terminal and Associated
Facilities (Register No.: AEIAR-106/2007), December 2006.
([57])
Cadwallader and Herring, Safety Issues with Hydrogen
as a Vehicle Fuel, Idaho National Engineering and Environmental Laboratory,
Lockheed Martin, September 1999.
([61])
DNV, Safety Case Report for Black Point Gas Supply Project, Report No.:
PP019678, Revision No.2, August 2013.
([62])
Center for Chemical Process Safety of the American
Institute of Chemical Engineer, Guidelines for Chemical Process Quantitative
Risk Analysis, Second Edition, 2000.
([63])
Centre for Chemical Process Safety of the American Institute of Chemical
Engineer, Estimating the Flammable Mass of a Vapour Cloud, John L. Woodward.
([65])
ERM/ BV, EIA for Desalination Plant at Tseung Kwan O – Feasibility Study
(Register No.: AEIAR-192/2015), November 2015.
([66])
ERM, EIA for South Island Line (East) – Hazard to Life Assessment (Register
No.: AEIAR-155/2010), Report No.: 248137/51/F, August 2010.
([67])
Birk, A.M., 1996, Hazards from Propane BLEVEs: An
update and Proposal for Emergency Responders, Journal of Loss Prevention in
Process Industries, Vol.9, No. 2, pp.173-181.
([68])
Center for Chemical Process Safety of the American
Institute of Chemical Engineer, Guidelines for Evaluating the Characteristics
of Vapor Cloud Explosions, Flash Fires, and BLEVEs,
1999.
([69])
Lees, F.P., Loss Prevention in Process Industries, 3rd Edition, Butterworth-Heinemann,
Oxford, UK, 2005.
([70])
Lees, F.P., Loss Prevention in Process Industries, 3rd Edition,
Butterworth-Heinemann, Oxford, UK, 2005.
([71])
Baum M. R., 1988, Disruptive Failure of Pressure Vessels – Preliminary Design Guidelines
for Fragment Velocity and the Extent of the Hazard Zone, Journal of Pressure
Vessel Technology – Trans ASME 1988:110:168-76.
([72])
Holden, P., Reeves, A., (1985). Fragmentation Hazards from Failures of
Pressurized Liquefied Gas Vessels. Assessment and Control of
Major Hazards. IChemE Symposium Series # 93 IChemE, Rugby, UK.
205-220.
([73])
Lees, F.P., Loss Prevention in Process Industries, 3rd Edition,
Butterworth-Heinemann, Oxford, UK, 2005.
([74])
T. Abbasi, S.A. Abbasi, The
boiling liquid expanding vapour (BLEVE): Mechanism, consequence assessment,
management, Journal of Hazardous Materials 141 (2007) 489-519.
([75])
ERM, EIA for South Island Line (East) – Hazard to Life Assessment (Register
No.: AEIAR-155/2010), Report No.: 248137/51/F, August 2010.
([76])
TNO, Methods for the Determination of Possible Damage to People and Objects
Resulting from Releases of Hazardous Materials (The Green Book), Report CPR
16E, The Netherlands Organisation of Applied Scientific Research, Voorburg, 1992.
([79])
DNV, Safety Case Report for Black Point Gas Supply Project, Report No.:
PP019678, Revision No.2, August 2013.
([80])
Hawksley, J.L., Some Social, Technical and Economic Aspects of the Risks of
Large Plants, CHEMRAWN III, 1984.
([82])
ERM, EIA for Liquefied Natural Gas (LNG) Receiving Terminal and Associated
Facilities (Register No.: AEIAR-106/2007), December 2006.
([83])
DNV, Safety Case Report for Black Point Gas Supply Project, Report No.:
PP019678, Revision No.2, August 2013.
([84])
Hawksley, J.L., Some Social, Technical and Economic Aspects of the Risks of
Large Plants, CHEMRAWN III, 1984.
([87])
Rijnmond Public Authority, A Risk Analysis of Six
Potentially Hazardous Industrial Objects in the Rijnmond
Area – A Pilot Study, COVO, D. Reidel Publishing Co.,
Dordrecht,1982.
([89])
Cadwallader and Herring, Safety Issues with Hydrogen
as a Vehicle Fuel, Idaho National Engineering and Environmental Laboratory,
Lockheed Martin, September 1999.
([90])
Mott MacDonald, EIA for Expansion of Hong Kong International Airport into a
Three-Runway System, (Register No.: AEIAR-185/2014), November 2014.
([92])
DNV, Safety Case Report for Black Point Gas Supply Project, Report No.:
PP019678, Revision No.2, August 2013.
([93])
ERM, EIA for Liquefied Natural Gas (LNG) Receiving Terminal and Associated
Facilities (Register No.: AEIAR-106/2007), December 2006.
([94])
ERM/ BV, EIA for Desalination Plant at Tseung Kwan O – Feasibility Study
(Register No.: AEIAR-192/2015), November 2015.