This project set out to develop a link between Oceanographic computer models and Computational Fluid Dynamics (CFD) models in order to improve state of the art modeling techniques used for resource assessments and tidal turbine siting for both single and multiple TISECs.
The research was completed in two phases:
- CFD Modeling and flume tank experiments of single and multiple turbine facsimiles in a straight channel;
- A case study modeling single and multiple turbines in Minas Passage and Petit Passage.
Experiments were completed in the flume tank at the University of Victoria to measure thrust and study downstream wake dissipation for turbine arrays. Porous disc mounted on a force-measurement rig in the flume tank were used to represent the turbines. Particle image velocimetry (PIV) measurement equipment was used to visualize and quantify the wake structures behind the disks. The PIV system provided very rich flow-field data for a variety of array configurations clearly showing disc interactions and wake structure. Data from the experimental results was subsequently used to validate the turbine thrust and wake field computed using the CFD and Ocean models.
It was found that the CFD simulations did a reasonable job in predicting the thrust force acting on porous discs in several different array configurations. For all cases considered as part of this project, the thrust was predicted within 8% error of experimental results. The CFD simulations also did an adequate job predicting the wake recovery behind porous discs; however, significant tuning of turbulence parameters was required to get a good match to experimental data. The fact that thrust forces for each of the turbines can be predicted with reasonable accuracy and the wake can be tuned will allow site developers to use this simplified method for planning the initial layout of turbine arrays.
A team workshop was hosted by the University of Victoria on July 10th, 2012 to brainstorm CFD – Ocean model coupling approaches. Two cross-coupling methods were identified at the workshop:
- Mid Field CFD model: 100s of meters or even few km in size (see Figure 1)
- Near Field CFD model: 10s of meters, only spanning one or few Ocean model elements around the turbine(s) (see Figure 2)
One of the key objectives of this project was to apply these cross coupling methods to tidal sites in Nova Scotia. At the project outset, it was anticipated that all of the work would be applied to modeling the flows through Minas Passage with inclusion of 4 turbines at the test berths. As the project progressed, and the two different cross coupling approaches were identified, the team decided to include Petit Passage as a case study for testing the mid-field CFD approach while Minas Passage was used for testing the near-field approach.
The near field-modeling approach was tested on the Minas Passage site by modeling four 16m diameter turbines, one at each of the FORCE test berths. Each of the four turbines was first modeled in CFD (as a porous disk) in a 200m x 200m area surrounding each berth with inclusion of detailed bathymetry. Inflow and turbulence conditions were sourced from the Ocean model for the peak flood. To simplify the analysis only the M2 tidal component was used to drive the system. At peak flood (nominally U=2.5m/s) the total estimated power production in the Ocean model was 5MW.
The near-field coupling method work shows great promise. The objective was to ensure consistency between the Ocean and CFD models and in large part this was achieved. This methodology has a range of potential applications including:
- estimation of total extractable power from a tidal system;
- informing wide tidal site selection,
- array layout and channel build-out;
- investigation of the impact of a tidal installation on current patterns and tidal range;
- Investigation of the impacts of tidal installations on one another, etc.
With this methodology in place, regulators, developers and other stakeholders in the tidal industry can virtually investigate any number of ‘what if’ scenarios for the installations of free stream turbines before ever driving a pile or laying cable.
The mid-field cross-coupling method was also successfully implemented and demonstrated for Petit Passage. The flow through Petit Passage was modeled for both ebb and flood conditions (6 h each). In general results from the CFD simulations showed good agreement with Ocean model data, especially where the flow is relatively uni-directional and not dominated by large eddies.
A methodology was also proposed and subsequently demonstrated for how best to use the mid-field modeling approach to identify suitable turbine deployment locations. A deployment location along the north eastern shore was identified as a possible turbine deployment site. A simulation was subsequently run with the inclusion of a turbine. This simulation demonstrated the potential for using CFD to calculate power extracted by the turbine over a tidal cycle as well as modeling the wake generated by the turbine. This methodology could therefore be extended to modeling tidal farm arrays with inclusion of turbine interaction effects.
At this point, the CFD model of Petit Passage is still considered preliminary because it has not been validated against field data. While ADCP measurements have been completed at several locations in Petit Passage, the data is not yet available for public release.
The team therefore succeeded in meeting initial project objectives by testing two separate methods of coupling CFD and Ocean models. Detailed models of both Minas Passage and Petit Passage now exist that can be used as tools by project developers for laying out turbine arrays and technology developers to better understanding local inflow conditions.