Setiawan, Iwan (2026) Strategi Peningkatan Keandalan Jaringan Distribusi Listrik melalui Implementasi FLISR dengan Metode FMEA dan Fuzzy Logic (Studi Kasus: PLN UP2D Jawa Barat). Masters thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 6022241029-Master-Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6022241029-Master-Thesis.pdf - Accepted Version Restricted to Repository staff only Download (11MB) | Request a copy |
Abstract
Sistem Fault Location, Isolation, and Service Restoration (FLISR) pada jaringan distribusi 20 kV sangat krusial untuk meminimalkan durasi pemadaman. Implementasi FLISR berbasis Survalent ADMS di GI Malangbong PLN UP2D Jawa Barat saat ini memiliki tingkat keberhasilan 70%, masih di bawah target. Analisis data 2022–2024 menunjukkan dari 141 kegagalan, 94,3% disebabkan oleh komponen lapangan, bukan logika sistem. Namun, belum ada metodologi risiko sistematis untuk memprioritaskan mitigasi dan memproyeksikan peningkatan keandalan secara kuantitatif. Penelitian ini mengembangkan kerangka terintegrasi yang menggabungkan Failure Mode and Effect Analysis (FMEA) dengan Sistem Inferensi Fuzzy Mamdani menggunakan MATLAB R2023b. Tujuannya adalah menghasilkan Total Risk Score (TR*) skala 0–100. Berdasarkan Focus Group Discussion (FGD) dengan delapan pakar dari PLN dan ITS, diidentifikasi 25 mode kegagalan dalam lima kategori: Baterai Cadangan (36,2%), Motorized Actuator (29,8%), Komunikasi & Sinyal (22,0%), RTU (6,4%), dan Logika ADMS (5,7%). Setiap mode dinilai berdasarkan parameter Severity (S), Occurrence (O), dan Detection (D). Sistem Fuzzy dengan 125 aturan IF-THEN menghasilkan skor TR* yang lebih akurat daripada metode RPN konvensional. Hasilnya, 76% dari 25 mode kegagalan mengalami perubahan peringkat risiko. Pergeseran signifikan terlihat pada MTR-05 yang turun 4 posisi, serta ADM-02 dan KOM-04 yang masing-masing naik 6 dan 5 posisi. Skor tertinggi (91,7) ditemukan pada BAT-01 dan BAT-02, dipicu oleh sensitivitas tinggi terhadap deteksi yang buruk (D=8–9). Empat strategi mitigasi dirumuskan: manajemen baterai (+9,2 pp), pemeliharaan motorized switch (+4,7 pp), penguatan komunikasi (+5,2 pp), dan optimasi ADMS/RTU (+5,9 pp). Implementasi kumulatif diproyeksikan meningkatkan success rate FLISR dari 70% menjadi 92,5% pada skenario moderat (rekomendasi), dengan potensi mencapai 95% pada skenario optimistis. Validasi melalui simulasi DIgSILENT PowerFactory 2023 menunjukkan penurunan SAIDI sebesar 21,9% menjadi 1,32 jam/pelanggan/tahun (hingga 28,4% menjadi 1,21 jam/pelanggan/tahun pada skenario optimistis) dan penurunan CAIDI sebesar 21,8% menjadi 2,40 jam/gangguan, sementara SAIFI tetap pada 0,55. Hasil ini membuktikan bahwa kerangka kerja yang diusulkan efektif dalam memenuhi standar keandalan SPLN 68-2:1986.
=======================================================================================================================================
The Fault Location, Isolation, and Service Restoration (FLISR) system in 20 kV distribution networks is vital for minimizing power outage durations. Currently, the Survalent ADMS-based FLISR at the GI Malangbong, PLN UP2D Jawa Barat, achieves a 70% success rate, falling short of operational targets. Analysis of 141 failure events during 2022–2024 reveals that 94.3% were caused by field component failures rather than ADMS logic. However, a systematic risk-based methodology to link failure modes with prioritized mitigation and validated reliability projections is still missing. This research develops an integrated framework combining Failure Mode and Effect Analysis (FMEA) with a Mamdani Fuzzy Inference System using MATLAB R2023b to generate a Total Risk Score (TR*) on a 0–100 scale. Based on Focus Group Discussions (FGD) with eight experts from PLN and ITS, 25 failure modes were identified across five groups: Backup Batteries (36.2%), Motorized Actuators (29.8%), Communication & Signals (22.0%), RTU (6.4%), and ADMS Logic (5.7%). Each mode was evaluated using Severity (S), Occurrence (O), and Detection (D) parameters.
The Fuzzy system, utilizing 125 IF-THEN rules, provides a more representative TR* compared to conventional RPN. Consequently, 76% of the failure modes experienced rank shifts. Significant changes occurred in MTR-05, which dropped 4 positions, while ADM-02 and KOM-04 rose by 6 and 5 positions, respectively. BAT-01 and BAT-02 recorded the highest TR* (91.7) due to high sensitivity to poor detection values (D=8–9). Four mitigation strategies were formulated: battery management (+9.2 pp), motorized switch maintenance (+4.7 pp), communication reinforcement (+5.2 pp), and ADMS/RTU optimization (+5.9 pp). Cumulative implementation is projected to increase the FLISR success rate from 70% to 92.5% under the moderate (recommended) scenario, with an optimistic upper bound of 95%. Validation via DIgSILENT PowerFactory 2023 simulations showed SAIDI decreasing to 1.32 hours/customer/year (−21.9%) under the moderate scenario (down to 1.21 hours/customer/year, −28.4%, under the optimistic scenario), SAIFI remaining at 0.55, and CAIDI dropping to 2.40 hours/interruption (−21.8%). These results successfully meet the SPLN 68-2:1986 reliability standards.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | FLISR, FMEA, Fuzzy Logic, Keandalan Distribusi, SAIDI, Distribution Reliability |
| Subjects: | T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3030 Electric power distribution systems |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20101-(S2) Master Thesis |
| Depositing User: | Iwan Setiawan |
| Date Deposited: | 31 Jul 2026 09:11 |
| Last Modified: | 31 Jul 2026 13:20 |
| URI: | http://repository.its.ac.id/id/eprint/141072 |
Actions (login required)
 |
View Item |