Turbine choice and optimization for a shoreline OWC wave energy plant

A.F.O. Falcão, J.C.C. Henriques & L.M.C. Gato

IDMEC, LAETA, Instituto Superior Técnico, Universidade de Lisboa, 1049-001 Lisbon, Portugal

ABSTRACT: The aerodynamic performance of the air turbine plays a major role in the success of the oscillat-
ing-water-column (OWC) technology for wave energy conversion. A case study was selected to investigate
this issue: the existing bottom-standing plant on the shoreline of the island of Pico, in Azores Archipelago.
The overall performance of the OWC plant was modelled as an integrated hydrodynamic and aerodynamic
process. Two different types of self-rectifying air turbines were considered: the single- and two-stage Wells
turbine and the biradial turbine. Results from model testing of both turbine types, together with well-known
tools from dimensional analysis, were employed to determine the aerodynamic performance, and to optimize
the turbine size and rotational speed. This allowed a comparison to be made between the two types of turbine
for this kind of application. In general, the biradial turbine was found to outperform the Wells turbine, espe-
cially the single-stage one.

1 INTRODUCTION and the constraints that such equipment may intro-

duce into the overall system are important, but are
The OWC device is widely regarded as the simplest
not analysed here.
and most reliable type of wave energy converter, and
A case study was selected to investigate these is-
the one that was object of the most extensive devel-
sues: the existing bottom-standing OWC plant con-
opment effort, with the largest number of prototypes
structed on the shoreline of the island of Pico, in
deployed into the sea (Heath 2012; Falcão and Hen-
Azores Archipelago, northern Atlantic Ocean.
riques 2016). The OWC comprises a partly sub-
In the study presented here, theoretical modelling
merged (fixed or floating) structure, open below the
is used based on linear water wave theory for the hy-
water surface, inside which air is trapped above the
drodynamics of wave energy absorption.
water free surface. The oscillating motion of the in-
Two different types of self-rectifying air turbines
ternal free surface produced by the incident waves
are considered: the Wells turbine and the biradial
makes the air to flow through a turbine that drives an
turbine (Falcão and Henriques 2016). Results from
electrical generator. Self-rectifying air turbines have
model testing of both turbine types, together with
the advantage of not requiring rectifying valves. The
well-known tools from dimensional analysis, are
Wells turbine is the best known self-rectifying air
employed to determine the aerodynamic perfor-
turbine for wave energy applications, but there are
mance, and to optimize the turbine size and rotation-
others, mostly of impulse type (Setoguchi and Takao
al speed. This allows a comparison to be made be-
2006; Curran and Folley 2008; Falcão and Gato
tween the two types of turbine in this kind of
2012; Falcão and Henriques 2016).
application. The Wells turbine is known to be ap-
The success of an OWC converter depends large-
proximately a linear turbine. In the case of the bira-
ly on its wave-to-wire conversion efficiency. Here,
dial turbine, a linearization was introduced to allow
the aerodynamic performance of the air turbine plays
the frequency-domain analysis to be employed in
a major role. It should not be forgotten that there is a
regular waves, and a stochastic analysis to be used in
coupling between the hydrodynamic process of wave
irregular waves.
energy absorption and the aerodynamic process that
takes place in the turbine. More precisely, this cou-
pling is affected by the relationship between pressure 2 AIR CHAMBER THERMODYNAMICS
in the air chamber and air flow rate through the tur-
bine (the so-called damping effect). This relationship We consider an OWC converter with a bottom-fixed
in turn depends on turbine geometry, size and rota- structure. Let pat  p(t ) ( pat  atmospheric pres-
tional speed. For these reasons, the overall perfor- sure) be the pressure (assumed uniform) of the air
mance of the OWC plant should be modelled as an inside the chamber, q(t ) the volume flow rate dis-
integrated hydrodynamic and aerodynamic process. placed by the motion of the inside free surface of
The performance of the electrical power equipment
water (positive for upward motion), and m(t ) the flow rate (in fact bidirectional and largely random).
mass of air inside the chamber. We may write Hysteretic effects are known to occur (they are par-
m  chV , where ch (t ) and V (t ) are respectively ticularly significant in the Wells turbine, see Setogu-
chi et al. 2003); such effects are ignored here.
the density and volume of air inside the chamber.
If the effects of variations in Reynolds number
Then, taking into account that dV dt  q , we ob-
and Mach number are ignored (which is in general a
tain, for the mass flow rate through the turbine reasonable assumption), the turbine performance can
w   dm dt (positive for outward flow), be represented in dimensionless form as (Dixon and
d Hall 2014; Falcão and Henriques 2016)
w  V ch  ch q . (1)
dt   f  (), (7)
The discharge process ( p  0, w  0) , when air is   f  (), (8)
flowing out of the chamber to the atmosphere, may where  ,  ,  are dimensionless coefficients of
be regarded as approximately isentropic, i.e. the spe- flow rate, pressure head and power, respectively, de-
cific entropy of the air remaining in the chamber is fined as
unchanged (although in general different from the w
atmospheric air entropy). The filling process  , (9)
0D3
( p  0, w  0) is more difficult to model, since the
p
air specific entropy increases (from its atmospheric  , (10)
value) due to viscous losses in the turbine, valves 02 D 2
and connecting ducts; the air specific enthalpy also 
Pt

