ABSTRACT 
Title of Thesis: RESILIENCE ASSESSMENT OF SUPPLY-
CHAIN NETWORK INFRASTRUCTURE 
  
 Sally Saleem, Master of Science, 2020 
  
Thesis directed by: Professor Bilal M. Ayyub, Department of Civil 
and Environmental Engineering 
This thesis proposes a discrete event simulation model to investigate the complex 
interactions among supply chain components and infrastructure (1) under normal 
conditions when the facility can maintain the continuity of its performance under 
uncertainty and (2) under disruption scenarios. The model shows the flow of the risk in 
the supply chain. A special focus is given to risks from both (1) the reliability of 
infrastructure affecting materials necessary to produce a final product and (2) failures 
of infrastructure that support the transportation of materials. Bayesian network 
provides necessary models to analyze the different types of infrastructure supporting 
the supply chain, including dependencies, in simplified terms. This thesis provides a 
quantitative assessment of supply chain resilience to characterize the impact of events 
that threaten supply chains due to uncertainty in infrastructure performance, including 
potential infrastructure failures. Results were used to demonstrate the development of 
mitigation plans to reduce the impact of infrastructure disruption on the supply chain 
to overcome the possible risks associated with infrastructure failure.  
 
 
 
RESILIENCE ASSESSMENT OF SUPPLY CHAIN NETWORK 
INFRASTRUCTURE 
  
by 
Sally Saleem 
 
Thesis submitted to the Faculty of the Graduate School of the 
University of Maryland, College Park, in partial fulfillment 
of the requirements for the degree of 
Master of Science 
2020 
 
 
 
 
Advisory Committee: 
Professor Bilal M. Ayyub, Chair/ Advisor 
Research Professor Chung C. Fu 
Assistant Professor Michelle Bensi 
Dr. Jennifer Helgeson (National Institute of Standards and 
Technology) 
 
 
 
 
 
 
 
 
? Copyright by 
Sally Saleem 
2020 
 
 
 
 
 
 
 
 
 
 
 
 
Acknowledgments 
First, I would like to express my sincere appreciation to my advisor, Professor Ayyub, 
for his continuous support of my research, for his unwavering guidance in writing the 
thesis, and for his patience. I hope that my work in this thesis can be a valuable 
contribution to the great work he is leading. I would not have been able to finish this 
work without him. I also want to thank him for believing in me and offering me this 
great opportunity. 
 
I would also like to acknowledge National Institute of Science and Technology for 
funding my work. My sincere thanks goes to Dr. Helgeson who provided me a great 
opportunity to join her team, without her valuable support it would not be possible to 
conduct this work. I also want to thank my teammate Yalda for all the time we worked 
together and the stories we shared. Sincere thanks go to my thesis committee, Professor 
Fu and Professor Bensi for taking the time out of their busy schedules to go over my 
work. And for the inspiring classes I attended with them.  
 
Also, I would like to thank Sara and my uncle Sufian who have been a great family to 
me here in the States. I am grateful for all of your support. A special thanks goes to 
Sara for reviewing this thesis. Finally, my very profound gratitude to my parents for 
their love and support. Thank you both for being there for me to achieve my dreams. 
My sisters, my brother, my grandmother, Sakeenah, my dearest friends: Mohammad, 
Saba, Dua?a, Sirgio, Thuy-An, Rohoollah, and Siqi, thank you for being there for me. 
ii 
 
Table of Contents 
Chapter 1: Introduction ................................................................................................. 1?
1.1 Background ......................................................................................................... 1?
1.2 Objective ............................................................................................................. 4?
1.3 Structure of the Work .......................................................................................... 5?
Chapter2: Supply Chain Risks and Infrastructure: A Literature Review ..................... 7?
2.1 Supply Chain ....................................................................................................... 7?
2.2 Risks in Supply Chains ....................................................................................... 9?
2.2.1 Definition of Risk ........................................................................................ 9?
2.2.2 Source of Risk ............................................................................................ 10?
2.3 Supply Chains and Infrastructure...................................................................... 11?
2.4. Supply Chain Resilience .................................................................................. 13?
Chapter 3: Methodology ............................................................................................. 16?
3.1 System Description ........................................................................................... 17?
3.1.1 Supply Chain Components ........................................................................ 19?
3.1.2 Infrastructure .............................................................................................. 19?
3.1.3 Failure Modes and Associated Risks ......................................................... 21?
3.2 Discrete Event Simulation ................................................................................ 25?
3.2.1 Background ................................................................................................ 25?
3.2.2 Methods and Software ............................................................................... 27?
3.3 Bayesian Network ............................................................................................. 33?
3.3.1 Background ................................................................................................ 33?
iii 
 
3.3.2 Methods and Software ............................................................................... 33?
3.3.3 Bayesian Network Limitations .................................................................. 37?
3.4 Study Approach ................................................................................................ 37?
3.4.1 Steady-State ............................................................................................... 38?
3.4.2 Imposed Disruption Scenarios ................................................................... 38?
3.4.3 Supply Chain Resilience Assessment ........................................................ 38?
3.4.4 Identification and Assessment of Mitigation Strategies ............................ 41?
Chapter 4: Case Studies .............................................................................................. 44?
4.1 Model Characteristics and Assumptions ........................................................... 44?
4.1.1 Raw Materials ............................................................................................ 45?
4.1.2 Reliability of Infrastructure ........................................................................ 47?
4.1.3 Manufacturing Process ............................................................................... 54?
4.1.4 Disruption Scenarios .................................................................................. 55?
4.2 Case Study 1: Food Processing ......................................................................... 55?
4.2.1 Model Construction ................................................................................... 55?
4.2.2 Scenarios and Response Plans ................................................................... 60?
4.2.3 Results and Analysis .................................................................................. 62?
4.3 Case Study 2: Medical Device Assembly ......................................................... 70?
4.3.1 Model Construction ................................................................................... 70?
4.3.2 Scenarios and Response Plans ................................................................... 75?
4.3.3 Results and Analysis .................................................................................. 81?
4.4 Discussion and Assessment of the Proposed Methodology Results ................. 89?
4.4.1 Risk and Resilience in Steady State ........................................................... 89?
iv 
 
4.4.2 Risk and Resilience in Disrupted State ...................................................... 90?
Chapter 5: Conclusion................................................................................................. 94?
Bibliography ............................................................................................................... 98?
 
v 
 
List of Tables 
 
Table 2.1 Definition of Risk ....................................................................................... 10?
Table 2.2 Definitions of Resilience ............................................................................ 14?
Table 3.1 Used blocks in the discrete event simulation model ................................... 30?
Table 4.1 Infrastructure variable names ...................................................................... 50?
Table 4.2 Reliability of the parent nodes of the BN of infrastructure ........................ 50?
Table 4.3 Reliability of communication based on the state of electrical power ......... 50?
Table 4.4 Reliability of water & wastewater based on the state of electrical power .. 51?
Table 4.5 Reliability of workforce based on the state of roadways and communication
..................................................................................................................................... 51?
Table 4.6 Raw materials of the chicken manufacturing facility ................................. 56?
Table 4.7 The number of materials provided by each infrastructure z?  .................. 57?
Table 4.8 The delay time and the capacity of Activity block in DES model for case 
study 1 ......................................................................................................................... 59?
Table 4.9 Total market distribution of the canned chicken sausage ........................... 60?
Table 4.10 Local market distribution of the canned chicken sausage ........................ 60?
Table 4.11 Fulfillment rate of chicken sausage distribution to grocery stores ........... 68?
Table 4.12 Resilience index of facility AB ................................................................. 70?
Table 4.13 Raw materials of the IV bags and medical kits ........................................ 71?
Table 4.14 The number of materials provided by each infrastructure z?  ................ 72?
vi 
 
Table 4.15 The total delay time and the capacity of Activity block in DES model for 
case study 2 ................................................................................................................. 74?
Table 4.16 Total market distribution of the IV bags ................................................... 75?
Table 4.17 Total market distribution of medical kits .................................................. 75?
Table 4.18 Fulfillment rate of IV bags distribution to the USA for design plan one . 87?
Table 4.19 Resilience index of facility AB ................................................................. 88?
 ?
 
 
  
vii 
 
List of Figures 
Figure 2.1 Time-series from Google Ngram Viewer to show the history of literature for 
the supply chain from Google Ngram site .................................................................... 8?
Figure 2.2 Supply chain and infrastructure ................................................................. 13?
Figure 3.1 Proposed methodology to understand the interaction between infrastructure 
and supply chain including the influence of infrastructure reliability on supply chain 
resilience ..................................................................................................................... 18?
Figure 3.2 Supply chain and the different types infrastructure supporting the supply 
chain???????????????????????????????.20?
Figure 3.3 Risk flow in the supply chain .................................................................... 22?
Figure 3.4 Discrete event simulation (de Lara et al. 2014) ......................................... 25?
Figure 3.5 Discrete event simulation terminology ...................................................... 29?
Figure 3.6 A Bayesian network .................................................................................. 35?
Figure 3.7 Disruption profile (Sheffi and Rice Jr 2005) ............................................. 40?
Figure 3.8 DES model development steps .................................................................. 43?
Figure 4.1 The characteristics and the assumptions that are required to construct the 
DES model .................................................................................................................. 45?
Figure 4.2 Raw materials in the DES model............................................................... 46?
Figure 4.3 Bayesian Network of infrastructure in the DES model ............................. 48?
Figure 4.4 Parent nodes and child nodes of the BN of infrastructure in the DES model
..................................................................................................................................... 52?
Figure 4.5 Analysis of BN of the different types of infrastructure supporting the supply 
chain in the DES model .............................................................................................. 53?
viii 
 
Figure 4.6 Delay time in the DES model .................................................................... 54?
Figure 4.7 Structure of facility AB in DES model ...................................................... 57?
Figure 4.8 Facility AB under normal state .................................................................. 61?
Figure 4.9 Facility AB under disrupted state .............................................................. 61?
Figure 4.10 Facility AB under a disrupted state with the mitigation plan .................. 62?
Figure 4.11 The amount of chicken provided to the facility ....................................... 64?
Figure 4.12 Fulfillment rate of chicken sausage distribution to grocery stores .......... 65?
Figure 4.13 Fulfillment rate of the chicken sausage distribution to retailers .............. 66?
Figure 4.14 Fulfillment rate of the chicken sausage distribution to restaurants ......... 66?
Figure 4.15 Fulfillment rate of the chicken sausage distribution to the global market
..................................................................................................................................... 67?
Figure 4.16 Fulfillment rate of the chicken sausage distribution to the storage facility
..................................................................................................................................... 67?
Figure 4.17 Structure of facility AC in DES model.................................................... 72?
Figure 4.18 Facility AC with design plan one under normal state ............................. 76?
Figure 4.19 Facility AC with design plan one under disrupted state .......................... 76?
Figure 4.20 Facility AC with design plan one under a disrupted state with a mitigation 
plan of using the inventory ......................................................................................... 77?
Figure 4.21 Facility AC with design plan one under a disrupted state with a mitigation 
plan of reallocating the resin film ............................................................................... 78?
Figure 4.22 Facility AC with design plan one under normal state ............................. 78?
Figure 4.23 Facility AC with design plan two under a disrupted state with a mitigation 
plan of using the inventory ......................................................................................... 79?
ix 
 
Figure 4.24 Facility AC with design plan two under a disrupted state with a mitigation 
plan of using the inventory ......................................................................................... 80?
Figure 4.25 Facility AC with design plan two under a disrupted state with a mitigation 
plan of reallocating the resin film ............................................................................... 80?
Figure 4.26 The amount of resin provided to the facility for design plan one ........... 82?
Figure 4.27 The amount of resin provided to the facility for design plan two ........... 82?
Figure 4.28 Fulfillment rate of IV bags distribution to the USA for design plan one 84?
Figure 4.29 Fulfillment rate of IV bag distribution to the local market for design plan 
one ............................................................................................................................... 84?
Figure 4.30 Fulfillment rate of IV bag distribution to other countries for design plan 
one ............................................................................................................................... 85?
Figure 4.31 Fulfillment rate of IV bag distribution to the USA for design plan two . 85?
Figure 4.32 Fulfillment rate of IV bag distribution to the local market for design plan 
two............................................................................................................................... 86?
Figure 4.33 Fulfillment rate of IV bag distribution to other countries for design plan 
two............................................................................................................................... 86?
 
x 
 
Chapter 1: Introduction 
Supply chain disruption due to the failure of infrastructure is a significant problem for 
supply chains world-wide (FEMA 2017 and Fuller 2011). The  global interconnected 
nature of supply chains amplify the consequences of any possible disruption (Goldberg 
2020).  It is an inevitable reality that decision-makers must consider and address. Most 
of the literature reviewed focused on analyzing the performance of a facility under a 
disrupted infrastructure state, where the facility is incapable of performing its primary 
goal of satisfying the customers in the right place and at the right time.  
 
This work is aiming to achieve an adequate understanding of the complex interaction 
between supply chain components and supporting infrastructure ? under disruptive 
conditions and additionally under normal conditions when the facility can maintain the 
continuity of its performance, albeit with uncertainty. This uncertainty depends on the 
infrastructure reliability, which affects the number of materials that can be processed 
to a final product. Bayesian network analysis is performed to model the dependencies 
of the different infrastructure types that support the supply chain. Discrete Event 
Simulation (DES) is used to analyze supply chain performance and to characterize the 
risk that threatens supply chains due to the failure of infrastructure by focusing on 
quantifying the impact of the disturbance. 
1.1 Background 
Hurricanes Harvey, Irma, and Maria caused around $265 billion damage to the United 
States economy during the 2017 hurricane season (FEMA 2017). Failure of supply 
1 
 
chains in various sectors caused that huge economic loss. Some supply chains have not 
recovered yet after Hurricane Maria (FEMA 2017). Failure of supply chains plays a 
fundamental role in world economies (Fuller 2011). Therefore, it is important to assure 
the continuity of supply chain performance. A disruption to the supply chain can lead 
to enormous economic losses. Natural disasters are one of the primary causes of supply 
chain disruption (Industry Star 2018). For the purposes of this thesis, the focus of 
disruption will be due to the occurrence of unpredictable events such as natural 
disasters. that negatively impacts people and society besides the supply chain 
performance (Madu and Kuei 2017).  
 
