4.4.2.3 Surrounding Topography
OWTF 2 is located about 30m above the Principal Datum (mPD). From this point the land slopes gently down towards Man Kam To Road. From site survey and desktop study it has been determined that the most populated areas around OWTF 2 are located around Man Kam To Road and most of the residential premises are low-rise village houses. During assessment the effect of topography was taken into account by using a surface roughness length parameter of 50 cm. This setting can reflect the numerous bushes and obstacles presents around the Project site area [10].
4.4.2.4 Meteorological Data
Meteorological
data is required for consequence modelling and risk calculation. Consequence
modelling (i.e. dispersion modelling) requires wind speed and stability class
to determine the degree of turbulent mixing potential whereas risk calculation
requires frequencies of occurrence for each combination of wind speed and
stability class. The meteorological data from the Ta Kwu Ling Weather Station
in 2010 was adopted in this HA. The data are transformed into a set of weather classes in accordance
with the TNO purple book [11] for daytime and night-time, and can be expressed
in a combination of wind speed and Pasquill stability
classes. Pasquill stability classes (A to F)
represent the atmospheric turbulence with class A
being the most turbulent class while class F is the least turbulent class [12]. The six most
dominant sets of wind speed-stability class combination for both daytime and
night-time were identified and the occurrence probability of each weather
class is summarised in Table 4.9 and Table 4.10. The average ambient temperature adopted in
the analysis is 23°C and relative humidity is 78%.
Table 4.9 Daytime Weather Conditions (Ta Kwu Ling Weather Station 2010)
|
Wind Direction |
2.5B |
1D |
4D |
7D |
2.5E |
1F |
Total |
|
0 |
8.71 |
1.45 |
2.25 |
0.14 |
0.9 |
1.97 |
15.42 |
|
30 |
2.49 |
0.83 |
0.66 |
0.17 |
0.21 |
0.81 |
5.17 |
|
60 |
2.94 |
0.47 |
0.07 |
0 |
0.12 |
1.02 |
4.62 |
|
90 |
21.48 |
2.8 |
3.51 |
0 |
1.19 |
2.23 |
31.21 |
|
120 |
6.34 |
1.59 |
3.25 |
0.05 |
1.57 |
2.18 |
14.98 |
|
150 |
1.92 |
0.71 |
0.17 |
0 |
0.19 |
1.12 |
4.11 |
|
180 |
2.14 |
0.66 |
0.21 |
0 |
0.07 |
1.07 |
4.15 |
|
210 |
5.63 |
0.9 |
0.28 |
0 |
0.21 |
0.66 |
7.68 |
|
240 |
3.04 |
0.36 |
0.17 |
0 |
0.05 |
0.55 |
4.17 |
|
270 |
1.38 |
0.28 |
0 |
0 |
0 |
0.4 |
2.06 |
|
300 |
1.28 |
0.19 |
0.02 |
0 |
0.02 |
0.38 |
1.89 |
|
330 |
2.47 |
0.59 |
0.19 |
0 |
0.17 |
1.09 |
4.51 |
|
Total |
59.82 |
10.83 |
10.78 |
0.36 |
4.7 |
13.48 |
100 |
Table 4.10 Night-time Weather Conditions (Ta Kwu Ling Weather Station 2010)
|
Wind Direction |
2.5B |
1D |
4D |
7D |
2.5E |
1F |
Total |
|
0 |
0 |
0 |
3.08 |
0.07 |
1.67 |
5.76 |
10.58 |
|
30 |
0 |
0 |
1.3 |
0.12 |
1.03 |
2.81 |
5.26 |
|
60 |
0 |
0 |
0.02 |
0 |
0.3 |
3.77 |
4.09 |
|
90 |
0 |
0 |
4.09 |
0.05 |
6.6 |
11.72 |
22.46 |
|
120 |
0 |
0 |
4.8 |
0.07 |
6.62 |
14.18 |
25.67 |
|
150 |
0 |
0 |
0.12 |
0 |
0.54 |
7.73 |
8.39 |
|
180 |
0 |
0 |
0.1 |
0 |
0.47 |
6.2 |
6.77 |
|
210 |
0 |
0 |
0.17 |
0 |
0.89 |
5.39 |
6.45 |
|
240 |
0 |
0 |
0.1 |
0 |
0.15 |
2.95 |
3.2 |
|
270 |
0 |
0 |
0 |
0 |
0 |
1.45 |
1.45 |
|
300 |
0 |
0 |
0.02 |
0 |
0.1 |
2.09 |
2.21 |
|
330 |
0 |
0 |
0.07 |
0 |
0.42 |
2.95 |
3.44 |
|
Total |
0 |
0 |
13.87 |
0.31 |
18.79 |
67 |
100 |
The percentage frequencies are plotted in the form of a wind rose in Figure 4.5.
4.4.2.5 Preliminary Design of
Generation, Transfer, Storage and Use of Biogas at OWTF 2
The Project is intended to be delivered under a Design Build Operate (DBO) contract arrangement. Therefore, more detailed design information will be provided by the appointed Contractor following tender award. For the purposes of the Feasibility Study, investigations and tendering, a preliminary design has been produced and has been used as the basis of this assessment. Future design development carried out by the appointed Contractor must observe all relevant legislation and guidance, and shall not exceed the risk levels agreed with the regulator under the approved environmental permit unless agreed by the authorities.
Proven biological treatment technologies have been adopted to recover reusable materials and energy, such as compost and biogas from source-separated organic waste. Biogas may be used for onsite heat and power production or for direct injection in the gas network. The preliminary site layout is shown in Figure 4.1. Figure 4.6 and Figure 4.7 show the proposed treatment process and mass balance diagram from the Preliminary Design.
An indicative summary of the treatment process is as follows:
The incoming organic waste will pass through a pre-treatment stage which will separate out unsuitable materials and reduce organic material to a homogenous size. Pre-treated material will then be directed to anaerobic digesters. It is proposed to have three vertical cylindrical shaped digesters with each size 5,572 m3, which can be concrete, steel or glass enamel tanks. Biogas will be generated continuously from digesters with production rate of around 33,285 Nm3/day.
There will be pressure relief valves installed on the digesters and gasholder protecting against overpressure and under-pressure (50 mbarg/-2 mbarg) and on the gas storage (38 mbarg). Over flow pipes will be provided on the digesters and gasholder for protection against overfilling.
The digesters will also produce solid waste material, which will be dewatered for composting. Wastewater from the dewatered compost will pass through a wastewater treatment plant.
Biogas
composition is dependent on the composition of the food waste and the final
design of the adopted system. From the preliminary design information, the
composition of biogas from Anaerobic Digestion (AD) process are extracted and
shown in Table
4.11 below.
Table 4.11 Composition of Biogas from Anaerobic Digestion (AD) Process
|
Composition
& Parameters |
Biogas
from AD |
|
|
Methane |
(%) |
62 (55~70) |
|
Carbon Dioxide |
(%) |
38 (30~45) |
|
Density |
(Kg/Nm3) |
1.2 |
|
Lower Caloric Value |
(MJ/Nm3) |
23 |
The
produced biogas will pass through a purification process which consists of a
chemical absorption unit and a purification unit. The process will remove the
hydrogen sulphide and carbon dioxide from the biogas. According to the
Biogas Utilisation Paper [32], the potential uses of the purified biogas are as
follows:
1. Generation of heat and electricity by Combined Heat and Power (CHP) system: The biogas will be treated and then used to generate electricity and heat through co-generation equipment. The heat and electricity generated will be used onsite. Any surplus electricity will be supplied to the power grid; or
2.
Purification
of biogas and connection to a gas pipeline: The biogas will be treated to remove impurities. Sulphur will be
removed by absorption and moisture, siloxanes, carbon
dioxide and other impurities will be condensed out by refrigeration. Then, the
purified biogas can be connected to an existing synthetic natural gas pipeline,
which runs from NENT landfill to Tai Po Gas Production Plant. Direct injection
would be achieved via a 5.5km (approx.) pipeline aligned along the Kong Nga Po
road to the south-west. The external pipeline will be developed, owned and
maintained by the gas utility company up to the appropriate connection point
within the OWTF 2 site boundary.
