TELKOMNIKA, Vol.16, No.1, February 2018, pp.

18~24
ISSN: 1693-6930, accredited A by DIKTI, Decree No: 58/DIKTI/Kep/2013
DOI: 10.12928/telkomnika.v16i1.5919 n 18

A Numerical Modeling for Study Marine Current in the Manado

Bay, North Sulawesi

Parabelem Tinno Dolf Rompas*, Jenly Dyliep Isria Manongko Universitas

Negeri Manado, Tondano, Indonesia
Jl. Kampus FT-Unima Tondano 95618, North Sulawesi, Indonesia, +62431322543 *Corresponding
author, e-mail: parabelemrompas@unima.ac.id

Abstract
This study is investigating about marine currents provided electrical energy through the numerical
model. The objective of this study is to know the power available distributions in the Manado Bay, North
Sulawesi, Indonesia. The Manado Bay was width 2200 m with 79 m of depth. In computation, we are made
grids in x and y horizontal were 7 m respectively, also for z vertical of four layers. The results shown that the
power available distributions in the Manado Bay at 0.1 Sv were 0.00-20.00 kW/m2 when low tide currents and
when high tide currents were 0.00-105 kW/m2. The values will enable for marine currents power plant in the
Manado Bay to future.

Keywords: numerical model, marine currents, the Manado Bay, power available

Copyright © 2018 Universitas Ahmad Dahlan. All rights reserved.

1. Introduction
Numerical methods are one completion to determine the power available in Manado Bay.
Already many researchers are investigating on numerical methods for ocean currents such as
Casulli and Cheng in [1] that study on numerical method in San Francisco Bay, California and
Lagoon of Venice, Italy. They simulated flooding and drying of tidal mud-flats in conjunction by
3D flows. Clement, et al., in [2] that developed numerical models in Bering Sea in North Pacific
which have investigated the short-term marine circulation and flux in a small geographic region.
Zarrati and Jin in [3] developed the mathematical model for 3D simulation into multi-layer model.
Their models were able to predict diverse three-dimensional flow conditions through the velocity
distribution and secondary flows. Rompas and Manongko [4] have studied on velocities of marine
current in Bunaken Strait, North Sulawesi, Indonesia by numerical simulation. Draper, et al., [5]
simulated energy potential of a tidal near a coastal headland using 2D depth averaged numerical
model. Numerical experiments of tidal currents near New River inlet, NC, USA are conducted by
Chen, et al., [6]. The Reynolds-averaged Navier-Stokes (RANS) equations as the basic equations
used to solve numerical modeling [7-9]. The numerical models of tidal currents are developed such
as in a coastal ocean (Maine Gulf, Fundy Bay, Minas Passage, and Minas Basin in North America)
with subgrid approximation [10], in the Gold Coast Seaway area, Gold Coast, Australia [11] with
simulations of flow, wave, and sediment transport. The evolution of tidal creek has studied by Gu,
et al., [12]. Elzalabani, et al., [13] have presented the mathematical modeling and simulation tidal
current energy that can be modeled as a stream of harmonics and applicated in building the tidal
barrage across a bay.
In this study, we have used a numerical model that developed from [1-3]. Also, we want to
investigate the power available distributions in the Manado Bay, North Sulawesi, Indonesia which
influenced by Singkil river, refer to Figure 3.
2. Theory
2.1. Mathematical Model
The governing equations used to get the distributions of the power available from basic
RANS Equations [14]. The model of Equations become:

Received July 6, 2017; Revised October 2, 2017; Accepted January 1, 2018

u u u u η
(1)

u v w g div νeff grad u fcorv

t x y z x

v v v v η
u v w g div νeff grad v fcoru (2)
t x y z y

u v w
0 (3)
x y z

where u (x,y,z,t), v (x,y,z,t), and w (x,y,,z,t) are the velocities in horizontal directions of x, y, and z-
direction respectively, η(x,y,t) is the elevation of free surface, t is the time, νeff is an effective
diffusion taking of account turbulent viscosity and dispersion, νeff ν νt , fcor is the Coriolis
parameter, assumed to be constant, and g is the gravity.

2.1.1. Turbulence Model

The equation of turbulence that used was turbulent viscosity from the mixing-length model
[15, 16] as follow:

1/2 ν lh4

2 u 2 2 vy 2 vx uy 2 4 uz 2

v 2 (4) t x

lv z

where lh and lv are horizontal and vertical mixing length scales respectively.

2.1.2. Boundary Conditions

In the numerical study here, the boundary conditions need to be set as [14]: by the bottom,
by the surface of the water [17], by boundaries which can be vertical impermeable structures (wall),
and by boundaries in the open sea [18]. The power available in the Manado Bay, we can be
obtaining from Equation [19-20]:

P ρ(v)310 3 (5)
where P in kW/m2, v is u2 v2 w2 (m/s) and ρ is density of water (kg/m3).

2.2. Numerical Model

The numerical method that used was semi-implicit finite difference for the 3D in Equation
(1), (2), and (3) was used by: [1], [3], [12], [15], and [21-23].
Figure 1 shows choices adopt in the vertical direction. The velocities are defined on the
edge of the mesh, we guessed virtual meshes to write the limiting conditions with the walls, and we
decorate the free surface with the grid.
The generally, we can be written semi-implicit discretization in Equation (1) and (2) in the
compact matrix form [1] as follow:

(6)
Ain 1/2, j Uin 11/2, j Gin 1/2, j g Δt ηin 11,j ηi,nj 1 ΔZin 1/2, j

Δx
(7)
Ai,nj 1/2Vi,nj 11/2 Gi,nj 1/2 g Δt ηi,nj 11 ηi,nj 1 ΔZi,nj 1/2

Δy

Then, the velocity vertical in Equation (3) becomes:

n 1 wi,nj, k1 1/2 Δzin 1/2, j,kuin 11/2, j,k Δzin 1/2, j,kuin 11/2, j,k

wi,j,k 1/2
Δx
(8)
Δz
i,nj 1/2,kvi,nj 11/2,k Δzi,nj 1/2,kvi,nj 11/2,k

Δy

If we are discretization Equation (5), then the power available becomes:

