See

discussions, stats, and author profiles for this publication at: http://www.researchgate.net/publication/226547696

A performance study on a direct drive hydro

turbine for wave energy converter

ARTICLE in JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY · NOVEMBER 2010

Impact Factor: 0.84 · DOI: 10.1007/s12206-010-0903-5

CITATIONS READS

5 27

5 AUTHORS, INCLUDING:

You-Taek Kim Young-Ho Lee

Korea Maritime and Ocean University Korea Maritime and Ocean University
33 PUBLICATIONS 51 CITATIONS 110 PUBLICATIONS 255 CITATIONS

SEE PROFILE SEE PROFILE

Available from: Young-Ho Lee

Retrieved on: 12 November 2015
 Journal of Mechanical Science and Technology 24 (11) (2010) 2197~2206
www.springerlink.com/content/1738-494x
DOI 10.1007/s12206-010-0903-5

A performance study on a direct drive hydro turbine for wave energy converter†
Young-Do Choi1, Chang-Goo Kim2, You-Taek Kim3, Jung-Il Song4 and Young-Ho Lee5,*
1
Department of Mechanical Engineering, Mokpo National University, Jeonnam, 534-729, Korea
2
Department of Mechanical Engineering, Graduate School, Korea Maritime University, Busan, 606-791, Korea
3
Department of Marine System Engineering, Korea Maritime University, Busan, 606-791, Korea
4
School of Mechatronics, Department of Mechanical Engineering, Changwon National University, Changwon, 641-773, Korea
5
Division of Mechanical and Information Engineering, Korea Maritime University, Busan, 606-791, Korea

(Manuscript Received July 1, 2008; Revised May 6, 2010; Accepted July 14, 2010)

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Abstract

Clean and renewable energy technologies using ocean energy give us non-polluting alternatives to fossil-fueled power plants as a
countermeasure against global warming and growing demand for electrical energy. Among the ocean energy resources, wave power
takes a growing interest because of its enormous amount of potential energy in the world. Therefore, various types of wave power sys-
tems to capture the energy of ocean waves have been developed. However, a suitable turbine type is not yet normalized because of rela-
tively low efficiency of the turbine systems. The purpose of this study is to investigate the performance of a newly developed direct drive
hydro turbine (DDT), which will be built in a caisson for a wave power plant. Experiment and CFD analysis are conducted to clarify the
turbine performance and internal flow characteristics. The results show that the DDT obtains fairly good turbine efficiency in cases with
and without wave conditions. Most of the output power is generated at the runner passage of Stage 2. Relatively larger amount of the
decreased tangential velocity at Stage 2 produces more angular momentum than that at Stage 1 and thus, the larger angular momentum at
the Stage 2 makes a greater contribution to the generation of total output power in comparison with that at Stage 1. Large vortex existing
in the upper-left region of the runner passage forms a large recirculation region in the runner passage, and the recirculating flow con-
sumes the output power at Region 2.
Keywords: Direct drive hydro turbine (DDT); Internal flow; Performance; Wave condition; Wave power
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

been introduced to improve the turbine performance because

1. Introduction
water flow has a higher energy density compared to that of air
Clean and renewable energy technologies using ocean en- flow [13-16]. Therefore, this study investigates the perform-
ergy give us non-polluting alternatives to fossil-fueled power ance of a newly developed direct drive hydro turbine (DDT)
plants and act as a countermeasure against global warming for wave energy converter.
and growing demand for electrical energy. Among ocean en- Fig. 1 shows a schematic view of the DDT, which will be
ergy resources, wave power takes a growing interest because built in a caisson of a wave power plant, and the turbine op-
of its enormous amount of potential energy in the world. eration mechanism was originally suggested by Fukutomi et al.
To develop wave energy, various types of wave energy con- [13]. When the water with waves enters the front guide nozzle,
verters [1-12] have been developed so far to use the energy the wave motions change to reciprocating flow in the internal
from ocean waves. Among the converter systems, the oscillat- flow passage of the turbine. The reciprocating flow in the
ing water column (OWC) type turbine is a representative con- turbine passage rotates the runner and thus generates output
verter using air as a working fluid to rotate the runner [1-8]. power. The runner in the turbine is designed to maintain unidi-
However, a suitable turbine type for the OWC is not general- rectional rotation even in the case of reciprocating flow in the
ized yet because of relatively low efficiency and high flow turbine passage.
induced noise by the OWC turbine systems. Therefore, the The shape of the front and rear nozzles has the key impor-
direct drive hydro turbine (DDT) type converter, which uses tance of unidirectionality of the rotor rotational direction. The
the water flow in the turbine passage as working fluid, has flow moves bidirectionally according to the wave movements.
†
This paper was recommended for publication in revised form by Associate Editor However, the turbine rotor should be rotated in one rotational
Won-Gu Joo direction because only unidirectional rotation of the rotor axis
*
Corresponding author. Tel.: +82 51 410 4293, Fax: +82 51 403 0381
can produce uniform electricity from the generator. Therefore,
E-mail address: [email protected]
© KSME & Springer 2010 the shape of the front and rear nozzles has the same dimen-
2198 Y.-D. Choi et al. / Journal of Mechanical Science and Technology 24 (11) (2010) 2197~2206

Table 1. Dimensions of test turbine for experiment.