Either one of the above options may be adopted for OWTF 2. The final biogas utilisation option will be confirmed at later stage of the Project. Both cases are considered in the assessment as a conservative approach.
4.5
Hazard
Identification
Potential hazardous scenarios associated with generation, transfer, storage and use of biogas in OWTF 2 have been identified. Historical incidents and relevant studies of similar facilities were reviewed to identify possible hazardous scenarios. In addition, the OWTF 1 HA Study [7] has been reviewed to ensure all the relevant hazardous scenarios are incorporated into this Study.
4.5.1 Review of Historical Incident Database
Review of Major Hazard Incident Data Services (MHIDAS) database, eMARS, FACTS, ARIA as well as internet searches was conducted, to further investigate the possible hazards from organic waste treatment facilities involving generation, transfer, storage and use of biogas or methane, anaerobic digesters or facilities of similar nature.
A total of 11 records were identified and grouped into different incident scenarios for further analysis. Details of incident data retrieved from these databases are shown in Appendix 4.1.
Table 4.12 summarises the incidents records related to biogas and methane.
Table 4.12 Summary of Biogas or Methane Incidents
|
Hazardous Scenario |
No. of Cases |
Country |
|
Methane Storage Tank Failure |
3 |
Turkey, India, Australia |
|
Methane Pipeline Failure |
2 |
UK, USA |
|
Anaerobic Digestion Plant
Failure |
6 |
Italy, France, Germany,
India |
4.5.2 Review of Relevant Studies
Relevant studies were reviewed in order to identify potential biogas impacts from explosion and flammability. Failure events and the respective hazardous scenarios associated with the biogas facilities have been identified in the approved HK EIA report “Organic Waste Treatment Facilities (OWTF), Phase 1 (EIA-176/2009)” [7] and “Development of a Biodiesel Plant at Tseung Kwan O Industrial Estate (TKOIE) (EIA-156/2008)” [13].
The OWTF 1 HA Study [7] evaluated the risk to construction workers and operational staff of the OWTF 1 due to the transport, storage and use of chlorine associated with the operations at Siu Ho Wan Water Treatment Works (SHWWTW). Chlorine is a poisonous, greenish-yellow gas described as having a choking odour. A set of hazardous scenarios associated with the transport, storage and use of chlorine at SHWWTW was included in the QRA for OWTF 1 [7], which also included the scenarios associated with the impact from biogas storage of OWTF Phase I on the chlorine store of SHWWTW. The hazardous scenarios of biogas identified in OWTF 1[7] were reviewed and adopted in this QRA study where applicable.
The Risk Assessment of the Biodiesel Plant at Tseung Kwan O Industrial Estate [13] assessed risk to life of the general public, including the workers of nearby plants, from the proposed facility during the operational phase of the biodiesel plant. Biogas is generated from an internal circulation (IC) reactor in the water treatment plant. Biogas is temporarily stored in a biogas buffer tank. Biogas consists mostly of methane and its properties are very similar to Natural Gas (NG). While it is non-toxic, in high concentrations it could lead to asphyxiation. A loss of containment can lead to jet fire (if stored/ transferred under sufficient pressure) or to an explosion if the gas accumulates in a confined space. [13] The possible biogas hazards identified in the biodiesel plant study were included in this Study where applicable.
4.5.3 Biogas Properties
Biogas
is a colourless flammable hydrocarbon gas at atmospheric conditions. Generally,
biogas has a methane content of 55% to 70% by volume. The physical and chemical
characteristics of biogas are modelled as a composition of 70-mol% methane and
30-mol% carbon dioxide [7] as
a conservative approach (flammability increases
with increase of methane content). Properties of biogas are very similar
to those of Natural Gas (NG), which is presented in Table 4.13.
Biogas can be further purified in the later processing stages. Carbon dioxide party removed increases the methane content to at least 80% by volume. 100% methane is used to model the purification unit as a conservative approach.
Table 4.13 Properties of Biogas (Natural Gas)
|
Property |
Values |
|
Flammability |
Extremely Flammable |
|
Auto-Ignition Temperature |
580°C |
|
Flash Points |
-188°C |
|
Melting Point |
-182.5°C |
|
Boiling Point |
-161.4°C |
|
Flammable Limits |
5% (Lower) – 15% (Upper) |
|
Vapour Density |
0.59-0.72 (air = 1) |
4.5.4 Spontaneous Failures
Gasholder Failure
The Preliminary Design recommends the use of a dry membrane fixed tank gasholder with steel containment for evening out variations in biogas production from the digesters. This type of gas holder typically consists of an external steel containment which forms the outer shape of the tank, as well as an internal membrane (a “dry bag”) which makes up the actual gas space. A non-return valve is installed at the inlet pipe to prevent gas from back-flow. Gas is discharged through the outlet pipe by suction blower. There are pressure relief valves on the gas holder for protection against the exceedence of designed gas storage pressure (38 mbarg), and over flow pipes on the gas holder for protection against overfilling. There will be emergency shut-off valves at the inlet and outlet pipes of the gas holder. In case of gas holder failure, the emergency shut-off valves can close the inlet and outlet pipes and the release of biogas to the atmosphere can be minimised.
A gas holder of dry bag type is different from column guided water-sealed gas holders, which do not have a gas holder crown. Therefore, tilting of tank top or blown seal failure will not occur in the operation of the gas holder. However, release of biogas could be from various parts of the gas holder or associated piping and devices. Possible hazardous outcomes include fireball, jet fire, flash fire and Vapour Cloud Explosion (VCE) [7].
The size of the gas storage, for evening out the variation in the gas production, is equal to 1 hour gas production. Assuming an hourly production rate of 1,387Nm3 per hour, the required gas storage capability will therefore be around 1,400 Nm3, with a design for 2 x 800m3 gas holders. The storage amount of the biogas will be around 1,920 kg. The maximum storage quantity is less than 15 tonnes. The quantity does not exceed the lower threshold quantity for existing Potentially Hazardous Installations (PHIs) for flammable gas and town gas installations in Hong Kong. Therefore, the proposed waste treatment facilities are not classified as a PHI.
Digester and Sulphur Absorption Vessel Failure
The preliminary design of the OWTF 2 incorporates three digester tanks with a combined volume of 15,858m3. The digesters consist of concrete, steel or glass enamel holding tanks, with either gas or top mounted mixing systems. Approximately 300tpd of organic waste slurry will enter the digestion tanks along with additional water to reduce the Dissolved Solid (DS) content from an estimated 22% to 10%. The estimated average residence time within the digesters is assumed to be 21 days. Providing space for mixing and gas production gives a volume requirement of approximately 15,000 m3 (biogas volume is about 858 m3) in total of the three digesters. Therefore, a biogas volume of 286 m3 will be assumed in each digester. Heating is required for heating of biomass during feeding of the digesters and for heat loss compensation from the digesters. The required heating will be provided via heat recovered from the CHP unit, or from a boiler. Pressure relief valves will be installed on the digester for protection against overpressure and under pressure (50 mbarg/ -2 mbarg) and over flow pipes on the gas holder protecting against overfilling.
Two 5m3 sulphur absorption
vessels are
provided downstream of the gasholders for the absorption of hydrogen sulphide in
the biogas. The
absorption vessels are made of steel and filled with absorbents, (Zinc oxide or
Iron oxide). An explosion proof blower will be used to
extract the biogas from gasholder to the sulphur absorption vessels at 400 mbarg. The absorbents are neither flammable nor
explosive so the major hazard will be from the release of biogas.
Failure of the digesters or the sulphur absorption vessels can
be caused by undetected corrosion, fatigue, material or construction defect.
