PREDICTIVE MAINTENANCE OF STEAM TURBINE ENGINES IN PALM OIL PROCESSING WITH THE DIGITAL TWIN CONCEPT APPROACH

• Main Author: Uwaisy Marchiningrum, Anranur
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/77439
• Institution: Institut Teknologi Bandung

When industrial equipment in the palm oil processing process is damaged, the problem that often occurs is the forced down time of the machine which results in a cessation of processing time and losses in palm oil processing. The turbine engine is one of the machines that often experiences damage to the point where it has a big impact with the cessation of the palm oil processing process. This study analyzes the damage factors that have a high criticality level in steam turbine engines, especially on the rotor wheel blades of the machine, and obtains indicators that affect the damage. Based on the results of the analysis obtained five indicators that affect the unbalance of the rotor wheel blades of steam turbine engines. A predictive maintenance mechanism is designed that focuses on five main indicators. Meanwhile, to complement the occurrence of lost data which reaches around 29 hours per week, the Digital Twin approach is used, namely using the LSTM forecasting model. The measurement results of the LSTM forecasting model using MSE are that the used steam temperature has an MSE value of 0.000479, 0.0034 oil pressure, 0.0037 cooling water temperature, 0.029 inlet steam pressure, and 0.011 used steam pressure. While the measurement results of the predictive maintenance model using the decision tree obtained a precision value of 1.00, a recall of 1.00 and an f-1 score of 1.00 and a running time of 0.07s. Based on the evaluation results, the entire system is represented in the form of graphs and tables to see the rise and fall of the values of the five indicators that affect damage to the rotor wheel blades of steam turbine engines which can be used as a decision support system to overcome machine forced down time and reduce processing system losses Palm oil.