P ρ (vi,nj, 1 3
k ) 10
3
(9)

n 1 2 2 2
where P is the availability power in the Manado Bay (kW/m2) and vi, j, k u v w is

velocity resultant with u 1 (ui,nj,k 1 uin 11, j,k ),v 1(vi,nj, k1 vi,nj 11,k ) and w 1 (wi,nj,k 1

wi,nj,k 1 1) are scalars,

2 2 2
respectively.
 Figure 1. Meshes and notations for computational (2D and 3D)

3. Method
Figure 2 shows flow chart for solution of a numerical model in calculating the velocities of
u , v and w respectively and the power available in the Manado Bay. The “calculate components of
the velocities (u, v and w) and power available” symbol are show a process for calculating of
horizontal velocities u in Equation 6 and v in Equation 7 with a linear threediagonal, whereas for
calculating velocity vertical w use Equation (8). Finally, calculating the power available where the
calculation used Equation (9). The “n” symbol shows the quantity of calculating with iteration do-
process until maximum iteration (T max). The “T > Tmax” symbol is the process to execute
determination when the iteration has been greater than maximum iteration, if no then process will
be go to “n” for continue to calculate again, and if yes then it go to “finish”.
The position of Manado Bay in Indonesia and numerical area are located in the Sulawesi
Sea with approximately 300 km2 of the area as shown in Figure 3 and width of about 2.2 km
between Bunaken Island and Sulawesi Island, and down to 79 m deep.
We are made two types of simulations in 3D-simulations with one discharge. There are
four layers to deep. In calculation, there are 174 x 318 mesh in x, y directions with Δx = Δy = 7 m.
Also, we are used four vertical layers and the integration time Δt = 0.4 sec as shown in Table 1, and
discharge is 0.1 Sv (1 Sv = 106 m3/s). Figure 4 illustrates the bathymetry 3-D and 2D of the Manado
Bay which used for numerical simulation.
Tide predictions were computed based on the Admiralty method using Harmonic
Constants taken from the Indonesia Sailing Direction and the results of the HydroOceanographic
surveys. Information about tide is needed for the safety of navigation as mention in navigational
Indonesia regulation number 21, 1992 [24]. Hourly heights of tide of 95 stations in Indonesia are
given for whole period in one year. Time used in local standard time. Predictions refer to Chard
Datum of Low Water Spring. Height should be added to charted depths, unless preceded by minus
sign (-) then they should be subtracted. Water heights are given in Meter. It is pointed that though
the prediction in the tide tables, in substance gives the actual movements of tide.

Table 1. Numerical parameter for 3D-simulations

G 9.81 m s-2 Ρ kg/m3
Cz 48 Δx m
τo days Δy m
τi day Δz m
Discharge 0.1 Sv Δt 0.4 sec
 Parameter Value Parameter Value

START

READ DATA

GENERATION OF THE MESH (dz)

GENERATION OF THE INDEX (ivf)

INITIAL CONDITIONS

BOUNDARY CONDITION S

CALCULATE ADVECTIONS IN U AND V

TURBULENCE VISCOSITY

CALCULATE FREE SURFACES

CALCULATE COMPONENTS OF THE VE LOCITIES (U, V AND W) AND

POWER AVAILABLE

PRINT RESULTS

no
T>T max

yes

STOP

Figure 2. Flow chart of a numerical model

 Figure 3. The position of Manado Bay in Indonesia and numerical area

(a) (b)

Figure 4. The bathymetry of the Manado Bay in 3D (a) and 2D (b)

4. Results and Discussion

We can see that the power available distributions when high tide currents (Figure 6b) at in
front of the Singkil river downstream (Figure 4b) where around 80-105 kW/m2 bigger than the
other area in around that of 1-60 kW/m2. It caused by existence of manger and average depth in the
place of ~5 m. Also, in West area, especially at centre area where power availabilities around 16-20
kW/m2. Whereas in North area where the power available still less unless in South area about 30-50
kW/m2. Figure 5 shows the tide predictions for 37 days from on 16 January to 21 February 2014
[24]. Predictions refer to Chart Datum 1.2 m under Mean Sea Level (MSL). The type of tides is
mixed, mainly semidiurnal and some of that are diurnal. If we see in Westside (left hand) where
there are power availabilities biggest around 10-15 kW/m2. When low tide currents (Figure 6. a) at
in front of the Singkil river, East-North, and East-South, where around 15-20 kW/m2. Whereas at
enter channel in Singkil river of 0.5-8 kW/m2. It is caused due to the confluence of the river and the
marine currents, also, the flow of river water to move freely into the sea with an average depth of 2
m.
 Figure 5. Tide predictions of the Manado Bay

(a) (b)

Figure 6. 2D-Simulated distributions of the power available when low tide currents (a) and high
tide currents (b) in the Manado Bay

(a)
(b)
Figure 7. 3D-Simulated distributions of the power available when low tide currents (a) and high
tide currents (b) in the Manado Bay

The distributions of the power available when low and high tide currents in the Manado
Bay (3D-simulation) at discharge 0.1 Sv respectively showed in Figure 7. Tide currents very
influence to the power available which very big at high tide current [25-27]. When low tide
currents, the distributions were 0.00-20.00 kW/m2 and 0.00-105.00 kW/m2 when high tide currents
in the Manado Bay.

4. Conclusion
Study on marine current in the Manado Bay, North Sulawesi, Indonesia through a
numerical model has been successfully accomplished. The maximum of power available at
discharge of 0.1 Sv when low and high tide currents were 20 and 105 kW/m2 respectively. The
results will be enabling to design the turbines that used in the marine current power plant in
Manado Bay in the future.
Acknowledgements
This work is supported by Dirlitabmas, Dirjendikti, Ministry of Education and Culture,
Republic of Indonesia. Authors also acknowledge the Rector of Universitas Negeri Manado,
Indonesia who has proposed this research grant.