Release of biogas could be from various parts of the process vessels as well as associated
piping and devices. Possible hazardous outcomes include fireball, jet fire,
flash fire and VCE.
Aboveground Inlet or Outlet Piping
Failure
Piping will be used to connect process vessels to the gasholder, compressor, and further purification unit and gas grid connection point. Failure along the on-site piping may be caused by undetected corrosion, fatigue, material or construction defect, or associated with flange gasket / valve leakage resulting in continuous gas release to the atmosphere. Failures of gaskets and valve leak only tend to give relatively small scale leakage and will not contribute to any off-site risk. Nevertheless, gasket and valve leak failure will be considered and be absorbed into pipework failure in the study. [8] For above ground piping, possible hazardous outcomes include jet fire, flash fire and VCE. For underground piping, possible hazardous outcomes include flash fire and VCE.
An
above ground pipeline of 300 mm diameter will be installed. The length of the
aboveground pipeline is estimated to be 100 m long by referring to the
equipment layout in the Project Site. A length of 200 m for the above ground
pipeline has been adopted in the study to represent a conservative approach.
Purification
Unit Failure
A
purification unit is used to condense carbon dioxide from the biogas, leaving a
majority of methane which is used for heat and power generation or gas
utilisation downstream. The major modes of failure in the purification unit are
similar to those of the above ground piping failures described in earlier
paragraphs.
Since
the purification unit mainly consists of a pipeline, a 100 mm diameter pipeline
at 20 barg with 100 m length has been adopted in the
study to represent a conservative approach.
4.5.5
External
Hazards
External hazards that are outside the control of the operating personnel could still pose a threat to the OWTF 2. Such hazards are termed as ‘external hazards’ because they are independent of the operations on-site but can lead to major hazard scenarios. This section discusses the credibility of loss of containment due to the external hazards with respect to Hong Kong’s geographical location.
Aircraft Crash
The OWTF 2 site is located about 32 km from the Hong Kong International Airport. The frequency of aircraft crash has been estimated using the HSE methodology, which is in line with approved “Kai Tak Development” EIA report [8] [17]. Details of frequency analysis are given in Section 4.6.2.
Helicopter Crash
Historical incidents show that helicopter accidents during take-off and landings are confined to a small area around the helipad, extending up to 200m from the centre of the helipad [13] [17]. 93% of accidents occur within 100m of the helipad. The remaining 7% occur between 100 and 200m of the helipad.
Since
the distance to nearest helicopter landing pad (namely Man Kam To Helicopter Landing Pad) is
about 1,000 m away from the Project site [31], as shown in Figure
4.12, only the background crash rate for
helicopters is considered in this report. Details of frequency analysis are
given in Section 4.6.2.
Vehicle Impact
Only authorised vehicles will be permitted to enter the OWTF 2 site, and speed will be restricted for vehicle movements around the site. Safety markings and marked crash barriers will be provided to the above ground piping, digesters and the gas holder near the entrance. The accident rate is calculated to be 1.80x10-07 severe car accident per km per year based on statistical data for Vehicle/ Object Crash accident involving medium and heavy goods vehicles in recent years, which is in line with approved EIA report “Hong Kong Section of Guangzhou - Shenzhen - Hong Kong Express Rail Link” [23]. Details of frequency analysis are given in Section 4.6.2.
Earthquake
Hong Kong is situated on the southern coast
of mainland China facing the South East China Sea. Hong Kong is not located
within the seismic belt and according to Hong Kong Observatory, earthquakes
occurring in the circum-Pacific seismic belt, which
passes through Taiwan and Philippines, are too far away to affect Hong Kong
significantly [14]. Although there has
not been any reported case of destructive earthquake in Hong Kong, loss of
containment incident due to earthquake was considered credible in this study.
The probability of earthquake occurrence at Modified Mercalli
Intensity Scale VII and higher in Hong Kong is low comparing to other regions,
and is estimated to be 1.0×10-5 per year [15]. The failure
probability of the equipment in an earthquake (both leak and rupture) is assumed to
be 0.01.
Details of frequency analysis are given in Section
4.6.2.
Landslides
The elevation of the Project site is about 30mPD and there is a registered slope (feature 3NW-C/C30) that lies at about 48mPD. According to Geotechnical Engineering Office's Slope Information System, the slope gradient is of 25° and is mainly made up of decomposed volcanic materials. In addition, there is a flat area (elevation of about 37mPD) in front of the slope feature toe which should act as a buffer zone limiting the hazard to the tanks/ facilities from major landslides in the area. Besides, the above ground piping, digesters and gas holders will be installed on foundations that can withstand the impact of landslide. Hence, landslides causing release of biogas are not considered further in this assessment. Details of the natural terrain study are given as below.
The Enhanced Natural Terrain Landslide Inventory and the Large Landslide Inventory have been used and the information obtained is shown in Figure 4.13. No landslides or slope failure are recorded to have occurred within the site or on the natural terrain which remains around the Site. A number of relict landslides have been recorded immediately to the south of the Site, but this natural terrain was removed by the site formation works carried out between 1982 and 1991. Thus, landslides causing release of biogas are not considered further in this assessment. These landslides, together with other landslides on similar natural terrain in the area, are usually located on the upper parts of the slopes, at the heads of ephemeral first-order drainage lines.
No Large Landslides are recorded in the immediate area of the Site. The nearest recorded Large Landslide is located approximately 300 m to the south-west, at an elevation of +75mPD at the head of a drainage line on the other side of the valley adjoining the Site (see Figure 4.13). This feature (3NWC063L) is recorded as a 30m x 25m active slide, with a sharp main scarp, hummocky morphology and absent or sparse vegetation.
Four incidents are recorded within the general area (see Figure 4.13), but only one of these (MW97/8/17) is in a position at the sloping area around the Site. This incident occurred on 7 August 1997 and comprised a small (0.5 cu m) failure in a soil cut slope. The cause of failure was logged by GEO as “infiltration” and “washout”. No previous instability or groundwater seepage was noted. The failure appears to have involved slope C453 which was later (August 1999) improved by Prescriptive Measures.
Guidelines for Natural Terrain Hazard Studies are contained in GEO (2003). “Inclusion” criteria help to identify whether a site requires screening in respect of natural terrain hazards, and are as follows:
(a) the proposed development involves Group 1, 2 or 3 facilities, and
(b) there is “hillside” sloping at more than 15° within 100m horizontally upslope from the site.
With respect to these criteria, the existing and proposed developments at the site fall within Group 2 (a) –built-up area (this definition includes “… incinerator… refuse transfer station… manned substation…”. However, criterion (b) is not met as there is no natural terrain overlooking the Site.
Under the guidelines in GEO (2003) there is no requirement for further screening or study of natural terrain hazards. “Natural terrain hazards” are defined as; open hillslope landslides; channelized debris flows; deep-seated slides; rock falls; and boulder falls.
Lightning
Lightning sparks could ignite combustible gas in air. OWTF 2 will be equipped with a lightning protection system that can effectively protect the OWTF 2 equipment from lightning. Lightning protection installations should be installed following IEC 62305, BS EN 62305, AS/NZS 1768, NFPA 780 or equivalent standards. [30] The installations will be protected with lightning conductors to safely earth direct lightning strikes. The double grounding system will be inspected regularly. Therefore, failures due to lightning strikes are to be covered by generic failure frequencies. [13]
External Fire
External fire means the occurrence of fire
event which leads to the failure of the equipment inside OWTF 2. The facilities
will be equipped with fire alarm and fire suppression system. In addition,
stringent procedures are implemented to prohibit smoking or naked flames to be
used on-site. However, hill / vegetation fires are relatively common in Hong
Kong and could potentially occur near OWTF 2. Details of frequency analysis are
given in Section 4.6.2.
Typhoon/ Tsunami
Loss of containment due to severe environmental event such as typhoon or tsunami (large scale tidal wave) is not possible. OWTF 2 will be designed to withstand wind load for local typhoon and the location of OWTF 2 means that it is not threatened by tsunami in Hong Kong. Thus, typhoon or tsunami causing a release of biogas is not considered further in this assessment.