Boeing Center (2017) indicated that 25% of the oil refiners in the United States were 
closed by the end of August after Hurricane Katrina hit the Gulf Cost of the United 
States, which caused massive disruption to the gasoline supply chain in the Southeast 
and the Midwest, thereby creating a rise in gas prices. 
 
Supply chains are vulnerable to natural disasters, such as earthquakes and hurricanes, 
due primarily infrastructure damage. The destructive flooding and high winds that 
accompany hurricanes often cause closure of ports, power outages, closed roads, water 
shortages, and communication loss. Disruptions propagate through the supply chains 
causing harm to some physical parts of the supply chain and a lot of chaos to the flow 
of materials and information. The delivery of raw materials or final products can be 
delayed, and manufacturing processing required to turn the raw materials into a good 
or service can be halted for a while. The consequences of damage to the supply chain 
2 
 
reach beyond the local origin of a given supply chain. Infrastructure damage sustained 
during hurricanes Katrina and Harvey serve as good examples of how damage to 
infrastructure can affect supply chain performance regardless of the distance between 
the point of the disaster and the supply chain.  Some airports managed to resume 
operations in a very short time, but around 1,000 flights over a day or two were delayed. 
Other airports were disrupted for weeks. Ports were highly impacted by the hurricane. 
Port Arthur was completely flooded. Some railways and highways were shut down 
halting many deliveries for FedEx, and USPS.  
 
Fuller (2011) mentioned that several floods happened in Thailand in 2011 that caused 
significant damage to the offices of Western Digital -- one of the pioneer companies in 
hard drive production. It provides around 25% of the world?s computer hard drives. It 
took the company approximately a year to recover and reach normal levels of 
production. During that year, customers from all over the world noticed around a 10 
percent increase in the price of hard drives. 
 
Hochfelder (2017) published statistics from the Business Continuity Institute (BCI-
Zurich) on disasters around the world in 2016: the Japanese earthquake, Canadian 
wildfires, hurricanes in Europe, and Hurricane Matthew in the US:  
? 8%   Productivity loss 
? 53% Increase in cost of working 
? 38% Reputation or image damage   
? 37% Loss of revenue  
3 
 
The magnitude of these consequences of natural disasters are unpredictable ahead of 
time. These unknown risks that threaten the supply chain make the supply chain more 
vulnerable (Christopher and Peck 2004) to disruption. Therefore, it is necessary to 
focus on increasing the resilience of the supply chain in addition to increasing its 
efficiency in order to minimize the impact of any possible risk that may impact any part 
of the supply chain. Resilience is the ability to withstand a potential disruption with 
minimal effects. A resilient supply chain can withstand those unseen disruption, or the 
consequences that follow on supply chain performance and recover quickly in a short 
period. A design plan in case of emergencies should be prepared to mitigate the risk 
and the consequences (Barroso et al. 2015). Other researchers believe supply chain 
design plays an pivotal role in supply chain resilience. Pettit et al. (2010) indicate that 
increasing the capability of the supply chain will enhance its resilience and decrease its 
vulnerability. The supply chain must be designed with a high readiness to face any 
disruption in an effective way to recover to its previous state before the disruption or 
to a better state (Ponomarov and Holcomb 2009). 
1.2 Objective 
The goal of the research is to develop a methodology for the resilience assessment for 
supply chain network infrastructure, to examine the role of infrastructure reliability on 
the performance and the resilience of the supply chain, and to propose mitigation plans 
for the impact of infrastructure failure by increasing the resilience of a supply chain 
under natural-hazard disruptions. 
 
4 
 
The objectives of this research are to: 
1. Develop a methodology for resilience assessment for supply chain network 
infrastructure. 
2. Identify the impact of the reliability of infrastructure on the supply chain 
performance. 
3. Identify the failure scenarios for a supply chain with a focus on infrastructure 
disruption. 
4. Evaluate the consequences of disruption by quantifying supply chain 
performance and assess the resilience of the supply chain during a certain 
window of time.  
5. Examine potential mitigation plans to increase the resilience of the supply 
chain. 
1.3 Structure of the Work 
The work has been divided into three parts:  
 
Part I. Acquisition of background information 
The focus herein is on linking the reliability of built infrastructure to the supply chain. 
This information was collected by way of a thorough literature review and interviewing 
workers and business owners who experienced discontinuity in their supply chain 
performance due to infrastructure disturbance. The literature review has focused on 
definition of risk in the supply chain due to infrastructure, to define the resilience of 
the supply chain, and to explore supply chain management decisions that may help to 
increase the resilience of the supply chain. 
5 
 
 
Part II. Methodology development 
Once the background information was collected, the methodology was established. 
This methodology is focused towards analyzing the interactions among different 
infrastructure types and supply chains including the influence of infrastructure 
reliability on the supply chain and to set a quantitative performance measure to assess 
the resilience of the supply chain was established. This step has been completed by 
taking the following steps: (1) defining the system ; (2) identifying the dependency 
among infrastructure using Bayesian network; (3) assessing the performance of the 
supply chain; (4) imposing a disruption scenario as a failure of one of the infrastructure 
types; (4) developing a discrete event simulation model for two case studies using 
Extendsim; (6) calculating the supply chain resilience; and (7) assessing mitigation 
strategies to manage associated risks and increase the resupply chain resilience.  
 
Part III. Evaluation and assessment of simulation results 
This final part of the work consists of running the model, tracking the logical progress 
of the work, and finally evaluating the output with the desired goal. Extendsim can 
show outputs in tables and graphics. The outputs show the movement of materials and 
information among the supply chain entities, the impact of supply chain disruption, and 
the vulnerability of entities.  
6 
 
Chapter2: Supply Chain Risks and Infrastructure: A 
Literature Review  
2.1 Supply Chain  
Supply chains (SCs) have attracted a great deal of attention in the last few years as 
businesses have realized the importance of the supply chain to their success and 
economists have studied the wide impact of supply chains on the economy. Most 
economists and researchers have focused only on increasing the profitability of supply 
chains, which can make them more vulnerable to the negative impact of uncertain 
events (Oliveira et al. 2017). The focus of the researcher has been shifted to the risk 
that threatens the supply chain resilience due to the negative consequences that 
followed some events, such as the attacks of September 11, 2001, Hurricane 
Katrina2005, Thailand flooding, and the Japanese earthquake (Schmitt and Singh 2012) 
and Hochfelder (2017). 
 
The term ?supply chain? arose in the 1980s when companies realized the importance 
of collaboration within their surroundings and beyond (Lummus and Vokurka 1999). 
The concept of a supply chain has become a popular topic since the 1990s (Cooper et 
al. 1997).  The term was originally implicitly introduced by Forrester (1958) who 
identified some essential management issue ?there will come general recognition of the 
advantage enjoyed by the pioneering management who have been the first to improve 
their understanding of the interrelationships between separate company functions and 
7 
 
between the company and its markets, its industry, and the national economy.? Figure 
2.1 provides the history of the term. 
 
 
Figure 2.1 Time-series from Google Ngram Viewer to show the history of literature 
for the supply chain from Google Ngram site, https://books.google.com/ngrams 
(2020) 
 
Several definitions of supply chains are available. The American Production and 
Inventory Control Society (APICS) Dictionary defines a supply chain as ?the processes 
from the initial raw materials to the ultimate consumption of the finished product 
linking across supplier-user companies? (Cox III et al. 1995).  
 
According to (Beamon) 1998, a supply chain is an organized manufacturing process to 
transfer raw materials into finished goods delivered to customers. Beamon considers 
supply chains to be global networks consisting of several organizations that team up 
together to assure a smooth flow of materials and information from suppliers to 
customers with the lowest possible cost to eventually gain customer satisfaction. 
 
8 
 
Other definitions are available, for example, Waters (2003) defined a supply chain as a 
collection of organizations that work together to move the materials needed for their 
process. The materials include raw ingredients, components, and finally finished goods. 
Waters (2003) pointed out that a supply chain is a group of organizations that include 
collecting, manufacturing, storing, distribution, and communication.  
 
All of the previous definitions of a supply chain unanimously agree that a supply chain 
is a combination of all the required resources, organizations, events, and processes to 
deliver a final product or service to the customer. This is the definition that is used in 
this thesis.  
 
2.2 Risks in Supply Chains  
2.2.1 Definition of Risk 
The focus on the risk that surrounds supply chains has increased since the occurrence 
of destructive events, such as the attacks of September 11, 2001and Hurricane Katrina 
(Schmitt and Singh 2012). Those events contributed to increasing attention to the 
ability of a supply chain to handle negative consequences. The concept of risk can be 
connected to uncertainty in the occurrence of an event (Ayyub 2014). Risk is about a 
negative effect if an event occurs. Table 2.1 shows some definitions found in the 
literature, all of which agree that risk relates to the following general questions: 
? What can go wrong when an event occurs? 
? How likely will that happen? 
9 
 
? What are the consequences? 
Table 2.1 Definition of Risk 
Definitions Reference  
?Risk is the negative deviation from the expected value of a (Wanger and Christoph 2008) 
certain performance measure, resulting in undesirable  
consequences for the local company.? 
?Risk is the expected outcome of an uncertain event, i.e., (Manuj and Mentzer 2008) 
uncertain events lead to the existence of risks.? 
 
?Risk is the effect of uncertainty on objectives. ? (ISO2009a)  
?Risk is the potential loss resulting from an uncertain exposure (Ayyub 2014) 
to a hazard or resulting from an uncertain event that exploits 
the system?s vulnerability.? 
 
Identifying the source of risk in a supply chain is an essential step to help to take some 
actions that might reduce the probability of the occurrence of a negative effect.  
2.2.2 Source of Risk 
Based on the literature review, risk originates from one of two sources or a combination 
of both: (1) external and (2) internal. External sources of risk to a supply chain refer to 
risk that occurs due to natural disasters, such as floods, earthquakes, and hurricanes, or 
human-made catastrophes whether intentional or not. Internal risks to a supply chain 
originate from a company's supply chain and the interaction among the different parts 
of the supply chain. This source can be related to human activities, finances, and 
materials, etc.   
10 
 
 
In this work, the focus is on risk sources related to infrastructure failure as a result of 
natural disasters. Once the source of the risk is identified, responsive action can be 
taken to reduce the impacts of that risk and enhance the supply chain performance. 
2.3 Supply Chains and Infrastructure 
The performance of a supply chain depends on the design of the chain (Barroso et al. 
2015): the components of the supply chain (suppliers, manufacturing and assembly 
plants, and distributors), and the connection among the different components. 
Infrastructure is involved in the transfer of materials, service, information, and 
operations. Some types of infrastructure are essential to the transfer of materials across 
different components of a supply chain, from the supplier to the customer. Those types 
of infrastructure are considered as primary parts of supply chain logistics. Supply chain 
logistics is about the positioning of resources or strategic management (Waters 2003), 
and  infrastructure is involved in this change of positioning. Movement of materials 
and information occurs through the infrastructure from a supplier to a customer (e.g., 
seaport, roadways, and airports). Other types of infrastructure are very important to a 
supply chain operation, such as the electrical power supply, communication, water 
plants, and wastewater.  
 
The availability of infrastructure is an important factor in the selection of physical 
locations when a company starts a business. Decision-makers typically conduct a deep 
analysis of geographical locations and their local climate to decide the number of 
suppliers, manufacturers, and distributors. Climate might affect the reliability of the 
11 
 
various types of infrastructure, such as electrical power, water supply, transportation, 
and communications.  
 
FEMA (2017) reported that during Hurricane Maria, supply chains were able to survive 
during the disaster because most of the businesses did not suffer major damage to their 
physical properties. The main challenges were caused by the damage that happened to 
the infrastructure. Figure 2.2 shows an example of a supply chain?s dependency on 
infrastructure. Each physical part of the supply chain is represented as a node, a link 
connects the nodes that indicate the flow of information and materials. The transfer of 
information and services occurred through infrastructures.  
 
Figure 2.2 shows the different types of infrastructure that are considered as primary 
supply chain components. The types of infrastructure in this thesis are limited to (1) 
electrical power; (2) communication; (3) roadways; (4) water and wastewater; (5) 
workforce. 
 
12 
 
 
Figure 2.2 Supply chain and associated infrastructure systems 
 
2.4. Supply Chain Resilience  
The resilience of a supply chain is its capability to survive negative impacts of unseen 
events and to bounce back to its original state or to move to a better state (Barroso et 
al. 2015). Table 2.2 provides several definitions of resilience. The definition of 
resilience that is used in this thesis is the one introduced by (PPD-21 2013) with the 
focus on natural disaster. 
 
 
13 
 
Table 2.2 Definitions of Resilience 
Definitions Reference 
?The ability of a system (supply chain) to return to its original (Christopher and Peck 2004) 
state or move to a new, more desirable 
state after being disturbed.? 
?The adaptive capability of the supply chain to prepare for (Priya Datta et al. 2007) 
unexpected events, respond to disruptions, and recover from 
them by maintaining continuity of operations at the desired 
level of connectedness and control over structure and 
function.? 
 
?Resilience means the ability to prepare for and adapt to (PPD-21 2013) 
changing conditions and withstand and recover rapidly from 
disruptions. Resilience includes the ability to withstand and 
recover from deliberate attacks, accidents, or naturally 
occurring threats or incidents.? 
?Resilience in supply chain context, defined as an ability of (Scholten et al. 2014) 
supply chains to recover from inevitable and unexpected 
disruptions.? 
?Resilience is the ability to prepare for and adapt to changing (Ayyub 2015) 
conditions and withstand and recover rapidly from 
disruptions.? 
 
