Quick Search  
  Advanced Search
Fault Tree Analysis of Fire and Explosion Accidents for Dual Fuel (Diesel/Natural Gas) Ship Engine Rooms
Yifeng Guan, Jie Zhao, Tengfei Shi, Peipei Zhu     
School of Naval Architecture & Ocean Engineering, Jiangsu University of Science and Technology, Zhenjiang 212000, China
Abstract: In recent years, China's increased interest in environmental protection has led to a promotion of energy-efficient dual fuel (diesel/natural gas) ships in Chinese inland rivers. A natural gas as ship fuel may pose dangers of fire and explosion if a gas leak occurs. If explosions or fires occur in the engine rooms of a ship, heavy damage and losses will be incurred. In this paper, a fault tree model is presented that considersboth fires and explosionsina dual fuel ship; in this model, dual fuel engine rooms are the top events. All the basic events along with the minimum cut sets are obtained through the analysis?The primary factors that affect accidents involving fires and explosions are determined by calculating the degree of structure importance of the basic events?According to these results, corresponding measures are proposedto ensure and improve the safety and reliability of Chinese inland dual fuel ships?
Key words: dual fuel ship     engine room     fault tree     fire and explosion     safety measures     gas leak    

1 Introduction

With rising fuel costs and a global interest in emission reduction, Liquefied Natural Gas (LNG) is a promising alternative fuel for ship engines. This technology uses natural gas as primary fuel and neat diesel as pilot fuel. The existing ship engines that use natural gas as their primary fuel produce less soot, CO2, NOX species, and SOX species in comparison with conventional diesel-only engines. Related studies by Chen (2013) and Mattarelli et al.(2014) indicated that dual fuel (diesel/natural gas) ships result in energy conservation and environmental protection. Consequently, there is a huge, promising market for dual fuel engines.

A recent high-profile explosion accident in Kaohsiung killed at least 26 people and injured another 260 people. The accident may have been caused by a combustible gas ignition. Therefore, safety concerns exist for using natural gas as a ship fuel. Natural gas could lead to jet fire, flash fire, vapour cloud explosion and other accidents after a gas leakage occurs; such accidents will significantly damage the surroundings, personnel, and equipment. If an accident occurs in the engine rooms, damages and losses will be enormous because the engine rooms are small and have some equipment.

2 Introduction and risk identification of dual fuel ship engine rooms 2.1 The fuel supply system of dual fuel ship engine

To have an intuitive understanding of the fuel supply system of a dual fuel ship engine, a newly-built 63-meter dual fuel ship is introduced here (Fig. 1). The ship is a bulk carrier with two engines manufactured by the Zibo diesel engine corporation. The power of each engine is 220 kW and the engine type is Z6170ZLC/S. The LNG storage tank is manufactured by Nantong CIMC, and the tank volume is 5 m3.

Figure 1 Dual fuel bulk carrier

The examplary engine adopts manifold intake before the air is turbocharged. Typically,a mixer is installed in front of the turbo in this type of engines (Li and Guan, 2013; Lei et al., 2014). The gas pipeline and mixer are interlinked, and the natural gas that is provided to the mixer is in a zero-pressure state. While the engine is running, the natural gas pumped out by vacuum is formed by the air flow in the mixer; the gas then forms a mixture with the air. The mixture goes through the inlet manifold into the cylinder to generate combustion power. The system’s components include the LNG storage tank, evaporator, pressure reducing valve, manual stop valve, principal gas valve, flame arrester, mixer, filter, and turbo. The system’s work process is shown in Fig. 2.

Figure 2 The working principle of manifold intake before air turbocharge
2.2 Brief introduction of dual fuel (diesel/natural gas) ship engine rooms

To receive a certification by the international classification societies, dual fuel ship engine rooms must include an inherently safe engine room and an Electro-Static discharge(ESD)-protected engine room (Wang et al., 2014). In consideration of the Chinese inland rivers, a concept for the enhanced safe engine room is proposed herein that adheres to the principle of having an equal risk level with the inherently safe engine room. The three types of engine rooms are introduced later in this paper. Inherently safe engine rooms require all gas supply pipes to be airtight and ring-fenced (e.g., using double-walled pipeline or using ventilation ducts). The joints of the gas pipeline and gas injection valves should be completely covered by conduits; this reduces the risk of pipeline leakage and improves the overall security level of the engine rooms. However, there are significant technical requirements for gas supply lines and engines of inherently safe engine rooms.