4.5.6 Possible Hazardous Scenarios and
Hazardous Outcomes
Referencing OWTF Phase I EIA [7], OWTF Phase I will be operated on a 24-hour basis daily, pre-treated material would be fed into the buffer tanks to start the anaerobic digestion process. From the buffer tanks, the material is pumped to the individual digester where a major portion of the organic material is converted into biogas. In the reference design, five vertical cylindrical digesters are provided, each with a design capacity of approximately 3,000m3. After digestion, the material from digesters is pumped to a dewatering facility and further treated by tunnel composting. All the post-treatment facilities were located in an enclosed building with air extraction system. In summary, anaerobic digestion (AD) and composting were recommended for OWTF 1.
The biogas is then treated and compressed to approximately 100 mbar and stored in a double membrane gasholder. The design of energy recovery system aims to convert the energy contained in the biogas to electricity and heat by the application of cogeneration units (Cogen Units). A stand-by flare was provided for burning the surplus biogas in emergency or under abnormal circumstances.
The gasholder is a spherical double membrane type and is different from column guided water-sealed gas holders in that it does not have a gas holder crown. Therefore, tilting of tank top or blown seal failure will not occur in the operation of the double membrane gasholder. It is concluded that release of biogas could be from various parts of the gasholder or associated piping and devices.
OWTF
Phase 2
For OWTF 2, a combination of anaerobic digestion (AD) and composting is considered appropriate [32]. Three vertical cylindrical shaped digesters each sized at 5,572 m3 are proposed, which will be constructed in steel or concrete. Biogas will be generated continuously from the digesters. Figure 4.6 shows the Process Flow Diagram (PFD) of OWTF 2.
Energy will be recovered under normal conditions, with emergency flare under emergency conditions. All biogas generated will either be exported to CHP for power generation, or further purified and exported to the gas utility distribution network. Before utilisation of biogas in either case, sulphur in biogas will be removed by absorption with two vessels of 5m3 volume.
The preliminary design of OWTF 2 predicts biogas production is about 33,285 Nm3/d, with a required buffer storage capacity of 1,387 Nm3 for 1 hour capacity onsite. As the design is still preliminary, a buffer storage capacity of 1,600 m3 is adopted in this HA study as a conservative approach.
Based on the review of OWTF 1 and available information of OWTF 2, both facilities will be using similar organic waste treatment technology, that is a combination of anaerobic digestion (AD) and composting. The sulphur absorption vessels are structurally similar to anaerobic digestion vessels. Table 4.14 identifies the possible hazardous scenarios and hazardous outcomes in OWTF 2, making reference to the OWTF 1 EIA Study [7].
Table 4.14 Possible hazardous scenarios and hazardous outcomes in OWTF 2
|
Potential Cause |
Reference ID in
Figure 4.6 |
Release Type |
Hazardous Outcome |
||||
|
Gasholder |
16 |
Rupture |
Fireball VCE Flash fire |
||||
|
Leak |
Jet fire VCE Flash fire |
||||||
|
Digester |
13 |
Rupture |
Fireball VCE Flash fire |
||||
|
Leak |
Jet fire VCE Flash fire |
||||||
|
Sulphur absorption vessels |
General impurities removal
equipment for biogas purification |
Rupture |
Fireball VCE Flash fire |
|
||||
|
Leak |
Jet fire VCE Flash fire |
|
||||||
|
Aboveground inlet or outlet piping
/ purification piping/ pump / non-return valve / flange |
General plant item |
Rupture / Leak |
Jet fire VCE Flash fire |
|
||||
|
Safety valve * |
General plant item |
Discharge due to overfilling |
Jet fire VCE Flash fire |
||||
Note: #Aboveground inlet or outlet piping will be assumed for all piping in this HA study that takes into account failure of piping will lead to direct release to the atmosphere, on conservative approach.
* Safety valve is a valve mechanism which automatically opens when the pressure exceeds pre-set conditions. Safety valve as a safety measure, the design shall take into account discharging any released fluid to a safe location to avoid hazardous outcome. Hence, safety valve causing a release of biogas shall not be considered as potential cause in this assessment.
Possible hazardous outcomes will be assessed using PHAST Professional version 6.53, to determine the risk impact, where the potential risk associated with the operation, layout and facilities threat posed to life and neighbouring property in a hazardous outcome at the Project. Details of consequence analysis are shown in Section 4.7.
4.6
Frequency
Analysis
Frequencies for each of the identified hazardous scenarios are estimated using the best available failure data or historical accident data in the process and gas industry. The frequencies documented in the relevant sources are reviewed and justified if necessary, to reflect the specific operation and risk reduction practices evident at the organic waste treatment facilities.
When the historic data on failure frequency is not available, failure frequencies of similar installations or events are adopted with suitable modifications based on the process conditions of OWTF 2. For example, the failure frequency of the fixed tank dry membrane type biogas holder is not readily available in literature, the failure frequency of double containment tank (which is available in TNO purple book [11]) having a similar structural arrangement will be used in this HA Study. Modification is made according to the specifications as required.
4.6.1
Spontaneous
Failures Frequencies
Gasholder Failure
The Preliminary Design recommends the use of a fixed steel tank dry membrane type gas holder for evening out variations in biogas production at the OWTF 2 site. This type of gas holder typically consists of an external cylindrical steel tank, and an internal membrane, which makes up the actual gas space. According to “Bevi Risk Assessments” published by National Institute of Public Health and the Environment (RIVM), the catastrophic rupture and leak failure leading to release to atmosphere of double containment tank are 1.25 x 10-8 per year and 1 x 10-4 per year respectively. [34]
Digester / Sulphur Absorption Vessel
Failure
The preliminary design of the OWTF 2
incorporates three digester tanks each with a volume of 5,572m3 and
a combined volume of 15,858m3. Each digester consists of a concrete, steel or glass enamel holding tank, with either
gas or top mounted mixing systems. The
system also incorporates two sulphur absorption vessels each with a volume of 5
m3. The catastrophic rupture and leak failure frequencies of
digester tank / sulphur absorption vessel are 1 x 10-5 per year and
1 x 10-4 per year respectively. [11]
Aboveground / Purification Unit Piping
Failure
Failure along the onsite piping may be caused by undetected corrosion, fatigue, material or construction defect, or associated with flange gasket / valve leakage resulting in continuous gas release to the atmosphere. For aboveground piping, catastrophic rupture and leak failure frequencies are 1 x 10-7 per metre per year (300 mm dia.) and 5 x 10-7 per metre per year (30 mm dia.) respectively. [11]. According to the OWTF 2 layout plan as shown in Figure 2.4, a length of approximately 150 m is measured between digesters and the gasholder. Nevertheless, a length of 200 m is assumed for the aboveground pipelines for a conservative approach.
A summary of the base event frequencies are shown in Table 4.15.
Table 4.15 Summary of Spontaneous Failures Frequencies
|
Events |
Frequency of Occurrence |
|
|
Rupture |
Leak |
|
|
Gasholder |
1.25 E-8 per year |
1.00 E-4 per year |
|
Digester/Sulphur
Absorption Vessel |
1.00 E-5 per year |
1.00 E-4 per year |
|
Aboveground Inlet or Outlet
Piping |
1.00 E-7 per metre per year |
5.00 E-7 per metre per year |
|
Purification Unit Piping |
3.00 E-7 per metre per year |
2.00 E-6 per metre per year |
4.6.2 External Event Frequencies
Aircraft
Crash
The OWTF 2 site is located around 32 km from the Hong Kong International Airport. The frequency of aircraft crash is estimated using the HSE methodology, which is in line with approved “Kai Tak Development” EIA report [8] [17].
