Multi-Rotor Turbines

The idea of large-scale multi-rotor turbines with two or more rotors atop a single support structure has inspired inventors since the 1920s. Current renewed interest in multi-rotor technology has a strong focus within large-scale offshore. Innovative product announcements and developments up until 2017 cover especially horizontal-axis turbines for bottom-fixed foundations as well as floating wind foundations, but no scaled and/or full scale prototypes yet.

The offshore wind leap into to the 12 – 15MW+ class is
ongoing with GE record holder announcing a 12MW giant with 220-metre rotor. A possible
alternative route in the new superclass could become multi-rotor turbines with
fresh opportunities including reduced rotor and head mass, faster scaling pace and
prolonged use of existing supply chains. Shorter time-to-market due to the use
of proven commercial turbine technology is another potential benefit, but the
road towards commercialization will not be easy.

Perhaps the world’s most famous multi-rotor turbine
designer was German engineer-inventor Hermann Honnef, who in 1931 presented a
concept with a 250-metre high tower and three turbines with 160-metre rotors.
However, his design and other typically skilful designs by fellow
engineer-inventors during the pre-WWII period, were never realised.

A key driver behind renewed multi-rotor turbine interest
is continuing turbine size increasement raising challenges to circumvent
negative consequences linked to the infamous Square Cube Law. This scaling law
dictates that for any turbine (installation) growing bigger, power output
scales with rotor diameter squared (P ~ D²), but mass increases with
diameter cubed (m ~ D³). Scaling turbines, by assuming unchanged
specific power rating (W/m²) and retaining main technology principles,
will make installations inherently heavier (Textbox 1).

Main contributing factors for the limited past
multi-rotor turbine successes are both technical and non-technical. A major
technological challenge is their complex dynamic behaviour. Rotor and head mass
are, for instance, not concentrated, but distributed over a collective plane. Early
multi-rotor pioneers further lacked necessary technical and financial means losing-out
in the competition with established suppliers continuously introducing next-generation
larger products.

Multi-rotor turbines inherently contain more components and (sub-)systems, but these are also typically smaller. Furthermore, higher quantities failure-prone critical components like bearings increases failure risk, putting additional pressure on ‘design for reliability’ issues. Re-utilizing validated components of single-rotor turbines in multi-rotor systems on the other hand promotes optimal use of existing supply chains, offering reduced risk-profile and prospects for lower LCOE.

The largest
multi-rotor turbine ever built was initially rated at 450kW. It incorporated six
75kW two-bladed variable-speed Lagerwey turbines with 15-metre rotor diameter
each, passive pitch control and flexible blade mountings. The three turbine levels each had two rotors, rigidly
attached to the tower and were complimented by a yaw system at ground level. The
installation commenced during 1988 at an industrial site in the port of
Rotterdam (the Netherlands), but vibration issues soon forced the bottom rotors
removal. The de-rated 300kW Quadro operated successfully for many years.

Early this century, Dutch company MultiWind BV conducted
a feasibility study for a 6MW offshore-dedicated turbine comprising three 2MW turbines
with 70-metre rotor, the biggest the industry could then provide. Other distinct
features of the patented MWT 6000 concept were 520W/m2 and three 120-degree interspaced
rotor arms attached to a central chassis with a (common) yaw system and a turn-able
rotor. The latter feature allowed turning individual rotors to the lowest
bottom position during installation and service. In full operating mode, one
rotor operated in top position with the other rotors left and right to the
tower. Following single rotor failure, the dysfunctional unit could be turned
in bottom position. This built-in redundancy would allow continued partial operation
at 66.7 per cent of total rated power. Despite generating a lot of wind
industry interest at the time, a prototype was never built.

A Vestas spokesperson during the 2016 introduction of its
innovative 900kW multi-rotor concept turbine said it could become ‘a blueprint for
larger-scale future products focused at specific markets’. The importance of
the Vestas initiative lies in the fact that it originates from one of the
world’s largest wind turbine suppliers with matching technological and
financial capabilities and strength.

The installation incorporates four variable-speed pitch-controlled
V29-225 kW turbines with 29-metre rotor, a popular model of the 1990s. Innovative features
include two vertical operational levels with at each level an independent yawing
chassis, an active yaw system and closely interspaced rotors in both vertical and
horizontal planes.

Two nacelles are mounted at each level to tubular-steel
arms and flexibly attached to a common turn-able chassis via steel tension
cables, a lightweight solution known from bridges. Flexible rotor mounting and elevated
individual yaw systems are two main differences with the 1988 Quadro rigid
design. The Vestas solution aimed at becoming a key enabler for substantial
dynamic loads reduction, reduced mass and system costs. The flexible systems design
approach was enabled by the availability of advanced modelling and design tools,
real-time and deterministic control strategies. The now completed one-year testing and validation
period focused at exploring the optimising potential and especially the impact
and system performance from three perspectives: aerodynamically, structurally and
loads. Main challenges were turbine dynamics and control, and a major control
challenge whether it can be fast enough to ensure safe operation under extreme
conditions and in the event of single-rotor failure. An overall challenge was
at minimizing structural system costs considering that scaled-down components
in multi-rotor turbines offer size and mass-related benefits through simplified handling and transportation requirements.

