The completion of the hydraulic air compressor (HAC) demonstrator at Dynamic Earth in Sudbury, Ontario marks the beginning of a series of research activities to increase the efficiency of compressed air production and build confidence in future commercial applications. Before any proper experiments could be conducted on the HAC Demonstrator a series of commissioning activities and testing was completed to i) calibrate the instruments, ii) check and understand losses, and iii) verify, or otherwise, some of the assumptions made during the system design.

The practical work associated with this master’s thesis included the development of a human machine interface (HMI) to allow for automated control of the HAC. Instrumentation and control equipment was installed and routed to a control panel providing conditioned power and routes for signals. Within the control panel, these are digitised and transmitted using TCP/ IP/ MODBUS protocol, operating over a TopServer (Software toolbox, 2009) OPC backbone. The OPC Client toolbox in MATLAB was adopted to interface with the OPC Server, and MATLAB’s App Designer adopted for authoring the HMI. All I/O functionality is thus routed to MATLAB in which a PID control loop was established between the HAC separator water level and the HAC’s compressed air motorized globe valve. Thus, a reliable, flexible, scientific control interface and data storage infrastructure was established for this novel compression plant as part of the master’s work. The HAC Demonstrator can now effectively run a variety of experiments while recording a wide range of data for analysis. To date, a series of 90 benchmark tests for compressor performance have been completed in a systematic manner on the demonstrator to create a database of real HAC operating conditions. This thesis thus represents the first formal publication of the HAC Demonstrator’s complete performance under the baseline operating conditions.

Previous predictions of the compressed air yield and efficiency of a HAC of this size have been made by Millar (2014), upgraded to weakly couple solubility loss by Pavese et al. (2016) and refined using Young’s (2017) detailed coupling of solubility and psychrometric phenomena. The predictions made by these models have been tested. The 1D hydrodynamic solubility models also predicted a small beneficial ‘airlift’ effect on compressor performance, due to exsolution of formerly dissolved compressed gas, that has also been reported upon.

One unexpectedly important factor that has been found to affect HAC performance that was not anticipated in any of the models included the absolute surface roughness of rubber lined pipe, in comparison to that of bare steel pipe. High precision experiments are reported upon that have produced reliable values for absolute surface roughness for rubber lining materials, that have now been adopted in the HAC models, and may be adopted more widely too. The occurrence of detrainment, water jet-free fall and air re-entrainment is speculated upon as the source of previously unreported loss in the air-water mixing process, based on pressure profiling observations undertaken over the complete performance envelope of the Dynamic Earth HAC Demonstrator.