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. The crash frequency per unit ground area (per km2) is calculated as:
(Equation
1)
Where N is the number of runway movements per year and R is the probability of an accident per movement (landing or take-off). FL(x,y) gives the spatial distribution of crashes and is given by:
For aircraft landing,
(Equation 2)
for x >-3.275 km
For aircraft take-off,
(Equation 3)
for x >-0.6 km
Equations 2 and 3 are valid only for the specified range of x values. If x lies outside this range, the impact probability is zero. Aircraft Crash Coordinate System is shown in Figure 4.8.
NTSB data [18] for fatal accidents in the US involving scheduled airline flights during the period 1986-2010 are given in Table 4.16. The 10-year moving average suggests a downward trend with recent years showing a rate of about 1×10-7 per flight. However, only 18.7% of accidents are associated with the approach to landing, 14% are associated with take-off and 4.7% are related to the climb phase of the flight [19]. The accident frequency for the approach to landings hence becomes 1.87×10-8 per flight and for take-off / climb 1.87×10-8 per flight. Arrival and departure flight paths of Hong Kong International Airport are shown in Figure 4.9 and Figure 4.10 respectively.
Table 4.16 U.S Scheduled Airline Accident Rate [18]
|
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 |
1.1 |
- |
|
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 |
0.3 |
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 |
|
2006 |
0.19 |
0.169 |
|
2007 |
0 |
0.139 |
|
2008 |
0 |
0.13 |
|
2009 |
0.1 |
0.122 |
|
2010 |
0 |
0.104 |
The number of flights from 2001 to 2011 is extracted from the Civil Aviation Department [20], and extrapolated to year 2017 by adopting an annual growth rate of 8% for aircraft movements based on Air Traffic Statistics at HKIA [21]. The number of flights at Chek Lap Kok for year 2017 is estimated at 529,707 The number of plane movements is illustrated in Table 4.17 below:
Table 4.17 Hong Kong International Airport Civil International Air Transport Movements of Aircraft
|
Year |
Landing |
Take-off |
Total |
|
2001 |
98,415 |
98,402 |
196,817 |
|
2002 |
103,355 |
103,346 |
206,701 |
|
2003 |
93,748 |
93,759 |
187,507 |
|
2004 |
118,662 |
118,646 |
237,308 |
|
2005 |
131,759 |
131,745 |
263,504 |
|
2006 |
140,203 |
140,177 |
280,380 |
|
2007 |
147,675 |
147,657 |
295,332 |
|
2008 |
150,577 |
150,561 |
301,138 |
|
2009 |
139,715 |
139,684 |
279,399 |
|
2010 |
153,277 |
153,257 |
306,534 |
|
2011 |
166,918 |
166,887 |
333,805 |
|
2017 |
|
|
529,707 (projected) |
The number of plane movements has been divided by 8 to take into account that half of movements are take-offs and only a quarter of landings use specific runways (i.e., runway 07R, 07L, 25L, 25R). This effectively assumes that each runway is used equally.
Considering landings on runway 25R for
example, the values for x and y are estimated to be 30 and 16 km
respectively. *x and y are measured by Geoinfo
map and the difference between north and south runway is assumed to be
negligible.
Applying Equation 2 gives FL= 1.96 ×10-11 km-2. Substituting this into Equation 1 gives:
Table 4.18 shows the
calculated impact frequency due to aircraft crash is 1.17 x 10-15 per year,
which is much less than 1.0 x 10-9 per year. The risk of aircraft
crash at the OWTF 2 site is therefore not considered further in the analysis.
Table 4.18 Aircraft Crash Frequency onto the OWTF 2 Site
|
Site |
Distance from Runway Threshold (km) |
Crash Frequency (/km2/yr)* |
OWTF2 Area (m2) |
Impact Frequency (/yr) |
|||||||
|
|
07L/25R |
07R/25L |
07L |
25R |
07R |
25L |
Total |
|
|
||
|
|
x |
y |
X |
y |
Take-off |
Landing |
Take-off |
Landing |
|
|
|
|
OWTF2 |
30 |
16 |
30 |
16 |
1.91E-21 |
2.43E-14 |
1.91E-21 |
2.43E-14 |
4.86E-14 |
24156 |
1.17E-15 |
Note: Size of the Project Area is referenced from Statutory Planning Portal [22].
Helicopter
Crash
Historical incidents show that helicopter accidents during take-off and landings are confined to a small area around the helipad, extending up to 200m from the centre of the helipad [13] [17]. 93% of accidents occur within 100m of the helipad. The remaining 7% occur between 100m and 200m of the helipad.
For
most sites, no consideration is given to helicopter crashes associated with
helipads when there are no helipads within 200m of the site [19]. Since the distance to nearest helicopter
landing pad (namely Man Kam To
Helicopter Landing Pad) is about 1,000m away from the Project site [31], as shown in Figure
4.12, only the background crash rate for
helicopters is considered in this report.
The
background crash rate for helicopters is assumed to be 1.00 x 10-5/km2/year [35], and hence, by
accounting for the areas of anaerobic digesters (254.5m2 x3), gas
holders (63.6m2 x2), aboveground pipeline area (80m2),
sulphur absorption vessels (7m2 x2) and purification unit area (150m2)
, the background crash frequencies of helicopter for anaerobic digesters, gas
holders, aboveground pipeline, sulphur absorption vessels and purification unit
are 7.635 x 10-9/year, 1.272 x 10-9/year, 8 x 10-10/year, 1.4 x 10-10/year, and 1.5
x 10-09/year respectively.
Vehicle
Impact
The accident rate is calculated to be 1.80 x 10-7 severe car accident per km per year based on statistical data for Vehicle/ Object Crash accident involving medium and heavy goods vehicles in recent years.
Making reference to Road Traffic Accident Statistics obtained from HKSARG Transport Department [33], the overall number of accident involvements per million vehicle-kilometres is given in Table 4.19 for Medium/ Heavy Goods Vehicles (M/HGVs).
Table 4.19 Hong Kong Vehicle Accident Involvements
|
Serious and Fatal Vehicle involvements |
2003 |
2004 |
2005 |
2006 |
2007 |
2008 |
2009 |
2010 |
2011 |
Average |
|
Invol rate: per million veh-km |
|
|
|
|
|
|
|
|
|
|
|
M/HGV |
0.79 |
0.89 |
0.89 |
0.86 |
0.82 |
0.8 |
0.76 |
0.83 |
0.91 |
0.84 |
|
Total involvements |
|
|
|
|
|
|
|
|
|
|
|
M/HGV |
1108 |
1197 |
1180 |
1155 |
1081 |
1045 |
907 |
1031 |
1141 |
1094 |
|
|
|
|
|
|
|
|
|
|
|
|
|
Fatal involvements |
|
|
|
|
|
|
|
|
|
|
|
M/HGV |
50 |
31 |
27 |
25 |
21 |
17 |
27 |
16 |
22 |
26 |
|
Serious injury involvements |
|
|
|
|
|
|
|
|
|
|
|
M/HGV |
255 |
291 |
257 |
212 |
188 |
176 |
147 |
163 |
196 |
209 |
|
Slight injury involvements |
|
|
|
|
|
|
|
|
|
|
|
M/HGV |
1136 |
1380 |
1412 |
1364 |
872 |
1176 |
1050 |
1205 |
1370 |
1218 |
|
|
|
|
|
|
|
|
|
|
|
|
|
Fatal vehicle involvements ratio |
|
|
|
|
|
|
|
|
|
|
|
M/HGV |
4.5% |
2.6% |
2.3% |
2.2% |
1.9% |
1.6% |
3.0% |
1.6% |
1.9% |
2.4% |
|
Serious injury involvements ratio |
|
|
|
|
|
|
|
|
|
|
|
M/HGV |
23.0% |
24.3% |
21.8% |
18.4% |
17.4% |
16.8% |
16.2% |
15.8% |
17.2% |
19.0% |
|
|
|
|
|
|
|
|
|
|
|
|
|
High impact accident involvement rate per million
vehicle km |
|
|
|
|
|
|
|
|
|
|
|
M/HGV |
0.04 |
0.02 |
0.02 |
0.02 |
0.02 |
0.01 |
0.02 |
0.01 |
0.02 |
0.02 |
|
Medium impact accident involvement rate per
million vehicle km |
|
|
|
|
|
|
|
|
|
|
|
M/HGV |
0.18 |
0.22 |
0.19 |
0.16 |
0.14 |
0.13 |
0.12 |
0.13 |
0.16 |
0.16 |
The data generally shows a constant overall accident involvement rate in the past 9 years. The statistics indicate an overall impact accident involvement rate of 0.18 (=0.02 + 0.16) involvements per million vehicle kilometre (pmvkm) for MGV/HGV. Therefore, the vehicle crash frequency is estimated to be 1.8×10-7 per vehicle kilometre per year.