Designed inflow velocity vin 0.43 m/s

Designed effective head H 0.05 m

Designed rotational speed N 45 min-1

Designed output power PT 22 W

Turbine length 700 mm

Turbine width bT 700 mm

Turbine height 370 mm

Front nozzle inlet width bn 700 mm

Front nozzle inlet height 250 mm

Fig. 1. Schematic view of test DDT model. Guide nozzle inlet width bgn 0.9 m

Guide nozzle inlet height 623 mm

sions but is installed in the symmetric positions to the axis; the
nozzles are installed at the upper wall side at front nozzle and Table 2. Dimensions of test runner for experiment.
at the bottom wall side at rear nozzle.
Even though this turbine can be applied to the wave energy Outer diameter D2 260 mm
converters both of the fixed and floating types, the present Inner diameter D1 165 mm
study is mainly focused on the application to the fixed type
runner width bT 700 mm
wave energy converter installed in a breakwater in the near-
shore waters. Outer blade angle α 30 deg.
Experiments using a DDT model in a 2-D wave channel and Inner blade angle β 90 deg.
numerical simulation using a commercial CFD code are
Blade number Z 30
carried out to investigate the performance and internal flow
characteristics of the turbine.
Wave gauge Wave gauge
Main purpose of CFD analysis for the basic shape of the di-
Torque meter (Unit : mm)
rect drive turbine is to investigate the internal flow and per-
formance of the turbine as well as to examine the local output
720

power distribution according to the tangential location of the

623
250
370

turbine rotor. Even though the present CFD analysis adopts

700
175

unidirectional inflow to the turbine passage, the results by

CFD analysis can explain the mechanism of output power 550 3650

occurrence by the inflow to the turbine. As the reverse flow to

Fig. 2. Schematic view of the test facility in a 2-D wave channel.
the turbine passage also causes almost the same tendency and
movement of the flow field, present CFD results by unidirec-
tional inflow can be a good reference to the actual test turbine the flow rotates the runner in one direction constantly.
performance by experiment with reciprocating inflow condi- All the experiments for the test turbine model are carried out
tion to the turbine passage. in a 2-D wave channel as shown in Fig. 2. The wave channel
has a total length of 35 m, width of 1m and depth of 1m. The
2. Experimental apparatus and numerical methods bottom slope of the wave channel consists of a 7 m flat section
2.1 Turbine model and experimental setup in the deep end from a wave maker and a slope of 1:100 to the
shallow end. The test turbine model is installed at the position
The design concept of this turbine is to apply the operation of 15 m downstream from the wave maker in the wave
mechanism of an axisymmetric cross-flow type hydro turbine channel. The piston-type wave maker, which is installed at the
to a wave energy converter [13]. The dimensions of the test right end of the water channel, can generate regular waves
turbine are summarized as shown in Tables 1 and 2. The front with various wave height and wave period. By the dis-
guide nozzle gathers waves into the turbine and changes the placement and velocity of wave paddle movements, the wave
waves to horizontally reciprocating flow in the turbine passage. condition can be controlled. Table 3 represents the dimensions
The rear water reservoir attached to the turbine plays the role of the wave channel and the test wave condition.
of accumulating the flow passing from the runner passage and Measurement instruments used in this study are summarized
returning the flow in contrary directions. The inflow and out- in Table 4. A torque meter is installed outside of the turbine
flow in the turbine passage make the reciprocating flow and and the output torque generated at the runner shaft is trans-
 Y.-D. Choi et al. / Journal of Mechanical Science and Technology 24 (11) (2010) 2197~2206 2199

Table 3. Dimensions of wave channel and wave condition. 4 2.0

Total length 35 m

Torque loss (Nㆍ m)

3 1.5
Width 1m

Power loss (W)

Power Loss
Torque Loss
Height 1m
2 1.0
Bottom slope ratio 1:100

Turbine location from wave maker 15 m

1 0.5
Water depth h 75 cm

Wave height Hw 20 cm 0 0.0

0 10 20 30 40 50 60 70
Wave length L 4.75 m -1
N (min )
Wave period Tw 2 sec.
Fig. 3. Verification of mechanical loss.
Wave type Regular

