Aristotle University Of Thessaloniki (AUTH)
Ανάπτυξη δυναμικών προτύπων διαχείρισης καταστροφικών κινδύνων σε συστήματα εφοδιαστικής αλυσίδας κλειστού βρόχου με διαδικασίες ανακατασκευής και ανακύκλωσης
Abstract
dc:descriptionThis doctoral dissertation examines risk/disaster management in Closed-loop Supply Chains (CLSCs) with remanufacturing and recycling processes. More specifically, a dynamic simulation model has been developed in order to check the response of a CLSC system under the influence of a disaster event by employing alternative disaster mitigation policies. The aim of this study is to explore a bundle of policies and identify best practices which favor the creation of a sustainable CLSC under the effect of alternative scenarios in terms of intensity and duration of the disaster event. In sum, the literature review aims to adumbrate, classify and evaluate the basic parameters examined, which affect the system in the processing of supply chains (SCs) under risk/disaster circumstances. Subsequently, a simulation model has been developed, which studies the dynamic behavior of a CLSC system with remanufacturing and recycling processes, in accordance with the methodological approach of the Theory of System Dynamics.More specifically, the dissertation draws alternative mitigation policies, which are primarily focused on the producer, against disaster events, taking the observations and the findings of the relevant literature into consideration. By using the simulation model, the alternative policies are examined under different disaster circumstances. For each circumstance, the processing of the simulation output reveals best practices. In detail, during the first stage (Section 1) of the dissertation, the systematic literature review was conducted. The purpose was to highlight the importance of sustainability in CLSCs functioning in the context of risk. This is reflected through a methodologically documented comparative taxonomy of the literature review, which aims not only to discover gaps in existing literature and to bring forward the academic contribution of the present dissertation, but also to propose mitigation policies against the effects of risk/disaster events as well as a methodological approach so as to develop a dynamic model aiming at sustainable development of CLSCs. Consequently, the dissertation attempts: a) a critical literature taxonomy in pillars based on the nature of the event hitting the system (risk or disaster event), as well as on the level of decision-making (strategic, tactical, operational); b) to identify and bring forth major and reliable policies addressing the effects of the risk/disaster event, which could be used on a modeling level; and c) to document a methodological approach in order to control and evaluate the effects of alternative policies, while focusing on the sustainability of a CLSC system. The main results from the literature review are the following: i) The capture of the functional features of CLSC systems with remanufacturing and recycling processes. ii) The capture of the main types of risk/disaster, their sources and their effects on SC and CLSC systems. iii) The emerging way of risk/disaster management in SC and CLSC systems. iv) The critical taxonomy of the findings of the existing literature, including all aspects of the sustainability of SC systems (forward, reverse, closed-loop) based on the following criteria: a) the nature of the risk/disaster event; b) the source of the risk/disaster event; c) the structure of the SC; d) the level of decision-making (strategic, tactical, operational); e) objective (viability and sustainability); and f) the method for addressing the effects of risk/disaster. v) The delineation of the most important policies for supporting the methodological framework for decision-making and, by extension, the design of viable SCs. vi) The absence of a holistic approach for the management of CLSCs under risk/disaster event and the documentation of the suitability of System Dynamics (SD) for use as a methodological tool, mainly for the following reasons: a. An SD approach allows to study the dynamic behavior of the system under alternative scenarios as regards the structure and the rules with which it functions, as well as under alternative scenarios as regards the influences from the external environment; b. SD has been established as a scientific field offering the possibility to design and analyze systems with increased structural and functional complexity; c. The CLSC system examined in the present dissertation demonstrates high complexity, but also many interactions between its involved partners. vii) The delineation of a uniform methodological framework for decision-making in order to manage disaster events at strategic, tactical and operational level with the aim to successfully achieve the transition to sustainable and viable SC systems. All in all, the systematic literature review has revealed the necessity of CLSC management under risk/disaster circumstances with the use of SD, aiming at the development of a sustainable system viewed as a whole and not in isolated parts. In the second stage (Section 3) of the dissertation, a detailed description of the dynamic model for the CLSC system with remanufacturing and recycling processes takes place. More specifically, the components of the CLSC system along with its operational mechanisms under normal circumstances (in the absence of disaster) are described. The CLSC system is divided into eight sub-systems. This division facilitates the description of the operations under normal-operation conditions, so that both the effects of the disaster event and the application of suitable mitigation policies can be easily understood later on. The model described in this stage will be the basis for the subsequent creation of other models in order to explore alternative policies (basic scenario). 