Only authorised vehicles will be permitted to enter the OWTF 2 site, and speed will be restricted for vehicle movements. Safety markings and marked crash barriers will be provided to the aboveground piping, digesters gasholders and purification unit, as shown in Figure 4.1 (1). Therefore, it is assumed that vehicle impact could only cause leak failure to digesters and gas holders, whereas it could cause both rupture failure and leak failure to aboveground piping [9].
Earthquake
Hong Kong is situated on the southern coast
of mainland China and facing the South East China Sea. Hong Kong is not located
within the seismic belt and according to Hong Kong Observatory, earthquakes
occurring in the circum-Pacific seismic belt, which
passes through Taiwan and Philippines, are too far away to affect Hong Kong
significantly [14]. Although there has
not been any reported case of destructive earthquake tremor in Hong Kong, loss
of containment incident due to earthquake was considered credible in this
study. The probability of earthquake occurrence at Modified Mercalli
Intensity Scale
(MMIS)
VII and higher in Hong Kong is low comparing to other regions, and is
estimated to be 1.0×10-5 per year [15]. The failure
probability (for
both leak and rupture) of the equipment in an earthquake is assumed to
be 0.01 [16].
External Fire
Although OWTF 2 is not located in a country
park, some of the surrounding terrain and vegetation is similar to that
typically found in country parks. According to the statistics from the
Agriculture, Fisheries and Conservation Department, the average number of hill
fires was 30 per year during the recent five years 2009 – 2013 (range: 16 to
51). Since the total area of country parks in Hong Kong was 43,394 Ha as in
2011 (most recent available figure), the frequency of hill fire in Hong Kong is
taken as 6.91 x 10-8 per m2 per year.
At the thermal radiation intensity of 37.5 kW/m2, damage to process equipment can
happen [24]. From the literature,
for a heat flux of 37.5 kW/m2, the
corresponding flame-to-structure distance is 25 m caused by burning in tree
canopy producing persistent flames [36]. For a conservative approach, 50 m is adopted
in this study to account for uncertainty (e.g. spreading of hill fire). The
resulting total area used in the frequency calculation (Figure 4.11) is thus the total area of vegetation
extending 50 m beyond the process area, which is 16119 m2.
In OWTF 2, the
facilities will be equipped with fire alarm and fire suppression system (including
fire alarms, fire detectors, sprinkler extinguishing
system and fire pumps) to protect the facilities against external fire. It is
considered that damage to gas holders, vessels and piping happens when there is
hill fire as well as failure of fire protection system. By taking into account the failure rate of fire protection system of 2.20 x 10-2
per year [24], the overall
frequency of damage to gas holders, vessels and piping is 2.45 x 10-5
per year respectively. It is assumed that damage to the process equipment
results in rupture failure and leak failure in equal probability. Hence, the
catastrophic rupture and leak failure frequencies of gas holder / digester tank
/ sulphur absorption vessel / aboveground pipeline are 1.23 x 10-5
per year respectively.
A summary of the base event frequencies are shown in Table 4.20.
Table 4.20 Summary of Base Event Frequencies
|
Events |
Frequency of Occurrence |
|
Aircraft Crash |
1.17E-15 per year |
|
Helicopter Crash |
1.00E-05 per km2 per year |
|
Vehicle
Impact |
1.80E-07 per vehicle-km per year |
|
Earthquake |
1.00E-05 per year |
|
External Fire |
6.91E-08 per m2 per year |
Fault Tree Analysis (FTA) was conducted to
evaluate the frequencies of the identified biogas release scenarios. FTA is the
use of a combination of simple logic gates, “AND” and “OR” gates, to synthesise
a failure model of the biogas facilities. Fault Tree Analysis is shown in Appendix
4.2. The
assumptions used in FTA are summarised in the following table (Table 4.21):
Table 4.21 Assumptions used in FTA
|
Items |
Assumed Value |
Justification |
|
probability of rupture
failure in helicopter crash |
1 |
On conservative approach |
|
length of access road |
0.16km |
Measured using the preliminary plot plan (Figure
4.1) |
|
no.
vehicle movements per day |
80 |
Conservative assumption based on the traffic
assessment report of OWTF2 (60 for waste trucks, 10 for other vehicles and 10
account for uncertainty). |
|
probability running into
gasholder / digesters / absorption vessels / pipelines / purification unit |
0.5 |
Following approved EIA report HATS 2A [9], and based on the fact that concerned
process vessels are only at one side of the road. |
|
probability damage to
gasholder / digesters / absorption vessels / pipelines / purification unit |
1 |
On conservative approach |
|
probability rupture failure
in car crash for pipeline / purification unit |
0.1 |
Following approved EIA report HATS 2A [9] |
|
probability leak failure in
car crash for pipeline / purification unit |
0.9 |
Following approved EIA report HATS 2A [9] |
4.6.3 Ignition and Explosion Probability
The probabilities of ignition and explosion following a release depend on several factors, i.e. presence of an ignition source, material that was released, and the rate and the duration of the release. Possible ignition sources include hot surfaces, static electricity, flame and hot particles from external fire, etc [24]. The ignition probabilities are further split between immediate ignition and delayed ignition in equal proportions [13]. Immediate ignition of biogas could lead to a fireball or jet fire, whereas delayed ignition could cause a flash fire or vapour cloud explosion. Table 4.22 shows the total ignition probabilities and explosion probabilities adopted from Cox, Lees and Ang [24] according to gas release size.
Table 4.22 Ignition and Explosion Probabilities for Gas Releases
|
Release Size |
Probability of Ignition |
Probability of Explosion |
|
Minor (< 1 kg/s) |
0.01 |
0.04 |
|
Major (1 – 50 kg/s) |
0.07 |
0.12 |
|
Massive (> 50 kg/s) |
0.3 |
0.3 |
Event Tree Analysis
(ETA) was developed to determine the possible hazard event outcomes from the
identified hazardous events and to estimate the hazard event frequencies from the
initiating release frequency. Event Tree Analysis is shown in Appendix 4.3.
4.6.4 Estimating Generic Frequencies
Generic frequency was estimated based on the historical incidents review identified the accidents involving generation, transfer, storage and use of biogas or methane, anaerobic digesters or facilities of similar nature. The generic accident frequency can be estimated through the information of the number of biogas plants works involved, the operating period and the total number of accidents occurred within the operating period. The objective of the generic frequency estimation is to confirm the appropriateness of adopting generic failure frequencies for this HA.
The
generic frequencies estimated based on European experience are 1.73x10-4
incident per plant-year, whilst the overall failure frequency for OWTF 2 HA is
2.27 x10-3 (according to FTA
shown in Appendix
4.2), which is greater than the estimated value
from the European historical incidents. Therefore, the frequencies in the OWTF
2 HA Study are considered reasonably conservative. Details of generic frequency
estimation are given in Appendix 4.4. Failure scenarios are considered in this
study, and modelled comparing to the generic failure frequencies. It is assumed
that the biogas facilities will be designed and constructed to the appropriate
standards so that generic failure frequencies are appropriate. [13]
4.7 Consequence Analysis
The consequence assessment estimates impact of each outcome in the area of concern. The consequence assessment consists of two major parts, namely:
¡ Source term modelling – to determine the appropriate discharge models to be used for calculation of the release rate, duration and quantity of the release; and
¡ Effect modelling – to determine dispersion modelling, fire modelling and explosion modelling from the input of source term modelling.