The gas supply pipes of ESD-protected engine rooms are not necessarily airtight or ring-fenced. The engines that generate propulsive power and electric power can be arranged in two or more engine rooms. When the fuel supply to any one engine is cut off, it should be able to maintain at least 40% each of the propulsion and normal power supply. The engine rooms should accommodate the necessary equipment. Incinerators, inert gas generators, and fuel boilers should not be arranged in the engine room. The complexity and high cost of this kind of engine room is not suitable for Chinese inland small vessels.

Enhanced safe engine rooms are proposed herein based on the inherently safe engine room; the proposed rooms take into account the situation of Chinese inland rivers and follow the principle that the risk level of dual fuel (Diesel/Natural Gas) ship engine rooms should equal the inherently safe engine room. The following measures were taken to prevent the risk of gas leakage: enhancement of ventilation ability, strengthening of combustible gas detection and alarms; and full penetration welding.

The cited example ship adopts an inherently safe engine room with the layout shown in Fig. 3.

Figure 3 The layout of an inherently safe engine room
2.3 Risk identification of Chinese inland dual fuel ship engine rooms

Although measures have been taken for risk aversion, the possibilities of dangers still exist. The security failures of dual fuel engine rooms include the following:

1) Leakage of combustible gas in the pipeline and from the engine. When the concentration of combustible gas reaches the LEL, dangerous situations may occur, e.g., fires and explosions.

2) Fire sources: To present the particularity of dual fuel ship engine rooms, the studies discussed herein focused only on accidents caused by gas leakages and fires. As long as an ignition source is present, accidents could be engendered.

3) Failure of risk avoiding devices: Failure of the detection and alarm systems may be caused by the following: system damage or an inadequate detection range; an abnormal valve switch caused by damage to the valves; or poor ventilation caused by a faulty power supply line or damaged ventilator. When a gas leakage occurs, it may result in gas accumulation, which increases the danger of explosion.

According to the above-mentioned reasons, a fault tree model was established wherein Fires and Explosions in Dual Fuel Engine Rooms (FEDFER) are the top events. Then, the influence factors for the accidents are determined and a qualitative analysis is made. According to these results, corresponding measures are proposed to improve and ensure the safety and reliability of Chinese inland dual fuel ships.

3 Qualitative analysis of the fault tree 3.1 Construction of FTA for FEDFER

Fault tree analysis (FTA) is a systematic approach to estimating the safety and reliability of a complex system; it is characterized by concision, visualization, and predictability (Fay, 2013; Li and Huang, 2012). The FTA approach starts with a top undesired event and branches backward through intermediate events until all possible root causes at the bottom are confirmed. To show the particularity of dual fuel ship engine rooms, only fires and explosions caused by a gas leakage are studied.

According to some relevant papers (Huang and Li, 2012; Fan et al., 2013; Liao et al., 2014; Xu and Gan, 2014) and safety manuals, the fault tree model is established as Fig 4:

Figure 4 The fault tree for FEDFER

Table 1 shows the event type of each symbol.

Table 1 Event type list
Symbols Event type
T Fire and explosion of dual fuel engine rooms
X The natural gas concentration reaching the LEL
F1 Fire sources
F2 Gas leakage
F3 Risk avoiding devices failure
F4 Open fires
F5 The detection and alarm systems failure
F6 Poor ventilation
X1 Electric gas welding
X2 Spontaneous combustion of cotton yarns
X3 Surroundings fire
X4 Short circuit
X5 Impact sparks
X6 Static sparks
X7 Smoking
X8 Pipeline leakage
X9 Valve leakage
X10 Joint leakage
X11 Engine leakage
X12 Detection and alarm systems damage
X13 Beyond the range of detection
X14 Gas cut-off valve failure
X15 The power supply line fault
X16 Damage of ventilator
3.2 The Minimum Cut Sets (MPSs)

MCSs relate the top event cause directly with the basic event causes. MCS is the smallest combination of basic events that, if they all fail, will cause the occurrence of the undesired event. Each MCS is a possible channel to the top event.

