Abstract

This contribution reports on the development of Performance Digital Twin for industrial Small Gas Turbines. The objective of this study was the development of automation systems with control and monitoring functionalities, capable of addressing the requirements of future gas turbine plants for increased availability and reliability by use of Digital Twin technology.
The project explored development of Performance Digital Twin based on Real-Time Embedded computing, which can be leveraged with Internet-of-Things (IOT) Cloud Platforms. The proposed solution was provided in a form of modular software for a range of hardware platforms, with corresponding functionalities to support advanced control, monitoring, tracking and diagnostics strategies.
The developed Digital Twin was designed to be used in off-line mode to assist the software commissioning process and in on-line mode to enable early detection of degradation and fault modes typical for gas path components. The Performance Digital Twin is based on a dynamic gas turbine model which was augmented with a Kalman tuner to enable performance tracking of physical assets.
To support heterogeneity of gas turbine Distributed Control Systems (DCS), this project explored deployment of Digital Twin on multiple platforms. In the paper, we discuss model-based design techniques and tools specific for continuous, discrete and hybrid systems.
The hybrid solution was deployed on PC-based platform and integrated with engine Distributed Control System in the field. Monitoring of gas turbine Performance Digital Twin functionalities has been established via Remote Monitoring System (STA-RMS). Assessment of deployed solution has been carried out and we present results from the field trial in this paper.
The discrete solution was deployed on a range of Programable Logical Controller (PLC) platforms and has been tested by integrating Digital Twin in virtual engine Distributed Control System network. The Performance Digital Twin was embedded in Single Master PLC and Master-Slave PLC configurations, and we present results from the system testing using virtual gas turbine assets. The IoT Platform MindSphere was integrated within virtual engine network, and in this contribution, we explore expansion of the developed system with Cloud based applications and services.