In recent decades, many attempts have been made to establish the cause-effect relationship model of accidents, while little work has been carried out to comprehensively consider the interdependence between the causal factors and their complex interactions with the accident outcomes. In this study, a novel accident analysis approach based on Bayesian networks (BNs) was proposed to achieve quantitative accident analysis and dynamic risk prediction of accident types and consequences. To develop the BN-based accident analysis model, a total of 1144 accident cases occurred in tank farm of China from 1960 to 2018 were collected. The BN model that can comprehensively characterize the dependencies among accident elements was established through structural learning based on accident case analysis and parameter learning based on EM algorithm. The reliability and validity of the BN model were verified by k-fold cross-validation method and comparison of predicted data with real data, and the results showed that the BN model had good classification and prediction performance. Furthermore, the established BN model was applied to the accident occurred in Huangdao, China. The analysis results show that not only the accident outcome can be accurately predicted, but also the hidden correlation can be deeply explored through the established BN model. The proposed method and findings can provide technical reference for accident investigation and analysis, and provide decision support for accident prevention and risk management.
ReniersGCozzaniV. Domino effects in the process industries: modelling, prevention and managing. Amsterdam: Elsevier B.V, 2013. pp.1–372.
2.
FabianoBCurròF. From a survey on accidents in the downstream oil industry to the development of a detailed near-miss reporting system. Process Saf Environ Prot2012; 90: 357–367.
3.
DrogarisG. Learning from major accidents involving dangerous substances. Saf Sci1993; 16: 89–113.
4.
KonstandinidouMNivolianitouZMarkatosN, et al. Statistical analysis of incidents reported in the Greek petrochemical industry for the period 1997–2003. J Hazard Mater2006; 135: 1–9.
5.
HemmatianBAbdolhamidzadehBDarbraRM, et al. The significance of domino effect in chemical accidents. J Loss Prev Process Ind2014; 29: 30–38.
6.
NivolianitouZKonstandinidouMMichalisC. Statistical analysis of major accidents in petrochemical industry notified to the major accident reporting system (MARS). J Hazard Mater2006; 137: 1–7.
7.
VinnemJEHestadJAKvaløyJT, et al. Analysis of root causes of major hazard precursors (hydrocarbon leaks) in the Norwegian offshore petroleum industry. Reliab Eng Syst Saf2010; 95: 1142–1153.
8.
KonstandinidouMNivolianitouZKefalogianniE, et al. In-depth analysis of the causal factors of incidents reported in the Greek petrochemical industry. Reliab Eng Syst Saf2011; 96: 1448–1455.
9.
Al-GhamdiAS. Using logistic regression to estimate the influence of accident factors on accident severity. Accid Anal Prev2002; 34: 729–741.
10.
MichalakiPQuddusMAPitfieldD, et al. Exploring the factors affecting motorway accident severity in England using the generalised ordered logistic regression model. J Saf Res2015; 55: 89–97.
11.
OnderS. Evaluation of occupational injuries with lost days among opencast coal mine workers through logistic regression models. Saf Sci2013; 59: 86–92.
12.
LiZWangEOuJ, et al. Hazard evaluation of coal and gas outbursts in a coal-mine roadway based on logistic regression model. Int J Rock Mech Min Sci2015; 80: 185–195.
13.
FengJZWangYPengJ, et al. Comparison between logistic regression and machine learning algorithms on survival prediction of traumatic brain injuries. J Crit Care2019; 54: 110–116.
14.
ChangL-YChienJ-T. Analysis of driver injury severity in truck-involved accidents using a non-parametric classification tree model. Saf Sci2013; 51: 17–22.
15.
RivasTPazMMartínJE, et al. Explaining and predicting workplace accidents using data-mining techniques. Reliab Eng Syst Saf2011; 96: 739–747.
16.
CombertiLDemichelaMBaldissoneG. A combined approach for the analysis of large occupational accident databases to support accident-prevention decision making. Saf Sci2018; 106: 191–202.
