Model Based Diagnostics of AE-T100 Micro Humid Air Turbine Cycle

Abstract

Micro gas turbines (mGT) are emerging power sources for distributed generation facilities with promising features like environment friendliness, high fuel flexibility, cost effectiveness and efficient cogeneration of heat and power (CHP). However, curtailed heat demand during summers reduces the plant operating hours per year and negatively affects the overall economic feasibility of a CHP project. The micro Humid Air Turbine (mHAT) cycle is one of the novel cycles to increase the electrical efficiency of the gas turbine by utilizing the exhaust gas heat in periods of low heat demand, thus avoiding the system shutdown. However, the water injection system can introduce additional pressure losses in the mGT cycle, which may lead to compressor surge and it may also affect the recuperator performance in the long run due to corrosion. Hence, numerical simulation and diagnostic tools are essential for cycle optimization of mHAT and prediction of performance degradation. This work is focused on the real time application of the AE-T100 model for the mHAT system located at the Vrije Universiteit Brussel (VUB), which is based on the T100 mGT equipped with a spray saturation tower. The AE-T100 model is a steady-state simulation tool for mGT cycles, which has been developed within a collaboration between the University of Genova (Unige) and Ansaldo Energia, and has been successfully applied at the Ansaldo Enegia test rig (AE-T100) for the diagnostic purpose. For this study, the basic AE-T100 model has been modified to simulate the humidified cycle according to the VUB plant configuration. The modified AE-T100 model has been validated against the experimental data obtained from the mHAT unit at nominal and part load. Once the model was validated using real operating conditions, it has been used for monitoring the recuperator performance over large number of tests in dry mode, conducted over the past five years, as well as initial tests in wet mode, from the VUB-mHAT system. This work has proved the modeling capability of the AE-T100 tool to simulate the mHAT cycle with reasonable accuracy and first diagnostic application of the AE-T100 tool, in dry mode. However, the lack of data available at present in wet mode does not allow to provide a complete and robust diagnostics of this novel cycle under wet operation. Hence, this preliminary analysis will provide basis for more detail diagnostics of the mHAT cycle using AE-T100 tool, over a longer time period under wet operation, in future.