All the definitions above agree that resilience is about preparing for unexpected events, 
so that a supply chain can withstand any possible disruption with a minimum of 
possible loss and recover from loss as soon as possible. Some disruptive events cannot 
be prevented, but their consequences can be mitigated. A supply chain can be designed 
to be resilient and withstand any possible interruption of its performance. This 
14 
 
performance measure will indicate how a supply chain acts under normal conditions 
and during a disruption. Chapter 3 explains the resilience assessment in more detail. 
This thesis includes the effects of reliability of infrastructure on the resilience of the 
supply chain under normal conditions and disturbances. The development of the 
methodology is shown in the next chapter.   
 
Finally, the risk of failure in a supply chain is inevitable. Some sources of risk are 
outside human control. With this in mind, the only way to deal with these potential 
risks is to be prepared for them with contingency plans in the event that such an event 
occurs. These plans will make the supply chain increasingly resilient. 
15 
 
Chapter 3: Methodology  
This chapter proposes a methodology to analyze the interactions among different 
infrastructure types and supply chains, including the influence of infrastructure 
reliability on the supply chain performance and to assess the resilience of the supply 
chain. Figure 3.1 summarizes the proposed methodology and the associated steps:  (1) 
define the system by identifying all the supply chain components including 
infrastructure and associated reliabilities; (2) identify the dependency among 
infrastructure and the impact of the dependency on the reliability of infrastructure using 
Bayesian network; (3) assess the performance of the supply chain by accounting for 
infrastructure reliability; (4) assess the performance of the supply chain under failed 
states of infrastructure types; (5) develop a discrete event simulation model using 
Extendsim; (6) assess the supply chain resilience and assess mitigation strategies to 
manage associated risks.  
 
The proposed methodology examines how infrastructure is linked to the supply chain 
and how variation in the reliability of infrastructure influences supply chain resilience. 
The results of the study propose solutions to minimize the impact including disruptions 
of the supply chain and to increase its resilience. The methodology depends on creating 
a quantitative model to analyze the interaction among supply chain components, the 
impact of infrastructure reliability, and disruption in the performance of the supply 
chain. The model has been generated using discrete event simulation. Simulation is a 
very effective tool to capture the interaction of system components and to test different 
response scenarios to lead to a more resilient performance. The different infrastructure 
16 
 
types supporting the supply chain are interdependent. The failure of one infrastructure 
may lead to a disruption in another one. Therefore, a Bayesian network has been used 
to capture the dependency and reflect its impact on supply chain performance. 
3.1 System Description  
This section includes a detailed description of the supply chain components that form 
the system analyzed in this thesis using DES. The system of the research consists of 
the various components of the supply chain, the infrastructure supporting it, the failure 
mode, and the associated risk. In this thesis, two case studies have been examined; each 
case study represents a manufacturing facility. Both cases include all the primary 
components of the supply chain and the different types of supporting infrastructure that 
allow the supply chain to properly operate under normal conditions (i.e., without 
disruption). Chapter 4 provides further explanations of each case study.  
 
17 
 
 
Figure 3.1 Proposed methodology to understand the interaction between infrastructure 
and supply chain including the influence of infrastructure reliability on supply chain 
resilience 
18 
 
3.1.1 Supply Chain Components  
The main components of a supply chain are suppliers, manufacturing processes, and 
distributors. For each case, a specific number of suppliers and the type of infrastructure 
they use to transfer the real materials to the facilities (i.e., roadways, airports, and 
seaports) are stated. Therefore, the infrastructure used in transferring the materials is 
critical to the continuity of the supply chain performance. Supply chain components 
are modeled using discrete event simulations as nodes connected by links ? where links 
are responsible for the movement of materials and information. 
 
Manufacturing processes are specific to the supply chain analyzed and depends on the 
study case under consideration. Each case study represents a different sector. The 
facility consists of multiple plants, where each plant is responsible for a specific step 
of the manufacturing process. The facility can also have a storage space/warehouse to 
keep the inventory. The number of outputs represents the final product of each facility. 
The distribution of the final product is divided into two markets: domestic and global. 
The percentage for each one is included in the model, but the distribution process is 
disregarded. 
3.1.2 Infrastructure  
Multiple components fall under the definition of supply chain infrastructure. The 
definition contains some physical aspects like the buildings of the manufacturers and 
distributors, and the informational aspects that are essential to run the supply chain 
(Fallon 2020). Supply chain infrastructure includes the infrastructure used to move the 
19 
 
products across the supply chain components and the other types of infrastructure that 
are essential to run the facility.?
 
Supporting infrastructure is one of the main components of supply chain logistics. That 
means it is responsible for the movement of materials and information from suppliers 
to intermediary manufacturers and customers by way of seaports, ports, and airports, 
etc. Infrastructure is also critical to manufacturing operations. The facility requires 
water, electrical power, and communication. Therefore, it is crucial to understand the 
role of infrastructure in the supply chain and the impact of potential disruption.  
Figure 3.2 shows how the different types of infrastructure are important to the 
interaction among supply chain components.  
 
 
Figure 3.2 Supply chain and the different types infrastructure supporting the supply 
chain 
 
20 
 
 Infrastructure performance under the stress of a disruption will highly impact the 
whole supply chain. The performance as a general term considers ?the capability of a 
system to fulfill the need of its functional requirements? (Ayyub 2014). Performance 
can be measured based on the problem of interest. The performance of infrastructure 
in this thesis is measured through the reliability of the infrastructure.  
3.1.3 Failure Modes and Associated Risks 
Supply chain resilience is the ability of the supply chain or any of its components to 
rebound from a setback in the occurrence of disruption (Schmitt and Singh 2012). 
Supply chain components must be prepared to face the risk of failure to maintain supply 
chain performance. 
 
One of the major steps in performing this work is to develop a strong understanding of 
how the risk flows across the system components. In other words, how the failure of 
one component can transfer to other components and eventually affect the outputs of 
the supply chain? In this research, the focus is on the risk that threatens the supply chain 
due to the variation of the reliability of infrastructure and the total failure of the 
infrastructure responsible for the logistics of materials/goods. 
 
Risk is embedded in the supply chain even under normal conditions due to the variation 
of the ability of infrastructure to provide the supply chain with services required by the 
supply chain to maintain its operations. For example, ports and roadways are 
responsible for the movement of materials; electrical power, water supply, and 
communication are susceptible to a disturbance at any time. This variation can affect 
21 
 
all the various parts of a system, and it will be reflected in the number of outputs. 
Another cause of risk is the total failure of the infrastructure that is critical to 
transportation can cause a loss of a supplier. Figure 3.3 shows different kinds of risk. 
 
 
Figure 3.3 Risk flow in the supply chain ? generalized representation  
 
3.1.3.1 Infrastructure Reliability 
The reliability of the infrastructure is identified as the performance of infrastructure. It 
measures its ability to serve the supply chain. This uncertainty potentially affects the 
services provided by the infrastructure of the supply chain. As a result, the type of 
materials and the relative magnitude of each sent by the supplier to the manufacturing 
facility of a supply chain will change due to infrastructure reliability. Probability is a 
22 
 
common way to describe the ability of the system (Ayyub 2014). Reliability is 
measured as the probability of the complementary event to failure. The reliability of 
infrastructure is calculated based on equation 3.1. 
Reliability 1 failure probability (3.1) 
This type of risk is captured in the model by reducing the value of critical raw materials 
by multiplying them by the reliability of infrastructure. Consider ?? as a random 
variable that represents the amount of critical raw material required by a facility at a 
specific time, and ?? is the random variable that represents the provided material to the 
facility based on the reliability of infrastructure. 
?? ???? (3.2) 
where,  
??: Reliability of infrastructure, 0 ?? 1 
??: The amount of required raw materials by the facility at time ? 
??: The amount of provided raw materials by infrastructure n at time  ?, 0 ?? 1 
 
Thus, each infrastructure type that supporting the supply chain produces a different 
amount of materials. The minimum amount of materials provided by infrastructure 
controls the raw materials delivered to the facility.  
 
In the discrete event model, the materials provided by each type of the supporting 
infrastructure will be presented as a random variable ?? uniformly distributed with a 
minimum value of ?? and a maximum value of ??. 
?? ? ??, ??  (3.3) 
23 
 
????? (3.4) 
z? ????? .  ? .???
 
where,  
z?: The minimum amount of provided raw materials by infrastructure at time ? 
 
3.1.3.2 Total loss of Suppliers Due to Failure in the Logistics Infrastructure   
The supply chain is highly influenced by the state of infrastructure, whether it is 
available or not. This controls the movement of materials from one component to 
another. Losing one component of the infrastructure responsible for logistics will 
negatively impact the supply chain.  This is what usually occurs during a natural 
disaster when seaports can be shut down and roadways can be blocked.  
 
The occurrence of a natural disaster can lead to complete loss of a primary part of the 
logistics system of the supply chain. Failure of infrastructure prevents suppliers from 
providing a facility with the required materials. The loss of a supplier starts at one 
component, preventing the facility of producing the required product during the 
disruption. The impact of the risk propagates through the supply chain components and 
eventually affects the number of outputs. This loss of infrastructure was modeled in the 
discrete simulation model under the imposed scenarios. Equation 3.5 represents this 
kind of risk in the model. 
?? 0 (3.5) 
24 
 
where,   
?? is the amount of provided raw materials at time ? 
3.2 Discrete Event Simulation  
3.2.1 Background 
In this section DES will be defined and the associated terms will be introduced and 
defined. Discrete event simulation is a technique to represent real-life problems as a 
system by building a model to track the behavior of system components and evaluate 
the system performance (Banks 2014). DES  is defined by a series of events. The event 
is any cause possibly leading to a change in the state of any components of the system 
that will produce different outputs during a defined unit of time. Figure 3.4 provides a 
graphical presentation of the DES process.  
 
 
Figure 3.4 Discrete event simulation (de Lara et al. 2014) 
 
25 
 
Discrete event simulation (DES) is widely accepted as a tool to analyze manufacturing 
systems. It is based on Monte Carlo methods (Brailsford et al. 2014).  
To understand discrete event simulation, several terms are introduced herein. A system 
is a group of components interacting with each other over time to perform certain goals 
that are dependent on the field of interest.  
 
The state of a system is defined as a collection of variables that contain all the necessary 
information that describes the system in terms of the impact of an event over a certain 
entity at a certain time. Model is a descriptive representation of a system by using 
mathematical, logical, and physical relationships that describe the interaction between 
the entities that form the model.  
 
An entity is any component/part of the system that requires representation and will 
cause an event to happen. So, the entity usually represents a physical object in real life. 
 
Finally, the event is the change that happens to an entity at a specific time which will 
affect the state of the entity. 
 
According to the previous definitions, the DES method examines the flow of materials/ 
information among supply chain components caused by the occurrence of an event that 
leads to the movement of an entity through many queues and services, once the service 
is available. Each component of the supply chain has specific attributes.  These 
attributes account for the logic behind the flow of the entity from the queue to a specific 
26 
 
service -- whether it is the time needed to serve the entity or a priority of one service 
over another like the sequences of a manufacturing process where specific tasks must 
be done before others.  
3.2.2 Methods and Software 
The DES model was built using a discrete event simulation software (Extendsim power 
tools for simulation ?by Imagine That Inc). ExtendSim is a powerful simulation tool 
allowing to build dynamic models of real-life from building blocks to represent the 
components of the system. It is user-friendly software with a graphical drag-and-drop 
block. This helps through the research to visualize the supply chain. The software has 
a very wide range of blocks with different roles in the model included in a specific 
library for discrete event simulation. It is easy to track the movement of information 
through the different blocks. Most of the blocks can be defined by the use to provide 
statistical information about the associated data. Blocks are connected through lines 
called connectors which direct the movement of entities from one block to another. The 
behavior of the model strongly relies on the inputs. Therefore, setting the input values 
and associated relationships was challenging to relate to real-world functioning as 
closely as possible. Tracking errors in the model was the responsibility of the user. 
Excel was used to extract some of the graphical presentations based on the need for the 
research.  
 
Simulation is a time-dependent process for the movement of entities as a part of a 
system. The movement of the entities depends on time; the link between the blocks 
represents the movement of the entities between the supply chain components. In this 
27 
 
thesis, simulation was used as a tool to gain knowledge of a supply chain in order to 
make decisions that will help to increase the resilience of a supply chain. Extendsim 
allows flexibility to change the model based on the problem requirements. The model 
deviates from real life due to some of the assumptions and simplifications that have 
been done. An ideal system should experience a very small accumulation of queues, 
with a delay time for the activity block within the assumed range.  
 
Simulation as a tool has some limitations. The model of the two cases is built based on 
research specific to each of the modeled manufacturing facility. Simplifications have 
been made so that the discrete event simulation serves the purpose of the research to 
investigate risks that threaten supply chains due to the uncertainty of the reliability of 
infrastructure. 
3.2.2.1 Model Terminology  
Each software has specific terminology related to the specific method emplyed. This 
section provides the main terminology for the DES model developed using Extendsim. 
 Figure 3.5 shows what the terminology used in the discrete event represents in the real-
life system. Understanding the terms helps to develop the discrete event model. 
? Model is a group of blocks and links to represent a system. Each block 
represents one of the supply chain components. 
? Entity is the item that is being transferred through the whole model. It is being 
processed in blocks and, eventually, it will be terminated from the model as an 
output. 
28 
 
? Attributes are characteristics to describe the entities as they move from one 
block to another; they can be changed as they move from one to another. For 
example, raw materials will be measured in a unit weight at the initial stage of 
the model when being moved from supplier to facility. Once the raw materials 
are processed to a final product, the materials can be described in a different 
unit (e.g., box, can).   
 
Figure 3.5 Discrete event simulation terminology 
 
3.2.2.2 Blocks 
The software provides a wide range of blocks. The model was built using a group of 
blocks connected by links. Links are referred to as connectors; they transfer entity and 
29 
 
information from one block to another.  Table 3.1 shows the most commonly used 
blocks in the research provided by the discrete event-specific library. 
 