17.
LiaoC-WPerngY-H. Data mining for occupational injuries in the Taiwan construction industry. Saf Sci2008; 46: 1091–1102.
18.
VermaAKhanSDMaitiJ, et al. Identifying patterns of safety related incidents in a steel plant using association rule mining of incident investigation reports. Saf Sci2014; 70: 89–98.
19.
AlikhaniMNedaieAAhmadvandA. Presentation of clustering-classification heuristic method for improvement accuracy in classification of severity of road accidents in Iran. Saf Sci2013; 60: 142–150.
20.
BayamELiebowitzJAgrestiW. Older drivers and accidents: A meta analysis and data mining application on traffic accident data. Expert Syst Appl2005; 29: 598–629.
ChengC-WLinC-CLeuS-S. Use of association rules to explore cause–effect relationships in occupational accidents in the Taiwan construction industry. Saf Sci2010; 48: 436–444.
23.
GohYMUbeynarayanaCU. Construction accident narrative classification: an evaluation of text mining techniques. Accid Anal Prev2017; 108: 122–130.
24.
HajakbariMSMinaei-BidgoliB. A new scoring system for assessing the risk of occupational accidents: A case study using data mining techniques with Iran's Ministry of Labor data. J Loss Prev Process Ind2014; 32: 443–453.
25.
ByunJ-ESongJ. A general framework of Bayesian network for system reliability analysis using junction tree. Reliab Eng Syst Saf2021; 216: 107952.
26.
GuoXJiJKhanF, et al. Fuzzy Bayesian network based on an improved similarity aggregation method for risk assessment of storage tank accident. Process Saf Environ Prot2021; 149: 817–830.
27.
LamCYCruzAM. Risk analysis for consumer-level utility gas and liquefied petroleum gas incidents using probabilistic network modeling: A case study of gas incidents in Japan. Reliab Eng Syst Saf2019; 185: 198–212.
28.
LamCYTaiK. Network topological approach to modeling accident causations and characteristics: analysis of railway incidents in Japan. Reliab Eng Syst Saf2020; 193: 106626.
29.
ZhangLWuSZhengW, et al. A dynamic and quantitative risk assessment method with uncertainties for offshore managed pressure drilling phases. Saf Sci2018; 104: 39–54.
30.
DarwicheA. Modeling and reasoning with Bayesian networks. Cambridge: Cambridge University Press, 2009.
31.
GeorgePGRenjithVR. Evolution of Safety and security risk assessment methodologies towards the use of Bayesian networks in Process Industries. Process Saf Environ Prot2021; 149: 758–775.
32.
AdedigbaSAKhanFYangM. Process accident model considering dependency among contributory factors. Process Saf Environ Prot2016; 102: 633–647.
33.
KabirSPapadopoulosY. Applications of Bayesian networks and Petri nets in safety, reliability, and risk assessments: A review. Saf Sci2019; 115: 154–175.
34.
LiXChenGZhuH. Quantitative risk analysis on leakage failure of submarine oil and gas pipelines using Bayesian network. Process Saf Environ Prot2016; 103: 163–173.
35.
XinPKhanFAhmedS. Dynamic hazard identification and scenario mapping using Bayesian network. Process Saf Environ Prot2017; 105: 143–155.
36.
RathnayakaSKhanFAmyotteP. SHIPP methodology: Predictive accident modeling approach. Part I: Methodology and model description. Process Saf Environ Prot2011; 89: 151–164.
37.
ZhouYLiCZhouC, et al. Using Bayesian network for safety risk analysis of diaphragm wall deflection based on field data. Reliab Eng Syst Saf2018; 180: 152–167.
38.
ZhangDYanXPYangZL, et al. Incorporation of formal safety assessment and Bayesian network in navigational risk estimation of the Yangtze River. Reliab Eng Syst Saf2013; 118: 93–105.
