Simulating impacts of disruption in a network of chemical manufacturing plants and supporting infrastructures

In our study, we looked at how the impact of disruptions may be modelled in a network of chemical manufacturing plants. Given that the number of plants involved is large, it is not time-feasible to model the operation of each plant in detail. However, after 1 round of quick analysis, we can selectively model the identified critical components in greater detail. Hence, the challenge is to develop a standard template that can capture sufficient information about a plant, to give meaningful result. The standard template can be applied to all plants in our study and also allows a non-technical user to easily represent a new plant and integrate into the existing model. Introduction In many countries, the chemical industry, as a whole, contributes significantly to the economy. When a company decides to set up a chemical plant, this decision will also influence its suppliers to build their plants in the vicinity. Apart from economical considerations, the physical necessity of delivering the chemical in quantity to respective plants makes it necessary to site their own plants near to its customers. Hence, the chemical industry generates a huge investment potential for the country. The stability and security of the country is one major factor that a company considers when it makes investment. With the recent global terrorist threats, this makes it even more necessary to ensure that a safe working environment is in place in order to attract investment. The fixed-asset investment of a chemical manufacturing plant is not a small amount. To recoup the huge investment cost, most plants will operate every hour of the day if possible. A disruption to the operation of the manufacturing plant will mean a large production loss, amounting to millions of dollars. The companies may suffer further monetary loss as recovery works (such as cleansing of toxic chemical spill and equipment replacement etc.) are always necessary after a disruption. For example, a sudden electricity disruption may damage some machines due to permanent solidification of the accumulated chemicals. It is logical to build redundancy or backup in a critical system. In a chemical plant, the feedstock storage will act as safety stock when a supplier is disrupted. It is common for companies to keep spare parts for a certain critical component of the manufacturing 1 This article does not reflect the viewpoints of DSO National Laboratories. All information and views expressed here are the sole responsibility of the authors. 2 Email of corresponding author: [email protected] process. Some plants will have multiple suppliers for a certain critical feedstock. Although this will reduce the impact of disruption, these measures create additional costs to the companies. Given limited resources and increased security concerns, it is impossible to protect the plant operation from all possible disruptions. It is costly for a company to maintain an electrical generator that can support a plant operation. Often, most companies will only maintain a backup generator to ensure safe shutdown of the chemical process during an emergency. To keep the safety stock, the companies have to build additional storage tank or rent spot tank from logistics companies. It is not economical to keep an excessive amount of feedstock. Due to the hazardous property of certain chemicals, plants are also not allowed under safety regulations to keep additional storage. This motivates the study to assess the impact of disruption in a network of chemical manufacturing plants. It will help to address some critical concerns of the companies and the government: Which public infrastructures (that support the chemical industry) should be protected? For a given mitigation measure, how much of the impact of a disruption be reduced? How much safety stock should a plant keep for each feedstock? What is the impact on a plant’s operation when a main customer demand is disrupted? There are two main contributions in this work. Firstly, it presents a general framework on how disruptions in a large network of manufacturing plants may be studied on a reasonable study time scale. Based on our limited literature survey, we have not found an article that discusses this issue. Secondly, we develop a model template that facilitates the modelling of the plant’s operation. This template is useful for a quick analysis of the plants. After the critical components are identified from the analysis, the analyst can then focus on modelling these critical components in greater detail. Methodology Problem definition and scope In this study, we are faced with a large number of chemical manufacturing plants. It is not feasible to model the chemical process of each plant in detail. Currently, as the companies are required to conduct own internal quantitative risk assessments before setting up their plants, we decide to focus on the physical flow interaction between plants. It is still complicated to model the physical flow between plants. A plant may receive hundreds of chemicals from different suppliers. The suppliers may be located overseas and the chemicals are delivered via road or sea links. A finished product from a company may also be exported to an overseas customer. As we are interested in modelling the impact of disruption within a geographical location, we decide to confine our study to this designated location. A supplier or customer located outside the area will not be considered. Instead, we will only include the transportation infrastructures that link the companies and their overseas suppliers / customers in the study. To further simplify the problem, we narrow our scope to those chemicals that are critical to the main process of the plants.