A Mean-Variance robust model to minimize operational risk and supply chain cost under aleatory uncertainty: A real-life case application in petroleum supply chain

https://doi.org/10.1016/j.cie.2022.107949Get rights and content

Highlights

  • Supply chain cost and risk reduction in petroleum supply chain.

  • Usage of a Mean-Variance Robust (MVR) optimization model under aleatory uncertainty.

  • Analysis under risk-seeking, risk-neutral, and risk-averse behaviors.

  • Operational risks are reduced with slight increase in total supply chain cost.

  • Penalty cost influences increasing supply chain cost along with risk aversion level.

Abstract

The purpose of the research work is to consider a multi-echelon, multi-product, and multi-modal petroleum supply chain network design problem along with its various sources of uncertainties (demand, supply, production, etc.), and to minimize both total supply chain cost and risk simultaneously. The problem is articulated as a robust optimization problem and the results are derived under various risk attitudes (viz. risk-seeking, risk-neutral and risk-averse behaviors). The non-linear problem is proposed as a Mean-Variance robust optimization problem. Two-stage stochastic programming is extended to incorporate the robustness and capture the risk aversion behavior. The scenario-based planning method is used for the estimation of uncertain parameters. A real-life case study of petroleum supply chain is conducted in the Indian scenario, considering the requisite challenges and constraints. The results show that total supply chain cost and risk demonstrate conflicting behavior with each other. Total supply chain cost increases with an increase in risk aversion level. Significant amount of operational risks can be reduced with slight increase in total supply chain cost. Penalty cost contributes the maximum to the increase in total supply chain cost along with risk aversion level.

Introduction

Modern supply chains (SC) are challenged by ever-growing competition and highly unstable business environment. Under such circumstances, the objective of making an efficient SC at the cost of effectiveness exposes the organizations to risk. The risks in SC lead to serious repercussions like inferior product quality, loss of property and machinery, loss of firm’s reputation, delivery delays, conflict among various shareholders and sharp decline in firm’s share price (Rahimi et al., 2019, Tarei et al., 2018).

Though supply chain risk management (SCRM) has emerged as a popular research domain in the recent years, the consideration of risk and uncertainty in the supply chain network design problem are typically rare (Govindan et al., 2017, Choi et al., 2017). Conventional Supply chain network design typically involves decision making regarding determination of optimal location and capacity of facilities to satisfy the market demand at the lowest cost under a static (or stable) environment. Stochastic programming models (Dantzig, 1955, Beale, 1955) developed to consider uncertainty by virtue of probabilistic information, have undergone many theoretical expansions. Although stochastic programming models have been used extensively for past few decades, the primary limitation of stochastic optimization is its inability to handle decision maker’s risk preference (or risk aversion behavior).

The supply chain network design (SCND) problem typically involves various short-term and long-term planning decisions, such as vehicle routing and facility location, respectively. The goal of strategic planning is to maximize the long-term economic performance of a SC by configuring its various decision variables. Given the configuration of a SC, the future condition of the decision parameters can not be known with certainty. This class of problem is typically called decision-making under uncertainty Uncertainty is primarily classified as aleatory and epistemic uncertainty. Aleatory uncertainty is caused by the presence of natural randomness within the system. Whereas epistemic uncertainty is arisen due to the lack of scientific knowledge or information about the process (Helton et al., 2010). The current research work considers the former type of uncertainty, which can be modelled by considering substantial numbers of scenarios, whose probability of occurrence are well known in advance. Govindan et al. (2017) have performed an extensive review of articles pertaining to supply chain network design under uncertainty, and emphasized on the consideration of SC risk in process industries such as biofuel and petroleum in future research works. Their review article also indicates a call for robust SC network design using weighted mean-risk approach.

Petroleum industries play a crucial role in the world economy because they supply the commodities to sustain the global energy supply. The petroleum supply chain (PSC) is one of the complex supply chains, which covers operational activities staring from geoseismic data analysis, exploration of crude, preprocessing, refining, storage, and distribution of petroleum products to the end consumers (Ghaithan et al., 2017).