References
[1] Casulli V, Cheng RT. Semi-implicit finite difference methods for three-dimensional shallow water flow.
International Journal for Numerical Methods in Fluids. 1992; 15: 629-648.
[2] Clement JL, Maslowski W, Cooper LW, Grebmeier JM, Walczowski W. Ocean circulation and
exchanges through the northern Bering Sea-1979-2001 model results. Deep-Sea Research II. 2005; 52:
3509-3540.
[3] Zarrati AR, Jin YC. Development of a generalized multi-layer model for 3-D simulation of free surface
flows. International Journal for Numerical Methods in Fluids. 2004; 46: 1049-1067.
[4] Rompas PTD, Manongko JDI. Numerical simulation of marine currents in the Bunaken Strait, North
Sulawesi, Indonesia. IOP Conference Series: Materials Science and Engineering. 2016; 128(012003): 1-
7.
[5] Draper S, Borthwick AGL, Houlsby GT. Energy potential of a tidal fence deployed near a coastal
headland. Philosophical Transaction of the Royal Society A. 2013; 371: 1-16.
[6] Chen J–L, Hsu T-J, Shi F. Raubenheimer, B. & Elgar, S. Hydrodynamic and sediment transport
modeling of New River Inlet (NC) under the interaction of tides and waves. Journal of Geophysical
Research: Oceans. 2015: 4028-4047.
[7] Caliskan U, Fuhrman DR. RANS-based simulation of wave-induced sheet-flow transport of graded
sediments. Coastal Engineering. 2017; 121: 90-102.
[8] Zhang C, Zheng J-H, Zhang J-S. Predictability of wave-induced net sediment transport using the
conventional 1DV RANS diffusion model. Geo-Marine Letters. 2014; 34: 353-364.
[9] Tang L, Lin P. Numerical modeling of oscillatory turbulent boundary layer flows and sediment
suspension. Journal of Ocean Engineering and Marine Energy. 2014: 1-12.
[10] Walters RA. A coastal ocean model with subgrid approximation. Ocean Modelling. 2016; 102: 45-54.
[11] Sedigh M, Tomlinson R, Cartwright N, Etemad-Shahidi A. Numerical modeling of the Gold Coast
Seaway area hydrodynamics and littoral drift. Ocean Engineering. 2016; 121: 47-61.
[12] Gu Y, Gong M, Li H. Application of Remote Sensing Axis Line Method in Xiaomiaohong Creek.
TELKOMNIKA Indonesian Journal of Electrical Engineering. 2014; 12(5): 3323-3330.
[13] Elzalabani MM, Fahmy FH, Nafeh AESA, Allam G. Modelling and Simulation of Tidal Current Turbine
with Permanent Magnet Synchronous Generator. TELKOMNIKA Indonesian Journal of Electrical
Engineering. 2015; 13(1): 57-64.
[14] Hervouet JM. Hydrodynamics of free surface flows: Modelling with the finite element method. England:
John Wiley & Sons Ltd. 2007: xiv-341.
[15] Stansby PK. A mixing-length model for shallow turbulent wakes. Journal of Fluid Mechanics. 2003;
495: 369-384.
[16] Cea L, French JR, Vazquez-Cendon ME. Numerical modelling of tidal flows in complex estuaries
including turbulence: An unstructured finite volume solver and experimental validation. International
Journal for Numerical Methods in Engineering. 2006; 67: 1909-1932.
[17] Chen X. A free-surface correction method for simulating shallow water flows. Journal of Computational
Physics. 2003; 189: 557-578.
[18] Treguier AM, Barnier B, De Miranda AP. An eddy-permitting model of the Atlantic circulation:
Evaluating open boundary condition. Journal of Geophysical Research: Oceans. 2001; 106: 1-23.
[19] Hydro BC. Green energy study for British Columbia-phase 2-mainland tidal current energy. Triton
Consultants Ltd.
http://www.llbc.leg.bc.ca/public/PubDocs/bcdocs/357590/environment3928.pdf. (10 February 2012).
[20] Luquet R, Bellevre D, Fréchou D, Perdon P, Guinard P. Design and model testing of an optimized
ducted marine current turbine. International Journal of Marine Energy. 2013; 2: 61-80.
[21] Stansby PK. Semi-implicit finite volume shallow-water flow and solute transport solver with k-Ɛ
turbulence model. International Journal for Numerical Methods in Fluids. 1997; 25: 285-313.
[22] Rodriguez C, Serre E, Rey C, Ramirez H. A numerical model for shallow-water flows: dynamics of the
eddy shedding. WSEAS Transactions on Environment and Development. 2005; 1: 280-287.
[23] Casulli V, Walters RA. An unstructured grid, three-dimensional model based on the shallow water
equations. International Journal for Numerical Methods in Fluids. 2000; 32: 331-348.
[24] Daftar pasang surut (tide tables). Kepulauan Indonesia (Indonesian Archipelago). Jakarta: Dinas Hidro-
Oceanografi TNI AL. 2014.
 [25] Rompas PTD, Taunaumang H, Sangari FJ. A Numerical Design of Marine Current for Predicting
Velocity and Kinetic Energy. Indonesian Journal of Electrical Engineering and Computer Science.
2017; 5(2): 401-409.
[26] Rompas PTD, Taunaumang H, Sangari FJ. A Numerical Model of Seawater Volume and Velocity
Dynamic for Marine Currents Power Plant in the Bangka Strait, North Sulawesi, Indonesia. IOP
Conference Series: Materials Science and Engineering. 2017; 180(012100): 1-7.
[27] Rompas PTD, Sangari FJ, Tanunaumang H. Study on Marine Current with Approach of a Numerical
Model for Marine Current Power Plant (PLTAL) in the Bangka Strait North Sulawesi. Proceedings2016
International Seminar on Application of Technology for Information and Communication, ISEMANTIC
2016. IEEE. 2017: 104-110.