Lt
. (11)
changes due to work performed in the turbine. A  0 D
3 5
 0 2 D 5
polytropic relationship between the pressure p  pat Here, D is rotor diameter,  is rotational speed (in
and the density  ch in the air chamber was proposed radians per unit time), Pt  Lt is turbine power
by Falcão and Henriques (2014a) output, Lt is turbine torque, p is pressure head (as-
pat  p pat
 k , (2) sumed to be equal to the pressure oscillation in the
 chk
 at OWC chamber), and  0 and c0 are air density and
where  at is the density of atmospheric air and k is speed of sound at reference conditions (normally
the polytropic exponent that is related to the average turbine entrance stagnation conditions). Note that
efficiency  of the turbine. A good approximation functions f  and f  characterize a given turbine
is (Falcão and Henriques 2016) geometry, independently of turbine size, rotational
k  0.13 2  0.27  1. (3) speed or gas density.
From Eq. (2), we obtain
dch at dp 4 OWC HYDRODYNAMICS
 1k . (4)
dt kpat ( pat  p)( k 1) k dt
Assuming linear water wave theory to be applicable,
If the absolute value p of the pressure oscilla-
we may write, for the volume flow rate displaced by
tion is much smaller than the atmospheric pressure the motion of the inner free surface,
pat , Eq. (4) may be linearized as q(t )  qr (t )  qe (t ) . (12)
dch at dp Here, qr (t ) is the radiation flow rate due to the air
 (5)
dt kpat dt pressure oscillation p(t ) inside the chamber in the
and Eq. (1) becomes absence of incident waves, and qe (t ) is the excita-
 dp
w  V0 at  at q, (6) tion flow rate due to the incident waves if it were
kpat dt p  0.
where V0 is the volume of air in the chamber in the In the simple case of incident regular waves of
absence of waves. frequency  and amplitude Aw , we may write, after
the transients related to the initial conditions have

