This paper presents a study that has been conducted within a gas turbine development project aiming at an upgrade of an existing twin-shaft gas turbine. There are a number of uncertainties, both in the original turbine as well as in the introduced modifications. These uncertainties must be taken into consideration when setting performance guarantee margins for the turbine. In order to quantify the impact of those uncertainties on the turbine performance, a probabilistic risk assessment was performed. Uncertain parameters for the compressor turbine were defined and upper and lower bounds as well as probability distributions were estimated. The MINITAB software was used to determine an experimental plan resulting in test points that were investigated with aerodynamic and secondary air flow tools. Based on results from this analysis, a second-order metamodel was determined. The metamodel was fed with 20000 random parameter combinations according to the parameter probability distributions, and efficiency, swallowing capacity and cooling air flow probability distributions were calculated. In a second step, the resulting values were used as uncertain input parameters to the gas turbine performance analysis considering the entire gas turbine including compressor and combustor. An experimental plan was determined based on calculated bounds for the compressor turbine efficiency, swallowing capacity and cooling air flows. A metamodel was calculated, and again 20000 parameter combinations were randomly generated based on the found parameter distributions. The parameters were fed to the metamodel of the entire gas turbine, and probability distributions for the gas turbine overall power and efficiency were found. Results from the investigations were used to set guarantee margins for power and efficiency.

This content is only available via PDF.
You do not currently have access to this content.