machinery occupies considerable portion to the total output

Table 4. Specifications of measurement instruments.
power loss. Therefore, this study has conducted verification
Product name test for the mechanical loss with the variation of rotational
Description Range
(Maker) speed and the test results are floated in Fig. 3.
midi LOGGER The measurement method of power loss and torque loss is
Data logger 8CH GL500A carried out by measuring the output torque from the turbine
(Graphtec) axis when the turbine is operated under the condition of no
Pressure transducer 0 – 200 kPa
PSHF0002KAAG water flow in the turbine by some constant rotational speeds.
(KONICS) If there exists output torque without no water flow in the tur-
CHT4-50 bine, the output torque is definitely caused by mechanical loss.
Capacitance wave gauge ± 25 cm
(KENEK) Power loss can be achieved by multiplying the torque loss and
3-D acoustic Doppler Vector Velocimeter angular speed.
0 – 7 m/s
velocimeter (NORTEC AS)
The measured torque loss by rotational speed is almost linear,
YDRM-5KM and the measured output power loss is gradually increasing by
Torque meter 0 – 50 N·m
(SETech)
the increase of rotational speed. The value of the output power
MP-981 loss is remarkable compared with the designed output power
Revolution counter 0 - 1,000 min-1
(ONO SOKKI)
(22 W). Hence, in this study, consumed output power by the
HDB-03
Air tension brake 0 – 20 N·m mechanical loss is restored to the total output power in order
(Hyundai Clutch)
to examine the net output power generated hydrodynamically
by the present DDT.
ferred to the torque meter by a timing belt and a pulley. Rota- The uncertainty estimates for each variable in the graphs are
tional speed of the test runner is measured using a revolution based on the method of Abernethy et al. [7]. The total uncer-
counter which is attached to the torque meter. An air tension tainty ( U ) of the variables can be found by combining sys-
brake is used to control the rotational speed. Two pressure tematic and random errors as : U = [ B 2 + (tS X ) 2 ]1/ 2 , where B
transducers are connected to the holes on the sidewalls both of is the systematic uncertainty, S X is the standard deviation of
the front and rear turbine nozzles to measure the differential the mean, and the degree of freedom t is determined to 2 for a
pressure between the front and rear turbine nozzles. Two ca- 95% confidence level (for a sample size greater than 50). The
pacitance wave gauges are located in the water channel at the systematic uncertainty B is estimated based on the calibra-
positions of 3,650 mm upstream and 550 mm downstream tion data and previous test experience, and the standard devia-
from the test turbine center. tion of the mean S X is computed from the raw test data.
The measured wave data by the wave gauge at the position Measurement uncertainties for turbine performance under a
upstream from the turbine guide nozzle is used for the exami- loaded condition are estimated to be Q = ± 1.39 percent, Hw =
nation of incoming wave condition, but the wave data meas- ± 1.0 percent, Tw = ± 1.4, PT = ± 1.5 percent and h = ± 2.23
ured at the rear water reservoir is used for determining the percent at the rotational speed of best efficiency, respectively.
incoming flow rate by averaging the time serial wave height in
the reservoir. 2.2 Numerical methods
For the accurate measurement of the output torque generated
in the turbine, verification of the mechanical loss caused by As the DDT will be installed in the wave power plant lo-
bearings and seals installed in the turbine is necessary because cated at the offshore or nearshore waters, the performance of
the mechanical loss occurring from relatively small rotating the turbine will be influenced considerably by the wave condi-
2200 Y.-D. Choi et al. / Journal of Mechanical Science and Technology 24 (11) (2010) 2197~2206

Table 5. Dimensions of test turbine for CFD analysis.

Designed inflow velocity vin 1 m/s

Designed Effective head H 0.5 m

Designed rotational speed N 45 min-1

Designed output power PT 500 W

Turbine nozzle width bT 450 mm

Turbine nozzle inlet height 400 mm

Runner outer diameter D2 450 mm

Fig. 5. Numerical grids of the turbine model for CFD analysis.
Runner inner diameter D1 293 mm

Runner outer blade angle α 30 deg. Table 5 indicates the dimensions of the test turbine for CFD
Runner inner blade angle β 90 deg. analysis. The dimensions of the test turbine for CFD analysis
are different from those of the test turbine for experiment (re-
Blade number Z 26
fer to Table 1). As the basic shape of the test turbine for CFD
analysis could not be realized for the test turbine model for
α experiment in water channel, design parameters of the test
Runner
turbine model for experiment are determined differently ac-
Nozzle cording to the turbine design method.
β
Region 2 Stage 2 Numerical grids of about 1.5x106 are adopted for the analy-
Outflow
sis of the calculation domain including the runner and nozzle
Inflow
as shown in Fig. 5. Fine hexahedral grids are employed at the
Region 1 whole flow field to ensure the high accuracy of calculated
Stage 1 results. To check the dependence of grid quality on the CFD
Nozzle
result, Choi et al. [20] have investigated the turbine perform-
ance using the results from the CFD analysis by various val-
ues of dimensionless wall distance, y+, in the calculation do-
Fig. 4. Calculation domain of the turbine and division of runner pas- mains of runner and nozzle. As the calculated results for the
sage.
dependence of y+ on the turbine performance hardly show
large difference under the range of y+ ≦50, the dimensionless
tions such as wave height, wave period and the ratio of wave wall distance is kept both below 15 and below 50 in the grid
height vs. water depth. However, when the turbine is designed regions of runner and nozzle passages, respectively.
for an expected installation site, the averaged turbine perform- Moreover, the dependency of three turbulence models in-
ance at the design point is very important even though the cluding SST turbulence model has been compared using a
turbine performance varies largely by the wave conditions. basic DDT model from the previous related study [20]. Ac-
Therefore, examination of the turbine performance by the cording to the study result, SST turbulence model showed no
operating conditions of no-wave conditions, such as unidirec- large difference from the other turbulence models. However,
tional inflow, constant flow rate and constant effective head, is the SST turbulence model has relatively good advantages of
necessary. For the detailed investigation of the performance realizing both high and low energy areas simultaneously in a
and internal flow of the DDT, a 3-D finite volumetric method complicated flow field with relatively low grid dependency.
based commercial CFD code of ANSYS-CFX [18] is adopted Therefore, this study adopts the SST turbulence model.
and the code is well known for the excellence of analyzing the As the boundary conditions of CFD analysis for the turbine
performance of turbomachinery. flow field, constant pressure at the inlet and averaged flow rate
Fig. 4 shows the calculation domain of the turbine. The at the outlet, no slip conditions on the passage walls of the
shapes of the runner and nozzle of the turbine in Fig. 4 are calculation domain are used. Several rotational speeds are
same as that in Fig. 2, but the dimensions of the turbine are adopted for the CFD analysis to examine the turbine perform-
different. In case of the determination of the turbine nozzle ance at optimum rotational speed. Moreover, the interface
shape, previous related study result by Choi et al. [19] is re- between the rotating parts and stationary parts for steady cal-
ferred. The criterion of dividing the runner passage into the culations is treated using the frozen rotor method of the
local stages and regions is determined by the zone of main- adopted commercial CFD code of ANSYS CFX [18]. All the
stream passage. The stage stands for a flow passage of main calculations are conducted under the conditions of unidirec-
stream but the region represents a stagnated flow zone. tional flow and steady state.
 Y.-D. Choi et al. / Journal of Mechanical Science and Technology 24 (11) (2010) 2197~2206 2201

