Back to the programme printer.gif Print

Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'Aerodynamics and rotor design' taking place on Wednesday, 12 March 2014 at 09:00-10:30. The meet-the-authors will take place in the poster area.

Peter Jamieson University of Strathclyde, United Kingdom
Peter Jamieson (1) F P Michael Branney (1) George Sieros (2) Panagiotis Chaviaropoulos (2) Petros Chasapogiannis (3) Spyros Voutsinas (3)
(1) University of Strathclyde, Glasgow, United Kingdom (2) CRES, PIKERMI, Greece (3) NTUA, Athens, Greece

Printer friendly version: printer.gif Print

Presenter's biography

Biographies are supplied directly by presenters at EWEA 2014 and are published here unedited

Wind energy professional for over 30 years leading a section involved in design development of wind turbines for James Howden in the 1980's, PJ joined Garrad Hassan in 1991 and founded their (wind) technology evaluation department. He is inventor of a number of patents relating to wind turbine rotors, author of numerous papers and of the book "Innovation in Wind Turbine Design". Since 2009 he has been working part time as a technology advisor in the Wind Energy Centre for Doctoral Training of Strathclyde University researching, teaching and co-supervising PhD students.


The structural design and preliminary aerodynamic evaluation of a multi-rotor system as a solution for offshore systems of 20 MW or more unit capacity


The challenges of up-scaling the prevailing 3 bladed, pitch regulated, variable speed wind turbine design to 10 and 20 MW were explored In the EU UPWIND project. The present work is part of the FP7 Innwind.EU Project whose overall objective is to develop high performance innovative designs of beyond state-of-the-art, 10-20 MW, deep water, offshore wind turbines. A multi-rotor system [1,2] is proposed as a concept that will realise the very large unit capacities of 20 MW or more that are desirable to reduce relative costs of O&M, foundations and electrical interconnection in offshore projects.


The approach adopted involved the following steps;
1. Define the overall aim: To design a multi-rotor system at a sufficient level of engineering to allow credible estimates of energy production, component masses and lifetime costs for comparative Cost Of Energy evaluation
2. Identify most relevant roles for the participants:
a. University of Strathclyde (UoS)- Concept design, loads calculations and technical coordination
b. GL Garrad Hassan (GH) – Development of main tool for design loads prediction
c. Centre for Renewable Energy (CRES) – Structure and yaw system design
d. National Technical University of Athens (NTUA)- Aerodynamic modelling of closely spaced rotors and impact of support structure blockage
3. Define an outline concept design: Unlike the large single turbine concept, there is no real disadvantage in up-scaling the multi-rotor system to unit capacities of 20 MW or more. 20MW rating was chosen as enabling direct comparison with 2 DTU reference turbines [2] and also with the UPWIND 20 MW design. Salient parameters of the individual rotors and an overall layout (based on previous work) were defined by UoS as a starting point for the support structure design of CRES
4. Identify key issues: Many key issues were known from previous work (see references of Chapter within [1]). Among them are questions about;
a. the aerodynamic performance of a system of rotors at close lateral spacing
b. design tools to predict loads in turbulent wind
c. structure aerodynamic blockage as compared with “tower shadow” of a single large turbine
d. structure weight and cost
e. engineering of a yaw system.
5. Consider design tools necessary to address the key issues: Multi-rotor systems are presently being considered for tidal turbines and a version of GL Garrad Hassan’s software, Bladed, being developed for this purpose was adapted to deal with a 45 rotor multi-rotor array. Specifically for the flow, at NTUA vortex methods are being applied to be combined with actuator disk CFD modelling together with CRES.
6. Define a work programme including outputs relevant for subsequent comparative cost of energy evaluation: This has been done and work stages completed and in progress are reported here.

Main body of abstract

The multi-rotor system comprises many rotors on a single structure in turn on a single foundation (or floating platform). A concept design (Figure 1) has evolved with 45 rotors of 41m diameter each rated at 444 kW providing a total rated power of 20 MW. As there is no requirement to realise capacity in a single giant turbine, 20 MW is not a limit and greater unit capacity may well be optimum.
Each turbine is pitch controlled and independently operated with a direct drive generator based on a PMG design of Magnomatics. A comparatively high tip speed 90m/s was chosen to maximise advantages of the multi rotor set in overall light weight and low cost of rotors and drive trains compared to equivalent large turbines although turbine weight loading of the structure is minor and there is no essential need for this.

Loads prediction
GL GH developed an enhanced version of their Bladed software to model 45 rotors on a single structure. The model enables simulation of performance in a turbulent wind field, includes structure aerodynamic effects in terms of a “tower shadow” from each tubular member and can also include limited structural dynamics. The loads as determined by Bladed from a reduced set of load cases (UoS, CRES) and associated calculations (UoS) according to IEC standards were provided to CRES as time series of 6 load components at each wind turbine hub centre.
Figure 2 is a striking illustration how unbalanced loads on the giant cantilevered blades of the 20 MW (UPWIND) turbine can produce a huge overturning moment while the corresponding moments of the 45 rotors in the multi rotor system sum to a load that is negligible by comparison.

