Hydrodynamic and Structural Performance of the Transverse Horizontal Axis Water Turbine

Prof. Guy Houlsby, University of Oxford

Research and Industry Workshop, University of Leicester6th May 2011

Outline

• Why renewable power?• Why tidal stream?• The THAWT concept• Betz limit can be exceeded• 1/20th Scale tests: hydrodynamics• CFD adds understanding of flow• Loads on blades

Increase of atmospheric CO2

300.00

320.00

340.00

360.00

380.00

400.00

1950 1960 1970 1980 1990 2000 2010

Ann

ual a

vera

ge a

tmos

pher

ic C

O2

(ppm

v)Data from Manua Loa Observatory, HawaiiSource: Scripps Institution of Oceanography

The "Keeling Curve"

May 2009 – first monthly average over 390 ppmv

Increase of atmospheric CO2

0.00

0.50

1.00

1.50

2.00

2.50

3.00

1950 1960 1970 1980 1990 2000 2010

Ann

ual i

ncre

ase

of a

tmos

pher

ic C

O2

(ppm

v/ye

ar)

Data from Manua Loa Observatory, HawaiiSource: Scripps Institution of Oceanography

Historical records of atmospheric CO2

Source: GRID-Arendal (United Nations Environment Programme)

0

20

40

60

80

100

120

140

160

1985 1990 1995 2000 2005 2010

Oil (10^6 tonnes / year)

Gas (10^9 m3 / year)

UK oil and gas production

Source: Office for National Statistics

Problem:• Climate change due to fossil fuel use• Diminishing supply of hydrocarbons

Solution:• Nuclear• Renewables

Why Tidal Stream Energy?

• Potential to supply at least 18TWh/yr (6%) of UK electricity requirements (source: Black & Veatch/Carbon Trust)

– Minimal environmental impact - unlike a barrage

– Power is highly predictable - unlike wind

• Resident UK-based expertise in marine engineering

• Export potential

Tidal Resource – UK Target Areas

Pentland Firth, 4m/s

Irish Sea, 2m/s

Bristol Channel, 2m/s

English Channel, 2m/s

Peak tidal flow

Source: DTI Atlas of UK Marine Renewable Energy Resources

NB: Tides are cyclic but totally predictable

Options for tidal stream power (1)• Axial flow turbines

(“underwater windmills”)– “Unducted”

» MCT (most developed)» TidalStream» Tidel» … at least 8 others

– “Ducted”» Lunar Energy» Open Hydro» … at least 8 others

– Fixing options:• Fixed foundation• Pivoted• Anchored

Options for tidal stream power (2)• Vertical axis turbines

– Blue Energy– Polo– … 4 other vertical axis devices

• Horizontal axis turbine– THAWT (Oxford development)– … one or two others?

• Oscillating devices– Stingray– Pulse Tidal– … other oscillating devices

• Weird variants– Tidal Sails– Atlantis “Aquanator”

Scalability of tidal devices

… stretch …

Scalability of tidal devices

THAWT

• Turn axis of Darrieus vertical axis wind turbine (VAWT) through 90 to lie horizontally across a tidal flow

• Stretch across the flow

THAWT ConceptTransverse Horizontal Axis Water Turbine

• Length limited only by stiffness of structure and width of channel• THAWT is scalable horizontally

Fluid Mechanics of Darrieus Cross-Flow Turbine

Driving mechanism:

Torque per blade = R (L sinα – D cos α)

A comparison: axial flow turbine v. THAWT

• Advantages:– fewer foundations– fewer (larger) generators– fewer moving parts

THAWTGeneric

Axial FlowTidal velocity (m/s) 2.5 2.5Depth (m) 20 20Rotor Dia. (m) 10 10Rotor length (m) 60Number of rotors 2 10Flow area intercepted sq. m) 1200 785Total length (m) 128.0 125.0Power Output (MW) 7.0 3.7Number of foundations 3 5Number of generators 1 10Number of primary seals 4 30Estimated manufacturing costs 60% 100%Estimated maintenance costs 40% 100%

Tests at Newcastle University at 1/20th scale• Tests on a single turbine bay

– 0.5m diameter– 0.875m length– up to 1m depth

• Power curves recorded using servo motor/generator control of turbine speed

• Performance in a range of realistic flow conditions explored

• Turbine optimisation explored

Froude number scaling of flow conditions• Choice between Froude number

or blade Reynolds number scaling

• Maximum power available to a device in an open channel flow only a function of Froude number and blockage ratio(G.T. Houlsby et al. (2008). Application of Linear Momentum Actuator Disc Theory to Open

Channel Flow, University of Oxford Internal report, OUEL 2296/08)

• Froude number scaling

• Lower Reynolds number flows result in poor hydrofoil performance and a conservative estimate of the full scale performance

Scale Model Full ScaleFlow depth 1m 20mFroude number range 0.09 ‐ 0.22 0.09 ‐ 0.22Flow velocity range (m/s) 0.3 – 0.7 1.3 ‐ 3.1

ChordubladeRe

ghu

Fr

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0 1 2 3 4 5 6

Kin

etic

effi

cien

cy

Tip speed ratio

BetzFr=0.09Fr=0.13Fr=0.18Fr=0.22

Effect of flow rate on Truss THAWT performance

3

21

AuP

k

Blade transient stall at low speeds

Reduced angle of attack and increased drag at high speeds

Peak power

What about the Betz limit?

