From Test Rig to Digital Twin: First Dynamic Model of TUM's Rev. ORC Published

News & events

From Test Rig to Digital Twin: First Dynamic Model of TUM’s Rev. ORC Published

The FlexGeo project has reached its next milestone with a new preprint presenting the first dynamic model of the reversible high-temperature heat pump (HTHP) and Organic Rankine Cycle (ORC) test rig, experimentally validated in both operating modes. Building on the experimental data set published last year, this work moves the project from understanding the rig at steady state towards describing its behaviour over time, a key step for automated mode switching, advanced control and, ultimately, a digital twin of the system.

The model was developed in the process simulation software APROS and coupled with dedicated Python sub-models for the reversible twin-screw machine, including a polynomial expander model with empirical corrections, a neural network for expander power prediction and an empirical pump correlation. Key findings include:

ORC Mode: The net electric thermal efficiency is reproduced with a mean absolute error of only 0.27 to 0.32 percentage points. During load changes, pump power, expander power and expander speed follow the measurements with correct timing and amplitude
Heat Pump Mode: The control-relevant district heating supply temperature is predicted within 1 K, and setpoint changes are tracked with deviations of only 0.13 K to 0.20 K. The mechanical COP is reproduced with a mean absolute error of 0.31 to 0.44
Real-Time Capability: In both modes the simulation runs faster than real time, a prerequisite for future bidirectional coupling with the physical test rig
Oil Management: Residual lubricating oil in the refrigerant circuit was identified as the main remaining challenge, causing a systematic overprediction of the refrigerant mass flow. This once again highlights oil handling as a central topic for reversible HTHP/ORC systems

The full preprint is available on the Elsevier SSRN server. You can access it here:

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7434335

With a validated dynamic model in place, the next steps focus on refining the heat pump mode through a deeper characterisation of the oil circulation and oil separator, and on designing an automated mode-switching control strategy that can subsequently be implemented on the test rig. Stay tuned for further updates as FlexGeo continues to bring reversible HTHP/ORC systems closer to market.