Releases from hazardous sources and their consequences are modelled with the well-established software PHAST Professional version 6.53.
4.7.1 Source Term Modelling
For instantaneous failure, the whole content release of a tank is modelled. In case of continuous release, release parameters such as release rate and exit velocity are calculated by a discharge model according to storage conditions. Release duration is based on capacity of the storage tank (2). For piping connecting to the reactor network, release duration is determined by the response time to completely isolate the system. For piping connecting to the storage tank, release duration is based on the time to empty the whole tank gas content for anaerobic digesters and the response time to completely isolate the gasholder. Release parameters together with release duration are then fed into the dispersion model to calculate the effect.
Process vessel, piping and storage vessel would be the major release sources. Relief pressure of pressure relief valves and isolation valves are used to estimate storage pressure in failure cases.
4.7.2 Potential Hazardous Outcomes and their
Effect Modelling
The following sub-sections briefly describe the types of hazard events arising from a loss of containment scenario at the OWTF 2.
The Unified Dispersion Model (UDM) model is
used for the dispersion calculation of biogas for non-immediate ignition
scenarios. The model takes into account various transition phases, from dense cloud
dispersion to buoyant passive gas dispersion, in both instantaneous and
continuous releases.
Fireball
For immediate ignition of an instantaneous gas release, a fireball can be formed. Fireball is more likely for immediate ignition of instantaneous release and heat is evolved by radiation. The principal hazard of fireball arises from thermal radiation. Due to its intensity, its effects are not significantly influenced by weather, wind direction or source of ignition. Sizes, height, shape, duration, heat flux and radiation are determined in the consequence analysis. A 100% fatality is assumed for anyone within the fireball radius.
Jet Fire
When a pressurised flammable gas is released and ignited immediately, a jet fire could occur. The momentum of the release carries the flammable substance forward in a long plume, giving a flammable mixture by entraining air. Combustion in a jet fire occurs in the form of a strong turbulent diffusion flame, which is heavily influenced by the momentum of the release. The major concern regarding jet fire is the heat radiation effect generated from the fire. The thermal effect to adjacent population is quantified in the consequence model.
Flash Fire
Following a hazardous gas release, it could form a flammable gas cloud initially located around the release point. If this cloud does not get ignited immediately, it could move in the downwind direction and be diluted as a result of air entrainment. Flash fire is the consequence of combustion of gas cloud resulting from delayed ignition. The flammable gas cloud can be ignited at its edge and cause a flash fire of the cloud within the Lower Flammable Limit (LFL) and Upper Flammable Limit (UFL) boundaries.
Major hazards from flash fire are thermal radiation and direct flame contact. Since the flash combustion of a gas cloud normally lasts for a short duration, the thermal radiation effect on people near a flash fire is limited. Any persons who are encompassed outdoors by the flash fire could be fatally injured. A fatality rate of 100% is assumed.
Vapour Cloud Explosion
When there is a large amount of pressurised gas rapidly releasing to the atmosphere from a pressurised tank, a vapour cloud could be formed, dispersed and mixed with the surrounding air. If the vapour cloud is passing through a confined/ semi-confined environment and gets ignited, the confinement could limit the degree of expansion of the burning cloud and create an overpressure and explosion. This type of explosion is called a VCE.
Thermal Radiation
Hazardous consequences, such as jet fire, flash fire, etc. is assessed using PHAST’s consequence models. Fatality probabilities of various hazardous event outcomes are evaluated at a number of end-point criteria in each type of hazard outcome. 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 effect. The probability of fatality, Pr, due to exposure to heat radiation, i.e. jet fire and fireball is given by the following probit relationship by Eisenberg et al. which provides one of the more conservative estimates [25]:
Pr =
-14.9 + 2.56 ln (Q4/3 x t)
Where,
Pr is the probit associated with the probability of fatality;
Q is the heat radiation intensity (kW/m2);
t is the exposure time (s)
4.7.3 Assessment Criteria for Biogas Hazards
Biogas rises and dilutes rapidly due to its buoyancy when it is released to the atmosphere. In case of instantaneous release of biogas, immediate ignition near the release source could lead to fireball. VCE would occur when vapour cloud is trapped between facilities and is ignited. Potential damage from a fireball and a vapour cloud explosion are caused by thermal radiation and overpressure respectively. Assessment criteria for the thermal radiation [24] and overpressure effects [26] are adopted and shown in Table 4.23 [7].
Table 4.23 Assessment Criteria for Biogas Hazards
|
Outcome |
Effect |
Assessment Criteria |
Damages |
|
Fire |
Thermal radiation intensity |
37.5 kW/m2 / Jet flame / Fireball/ LFL |
Process equipment damage |
|
VCE |
Overpressure |
0.2 bar (about 3 psi) |
Damage to heavy machinery |
The fatality probability for VCEs is taken from CIA guidelines [27] as shown in Table 4.24. The indoors fatality probability is higher because of the increased risk from flying debris such as breaking windows [13].
Table 4.24 End Point Criteria for Vapour Cloud Explosions
|
Overpressure (psi) |
Fatality Probability
(outdoors) |
Fatality Probability
(indoors) |
|
5 |
0.09 |
0.55 |
|
3 |
0.02 |
0.15 |
|
1 |
0.00 |
0.01 |
The effective hazardous distances are quantified by DNV’s PhastRisk v6.53 Multi-energy model [29] available in the PhastRisk. This model is used to estimate the overpressure effect of vapour cloud explosion. Referencing to the guidance suggested by Kinsella [28] was adopted to determine the confined strength via the determination of the blast strength class. Blast strength category is divided into 12 categories. Blast strength category is a combination of ignition strength, obstruction, existence of parallel plane confinement / unconfinement. “Blast strength category” is used for determining the blast strength class. “Blast strength category” 1 represents high in ignition strength, obstruction and confinement. The lower blast strength category is, the higher the blast strength class. The highest blast strength class 10 is equivalent to detonation of TNT explosive. Thus, high blast strength class implies high initial overpressure. Hazard distance in a VCE increases with the increase in initial overpressure. Blast strength category 3 (equivalent to confined strength between 5 and 7) is estimated based on the following assumptions:
¡ High Obstruction – 50% volume blockage ratio
¡ Existence of parallel plane confinement – vertical walls
¡ Low ignition strength – ignition sources such as spark (mechanical or electrical), flare stack, hot surface
4.7.4 Consequence Distances associated with
OWTF 2
Considering the 300mm diameter hole size scenario, the whole content of the gasholder will release to the atmosphere in 3.7 minutes. For the 30mm or less equivalent hole size leak scenario, it is assumed the amount of gas in a VCE is equivalent to 10 minute discharge without being noticed. The consequence distances obtained from PhastRisk modelling for identified release scenarios are shown in Appendix 4.5.
For the flash fire events, it shall be noted that flash fire could spread to 146.8m upon purification unit failure, which would only last for a few seconds. Moreover, the perimeter of the fire reduces rapidly when biogas is ignited and consumed in the fire. For the fireball case, the radius of fireball is 29m and the duration is less than 5 seconds upon biogas gasholder rupture. In the worst jet fire event, the maximum jet fire flame length is 94.7m due to full bore rupture of the purification unit.
4.8
Risk Assessment
Risk Summation
Individual risk can be characterised by summation of the results of meteorological data, frequency estimation and consequence analysis, risk levels of the assessed scenarios can be presented by individual risk contours.
Societal risk of the assessed scenarios can be characterised by combining the results of population data, meteorological data, frequency estimation and consequence analysis, in terms of F-N curves.
The above steps are done by using MPACT in the SAFETI software suite v6.53.
