Analisis Time Series Dan Perlakuan Flushing Fouling Pada Sistem Pendingin Turbin Generator Untuk Peningkatan Keandalan PLTA 286 MW

Haq, Ahmad Habib Ul (2026) Analisis Time Series Dan Perlakuan Flushing Fouling Pada Sistem Pendingin Turbin Generator Untuk Peningkatan Keandalan PLTA 286 MW. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6032231096-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Sistem pendingin pelumas bearing merupakan komponen vital dalam menjamin kontinuitas operasional unit PLTA. Namun, akumulasi fouling pada penukar panas sering kali memicu kenaikan suhu oli yang sulit diprediksi secara manual akibat fluktuasi kualitas air sungai yang dinamis. Pentingnya studi ini terletak pada urgensi transformasi manajemen pemeliharaan dari metode konvensional berbasis waktu menjadi berbasis kondisi. Keberhasilan implementasi metode ini berdampak langsung pada pencegahan kegagalan unit akibat overheat serta efisiensi biaya perawatan melalui penjadwalan yang lebih terukur. Penelitian ini bertujuan untuk mengatasi ketidakpastian laju pembentukan kerak yang mengakibatkan inefisiensi pada jadwal pembersihan (flushing). Masalah diselesaikan dengan pendekatan analisis data aktual operasional melalui perbandingan model statistik lanjut. Tiga algoritma utama, yaitu Holt-Winters, Prophet, dan ARIMAX, dievaluasi untuk menemukan model prediktif yang paling representatif terhadap dinamika fouling resistance (Rf) di lapangan. Hasil penelitian mengonfirmasi bahwa model ARIMAX unggul dengan tingkat akurasi sangat tinggi, ditunjukkan oleh nilai Mean Absolute Percentage Error (MAPE) sebesar 4,22%. Analisis termodinamika menunjukkan setiap kenaikan 0,001 unit Rf berkontribusi pada peningkatan suhu oli sebesar 0,727°C. Dengan menetapkan ambang batas kritis suhu pada 65°C, diperoleh perhitungan bahwa interval pemeliharaan paling optimal tercapai setiap 47,4 hari operasional. Simpulannya, sinergi antara analisis termodinamika dan pemodelan ARIMAX mampu memberikan estimasi jadwal pemeliharaan yang presisi. Pendekatan ini terbukti efektif sebagai instrumen pengambilan keputusan bagi operator untuk menjaga keandalan unit sekaligus mengoptimalkan siklus operasional pembangkit secara berkelanjutan.
========================================================================================================================
The lubrication cooling system for unit bearings is a vital component in ensuring the operational continuity of a Hydroelectric Power Plant (HEPP). However, fouling accumulation on heat exchangers frequently triggers oil temperature rises that are difficult to predict manually due to dynamic river water quality fluctuations. The significance of this study lies in the urgency of transforming maintenance management from conventional time-based methods to condition-based strategies. Successful implementation of this method directly impacts the prevention of unit failure caused by overheating and improves maintenance cost efficiency through more measurable scheduling. This research aims to address the uncertainty of fouling formation rates, which leads to inefficiencies in flushing schedules. The problem is resolved using an actual operational data analysis approach by comparing advanced statistical models. Three primary algorithms—Holt-Winters, Prophet, and ARIMAX—were evaluated to identify the most representative predictive model for the dynamics of fouling resistance (Rf) in the field. The research results confirm that the ARIMAX model is superior, with a very high level of accuracy, as indicated by a Mean Absolute Percentage Error (MAPE) of 4.22%. Thermodynamic analysis shows that every 0.001 unit increase in Rf contributes to an oil temperature rise of 0.727°C. By setting a critical temperature threshold of 65°C, calculations reveal that the most optimal maintenance interval is achieved every 47.4 operational days. In conclusion, the synergy between thermodynamic analysis and ARIMAX modeling provides precise maintenance schedule estimations. This approach proves effective as a decision-making instrument for operators to maintain unit reliability while optimizing the power plant's operational cycles sustainably.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Fouling Resistance, Pendingin Bearing, ARIMAX, Interval Optimal, PLTA, Fouling Resistance, Bearing Cooler, ARIMAX, Optimal Interval, HEPP
Subjects:	T Technology > TJ Mechanical engineering and machinery > TJ174 Maintenance and repair of machinery
Divisions:	Interdisciplinary School of Management and Technology (SIMT) > 61101-Master of Technology Management (MMT)
Depositing User:	Ahmad Habib Ul Haq
Date Deposited:	05 Feb 2026 07:46
Last Modified:	05 Feb 2026 07:46
URI:	http://repository.its.ac.id/id/eprint/132192

Actions (login required)

View Item