View publication stats

 Numerical simulation of marine currents in the Bunaken
Strait, North Sulawesi, Indonesia

P T D Rompas* and J D I Manongko

Faculty of Engineering, Universitas Negeri Manado, Tondano, North Sulawesi, 95618, Indonesia

*Corresponding author: parabelem_rompas@yahoo.com

Abstract. This study intended for the generation of hydroelectric power at suitable area of the strait
in order to provide electric current to a close environment. The project uses a threedimensional
model of taking flow into account the variation of hydrostatic pressure in the liquid vertical layers.
We brought back to a two-dimensional calculation using the shallow water equations. The
objectives of the study are getting simultaneous obtaining the velocities of currents by the
component of velocities and distributions of the kinetic energy from the numerical results. The
Bunaken strait is 5280 m width for an average depth of 130 m. Numerical calculation is simulated
using horizontal meshes of 60 side meters. The numerical solutions obtained by using a time step
of one second. It found that there was no great difference between 2D and 3D numerical
simulations because the effect of flow velocity in the vertical direction is very small. The numerical
results have shown that the average current velocities when low and high tide currents are 1.46 m/s
and 0.85 m/s respectively. The kinetic energy ranged from 0.01 to 2.54 kW/m 2 when low and high
tide in the Bunaken strait area at discharge of 1 Sv, whereas at discharge 2 Sv, 0.11-17.40 kW/m2
and 0.11-2.77 kW/m2 (when low and high tide currents). These results can used in the design of
turbines for power generation marine currents in the Bunaken strait at depths below 60 meters.

1. Introduction
Construction of electrical power plants in Indonesia is urgently needed to overcome the shortage of
electrical energy to date. One of the components for the construction of the power plant is marine current turbine [8,
9]. Marine current turbine designs requiring variable of current velocity which is directly proportional to the kinetic
energy. Many ways to get the data of current velocity, one way is by numerical model [2]. A numerical model of
marine currents in Bunaken strait used a semi-implicit finite difference method for the numerical solution of three-
dimensional shallow water flows. Several numerical methods with solution of shallow water equations used in
practical applications [1, 2, 3, 5]. Several existing numerical model for two and three dimensional shallow water flow
simulations based on an alternating direction implicit ADI method. In semi-implicit method, only the barotropic
pressure gradient in the momentum equations and the velocity divergence in the continuity equation taken implicitly.
Each time step a linear five-diagonal system solved in new the water surface elevations for the entire domain are the
unknowns. The model is generally explicit with the exception that the vertical eddy viscosity terms discretized
implicitly. In the model formulation, the governing system of equations split into an external and an internal mode
[2]. Momentum exchanges between vertical layers expressed in a set of tri-diagonal matrix equations relating the
discrete horizontal velocities in each vertical level to the gradient of the water surface elevations [5].

Content from this work may be used under the terms of theCreativeCommonsAttribution 3.0 licence. Any further
distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
This paper is more majoring to study the velocities of current and know the availability of kinetic energy in
the Bunaken Strait. In addition, intended for the installation of marine current turbine in the place more adapted strait
in order to provide electrical current to the close environment.

The objectives of the study are getting simultaneous obtaining the velocities of currents by the component of
velocities and distributions of the kinetic energy from the numerical results.
2. Experimental Setup

2.1 Mathematical Model

The model of three-dimensional equations that developed from the Navier-Stokes equations after turbulent
averaging and under assumption that the pressure is hydrostatic [2, 3], the equations as follow:
2 2

u u u u u u u (1)
u v w g x2 y2 z t z f v.
t x y z x
v v v v 2
v 2
v v (2)
u v w g x2 y2 z t z f u.
t x y z y
u v w
0 x (3)
y z
where u(x,y,z,t), v(x,y,z,t), and w(x,y,,z,t) are the velocity components in the directions represent x, y, and z
respectively,
η(x,y,t) is the free surface, t is the time, and t are the eddy viscosity coefficients of horizontal and vertical
respectively, f is the Coriolis parameter, assumed to be constant, And g is the constant gravitational acceleration.

The current computing power does allow taken into the account direct one by using the Reynolds Average
Navier-Stokes Equations (RANS) [1].

2.1.1. Turbulence model.

A formula for turbulent viscosity is the standard form as defined:

4 u2 v2 v u2 4 u2 v2 (4)

t lh 2( x) 2( y) ( x y) lv ( z) ( z)

Where: lv = κ(z-zb), for (z-zb)/h < λ/κ; lv = λh, for λ/κ < (z-zb)/h < 1, for lh = β lv, for the horizontal length
scale is larger, κ is the von Karman's constant (κ = 0.41), λ is a constant (λ = 0.09); (z-zb) is the distance from the
wall, h is the boundary layer thickness assumed to be equal to the water depth,
lv and lh are the vertical and horizontal length scales,
In addition, the constant β has to be determined from comparison with experiment.

2.1.2. Boundary condition.

Some types of boundary conditions are required as the boundary conditions at the free surface are specified
by the prescribed wind stresses of directions x and y, and a slip boundary u z/ v z/ 0. At the bottom
stress can be related to the turbulent law of the wall, a drag coefficient associated with using a Chezy formula [2].
Velocity on a solid wall is a no-slip condition, and on the open boundary, we used principally two condition, the first
is Neumann method and the second is a condition radiation which derived from Orlanski’s algorithm that developed
by Treguier et al. [4].
The kinetic energy in the Bunaken strait (PA), we can be obtaining from equation [6]:
3 3
P1
PA v10 (5)
A2
Where PA in kW/m2, v is the velocity resultant of marine current (m/s) and ρ is seawater density (kg/m3).
 2.2 Numerical Model
Semi-implicit finite difference method for the numerical solution of the three-dimensional Equation 1 and
Equation 2 used by Casulli & Cheng [2], and Stansby [3] in the computation of shallow water flows. The free surface
flow equations can derive in which the gradient of surface elevation in the momentum equations and the velocity can
discretized implicitly. The convective, Coriolis and horizontal viscosity terms in the momentum equations discretized
explicitly, but in order to eliminate a stability condition due to the vertical eddy viscosity, the vertical mixing terms
discretized implicitly.