Table 3.1 Used blocks in the discrete event simulation model  
 
Number Block name Block symbol Use definition 
 
1.  Create block  Generates (entities) randomly, 
or by a schedule by setting the 
time between the arrival of 
items. This is the initial start that 
leads to a change in the state of 
 the model causing the 
movement of generated items. 
2. Queue block  Keeps the entities until the next 
block is available. All blocks in 
this work are FIFO (first in-first 
out). The first entity that arrives 
is the first one to leave. 
 
3. Activity block  Executes a specific task with a 
predetermined amount of time. 
Time can be constant or random.  
 
 
 
30 
 
 
Table 3.1 Used blocks in the discrete event simulation model (cont.) 
Number Block name Block symbol Use definition 
4.  Batch block  Assembles different entities into 
one entity. It serves the purpose of 
the assembly line in the model. 
 
 
5. Select item in   Controls the path of entities going 
 into a block  
 
 
6. Select item out  Controls the path of entities going 
out of a block. 
 
 
31 
 
3.2.3 Simulation Limitations 
The work has been done to make the model as logical as possible. However, there are 
certain issues that we cannot control. We also should be aware of all the assumptions 
and simplifications that have been done to build the model. Many of the faced 
challenges have been solved by adding blocks. It is impossible to reach the perfect 
model; thus, it is helpful to point out the most basic limitations.  
 
Queues are important to the simulation model. This refers to It is the block where an 
entity is held when the activity block is unavailable. Therefore, it is helpful to include 
queues wherever there is a chance for the activity block to be occupied. This will 
prevent the accumulation of entities in the create block. Then the create block can keep 
generating items according to the original plan. In real-life, different actions would be 
taken, such as increasing the number of workers or rescheduling. 
 
The reliability of the services provided by infrastructure is presented in a create block 
at the beginning of the model showing the number of critical materials that are served 
by the infrastructure. The variation in infrastructure efficiency makes the model more 
realistic. The resulting amount of materials would thus include in real life the impact 
of infrastructure on each separate step. 
32 
 
3.3 Bayesian Network 
3.3.1 Background 
Bayesian Network (BN) theory is considered an effective approach to represent risk 
flow in a supply chain (Schmitt and Singh 2012). As mentioned earlier in section 3.2, 
infrastructure forms a primary component of the supply chain logistics. Therefore, the 
BN can be used to model the risk propagation in supporting infrastructure due to the 
variation in the reliability of services that infrastructure can provide to the supply chain.  
 
As mentioned in Section 3.1, the performance of all components of infrastructure is 
interdependent. The state of some infrastructure components depends on the state of 
others, which will eventually impact the reliability of infrastructure. For example, the 
failure of electrical power has a high impact on the water supply because most high 
elevation areas get their water through water pumps run by electrical generators. Other 
examples of the interdependency of different infrastructure components are workforce, 
communication, and roadways. The workforce needs to have safe roadways to 
commute to a facility and a reliable communication system. It is crucial to include a 
comprehensive analysis of the interconnectivity of all supporting infrastructure 
components when completing a performance assessment and making recovery 
decisions in supply chain design. 
3.3.2 Methods and Software 
A Bayesian network is an effective way to determine the conditional probability of a 
specific infrastructure type given the information about the state of another 
33 
 
infrastructure. This prior information is usually built on observation, expert opinion, 
engineering judgment, or a physical model (Bensi 2010). A model based on statistical 
analysis using a Bayesian network can always be updated once a new observation is 
obtained.  
 
BN is built based on the ?Bayes? theorem (Bensi 2010).  The Bayes rule is used to 
describe how the probability of an event changes based on prior knowledge gained 
about the occurrence of an event. Equation 3.6 describes the formula of the Bayes rule 
and some primary terminology to help to understand the rule. 
? ?|? ? ? ? ? ? ?|? (3.6) ? ? ? ? ? ? ?  
 
where,  
? ?|?  : The conditional probability of event A given that event B has occurred 
? ? ? ?  : The joint probability of events A and B 
? ?  : The marginal probability of event A 
? ?|? : The likelihood of the observed event B 
? ?  : The marginal probability of event B 
 
Typically, the prior information about event A represents a belief about the probability 
of the prior probability. A similar term is used for the calculated probability (the 
posterior probability). It represents a belief about the probability after considering the 
observation. Obtaining an observation about event B narrows the sample space of the 
34 
 
possible observation. Therefore, a normalizing factor calculated based on the Theorem 
of Total Probability appears in the denominator of equation 3.6. 
 
A Bayesian network captures the dependency among variables by using probability 
dependencies (Kabir et al. 2015). It is a probabilistic graphical modeling method to 
represent knowledge of uncertainty. It combines graph theory, probability theory, and 
conditional probability. It consists of nodes to represent the random variables in a 
system and line and arrows connect the nodes in order to describe the probabilistic 
dependency that connects the random variables. Identifying the nodes and links in the 
BN must be carefully performed to avoid any misleading and unnecessary 
complications to the model (Bensi 2010). A BN works as an analytical tool to compute 
the probability of the variable of interest -- based on the condition of a previously 
observed variable (Ojha et al. 2018). The state of the variable of interest depends on 
the state of the previously observed variable. 
 
Figure 3.6 A Bayesian network 
 
 
In the graphical representation of a BN, there are three kinds of nodes:  
35 
 
? Child nodes: the nodes with edges directed into them (X1, X2, X3), 
? Parent nodes: nodes from which the arrows direct out of them (X1, X3, X4), and 
? Root nodes: the nodes without arrows directed into them (X4). 
 
For a system A that consists of multiple random variables ??, ??,x?, ? . , ?? , the 
probability of the system is defined in this work a Bayesian network has been used to 
find the reliability of a specific infrastructure and its impact on the supply chain 
performance based on a prior probability of other infrastructure. Because the network 
of the infrastructure that supports the supply chain is small, an Excel sheet has been 
used to calculate the probability of each node based on the state of its parent nodes. 
More details will be explained in Chapter 4. 
? ? ? ??, ??, ??, ? , ??  (3.7) 
 
? ? ? ??|?????? ?? ?  ? ??|?????? ?? ?. . .? ? ??|?????? ??  (3.8) 
  
(3.9) 
? A ??|?????? ??  
To describe the relationship between a child node and all the possible parent nodes a 
Conditional Probability Table (CPT) is constructed for each of the child node. The CPT 
shows the probability of a node to be in a state, given the knowledge about the state of 
the parents. 
 
36 
 
In this work a Bayesian network has been used to find the reliability of a specific 
infrastructure and its impact on supply chain performance based on a prior probability 
of other infrastructure. As the network of the infrastructure that supports the supply 
chain is small, an Excel sheet has been used to calculate the probability of each node 
based on the state of its parent nodes. More details will be explained in Chapter 4. 
3.3.3 Bayesian Network Limitations 
A Bayesian network is an effective tool to capture the dependency between the different 
parts, but it is highly dependent on the assumptions that we make about the parent nodes 
of certain infrastructure. Assumptions have been made based on literature and some 
research related to our interest. 
 
Bayesian networks rely on conditional independence. As the number of parent nodes 
increases, the conditional probability table can become overly complicated. To avoid 
such a problem in this study, some deliberate assumptions have been made about 
connections between different kinds of infrastructure. Those assumptions make some 
of the prior knowledge about nodes biased. However, this is acceptable for the purpose 
of achieving the goal of this study. 
3.4 Study Approach  
Once the supply chain components and the reliability of infrastructure are defined and 
the risk included in the supply chain has been identified, the analysis of the supply 
chain performance can be started. A discrete event simulation model can be constructed 
using Extendsim to replicate a real-life system under normal conditions for different 
37 
 
design strategies to understand the system as a whole and the role of infrastructure in 
the supply chain. Then disruption can be imposed as a different scenario to the system 
to examine the supply chain performance under disruption and measure its resilience 
and explore the available alternatives that can be applied in the supply chain to react to 
disruption.  
3.4.1 Steady-State 
The first part of the simulation is to build a model to simulate a real-life supply chain 
for facilities. This has been done after several trials to ensure that each component of 
the supply chain is working correctly to achieve the target goal of the outputs. The 
steady-state analysis shows how a facility runs under normal conditions and how the 
supply chain is influenced by the uncertainty of the reliability of infrastructure. The 
models under steady state serve as a reference for the validation of other scenarios. The 
number of outputs in each case study depends on time.  
3.4.2 Imposed Disruption Scenarios 
Disruption scenarios were imposed on the system by halting a facility from receiving 
raw materials. The disruption scenarios in both cases occur due to a total loss of 
supporting infrastructure of the logistics system that causes a total loss of a supplier 
due to a serious failure in a seaport or a roadway that will prevent the movement of 
ships and trucks. Further details will be provided under each case study. 
3.4.3 Supply Chain Resilience Assessment 
Supply chain is a relatively new term that arose in the 1980s to describe new 
management disciplines (Christopher and Peck 2004). A supply chain is a series of 
38 
 
steps that are performed by collaboration among suppliers, manufacturers, and 
distributors until the goods finally reach customers. The main goal of a supply chain is 
to satisfy customers by ensuring the right quantity and proper quality at the right time 
and right location. Thus, it is very important that supply chains maintain resilience.  
 
There is no definite answer to how we can determine supply chain resilience. Several 
aspects fall under supply chain resilience (Barroso et al. 2015): The size of the change 
that occurs to a system and still maintains the same control on the function of the 
system; the ability of the system to operate; and finally the ability of the supply chain 
to adapt to any change. Therefore, it is an essential step to define the system that we 
are interested in and to recognize the failure mode.  
 
Risk of disruption can affect supply chain performance.  This effect can be quantified 
through certain measures: profits, customer satisfaction, sales, and production level. 
Sheffi (2006) provides a graph for the disruption profile of the supply chain 
performance. The system goes through different stages. The performance declines 
during the disruption, but as responses and actions are executed, the system?s 
performance will gradually increase. 
 
Figure 3.7 shows that there are eight stages of responding to risk: initial preparation, 
the occurrence of disruption, first response, delayed impact, full impact, preparation 
for recovery, recovery, and long-time impact. 
 
39 
 
The disruption profile facilitates the visualization of the risk on the performance of the 
supply chain. It can be applied to assess the performance of the supply chain over time.  
 
 
 
Figure 3.7 Disruption profile (Sheffi and Rice Jr 2005) 
 
In this thesis, the supply chain performance under different scenarios is used as a 
resilience measure. The supply chain performance is defined through the fulfillment 
rate. 
 
As the ultimate goal of the supply chain is to satisfy customers, we have chosen the 
number of outputs over a unit of time compared to the supply chain target over the 
same unit of time to track the propagation of any disruption through the movement of 
materials and information which will affect the number of outputs. This will also allow 
40 
 
the comparison of the supply chain under different scenarios. The fulfillment rate is 
being used as a term to refer to the supply chain performance. 
???????????? ?????? ?? ??????? (3.10) ???????  
Then, the supply chain resilience can be described through an index between 0 and 
1(Schmitt and Singh 2012). The closer the index is to 1, the more resilient is the supply 
chain. 
 
?
?? ? ?? ??
?
1 ? ? ? ?? (3.11) ?? ?? ?? ?? ?? ??  ?
?
?? 1 ? ? 1 ??/??? ?? ??  
 
where,  
??: Resilience index of the supply chain at time ? 
?: Performance of a company when it is not affected by a disruption  
??: Performance of a company at time ? 
??: is the upper limit of the time period that is used to calculate the resilience  
??: is the lower limit of the time period that is used to calculate the resilience  
3.4.4 Identification and Assessment of Mitigation Strategies  
There are several ways to respond to risk. These ways include risk acceptance, 
avoidance, transfer, and mitigation (Ayyub 2014). The resilience of the supply chain 
depends on responding to the disruption by being prepared for it.  
 
41 
 
A resilient supply chain is capable of responding to risk. This can be done through the 
design of a supply chain or some strategies that can mitigate the risk impact or even 
remove it.  
The models in this research were built to test resilience through specific design plan 
and other mitigation plans. 
? Design plans 
? Redundancy: the supply chain has more than one supplier. 
Each supplier relies on a different transportation system. 
? Mitigation plans 
? Conservation: to maintain the same level of production 
without substituting for the loss during a disruption. 
? Input reallocation: to replace the proportion of inputs to a 
critical output. (e.g., assign all the raw materials for a high-
demand product). 
? Output substitution: to change the distribution of the 
products (e.g., provide all the outputs of a specific market).   
? Share resources: to collaborate with other supply chains and 
try to get raw materials during the disruption period. 
 
Once all the methodology steps have been completed, the case study models can be 
developed. Both cases are being assumed to be located in the same area. The reliability 
of the types of infrastructure supporting the supply chains will be the same. Figure 3.8 
42 
 
provides a summary of the steps that will be followed in Chapter 3 to develop the DES 
model. 
 
Figure 3.8 DES model development steps 
 
 
43 
 
Chapter 4: Case Studies 
Assembly facilities of two different manufacturing sectors are considered and modeled 
in this chapter for the purpose of examining their respective supply chains and 
underlying physical infrastructure. DES model has been developed for each case. Both 
cases are assumed to be located in the same geographical area. The first supply chain 
is a food processing manufacturer. The second case study is a medical device assembly 
company.  
 
The purpose of the simulation is to build an appropriate model in order to analyze the 
supply chain and make future decisions to increase the supply chain resilience. 
Campuzano and Mula (2011) provide a list as a guideline for simulation models; these 
include: 
? Not constructing a complex model, but a simple one that works. 
? Trying to understand the problem to find the right technique. 
? A model cannot be better than its inputs. 
? Models cannot replace decision-makers; they are a tool to be used in order to 
understand.  
4.1 Model Characteristics and Assumptions 
The accuracy of simulation results is highly dependent on the system inputs. In this 
thesis, the input data are divided into two groups. The first group includes all the 
variables that are altered by the user to construct a specific setting. These variables 
include the number of raw materials that will change throughout the model to produce 
44 
 
the outputs, the execution time for tasks and the number of workers, and finally the 
routing of raw materials and outputs. The second group is the disruption scenarios 
inputs. They are data imposed on the model only to test the system response. It is being 
done by stopping the generation of raw materials for a period of time. Figure 4.1 shows 
the characteristics and the assumptions required to construct the DES model. 
 
