A multi-echelon, multi-period strategic planning supply chain model for natural gas with stochastic demand

• Main Author: Rubio, Ramon Felipe D.
• Format: text
• Language: English
• Published: Animo Repository 2006
• Subjects: Natural gas; Natural gas reserves; Industrial Engineering
• Online Access: https://animorepository.dlsu.edu.ph/etd_masteral/3393; https://animorepository.dlsu.edu.ph/context/etd_masteral/article/10231/viewcontent/CDTG004071_P.pdf
• Institution: De La Salle University

Although natural gas is one of the more ideal alternative fuels, the lack of refueling infrastructure pose impediments to the future spread of natural gas vehicles. Several supply chain models reviewed differ with the natural gas supply chain in that they have differing number of echelons, are limited only by series connections, involve distance and demand information modeling, stocking policies, lack of suppliers for warehouses and processing plants, and so on. There are unique features of the natural gas supply chain that cannot be found in traditional supply chain models, such as the depleting supply wells and the network of pipelines gas has to go through. There is a lack of a backbone type of model for the natural gas supply chain that could be flexible depending on the needs of the user, where adding parameters to a simple yet complete model that captures the behavior of the natural gas supply chain can be done. To address this, a multi-echelon, multi-period strategic planning supply chain model for natural gas with stochastic demand was formulated. The model focuses on four echelons of the natural gas supply chain – the supply wells, the processing plants, the storage facilities, and the fueling centers. Focus is given on the relevant costs such as capital costs for opening facilities, operating expenses, the holding costs for well and storage facilities, and extraction, transportation, and delivery costs. The model's objective was to minimize total system costs, with decision variables of how much to extract from wells, how much to process in each plant, and how much to deliver to customers. Site selection among candidate sites is also a key decision in the model. The model was run in General Algebraic Mathematical Model System (GAMS) solver. The model was able to capture the expected behavior, and it is shown that a combination of low costs in operating expense, capital, and holding costs with their respective discounted rates help the model decide and are relevant parameters for the decision making, as well as the depleting nature of the supply wells. Through the sensitivity analysis, it was determined that total costs increase primarily when more wells are used compared to having more storage facilities. Opening of storage facilities is more desirable compared to having more wells opened. Wells are opened primarily to keep more gas in the storage facilities, and also when there is little extractable amount to fulfill demand. Plants are opened as more wells are opened, depending on the set of wells that have to be accommodated and processing plants cannot serve all wells. Storage facilities are ideally used more to store compared to supply wells, as they are not only nearer to the customers but also they are less expensive to maintain. In response to varying demands, the model is able to become flexible and decides based on the costs of the facilities, explores the trade offs between varying expenses such as high capital but low transfer cost, and vice versa. For future studies, it can be recommended that the focus would be on the stochastic nature of the demand with weights given to unmet demand and overstocking, the stochastic nature of the wells capacities, whose uncertain probability when it comes to having a good supply of natural gas also affects decision making in the supply chain, and exploration of more sources of gas in terms of imports could be explored to determine how it could affect the opening of more wells. Also, demand points could be added that would be direct from the plant and not having to pass through the storage facility, as is the case in some scenarios.