Modelling A Safety Management System Using 2 System Dynamics at the Bhopal Incidental Event 3

Abstract: In Safety Management System (SMS), the risk management plays a key role for the 11 prevention of accidents. The aim of this paper is to propose a Safety Management model using a 12 system dynamic approach to update conventional industrial safety into the new industrial safety 13 4.0 that is time to developed. This study analyzes some safety 4.0 aspects lacked in the Bhopal 14 incidental event by considering different data detected in the industrial Plant. The model proposed 15 in this paper discusses the relationships among the main causes that have contributed to the 16 occurrence of the incidental event studied, such as broken safety devices, inadequate personnel 17 experience, operator decisions, manager production strategy, policy decision, as deduced from the 18 relevant literature about Bhopal incidental dynamic. 19


23
In 1984, the well-known Bhopal disaster involved the chemical plant of the American multinational 24 Union Carbide, specialized in production of pesticides. The explosion of a methyl isocyanate storage 25 tank inside the plant caused the death of about 3000 people and poisoned tens of thousands more.

26
The incidental event started a long process of research and investigation concerning the causes and 27 effects [1], which led to the consideration that the accident could be avoided. In fact, a lack of 28 maintenance and precautions in safety operations was detected. Moreover, according to these 29 studies the large multinational Union Carbide had little concerning with safety in an emerging 30 country where the technology used was untested and faulty. A rare incident as Bhopal is not a 31 random event, but it is far more predictable, and the causes are accessible and the events potentially 32 avoidable [2].

33
Bhopal incidental event is one of the most important examples of the lack of safety due to no 34 maintenance policies, no human continuous training, no plan investments. For these reasons, the 35 Bhopal accident is well suited to highlight the impact that maintenance and the human factor have 36 on the safety management system.

37
In particular, if the maintenance was efficient, the extent of the accident would have been more

39
The Bhopal disaster has led to regulations and awareness for process safety-related activities

49
The System Dynamics is used to highlight the interrelations among the factors that mainly affect 50 maintenance and its effects on safety. This leads to highlighting the impact that maintenance and the 51 human factor could have on the safety of a high-risk plant, thus being key factors in an appropriate

57
Literature tries to support the integration among concepts and Industry 4.0 technologies, as well as 58 PSEP (Process Safety and Environmental Protection) for sustainable production systems [11].

59
Undoubtedly receiving data in real time allows the optimization of the ERP [12].

60
The paper is organized as follows: Section 2 presents the methodological approach; in Section 3 the 61 model is analyzed through a real case study concerning Bhopal incident; results are analyzed in 62 Section 4. A discussion about how Industry 4.0 could reduce system failure probability is made in 63 Section 5. Finally, conclusion of the research is summarized in Section 6.

147
The interaction of all these variables, above described, leads to the creation of DANGEROUS

171
The relationship between the safety plant level and its workforce is shown in Figure 6.  In Figure 5, the value of the minimum work force required under standard 176 conditions, in order to prevent accidents in the plant, is also shown. Based on this 177 value function has been defined in order to prevent the possibility of 178 non-inspection of the plant. 179 The function of human error has been defined separately using the HCR method 180 (Figure 6) [24]. 181 182 Figure 6SD diagram -HCR loop

183
It is necessary also to consider the effects/actions on the maintenance plan to close 184 the loop of this subsystem. In fact, operators could make errors by carrying out 185 maintenance. These errors may be reduced by improving the operators' skills 186 (Figure 7). 187 188 Figure

198
The results analysis permits to focus on different aspects of the Bhopal plant: as shown in Figure 10,

213
Obviously, the quantities released will not always be the same; the greater quantity of material 214 will be released in the first days of activity. This leads to the greater outlay of money in order to 215 support the costs of restoring plant, insurance and image.

216
As shown in Figure 12, these costs are significantly higher than the costs that the plant has 217 normally planned to support for the standard maintenance and personal management activities. On the other hand, as it is noticeable in Figure 13, the ordinary costs of the plant 221 (maintenance, personnel) are linear and are much lower than the plant losses due to 222 accidents of a 1:50 ratio. After the first major release, many micro losses generate daily costs derived from 226 loss function that still significantly increases the costs that the plant must support 227 with a 1:20 ratio. 228 While the loss function turns out to the end of the economic loop of this model, the 229 investments in maintenance and human resources are at the starting points by 230 influencing downstream the whole operation of the plant. 231 As far as human management is concerned, the total workforce is defined by the 232 investments made for staff. 233 The Figure 14 shows the maintenance-planning index follows the trends of human 234 error according to the planning. The SYSTEM DAMAGE FACTOR ( Figure 15) reaches a peak value with the release 238 of greatest quantity of material, which also corresponds to a peak of the 239 management system factor, in which loss inspection likelihood and effectiveness 240 maintenance index flow), while the overall damage factor immediately settles at 241 maximum alert values even before the accident occurs. 242 243 Figure  Despite the use of Industry 4.0 concept and technology, some risks may arise from 271 these implementations, as highlighted by Badri, Boudreau-Trudel and Souissi in a 272 recent paper [27].

273
The use of System Dynamics allows to understand the importance of complex 274 system and underlines the interactions among different management areas and can 275 help to prevent possible critical events. 276 However, the proposed complex system can be face up increasing the operators' 277 economic resources in terms of specific technical training. This bound could restrict 278 the model application in the small industrial companies. 279 In future, the general model proposed could be adapted to different complex plants 280 according to a Safety Management approach towards Safety 4.0. This will allow to 281 detect potential hazardous situations and to control risk. 282