Structure Design (CRES)
In an iterative design of the tubular structure, CRES considered selections from 15 standard tube cross-sections, (IMAT in Figure 3) the diameter over thickness ratio of which has been externally optimized for lateral and local buckling resistance. Given the space-frame topology its members are then selected out of the 15 options so to withstand ultimate loading from a reduced set of IEC load cases comprising a) wind loading in the designing 50 year gust case over 360° wind direction in 1o intervals around the structure and b) extreme turbulence loading under IEC DLC 1.3, following [3]. The mass of the optimized structure design was found just below 3000 t, suggesting that the structure should not be at a cost that would penalize COE. The equivalent design stresses (in MPa) over the members of the structure are shown in Figure 4.

Electrical design (UoS)
Top level evaluation of electrical design options considered clustering and DC interconnection of the multi-rotor turbines (not considered advantageous), definition of a wiring scheme (Figure 5) considering mass/cost and impact of faults on system output. To avoid coherent shut down of rotors in a grid loss which would have led to designing higher load levels, small resistive dump loads of varying magnitude are in circuit with each turbine during shutdown.

Aerodynamic evaluation (NTUA)
The aerodynamic performance of a 7 rotor system has been assessed using NTUA’s free wake code GENUVP. The blades are modeled as lifting surfaces while the wake is approximated by freely moving vortex particles [4]. The response signals showed that oscillations are generated at the blade passing frequency. In Figure 6 contours of the instantaneous axial flow are shown for a wind speed of 7m/s. The open areas in between the rotors exhibit high acceleration resulting from the blocking of the flow.

Processing of the results has shown that with respect to the isolated rotor, the thrust and torque of the rotors increase (Figures 7 & 8). At rotor level the changes on the central rotor are higher, while at system level the overall increase is smaller as a result of averaging. The level of increase in power is substantial at low wind speeds (13% at 7m/s). It is attributed to the blocking of flow due to the close placement of rotors. The level of increase is probably overestimated by a potential flow solver. In order to cross check the results CFD simulations based on the actuator disk model have been scheduled.


The work is in progress and there are not final conclusions to be drawn although a number of clear results have already emerged. Fatigue load evaluation is still to be completed but considered unlikely to be problematic. The structure has many closely spaced frequencies but fatigue load ranges are small and excitation is rarely coherent.
a) Although each individual rotor of the multi rotor array is designed essentially in the usual way and subject to the usual loads commensurate with its size and rating which increase with wind turbulence, the loads on the structure arising from the rotors are generally reduced by increasing turbulence which randomises load inputs.
b) The structure is designed by aerodynamic loading on its own members in the Class 1 storm conditions albeit with DLC 1.3 giving rather similar load levels. It is likely that for multi-rotor systems above 20MW (assuming the unit rotor size does not change) rotor loading will not be designing support structures
c) It would appear that a structure design complaint with IEC Class 1 can be achieved within a total mass of 3000 t. This should not penalise COE and based on previous work [1], very large savings on cost of rotors and drive trains of the multi-rotor system as compared with equivalent large turbine(s) are expected
d) Aerodynamic evaluation of a 7 rotor array by NTUA suggests that there is not an adverse interaction on power performance at very close spacing (rotors 5% of diameter apart). Previous CFD and wind tunnel work also on a 7 rotor array [1] did not measure any net loss or gain. Present results of NTUA with a potential flow solver show some substantial gain in performance at low wind speeds and further checks will be done with CFD.

Learning objectives
1. Understand the consequences of placing rotors in close spacing on performance and loads using advanced tools
2. Assess the multi-rotor concept and its advantages at 20MW scale and even beyond, based on state of art engineering design tools.
3. Optimize the design of the multi-rotor support structure
4. Quantify the cost effectiveness of the concept which in comparison to single rotor configurations has an expensive support structure in the benefit of the rotor/nacelle CAPEX.

1. P Jamieson, Innovation in Wind Turbine Design, A John Wiley & Sons, Ltd., Publication, ISBN 978-0-470-69981-2, 2011
2. P Jamieson, M Branney (2012) “Multi Rotors; A Solution to 20 MW and beyond”, Energy Procedia Vol 24, p 52-59
3. P. Chaviaropoulos, "Development of a State-of-the-art Aeroelastic Simulator for Horizontal Axis Wind Turbines. Part 1: Structural Aspects", Journal Wind Engineering, Vol 20, No. 6, pp. 405-421, (1996).
4. S. G. Voutsinas (2006) “Vortex Methods in Aeronautics: How to make things work”, Int. Journal of Computational Fluid Dynamics, Vol 20, No 1