Throughout the last century, there has been a growing public awareness about environmental issues and serious effects of greenhouse gas (GHG) emissions, such as global warming and climate change. At the governmental level, the Kyoto Protocol (1997), which was signed by 37 industrialised countries, has urged them to decrease their GHG emissions. Changes in environmental consciousness have also an impact on the industrial sectors. Regarding the environmental concerns and regulations that have been enacted, it is critical for industries to put effort into emission mitigation.
As a result, decision-makers must establish the best environmental management strategies while keeping the economic aspect in mind. In this regard, we focus on the topic of integrating environmental concerns in the management of manufacturing systems using a stochastic optimal control approach. In practice, the manufacturing sector has a particularly dynamic behaviour due to the multiple variables and random events such as random failure and repair times. In this thesis, we address the production planning problem of unreliable manufacturing systems integrated with environmental aspects.
In the first part, we focus on the production planning problem of manufacturing systems consisting of high-emitting machines (HEMs), which are willing to gradually invest in lowemitting machines (LEMs). New cost-effective and environmentally friendly control policies are developed for such companies to help the decision-maker in synchronising HEM and LEM. Accordingly, we develop two new policies under which both HEM and LEM can run at the same time to boost system availability while simultaneously reducing inventory, backlog, and generated emissions. The simulation modeling, experimental approach, and response surface methodology are used to obtain the optimal parameters of these control policies. Finally, various experiments conducted for a wide range of cost and system parameters, revealed that our control policies result in considerable cost saving and GHG emissions reduction.
In the second part, we address the problem of trading and production planning for failure-prone manufacturing systems regulated by cap-and-trade scheme. In the case of the carbon price, which randomly changes in the market, it is difficult and complicated for businesses to determine their production rate and trading volume in order to make a profit from the carbon market. We develop a new joint production and trading control policy to help managers in determining when to buy/sell allowances or increase/decrease the production rate in order to minimize the total cost and reduce GHG emissions. Finally, sensitivity analysis and a comparative study are conducted to demonstrate how our proposed joint policy leads to significant cost savings and emissions decrease when compared to existing policies adapted from the literature.
In the third part, we integrate the environmental considerations into failure-prone manufacturing/remanufacturing systems taking variate quality returns into account. Specific quality conditions (high- or low-quality) are assigned to returned products. The impact of the returns’ quality on GHG emissions and production rates is also taken into consideration. We develop two production control policies to synchronise manufacturing and remanufacturing considering both high- and low-quality returns in order to achieve environmental and economic goals. The optimal value of parameters is calculated using simulation, design of experiment, and response surface methodology. Based on the results, our proposed policies outperform the existing policies adapted from the literature in terms of both optimal total costs and emissions reduction.
In the fourth part, we focus on unreliable manufacturing systems which consume high-emitting fuels (e.g., coal) and tend to reduce emissions by switching to cleaner fuels (e.g., natural gas (NG)). Indeed, production using a cleaner fuel type reduces GHG emissions but results in higher fuel cost. We also take the random fluctuation of NG price into account. We developed production control policies for failure-prone manufacturing systems with the possibility of fuel switching to minimize total costs while reducing GHG emissions. In this vein, two production control policies are developed to achieve both economic and environmental goals. According to our proposed policies, decisions on fuel switching are made based on the random fluctuation of NG price, and the key elements of the system including its state, inventory, and emissions level. Our proposed policies could reduce emissions while minimizing total costs including inventory, backlog, fuel, and emission costs. The optimal value of control factors is determined using simulation, design of experiment, and response surface methodology. Our proposed control policies surpass the policies derived from the literature in terms of both optimal total costs and emissions reduction.
In conclusion, this thesis contributes to the field of production planning and control of unreliable manufacturing systems in several ways. Indeed, we look at the environmental aspects of such systems from multiple perspectives (low-emitting technologies, carbon trading, reverse logistics, fuel switching). Efficient structures of production control policies are developed, contributing to a better understanding of the dynamic and stochastic behavior of such systems. Comparative study and in-depth sensitivity analysis are performed to validate the structure of developed control policies.
| Date | 23 Sept 2022 |
|---|
| Original language | American English |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | Ali Gharbi (Supervisor) & Mustapha Ouhimmou (Co-supervisor) |
|---|