3 AIR TURBINE AERODYNAMICS  

died out, qe (t )  Re Qeei t , where Re() stands for
real part of, and Qe ( ) is an (in general complex)
Unlike the turbines in most other applications, the
excitation flow rate coefficient that is assumed
air turbine in an OWC is subject to a highly unsteady
known for the structure and surrounding wall ge- Here V0 is the volume V of the air chamber in calm
ometry and for incident wave amplitude and direc- water. The power absorbed from the waves is
tion. Pabs (t )  p(t )q(t ) . Its time-averaged value is
Real (unidirectional) irregular waves in a given
sea state may be represented with good approxima- Pabs  12 Re(QP*) (where the asterisk denotes com-
tion as a superposition of regular waves, by defining plex conjugate).
a one-dimensional variance density spectrum S ( )
4.2 Stochastic model for irregular waves
4.1 Frequency domain analysis We assume now that the local wave climate may be
We consider first the simple case of regular incident represented by a set of sea states, each being a sta-
waves of frequency  and unit amplitude Aw . The tionary stochastic ergodic process. In a sea state, the
surface elevation  at a given observation point is
air turbine is assumed linear at constant rotational
speed, i.e. we may write   , where  is a di- supposed Gaussian. We assume a one-dimensional
variance density spectrum S ( ) (Holthuijsen
mensionless constant, and assume the reference den-
sity  0 to be the atmospheric air density  at . It fol- 2007). Within the framework of linear water wave
theory and if, in addition, a linear turbine is as-
lows, from Eqs (9) and (10), that w  Kat p, where
sumed, the wave-to-pneumatic conversion process
K  Dat11 . Note that, unlike  , K is not inde- may also be regarded as linear. Consequently, the
pendent of turbine size or rotational speed. Equation pressure oscillation p(t ) in the chamber is also a
(6) becomes Gaussian process, and a pressure spectrum S p ( )
V dp may be defined. It is related to the wave spectrum
q 0  Kp . (13)
kpat dt S ( ) by (Falcão and Rodrigues 2002)
This relationship between the volume flow rate q 2
S p ( )  2 ( ) ( ) S () , (17)
displaced by the OWC free surface and the air pres-
sure oscillation p represents a linear power take-off where ( ) is given by Eq. (16), and the excitation
system, provided that the rotational speed  , and flow rate coefficient ( ) is equal to the value of
hence K, are independent of time. This requires a Qe ( ) for waves of unit amplitude. We denote by
sufficiently large rotational inertia of the rotating el-
ements.  p and  2p the standard deviation (or root mean
In this a case we may employ the frequency do- square) and the variance of the air pressure oscilla-
main analysis and write tions, respectively. It is
q(t ), qe (t ), qr (t ), p(t ) ReQ, Qe , Qr , Peit . (14) 
 2p   S p ( )d . (18)
0
Here Re() stands for real part of (a notation that
The probability density function f p ( p ) of the
will be omitted in what follows) and Q, Qe , Qr and
Gaussian pressure oscillation is given by
P are complex amplitudes. We further write
 p2 
Qr  (G  iB) P , where G and B are real coeffi- f p ( p) 
1
exp   2  . (19)
cients that depend on OWC geometry and on wave 2  p  2 p 
 
frequency  (but not on wave amplitude), and G is
The time-averaged power output of the turbine Pt
non-negative. We call G the radiation conductance
and B the radiation susceptance (Falnes 2002). These may be obtained by integration

hydrodynamic coefficients are assumed known as Pt   f p ( p ) Pt ( p )dp
functions of the frequency  for the structure and 

surrounding wall geometry, from theoretical or nu-  at  3 D 5   p2    (20)

 f  p 
merical modelling or from experiments. The linear-  
2  p  
exp
 2 2      2 D 2 
dp.
ized frequency-domain version of Eq. (13) becomes  p  at 
P  Qe , (15) Its dimensionless value is defined as
where Pt
 . (21)
  V 
1  at 3D 5
   K  G  i 0  B  . (16) The function f  ( ) in Eq. (20) is defined by Eq. (8).
  k pat 
The time-averaged power available to the turbine is
Pavai  K 2p . The average efficiency of the turbine is RELIEF VALVE

defined as   Pt Pavai . It is convenient to consider AIR TURBINE

 and  as functions of the dimensionless value of
the standard deviation of the pressure oscillation (see
Falcão and Rodrigues 2002)
p
  . (22)
 at  2 D 2 12 m

5 TEST CASE: THE PICO PLANT

Figure 2. Longitudinal cross-section of the plant.
5.1 Plant description
The OWC plant on the island of Pico (Falcão 2000) The concrete structure of the chamber (square
was chosen as a test case for the numerical simula- planform with inside dimensions of 12 m  12 m at
tions based on the theory presented above. The plant mean water level) was built in-situ on rocky bottom
is located on the northern shoreline of the island, at a (about 8 m water depth), spanning a small natural
site named Porto Cachorro (3832' N, 2832' W) . harbour (gully) (Fig. 1). The power plant is equipped
with a horizontal-axis Wells turbine-generator set