(a) (b)
Figure 1. Schematic diagram of computational mesh and notations

Figure 1 shown that a spatial mesh which consists of rectangular cells of length Δx, width Δy and height Δzk
is introduced. Each cell is numbered at its centre with indices i, j and k. Figure 1 (a) show that the discrete u-velocity
is then defined at half-integer i, j and k; v is defined at integers i, k, and half-integer j; w is defined at integers i, j, and
half-integer k. Then figure 1 (b) show that η is defined at integers i and j. The water depth h(x,y) is specified at the u
and v horizontal points. So that a general semi-implicit discretization of the momentum equations in Equation 1 and
Equation 2 can be written in the more compact matrix form as

n n 1 n n 1n 1 n
t (6)
A Ui 1/2, j i 1/2, j Gi 1/2, j g i 1, j i j, ΔZi 1/2, j x

n n 1 n t n 1 n 1 n (7)
A Vi j, 1/ 2 i j, 1/ 2 Gi j, 1/ 2 g i j, 1 i

j, ΔZi j, 1/ 2 y

n 1
Equation 6 and Equation 7 are linear tri-diagonal systems. For determine the free surface i j , can be

written in the matrix notation form

 n 1 n 1 t T n 1 T n 1

i j, , i j, , ΔZi 1/ 2, j Ui 1/ 2, j ΔZi 1/ 2, j Ui 1/ 2, j

x (8)

t T n 1 T n 1

ΔZi j, 1/ 2 Vi j, 1/ 2 ΔZi j, 1/ 2 Vi j, 1/ 2 y

The available energy that investigated in this study is the kinetic energy (kW/m2). The first, we will back at
the equation of the kinetic energy which is equation of the marine current power in the Bunaken strait can be
discretized from Equation 5 becomes:
n 1 3 3
P1 (9)
PA (vi j k, , ) 10
A2

3. Results and Discussion

3.1. The Domain Presentation of the Bunaken Strait

The Bunaken strait is located between the Pacific ocean and the Sulawesi sea (Celebes sea) whose area is
approximately 200 km2 (Figure 2), with a minimum width between Bunaken island and Sulawesi island about 5.28
km and the average depth of 130 m.

The three-dimensional current circulation in the Bunaken strait is simulated using the present model with
174 x 318 x 4 finite difference meshes of equal Δx = Δy = 60 m and Δx = 20 m. The numerical solutions used an
integration time Δt = 1 sec and inlet volume transports (discharges) are 1 Sv to 2 Sv (1 Sv = 10 6 m3s-1). Figure 3
illustrates the bathymetry (a) and the meshes (b), the 3D of the Bunaken strait used for numerical simulation. The
water depth distributions show the complex areas.

Currents in the Bunaken strait closely related with ocean currents and Indonesian throughflow. According to
Brown et al. [7] that the global wind system would be if the Earth were completely covered with water, and the
atmospheric circulation transports heat from low to high latitudes and the same is true in the oceans, where surface
currents warmed in low latitudes carry heat polewards, while currents cooled at high latitudes flow equatorwards.
Marine currents in the Bunaken strait consist of two i.e. high tide currents and low tide currents. In the morning,
occurred high tide currents where input currents from from Maluku Sea (see right of Figure 2). The currents then go
out to Manado bay and then they go out to Sulawesi Sea. In the evening, occurred low tides currents that are inversed
on obtaining of high tide currents. The currents obtain from Sulawesi Sea and Manado bay. The currents then go
inside pass in the left of Bunaken strait. Then the currents go out to the right of Bunaken strait and then go out to
Maluku Sea.

Sulawesi Sea

Maluku Sea

Bunaken Island

Bunaken Strait
Sulawesi
Manado Bay Island
 (a)

(b)
Figure 3. The 3D of the Bunaken strait with water depth: (a) Bathymetry and (b) Meshes.

Figure 2. Location of the Bunaken strait in North Sulawesi, Indonesia.

In 2D simulations, we made two type of simulations with two variations of discharge. In first simulation, we
have conducted when low tide current where each simulation has considerate with constant discharge inside. In
second simulation, when high tide currents with condition discharge same as in first simulation. Parameter of
discharge varies from 1 to 2 Sv with classifications are 1 Sv and 2 Sv.

Like in 2D simulations, 3D simulations we also have made two type of simulations with two variations of
discharge. In 3D simulations, there are 4 layers of seawater column which each layer depth of 20 m (Δz),
respectively. Whereas in 2D simulations only one layer of seawater column which layer depth is maximum depth.

3.2. Numerical Simulation of Marine Currents in the Bunaken Strait

The current velocity is a key factor in the design of a marine current power plant, since it sets the limits for
both the power output as well as the forces acting on the turbine and support structures [8]. Figure 4 shows simulated
of velocity component distributions at seawater column when low (a) and high (b) tide currents. Generally, when low
tide currents which water enters from left side section and then flows go to section of right side. A small part to top
side section which previous rotate form two eddies like elliptic diameter at centre. The average velocity at enter of
Bunaken strait is 1.46 m/s (figure 4a). On the contrary, when high tide current (figure 4b), current enters from right
side section go to section of left side and a small part to top side section which previous happened eddy is very small
at center east area near Bunaken island. The average velocity at enter of Bunaken strait is 0.85 m/s.

(a) (b) (c) (d)

Figure 4. Simulated of velocity component (2D) and kinetic energy (3D) distributions at seawater column
when the tide currents of low (a and c) and high (b and d) at discharge of 1 Sv.
 (a) (b) (c) (d)
Figure 5. Simulated of velocity component (2D) and kinetic energy (3D) distributions at seawater column
when the tide currents of low (a and c) and high (b and d) at discharge of 2 Sv.
Simulated (2D) of velocity component distributions at seawater column when low (a) and high (b) tide
currents shown Figure 5. Water flows when tide currents but the average velocities are different. When low tide current
and high tide current at enter of Bunaken strait are 2.6 m/s and 2.0 m/s respectively.

The distributions of the kinetic energy when the tide currents of low (c) and high (d) in the Bunaken strait at
discharge of 1 and 2 Sv respectively showed in figure 4 and figure 5. Discharge influence to the kinetic energy is
very big where ever greater of discharge then ever greater also kinetic energy. Discharge of 1 Sv shows that there are
about 0.05-2.54 kW/m2 (when low tide currents) and 0.01-0.38 kW/m2 (when high tide currents) kinetic energies,
whereas at discharge of 2 Sv, 0.41-17.40 kW/m2 (when low tide currents) and 0.11-2.77 kW/m2 (when high tide
currents) available in the Bunaken strait.