3. Results and discussions 100 1000 1.60

90 900 PT 1.44
3.1 Performance curves and internal flow of the DDT unidi- 80 800 ηT 1.28
rectional steady flow 70 700 H 1.12
60 600 0.96
3.1.1 Performance curves and output power

PT (W)
ηT (%)

H (m)
50 500 0.80
Fig. 6 reveals the performance curves of the turbine model 40 400 0.64
by CFD analysis under the condition of unidirectional steady 30 300 0.48
flow. The performance curves result from the calculation con- 20 200 0.32

ditions that constant pressure at the inlet and averaged flow 10 100 0.16

rate at the outlet, no slip conditions on the passage walls of the 0 0

0 20 40 60 80 100 120
0.00

calculation domain. Several rotational speeds are adopted for -1

N (min )
the CFD analysis to examine the turbine performance at opti-
Fig. 6. Performance curves by CFD analysis.
mum rotational speed. The abscissa means rotational speed N,
and the ordinate indicates the efficiency ηT, output power PT 700
and effective head H. PT
600
The turbine efficiency ηT is given by the following equation : 621.2(100 %)
104.7(16.8%)
500 485.3(78.1%)
36.6( 5.9%)
400 - 5.4(- 0.9%)
PT Tω

PT (W)
ηT = = . (1) 300
Pw ρ gQH
200

As the rotational speed increases, effective head H , which is 100

calculated from the differential pressure between the measur- 0
ing points at the front and rear nozzles of the turbine, increases
-100
almost proportionately with a steep slope. Output power PT
Total Stage1 Stage2 Region1 Region2
increases steeply by the increase of rotational speed but de-
creases at the range of N > 90 min-1. In case of efficiency, after Fig. 7. Output power by CFD analysis at the best efficiency point
a steep increase of the efficiency until the rotational speed of (N=50 min-1).
N=40min-1, the slope of the efficiency curve becomes gentle
and the efficiency curve decreases gradually by the increase of controlled by the unique shape of the nozzle passage and the
rotational speed. Best efficiency reaches to 51.7% at the rota- local output powers are governed by the nozzle shape as well.
tional speed of N = 50 min-1.
The result implies that as available effective head H by wave
height in an actual ocean environment can’t be obtained 3.1.2 Velocity vectors and distributions
enough to give remarkable effect on the turbine efficiency, To examine the relation between the performance and inter-
rotational speed and incoming water flow rate should be in- nal flow, velocity vectors in the turbine internal passage and
creased to generate an optimum output power. However, as averaged velocity distributions at the runner blade passage are
the rotational speed is restricted by the structural condition of examined as shown in Fig. 8.
the turbine runner, the cross-sectional area of the turbine inlet Fig. 8(a) shows velocity vectors in the internal flow field of
is designed as wide as possible to receive enough water flow the turbine model. Except for the relative velocity vectors in
rate for a required output power. the runner blade passage, absolute velocity vectors are shown
To investigate the output power obtained at the runner, total in the other flow passages. It is clear that flow velocity be-
output power and the local output powers, which are calcu- comes accelerated within the runner center region after pass-
lated at each stage and region as defined in Fig. 4, are com- ing through Stage 1, and then the flow enters the runner blade
pared with each other as shown in Fig. 7. In case of cross-flow passage at the Stage 2 with higher velocity than that at Stage 1.
hydro turbine for hydro power generation, Stage 1 produces A large vortex exists at the upper-left zone of the runner in-
about 70% of output power and about 30% of output power is ternal passage and the vortex forms a wide recirculating flow
obtained at Stage 2 [21, 22]. However, the present study re- within the runner passage. The recirculating flow region
veals that output powers of 16.8% at the Stage 1 and 78.1% at gives considerable influence on the output power of this tur-
the Stage 2 are generated, respectively. In addition, 0.9% of bine because the runner blades at the Region 2 consume the
the output power is consumed at the Region 2. It is conjec- output power as indicated in Fig. 7. Moreover, there exists
tured that the unique proportion of the output power might be another large vortex at the bottom region in the rear nozzle
resulting from the shape of the present DDT nozzle, which is passage.
axisymmetric each other at the front and rear nozzles. There- Fig. 8(b) indicates the averaged velocity distributions, which
fore, the flow passing through the runner passage is strongly are averaged from the whole blade passages at Stages 1 and 2
2202 Y.-D. Choi et al. / Journal of Mechanical Science and Technology 24 (11) (2010) 2197~2206