WAVES
N rated 400 kW. The plant is connected to the island’s
electrical grid.
The radiation conductance G, the radiation sus-
ceptance B and the excitation flow coefficient
  Qe Aw were computed, as functions of fre-
quency  , by a modified version of the code
AQUADYN, based on the boundary element method
PLANT (BEM) (Brito-Melo et al. 2001). In the definition of
 , Aw is the wave amplitude in deep water; the
dominant wave direction offshore was assumed. The
20 m hydrodynamic coefficients G, B and  are plotted
Figure 1. Porto Cachorro site, showing the location of the plant. versus wave frequency  in Fig. 3. The irregularity
of the curves reflects the very irregular configuration
The coastline at Porto Cachorro (Fig. 1) is very of the rocky bottom and side walls in the vicinity of
indented over a distance of a few hundred metres. It the plant. The volume of air in the chamber, in the
comprehends several small harbours and recesses absence of waves and under mid-tidal conditions, is
with almost vertical basaltic walls and water depths V0  1050 m3.
of about 6 to 9 metres. There is no sand. The exact 0.025
site where the plant was actually built was chosen 0.020
because of natural wave energy concentration (hot   5  105
0.015
spot) and also because of relatively easy access by
0.010
land. The absence of a continental shelf makes it
0.005 G
possible for the waves to reach the vicinity of the
shoreline with relatively little energy dissipation. 0.000
Detailed information on wave climate was available 0.005 B
for the chosen site from measurements offshore 0.010
(Waverider buoy in 100 m-deep water). The site is 0.0 0.5 1.0 1.5 2.0 2.5
partly shielded by the presence of the two neighbour-  (rad/s )
ing islands of Faial and São Jorge, which however Figure 3. Hydrodynamic coefficients G (radiation conduct-
does not substantially prevent the propagation of the ance), B (radiation susceptance) and   Qe Aw (excitation
waves from the predominant N-NW direction. The flow rate coefficient) for the plant
average tidal amplitude is about 0.5 m.
5.2 Wave climate guide vanes
Wave measurements were performed off the plant
site with a Waverider buoy. A simplified description
of the local wave climate was adopted here for the
calculations. It consists of a set of 9 sea states, each
defined by the significant wave height H s , the ener-
rotor
gy period Te and the frequency of occurrence  Inlet/outlet ducts
(Table 1), together with a Pierson-Moskowitz vari- Figure 4. Schematic representation of the biradial turbine with
ance density spectrum (Goda 2002) sliding guide vanes.

 
S ( )  262.6 H s2 Te 4 5 exp  1052 (Te  )  4 , (23) The chosen Wells turbine geometry was that of
the model, code named GV6, tested at Universität
where H s is the significant wave height and Te is Siegen, Germany (Starzmann 2012). This may be re-
the energy period, defined in the usual way. The en- garded as a highly efficient state-of-the-art Wells
ergy flux of the waves per unit crest length in deep turbine. The tested turbine had a rotor of diameter
water is D  0.4 m , hub-to-tip ratio 0.43, 5 blades of increas-
 g2 2 ing chord from hub to tip, and solidity at hub equal
Pwave  H s Te . (24)
64 to 0.67. The turbine had a row of guide vanes on
Taking   1025 kg/m 3 and g  9.8 m/s 2 , the annu- each side of the rotor. Dimensionless performance
curves of flow rate , power output  and aerody-
al-averaged wave power is
9 namic efficiency  versus pressure head  are giv-
Pwave   Pwave,i i  18.2 kW/m. (25) en in Figs 5 to 7. The average values  and  , in
i 1
random waves, of the dimensionless power output
Table 1. Energy period Te,i , significant wave height H s,i and
and of the efficiency are plotted in Figs 6 and 7 ver-
frequency of occurrence i of the nine sea states i  1 to 9.
sus the standard deviation   of the pressure oscil-
lation.
i Te,i (s) H s,i (s) i  Pwave,i (kW/m)

1 9.0 0.8 0.25 2.82

2 9.5 1.2 0.2 6.70 

3 10.0 1.6 0.177 12.5

4 10.5 2.0 0.145 20.6
5 11.0 2.4 0.10 31.0 
6 11.5 2.9 0.07 47.4 Figure 5. Dimensionless plot of the flow rate  versus the
pressure head  for the single-stage Wells turbine.
7 12.0 3.4 0.045 68.0
8 12.5 4.0 0.007 98.0
9 13.0 4.5 0.006 129.0