We can see generally that the values of current velocities and kinetic energy are bigger when low tide
currents. That's because the cross sectional area of water that entered the Bunaken strait is smaller when low tide
currents than when high tide currents.

4. Conclusions
Numerical simulation of marine currents in the Bunaken strait, North Sulawesi, Indonesia has presented.
The values of the kinetic energy obtained by calculations could be enabling to choose a suitable place for installing
the turbines adapted well for a future undersea electricity power plant in the Bunaken strait. The numerical results
have shown that the values of current velocities and kinetic energy are bigger when low tide currents. At the
discharge of 1 Sv, the average current velocities are 1.46 m/s (when low tide currents) and 0.85 m/s (when high tide
currents), the distributions of the kinetic energy are 0.05-2.54 kW/m2 (when low tide currents) and 0.01-0.38 kW/m2
(when high tide currents). Whereas at discharge of 2 Sv, the average current velocities are 2.6 m/s (when low tide
currents) and 2.00 m/s (when high tide currents), the distributions of the kinetic energy are 0.41-17.40 kW/m2 (when
low tide currents) and 0.11-2.77 kW/m2 (when high tide currents). These results can used in the design of turbines for
power generation marine currents in the Bunaken strait at depths below 60 meters.

References
[1] Broomans P 2003 Numerical Accuracy in Solution of the Shallow-Water Equations (Master thesis, TU Delft
& WL, Delft Hydraulics)
[2] Casulli V and Cheng R T 1992 Semi-implicit finite difference methods for three-dimensional shallow water
flow International Journal for Numerical Methods in Fluids 15 629-648
[3] Stansby P K 1997 Semi-implicit finite volume shallow-water flow and solute transport solver with k-ε
turbulence model International Journal for Numerical Methods in Fluids 25 285313
[4] Treguier A M, Barnier B, and De Miranda A P 2001 An eddy-permitting model of the Atlantic circulation:
Evaluating open boundary condition J. Geophy. Res. Oceans 106 (C10) (2211522129) 1-23
[5] Zarrati A R and Jin Y C 2004 Development of a generalized multi-layer model for 3-D simulation of free
surface flows Int. J. Numer. Meth. Fluids 46 1049-1067
[6] Luquet R, Bellevre D, Fréchou D, Perdon P, and Guinard P 2013 Design and model testing of an optimized
ducted marine current turbine International Journal of Marine Energy 2 61-80
[7] Brown E, Colling A, and Park D 2001 Ocean circulation (second edition, The open university, England:
Walton Hall, Milton Keynes, MK7 6AA and Butterworth-Heinemann)
[8] Thomas K 2007 Low Speed Energy Conversion from Marine Currents (Ph.D. Thesis, Acta Universitatis
Upsaliensis, Uppsala: Digital comprehensive summaries of Uppsala Dissertations from the Faculty of Science
and Technology)
[9] Fraenkel P L 1999 New Development in Tidal and Wave Power
Technologies (Presented at Towards a renewable future: silver jubilee
conference, Brighton, UK)

