WP3 – Power Take-Off (PTO)

In this task, UGent and Uppsala will investigate several control strategies for different fluid machines as defined in WP2. UGent will investigate how the multiple control systems on the machine side, e.g.,  single or double generator/motor control with Maximum Power Point Tracking (MPPT), blade pitching control (if chosen for the design turbine in WP2) and gate control can be used together to achieve the desired power reference tracking. In this, the whole operating range of the system will be taken into account, i.e., both pumping and turbine mode for different head heights and transitions from one operating point to another.

Here, the boundary conditions are determined that need to be taken into account in the development of the control system to limit the impact of control actions on the wear and fatigue loads. Together with dynamic limitations, this will lead to desirable limits on the ramping rates. This task is in collaboration with the grid control (WP6) and turbine (WP2) work packages.

In this task, several topologies are investigated with a focus on their technical feasibility, scalability, efficiency and flexibility towards the control system. Regarding flexibility to both the grid and the machine, it is crucial to investigate to which extent each architecture is capable of delivering which control action, e.g., field oriented control, reactive power control, desired control dynamics, harmonic mitigation, etc. For the counter-rotating double-rotor turbine concepts, a double generator with double AC/DC converter is envisaged, coupled on a shared DC-bus with a single inverter.

For the three different runner technologies UGent will study the most optimal drivetrain architecture given the system constrains such as mechanical integration, efficiency, economic cost, etc. This study includes both the selection criteria of suitable off-the-shelf powertrain components (electric motor/generator, gearbox, cooling system) as well as the design of custom made components e.g. direct drive electric motor/generator. This study will require scalable parametrized numerical models for both the off-the-shelf and custom-made drivetrain components.

Especially the powertrain design for a RIM-type runner is challenging in terms of geometry constraints. Here, a custom design based on the UGent direct-drive axial flux permanent magnet synchronous motor/generator technology can be studied. This motor/generator topology has an inherent ring structure and combines an excellent efficiency (even at a very low speeds <60rpm) with a lightweight construction.

Different powertrain topologies e.g. geared vs direct drive will be compared and optimized to achieve the required technical system performance parameters. This will result in a pareto optimal front that facilitates the selection of a powertrain architecture and will also help to understand the influence of key parameters e.g. the influence of the gear ratio vs. the dimensions of the motor/generator.

In the whole design process, the powertrain’s design is optimized for a scenario (“drive cycle”, “use cycle”) rather than for a nominal set-point, i.e., full cycle optimization. These scenarios are defined by the grid control, determined in WP5.

The availability of these parametrized and scalable models for powertrain components will result the proper design of the +/-50kW device under test, but will also lead to a virtual design of a 10MW PTO.

A full multi-dimensional model-based predictive control system is developed to achieve accurate tracking of the reference power, as provided by the grid-side control developed in WP5. The possible control dimensions are blade pitching (if chosen in WP2), gate control and the torque setpoint(s) of the single or double generator. In the case of a double generator, the rotational speeds of both machines will be considered as independent control variables, as a speed difference can have a positive impact on the performance of counter-rotating turbine systems. This impact will be investigated with CFD simulations in WP2.  Starting from the results of task 3.1, in this control system, the boundary conditions regarding wear, impact on fatigue loads and limitations on ramping rates of Task 3.2 are incorporated. The limitations of the power-electronic architecture (Task 3.3) and the PTO hardware (Task 3.4), are also incorporated. The control is developed based on two modelling techniques of the turbine rotor:

• Quasi-static modelling based on simulated or measured power-coefficient characteristic curves (e.g., power coefficient C_P, lift coefficient C_L, drag coefficient C_D, etc.)
• Blade-Element-Method (BEM) based modelling of the rotor. The BEM model is developed based on CFD simulations.

The results of the Tasks 3.3, 3.4 and 3.5 are validated on a +/-50kW dry test setup. For this, a power-electronic prototype is constructed based on Task 3.3. An AF-PMSM is constructed based on Task 3.4. The control concept of Task 3.5 is implemented. A set of realistic validation scenarios are defined as test cases to perform the experimental validation. On the test rig the turbine will be replaced with another electrical machine that will mimic the behaviour of the turbine (input from WP2).