Figure 4.1 The characteristics and the assumptions that are required to construct the 
DES model 
 
ExtendSim is a unit-less software which can work with undefined time. Thus, it is 
important to decide about the time unit once we start to build the model, and we should 
maintain the unit throughout the model to get the correct relationship. ExtendSim gives 
the flexibility to change the unit of time inside the activity block to be more realistic. 
Each of the case studies has a different time unit based on the manufacturing sector and 
the time between received raw materials. 
4.1.1 Raw Materials 
Most supply chains require different kinds of intermediate input materials to produce 
their final products. Critical raw materials are identified in the model. The critical 
45 
 
materials include the effect of the reliability of infrastructure. As a result, critical 
materials ultimately control the model. This will be explained in detail for each case 
study.  
 
The number of raw materials in each case study is modeled as a random variable based 
on a uniform distribution with a minimum value equal to the amount of provided raw 
materials by infrastructure n at time ? and a maximum value equal to the amount of 
required raw materials by the facility at time ? as was explained in equation 3.3 and 
equation 3.4 and as illustrated in Figure 4.2. In each case study, a table has been 
constructed for the critical raw materials that can be provided to the facility through 
each infrastructure system.  
 
Figure 4.2 Raw materials in the DES model 
 
46 
 
4.1.2 Reliability of Infrastructure  
The reliability of infrastructure that supports the supply chain is calculated based on 
the probability of failure of infrastructure. The included components of infrastructure 
in both study cases are roadways, electrical power, water and wastewater, workforce, 
and communication.  
 
The purpose of constructing a BN is to define the reliability of all the types of 
infrastructure supporting the supply chain. Since BN relies on probability theory, the 
nodes that represent the infrastructure components will have a probabilistic nature. The 
impact of the reliability of infrastructure in the discrete event simulation will be 
captured by multiplying the reliability of infrastructure by the amount of the required 
critical material by the facility. The different types of infrastructure will be modeled as 
nodes and arrowed lines, as described in the following three steps.  
1. Identifying the parent nodes and the child nodes in the network based on the 
definitions of each kind of node was explained in section 3.3.2. Figure 4.3 
illustrates the different types of infrastructure that support the supply chain in 
the two case studies. Assumptions have been made about how each 
infrastructure component is connected to others in the network. 
2. Calculating the reliability of the parent nodes: The reliability of the 
infrastructure can be calculated based on the calculation of the failure 
probability of the infrastructure. In the discrete event model of each case 
study, the parent nodes are roadways, electrical power, and communication. 
The failure probability of communication depends also on electrical power. 
47 
 
The workforce assumed to be controlled by the infrastructure. Pandemics are 
disregarded in this model (Telford and Kimberly 2020). The two facilities are 
assumed to be in an area where they experience blockage of the main highway 
roads around 15 days during the year and an electrical power outage of 7 days 
during the year due to disruption by natural hazards. Poisson distribution is 
being used to represent the occurrence of infrastructure disruption over time. 
The failure of each infrastructure disruption has a different rate as mentioned 
in Section 3.1.3.1. 
 
 
Figure 4.3 Bayesian Network of infrastructure in the DES model 
 
  
? 7 365 0.01917 0.02 
(4.1) 
where, 
48 
 
?   is the rate of failure of roadways. 
??????????? ?? ???????? 1 0.02 0.98 4.2  
 
? 14 365 0.04109 0.04 
(4.3) 
where, 
?   is the rate of failure of roadways. 
??????????? ?? ???????? 1 0.04 0.96 (4.4) 
3. Construct the conditional probability table (CPT). Once the reliabilities 
associated with the roadways and electrical power systems is calculated, the 
CPT of the reliability of a child node can be constructed. The reliability of 
each child node depends on the state of the parent node whether it is reliable 
or not (failure). Table 4.1 shows the variable names for each infrastructure. 
 
We are going to refer to the reliability of infrastructure by the probability of the letters 
which appear in Table 4.1 and to the probability of failure using the same letters but 
with a dash on the top of the letter (e.g., R if the roadways are available , and ? if the 
roadways are unavailable). Figure 4.4 shows how the state of the parent node affects 
the reliability of a child node.  
 
The reliability and the failure probability of the parent node illustrated in Table 4.2-4.5 
were calculated based on equation 4.2. 
 
49 
 
Table 4.1 Infrastructure variable names 
R Roadways 
E Electrical power 
W Water & wastewater 
C Communication 
F Workforce 
  
The accuracy of BN analysis is highly dependent on the understanding of the 
relationship among the nodes. The relationship between every two components of 
infrastructure is being interpreted through probability. The following three tables show 
the probability of the infrastructure being reliable with the prior information about the 
state of the parent infrastructure given. All the values shown in the next three tables are 
based on the assumptions for the simulation. 
 
Table 4.2 Reliability of the parent nodes of the BN of infrastructure   
Infrastructure Reliability  
Electrical power 0.96 
Roadways 0.98 
 
Table 4.3 Reliability of communication based on the state of electrical power 
State of electrical power Reliability of communication P(C) 
Available, ? ? ?|?  0.92 
 
Unavailable, ? ? ?| ?  0.85 
 
50 
 
Table 4.4 Reliability of water & wastewater based on the state of electrical power 
State of electrical power Reliability of water and wastewater P(W) 
Available, ? ? ?|?  0.95 
Unavailable, ? ? ?| ?  0.8 
 
Table 4.5 Reliability of workforce based on the state of roadways and communication 
State of roadways State of communication Reliability of workforce P(F) 
Available, ? Available, ? ? ?|? ? ?  0.95 
Available, ? Unavailable, ??? ? ?|? ? ???  0.7 
Unavailable, ? Available, ? ? ?|???? ? ?  0.85 
Unavailable, ? Unavailable, ??? ? ?|???? ? ???  0.25 
 
51 
 
 
Figure 4.4 Parent nodes and child nodes of the BN of infrastructure in the DES model 
 
52 
 
Now the CPT can be created for each child node as explained in Figure 4.5 as indicated 
in the equations below: 
? ? ? ?|? ? ? ? P C|/E ?*P E?  (4.5) 
? ? ? ?|? ? ? ? ? ?|??? ? ? ????  (4.6) 
? ? ? ?| ? ? ? ? ?| ? ? ? ? ?| ? ? ??? ? ?| ? ? ???  (4.7) 
? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?   
? ? ? ? ? ??? ? ? ? ? ??? ? ? ? ? ? ??? ? ? ? ? ???  
 
 
 
Figure 4.5 Analysis of BN of the different types of infrastructure supporting the 
supply chain in the DES model 
 
 
53 
 
4.1.3 Manufacturing Process 
The main goal of the supply chain is to turn raw materials into goods and/or services. 
This process includes multiple manufacturing steps that depend on the sector. In each 
supply chain case study, there are several tasks. Each task is  modeled as an activity 
block with a specific number of workers referred to it as the activity block capacity and 
delay time to perform the task. Delay time can be a constant or random variable. Some 
of the tasks are performed by workers and others by machines, so the delay time is not 
constant and is susceptible to some change. To capture the uncertainty of the unknown 
delay, triangular distribution has been chosen to represent the delay time. It is defined 
by three values: the minimum value, the maximum value, and the most likely value as 
indicated in Figure 4.6. 
  
Figure 4.6 Delay time in the DES model 
 
 
54 
 
4.1.4 Disruption Scenarios 
To expose the DES model to a disrupted scenario, we impose the disruption as a cut in 
the supply of critical raw materials due to a total loss of a seaport or a roadway.  As a 
result, the facility will activate some contingency plans to deal with the disruption, for 
example, reliance on inventory or changes made in the kind and distribution of goods.  
 
A comparison among the different scenarios can be done by identifying the supply 
chain performance, by measuring its ability to satisfy the facility target of production, 
and by finding the resilience index for each case to help make the right decision in order 
to strengthen the supply chain resilience to face any possible disruption and to recover 
its original performance in a short period of time. 
4.2 Case Study 1: Food Processing 
Company AB is an assembly distribution system. Figure 4.7 depicts the structure of the 
supply chain. It consists of one supplier, a manufacturing center that has multiple plants 
and multiple distribution destinations. This supply chain supports one product. The 
product is canned chicken sausage. The raw materials are delivered daily to the facility 
through one of the main highways that connect the suppliers with the manufacturing 
plants.  
4.2.1 Model Construction 
The simulation time-unit was set to hours indicating that a run over 648 hours is equal 
to 27 days. The output unit is the number of five O.Z cans produced. The facility runs 
two shifts. Each shift is 8 hours long. 
55 
 
4.2.1.1 Raw Materials  
The generation of materials represents the daily amount of materials that can be 
provided by a supplier towards production of the final good. Raw materials are divided 
into two groups: the critical materials that will be controlled by the infrastructure 
reliability and the non-critical ones. For this case study, the frozen chicken is the critical 
material. The required raw materials are chicken, metal cans, spices, and other 
ingredients as illustrated in Table 4.6. 
 
Table 4.6 Raw materials of the chicken manufacturing facility 
Raw materials Variables Amount 
Chicken ? 60000 lbs./day 
Metal cans ?1 10560 lbs./day 
Spices and others ?1 180000 cans/day 
  
The amount of critical material is the minimum amount of the generated items served 
by each infrastructure based on uniform distribution. Each infrastructure will generate 
several items based on equation 3.3. Table 4.7 shows the number of materials provided 
by each type of infrastructure. 
 
56 
 
  
 
Figure 4.7 Structure of facility AB in DES model 
 
Table 4.7 The number of materials provided by each infrastructure ??  
Infrastructure ??  ?	 ? ?? (lbs./day) ? (lbs./day) 
Roadway  ?? 0.98 60000 58800 
Electrical power ?? 0.96 60000 57600 
Water & wastewater ?? 0.944 60000 56640 
Communication ?? 0.9172 60000 54762 
Workforce ?? 0.9361 60000 56166 
 
4. 2.1.2 Manufacturing Process 
The chicken experiences multiple manufacturing steps until it becomes ready to be 
distributed to the market. The service time for each task is presented as triangular 
distribution. The tasks that are performed by workers are more susceptible to 
57 
 
uncertainty than the automated ones. Then, when the materials arrive at the facility, 
they will be unloaded and inspected, as shown in Table 4.8. The table also shows the 
delay time of each task that is defined in each block of the activity block in the DES 
model, and the capacity of each block refers to the number of entities that the block can 
serve. Batch block is used to assemble the different materials to form one can of chicken 
sausage. Each one can of sausage consists of chicken, metal cans, spices, and other 
ingredients as indicated in equation 4.8. The number of produced cans relies on time. 
Each 1 lb. of chicken is enough to produce 3 cans.  
? 1 13 ?? 25 ?? ?? 
(4.8) 
where, 
?: 1 can of chicken sausage 
??:1 lb. of chicken 
??:1 lb. of spices and other ingredients 
??: 1 metal can 
The resulting function of produced cans at any time in days is ? ? .  
F ? 3tZ. (4.9) 
where, 
?: The time in days 
Z: The minimum amount of received chicken based on the reliability of infrastructure.  
 
58 
 
Table 4.8 The delay time and the capacity of Activity block in DES model for case 
study 1 
Task Capacity Min Max Most Likely 
(hrs.) (hrs.) (hrs.) 
Transportation of chicken to the 1 shipment 1 1.5 1 
facility 
Arrival of chicken to the facility/ 1 shipment 2 2.5 2 
unloading 
Inspection of the chicken 1 shipment 1 1.5 1 
Mixing 6000 cans 0. 67 0.7 0.67 
Cooking  6000 cans 1.6 1.6 1.65 
Cooling 6000 cans 0.67 0.7 0.67 
Sterilizing 6000 cans 0.83 0.85 0.83 
Inspection 100 cans 1.62 1.8 1.8 
Packaging and labeling 100 cans 3.6 3.9 3.6 
 
 
4.2.1.3 Distribution 
To model the supply chain performance behavior under disruption, we will be focusing 
on the number of outputs that are allocated to each market. Transportation of the final 
product is disregarded in this project. The number of outputs helps to assess the supply 
chain?s ability to satisfy customers in the local and global markets. 
 
Outputs are listed by the number of cans and by categorizing the market for the final 
good into three groups: the local market, the global market, and inventory in the 
warehouses. The internal shares of the local market have been shown in detail because 
59 
 
that will affect some of the strategies that the facility will be applying during a 
disruption. 
Table 4.9 Total market distribution of the canned chicken sausage 
Market Percentage 
Local market 60% 
Inventory 10% 
 
Table 4.10 Local market distribution of the canned chicken sausage 
Destination Percentage 
Grocery  50% 
Retailers 40% 
Restaurants 10% 
 
4.2.2 Scenarios and Response Plans 
For this manufacturing supply chain, the simulation is being done for 648 hours 
(equivalent to 27 days), and the supply chain performance is being tracked every day 
because the materials arrive daily to the facility.  
 
The manufacturing plant has only one local supplier for the chicken. The chicken is 
sent from the supplier to the facility through a highway. Three scenarios are being 
tested: the steady state, disrupted state, and finally the disrupted state while undergoing 
a response plan. 
? Steady state: the amount of frozen chicken is influenced by the reliability 
of infrastructure under normal conditions. 
60 
 
54762 ? 60,000 (critical material) (4.10) 
 
Figure 4.8 Facility AB under normal state 
 
? Disrupted state: affected by a transportation disruption of seven days, 
which causes an interruption in the flow of critical materials from the 
supplier to the facility for seven days. 
? 0 (critical material) (4.11) 
 
 
Figure 4.9 Facility AB under disrupted state 
 
? Response plans for the disrupted state (shared resources). 
The facility can adapt to the disruption by keeping the same target of 
cans and getting the critical materials from another supplier by agreeing 
to useless amount of raw materials. The supply chain will be willing to 
share the resources of a new supplier with other supply chains ahead of 
time by paying a premium for the new supplier  (van Donk and van der 
Vaart 2005). This agreement provides the facility with certain amount of 
materials during the loss of its main supplier. In this case study, the 
61 
 
assumption is that the facility runs during the disrupted days by relying 
on a local resource that can provide the facility 50-70% of their regular 
requirement of the critical materials for eight days.  
30000 ? 42000  (critical material) (4.12) 
 
Figure 4.10 Facility AB under a disrupted state with the mitigation plan 
 
4.2.3 Results and Analysis  
4.2.3.1 Steady State   
The objective of this work is to quantify the resilience of the supply chain to risk due 
to the reliability of infrastructure and a total failure of the infrastructure responsible for 
logistics (seaport/roadway). The resilience index is calculated for the steady state and 
disrupted state before taking any response action and after taking a response. It is 
calculated for a time window between 1 and 27 days.  
 