Velocity
Velocity Recirculating flow by vortex Pressure

Vortex
(m/s)
(Pa)
(a) Velocity vectors
(a) Static pressure contours
Suction Pressure
side side 3
5
4
2
3
vθ / u2

2
1
1

Cp
0
vr / u2

0
Stage
-1
1
-2 2
-1
-3 Stage 1 inlet Stage 1 inlet
Stage 1 outlet Stage 1 outlet
-4 vθ / u2 Stage 2 inlet vr / u2 Stage 2 inlet -2
Stage 2 outlet Stage 2 outlet
-5 0.6 0.7 0.8 0.9 1.0
0.0 0.2 0.4 0.6 0.8 1.0 r/r2 2)
Radius(r/r
θ1, θ2
(b) Averaged pressure coefficient distribution on the surface around
(b) Averaged velocity distribution at the runner blade passage the runner blade
Fig. 8. Velocity vectors and distributions in the turbine by CFD analy- Fig. 9. Pressure contours and distributions in the turbine by CFD
sis (N=50 min-1). analysis (N=50 min-1).

to the one blade passage. Tangential velocity vθ at the inlet of

3.1.3 Pressure contours and pressure distributions
Stage 1 decreases when the flow passes through the blade
passage. As the flow passes through the center zone of the Fig. 9 shows static pressure contours in the turbine flow field
runner, the relatively low tangential velocity at the outlet of and averaged pressure distributions on the surface around the
Stage 1 increases to the tangential velocity at the inlet of Stage runner blade. The static pressure contours in Fig. 9(a) show
2. The tangential velocity at the inlet of Stage 2 decreases that relatively high pressure at the outlet of the turbine nozzle
again when the flow passes through the blade passage of the decreases sharply along the runner passages both at Stage 1
stage. and Region 1. The pressures at the runner blade passage of
While the amount of decreased tangential velocity (Δvθ) be- Stage 2 and at the rear turbine nozzle are relatively very low.
tween the inlet and outlet of the runner blade passage at the
two stages is proportional to the amount of angular momen- Moreover, distributions of pressure coefficient by averaged
tum in the runner passage (ΔLAM = rMΔvθ), the decreased tan- pressure around the surface of the runner blades at Stages 1
gential velocity between the inlet and outlet of the runner and 2 are compared with each other in Fig. 9(b). The area
blade passage at the two stages changes to output power at the filled with the curves of pressure coefficient means the pres-
local stages. Therefore, the reason for the relatively larger sure difference between pressure side and suction side of the
output power at Stage 2 than that at Stage 1 in Fig. 7 can be runner blade. Therefore, a wider area filled with the pressure
explained by the present calculated result for the velocity dis- coefficient curve produces higher output power in the runner
tributions. A relatively larger amount of the decreased tangential passage. From the result in Fig. 9(b), it is obvious that there is
velocity at Stage 2 produces a greater amount of angular mo- a wide difference between the areas by the two stages. Hence,
mentum than that at Stage 1 and thus, the larger angular mo- it can be conjectured that the high tangential velocity at the
mentum at Stage 2 makes a greater contribution to the genera- inlet of Stage 2 is converted to the relatively large output
tion of total output power in comparison with that at Stage 1. power at the stage. In other words, the pressure distributions
 Y.-D. Choi et al. / Journal of Mechanical Science and Technology 24 (11) (2010) 2197~2206 2203

(a) No load condition (N=66 min-1)

Fig. 10. Water surface elevation generated by 2-D wave maker (Hw=20
cm, Tw=2 sec., h=75 cm).

suggests that primary contribution to the output power at

Stage 2 is mainly made by the effect of angular momentum
acting on the surface of the runner blade at the stage.

