When the swept area of an array of current turbines occupies a substantial fraction of a channel’s cross-section, their unit performance is augmented. In moderate confinement, power coefficients for individual turbines may exceed the Betz-Lanchester-Zhukowsky limit, and, in practice, exceed unity by drawing on both kinetic and potential energy embodied in the flow. This effect has long been recognized in the experimental fluids community, though is often treated as an artifact to be controlled for. The potential benefits of confinement for current turbines has been recognized since Garrett and Cummins’ seminal work, but most of the subsequent studies on this topic have been theoretical or abstracted to ideal channel geometries. Contrary to the vertical exaggeration used to render tidal and river channel cross-sections, these channels are typically orders of magnitude wider than they are deep and, consequently, have a fundamentally rectangular aspect ratio. Because of this, achieving meaningful levels of confinement with axial-flow (“horizontal-axis”) turbines requires filling a square box with round pegs. While this is possible, the rectangular aspect ratio of cross-flow (“vertical-axis”) is more favorable.
As part of ARPA-E’s Submarine Hydrokinetic and Riverine Kilo-megawatt Systems (SHARKS) program, we are developing cross-flow turbines and control strategies that can exploit confinement to reduce the levelized cost of energy. An initial step in that process is understanding what level of confinement is practically achievable in tidal and river channels, subject to the constraints of (1) adopting a uniform module layout that minimizes the cost of site-specific customization, (2) minimizing the total number of modules to reduce installation, operations, and maintenance complexity, and (3) providing sufficient overhead, seabed, and inter-module clearance to permit marine animal and vessel traffic. We consider an array of turbines to consist of one or more stacks of vertically joined modules across a channel transect. Results of a parameter sweep for the SHARKS S1 site (Igiugig, AK - riverine) and S4 site (Angoon, AK – tidal current) indicate optimal layouts can achieve channel blockages of 30-40% using modules 2-4 m height and 40-50 m width. In conjunction with logistical considerations, this motivates an array design based on cross-flow turbine units that can be packaged in a standard shipping container and, on site, joined to form modules and module stacks. To maximize the swept rotor area within a module, creative approaches using co-design principles, are needed to balance the arrangement of rotor, power take-off, and support structure.
This presentation lays out the module optimization process, the overall architecture of a confinement-exploiting array of cross-flow turbines, and the engineering considerations for the individual turbines that make up the modules. In addition, potential approaches to array-level control in time-varying inflow velocity and water-depth are outlined.