Like any other SC, risk in PSC can be divided into two broad categories, operational risk and disruption risk (Sabouhi et al, 2018). Operational risks originate from the inherent uncertainties that inevitably exist in supply chains due to lack of coordination between multiple entities (Sreedevi & Saranga, 2017). For instance, slow demand of petroleum products by BRICS (Brazil, Russia, India, China, And South Africa) economies (Planche et al., 2016), trade restrictions (sanction) on Iran, Russia and Venezuala (Brown, 2020), crude oil price drop, transfer price, international taxation, diplomatic relations between oil importing, exporting countries, etc., are considered operational risks (Seshasayee, 2019). Disruption risks, on the other hand, are inadvertent events which restrict the flow/movement of material/information/capital in SC caused by natural, man-made or technological threats (Shekarian et al., 2020). For instance, oil spill in the Gulf of Mexico (Tabuchi & Migliozzi, 2021). Hurricane Katrina destroyed two major pipelines of the USA which lead to an unprecedented crises of oil for several months (Pan & Karp, 2005). A recent terrorist attack in the trans-regional pipeline connecting from Egypt to Jordan, Syria, and Lebanon had caused fifty percent power plants to shut down in Syria (TOI, 2021). Though it seems that disruption risks have pose more impact as compared to operational risk, the findings of Tarei et al. (2018) indicates that operational risks are pertinent in case of Indian PSC over a long-term. Though there has been extensive research on the oil industry optimization and oil supply chain under uncertainty, there are few studies considering supply chain risk management (Azarakhsh et al., 2021, Zhai and Yu, 2019).

Responding to the research issues and limitations mentioned in the previous paragraphs, we have limited the scope of the current work to operational risk only which is primarily caused by aleatory type uncertainty (Interest readers are suggested to refer to Aldrighetti et al., 2021, Xu et al., 2020, Shekarian et al., 2020, DuHadway et al., 2019; and Paul et al., 2019 for disruption risk). The proposed risk averse SC network model (focusing on aleatory type uncertainty) is articulated as a Mean-Variance Robust (MVR) optimization model, which was initially proposed by Markowitz to find out the efficient frontier of financial assets to maximize the expected return at a given level of risk (Markowitz, 1952, Chiu and Choi, 2016). We have validated the mathematical model by considering petroleum SC as an empirical case study in India. We have discussed and compared the results from different risk attitude behavioral perspective, i.e., risk seeking and risk averse.

This research contributes to the literature of PSC mainly in four ways. First, we develop a MVR optimization model focusing on the SC operational risk caused by aleatory uncertainty. Second, we explore and validate the model using a real-life case of PSC in the Indian context. Third, we analyze the results from the behavioral perspective of the decision-maker(s) (i.e., risk seeking, risk neutral, and risk averse attitude). Fourth, we provide multiple feasible solutions to balance the trade-off between SC operational cost and SC risk. By developing MVR model under aleatory uncertainties in PSC network, this study complements the extant literature of PSC that highlights the need to study uncertainties in PSC network (Azarakhsh et al., 2021, Govindan et al., 2017, Mohammadi et al., 2020) and underlines importance of operational risks in PSC (Tarei et al., 2018). As real-life situations are more uncertain and riskier, this study adds value to the literature that considers deterministic measures of PSC (Attia et al., 2019, Leiras et al., 2013; Lima et al., 2021b; Azadeh et al., 2017). We further add value to the PSC literature by bringing behavioral aspects under aleatory uncertainties in PSC in India. India which is one of the fastest emerging economies is the third-largest consumer of crude oil and petroleum products, with 4.7 percent of the global oil consumption in 2018.

The remainder of the paper is structured as follows. Section 2 provides a brief review of literature review, pertaining to the management of uncertainty focusing on petroleum industries. Section 3 mentions the theoretical foundation of MVR optimization model. Section 4 contains the assumption and the problem definition, objective function and constraints of the petroleum SC related to the case country. Section 5 presents a brief introduction of the case study and data collection procedure. Results and implications derived from the study are discussed in section 6. Finally, the paper ends with conclusions, limitations and future research directions in section 7.