For a recap on main project
outcomes, Peter Lindholst, Vice
President, Concept Development and Erik Carl Lehnskov Miranda, Director,
Mechanical, Loads & Control Technologies, said: “The reason for
installing the multi-rotor concept demonstrator turbine at Risø was to learn
fast. Considerable knowledge has been gained about controllability, aerodynamic
interaction between the rotors and our ability to simulate loads and dynamics
for multi-rotor turbines. We can now conclude that no surprises have arisen. In
other words, we are able operate the rotors in a predictable stable manner.
Verification of the load simulation tool is progressing according to schedule
and results show an excellent match between simulations and measurements.
Although rotor interaction is limited it seems to have slightly positive impact
on the power curve. Simulation tools had to be further developed for predicting
this positive effect. Measurements to investigate further are being conducted
during 2018.”

The Vestas multi-rotor layout creates multiple
possibilities for systems enhancement. One could be varying rotor speed between
left and right rotors at each level offering active yaw control support, potentially
reducing yaw system complexity, mass and cost.

The above mentioned control principle was introduced
by Germany’s aerodyn-engineering during mid-2017 for a fully integrated 15MW
floating concept incorporating twin two-bladed downwind turbines with 150-metre
rotor diameters (425W/m2) and
semi-submersible floater. This (Super Compact Drive) SCDnezzy2 concept also presents a radical vision on how large-scale
floating offshore power plants could look like in 2025 or earlier. A twin-rotor
concept was selected in ‘offering perhaps the best compromise solution.’

SCDnezzy2 rotors
are centre-to-centre interspaced at one full rotor diameter plus 2metres, providing
302-metre installation width. The
rotors counter-rotate to balance opposing Cariolis forces acting upon them, and
relative rotor blade positions during operation are 90 degrees offset. One
rotor is thus horizontal when the other is vertical, a measure aimed at
minimizing blade interactions causing tip-vortices related performance loss.

The Y-shaped floater incorporates a dual-mode
single-point catenary mooring and rotatable yawing system for the full
installation, eliminating turbine yaw systems. Individually controllable turbine
rotor speed provides active yaw support to the otherwise mainly passive yaw
system.

Scaling up a single-rotor SCDnezzy 7.5MW with unchanged specific power rating to 15MW, would according aerodyn have increased head mass by a factor 2.6 (instead of 2.0 with SCDnezzy2). Applying again 425W/m2 gives 212-metre rotor diameter, and would result in around 30-metre higher centre of gravity for the single-rotor variant with unchanged wave clearance. The combination of SCDnezzy2’s reduced head mass, lower centre of gravity, and additional benefits limits floater cost increment to 25 – 30 per cent compared to a reference SCDnezzy 7.5MW according aerodyn.

Emeritus Prof. Friedrich Klinger of Germany’s Saarbrücken University of Applied Sciences, multi-rotor pioneer and leader of the INNOWIND wind research group in 2012 completed a feasibility study of a 15MW lightweight five-rotor concept. It comprises a three-legged lattice-type tower with the two left-and-right horizontal turbine levels and one central upper turbine. Each turbine is rigidly mounted to the tower and a collective yaw system is at the tower base, eliminating individual turbine yaw systems. The donor turbines are 3MW direct drive Siemens Gamesa SWT-3.0-101 units with 101-metre rotor diameter (375W/m2). Klinger and his team conducted total mass comparisons for two 15MW options, a scaled-up lightweight single-rotor SWT-3.0-101 and a multi-rotor turbine with five rotors, Table 1.

Multi-rotor turbines can finally benefit from a ‘multiplier’ effect during scaling. Fitting a three-rotor system with 7MW single-rotor turbines would already produce 21MW, and a four-rotor system with the latest 9MW+ units a striking 36MW+! Only time will tell whether such grand visions will become reality, with many practical questions left like on viable fast-track development paths for maximizing LCOE benefit and simultaneously limit developer risk perception and real risk.

Scaling-related impact is hidden by continues improvements in blade and turbine technology. This allows minimizing nacelle and rotor mass increment thanks to the availability of powerful computers, advanced design methods and the latest software-modelling tools. In addition, common engineering principles always promote applying optimizing measures like switching from solid and semi-solid shapes to tubular and slotted components shapes and other ‘open’ structures whenever possible and feasible. Advances in wind power science and technology equally benefit large-scale single-rotor and multi-rotor concept further development, but with different maximum gains in multiple specific areas.

The EU supported technology development and demonstration project INFLOW’s floating concept evolved into an unusual 5MW TWINFLOAT concept with two narrow interspaced two-bladed contra-rotating vertical-axis Darrieus rotors. Such configuration under controlled conditions offers enhanced aerodynamic performance due to increased air flow rate through the rotors known as ‘coupled-vortex effect.’ Keeping TWINFLOAT’s coupled rotor plane always stable and ‘perfectly’ perpendicular to the prevailing wind directions under all operating conditions is a challenge but essential for optimal wind flow and performance. The project’s status is unknown.

This article was previously published in the Offshore WIND magazine, issue 2, 2018.