5.3 Air turbine

Two types of self-rectifying air turbines were con-
sidered in this simulation. The first one is the well-
known Wells turbine, equipped with guide vanes.
Both the more usual version with a single rotor plane ,  
and a row of guide vanes of each side of the rotor, Figure 6. Dimensionless plot of the turbine power output 
and the two-stage version were modelled. The other versus the pressure head  for the single-stage Wells turbine
type is the recent biradial impulse turbine, described (solid line). The broken line represents the average power 
in Falcão et al. (2013a, 2013b), Fig. 4. In both cases, versus   (rms of  ).
results from model testing are used here for the tur-
bine performance curves.
 were computed assuming a Gaussian distribution for
the pressure oscillation. Figure 11 shows that the

0.14
0.12
0.10
0.08
0.06
0.04
,  
Figure 7. Dimensionless plot of the turbine efficiency  versus 0.02

the pressure head  for the single-stage Wells turbine (solid 0.00
0.0 0.2 0.4 0.6 0.8 1.0
line). The broken line represents the average efficiency  ver- ,  
sus   (rms of  ). Figure 9. Dimensionless plot of the turbine power output 
versus the pressure head  for the biradial turbine (solid line).
The broken line represents the average power  versus  
Figure 5 shows that the adopted Wells turbine is
(rms of  ).
very approximately linear, with a flow rate versus
pressure head relationship given, in dimensionless 0.80
form, by   , with   1.0 . Figure 6 shows that
0.75
the power output drops sharply when the pressure
head exceeds a critical value (crit  0.072) . This is 0.70

due to strong aerodynamic stalling at the rotor blades 0.65

and is typical of most Wells turbines. 0.60

A two-stage Wells turbine was also considered,
0.55
with guide-vanes between rotors, possibly as de- 0.0 0.2 0.4 0.6 0.8 1.0
scribed in Falcão and Gato (2012) and Arlitt et al. ,  
(2013). Since no data from model testing are availa- Figure 10. Dimensionless plot of the turbine efficiency  ver-
ble, it was assumed that the dimensionless perfor- sus the pressure head  for the biradial turbine (solid line).
mance curves for the two-stage turbine can be ob- The broken line represents the average efficiency  versus
tained from the curves for the corresponding single-   (rms of  )
stage turbine,   f  (),   f  () and
  f (), by setting  2  f ,2 ()  2 f  ( 2), were computed assuming a Gaussian distribution for
the pressure oscillation. Figure 8 shows that the
 2  f ,2 ( )  f  (  2) ,   f ,2 ( )  f (  2). flow-versus-pressure-head curve of the biradial tur-
Identical transformations apply to the time-averaged bine is not rectilinear. Since the frequency-domain
quantities  and  as functions of   . analysis and the stochastic approach require the tur-
0.20 bine to be linear, in the computations the experi-
mental curve was replaced by a straight line through
0.15 the origin    , with   0.282 . This straight

line crosses the experimental line at abscissa
0.10
  0.35, for which the efficiency  is maximum.
0.05 This is also approximately equal to the value of  
(i.e. the root-mean square of  ) for which the aver-
0.00
0.0 0.2 0.4 0.6 0.8 1.0
age efficiency  is maximum. The time-averaged
,   power output from the linearized turbine was com-
Figure 8. Dimensionless plot of the flow rate  versus the
pressure head  for the biradial turbine (solid line). The bro- puted from    (  ) 
2
.
ken straight line represents   0.282 .