Section snippets

Review of literature

In this section, we present the review of literature that motivated to conduct this research work. Since the literature, pertaining to mathematical modelling and optimization under risk and uncertainty is too broad (interested readers are suggested to refer the review paper of Govindan et al., 2017, Choi et al., 2017); we hereby limit the discussion of this section to petroleum and oil specific domain only.

Petroleum or oil SC can be broadly categorized into three segments, viz. upstream,

Mean-Variance robust optimization formulation

In this section, the proposed MVR model is explained in the light of Indian PSC.

Case introduction

To demonstrate the efficacy of the proposed MVR model, we consider a realistic case of the PSC in the eastern region of India. The oil and gas sector in India is among the country's six core industries, which contributes about 15 percent to the overall GDP of the country (Department of Industrial Policy and Promotion, Ministry of Petroleum and Natural Gas, 2017). Toward the start of 2015, India had 635 million metric tons (MMT) of demonstrated oil saves. With close to half of the country's

Results and discussions

In this section, we present the numerical results of the illustrated case study to evaluate the efficacy of the MVR model. The MVR model is coded in GAMS (General Algebraic Modeling System) and executed by CPLEX solver to find global optimality using the parallel computing platform of NEOS (Network Enabled Optimization Solution) server (Czyzyk & Moré, 1998) hosted at www.neos-server.org/neos. The proposed model is reasonably complex in terms of number of variables, number of constraints, and

Conclusion

The growing concern for risk in SC has received a global research attention. In this regard, this research work develops a MVR robust optimization framework to minimize operational risk, which is primarily caused by aleatory type of uncertainty. A real-life case study of Indian petroleum SC is conducted to validate the MVR framework, with the requisite constraints. Five uncertain scenarios are constructed based on the domestic case country’s possible (future) economic conditions. The risk

CRediT authorship contribution statement

Pradeep Kumar Tarei: Conceptualization, Methodology, Writing – original draft. Gopal Kumar: Conceptualization. M. Ramkumar: Conceptualization.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (87)

  • M. Ghatee et al.

    Optimal network design and storage management in petroleum distribution network under uncertainty

    Engineering Applications of Artificial Intelligence

    (2009)
  • K. Govindan et al.

    Investigating risk and robustness measures for supply chain network design under demand uncertainty: A case study of glass supply chain

    International Journal of Production Economics

    (2017)
  • K. Govindan et al.

    Supply chain network design under uncertainty: A comprehensive review and future research directions

    European Journal of Operational Research

    (2017)
  • A.A. Hamisu et al.

    Refinery scheduling of crude oil unloading with tank inventory management

    Computers & Chemical Engineering

    (2013)
  • G. He et al.

    A methodology for the optimal design of gathering pipeline system in old oilfield during its phased development process

    Computers & Industrial Engineering

    (2019)
  • Y. Kazemi et al.

    Modeling downstream petroleum supply chain: The importance of multi-mode transportation to strategic planning

    Transportation Research Part E

    (2015)
  • L.Y. Koo et al.

    Decision support for integrated refinery supply chains

    Computers & Chemical Engineering

    (2008)
  • R. Kouwenberg

    Scenario generation and stochastic programming models for asset liability management

    European Journal of Operations Research

    (2001)
  • X. Li

    An integrated modeling framework for design of logistics networks with expedited shipment services

    Transportation Research Part E: Logistics and Transportation Review

    (2013)
  • C. Lima et al.

    Stochastic programming approach for the optimal tactical planning of the downstream oil supply chain

    Computers & Chemical Engineering

    (2018)
  • C. Lima et al.

    Designing and planning the downstream oil supply chain under uncertainty using a fuzzy programming approach

    Computers & Chemical Engineering

    (2021)
  • S.A. MirHassani et al.

    Implications of capacity expansion under uncertainty in oil industry

    Journal of Petroleum Science and Engineering

    (2011)
  • S.A. MirHassani et al.

    Operational scheduling of refined product pipeline with dual purpose depots

    Applied Mathematical Modelling

    (2013)
  • S. Misra et al.

    A multi grid discrete time based framework for maritime distribution logistics & inventory planning for refinery products

    Computers & Industrial Engineering

    (2020)
  • S. Moradi et al.

    Transportation planning for petroleum products and integrated inventory management

    Applied Mathematical Modelling

    (2015)
  • E. Olivares-Benitez et al.