The condition of the steady state was tested by running the model for 27 days to achieve 
the daily target of facility AB of the canned chicken sausage. We calculate the 
performance of the facility by dividing the number of outputs over the daily target of 
the facility. The fulfillment rate of the steady state shows the impact of the reliability 
62 
 
of infrastructure. During the steady state, the facility is provided with raw materials 
from the supplier.  
 
Figure 4.11 shows the amount of the critical materials provided by the supplier based 
on the reliability of infrastructure under a steady state. Here we can notice that the 
amount of critical materials is not a constant value. This result aligns well with the 
assumption that materials will follow a uniform distribution with specific minimum 
and maximum values. The figures also show the influence of the loss of suppliers due 
to the disruption of the roadways when the facility will not receive any materials during 
the seven days. And finally, it shows the impact of adapting the facility to share 
resource plans where the amount of the chicken has increased from zero to almost 50% 
of the original one. 
4.2.3.2 Disrupted State 
Once we were certain that the model was operating well under the normal state, we ran 
the model for 27 days  and imposed a disruption scenario to the facility by losing all 
the critical materials provided by the supplier due to a natural hazard that caused seven 
days of a block to the main highway that is used to transport the critical materials to 
the seaport. This disruption prevented the facility from receiving the materials from 
supplier 1 as shown in Figure 4.11. They provided zero materials. As a result, the 
fulfillment rate dropped to zero. 
 
63 
 
The?Amount?of?the?critical?raw?materials?provided?to?the?
70000 facility?
60000
50000
40000
30000
20000
10000
0
0 5 10 15 20 25 30
Time (day)
Steady?state Disrupted?state Disrupted?state?with?(shared?resources)  
Figure 4.11 The amount of chicken provided to the facility 
 
4.2.3.3 Disruption and Mitigation Response   
The production of the canned chicken sausage is being tracked over time to capture the 
impact of the disruption and test the different mitigation responses. The steady state is 
being used as a reference to validate the other scenarios. For the disrupted state, we 
included the calculation when no backup plan is applied. We evaluate the impact of 
shared resources with other supply chains. Sharing resources is a mitigation plan used 
in this model. We assume this facility is of medium size and has medium capabilities. 
Therefore, it will rely on a local supplier to gain 50-70% of its original needs. In this 
case, the facility will receive a different amount of materials based on the amount that 
the temporary resource can provide. 
 
The performance of the facility under the different scenarios and response plans for 
each of the specific destinations is illustrated in the following figures based on equation 
64 
 
Frozen chicken (lbs.)
3.9, by dividing the number of outputs over the daily target of the company. The 
performance of the facility varies under the steady state due to the reliability of the 
different types of infrastructure supporting the supply chain. Table 4.11 shows a sample 
of the fulfillment rate of the facility for the output distribution under the different 
scenarios.  
 
Chicken?suasage?distribution?to?grocery?stores
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Time in days
Steady?state Disrupted?state Disrupted?state?with?shared?resources
 
Figure 4.12 Fulfillment rate of chicken sausage distribution to grocery stores 
65 
 
Fulfillment Rate %
Chicken?sausage?distribution?to?retailers
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Time in days
Steady?state Disrupted?state Disrupted?state?with?shared?resources
 
Figure 4.13 Fulfillment rate of the chicken sausage distribution to retailers 
 
Chicken?sausage?distribution?to?restaurants
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Time in days
Steady?state Disrupted?state Disrupted?state?with?sharing?resources
 
Figure 4.14 Fulfillment rate of the chicken sausage distribution to restaurants 
 
66 
 
Fulfillment Rate % Fulfillment Rate %
Chicken?sausage?distribution?to?the?global?market
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Time in days
Steady?state Disrupted?state Disrupted?state?with?shared?resources
 
Figure 4.15 Fulfillment rate of the chicken sausage distribution to the global market 
 
Chicken?suasage?distribution?to?the?storage?facility
140.00
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Time in days
Steady?state Disrupted?state Disrupted?state?with?shared?resources
 
Figure 4.16 Fulfillment rate of the chicken sausage distribution to the storage facility 
 
 
67 
 
Fulfillment Rate % Fulfillment Rate %
4.2.3.4 Resilience Index   
The resilience index of a supply chain AB is calculated based on the concept of the 
resilience profile presented in Chapter 3. The resilience index depends on the 
performance of the supply chain under different scenarios. In the steady-state 
assumption, the supplier can deliver the materials between day 1 and day 27. The 
amount of materials received by facility AB varies from one day to another due to the 
reliability of the infrastructure; therefore, the fulfillment rate is still high between 95-
100% as illustrated in Figures 4.12 to 4.16. The resilience index is calculated for the 
time window between day 1 and 27 based on equation 3.10 as shown in Table 4.12. 
When the disruption occurs and the fulfillment rate reaches zero, this indicates that the 
facility is not capable of withstanding the negative impact of the loss of the supplier. 
Consequently, the resilience index becomes 0.65.  
 
Table 4.11 Fulfillment rate of chicken sausage distribution to grocery stores 
Fulfillment Rate % 
Day Steady state Disrupted state Disrupted state (shared resources) 
1 99 98 98 
2 99 98 98 
3 94 95 95 
4 99 0 50 
5 94 0 50 
6 94 0 50 
7 100 0 50 
68 
 
8 94 0 50 
9 98 0 50 
10 94 0 49 
11 96 0 50 
12 97 101 101 
13 94 96 96 
14 94 96 96 
15 101 96 96 
16 94 93 93 
17 94 97 97 
18 99 94 94 
19 94 96 96 
20 99 96 96 
21 94 93 93 
22 94 96 96 
23 99 96 96 
24 94 92 92 
25 99 96 96 
26 94 96 96 
27 99 99 99 
 
When the facility adopts a mitigation plan to respond, the resilience index increases, 
too. This strategy will increase the resilience of the supply chain and it will allow the 
69 
 
facility to keep its production although not at the same level as the original one. 
Optimally, the facility will be able to satisfy its customers.  
 
Table 4.1 Resilience index of facility AB 
Scenario Resilience index 
  
Steady state 0.96 
A disrupted state without a mitigation plan 0.65 
 
A disrupted state with shared resources 0.82 
 
 
4.3 Case Study 2: Medical Device Assembly 
Company AC is another assembly distribution system for medical devices located in 
the same geographical area. The company?s manufacturing system has two lines of 
production. The first is IV bags made from resin film and the associated plastic for 
tubing. The second production processes two types of medical kits. Figure 4.17 
explains the structure of the supply chain. It consists of one supplier, a manufacturing 
center that has multiple plants, and multiple distribution destinations. The facility also 
has a warehouse to keep an inventory of the raw materials. Material is shipped to the 
facility every six days through a seaport. 
4.3.1 Model Construction 
The simulation time-unit was set to hours indicating that a run over 2016 hours, which 
is equal to 14 weeks. The output unit is the number of intravenous (IV) bags and 
medical kits. The facility runs one 12-hour shift. 
70 
 
4.3.1.1 Raw Materials  
The generation of materials represents the number of raw materials that can be provided 
by a supplier every six days. Raw materials are divided as in the previous case study 
into two groups: the critical materials that will be controlled by infrastructure reliability 
and the non-critical ones. The resin film is the critical material in this case study; the 
other required raw material is plastic as illustrated in Table 4.13. 
 
Table 4.13 Raw materials of the IV bags and medical kits 
Raw materials Variables Amount 
Resin film ? 40000 lbs./6days 
Plastic s 10000 lbs./6days 
 
The amount of the critical material (i.e., resin film) is the minimum amount of the 
generated items served by each infrastructure. This has been modeled by using a 
uniform distribution. Each infrastructure will generate several items based on equation 
3.3.  Table 4.14 shows the number of materials provided by each type of infrastructure. 
 
71 
 
 
Figure 4.17 Structure of facility AC in DES model 
 
Table 4.14 The number of materials provided by each infrastructure ??  
Infrastructure ??  ? ? ?? (lbs./day) ? (lbs./day) 
Roadway  ?? 0.98 40000 39200 
Electrical power ?? 0.96 40000 38400 
Water & wastewater ?? 0.944 40000 37760 
Communication ?? 0.9172 40000 36688 
Workforce ?? 0.9361 40000 37444 
 
4. 3.1.2 Manufacturing Process 
The resin film and the plastic need to be processed until they become IV bags and 
medical kits. The service time for each associated process task is modeled as triangular 
distribution. As mentioned in section 4.2, the tasks that are performed by workers are 
72 
 
more susceptible to uncertainty than the automated ones. Table 4.15 shows all the 
manufacturing tasks that are required to produce the IV bags and the medical kits. The 
shipment that arrives at facility AC will be unloaded and inspected. 
Table 4.15 shows the delay time of each task that is defined in each block of the activity 
block in the DES model and the capacity. Batch block is used to assemble the different 
materials to form one IV bag. Each IV bag consists of resin film and plastic as indicated 
in equation 4.13. The number of produced IV bags relies on time.  
? 1? 2? (4.13) 
?1 1? (4.14) 
?2 2? (4.15) 
where, 
v: 1 IV bag 
r: 1 lb. of resin film 
b: 1 tubing= ? 
?: 1lb.of plastic= 8 tubings 
?1: kits made of resin film 
?2: kits made of plastic 
The resulting function of produced IV bags and medical kits at any time in days is ? ? .  
? ? 0.7?? (4.16) 
where, 
?: The time in weeks 
z?: The minimum amount of received resin film based on the reliability of infrastructure 
at time ? 
73 
 
Table 4.15 The total delay time and the capacity of Activity block in DES model for 
case study 2 
Task Capacity Min Max Most Likely 
(hrs.) (hrs.) (hrs.) 
Transportation of the resin film 1 shipment 1 1.5 1 
Arrival of raw materials to the 1 shipment 1 2.5 2 
Inspection of raw materials 1 shipment 1 1.5 1 
Assembly and filling with fluid 10 bags 14 14.5 14 
Sterilizing 500 bags 7 7.5 7 
Parts inspection 10 bags 23.8 28 28 
Leakage inspection 6 bags 23.3 23.5 23.3 
Packaging & labeling 10 bags 28 35 42 
 
4.2.1.3 Distribution 
The goal of the facility AC is to meet the target number of IV bags and medical kits 
allocated to each market. There are four relevant markets for this case study:  the U.S. 
market, the local market, other countries or whether to other facilities of the parent 
company for further processing. Transportation of the final product is disregarded in 
this project. The number of outputs is essential to the assessment of the supply chain 
performance. 
 
Outputs are listed by the number of IV bags and kits made of resin film and plastic and 
by categorizing the market of the IV bags into three groups: the local market, USA 
market, and other countries. In contrast, the medical kit bags are sent to two other 
facilities for further processing.  
 
74 
 
Table 4.16 Total market distribution of the IV bags 
Market Percentage 
USA market 50% 
Local market 45% 
Other countries 5% 
 
Table 4.17 Total market distribution of medical kits 
Destination Percentage 
Facility 1 60% 
Facility 2 40% 
 
4.3.2 Scenarios and Response Plans 
For this manufacturing supply chain, the simulation runs for 2020 hours, and the supply 
chain performance is tracked every 6 days because of the same amount of time that the 
facility receives the raw materials. We will discuss two design plans. The first one is 
the original assumption that the manufacturer has only one resin film supplier that sends 
the materials through a seaport. The second plan is to assume that the company relies 
on two suppliers. Three scenarios are being tested for the two plans: the steady state, 
the disrupted state, and finally the disrupted state while undergoing a response plan. 
4.3.2.1 Design Plan One 
? Steady state: the amount of resin film is influenced by the reliability of 
infrastructure under normal conditions. The facility has one supplier that 
sends the materials through the seaport. 
75 
 
36688 ?? 40,000 (critical material) (4.17) 
 
Figure 4.18 Facility AC with design plan one under normal state 
 
? Disrupted state: affected by a transportation disruption for two weeks, 
which causes an interruption in the flow of the resin film from the supplier 
to the facility.  
? 0 (critical material) (4.18) 
 
 
 
Figure 4.19 Facility AC with design plan one under disrupted state 
 
? Response plans for the disrupted state (Using inventory) 
The facility can mitigate the disruption in getting critical materials by 
keeping the same weekly target of outputs and relying on the inventory of 
76 
 
raw materials. In this case, the supply chain will be able to produce IV bags 
and medical kits. The stored amount of materials depends on how much and 
for how long the facility has been storing the materials when the disruption 
occurs. In this case study, an assumption has been made that when the run 
starts, the inventory in the warehouse is zero. The disruption occurs during 
the third week. The amount of raw materials is a function of time.  
?? ? 0.3? ? critical material) (4.19) 
where,  
?: number of the last week before the occurrence of the disruption 
 
 
 
Figure 4.20 Facility AC with design plan one under a disrupted state with a mitigation 
plan of using the inventory 
 
Another mitigation plan the facility can follow to mitigate the impact of the 
disruption is to do a reallocation of the raw materials obtained from the 
inventory during the disruption. The facility will produce just the IV bags from 
the resin film. More details will be discussed in the results.  
 