1.Raw materials supplier's sub-system 2. Spare parts producer's sub-system 3. Manufacturer's sub-system 4. Product purchase (wholesaler, retailer) sub-system 5. Used products collection sub-system 6. Disassembly and remanufacturing sub-system 7. Recycling sub-system 8. Profit calculation sub-system. Concerning the remanufacturing process, it is supposed that the process retrieves as-good-as-new parts. The forward channel includes five partners, the Raw material supplier, the Parts producer, the Manufacturer, the Wholesaler and the Retailer. The Raw material supplier obtains materials from the raw material market. The Parts producer produces parts through two inflows of raw material, original raw materials from the Raw material supplier and material recovered from recycling taking place in the reverse channel. Priority is given to supply with recycled materials. The Manufacturer produces end products by assembling n=1,2, ...., N parts, obtained through two inflows: i) original parts provided by the Parts producer and remanufactured parts by the remanufacturing process. Remanufactured parts are prioritized against the use of original parts. The Wholesaler buys end products from the Manufacturer, while the Retailer from the Wholesaler in order to satisfy market demand. The reverse channel includes the following activities: The collection of used products, conducted by the Collector; their disassembly, conducted by the Disassembly center, where remanufacturing, recycling and disposal at landfills activities take place. More specifically, used products are collected by the Collector for reuse and are then transported and sold at the Disassembly center. The condition and quality of the parts obtained through disassembly is checked so that they are distributed for remanufacturing, recycling or disposal. The condition of those parts deemed unsuitable for remanufacturing is checked for recycling. Parts acceptable for recycling are classified as recyclable parts, while those rejected are sorted for disposal to landfills. Throughout the detailed analysis of the individual sub-systems composing the main system, it is understood that the model possesses high structural complexity, as it includes 22 state variables and 23 feedbacks. In addition, the mathematical representation of the model shows a high interaction between the partners, which makes the system vulnerable to disruptions, such as those resulting from a disaster event. Finally, in the second stage of the dissertation, a validity check of the model was performed, by applying tests suggested by the SD literature. The dynamic modeling of the disaster event hitting the system is presented in the third stage of the dissertation (Section 4), along with a description of its major features. The source of the disaster is environmental. In particular, the occurrence of the disaster results in the production rate a sudden decrease to a great extent (sometimes even in zero production rate). Furthermore, it results in significant delays as regards delivery times, as well as in times of critical operational parameters (e.g. inventory adjustment times). When the phenomenon occurs, all partners to the CLSC are affected. In particular, it is assumed that the Manufacturer is directly affected, while the rest of the partners are indirectly affected. In addition to the initial hypothesis, where it is considered that only the manufacturer is affected directly, the case that one or some of the other partners are affected directly is also considered, resulting in total interruption of operation during the phenomenon.To address the effects of the occurrence of the disaster event, alternative mitigation policies have been designed, based on the findings of the literature review (stage 1). From a technical perspective, each policy is designed as a feedback mechanism, which attempts to restore selected crucial state variables from actual values reached under undesirable circumstances of instability, due to the occurrence of disaster, to equilibrium with desired values. In other words, control mechanisms (CMs) for the dynamic behavior of the system are designed. These mechanisms are activated when the system falls in instability phase due to the disaster event, during the transient period In particular, four policies at the Manufacturer's sub-system have been designed (CM1-CM4). Each policy corresponds to a different Business Scenario (BS). In BS1, system response is examined through the development and maintenance of different levels of safety stock during the pre-disaster period (active CM1). Scenario BS2 adopts a policy that increases the Manufacturer's productivity (through an additional shift) during the recovery period only (active CM2). Scenario BS3 (active CM3) sees the Manufacturer adopting a policy that creates additional stock as a remote inventory before the occurrence of the event, so