OPTIMIZATION OF DIMETHYL ETHER (DME) DISTRIBUTION IN SUMATRA THROUGH VESSELS AND PIPELINE USING MIXED INTEGER LINEAR PROGRAMMING
The increase in Indonesia's LPG demand has not been followed by equivalent domestic LPG production. The government plans to develop dimethyl ether (DME) as an alternative to LPG to overcome the shortage of LPG supply. PT X, as the main stakeholder for DME distribution, needs to create a new sup...
Saved in:
Main Author: | |
---|---|
Format: | Final Project |
Language: | Indonesia |
Online Access: | https://digilib.itb.ac.id/gdl/view/73622 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Institut Teknologi Bandung |
Language: | Indonesia |
id |
id-itb.:73622 |
---|---|
spelling |
id-itb.:736222023-06-22T09:32:31ZOPTIMIZATION OF DIMETHYL ETHER (DME) DISTRIBUTION IN SUMATRA THROUGH VESSELS AND PIPELINE USING MIXED INTEGER LINEAR PROGRAMMING Winata, Devica Indonesia Final Project supply chain operations, distribution schedule, mixed integer linear programming, inventory control INSTITUT TEKNOLOGI BANDUNG https://digilib.itb.ac.id/gdl/view/73622 The increase in Indonesia's LPG demand has not been followed by equivalent domestic LPG production. The government plans to develop dimethyl ether (DME) as an alternative to LPG to overcome the shortage of LPG supply. PT X, as the main stakeholder for DME distribution, needs to create a new supply chain plan from strategic to operational level. In a separate study, a new distribution network has been designed involving one factory, one hub, and three depots. DME will be distributed using a combination of pipes and vessels to 6 predetermined provinces. With the new distribution network, PT X needs to design a scheduling strategy to optimize daily DME distribution through pipes and vessels which can minimize operational costs and maintain inventory levels above the standard coverage days of 3 days. This study solves PT X distribution problem using discrete time mixed-integer linear programming (MILP) approaches. The problem falls under single product, multi echelon, and multi modal problem. The optimization model provides two decisions which are the distribution time and the amount of DME distributions to each demand point. Completion with the branch and bound algorithm takes too long so a 5-hour stopping rule is used. In addition, two heuristic algorithms are proposed, namely mode capacity maximization and rolling horizon algorithms. The modal capacity maximization algorithm maximizes the modal capacity utilization to minimize costs. The rolling horizon algorithm splits the solving horizon into smaller ones to reduce computation time. The mode capacity maximization algorithm produces the best solution with the lowest computational effort. The algorithm generates a distribution schedule with a total cost of 21,042,566 USD/year or 20.8 USD/ton. Pipeline distribution is scheduled every 2.6 days on average with an average parcel size of 8,153 tons DME/distribution. Total vessel calls are 477 times a year with 170 times to Dumai, 177 times to Panjang, and 130 times to Tel. Kabung. The time interval between distribution via pipe or vessel can vary from 1 day, 2 days, or even 7 days. The daily inventory level meets the standard coverage days of 3 days at all demand points. A Daily DME Distribution (DDD) tool was developed for PT X to generate DME distribution schedules daily. Lastly, this study provides several recommendations to PT X to ensure the smooth distribution of DME. text |
institution |
Institut Teknologi Bandung |
building |
Institut Teknologi Bandung Library |
continent |
Asia |
country |
Indonesia Indonesia |
content_provider |
Institut Teknologi Bandung |
collection |
Digital ITB |
language |
Indonesia |
description |
The increase in Indonesia's LPG demand has not been followed by equivalent domestic LPG production. The government plans to develop dimethyl ether (DME) as an alternative to LPG to overcome the shortage of LPG supply. PT X, as the main stakeholder for DME distribution, needs to create a new supply chain plan from strategic to operational level. In a separate study, a new distribution network has been designed involving one factory, one hub, and three depots. DME will be distributed using a combination of pipes and vessels to 6 predetermined provinces. With the new distribution network, PT X needs to design a scheduling strategy to optimize daily DME distribution through pipes and vessels which can minimize operational costs and maintain inventory levels above the standard coverage days of 3 days.
This study solves PT X distribution problem using discrete time mixed-integer linear programming (MILP) approaches. The problem falls under single product, multi echelon, and multi modal problem. The optimization model provides two decisions which are the distribution time and the amount of DME distributions to each demand point. Completion with the branch and bound algorithm takes too long so a 5-hour stopping rule is used. In addition, two heuristic algorithms are proposed, namely mode capacity maximization and rolling horizon algorithms. The modal capacity maximization algorithm maximizes the modal capacity utilization to minimize costs. The rolling horizon algorithm splits the solving horizon into smaller ones to reduce computation time.
The mode capacity maximization algorithm produces the best solution with the lowest computational effort. The algorithm generates a distribution schedule with a total cost of 21,042,566 USD/year or 20.8 USD/ton. Pipeline distribution is scheduled every 2.6 days on average with an average parcel size of 8,153 tons DME/distribution. Total vessel calls are 477 times a year with 170 times to Dumai, 177 times to Panjang, and 130 times to Tel. Kabung. The time interval between distribution via pipe or vessel can vary from 1 day, 2 days, or even 7 days. The daily inventory level meets the standard coverage days of 3 days at all demand points. A Daily DME Distribution (DDD) tool was developed for PT X to generate DME distribution schedules daily. Lastly, this study provides several recommendations to PT X to ensure the smooth distribution of DME. |
format |
Final Project |
author |
Winata, Devica |
spellingShingle |
Winata, Devica OPTIMIZATION OF DIMETHYL ETHER (DME) DISTRIBUTION IN SUMATRA THROUGH VESSELS AND PIPELINE USING MIXED INTEGER LINEAR PROGRAMMING |
author_facet |
Winata, Devica |
author_sort |
Winata, Devica |
title |
OPTIMIZATION OF DIMETHYL ETHER (DME) DISTRIBUTION IN SUMATRA THROUGH VESSELS AND PIPELINE USING MIXED INTEGER LINEAR PROGRAMMING |
title_short |
OPTIMIZATION OF DIMETHYL ETHER (DME) DISTRIBUTION IN SUMATRA THROUGH VESSELS AND PIPELINE USING MIXED INTEGER LINEAR PROGRAMMING |
title_full |
OPTIMIZATION OF DIMETHYL ETHER (DME) DISTRIBUTION IN SUMATRA THROUGH VESSELS AND PIPELINE USING MIXED INTEGER LINEAR PROGRAMMING |
title_fullStr |
OPTIMIZATION OF DIMETHYL ETHER (DME) DISTRIBUTION IN SUMATRA THROUGH VESSELS AND PIPELINE USING MIXED INTEGER LINEAR PROGRAMMING |
title_full_unstemmed |
OPTIMIZATION OF DIMETHYL ETHER (DME) DISTRIBUTION IN SUMATRA THROUGH VESSELS AND PIPELINE USING MIXED INTEGER LINEAR PROGRAMMING |
title_sort |
optimization of dimethyl ether (dme) distribution in sumatra through vessels and pipeline using mixed integer linear programming |
url |
https://digilib.itb.ac.id/gdl/view/73622 |
_version_ |
1822007162516275200 |