3.2 Experimental results for test turbine performance by

reciprocating flow

3.2.1 Time serial output data by the DDT

As the DDT suggested in this study will be operated in the
ocean, the effect of wave conditions on the turbine perform- (b) Loaded condition (N=43 min-1)
ance should be considered. Therefore, a performance test is Fig. 11. Time serial output data by the turbine (Hw=20 cm, Tw=2 sec.,
carried out with consideration for the wave conditions embod- h=75 cm).
ied by a regular wave maker in the 2-D wave channel.
Fig. 10 shows the wave measured at the position 3650 mm
upstream from the turbine center as well as the input wave by
the wave maker. The measured wave shows higher and nar-
rower wave shape than that of the input wave at the water
surface elevation ηT over 0. The transformation of the wave
shape from the input wave to the measured wave is mainly
influenced by the reflection wave at the inlet of the guide noz-
zle. The reflection wave at the guide nozzle inlet mixes with
the regular wave propagated from the wave maker; then, the
shape of regular wave changes to the shape of the measured
wave as shown in Fig. 10.
Moreover, the measured time serial output data from the
DDT model by both the no load and loaded conditions is Fig. 12. Performance curves of the test turbine (Hw=20 cm, Tw=2 sec.,
h=75 cm).
shown in Fig. 11. Rotational speed of the runner varies regu-
larly according to the wave movement and the averaged rota-
tional speed N decreases from 66 to 43 min-1 by the applied tional speed of N > 43 min-1.
load to the runner axis. However, as the load is applied to the Meanwhile, the change of the efficiency ηT and output power
turbine, outstanding increase of the instant output power PT PT by rotational speed shows steep increase and decrease with
instant, instant output torque Tinstant and instant effective head
the turning point of a sharp peak at the rotational speed of N =
Hinstant can be recognized. 43 min-1. This result implies that when the wave enters to the
turbine internal passage with a certain wave height and a wave
period, the effective head H and flow rate Q change little by
3.2.2 Performance curves rotational speed but best efficiency and maximum output
Fig. 12 shows performance curves of the turbine model by torque occur at an optimum range of the rotational speed.
the wave condition of wave height Hw=20 cm and wave pe- Therefore, the designed point of the turbine should be
riod Tw=2 sec. Effective head H and flow rate Q reveal linear determined with consideration for the wave condition at the
decrease with gentle slope by the increase of rotational speed expected installation site to achieve optimum performance
N but output torque T decreases with steep slope at the rota- during the turbine operation.
2204 Y.-D. Choi et al. / Journal of Mechanical Science and Technology 24 (11) (2010) 2197~2206

Table 6. Efficiency analysis of the turbines for wave energy converter.

Designed Guide Significant Wave Wave Guide nozzle Turbine Generator Total
Project Institute Turbine Place
output power nozzle wave height period energy(1) efficiency efficiency efficiency efficiency
Name (Country) type of exp.
(PD) width (H1/3) (T1/3) (Pwave or Psea) (ηgn) (ηT) (ηG) (ηoverall)
Water
- - - - - Max. 80.0% - - -
LIMPET WaveGen channel
OWC
[5] (U.K.) 500 kW 40.0%
Ocean 20 m - - 12 kW/m 64.0% - 8.7%
(250 kw×2) (D1=2.6 m)
1.0 s
Water 3 cm - 48.0%
Mighty - 1.5 m - - Max. 55.0% - -
JAMSTEC channel 15 cm (D1=298 mm)
whale OWC 2.5 s
(Japan)
[6] 45.0% 90.0%
Ocean 50 kW 30 m 1.5 m 7.0 s - Max. 33.0% 13.2%
(D1=1.7 m) (estimated)
1.34 s
Water
Sakata - 0.8 m 0-52 cm - - Max. 70.0% - - -
PARI channel
port OWC 3.35 s
(Japan)
[7] 35.6%
Ocean 60 kW 20 m 2.27 m 7.4 s 15.9 kW/m 23.1% 92.9% 7.7%
(D1=1.337 mm)
Korea
48.6% 12.03%
Present Maritime Water 27.5% 90.0%
DDT 22 W 0.9 m 20 cm 2.0 s 80.68 W (D1=260 mm) (27.3%
study University channel (50% designed) (estimated)
(60% designed) designed)
(Korea)
Remarks (1) Reference equation for wave energy estimation : equation (2) for water channel, equation (3) for ocean