The corresponding curves for the biradial turbine 6 NUMERICAL RESULTS

were obtained from testing a model, as described in
detail in Falcão et al. (2013b), and are plotted in Figs The following values were adopted:
8 to 10. The curves (dotted lines) in Figs 9 and 10   1025 kg/m 3
for the sea water density,
for time-averaged values of power and efficiency at  1.25 kg/m 3 for the atmospheric air density,
g  9.8 m/s 2 for the acceleration of gravity and ed by the blade tip speed constraint than is the sin-
gle-stage turbine.
k  1.25 for the polytropic exponent.
Maximum energy production is not in general the
most appropriate criterion for the investor, who will
be looking for maximum profit (rather than maxi-
mum produced energy). Profit is related, not only to
the amount of energy produced annually (which in
turn depends on turbine size and generator rated
power capacity), but also to the investment required
to purchase, install and maintain the equipment
(which increases with turbine and generator size),
and to the price at which the unit electrical energy is
paid to the producer. These issues were addressed in
detail in Falcão (2004) where several scenarios were
D (m) analysed. Here we will simply assume that the best
Figure 11. Annual-averaged power output Pt,annual versus choices are turbine sizes somewhat smaller than the
turbine rotor diameter D for the single-stage Wells turbine sizes that maximize the annual energy production.
(Wells 1), the two-stage Wells turbine (Wells 2) and the biradi- For the more detailed analysis of power performance
al turbine. and turbine control that follows, we selected
D  2.0 m for the biradial turbine rotor diameter
Figure 11 shows results for the annual-averaged and D  2.5 m for the single- and two-stage Wells
power output Pt,annual of the Pico plant, when turbine rotor diameter.
equipped with a biradial turbine (rotor diameter The effect of the rotational speed constraints on
D  1.5 , 1.75, 2.0, 2.25, 2.5, 2.75 and 3.0 m) and the efficiency (and hence on power output) of the
Wells turbine (1 and 2 stages, rotor diameter Wells turbines is clearly shown in Fig. 12. This con-
D  2.0 , 2.25, 2.5, 2.75, 3.0, 3.25 and 3.5 m). A straint negatively affects the performance of the sin-
constraint of maximum allowed rotor blade tip speed gle-stage turbine in all but the two least energetic sea
D 2  180 m/s was imposed, in order to avoid ex- states, whereas the two-stage turbine is affected only
in sea states i  5 to 9.
cessive centrifugal stresses in the turbine rotor, and
to avoid shock waves in the air flow about the rotor
blades. (This constraint applies particularly to Wells
turbines.) In every case, the rotational speed was op-
timized for the turbine and sea state under considera-
tion with the aid of the subroutine FindMaximum of
Mathematica. 
It can be seen in Fig. 11 that Pt,annual reaches a
maximum of 252 kW for a biradial turbine with ro-
tor diameter D  2.6 m , 185 kW for a single-stage
Wells turbine with D  3.35 m , and 230 kW for a
two-stage Wells turbine with D  3.5 m . In no case 5 10 50 100
(rotor diameter and sea state) did the rotor blade tip Pwave (kW/m)
speed of the biradial turbine reach the 180 m/s which Figure 12. Average turbine efficiency  versus wave power
was set as a limit for the Wells turbines. This con- Pwave for the single-stage Wells turbine D  2.5 m (Wells 1),
straint results in a reduction in power output of the the two-stage Wells turbine D  2.5 m (Wells 2) and the bira-
single- and two-stage Wells turbines of all diameters dial turbine D  2.0 m . The plotted points on each curve rep-
(2.0 to 3.5 m) for several of the nine sea states of resent the 9 sea states.
Table 1. This is more severe for the single-stage
Wells turbine than for the two-stage Wells turbine,
and for the turbines with smaller rotor diameter D. 7 CONCLUSIONS
Obviously, in the case of the two-stage Wells tur-
bine, the pressure head p is divided equally between Results based on linear water wave theory were ob-
the two stages, which are less loaded and less affect- tained for the performance of the OWC plant on Pi-
co island. Non-linear effects due to real fluid (vis-
cosity) effects (including vortex shedding), non- OWC-shoreline devices including the effect of surrounding
small wave amplitude and wave breaking were not coastline and non-flat bottom. Int. J. Offshore Polar Engng
11:147–154.
accounted for. Such effects are known to be im- Curran, R., Folley, M. (2008). Air turbine design for OWCs.