77 
 
 
Figure4.21 Facility AC with design plan one under a disrupted state with a mitigation 
plan of reallocating the resin film 
 
4.3.2.2 Design Plan Two 
? Steady state: the amount of resin film is influenced by the reliability of 
infrastructure under normal conditions. The facility has two suppliers that 
send the materials through a seaport and roadway 
?? ?? ?? (critical material) (4.20) 
where, 
?1: The amount of resin film provided by supplier 1 (70% of the required materials by 
the facility) 
?2: The amount of resin film provided by supplier 2 (30% of the required materials by 
the facility) 
 
Figure 4.22 Facility AC with design plan one under normal state 
78 
 
? Disrupted state: affected by a transportation disruption for two weeks, 
which causes an interruption in the flow of the resin film from supplier 1 to 
the facility.  
?1 0 (4.21) 
?? ?? (4.22) 
 
 
Figure 4.23 Facility AC with design plan two under a disrupted state with a 
mitigation plan of using the inventory 
 
? Response plans for the disrupted state (using inventory) 
The facility is still receiving 30% of the amount of raw materials by 
relying on the inventory of raw materials. The amount of raw materials 
depends on the amount provided by supplier 2 and the stored amount of 
materials in this case study. An assumption has been made that when the 
run starts, the inventory in the warehouse is zero. The disruption occurs 
during the third week. The amount of raw materials is a function of time 
and the amount provided by supplier 2. 
?? ?? ? 0.3? ? critical material) (4.23) 
79 
 
where,  
?: number of the last week before the occurrence of the disruption  
 
Figure 4.24 Facility AC with design plan two under a disrupted state with a 
mitigation plan of using the inventory 
 
The second mitigation is the reallocation for the raw materials obtained from 
the inventory. The facility will produce just the IV bags from the resin film. The 
difference in this design plan from the previous one is that the facility is still 
receiving materials from supplier 2 and also all the materials will be allocated 
for the production of the IV bags. 
 
 
Figure4.25 Facility AC with design plan two under a disrupted state with a mitigation 
plan of reallocating the resin film 
80 
 
 
4.3.3 Results and Analysis  
4.3.3.1 Steady State   
To create the steady state condition for each design plans we ran the model for 14 weeks 
in order to match the weekly target of facility AC of the IV bags and the two kinds of 
kits. We tracked the performance of the facility by dividing the number of outputs over 
the weekly target of the facility. The fulfillment rate of the steady state of the two-
design plan clearly shows the impact of the reliability of infrastructure. During the 
steady state, the facility is provided with resin film from the supplier. Figure 4.26 shows 
the amount of the provided resin film under normal state and disrupted state conditions. 
The amount of critical materials is not a constant value. This result meets the 
assumption that materials will follow a uniform distribution with specific minimum 
and maximum values. 
4.3.3.2 Disrupted State 
For this part of the research, we ran the model for 14 weeks and hit the facility with 
disruption due to a natural hazard that caused a two-week shutdown to the seaport. This 
disruption prevented the facility form receiving the resin film from supplier 1 in the 
two-design plan as shown in Figure 4.26 and Figure 4.27. When the facility has only 
one supplier, the materials delivered during the disruption is zero. As a result, the 
fulfillment rate reaches zero under design plan one whereas in the plan with two 
suppliers, the facility can keep a low level of production. 
81 
 
The?Amount?of?the?critical?raw?material?provided?to?facility?AC
45000
40000
35000
30000
25000
20000
15000
10000
5000
0
0 2 4 6 8 10 12 14 16
Week
Steady?state Disrupted?state
 
 Figure4.26 The amount of resin provided to the facility for design plan one  
 
The?Amount?of?the?critical?raw?material?provided?to?facility?AC
45000
40000
35000
30000
25000
20000
15000
10000
5000
0
0 2 4 6 8 10 12 14 16
Week
Steady?state Disrupted?state  
Figure 4.27 The amount of resin provided to the facility for design plan two 
82 
 
Resin film in lb. Resin film in lb.
4.3.3.3 Disruption and Mitigation Response   
We track the disruptions of specific products to observe their impact and test the 
different mitigation responses and the different design plans for the two-design plans. 
As a reference, we included the calculation when no backup plan is applied. First, we 
evaluated the impact of using the inventory to produce IV bags and medical kits. The 
facility uses 50% of its inventory the first week, assuming that it doesn?t know how 
long the disruption will stay and another 50% the second week. The difference between 
the two-design plans is that when the facility has two suppliers, the facility still has a 
partial amount of the materials (30%). Thus, the production will never reach zero. The 
second mitigation plan is the decision to relocate the raw materials by producing the 
product with high demand or priority. In this case, the facility keeps the production of 
the IV bags and halts the production of the medical kits made of the resin film. In this 
way, all the resin film in the inventory and the received materials in design plan two 
will be allocated to produce IV bags. The testing results of the two-design plans are 
graphically presented for the 14 weeks. Table 4.18 shows a sample of the fulfillment 
rate of the IV bags distributed to the USA. 
 
83 
 
Distribution?of?IV?bags?to?USA
140.00
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0 2 4 6 8 10 12 14 16
Steady?state Time in days
Disrupted?state
Disrupted?state?with?using?the?inventory
Disrupted?state?using?the?inventory?and?reallocation?of?the?raw?materials  
Figure 4.28 Fulfillment rate of IV bags distribution to the USA for design plan one 
 
Distribution?of?IV?bags??to?the?local?market
140.00
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0 2 4 6 8 10 12 14 16
Steady?state
Disrupted?state Time in days
Disrupted?state?with?using?the?inventory
Disrupted?state?using?the?inventory?and?reallocation?of?the?raw?materials  
Figure 4.29 Fulfillment rate of IV bag distribution to the local market for design plan 
one 
 
84 
 
Fulfillment Rate % Fulfillment Rate %
Distribution?of?IV?bags??to?other?countries
140.00
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0 2 4 6 8 10 12 14 16
Steady?state Time in days
Disrupted?state
Disrupted?state?with?using?the?inventory
Disrupted?state?using?the?inventory?and?reallocation?of?the?raw?materials
 
Figure 4.30 Fulfillment rate of IV bag distribution to other countries for design plan 
one 
 
Distribution?of?IV?bags?to?USA
140.00
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0 2 4 6 8 10 12 14 16
Steady?state
Disrupted?state Time in days
Disrupted?state?with?using?the?inventory
Disrupted?state?using?the?inventory?and?reallocation?of?the?raw?materials
 
Figure 4.31 Fulfillment rate of IV bag distribution to the USA for design plan two 
 
85 
 
Fulfillment Rate % Fulfillment Rate %
IV?bags?Distribution?to?Local?Market
140.00
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0 2 4 6 8 10 12 14 16
Steady?state
Disrupted?state Time in days
Disrupted?state?with?using?the?inventory
Disrupted?state?using?the?inventory?and?reallocation?of?the?raw?materials  
Figure 4.32 Fulfillment rate of IV bag distribution to the local market for design plan 
two 
 
Distribution?of??IV?bags??to?other?countries
140.00
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0 2 4 6 8 10 12 14 16
Steady?state Time in days
Disrupted?state
Disrupted?state?with?using?the?inventory
Disrupted?state?using?the?inventory?and?reallocation?of?the?raw?materials  
Figure 4.33 Fulfillment rate of IV bag distribution to other countries for design plan 
two 
 
  
86 
 
Fulfillment Rate % Fulfillment Rate %
 
4.18 Fulfillment rate of IV bags distribution to the USA for design plan one 
Fulfillment Rate % 
Day Steady state Disrupted state Disrupted state Disrupted state using 
using inventory inventory and 
reallocation of the raw 
materials 
1 95 101 101 101 
2 104 94 94 94 
3 108 105 105 105 
4 98 0 69 125 
5 98 0 69 125 
6 101 52 52 52 
7 95 100 100 100 
8 103 102 102 102 
9 84 105 105 105 
10 106 96 96 96 
11 97 96 96 96 
12 86 99 99 99 
13 106 105 105 105 
14 85 92 92 92 
 
4.3.3.4 Resilience Index   
The resilience index of a supply chain company is found based on the concept of the 
87 
 
resilience profile presented in Chapter 3. The performance of the supply chain exceeds 
its maximum value by applying some of the mitigation plans. Since the resilience index 
is a value between 0 and 1, the maximum fulfillment rate that can be considered in the 
equation will be equal to or less than 100%. In the steady-state assumption, the supplier 
can deliver the materials between week 1 and week 14. 
 
The resilience index for the two-design plans under the normal state is similar; the 
impact of the redundancy in having two suppliers appears under the disruption. The 
resilience index for the facility under the disruption is 0.79. It reached to 0.85 when the 
facility depended on two suppliers. The table shows the different resilient indices for 
the two-design plan under tracking the different mitigation plans. The resilience index 
for both designs when using the inventory and reallocation are the same due to the 
assumption that made earlier that resilience index is equal to or less than 100%. But the 
fulfillment rate for design two when the facility has two suppliers at different locations 
is higher than the first design plan where the facility has only one supplier.  
 
Table 4.19 Resilience index of facility AB 
Resilience index 
Scenarios Design plan one Design plan two 
Steady state 0.95 0.95 
A disrupted state without a mitigation plan 0.79 0.85 
A disrupted state with using the inventory 0.90 0.93 
A disrupted state with using the inventory and 0.95 0.95 
reallocation of the raw materials 
88 
 
4.4 Discussion and Assessment of the Proposed Methodology Results  
The steady state in each model represents a baseline to judge the performance of each 
supply chain under disruption by applying the equations presented under each case 
study. The graphical presentation, the calculation of the fulfillment rate, and the 
resilience index under disruption compared to the steady state works as an effective 
indicator of the resilience of the supply chain. The discrete event model succeeds in 
presenting a clear picture of how a supply chain operates during a normal state.  
4.4.1 Risk and Resilience in Steady State 
One of the primary steps of the simulation process for the two case studies is to develop 
the risk in the supply chain for the steady state due to the variation in the reliability of 
the different types of infrastructure. For the first case study, Figure 4.11 shows how the 
arrival of the frozen chicken for 27 days is not constant. Everyday a different type of 
infrastructure controls the amount of the critical material provided to the facility. Thus, 
the number of the canned chicken sausage differs from one day to another. This is also 
true for the medical device assembly company. The amount of the critical material 
provided to the facility varies based on the reliability of the different types of 
infrastructure. The amount of resin film for a facility relying on one supplier is shown 
in Figure 4.26 and for a facility relying on two suppliers is shown in Figure 4.27.   
 
Based on the fulfillment rate of the food processing facility over a time window of 1 
day to 27 days, a resilience index of 0.95 is obtained.  For the medical device assembly, 
considering the fulfillment rate over a time window of 1 week to 14 weeks, a resilience 
index of 0.96 is obtained for both design plans. For both, supply chain resilience 
89 
 
depends on the reliability of the different types of infrastructure. Under the normal 
conditions of the steady state, the two facilities can increase their resilience by 
considering more reliable types of infrastructure. The risk that threatens the supply 
chain performance under normal conditions points out the importance of relying on 
more reliable types of infrastructure. 
4.4.2 Risk and Resilience in Disrupted State 
The second kind of risk that we dealt with in the simulation is due to the failure of 
infrastructure responsible for the transportation of the raw materials to the main facility. 
The simulation of the supply chain showed that the risk due to the failure of 
infrastructure depends on multiple factors: the number of suppliers; the reliability of 
the infrastructure critical to the logistics; the state of the system once the disruption 
occurs in terms of inventory; and other available mitigation plans. Time plays a 
significant role. The time of the occurrence of the disruption and its duration affected 
the supply chain?s ability to respond to the negative effect of the disruption. The amount 
of stored materials in the inventory depended on the time of occurrence.  
 
 
The food processing facility case shows how a small supply chain can be susceptible 
to a huge risk. The supply chain receives raw materials daily. The probability of the 
failure of the supply chain due to the failure of the infrastructure supporting the 
transportation of the material is huge. Once the failure of the main highway occurs, the 
facility is incapable of producing enough canned chicken sausage to meet their daily 
target. The fulfillment rate reached zero for 7 days. The resilience index dropped from 
90 
 
0.95 to 0.65 during the disruption, but if the facility applies a mitigation plan, such as 
sharing the resources of a local supplier who can provide the facility with 50-70% of 
the original required amount by the facility, the resilience index can reach to 0.82 over 
the same time window. The mitigation plan that was implemented was effective; it 
increased the resilience index to 0.82.  However, it is still risky since the facility will 
not receive a definite amount of materials.  
 
In the second case study, we notice the difference in the supply chain performance 
under disruption of the seaport for two different design plans.  In the case when the 
supply chain loses its only supplier due to the failure of the seaport, the resilience index 
decreases from 0.95 to 0.79. When the facility has two suppliers, one provides the 
facility with 70% of the required amount of the critical material through a seaport and 
the other one provides 30% through roadways. The resilience index exceeds 0.79. It 
reaches to 0.85 as indicated in the table. This shows how redundancy is effective in 
facing the negative effect of disruption. It didn?t totally avoid it, but it decreased its 
influence.  
 
The medical device assembly case study features a larger facility that is better equipped 
to face disruption. Two mitigation plans have been used. The first one is to use the raw 
materials stored in the inventory; that increased the resilience index to 0.9 in the case 
of one supplier and 0.93 in the facility with two suppliers. The other mitigation plan 
the facility can apply is to use the inventory as well as to reallocate some of the critical 
material to produce a specific product. In the medical device case, the decision was 
91 
 
made to allocate all of the resin film to produce the IV bags, thereby resulting in the 
production of more IV bags and halting the production of kits made of resin film. In 
this case, the facility will be able to meet the required fulfillment rate and in the case 
of two suppliers, it exceeds 100% as shown in all the previous figures that show the 
fulfillment rate. As a result, the facility was able to return to its original performance. 
The fulfillment rate for the supply chain with two suppliers is higher, but for resilience 
calculation purposes, we assumed the maximum fulfillment rate is 100%. Therefore, 
we got the same resilience index for the two design plans. The fact that the facility is 
more prepared to survive the disruption by allocating inventory for the critical raw 
materials shows that this is a very successful strategy. Simulation was adequate tool to 
provide us with the output number and allow us to track that number under different 
scenarios.  
 
