Abstract:
New Zealand is experiencing the need to meet net-zero GHG emissions and
growing energy demands. A key aim of the Ahuora research programme, that this
work is contributing to, is the development of tools such as Digital Twins to decarbonise New Zealand’s process heat sector and expand energy generation capabilities. The organic Rankine cycle (ORC) is a thermodynamic process capable of utilising low-temperature heat sources such as geothermal reservoirs and waste heat to
generate power.
This thesis investigates the dynamic modelling of vaporisers in ORC systems
to support the development and understanding of Digital Twins for New Zealand
industries. Its focus is the dynamic modelling of an industrial scale geothermal vaporiser where a model developed using Aspen HYSYS and a First Principles model
developed using Python were compared using various metrics. In terms of Digital
Twin model fidelity, the Python model and Aspen HYSYS model would be categorised as a 3-2-1D model in the behaves-like, looks-like and connects-to attributes.
In performance, good agreement between the temperature predictions was shown
for outlet geothermal brine, vapour and working fluid of the Aspen HYSYS model
with a percent error of 0.02%, 0.06% and 0.04%, respectively, for the validation case.
The Python model showed a percent error of 0.69%, 2.37% and 0.003%, respectively,
for the validation case. In comparing the ease of use of these software tools, a tradeoff would need to be made in deciding which modelling software to use, weighing
the monetary investment of using Aspen HYSYS and its high level of quality assurance against the time investment of using Python.
The First Principles model will be used to aid in development of a vaporiser
model in the Ahuora platform.