3.2.3 Efficiency analysis of the DDT ηoverall = η gn × ηT × ηG . (4)

The incoming wave motions to the DDT model change to
reciprocating water flow by the guide nozzle, and the recipro- where the converting efficiency of the guide nozzle (ηgn) can
cating flow in the turbine passage rotates the runner when the be calculated by ηgn = Pw / Etotal in case of water channel ex-
flow passes through the runner blade passages. Therefore, periment, and the incoming wave energy at the guide nozzle
local efficiencies at the guide nozzle (ηgn) and the turbine inlet (Etotal) can be calculated by Etotal = Pwave × bgn. In addition,
model (ηT), as well as total efficiency (ηoverall) are investigated wave energy at the ocean with irregular waves is calculated by
to examine the performance of the turbine system in detail. Etotal = Psea × bgn.
When a regular wave is given in the wave channel, wave Therefore, the efficiencies at the turbine local regions and
power Pwave is calculated by the following equation [14]: total efficiency by the present turbine system are summa-
rized in Table 6. For the comparison of the DDT efficiency
⎡ 4π h ⎤ with other types of the turbine for wave energy converters,
1 ⎢ ⎥
2 L L some other research results are summarized as well. Even
Pwave = ρ gQH i bgn ⎢1 + ⎥ . (2)
16 Tw ⎢ sinh ⎛ 4π h ⎞ ⎥ though the test environments for each turbine system are
⎢⎣ ⎜ ⎟
⎝ L ⎠ ⎥⎦ different, turbine efficiency of the present DDT model is
remarkable compared with those of the other turbine systems.
Moreover, in the case of irregular wave in the ocean, the However, for greater improvement of the total efficiency
wave power Psea is given by the following equation [23]: ηoverall to the designed value, optimization for the shapes of
both the guide nozzle and the turbine flow passage is neces-
ρ g 2 H s2Tz H 2T sary.
Psea = 2
× 0.001 = 0.492 s2 z . (3)
64π K sm K sm
4. Conclusions
where (1) The present direct drive hydro turbine system shows fairly
⎡ 4π h ⎤ good turbine efficiencies of 51.7% at the best point by uni-
1 1 ⎢ ⎥
K sm = , n = C , C = ⎢1 + L ⎥ directional steady flow without input wave condition and
2n tanh km h 2 ⎢ sinh ⎛ 4π h ⎞ ⎥ 48.6% by reciprocating flow with an input wave condition
⎢⎣ ⎜ ⎟
⎝ L ⎠ ⎥⎦ in a 2-D wave channel. For greater improvement of total
efficiency for the turbine system, optimization for the
While, total efficiency ηoverall of the present DDT model is shapes of both the guide nozzle and the turbine flow pas-
calculated by the following equation: sage is necessary.
 Y.-D. Choi et al. / Journal of Mechanical Science and Technology 24 (11) (2010) 2197~2206 2205

(2) Most of the output power generated at the runner blade ηgn : Guide nozzle efficiency
passage of the Stage 2 and the output power generated in ηoverall : Generator efficiency
the local runner passage reaches to 78.1% of total output ηw : Water surface elevation
power. Relatively larger amount of the decreased tangential θ1 : Normalized peripheral blade position in one blade
velocity at Stage 2 produces more angular momentum than passage at Stage 1
that at Stage 1 and thus, the larger angular momentum at θ2 : Normalized peripheral blade position in one blade
Stage 2 makes a more contribution to the generation of total passage at Stage 2
output power in comparison with that at Stage 1. ρ : Density of water
(3) A large vortex existing in the upper-left region of the run- ω : Angular velocity
ner passage forms a large recirculation region in the runner
passage and the recirculating flow consumes the output
References
power at the Region 2.
(4) Designed point of the turbine should be determined in [1] T. Setoguchi, S. Santhakumar, H. Maeda, M. Takao and K.
consideration for the wave condition at the expected instal- kaneko, A review of impulse turbines for wave energy con-
lation site to achieve optimum performance during the tur- version, Renewable Energy 23 (2001) 261-292.
bine operation. Among the design parameters for the test [2] T. Setoguchi, S. Santhakumar, M. Takao, T. H. Kim and K.
turbine, key importance is kept to the rotational speed in Kaneko, A performance study of a radial turbine for wave
this study. Best efficiency is located in the adjacent of rota- energy conversion, Proc. Instn Mech Engrs Part A : Power
tional speed of N=45min-1 as expected in the designed pa- and Energy 216 (1) (2002) 15-22.
rameter. [3] R. Curran, Productivity of ocean-wave energy converters :
turbine design, J. Energy Engineering 128 (2) (2002) 13-31.
[4] D. G. Dorrell, M-F. Hsieh and W. Fillet, Segmented small
Nomenclature------------------------------------------------------------------------
oscillating water columns using in-line savonius rotors, Proc.
bT : Width of nozzle and runner the 7th International Offshore and Polar Engineering Con-
bgn : Width of guide nozzle inlet ference, Lisbon, Portugal (2007).
Cp : Pressure coefficient (= ( p − pref ) / ρ gH ) [5] Research into the further development of the LIMPET
g : Acceleration of gravity shoreline wave energy plant, Supplement to the Final Report
H : Effective head to DTI Sustainable Energy Programme, Wavegen, U.K.
Hi : Incoming wave height (2003).
Hs : Significant wave height [6] Research and development of wave energy utilization tech-
Hw : Wave height nology - development of offshore floating type wave energy
h : Water depth converter ｢Mighty Whale｣, JAMSTEC (in Japanese), Japan
km : Root mean square wave number (2004).
Ksm : Root mean square shallow water coefficient [7] S. Takahashi, T. Adachi, H. nakada, H. Ohneda, H. Katou
L : Wave length and M. Shikamori, Field experiment of a wave power ex-
LAM : Angular momentum tracting caisson breakwater - data analysis of wave forces
M : Mass of water and wave power conversion, Report of the Port and Harbour
N : Rotational speed Research Institute, 31 (2) (1992) 21-54.
PT : Turbine output power [8] J-S. Moon, K-Y. Hong, S-H. Shin, B-S Hyun, H-J Ryu and
Pw : Water input power S-J Park, Oscillating flow field analysis as shape of air
Pwave : Waver energy by regular wave chamber in OWC-type wave energy conversion, J. of Ko-
p : Static pressure rean Navigation and Port Research, 31 (1) (2007) 29-33.
pref : Reference static pressure at the turbine rear nozzle [9] J. Tedd, J. P. Kofoed, M. Jasinski, A. Morris, E. Friis-
Psea : Wave energy by irregular sea wave Madsen, R. Wisniewski and J. D. Bendtsen, Advanced con-
Q : Volumetric flow rate trol techniques for WEC wave dragon, Proc. the 7th Euro-
r : Runner radius pean Wave and Tidal Energy Conference, Porto, Portugal
r2 : Outer radius of runner (2007).
T : Output torque [10] S. J. Beatty, B. J. Buckham and P. Wild, Modeling, design
Tw : Wave period and testing of a two-body heaving wave energy converter,
Tz : Zero up-crossing period Proc. the 7th International Offshore and Polar Engineering
u2 : Runner rotational speed at outer radius Conference, Lisbon, Portugal (2007).
vr : Radial velocity [11] Brochure of pelamis P-750 wave energy converter, Pelamis
vθ : Tangential velocity Wave Power Ltd. (2008).
ηT : Turbine efficiency [12] M. Folley, T. J. T. Whittaker and A. Henry, The effect of
ηG : Generator efficiency water depth on the performance of a small surging wave en-
2206 Y.-D. Choi et al. / Journal of Mechanical Science and Technology 24 (11) (2010) 2197~2206