321

2716

AuP

For wind turbine in free air, mass/momentum analysis shows that stream tube area increases and velocity decreases through the turbine.

Max power is the Betz limit

Betz limit does not apply to tidal flows

Mixing

41 2 5

ub4

h5

At1

At4

ub4

u5ut4

h

h1 = h

u1 = u

3

AT

X

hbhugP

Head Efficiency Mass flow rate Change in depth

Power extracted from turbine best represented by Head efficiency

Tidal flows are subcritical

Downstream Mixing and Efficiency

022

123

21 22

223

BFrChhBFrCFr

hh

hh TT

1

22

/12/11

hhhhFr

hgubhP

PPP

W

Depth change may become an extremely important measure in light of environmental considerations

CFD : Computational Fluid Dynamics

• Calculation of flow through turbine is difficult– Unsteady, three dimensional, free surface, wide

range of length scales• Simplified analysis to calculate:

– Power– Forces on blades

CFD : Computational Fluid Dynamics

25 -1.2

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

0 30 60 90 120 150 180 210 240 270 300 330 360

Equ

ival

ent l

ift c

oeff

icie

nt C

L

Angular position (degrees)

Phase 2 testing at Newcastle

Power curves

-0.2

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5

Pow

er C

oeff

icie

nt C

P

Tip Speed Ratio TSR

-0.2

0

0.2

0.4

0.6

0.8

0 1 2 3 4 5

Pow

er C

oeff

icie

nt, C

P

Tip Speed Ratio, TSR

Exceeding the Betz limit

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0.000 0.050 0.100 0.150 0.200

Pow

er c

oeff

icie

nt, C

P

Froude number, Fr

B = 0.48, ForwardB = 0.48, ReverseB = 0.5, COMSOLB = 0.6, ForwardB = 0.6, ReverseB = 0.625, COMSOL

Betz limit

Measurements of bending moment

-15

-10

-5

0

5

10

15

0 30 60 90 120 150 180 210 240 270 300 330 360

Ben

ding

mom

ent (

Nm

)

Angle (degrees)

End 1MidpointEnd 2

Measured loads on blades

-80

-60

-40

-20

0

20

40

60

80

0 30 60 90 120 150 180 210 240 270 300 330 360

Dis

trib

uted

load

(N/m

)

Angle (degrees)

-1.2

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

0 30 60 90 120 150 180 210 240 270 300 330 360E

quiv

alen

t lift

coe

ffic

ient

CL

Angular position (degrees)

Measured “Lift Coefficients”

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2.5

3

0.000 0.050 0.100 0.150 0.200

Lift

coe

ffic

ient

at p

eak,

CL

Froude number, Fr

B = 0.48, MaxB = 0.48, MinB = 0.5, COMSOL, MaxB = 0.5, COMSOL, MinB = 0.6, MaxB = 0.6, MinB = 0.625, COMSOL MaxB = 0.625, COMSOL, Min

221 uc

wCL

Performance in a reversed flow

-5

0

5

10

15

20

25

30

0 1 2 3 4 5

Pow

er (W

)

Tip Speed ratio

Standard

Reversed

Truss device tested in standard and reversed configurations at Fr = 0.18

Varying the fixed offset blade pitch

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

-6 -4 -2 0

Pow

er c

oeffi

cien

t, C

P

Pitch angle (degrees)

S = 0.25, Fr = 0.109S = 0.25, Fr = 0.143S = 0.25, Fr=0.172S = 0.25, Fr = 0.16, COMSOLS = 0.125, Fr = 0.109S = 0.125, Fr = 0.144

1/20th Scale Test Conclusions• Device is capable of exceeding the Betz

limit by utilising blockage effects

• THAWT has comparable performance to a parallel bladed equivalent

• Performance of the device may be improved using a fixed offset pitch

• Due to Reynolds number issues the results here are a conservative estimate of the full scale performance of the device

• Blade loading measurements have been made

100 MW from 1 km long array across Bristol Channel

• Linear array extracts potential energy by surface level drop• Greater power than from kinetic energy only• Exceeds “Betz” wind turbine theory limit• Bristol channel example:

• 10m diameter• 1km long array in 20m deep water • 2m/s flow • 0.5m total head drop• 53% efficiency (modified theory) would give 102 MW

THAWT in the Bristol Channel

1 GW installation?

Acknowledgements• Oxford Tidal Engineering Group:

– Staff• Richard Willden, Alistair Borthwick, Martin Oldfield, Malcolm McCulloch

– Students/RAs• Ross McAdam, Claudio Consul, Scott Draper, Yihui Shi, Esteban

Ferrer, Clarissa Belloni, Tom Adcock, Simon McIntosh, Sena Serhadlioglu, Conor Fleming, Justine Schluntz, Chris Vogel, Taka Nishino

– THAWT Commercial Team• Sean Westrope, Peter Dixon, Stuart Wilkinson, James Mallinson

• Sponsors:– EPSRC, Oxford University UCSF, Isis Innovation Ltd., RCUK, John Fell

OUP Fund, Rhodes Trust, Energy Technologies Institute, Royal Academy of Engineering, German Academic Exchange Service, Oxford Martin School, TSB, Kepler Energy Ltd.

Conclusions

• Tidal stream could provide at least 6% of UK electricity• Efficient and robust devices like THAWT need to be

developed• The available resource needs to be properly understood

38