Session: 20-05 Operation and testing in Oil and Gas Applications

Submission Number: 174620

Frame 5 Single-Shaft Gas Turbine Dynamic Model for Enhanced Plant Operator Training 

Digital transformation is progressively reshaping business models and industrial processes. One of the most concrete examples of this shift is the growing focus on advanced and accurate models to be used for different purposes, ranging from upfront support to product design and development to in-field verification of the expected product behavior by means of so-called digital twins. Another interesting application of dynamic models is in support of an Operator Training System (OTS), designed to train customer operators within a virtual environment that can accurately replicate the behavior of the real plant even before it is actually constructed or put into operation.

This study focuses on development of a dynamic model of a Frame 5 Single Shaft gas turbine, capable of operating under various boundary conditions, including ambient temperature variations, filter clogging, and changes in exhaust pressure. High model accuracy is also essential to support estimation of exhaust gas composition.

The model was integrated into an overall industrial plant model and is therefore tailored to the specific configuration of the real plant, which includes among others a steam turbine used as a starter, steam injection, and a recovery boiler connected to the exhaust duct.

Developed in MATLAB/Simulink, the model was statically validated against a NPSS based Cycle Deck operating points and dynamically validated using plant real data trends collected on site.

A model-based design approach was adopted to develop a static model of the gas turbine. Thermodynamic isentropic equations were applied to describe the compression, combustion, and expansion phases, ensuring a realistic representation of the gas turbine thermodynamic cycle. Considering an isentropic compression equation corrected for efficiency, is possible to relate pressure and temperature at axial compressor discharge, thus it is sufficient to have only one of them to compute the other one. From a thermodynamic point of view what happens during the combustion can be seen as an energy exchange between fuel, steam and the total flow through the combustion chamber, so a suitable equation weighting the different contributions was adopted. The power turbine behavior is similar to the axial compressor, the former expands the combustion gases, while the latter compresses the air; this thermodynamic transformation has been described using an isentropic expansion equation corrected for efficiency.  

The compression, combustion, and expansion coefficients are identified by solving a constrained optimization problem using Genetic Algorithms, with the aim of minimizing the deviations between the model and the reference benchmark (Cycle Deck).

Once the static thermodynamic model has been developed and validated it becomes possible to calculate the enthalpies and the overall power balance, which is essential for modeling the system’s dynamic behavior.

Thanks to its flexibility, the model can be integrated with various simulation frameworks. However, for simplicity and compatibility with existing platforms, it has been implemented directly inside the turbine control system PLC software in a closed-loop configuration and integrated into a commercial process simulation software (Honeywell Unisim) already in use for this purpose.

The paper will discuss in detail the features, implementation aspects, and added value provided by the model, which has been praised for its accuracy, versatility, and overall impact compared to the previously used solution for this application.

Presenting Author: Alessio Tumminello Nuovo Pignone SRL - Baker Hughes

Presenting Author Biography: I’m a Control Software Engineer at Nuovo Pignone – Baker Hughes, where I focus on gas turbine control software and dynamic simulations. I hold a Master’s degree in Automation and Robotics Engineering and have more than two years of experience in the oil and gas industry, working on control strategies and modeling for gas turbine applications.

Authors:

Alessio Tumminello Nuovo Pignone SRL - Baker Hughes
Tiziano Roma Nuovo Pignone SRL - Baker Hughes
Lorenzo Giovanardi Nuovo Pignone SRL - Baker Hughes