The MCSs equation is as follows:

$ T={{F}_{1}}{{F}_{2}}{{F}_{3}}X $ (1)

When each basic event is substituted into (1), the following is obtained:

$ \begin{align} &T=({{F}_{4}}+{{X}_{4}}+{{X}_{5}}+{{X}_{6}}+{{X}_{7}}) \\ &({{X}_{8}}+{{X}_{9}}+{{X}_{10}}+{{X}_{11}})({{F}_{5}}+{{X}_{14}}+{{F}_{6}})X\text{=} \\ &({{X}_{1}}+{{X}_{2}}+{{X}_{3}}+{{X}_{4}}+{{X}_{5}}+{{X}_{6}}+{{X}_{7}}) \\ &({{X}_{8}}+{{X}_{9}}+{{X}_{10}}+{{X}_{11}})({{X}_{12}}+{{X}_{13}}+{{X}_{14}}+{{X}_{15}}+{{X}_{16}})X \\ \end{align} $ (2)

In this study, all the MCSs are obtained using Boolean algebra. The proposed fault tree yields 140 MCSs for just 17 basic events, which indicates that there are 140 kinds of basic means of combination leading to the top event. {X1,X8,X12,X}, one of the MCSs, is used as an example to show that when a pipeline leaks, the detection and alarm systems are damaged, which increases the concentration of natural gas and air mixture beyond the LEL. If there is an electric gas welding operation, fire and explosion accidents will occur in the engine room.

3.3 The minimum path sets

In contrast with the concept of cut sets, the Minimum Path Sets (MPSs) is the smallest combination of basic events. If all basic events do not fail, the undesired event will not occur. Seeking the MPSs is to use its duality with the MCSs. According to the duality rules of Boolean algebra A·B=A + B and A + B=A·B, the MPSs are obtained as follows:

$ {T}'={{F}_{1}}^{\prime }+{{F}_{2}}^{\prime }+{{F}_{3}}^{\prime }+{X}'=\prod\limits_{i=1}^{7}{{{{{X}'}}_{i}}}+\prod\limits_{i=8}^{11}{{{{{X}'}}_{i}}}+\prod\limits_{i=12}^{16}{{{{{X}'}}_{i}}}+{X}' $ (3)

The MPSs, a total of four, are {X1,X2,X3,X4,X5,X6,X7}, {X8,X9,X10,X11}, {X12,X13,X14,X15,X16}, and {X}. To ensure that the top event of FEDFER does not occur, all events in at least one MPS must not occur.

3.4 The analysis of structure importance degree

The degree of structure importance identifies the basic events that contribute most to the probability that the top event will occur; consequently, they should be given priority for improvement. The following quadratic approximate calculation formula is used to calculate the structure importance degree coefficient (Du, 2012; Tu, 2012; Ren et al., 2011; Shi et al., 2013):

$ {{I}_{\varphi \left( {{X}_{i}} \right)}}=1-\prod\limits_{{{X}_{i}}\in {{k}_{j}}}{(1-\frac{1}{{{2}^{{{n}_{j}}-1}}}}) $ (4)

where $ {{I}_{\varphi \left( {{X}_{i}} \right)}} $ is the structure importance coefficient of ith basic event; kj is the total MCSs; nj is the total number of basic events in MCS kj, where ith basic event rests; and $ {{X}_{i}}\in {{k}_{j}}$ means ith basic event belongs to the jth MCS. By the above equation, the following was obtained:

$ {{I}_{\varphi \left( {{X}_{1}} \right)}}={{I}_{\varphi \left( {{X}_{2}} \right)}}=\ldots {{I}_{\varphi \left( {{X}_{7}} \right)}}\text{=}1-{{(1-\frac{1}{{{2}^{4-1}}})}^{20}}=0.\text{93}0\text{8 } $ (5)
$ {{I}_{\varphi \left( {{X}_{8}} \right)}}={{I}_{\varphi \left( {{X}_{9}} \right)}}={{I}_{\varphi \left( {{X}_{10}} \right)}}={{I}_{\varphi \left( {{X}_{11}} \right)}}\text{=}1-{{(1-\frac{1}{{{2}^{4-1}}})}^{35}}=0.\text{99}0\text{7} $ (6)
$ {{I}_{\varphi \left( {{X}_{12}} \right)}}={{I}_{\varphi \left( {{X}_{13}} \right)}}=\ldots {{I}_{\varphi \left( {{X}_{16}} \right)}}=1-{{(1-\frac{1}{{{2}^{4-1}}})}^{28}}=0.\text{9762} $ (7)
$ {{I}_{\varphi \left( X \right)}}=1 $ (8)

Then:

$ \begin{align} &{{I}_{\varphi \left( {{X}_{1}} \right)}}={{I}_{\varphi \left( {{X}_{2}} \right)}}=\ldots {{I}_{\varphi \left( {{X}_{7}} \right)}}<{{I}_{\varphi \left( {{X}_{12}} \right)}}={{I}_{\varphi \left( {{X}_{13}} \right)}}=\ldots {{I}_{\varphi \left( {{X}_{16}} \right)}}< \\ &{{I}_{\varphi \left( {{X}_{8}} \right)}}={{I}_{\varphi \left( {{X}_{9}} \right)}}={{I}_{\varphi \left( {{X}_{10}} \right)}}={{I}_{\varphi \left( {{X}_{11}} \right)}}<{{I}_{\varphi \left( X \right)}} \\ \end{align} $ (9)
4 Results analysis and safety measures

The above results suggest that the fault tree for FEDFER has 140 MCSs. The means that lead to the top event are numerous; therefore, the system has a certain risk level.

First, from the point of the MPSs, as long as the event of the combustible gas concentration reaching the LEL(X) does not occur, the accident of FEDFER can be guaranteed not to happen. Second, if the 7 basic events that would create fires or the 4 basic events that would cause gas leakages or the 5 basic events that would cause risk-avoiding devices to fail do not occur, fire and explosion accidents will not occur.

Without considering the probability that each basic event will occur, the structure importance degree coefficient of the natural gas concentration reaching the LEL is the biggest one, which means that reaching the LEL is the most important condition triggering the accident of FEDFER. It is followed by the structure importance degree of 4 basic events (X8, X9, X10, and X11) that would cause a gas leakage. The 5 basic events (X12, X13, X14, X15, and X16) lead to the failure of risk-avoiding devices failure. The 7 basic events (X1, X2, X3, X4, X5, X6, and X7) that would lead to a fire are in the sub-important position. In view of the above analysis, to effectively prevent the accident of FEDFER, the following measures should be taken:

1) The natural gas concentration in the engine room should be monitored along with the improved maintenance and repair of detection alarm devices and safety valves. The scope of monitoring should be expanded to avoid detection blind areas;

2) The pipelines, valves, joints, and engines should be regularly checked for protection against corrosion perforation. Un-tight seals should be prevented, and the welding quality in the joint should be strictly ensured;

3) The safety attachments of valves and flanges should be chosen correctly to ensure the intrinsic safety of equipment;

4) The ventilation of the engine room should be maintained, and the ventilation system should be regularly checked to ensure that the power supply circuit is in a good condition and the ventilator is operating normally;

5) The safety management and fire detection systems should be strengthened, and smoking in or near the engine room should be forbidden. It is vital to strengthen the management of engine room combustibles;

6) The rules and regulations of hot work in the engine rooms should be strictly implemented;

7) It should be forbidden to use the ironware to hammer the ground, pipeline, and equipment. When the equipment is in need of repair, explosion-proof tools must be used;

8) Anti-static overalls and personal labor protection should be worn in accordance with safety regulations while working;

9) Necessary fire fighting devices should be installed, and it must be ensured that these devices could be easily accessed and operated when dangers occur; and

10) Great importance should be attached to the human factors in the accident. To effectively train the staff, the training of good professional quality and related skills should be developed thus enabling the personnel to take appropriate and effective measures to avoid accident risks.