ergy converter, Ocean Engineering 34 (2007) 1265-1274. bine by the shapes of nozzle and runner blade, J. of Fluid
[13] J. Fukutomi and Y. Nakase, A study of turbine for wave Science and Technology, 3 (3) (2008) 398-409.
power generation, Proc. of the 1st Pacific / Asia Offshore [23] H-J. Ryu, K-H. Hong, S-H. Shin and D-Y. Kim, A study on
Mechanics Symposium, Seoul, Korea (1990) 193-198. wave condition and design wave for the optimal design of
[14] G. Orer and A. Ozdamar, An experimental study on the wave power generator, Proc. of Korea Ocean Science and
efficiency of the submerged plate wave energy converter, Technology Association Conference (in Korean), Korea
Revewable Energy 32 (2007) 1317-1327. (2007) 1800-1807.
[15] S. Kondo, K. Taniya, K. Takahashi, H. Watabe and K.
Okuda, Study on the wave energy absorption devices in-
stalled in the breakwater facility - wave power turbine type,
Proc. the 28th Coastal Engineering Conference (in Japa- Young-Do Choi received his B.E. and
nese), (1981) 381-385. M.E. in Mechanical Engineering from
[16] H. Watabe, S. Kondo, K. Taniya, E. Takeda and S. Kuroi, Korea Maritime University, Korea in
Study on the wave energy absorption devices installed in the 1996 and 1998, respectively. He then
breakwater facility - pendulum type, Proc. the 29th Coastal received his Dr. Eng. degree from
Engineering Conference (in Japanese), (1982) 486-490. Yokohama National University, Japan
[17] R. B. Abernethy, R. P. Benedict and R. B. Dowdell, ASME in 2003. Dr. Choi is currently a fulltime
measurement uncertainty, ASME J. Fluids Eng. 107 (1985) lecturer at the Department of Mechani-
161-164. cal Engineering, Mokpo National University in Jeonnam,
[18] ANSYS-CFX Documentation Ver. 11, ANSYS Inc. (2007). Korea. His research interests include ocean energy, wind
[19] Y-D. Choi, C-G. Kim, Y-T. Kim and Y-H. Lee, A study on power, small hydro power, fluid machinery, PIV and CFD.
the nozzle shape of a cross-flow type hydro turbine for wave
power generation, J. of Fluid Machinery (in Korean) 11 (3) Young-Ho Lee received his B.E. and
(2008) 30-35. M.E. degrees from Korea Maritime
[20] Y-D. Choi, Y-J. Cho, Y-T. Kim and Y-H. Lee, Perform- University, Korea. He then received his
ance and internal flow of a cross-flow type turbine for wave Dr. Eng. degree from the University of
power generation, J. of Fluid Machinery 11 (3) (2008) 22-29. Tokyo, Japan. Dr. Lee is currently a
[21] A. A. Fiuzat and B. P. Akerkar, Power outputs of two Professor at the Division of Mechanical
stages of cross-flow turbine, J. Energy Engineering 117 (2) and Information Engineering, Korea
(1991) 57-70. Maritime University in Busan, Korea.
[22] Y-D. Choi, J-I. Lim, Y-T. Kim and Y-H. Lee, Performance His research interests include ocean energy, wind energy,
and internal flow characteristics of a cross-flow hydro tur- small hydro power, fluid machinery, PIV and CFD.