Acknowledgments
The authors wish to express their appreciation to Dirlitabmas, Dirjendikti, Ministry of Education and
Culture of the Republic of Indonesia which had finance all of the research activities, and Rector of Manado
State University, Indonesia who has proposed
 Study on Marine Current with Approach of a
Numerical Model for Marine Current Power Plant
(PLTAL) in the Bangka Strait North Sulawesi
Parabelem Tinno Dolf Rompas education of
information and communication technology universitas
negeri manado
Tondano, Indonesia
parabelem_rompas@yahoo.com
Heindrich Tanunaumang
education of physics
universitas negeri manado
Tondano, Indonesia
hein.taunaumang@yahoo.com
Abstract—Study on marine current with approach of a numerical
model for marine current power plant (PLTAL) in the Bangka
strait North Sulawesi has been investigated. Construction of
power plant is needed to overcome the shortage of electricity in
North Sulawesi. Before building the electrical energy it would
require a feasibility study which aim to ensure the certainty of
the construction of power plant. One of them is through the study
of marine currents in the design of a numerical model. The
objective of this investigates for long-term is to get a profile of
marine current turbines as one component in the construction of
marine current power plant in the Bangka strait. Specific targets
to achieve are to get the first; data such as tide, sea water and air
temperature on the surface, the wind speed above sea level, a
map of the Bangka Strait and bathymetry, the second; a design of
numerical model and kinetic energy distributions. The method
used was initially literature study, survey in the research
location, measurements of data such as tide, temperatures of sea
water and the air above the surface, wind speed above sea level,
bathymetry of the Bangka strait, finally are the analysis of data
measurements and design of a numerical model in the form of
numerical program. The results showed that the data tide from
January 16 until February 21, 2016 the maximum and minimum
of 2.4 m and 0.3 m respectively that oscillates at datum line of 1.2
m. Numerical program developed from the semi implicit finite
difference method for shallow water in two and three dimensions
by the basis algorithm that consists of three fractional steps are
advection step, diffusion step, and pressure-continuity step. The
numerical program will be a product in analyzing potential
kinetic energy as the prime mover of turbines for marine current
power plant in the Bangka strait.
Keywords—numerical modeling; numerical simulation; marine
current turbines; PLTAL
Ferry Jhony Sangari
education of electrical
engineering universitas negeri
manado
Tondano, Indonesia ferry_sangari@yahoo.com
I. INTRODUCTION
Development of electricity power plant is a part of a
whole development in North Sulawesi because the electricity
consumption would go up along with the increasing of public
activity and a prosperous people (as the economy has grown
rapidly in North Sulawesi the last years, so has the demand for
electricity). Public utilizes electricity for many purpose such as
household requirement as well as economics trade. Therefore
supplying adequate amount of electricity and existence of
continuities electricity power should help to maintain conducive
social and economic activity, and to motivate public economic
growth. When the electricity is insufficient, the electricity
power will be put out to balance the supply for consumer.
Putting out of electricity has been occurring several times in
North Sulawesi, this case has influenced by the development
and investment.
The ideal locations for power station installation of the
current energy have velocities of current two directions
(minimum bidirectional) 2 m/s [12]. The ideal is 2.5 m/s or
more. One way (river/current of geotropic) is minimum 1.21.5
m/s. The deepness not less than 15 m and the most at 40 or 50
m. Close to coast so that energy can be channeled with low
expense. They have add for wide that more than one turbine can
be attached, not sea transport and the fish arrest area.
A numerical model of marine currents in Bangka strait
used a semi-implicit finite difference method for the numerical
solution of three-dimensional shallow water flows. Several
numerical methods with solution of shallow water equations are
used in practical applications [3], [4], [7]. In semi-implicit
methods only the barotropic pressure gradient in the momentum
equations and the velocity divergence in the continuity equation
are taken implicitly. Each time step a linear five-diagonal
system is solved in the new water surface elevations for the
entire domain are the unknowns. The model is generally explicit
with the exception that the vertical eddy viscosity terms are
discretized implicitly. In the model formulation the Free surface equation
governing system of equations is split into an external
and an internal mode [2]. Momentum exchanges between
vertical layers are expressed in a set of tri-diagonal matrix ∂η+ ∂ η + ∂ η
− h udz ∂ y − h vdz =0 (4)
equations relating the discrete horizontal velocities in
each vertical level to the gradient of the water surface ∂t ∂x
elevations [11]. A formal expression for the solution of Where, νeff is an effective diffusion taking of account turbulent
these tri-diagonal systems can be written in terms of the viscosity and dispersionνeff = +ν νt . This effective diffusion is
barotropic pressure gradient. Substituting the formal given by means of a model of turbulence adapted to the
solutions into the vertically integrated continuity equation problem considers. Equation (1) to (4) will be those considered
gives rise to a linear five-diagonal system whose only in the continuation of the report.
unknowns are the water surface elevation over the Power is just energy divided by time, so the power
domain of interest. Such a system is symmetric and available from the seawater current [13]-[17] can be expressed
positive definite and can be solved uniquely and as:
efficiently by using a conjugate gradient method. By
direct substitution of the barotropic pressure gradient
P = Ek = 1ρ.v .3 A (5) dt
known at the advanced time level, the horizontal velocity
for each vertical layer can be computed. The vertical 2
velocity component can be found by integration of the Where, P is the power available from the seawater current in
continuity equation. This paper is more majoring to study Watt.
the velocities of current and know the availability of In this study we will calculate the power of marine
kinetic energy in the Bangka Strait. This study is intended current in the Bangka strait per unit cross-sectional area (m2),
for the installation of turbines in the place more adapted thus, from equation 5 we can be obtain:
strait in order to provide electrical current to the close
environment. PA = =P 1ρv 103 −3 (6)
The objective of this investigates for long-term is A2
to get a profile of marine current turbines as one component Where, PA is the power per cross-sectional area in kW/m2 and v
in the construction of marine current power plant in the is the velocity resultant of marine current that defined as v = u v
Bangka strait. Specific targets to achieve are to get the first;
w2 + +2 2 with u , v and w respectively are scalars of the
data such as tide, sea water and air temperature on the
surface, the wind speed above sea level, a map of the velocitiesu , v and w respectively, and ρ = 1024 kg/m3 [18] (at
Bangka Strait and bathymetry, the second; a design of 20 (C) and salinity of 34).
numerical model and kinetic energy distributions. B. Turbulence model
II. MODEL EQUATIONS A formula for turbulent viscosity is the standard form
as defined from the mixing-length model with assuming (∂w / ∂z )
A. Basic equations
Under the assumptions of hydrostatic pressure,
2
<< (∂u / ∂x ) 2 + (∂v / ∂y ) 2 , ∂w / ∂y << ∂v / ∂z
and
and by using the decomposition of preceding Reynolds, the for shallow water flows where vertical velocity w
∂w / ∂x << ∂u / ∂z
realized average Navier-Stokes equations are written [6]: is small was used by Stansby [9] and Cea [4]. The eddy
Continuity equation viscosity is computed at each point from the horizontal and
vertical component velocity gradients and length scales for
∂u + ∂v + ∂w =0 (1) horizontal and vertical motion, giving a formula for turbulent
∂x ∂y ∂z viscosity as:

∂u
∂t
∂ ∂ ∂
+ u u + v u + w u =− g
∂x ∂y ∂z
∂η
∂x
( )
+ divν eff gradu() + f cor v (2) υt =
+∂∂uy)2
lh4
+lv4
2(∂∂ux)2 +2(∂∂yv)2 +(∂∂vx
(∂∂uz)2 +(∂∂vz)2 (7)
∂v
∂t
∂ ∂ ∂
+ u v + v v + w v =− g
∂x ∂y ∂z
∂η
∂y
( )
+ div ν eff gradv() − f cor u (3)

Momentum equation for lv = κ(z-zb), for (z-zb)/h < λ/κ; lv = λh, for λ/κ < (z-zb)/h < 1;
and lh = β lv, for the horizontal length scale is larger; where κ is
the von Karman's constant (κ = 0.41), λ is a constant (λ = 0.09),
(z-zb) is the distance from the wall, h is the boundary layer
thickness assumed to be equal to the water depth, lv and lh are
the vertical and horizontal length scales, and the constant u in++11/ 2 , j M,
β has to be determined from comparison with experiment.
n+1
C. Boundary conditions
For the problem studied in this paper, some u i + 1 / 2 , j M, − 1 n+

types of boundary conditions are required. These are 1 = n+1

imposed as follows: (i) the boundary conditions at the
free surface are specified by the prescribed wind stresses U i+1/2,j u i + 1 / 2 , j M, −2 ,
of directions x and y, and a slip boundary∂ ∂ =∂ ∂ =u z/ v z/ :
0 ; (ii) the boundary conditions at the bottom stress can be
related to the turbulent law of the wall, a drag coefficient
associated with using a Chezy formula [2]; (iii) the u in++11/ 2 , j m,
boundary conditions for velocity on a solid wall is a no-
slip condition [6], and on the open boundary, we used
principally two condition, the first is Neumann method
and the second is a condition radiation which derived
from Orlanski’s algorithm [10].