    A metaheuristic algorithm to solve the selection of transportation channels in supply chain design

    International Journal of Production Economics

    (2013)
  • F. Oliveira et al.

    Accelerating Benders stochastic decomposition for the optimization under uncertainty of the petroleum product supply chain

    Computers & Operations Research

    (2014)
  • J.M. Pinto et al.

    Planning and scheduling models for refinery operations

    Computers & Chemical Engineering

    (2000)
  • M.S. Pishvaee et al.

    Robust possibilistic programming for socially responsible supply chain network design: A new approach

    Fuzzy Sets System.

    (2012)
  • R. Qiu et al.

    A multi-scenario and multi-objective scheduling optimization model for liquefied light hydrocarbon pipeline system

    Chemical Engineering Research and Design

    (2019)
  • M. Rahimi et al.

    A stochastic risk-averse sustainable supply chain network design problem with quantity discount considering multiple sources of uncertainty

    Computers & Industrial Engineering

    (2019)
  • S.M. Santos et al.

    Information, robustness, and flexibility to manage uncertainties in petroleum field development

    Journal of Petroleum Science and Engineering

    (2021)
  • H. Sadjady et al.

    Two-echelon, multi-commodity supply chain network design with mode selection, lead-times and inventory costs

    Computers & Operations Research

    (2012)
  • R. Sreedevi et al.

    Uncertainty and supply chain risk: The moderating role of supply chain flexibility in risk mitigation

    International Journal of Production Economics

    (2017)
  • M. Tavana et al.

    A fuzzy multi-objective multi-period network DEA model for efficiency measurement in oil refineries

    Computers & Industrial Engineering

    (2019)
  • B. Wang et al.

    Optimization of a downstream oil supply chain with new pipeline route planning

    Chemical Engineering Research and Design

    (2019)
  • Abdellaoui, Wassila, Mehdi Souier, M’hammed Sahnoun, and Fouad Ben Abdelaziz. (2021). Multi-period optimal schedule of...
  • Abdussalam, O., Trochu, J., Fello, N., & Chaabane, A. (2020, November). Supply Chain Planning in the Petroleum...
  • S. Azarakhsh et al.

    Design of a sustainable integrated crude oil manufacturing network with risk cover and uncertainty considerations: A case study

    Journal of Ambient Intelligence and Humanized Computing

    (2021)
  • R. Baraiya et al.

    Transportation cost optimization model for the petroleum industry

  • E.M.L. Beale

    On minimizing a convex function subject to linear inequalities

    Journal of Royal Statistics Society

    (1955)
  • H. Beiranvand et al.

    A robust crude oil supply chain design under uncertain demand and market price: A case study

    Oil & Gas Science and Technology – Revue d’IFP Energies nouvelles

    (2018)
  • J.R. Birge et al.

    Introduction to stochastic programming

    (2011)
  • Cited by (21)

    • The IoT-enabled sustainable reverse supply chain for COVID-19 Pandemic Wastes (CPW)

      2023, Engineering Applications of Artificial Intelligence
    • Managing healthcare supply chain through artificial intelligence (AI): A study of critical success factors

      2023, Computers and Industrial Engineering
      Citation Excerpt :

      Further, AI technology can impart energy efficiency (Zhang et al., 2018b), cost reduction (Catalini and Gans, 2016), better inventory management and fraud reduction (Clauson et al., 2018), and security and traceability to the supply chain, thus generating superior performance and better customer and partner trust for businesses (IBM, 2018). An adequate supply chain reduces cost and risk (Tarei et al., 2022), improves quality, speed, and dependability (Munir et al., 2020), ensures flexibility (Azadegan et al., 2020), and enhances sustainability (Yousefi and Tosarkani, 2022). Although attaining an effective supply chain remains a challenge across industries and businesses (Schuetz and Venkatesh, 2020), the involvement of patient safety and related health outcomes generate greater complexity in the healthcare supply chain (HSC), which gets aggravated by the presence of multiple participants falling under different jurisdictions (Kwon et al., 2016; Gardas, 2022).

    View all citing articles on Scopus
    View full text