Wave energy converter power take-off characterization: comparing dynamometer and field data

Abstract

Through dynamometer testing of a wave energy converter’s (WEC’s) power take-off system (PTO), we characterize friction and establish a relationship between shaft speed and torque. This characterization is conducted through oscillation of the shaft with a velocity-controlled motor, yielding some key differences between lab tests and response of the system in the field. In this work, we develop an understanding of these differences, provide both linear and nonlinear representations of the PTO that can be used in hydrodynamic modeling, and assess how accurately these representations are reflected in data collected in the field. 

The WEC used in this testing is the TigerRAY, built by CPower as part of a WEC-UUV (uncrewed underwater vehicle) system funded by the Naval Facilities Engineering Systems Command. The full device is a two-body point absorber with a subsurface heave plate that acts to hydrodynamically stabilize the system and serves as a docking and charging station for the UUV. The system is undergoing its second year of development and testing, which includes both at-sea trials and shore-side dynamometer experiments at the University of Washington Applied Physics Lab. The dynamometer system records torque at the WEC drive shaft, as well as motor shaft speed. Data acquisition on board the WEC simultaneously records key generator information, such as current, voltage, and angular shaft position. 

Through comparison of lab and field data, we draw conclusions about the limits of a velocity-controlled dynamometer. Since this control scheme allows near-instantaneous changes in applied torque, it pushes the WEC through electrical system-induced torque spikes more quickly than is possible in the field. Reduction in the control gains of the dynamometer smooths the applied torque curve at the expense of added noise in the velocity profile. We weigh these tradeoffs using data obtained from field testing of the device in a real wave field. Additionally, we identify key characteristics in the dynamometer data and how they impact device performance in the field. Finally, we make recommendations about how to move forward with future dynamometer testing given these lessons learned.