39.
WangWShenKWangB, et al. Failure probability analysis of the urban buried gas pipelines using Bayesian networks. Process Saf Environ Prot2017; 111: 678–686.
40.
TangKParsonsDJJudeS. Comparison of automatic and guided learning for Bayesian networks to analyse pipe failures in the water distribution system. Reliab Eng Syst Saf2019; 186: 24–36.
41.
CNPC. Selection of 2003-2005 incident cases of the China national petroleum. Beijing: China Petrochemical Press, 2006.
42.
CNPC. Selection of typical incident cases in 2006. Beijing: China Petrochemical Press, 2007.
43.
CNPC. Selection of typical incident cases in 2007. Beijing: China National Petroleum Corporation Safety and Environmental Protection Department, 2008.
44.
CNPC. Selection of typical incident cases in 2008. Beijing: China National Petroleum Corporation Safety and Environmental Protection Department, 2009.
45.
CNPC. Selection of typical incident cases in 2009. Beijing: China National Petroleum Corporation Safety and Environmental Protection Department, 2010.
46.
FanJ. Accident analysis of thousand cases in tank farm. Beijing: China Petrochemical Press, 2005.
47.
HouX. Accident analysis of petrochemical enterprises. Beijing: China Petrochemical Press, 2014.
48.
HouLWuXWuZ, et al. Pattern identification and risk prediction of domino effect based on data mining methods for accidents occurred in the tank farm. Reliab Eng Syst Saf2020; 193: 106646.
49.
WuXHouLWenY, et al. Research on the relationship between causal factors and consequences of incidents occurred in tank farm using ordinal logistic regression. J Loss Prev Process Ind2019; 61: 287–297.
50.
WuXHouLWuZ. Quantitative analysis of the cause-effect relationship of incidents occurred in tank farm based on the method of logistic regression. Pressure vessels and piping conference: American Society of Mechanical Engineers, 2019, p. V007T07A23.
51.
KidamKHussinNEHassanO, et al. Accident prevention approach throughout process design life cycle. Process Saf Environ Prot2014; 92: 412–422.
52.
de DianousVFiévezC. ARAMIS project: A more explicit demonstration of risk control through the use of bow-tie diagrams and the evaluation of safety barrier performanc. J Hazard Mater2006; 130: 220–233.
SobralJGuedes SoaresC. Assessment of the adequacy of safety barriers to hazards. Saf Sci2019; 114: 40–48.
55.
AbdolhamidzadehBAbbasiTRashtchianD, et al. Domino effect in process-industry accidents – an inventory of past events and identification of some patterns. J Loss Prev Process Ind2011; 24: 575–593.
56.
CozzaniVGubinelliGSalzanoE. Escalation thresholds in the assessment of domino accidental events. J Hazard Mater2006; 129: 1–21.
57.
PaltrinieriNTugnoliABustonJ, et al. Dynamic procedure for atypical scenarios identification (DyPASI): A new systematic HAZID tool. J Loss Prev Process Ind2013; 26: 683–695.
DempsterAPLairdNMRubinDB. Maximum likelihood from incomplete data via theEMAlgorithm. J R Stat Soc Ser B1977; 39: 1–22.
60.
LauritzenSL. The EM algorithm for graphical association models with missing data. Comput Stat Data Anal1995; 19: 191–201.
61.
MarcotBGHaneaAM. What is an optimal value of k in k-fold cross-validation in discrete Bayesian network analysis?Comput Stat2021; 36: 2009–2031.
62.
HuJLTangX-WQiuJ-N. Assessment of seismic liquefaction potential based on Bayesian network constructed from domain knowledge and history data. Soil Dyn Earthq Eng2016; 89: 49–60.
63.
MengHAnXXingJ. A data-driven Bayesian network model integrating physical knowledge for prioritization of risk influencing factors. Process Saf Environ Prot2022; 160: 434–449.