Results
In the gasholder rupture event, the fireball duration lasts for about 5 seconds of radius 28.9m according to modelling results. The hazardous distances of flash fire and VCE in the worst case are assessed to be 146.8 m and 80.6 m respectively.
The individual risk (IR) contours associated with the OWTF 2 are shown in Figure 4.14. The maximum individual risk remains below 1x10-5 per year at the site boundary and hence meets the HKRG requirements.
For the societal risk of the 2017 scenario, the potential loss of life (PLL) for the OWTF 2 is 6.42 x 10-6 per year. The PLL value is very low for 2017 scenario, given the low off-site population in the vicinity.
In view of the uncertainties on the population intake (e.g. future Kong Nga Po Comprehensive Development Area, Hung Lung Hang Residential Population, Man Kam To Development Corridor and Sandy Ridge Crematorium and Columbarium Facilities) during the operational phase of the OWTF 2, the impact of OWTF 2 to the proposed developments and their population are also evaluated. The potential loss of life (PLL) for the OWTF 2 with proposed developments is 8.48 x 10-6 per year. It can be observed that the risks are higher comparing to 2017 scenario without proposed developments population intake. This is because the population intakes for all the proposed developments including Kong Nga Po Comprehensive Development Area, Hung Lung Hang Residential Population, Man Kam To Development Corridor and Sandy Ridge Crematorium and Columbarium Facilities are counted in this scenario in 2017 on conservative side.
Figure 4.15 shows the FN Curves for the 2017 scenario and the 2017 scenario with proposed developments. It can be seen that the societal risk for both scenarios are low and within the acceptable region as per HK EIAO Societal Risk Guideline. The risk increases for the case with proposed developments during the operation phase due to increase in surrounding population, but the risks are still low and in the acceptable region.
4.9 Recommendations
Although the
risks for both scenarios are within the acceptable region and thus no
mitigation measures are necessary, the HA has assumed that the following “Good
Practices” and “recommended design measures” for the safe operation of OWTF 2
shall be carried out as far as reasonably practicable.
¡ All electrical equipment inside the building will be classified in accordance with the electrical area classification requirements. No unclassified electrical equipment will be used during operations or maintenance.
¡ Reference can be made to Codes of Practice and guidance issued in Europe that applies to places where explosive atmospheres may occur (called ‘ATEX’ requirements). These are covered as part of the European Directive: the Explosive Atmospheres Directive (99/92/EC) and the UK regulations, Dangerous Substances and Explosive Atmospheres Regulations 2002 (DSEAR). Where potentially explosive atmospheres may occur in the workplace, the requirements include, identifying and classifying (zoning) areas where potentially explosive atmospheres may occur; avoiding ignition sources in zoned areas, in particular those from electrical and mechanical equipment; where necessary, identifying the entrances to zoned areas; providing appropriate anti-static clothing for employees; and before they come into operation, verifying the overall explosion protection safety of areas where explosive atmospheres may occur.
¡ All safety valves design shall take into account discharging any released fluid to a safe location, or stopping misdirection of fluid flows in order to avoid hazardous outcome.
¡ Safety markings and crash barriers will be provided to the aboveground piping, digesters and the gas holder near the entrance.
¡ Lightning protection installations will be installed following IEC 62305, BS EN 62305, AS/NZS 1768, NFPA 780 or equivalent standards.
¡ A 10m high boundary wall with fire resistance will be provided in the vicinity of the digester tanks, gasholders and gas purification equipment to protect the equipment against external fires, and to provide some protection to external areas from the effects of fire/explosion.
¡ Suitable fire extinguishers will be provided within the site. An External Water Spray System (EWSS) will be installed in appropriate areas, such as around the gasholders, gas purification, desulphurisation units, and digester areas. The facilities will also be equipped with fire and gas detection system and fire suppression system. Stringent procedures are implemented to prohibit smoking or naked flames to be used on-site.
¡ Fixed crash barriers will be provided in areas where process equipment is adjacent to the internal roadway to protect against vehicle collision. Adequate warning signage and lighting will also be provided and maximum speed limit will also be in place.
Implementation of risk mitigation measures is not required since the risk level is at the acceptable level.
4.10 Conclusion
A QRA for the proposed OWTF 2 was performed on the existing, committed and planned off-site population. Risks associated with the operational phases of the facility are evaluated to be within the acceptable region of the HK EIAO Societal Risk Guideline, for the scenarios with and without population intakes from proposed developments in the vicinity of the Project site.
Good safety practices and recommended design measures for operation of the OWTF 2 that have been assumed within the HA are summarised in Section 4.9. Implementation of risk mitigation measures is not required since the risk level is at the acceptable level.
Reference
[1] Environmental Impact Assessment Study Brief “Development of Organic Waste Treatment Facilities, Phase 2" (ESB 226/2011).
[3] 2011 Hong Kong Population Census, Census and Statistics Department, HKSARG, February 2012
[4] Projections of Population Distribution 2010-2019, Planning Department, HKSARG, December 2010
[5] Land Use Planning for the Closed Area-Feasibility Study Final Report, Planning Department, HKSARG, July 2010
[6] The Annual Traffic Census 2011, Transport Department, HKSRAG, August 2012
[7] Approved HK EIA report “Organic Waste Treatment Facilities, Phase I”, (EIA-176/2009)
[8] Approved HK EIA report “Kai Tak Development”, (EIA- 157/2008)
[9] Approved HK EIA report “Harbour Area Treatment Scheme (HATS) Stage 2A”, (EIA-148/2008)
[10] Meteorological Monitoring Guidance for Regulatory Modeling Applications, USEPA, Feb 2000.
[12]
Air Dispersion Modelling Conversions and
Formulas
http://www.air-dispersion.com/formulas.html#stability
[13] Approved HK EIA report “Development of a Biodiesel Plant at Tseung Kwan O Industrial Estate”, (EIA-156/2008)
[14] Hong Kong Observatory, Hong Kong Special Administrative Region Government
[15] Approved HK EIA Report “Proposed Headquarters and Bus Maintenance Depot in Chai Wan”, (EIA-060/2001)
[17]
Byrne, J. P., The Calculation of Aircraft Crash Risk in the UK, HSE\R150, 1997.
[18]
Aviation Statistical Reports, US National
Transportation Safety Board
http://www.ntsb.gov/data/table6.html
[20]
Civil Aviation Department, Hong
Kong International Airport Civil International Air Transport Movements of
Aircraft, Passenger and Freight (2001 - 2011)
http://www.cad.gov.hk/english/p-through.htm
[22]
HK Statutory Planning Portal
http://www.ozp.tpb.gov.hk/default.aspx
[23] Approved HK EIA report “Hong Kong Section of Guangzhou - Shenzhen - Hong Kong Express Rail Link”, (EIA-169/2009)
[24] Lee’s Loss Prevention in the Process Industries, 3rd Edition, 2005
[31] HK GeoInfo Map, powered by HKSARG Geospatial Information Hub (GIH) of the Lands Department http://www.map.gov.hk
[33] Road Traffic Accident Statistics, powered by HKSARG Transport Department http://www.td.gov.hk/en/road_safety/road_traffic_accident_statistics/index.html
[34] Bevi Risk Assessments, published by National Institute of Public Health and the Environment (RIVM), 2009
[35] Approved HK EIA report “Tsuen Wan Bypass, widening of Tsuen Wan Road between Tsuen Tsing Interchange and Kwai Tsing Interchange, and associated junction improvement works”, (EIA-152/2008)
Footnote:
(1)
The location of above ground piping, digesters and gas holders and purification
unit and crash barriers are shown according to the best information currently
available.
(2) Referencing OWTF Phase I EIA [7], for a 300mm diameter hole size scenario, the whole content of the gasholder releases to the atmosphere is less than 10 minutes. It was assumed that the amount of gas in a VCE is the same as the rupture scenario. For a 30mm equivalent hole size leak scenario, it was assumed the amount of gas in a VCE is equivalent to 10 minute discharge without being noticed.