Dynamic performance of power generation systems for off-shore oil and gas platforms

On off-shore oil and gas platforms two or more gas turbines typically support the electrical demand on site by operating as a stand-alone (island) power system. As reliability and availability are major concerns during operation, the dynamic performance of the power generation system becomes a crucial aspect for stable operation and prevention of unwanted shut down in case of disturbances in the local grid.

This paper aims at developing and validating a dynamic model of the gas turbine-based power generation system installed on the Draugen off-shore oil and gas platform (located in the North Sea, Norway). The dynamic model of the SGT-500 gas turbine includes dynamic equations for the combustion chamber and for the high pressure, low pressure and turbine shafts.

The low and high pressure compressors are modeled by using quasi steady-state conditions by scaling the maps of axial compressors employing a similar design point. For the turbines, the Stodola equation as well as a correlation relating the isentropic efficiency and the non-dimensional flow coefficient is utilized.

The model is implemented in the Modelica language. The dynamic model of a single SGT-500 gas turbine is first verified by comparing the transient response for a given load variation with the results of a non-physical Matlab model developed by the gas turbine manufacturer and adapted to the power set-point of the original engine installed on Draugen.

Subsequently, the complete power generation system consisting of three gas turbines is simulated during transient operation and the results are compared with operational data provided by the platform operator. The model is also applied to evaluate the transient response of the system during peak loads.

The results suggest that the highest accuracy (average relative error ~1%) arises on the prediction of the rotational speed of the high pressure shaft, while the largest deviation (average relative error ~20%) occurs in the evaluation of the pressure at the outlet of the low pressure turbine. As waste heat recovery units (e.g. organic Rankine cycles) are likely to be implemented in future off-shore platforms, the proposed model may serve in the design phase for a preliminary assessment of the dynamic response of the power generation system and to evaluate if requirements such as minimum and maximum frequency during transient operation and the recovery time are satisfied.

Furthermore, as the model is based on physics it can be coupled with the measuring instruments to monitor the thermodynamic variables at the inlet and at the outlet of each engine component.