Antara, Gede Bagus Dwi Suasti (2026) Pendekatan Hybrid Clustering, Optimization, and Selection Mapping (H-COSM) Framework untuk Desain Strategis Rantai Pasok LNG Skala Kecil di Negara Kepulauan. Doctoral thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of Disertasi]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text (Disertasi)
7019202001_DOCTORAL.pdf - Accepted Version Restricted to Repository staff only Download (8MB) | Request a copy |
Abstract
Wilayah Indonesia bagian timur yang meliputi Papua, Papua Barat, dan Maluku mencatatkan pertumbuhan permintaan listrik tertinggi nasional sebesar 12,9% per tahun, namun sebagian besar pembangkit listriknya masih bergantung pada diesel oil yang mahal, meskipun kawasan ini menyimpan cadangan gas alam terbesar di Asia Tenggara. Distribusi LNG ke puluhan pembangkit berkapasitas kecil (5–40 MW) yang tersebar di kepulauan dengan perairan dangkal dan infrastruktur terbatas tidak dapat mengikuti model distribusi LNG konvensional. Penelitian ini mengembangkan Kerangka Kerja H-COSM (Hybrid Clustering, Optimization, and Selection Mapping) sebagai arsitektur metode hibrida yang memetakan setiap subpermasalahan distribusi LNG ke metode analitik yang paling sesuai karakteristik matematisnya. Dua studi kasus diterapkan: (1) perancangan distribusi LNG untuk 30 pembangkit di Papua dan Maluku, di mana K-Means++ digunakan untuk klasterisasi spasial dan Greedy Capacitated Vehicle Routing Problem (CVRP) untuk optimasi rute dan armada; dan (2) penentuan lokasi fasilitas Ship-to-Ship bunkering LNG di perairan Batam dengan menggunakan metode hybrid Analytic Hierarchy Process (AHP) - Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS). Hasil penelitian menunjukkan empat temuan utama. Pertama, konfigurasi lima klaster (k=5) merupakan solusi optimal untuk menyelesaikan penghalang geografis di wilayah kepulauan; model bisnis distribusi berbasis klaster ini layak secara ekonomi dengan IRR 12% (> WACC 10,57%) pada tolling fee USD 6,00/MMBtu dan umur kontrak 20 tahun. Kedua, Algoritma Greedy CVRP menghasilkan skema distribusi optimal yang menggunakan Floating Storage Unit di Fakfak sebagai hub antara, dilayani armada kapal tipe-C berkapasitas 2.500–20.000 m³ dan tongkang LNG untuk lokasi perairan dangkal. Ketiga, empat model konseptual terminal penerima LNG dirancang untuk mengakomodasi variasi kondisi lapangan—dari terminal standar hingga hub mini dan terminal berbasis truk, yang mana seluruhnya mengacu standar NFPA 59A. Keempat, Kabil ditetapkan sebagai lokasi STS bunkering paling optimal di perairan Batam (Ci = 0,77), stabil pada hampir seluruh skenario analisis sensitivitas; jika tiga lokasi dikembangkan bersamaan, kombinasi Kabil, Tanjung Uban, dan Batu Ampar merupakan pilihan terbaik. Ketiga hipotesis penelitian terkonfirmasi, memvalidasi H-COSM sebagai kerangka kerja yang modular, transparan, dan dapat diadaptasi untuk merancang rantai pasok energi di negara kepulauan lainnya.
==================================================================================================================================
Eastern Indonesia, comprising Papua, West Papua, and Maluku, has the highest annual growth rate of electricity demand in the nation at 12. 9%. Nevertheless, most power plants in this region depend on costly diesel fuel, despite the area being endowed with Southeast Asia's most substantial natural gas reserves. The distribution of liquefied natural gas (LNG) to numerous small-capacity power plants (ranging from 5 to 40 MW) dispersed across shallow-water archipelagic terrains with limited infrastructure cannot adhere to traditional LNG distribution models. This research introduces the H-COSM Framework (Hybrid Clustering, Optimization, and Selection Mapping), a hybrid methodological architecture that assigns each LNG distribution subproblem to the most suitable analytical approach from a mathematical perspective. Two case studies are conducted: (1) designing an LNG distribution network for thirty power plants in Papua and Maluku utilizing K- Means ++ for spatial clustering and the Greedy Capacitated Vehicle Routing Problem (CVRP) for route and fleet optimization; and (2) selecting a site for a Ship- to- Ship LNG bunkering facility in Batam waters employing a hybrid Analytic Hierarchy Process (AHP) - Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) method. Four principal findings are identified. First, a configuration comprising five clusters (k = 5) is determined to be optimal, accurately reflecting actual geographic obstacles; the cluster- based distribution business model proves economically viable at an internal rate of return (IRR) of 12%, exceeding the weighted average cost of capital (WACC) of 10. 57%, with a tolling fee of USD 6. 00 per MMBtu over a twenty- year contract period. Second, the Greedy CVRP algorithm facilitates the development of an optimal distribution scheme employing a Floating Storage Unit in Fakfak as an intermediate hub, serviced by Type-C LNG vessels with capacities ranging from 2,500 to 20,000 m³, alongside LNG barges designed for shallow-water locations. Third, four conceptual models of LNG receiving terminals are developed to accommodate variability in field conditions, encompassing standard terminals, mini-hubs, and truck-based terminals, all conforming to NFPA 59A standards. Fourth, Kabil is identified as the most suitable site for Ship- to- Ship (STS) LNG bunkering in Batam waters (Ci = 0. 77), demonstrating robustness across various sensitivity analysis scenarios; if three sites are developed concurrently, the combined utilization of Kabil, Tanjung Uban, and Batu Ampar is recommended. All three research hypotheses are supported, validating H-COSM as a modular, transparent, and adaptable framework for energy supply chain design in archipelagic nations.
| Item Type: | Thesis (Doctoral) |
|---|---|
| Uncontrolled Keywords: | AHP-TOPSIS, Capacitated Vehicle Routing Problem, Greedy Algorithm, H-COSM Framework, K-Means++, LNG supply chain |
| Subjects: | T Technology > T Technology (General) > T57.6 Operations research--Mathematics. Goal programming |
| Divisions: | Faculty of Marine Technology (MARTECH) > Marine Engineering > 36001-(S3) PhD Theses |
| Depositing User: | Gede Bagus Dwi Suasti Antara |
| Date Deposited: | 05 Aug 2026 09:41 |
| Last Modified: | 05 Aug 2026 09:41 |
| URI: | http://repository.its.ac.id/id/eprint/144076 |
Actions (login required)
 |
View Item |