PENGEMBANGAN SISTEM PEMANTAUAN KONDISI UNTUK KESELAMATAN ROTATING MACHINE DI PWR DENGAN MOTOR CURRENT SIGNATURE ANALYSIS

Abstract

Pemantauan kondisi rotating machine sangat diperlukan untuk menjamin keselamatan operasi sekaligus untuk meningkatkan efisiensi operasi di PWR. Salah satu teknik pemantauan kondisi terbaik yang dewasa ini dipilih karena mudah, non-invasive dan murah dalam implementasinya adalah Motor Current Signature Analysis (MCSA). Namun sayangnya penelitian aplikasi teknik ini untuk perangkat keras yang compact, low cost, berkelas industri dan layak untuk aplikasi pembangkit daya bertenaga nuklir sangat terbatas. Penelitian ini bertujuan untuk mengembangkan metode pemantauan kondisi berbasis MCSA dengan perangkat keras berkelas industri yang kompak untuk pembangkit daya tenaga nuklir. Penelitian meliputi aspek pengembangan perangkat keras real-time berbasis FPGA-CompactRIO, pembuatan modul untuk penampil early warning, pengujian unjuk kerja algoritma perangkat kerasnya, analisis spektrum berbagai kerusakan komponen motor elektrik, serta pengujian kinerjanya dalam mendeteksi berbagai kerusakan. Sistem pemantauan mampu mengeksekusi dengan total waktu eksekusi berkisar 164 ms, berhasil mendeteksi spektrum frekuensi berbagai kerusakan di motor induksi seperti stator shorted turn berkisar 75%, rotor broken bar 95%, eccentricity 65%, dan mechanical misalignment 85%, termasuk gangguan catu daya voltage unbalance 100%. Berdasarkan unjuk kerja perangkatnya, sistem pemantauan kondisi rotating machine ini menjadi salah satu alternatif terbaik untuk sistem pemantauan berbagai perangkat pemantauan di reaktor nuklir.

Kata kunci : Pemantauan kondisi, rotating machine, Motor Current Signature Analysis (MCSA), Field Programmable Gate Array (FPGA)

Condition monitoring of rotating machine is essential to guarantee the safety operation as well as to improve the efficiency of nuclear power plants operations. One of the promising condition monitoring techniques which has been preferred currently since it is simple, non-invasive and inexpensive is Motor Stator Signature Analysis (MCSA). However, the investigation of the MCSA technique using a compact, low cost, and having industrial class hardware which is capable for nucear power plant applications has been limited. The research is aimed to develop condition monitoring method based on MCSA utilizing a compact industrial class for nuclear power plant. The investigation includes development of condition monitoring based on real-time FPGA-CompatRIO hardware, development of a custom built display module for early warning system, testing of the monitoring hardware, fault frequency analysis of electric motors including the performances of fault detections. The condition monitoring system is able to execute a fault detection task around 164 ms, to recognize accurately fault frequencies of stator shorted turn for about 75%, broken rotor bar around 95%, eccentricity 65%, mechanical misalignment 85%, including supply voltage unbalances 100%. The condition monitoring system based on its performance assessments could become a suitable alternative not only for rotating machines but also condition monitoring for other nuclear reactor components.

Keywords : Condition monitoring, rotating machine, Motor Current Signature Analysis (MCSA), Field Programmable Gate Array (FPGA)