Minesto says an upgraded power-take-off system fitted to its Dragon 4 tidal power plant has increased electricity generation across the tidal flow cycle at Vestmannasund in the Faroe Islands, producing record summer output during several months of comparative operation.
The 100 kW Dragon 4 operated by Minesto uses a tethered subsea kite that flies a controlled path through the tidal stream, driving a turbine and generator as the device moves faster than the surrounding water. Electricity is transferred through the tether and subsea connection to shore.
The latest change concerns the power-take-off, or PTO, which converts mechanical input from the turbine into electrical output. Minesto says measured tidal flows at Vestmannasund proved stronger than historical modelling had indicated, while the hydrodynamic performance of its second-generation Dragon platform allowed the machine to encounter more available energy than the original PTO could fully use.
Engineers responded by developing a more powerful PTO. The company’s 10 August data compares the same Dragon 4 unit, Saga, before and after the upgrade, with Minesto stating that the PTO is the only hardware difference between the two datasets. The comparison covers operation in the second and third quarters of 2026.
That is a useful test arrangement because marine-energy output is sensitive to the resource passing through the machine. Tidal cycles are predictable in timing, but flow velocity changes continuously, and the kinetic power available in moving water rises steeply with velocity. Comparing different devices or different sites can therefore obscure whether a production gain came from hardware or from better conditions.
Using the same kite at the same production site reduces that uncertainty, although it does not make the company’s performance claims independently verified. Minesto says the upgraded system increased energy production throughout the tidal flow cycle while also reducing system costs, but it has not published a single percentage improvement that would allow a simple before-and-after efficiency figure to be applied.
The production graph released with the announcement instead shows daily energy yield from the original and upgraded configurations against predicted tidal speed. According to Minesto, output from the revised system more closely tracks the available resource, allowing the generator to capture energy that the earlier PTO left unused when flows were stronger.
That distinction is important for tidal generation because maximum power rating tells only part of the commercial story. A device earns revenue from energy produced over time, not from the largest instantaneous number reached during a favourable current. Improving the useful production curve through more of each tidal cycle can therefore be more valuable than raising a single peak figure.
Dragon 4 is a microgrid-scale machine rather than a utility-scale turbine. Minesto lists the unit at 100 kW and approximately 2.5 tonnes, and has used it as part of the development path towards the larger Dragon 12 platform. The smaller machine provides a comparatively accessible way to test controls, hydrodynamics, power conversion and operating methods before those lessons are applied to larger hardware.
The kite architecture differs from fixed tidal-stream turbines mounted directly in high-flow channels. Minesto’s device is tethered to the seabed and follows a controlled underwater trajectory intended to increase the relative flow through its turbine. The company’s commercial proposition is that this can extract useful power from lower-flow tidal streams and ocean currents that are less suitable for conventional fixed machines.
That approach introduces its own mechanical and controls burden. The kite has to follow a repeatable flight path while current speed and direction change, and the tether must provide both mechanical restraint and an electrical route to shore. Mooring hardware, control software, turbine behaviour, generator loading and the PTO have to operate as one system in an environment where maintenance access is expensive.
Reliability therefore remains as important as energy conversion. Marine renewable projects have historically struggled with the cost of installation, retrieval and intervention because a relatively modest equipment fault can require vessels, suitable sea conditions and specialist crews before repairs can begin. A production gain is commercially useful only if it is not offset by more complex hardware or higher maintenance demand.
Minesto says the upgraded PTO also reduces system cost, although no detailed cost breakdown has been published. That claim will need to be judged alongside component life, retrieval frequency and the economics of manufacturing multiple machines. For array-scale projects, modest differences in hardware cost and maintenance time are multiplied across every device installed.
The Faroe Islands remain central to Minesto’s development work because predictable tidal generation could complement other renewable sources on an isolated power system. The company has used Vestmannasund to gather operating data and validate simulation models, with the latest PTO comparison adding another dataset to that engineering process.
Simulation accuracy matters when developers size equipment for a particular site. Underestimating the resource can leave conversion hardware unable to exploit available energy, as Minesto says happened with the original PTO; overestimating it can produce an unnecessarily expensive machine that rarely reaches its intended operating point. Better agreement between predicted and measured output reduces both risks.
The upgraded Dragon 4 does not settle the commercial case for tidal power. Arrays still have to combine reliable devices, subsea infrastructure, grid connections, maintenance logistics and finance at a cost that can compete with other forms of generation. It does, however, provide an example of the less glamorous work required after a marine-energy concept reaches the water: measure what the site actually delivers, find the component that is limiting output, change it, and run the machine long enough to see whether the improvement survives more than a single tide.
