Optimization of a 900 mm Tilting-Pad Journal Bearing in Large Steam Turbines by Advanced Modeling and Validation

Abstract

Modernization of steam turbine components can extend the life of a power plant, decrease maintenance costs, increase service intervals and improve operational flexibility. However, this can also lead to challenging demands for existing components such as bearings, e.g., due to increased rotor weights. Therefore, a careful design and evaluation process of bearings is of major importance. This paper describes the advanced modeling methods applied for the optimization of a novel 900 mm three-pad tilting pad journal bearing followed by validation results that showed a high bearing temperature sensitivity to the fresh oil supply temperature during operation. The bearing was especially developed to cope with increased rotor weights within the framework of low pressure steam turbine modernizations at two similar 1000 MW nuclear power plants. With a static bearing load of approximately 2.7 MN at a rotor speed of 1500 rpm, it represents one of the highest loaded applications for tilting pad journal bearings in turbomachinery worldwide. After identification of the reasons for the sensitivity, advanced modeling methods were applied to optimize the bearing. For this purpose, a more comprehensive bearing model was developed taking into account the direct lubrication at the leading edge of the pads and the thermo-mechanical pad deformation. For the latter, a co-simulation between the bearing computation code and structural mechanics software was performed. The results of the entire analyses indicated modifications of bearing and pad clearance, pad pivot position, circumferential and axial pad length as well as pad thickness. Furthermore, the oil distribution into the pads was optimized by modifying the orifices within the bearing. The optimized bearing was then implemented on both units and proved its excellent operational behavior at increased fresh oil supply temperatures of up to 55°C. In addition, inspections during scheduled outages after 18 months of operation and subsequent restarts with reproducible bearing behavior confirmed the robustness of the optimized bearing. In conclusion, the application of advanced modeling methods proved to be the key success factor in the optimization of this bearing, which represents an optimal solution for large steam turbine and generator rotor train applications.