Simulation captured the interaction between the different parts of the supply chain 
including the reliability of infrastructure and presented the influence of the reliability 
in a very effective way through the different amounts of materials served by each 
infrastructure.  
 
Discrete event simulation is an effective manner by which to present the amplified risk 
movement starting from the supplier to the end of the supply chain by providing the 
number of outputs as the defined unit of time. 
 
In summary, we can point out these important points about the two studies: 
92 
 
The implementation of different scenarios results in different resilience indices for each 
supply chain. 
 
The Supply chain resilience index is high for the supply chain that is prepared for the 
disruption and where the supply chain is built to mitigate any possible risk before it 
occurs. 
 
Acquiring an uncertain amount of material as a mitigation plan for the food processing 
case still carries some risk. However, it helped to increase the resilience index of the 
supply chain. 
 
In the second case study, the design plan had redundancy by relying on two suppliers 
where each supplier uses a different type of infrastructure to transfer the materials. This 
allowed the facility to exceed the weekly target once the mitigation plan of using the 
inventory and reallocating the materials was applied.  
 
93 
 
Chapter 5: Conclusion  
The primary contributions of this work are four-fold and described below.  
  
1. Developing an appropriate model to represent the supply chain performance under 
various scenarios, and to show the flow of the risk among the supply chain 
components.  
The Discrete event simulation that has been done in this thesis provides us with a 
realistic model to understand supply chain performance under uncertainty. It shows 
how risk starts at one component of the supply chain and transfers to other components. 
Discrete event simulation allows us to examine the supply chain performance under 
different scenarios and allows us to notice how the risk profile changes based on the 
design of the supply chain and the reaction plans that are taken. The fulfillment rate 
that has been used as a performance indicator provides information about the 
amplification of the disruption as it moves from the start point of the disruption until it 
reaches to the customers. Although the work can?t provide an explicit solution, it can 
help to understand the behavior of the supply chain and the impact of uncertain risks 
that can occur at any time due to the failure of infrastructure. 
 
2. Improving the understanding of the role of reliability of infrastructure on the 
supply chain performance under steady state and disrupted ones. 
The reliability of infrastructure is a major factor in the resilience of the supply chain. It 
plays an essential role in the ability of a supply chain to fulfill its goal. The failure of 
infrastructure cannot be avoided, but we can prepare the supply chain to face risk and 
94 
 
decrease its vulnerability to failure. The resilience of a supply chain is achieved by 
decreasing the probability of the occurrence of the disruption in the facility or by 
reducing its effects on supply chain performance. Thus, the company must prepare for 
specific risks. To increase supply chain resilience due to the failure of infrastructure 
critical for transportation, we should understand the high dependency among the other 
types of infrastructure. The Bayesian network was an effective method in representing 
that dependency. However, the results of Bayesian analysis are biased due to our 
assumption about the infrastructure responsible for the failure of other types. However, 
this is still acceptable in order to accommodate our interest in analyzing a specific role 
of infrastructure. 
 
3. Providing a quantitative assessment of the resilience of supply chain.  
A resilience index has been used throughout the simulation. The role of the resilience 
index is substantial in measuring the ability of the supply chain to handle any possible 
risk. The company?s resilience index was identified in this research by measuring the 
supply chain's ability to meet its goal during a specific period. The fulfillment rate as 
an indicator over time is a measure of the supply chain?s ability to meet its goal during 
a period of disruption and recovery. The resilience index helps the company to test 
multiple scenarios and make decisions about the right strategies that will increase the 
resilience of the supply chain. 
 
4. Enhancing adaption of the supply chain to overcome the possible risk of 
disruption due to the failure of infrastructure. 
95 
 
Mitigation plans are important although their implementation is not always possible. 
The activation of the plans depends on the time of the occurrence, the duration, and the 
state of the system. Therefore, it is necessary to identify the risk that we are preparing 
for. In this research, we have been dealing with the reliability of different types of 
infrastructure and the total loss of the infrastructure critical to transportation. 
Simulation allows conducting a study on a possible scenario. This will enable any 
supply chain to identify any risks that might affect the supply chain and the probability 
of these risks to affect it. Once the source of risk is known, a specific mitigation plan 
can be applied. 
 
And finally, the thesis is aligned with the general belief that a supply chain is as strong 
as its weakest component. Therefore, it is important to focus on increasing the 
resilience of that supply chain by making the components and the link to other 
components more robust by increasing redundancy in the supply chain and developing 
mitigation plans that can be applied once the disruption occurs. The supply chain 
should rely on more reliable infrastructure to decrease the influence of the disruption 
to maintain a continuous performance.  
 
Future work can be done to analyze the supply chain under a total failure of other 
components of the supporting infrastructure and their impact on the supply chain.  
 
The input data are logical, but for future work, it will be useful to enhance the data by 
including the transportation details (e.g. distance, time, and more details about the 
96 
 
trucks or ships). It will also be helpful if this model can be applied to an existing case 
study to enhance the model and overcome some of the limitations.  
97 
 
Bibliography 
Ayyub, B. (2014). Risk Analysis in Engineering and Economics, 2nd Ed., CRC Press. 
Ayyub, B. (2015). ?Practical Resilience Metrics for Planning, Design, and Decision 
Making.? ASCE-ASME Journal of Risk and Uncertainty in Engineering 
Systems, Part A: Civil Engineering, 1, 04015008. 
Barroso, A. P., Machado, V. H., Carvalho, H., and Cruz Machado, V. (2015). 
?Applications of Contemporary Management Approaches in Supply Chains.? 
Applications of Contemporary Management Approaches in Supply Chains, H. 
Tozan and A. Ert?rk, eds., InTech. 
Beamon, BM 1998, ?Supply chain design and analysis: models and methods?, 
International Journal of Production Economics, Vol. 55, no.3, pp.281?294.  
Boeing Center. (2017). ?The Supply Chain Manager?s guide to the Hurricane Harvey 
aftermath.? <https://olinblog.wustl.edu/2017/09/supply-chain-managers-
guide-hurricane-harvey-aftermath/> (accessed Feb. 15, 2020). 
Campuzano, F., and Mula, J. (2011). Supply Chain Simulation. Springer London, 
London. 
Christopher, M., and Peck, H. (2004). ?Building the Resilient Supply Chain.? The 
International Journal of Logistics Management, 15(2), 1?14. 
Cooper, M., Lambert, D., and Pagh, J. (1997). ?Supply Chain Management: More Than 
a New Name for Logistics.? International Journal of Logistics Management, 
The, 8, 1?14. 
Cox III, J. F., Blackstone Jr, J. H., and Spencer, M. S. (1995). APICS Dictionary. 
de Lara, J., Guerra, E., and Reiko Heckel. (2014). ?Domain-specific discrete event 
modelling and simulation using graph transformation.? ResearchGate, 
<https://link.springer.com/article/10.1007/s10270-012-0242-3> (accessed Feb. 
0, 2020). 
Fallon, C. (2020). ?Infrastructure: Supply chain?s missing link ? Technology ? 
CSCMP?s Supply Chain Quarterly.? 
<https://www.supplychainquarterly.com/topics/Technology/20150331-
infrastructure-supply-chains-missing-link/> (accessed Apr. 11, 2020). 
Federal Emergency Management Agency (FEMA), 2017.?2017 Hurricane Season 
FEMA After-Action Report.? Washington, DC.  
98 
 
Forrester, J. W. (1958). Industrial Dynamics A Major Breakthrough for Decision 
Makers. Harvard Business Review, 36, 37-66. Harvard Business Review. 
Fuller, T. (2011). ?Pervasive Thailand Flooding Cripples Hard-Drive Suppliers - The 
New York Times.? <https://www.nytimes.com/2011/11/07/business/global/07iht-
floods07.html> (MAR. 17, 2020).  
 
Industry Star. (2018). ?5 Supply Chain Disruptions & Mitigation Tactics.? Industry 
Star Solutions."< https://www.industrystarsolutions.com/blog/2018/09/5-
supply-chain-disruptions-reduce-impact/> (APR. 26, 2020). 
 
Justin, G. (2011). ?Interconnected Nature of Supply Chains Makes Disruptions Worse 
But Fixable, Says Professor Heese - SCM | Supply Chain Resource Cooperative 
(SCRC)? < https://scm.ncsu.edu/scm-articles/article/interconnected-nature-of-
supply-chains-makes-disruptions-worse-but-fixable-says-professor-heese> 
(APR. 25, 2020). 
 
Google Ngram Viewer ,?Supply chain?< 
https://books.google.com/ngrams/graph?content=supply+chain&year_start=1
800&year_end=2008&corpus=15&smoothing=3&share=&direct_url=t1%3B
%2Csupply%20chain%3B%2Cc0> (accessed 11 MAR 2020). 
 
Hochfelder, B. (2017). ?How to rebound from natural disasters | Supply Chain Dive.? 
<https://www.supplychaindive.com/news/natural-disaster-risk-management-
hurricane-earthquake-catastrophe/448581/> (accessed Apr. 1, 2020). 
International Organization of Standardizations (ISO), 2009b, Risk Management?Risk 
Assessment Techniques, ISO Standard IEC/FDIS 31010, iso.org, Geneva 
Lummus, R. R., and Vokurka, R. J. (1999). ?Defining supply chain management: a 
historical perspective and practical guidelines.? Industrial Management & Data 
Systems, 99(1), 11?17. 
Madu, C. N., and Kuei, C. (2017). Handbook of Disaster Risk Reduction & 
Management. WORLD SCIENTIFIC. 
Manuj, I., and Mentzer, J. T. (2008). ?Global supply chain risk management strategies.? 
International Journal of Physical Distribution & Logistics Management, 38(3), 
192?223. 
Oliveira, L., de Sousa, J. P., and Claro, J. (2017). ?Dealing with uncertainty in modern 
supply chains: vulnerability and risk management.?.Academic Press. 
99 
 
Pettit, T. J., Fiksel, J., and Croxton, K. L. (2010). ?ENSURING SUPPLY CHAIN 
RESILIENCE: DEVELOPMENT OF A CONCEPTUAL FRAMEWORK.? 
Journal of Business Logistics, 31(1), 1?21. 
Ponomarov, S. Y., and Holcomb, M. C. (2009). ?Understanding the concept of supply 
chain resilience.? The International Journal of Logistics Management, Emerald 
Group Publishing Limited, 20(1), 124?143. 
Priya Datta, P., Christopher, M., and Allen, P. (2007). ?Agent-based modelling of 
complex production/distribution systems to improve resilience.? International 
Journal of Logistics Research and Applications, 10(3), 187?203. 
 Presidential Policy Directive (PPD), 2012. ?Critical Infrastructure Security and 
Resilience whitehouse.gov.?<https://obamawhitehouse.archives.gov/the-press-
office/2013/02/12/presidential-policy-directive-critical-infrastructure-security-
and-resil> (accessed Apr. 17, 2020). 
Schmitt, A. J., and Singh, M. (2012). ?A quantitative analysis of disruption risk in a 
multi-echelon supply chain.? International Journal of Production Economics, 
139(1), 22?32. 
Scholten, K., Sharkey Scott, P., and Fynes, B. (2014). ?Mitigation Processes ? 
Antecedents for Building Supply Chain Resilience Capabilities.? Supply Chain 
Management, 19. 
Sheffi, Y., and Rice Jr, J. B. (2005). ?A Supply Chain View of the Resilient Enterprise.? 
MIT Sloan Management Review, <https://sloanreview.mit.edu/article/a-
supply-chain-view-of-the-resilient-enterprise/> (accessed Mar. 20, 2020). 
Telford, T., and Kimberly, K. (2020). ?As they rushed to maintain U.S. meat supply, 
big processors saw plants become covid-19 hot spots, worker illnesses spike.? 
The Washington Post, < 
https://www.washingtonpost.com/business/2020/04/25/meat-workers-safety-
jbs-smithfield-
tyson/?pwapi_token=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJjb29ra
WVuYW1lIjoid3BfY3J0aWQiLCJpc3MiOiJDYXJ0YSIsImNvb2tpZXZhbH
VlIjoiNTk2YmVhOTJhZTdlOGE0NGU3ZTE3NDc3IiwidGFnIjoid3BfbmV3
c19hbGVydF9yZXZlcmUiLCJ1cmwiOiJodHRwczovL3d3dy53YXNoaW5nd
G9ucG9zdC5jb20vYnVzaW5lc3MvMjAyMC8wNC8yNS9tZWF0LXdvcmtlc
nMtc2FmZXR5LWpicy1zbWl0aGZpZWxkLXR5c29uLz93cG1rPTEmd3Bpc
3JjPWFsX25ld3NfX2FsZXJ0LWVjb25vbXktLWFsZXJ0LW5hdGlvbmFsJn
V0bV9zb3VyY2U9YWxlcnQmdXRtX21lZGl1bT1lbWFpbCZ1dG1fY2FtcG
FpZ249d3BfbmV3c19hbGVydF9yZXZlcmUifQ.4IGXIstGHpJdE9UTeXsEB
hCqfYc651G9Q7Mwv_iMOPA&utm_campaign=wp_news_alert_revere&ut
m_medium=email&utm_source=alert&wpisrc=al_news__alert-economy--
alert-national&wpmk=1 > (Apr. 27, 2020).van Donk, D. P., and van der 
100 
 
Vaart, T. (2005). ?A case of shared resources, uncertainty and supply chain 
integration in the process industry.? International Journal of Production 
Economics, 96(1), 97?108. 
Wanger, S. M., and Christop, B. (2008). ?AN EMPIRICAL EXAMINATION OF 
SUPPLY CHAIN PERFORMANCE ALONG SEVERAL DIMENSIONS OF 
RISK.? JOURNAL OF BUSINESS LOGISTICS, 29(29) 307?325. 
Waters, D. (Ed.). (2003). Global logistics and distribution planning: strategies for 
management. Kogan Page, London. 
 
101