Simulation of Gas Channel Temperatures during Transients for SGT-800

Abstract: Siemens Industrial Turbomachinery has developed dynamic gas turbine models to compute performance parameters such as temperature, pressure, mass flow and power during transients. Results from the current dynamic model correspond to measured data except for gas channel temperatures that change too fast during transients. The measurement of gas channel temperature is also a source of error due to the fact that probe temperature is measured instead of desired real temperature in the gas channel. It is of great interest to be able to simulate both real and probe temperatures in the gas channel to ensure that customer and development projects receive correct data. The objective of this thesis is to implement heat soak to the existing dynamic gas turbine model of Siemens SGT-800. Probe models that compute temperature in probes at all cross sections along the gas channel were implemented to the existing gas turbine model. A general method to calculate heat soak in gas turbines was discovered from literature and its reliability was proven by in-house reports. The method to calculated heat soak was implemented to SITs existing dynamic gas turbine model of Siemens SGT-800. When the heat soak model had been applied to the existing dynamic gas turbine model in Dymola and it was running in a correct way, the model was compared to measured data to ensure its agreement. To tune in the difference in gas channel temperature between measured data and the gas turbine model, parameters such as thermal mass, area and heat transfer coefficient were changed in the heat soak model. To emulate the inertia in the probe that occurs due to the metal encapsulation of the thermocouple, a transfer function was implemented. The probe inertia was tuned in towards measure data. Implementing heat soak and probe inertia to gas channel temperatures in the gas turbine model resulted in good agreement to measured data during start operations. During stop and trip operations the gas channel temperature reacted too fast in the gas turbine model and further work is recommended. An additional thermal mass which is affected by a lower temperature might contribute to integrating inertia to the gas channel temperature. The mass flow during barring speed is different for measured data and the dynamic gas turbine model during stop and trip operations. Mass flows during barring speed and its effect on gas channel temperatures need further investigation.