Search

Liberalized electricity market conditions and concentrating solar power technologies call for increased power plant operational flexibility. Concerning the steam turbine component, one key aspect of its flexibility is the capability for fast starts. In current practice, turbine start-up limitations are set by consideration of thermal stress and low cycle fatigue. However, the pursuit of faster starts raises the question whether other thermal phenomena can become a limiting factor to the start-up process. Differential expansion is one of such thermal properties, especially since the design of axial clearances is not included as part of start-up schedule design and because its measurement during operation is often limited or not a possibility at all. The aim of this work is to understand differential expansion behavior with respect to transient operation and to quantify the effect that such operation would have in the design and operation of axial clearances. This was accomplished through the use of a validated thermo-mechanical model that was used to compare differential expansion behavior for different operating conditions of the machine. These comparisons showed that faster starts do not necessarily imply that wider axial clearances are needed, which means that the thermal flexibility of the studied turbine is not limited by differential expansion. However, for particular locations it was also obtained that axial rubbing can indeed become a limiting factor in direct relation to start-up operation. The resulting approach presented in this work serves to avoid over-conservative limitations in both design and operation concerning axial clearances.

In order to improve the startup flexibility of steam turbines, it becomes relevant to analyze their dynamic thermal behavior. In this work, the relative expansion between rotor and casing was studied during cold-start conditions. This is an important property to monitor during startup given that clearances between rotating and stationary components must be controlled in order to avoid rubbing. The investigation was performed using a turbine thermal simplified model from previous work by the authors. The first step during the investigation was to extend and refine the modeling tool in order to include thermomechanical properties. Then, the range of applicability of the model was validated by a twofold comparison with a higher order finite element (FE) numerical model and measured data of a cold start from an installed turbine. Finally, sensitivity studies were conducted with the aim of identifying the modeling assumptions that have the largest influence in capturing the correct thermal behavior of the turbine. It was found that the assumptions for the bearing oil and intercasing cavity temperatures have a large influence ranging between ±25% from the measured values. In addition, the sensitivity studies also involved increasing the initial temperature of the casing in order to reduce the peak of differential expansion. Improvements of up to 30% were accounted to this measure. The studies performed serve as a base toward further understanding the differential expansion during start and establishing future clearance control strategies during turbine transient operation.

Solar steam turbines are subject to high thermal stresses as a result of temperature gradients during transient operation, which occurs more frequently due to the variability of the solar resource. In order to increase the flexibility of the turbines while preserving lifting requirements, several operational modifications for maintaining turbine temperatures during offline periods are proposed and investigated. The modifications were implemented in a dynamic thermal turbine model and the potential improvements were quantified. The modifications studied included: increasing the gland steam pressure injected to the end-seals, increasing the back pressure and increasing the barring speed. These last two take advantage of the ventilation and friction work. The effects of the modifications were studied both individually as well as in different combinations. The temperatures obtained when applying the combined modifications were compared to regular turbine cool-down (CD) temperatures and showed significant improvements on the startup times of the turbine.