5 Conclusions

The method of fault tree analysis is used in this paper to model FEDFER. With this analysis, the laws for the occurrence of FEDFER are mastered and effective measures are proposed to prevent accidents. In the operation process of the ship, the relevant suggestions and rules should be strictly implemented. The incidence of FEDFER should be reduced to the utmost extent possible.

Due to lack of historical data regarding the probability of the basic events occurring, only a qualitative analysis of the accidents is performed in this paper. If reliable probability values of basic events can be obtained, accurate analysis results can be obtained. Therefore, it is necessary to establish and perfect the information database of dual fuel ships and strengthen the risk assessment research of random fuzzy factors and the probability model research on the dual fuel engine rooms.

References
Chen Wei, 2013. Speed up the application of dual fuel (diesel/ natural gas) power technology to ships in Yangzi River. China Water Transport, 13(8), 3–5.
Du Yanan, 2012. Consequences analysis of affiliated pipeline leakage of dual fuel ships. Master thesis, Jiangsu University of Science and Technology, Zhenjiang, 117-120. (in Chinese)
Fan Hongjun, Zhang Hui, Xu Jianyong, 2013. Pool fire danger distance analysis of barge. J. China Shipbuilding (Chinese edition), 52(4), 77–82.
Fay JA, 2003. Model of spills and fires from LNG and oil tanker. Journal of Hazardous Materials, B96, 171–188.
Huang Zhenghua, Li Jianhu, 2012. Assessment of fire risk of gas pipeline leakage in cities and towns. Procedia Engineering, 3(1), 77–82.
Lei Shi, Jian Shuai, Kui Xu, 2014. Fuzzy fault tree assessment based on improved AHP for fire and explosion accidents for steel oil storage tanks. Journal of Hazardous Materials, 8(7), 529–538.
Li Jianhua, Huang Zhenghua, 2012. Fire and explosion risk analysis and evaluation for LNG ships. Procedia Engineering, 6(3), 70–76.
Li Yueyang, Guan Yifeng, 2013. The desgn of dual fuel (diesel/ natural gas) ships gas fuel system, J. Jiangsu Ship, 30(5), 17–19.
Liao Suliang, Han Li, Liu Xinpeng, GuoKaihua, 2014. Safety quantitative risk analysis of LNG filling ships. Chinese Safety Science and Technology (Chinese edition), 10(11), 187–192. DOI:10.11731/j.issn.1673-193x.2014.11.032
Mattarelli E, Rinaldini CA, Golovitchev V, 2014. CFD-3D analysis of a light duty dual fuel (diesel/natural gas) combustion engine. J. Energy Procedia, 10(9), 929–937.
Ren Wei, Pan Xinxiang, Liu Xinjian, 2011. Safety assessment of main engine debugging during sea trial of newly-built ship based on FSA. J. China Shipbuilding, 52(1), 20–25. DOI:10.3969/j.issn.1000-4882.2011.01.019
Shi Guozheng, Zhang Hui, Fan Hongjun, 2013. Fuel injection mode research of LNG ship. J. Ship & Ocean Engineering, 6(14), 16–24.
TuHangming, 2012. Risk assessment studies of gas fuel ship power system.Master thesis. Dalian Maritime University, Dalian, 59-63. (in Chinese)
Wang Daqing, Zhang Peng, Chen Liqiong, 2014. Fuzzy fault tree quantitative analysis of LNG storage tank leakage with consideration of event interdependencies. J. China Safety Science Journal, 24(1), 1390–1398. DOI:10.3969/j.issn.1003-3033.2014.01.016
Xu Jianyong, GanShaowei, 2014. Development status and key technologies outline of LNG ships. Wuhan Specifications Institute, Wuhan. (in Chinese)
0

Article Information

Yifeng Guan, Jie Zhao, Tengfei Shi, Peipei Zhu
Fault Tree Analysis of Fire and Explosion Accidents for Dual Fuel (Diesel/Natural Gas) Ship Engine Rooms
Journal of Marine Science and Application, 2016, 15(3): 331-335
DOI: 10.1007/s11804-016-1366-6

Article History

Received date: 2015-11-18
Accepted date: 2016-03-02