that the Wholesaler is supplied throughout the occurrence of the event. In scenario BS4 (active CM4), partnerships with third-party suppliers are considered by the Manufacturer. These partnerships vary according to the percentage of wholesale orders that needs to be fulfilled, which is set down by each agreement. Concerning the rest of the parties to the SC, three policies have been implemented. These relate to the Parts producer's sub-system (CM5), the disassembly and remanufacturing sub-system (CM6) and the collection of used products sub-system (CM7) respectively. These policies are activated during the occurrence of the phenomenon in order to accumulate temporary inventory, which will help meet demand in the post-disaster period. Subsequently, the alternative scenarios regarding the effects of the disaster phenomenon on the system have been described. These scenarios relate to the intensity and the duration of the phenomenon. More specifically, alternative scenarios are described as regards the reduction of the Manufacturer's production rate and the increase of delivery time. The reduction of production rate in the occurrence of the phenomenon may be partial or total. As regards product demand, it may follow three demand patterns in the occurrence of a disaster event: i)Demand pattern 1: Demand marks a sudden decrease throughout the duration of the phenomenon. In the post-disaster period, demand returns to pre-disaster value levels. ii) Demand pattern 2: Demand is unaffected by the phenomenon. iii) Demand pattern 3: Demand marks a sudden increase in the occurrence of the phenomenon. This increase persists for as long as the phenomenon lasts. In the aftermath of the phenomenon, demand returns to pre-disaster value levels.Regarding the duration of the disaster phenomenon, three cases are identified: small-scale phenomen, medium-scale phenomena, large-scale phenomena. Finally, indices calculating changes in crucial state variables in comparison with the basic scenario (BBS) have been established so as to measure the efficiency of each policy at the Manufacturer's level, along with the application (or lack thereof) of policies related to the rest of the partners. In particular, these indices examine the economic performance of the system and the inventory level at the Manufacturer level. The fourth stage of the dissertation (Section 5 and Section 6) tests the model in order to explore the theoretical perspective and the effect of the COVID-19 pandemic on a CLSC system and to identify best practices among previously presented policies. Owing to the nature of the phenomenon, its duration was assumed to be steady and equal to 6 weeks, while the demand patterns that were examined relate either to the reduction of product demand (Pattern 1) or to absence of any effect on product demand (Pattern 2). At a Manufacturer level, the output demonstrates that different best practices emerge to address disruptions caused by the COVID-19 outbreak both at an economic and at an inventory level. At an economic level, choices depend on the targeted result. If the target combines: a) restoration of the system to its pre-COVID-19 state from a cost-efficiency perspective; and b) achieving (as much as possible) total profit compared to basic scenario (BBS), then we should first and foremost examine the percentage of production rate reduction and demand reduction. In case the reduction of production rate and demand is either 100% and 20% respectively or 50% and 0% respectively, contracting with third-parties is revealed as the best policy choice. In the rest of the cases, best policy choices revolve around the preservation of basic stock during the pre-COVID-19 period and the use of additional shift during the recovery period. At an inventory level, if the Manufacturer’s target is to restore the stock to its pre-COVID-19 levels as soon as possible, basic stock preservation policies will have to be implemented. Especially as regards the faster transition of inventory level to equilibrium state, those policies will reveal the speed with which the system responds to reductions in the values of inventory adjustment time. Finally, a remote inventory policy for the Manufacturer to supply the Wholesaler during the recovery period is revealed as the worst policy choice from the perspective of profit. The dynamic model developed in the present doctoral dissertation could be a useful tool for CLSC managers so as to create a viable system under disaster circumstances. The configuration of the dynamic model, which should be adapted to the specificities of each CLSC case study, by using appropriate values is a necessary precondition. Last but not least, expanding the model by developing endogenous mechanisms for defining values to the parameters involved in CMs could be an interesting question for further academic research.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Katsoras, Efthymios
- Κατσώρας, Ευθύμιος
Subjects
dc:subject × 13- Εφοδιαστικές αλυσίδες κλειστού βρόχου
- Διαχείριση καταστροφικών κινδύνων
- Δυναμική συστημάτων
- Πολιτικές μετριασμού
- Closed loop supply chains
- Disaster management
- System dynamics
- Mitigation policies
- Επιστήμη Μηχανολόγου Μηχανικού
- Επιστήμες Μηχανικού και Τεχνολογία
- Μηχανολογική μηχανική
- Mechanical Engineering
- Engineering and Technology
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/51995