The Limits of Reduced Order Current Energy Converter Modeling

Abstract

Reduced-order models for mesoscale current energy converter (CEC) modeling allow for tractable computation times for investigations of array configurations on power performance and environmental effects to support design optimization. The CEC representation in these models take the form of actuator discs in codes such as SNL-Delft3D-CEC-FM treating the rotating CEC blades as momentum sinks. In the first-of-its-kind, whole-plant optimization software, DTOcean, the hydrodynamic modelling of CECs is reduced one step further by superimposing wake models based on normalized CFD simulations onto a set of pre-computed velocity fields, to provide power estimates. DTOcean is a new tool and the amount of verification and validation evidence gathered is presently limited. To gain additional confidence and industry buy-in to the software penetration, this study investigated a primary component of levelized cost of electricity (LCOE) calculation, annual energy production (AEP), through an analytic calculation of power using the results of an identical simulation in SNL-Delt3D-CEC-FM. Three configurations of an 8-turbine array are studied with DTOcean where two rows of 4-turbines are spaced (unstaggered) 5-, 10-, and 20-Diameters apart and the AEP was calculated; The energy calculation in SNL-Delft3D-CEC-FM were more computationally expensive for the mesoscale domain making the optimization of solely an arrays power production using the wake superposition method implemented DTOcean attractive. The codes however are complementary as SNL-Delft3D-CEC-FM simultaneously investigates environmental effects of varying array configurations while DTOcean considers all aspects of array costs through its lifetime to optimize LCOE from a whole-plant perspective.