Modeling of Warm-Keeping Process with Hot Air in Steam Turbines

Abstract

Steam turbines in conventional power plants have to deal with an increasing number of start-ups due to the high share of fluctuating power input in renewable power generation. As a result, the development of new methods for flexibility improvements—such as reductions in start-up time and the effects these start-ups have on turbine lifetime—have become increasingly important. In pursuit of this objective, General Electric has developed a concept for both the pre-warming and warm-keeping of a high-pressure (HP) / intermediate-pressure (IP) steam turbine with hot air: hot air is passed through the turbine while the turbine is rotated by the turning engine. Due to the high impact of transient flow phenomena on heat transfer during turbine warm-keeping operations, the reliable modeling of the time-dependent temperature distribution within thick-walled components is required as a tool for the optimization of these operations. Due to the extremely high computational effort required for conventional transient Conjugate Heat Transfer (CHT) simulations, alternative fast calculation approaches must be developed. The applied methodology for modeling warm-keeping turbine operations with hot air is presented in this paper.Furthermore, the key modeling steps have been analyzed. A fast transient CHT simulation approach called the Equalized Timescales (ET) method was developed to investigate heat transfer in the fluid and blades. Moreover, the setup of ET simulations was optimized with regard to accuracy and computing time. As a result, several operating points characterizing the turbine warm-keeping operational range were calculated for a single stage model. A sensitivity analysis regarding the heat transfer between fluids and solids was conducted to identify the most relevant surfaces. The ET method was then expanded to a numerical 3-stage turbine model in order to determine a HTC characteristic map for heat transfer in warm-keeping operations. This enables fast calculation of heat transfer rates and, consequently, computationally efficient determination of temperature distribution in warm-kept steam turbines. For comparison, the distribution of HTC was additionally calculated for one operating point of a 5-stage turbine model. Finally, the contact heat transfer in blade roots, which is believed to have a high impact on the temperature distribution of the rotor, was experimentally assessed in a test rig. The description of thetest rig and the methodology of determination of the thermal contact resistance (TCR), as well as the impact of TCR on the temperature distribution in the rotor are presented.