portant in the more energetic sea states. This is espe- In: Cruz J, editor. Ocean Wave Energy. Berlin: Springer,
cially the case of wave breaking which is observed p. 189–219.
to occur due to local water depth of about 8 m in the Dixon, S.L., Hall, C.A. (2014). Fluid Mechanics and Thermo-
immediate vicinity of the plant and about 15 m fifty dynamics of Turbomachinery. 7th edition. Amsterdam:
Elsevier.
metres away. For these reasons, the numerical results Falcão, A.F. de O. (2000), The shoreline OWC wave power
in the more energetic sea states should be regarded plant at the Azores. Proc. 4th European Wave Energy
as qualitative, rather than quantitative, representa- Conf., Aalborg, Denmark, Paper B1.
tions of reality. Falcão, A.F. de O. (2004). Stochastic modelling in wave pow-
Except in the less energetic sea states, the per- er-equipment optimization: maximum energy production
versus maximum profit. Ocean Eng. 31:1407–1421.
formance of the Wells turbine, especially the single- Falcão, A.F. de O., Rodrigues, R.J.A. (2002). Stochastic mod-
stage version, was found to be severely affected by elling of OWC wave power performance. Appl. Ocean. Res.
the constraints on rotor-blade tip speed. This is not 24:59–71.
the case of the biradial turbine. Such behaviour con- Falcão, A.F.O., Gato, L.M.C. (2012). Air turbines. In: Sayigh
AA, editor. Comprehensive Renewable Energy, vol. 8,
firms what should be expected from Wells turbines,
Ocean Energy. Oxford: Elsevier, p. 111–149.
known for their high rotational speed and low Falcão, A.F.O., Gato, L.M.C., Nunes, E.P.A.S. (2013a). A
torque, compared with self-rectifying impulse tur- novel radial self-rectifying air turbine for use in wave ener-
bines, of which the biradial turbine is a radial-flow gy converters. Renew. Energy 50:289–298.
version. Falcão, A.F.O., Gato, L.M.C., Nunes, E.P.A.S. (2013b). A
novel radial self-rectifying air turbine for use in wave ener-
The fact that, unlike the Wells turbine, the biradi-
gy converters. Part 2. Results from model testing. Renew.
al turbine (if its rotational speed is adequately con- Energy 53:159–164.
trolled) is capable of performing efficiently even in Falcão, A.F.O., Henriques, J.C.C. (2014a). Model-prototype
very energetic sea states could be taken as a reason similarity of oscillating-water-column wave energy convert-
to couple it to an electrical generator whose rated ers. Int. J. Marine Energy 6:18–34.
Falcão, A.F.O., Henriques, J.C.C. (2016). Oscillating-water-
power would match the turbine power capability.
column wave energy converters and air turbines: A review.
However, one should not ignore the low frequency Renew. Energy 85:1391–1424.
of occurrence of the most energetic sea conditions. Falnes, J. (2002). Ocean Waves and Oscillating Systems. Cam-
Besides, unlike the turbine, the efficiency of the bridge: Cambridge University Press.
electrical equipment is known to be relatively poor Goda, Y. (2000). Random Seas and Design of Maritime Struc-
tures, 2nd edn. Singapore: World Scientific.
at (the frequently occurring) small power loads. For Heath, T.V. (2012). A review of oscillating water columns.
several reasons, including equipment costs, the spec- Phil. Trans. R. Soc. A 370:235–245.
ified rated power of the electrical equipment should Holthuijsen, L.H. (2007). Waves in Oceanic and Coastal Wa-
be somewhat lower that the power capacity of the bi- ters. Cambridge: Cambridge University Press.
radial turbine in the most energetic sea states. Setoguchi, T. Takao, M. (2006). Current status of self-
rectifying air turbines for wave energy conversion. Energy
In general, the biradial turbine was found to out- Conv. Manag. 47:2382–2396.
perform the Wells turbine, especially the single- Starzmann, R. (2012). Aero-acoustic Analysis of Wells Tur-
stage one. Naturally, the decision on which type of bines for Ocean Wave Energy Conversion. Doctoral thesis,
turbine to choose should also take into consideration Universität Siegen, Germany. Fortschr.-Ber. VDI Reihe 7,
equipment costs, an aspect that is not addressed here. Nr 500. VDI-Verlag, Düsseldorf.

Acknowledgements. This work was funded by the

Portuguese Foundation for the Science and Technol-
ogy (FCT) through IDMEC, under LAETA Pest-
OE/EME/LA0022. Author JCCH was supported by
FCT researcher grant No. IF/01457/2014.

REFERENCES
Arlitt, R., Banzhaf, H.U., Startzmann, R., Biskup, F. (2013).
Air turbine for wave power station. European patent No
EP2232058.
Brito-Melo, A., Hofmann, T., Sarmento, A.J.N.A., Clément,
A.H., Delhommeau, G. (2001). Numerical modelling of