III. NUMERICAL MODEL

Semi-implicit finite difference method for the
numerical solution of the three-dimensional in (1) to (4)
was used by Casulli & Cheng [2], Stansby [8], and Chen
[5] in the computation of shallow water flows. Equation (2)
and (3) will be derived in which the gradient of surface
elevation in the momentum equations and the velocity in
the free surface in (4) will be discretized implicitly. The
convective, Coriolis and horizontal viscosity terms in the
momentum equations will be discretized explicitly, but in
order to eliminate a stability condition due to the vertical
eddy viscosity, the vertical

mixing terms will be discretized implicitly.

Fig. 1. Schematic diagram of computational mesh and notations

Fig. 1 shown that a spatial mesh which consists of

rectangular cells of length Δx, width Δy and height Δzk is
introduced. Each cell is numbered at its centre with indices
i, j and k. The discrete u-velocity is then defined at half-
integer i, j and k; v is defined at integers i, k, and half-
integer j; w is defined at integers i, j, and half-integer k.
Then η is defined at integers i and j. The water depth h(x,y)
is specified at the u and
 v in, +j 1+1 / 2 ,M

+1 ΔzM
v
in, j +1 / 2 ,M −1

− Δν ν Δ+ Δ +( 2 2
)
n +1 = i jn, +1+1 / 2 ,M − 2 ΔzM −1
V i j, +1 / 2 v , 0 Δzmm−1−1/2/2t Δ +zm Δmz+m1/+21/2t g t u vCz2
: Δ =Z ΔzM −2

n +1 :
v i , j +1 / 2 ,m Δzm Where m and M denote the k-index of the bottom and the top
,

Δz M Fu in+1 / 2, j M, +Δtτxw

Δz M −1 Fu in+1 / 2, j M, −1
n = Δ
n

G i+1 / 2, j z M − 2 Fu i+1 / 2, j M, − 2
:

Δz Fum in+1 / 2, j m,

,
Δz M Fv i jn, +1 / 2,M +Δtτyw

Δz M −1 Fv i jn, +1 / 2,M −1

G i jn, +1 / 2 = Δz M − 2 Fv i jn, +1 / 2,M − 2

:

Δz Fvm i jn, +1 / 2,m

Δ +νM−1/2Δt − ΔνM−1/2 t

zM ΔzM−1/2 ΔzM−1/2

= − ΔνM−1/2 t Δ +zM νM−1/2Δt+νM−3/2Δt − ΔνM−3/2 t

0

A ΔzM−1/2 ΔzM−1/2 ΔzM−3/2 ΔzM−3/2

: : : : is the Chezy's friction
v horizontal points. So that a general semi-implicit finite difference stencil respectively, Cz
discretization of the momentum equations in (2) and (3) can τ w τw
be written in the more compact matrix form as coefficient, x and y are wind stresses, and F is non-linear

t
(
A in+1/ 2, j U in++11/ 2, j =G in+1/ 2, j − g Δ η ηin++11, j − i jn,+1 ΔZ in+1/ 2, j ) finite difference operator and an explicit.
Δx (8) Equation (8) and (9) are linear tri-diagonal systems. For

( )
A i jn, +1/ 2 Vi jn, ++11/ 2 =G i jn, +1/ 2 − g Δt η ηi jn,++11 − i jn,+1 ΔZ i jn, +1/ 2 (9) determine the free surface ηi jn,+1 as in (4) can be written in the
Δy
matrix notation form

where U, V, ΔZ, G and A are defined as: η ηn+1 = n+1 −ΔΔt [( )T n+1 −( )T n+1 ]
i j, , i j, , i+1/ 2, j Ui+1/ 2, j ΔZi−1/ 2, j Ui−1/ 2, j ΔZ

x (10)

−ΔΔt [(ΔZ i j, +1/ 2 )

T Vi jn, ++11/ 2 −(ΔZi j, −1/ 2 )T Vi jn, +−11/ 2 ]y
 The available energy that investigated in this study is the
START
available power per m2 (kW/m2). The first, we will back at the
equation of the available power which is equation of the marine
READ DATA
current power in the Bangka strait can be discretized from (6)
become:

GENERATION OF MESH = =P 1ρ n+1 3

−
3
(12)
PA (vi j k, , ) 10
GENERATION OF THE INDEX
A2
Where PA is the marine current power (kinetic energy) in the
INITIAL CONDITIONS Bangka strait in kW/m2 and vi j kn, ,+1 = u 2 + +v2 w 2 is velocity

1
with ,
n resultant 2 u = (ui j kn, ,+1 +uin++1, ,1j k ) v
1 1
= (vi j kn, ,+1 +vi jn, ++11,k ),
2
BOUNDARY CONDITIONS and 2
w = (wi j kn, ,+1 +wi j kn, ,+1 +1) are scalars, respectively.
CALCULATE ADVECTIONS IN U & V

IV. RESEARCH METHOD

TURBULENCE MODEL The method used was initially literature study; survey in the
research location; measurements of data such as tide (January
and February, 2016), temperatures of sea water and the air
CALCULATE FREE SURFACE above the surface, wind speed above sea level; bathymetry of
the Bangka strait; finally are the analysis of data measurements
and design of a numerical model in the form of numerical
CALCULATE VELOCITIES (U,V,W) AND program.
KINETIC ENERGY Fig. 2 shows steps of a numerical model in calculating the
velocities ofu , v and w respectively and the power of marine
PRINT RESULTS current in the Bangka strait per cross-sectional area.

V. RESULTS AND DISCUSSION

no The Bangka strait is located between the Pacific Ocean
T>TMAX
and the Sulawesi sea whose area is approximately 200 km2 (Fig.
yes
4), with a minimum width between Sahaong foreland (in
FINISH Bangka island) and Mokotamba foreland (in Likupang town)
about 5.5 km and down to 69 meters deep (the average depth of
Fig. 2. Flow chart of a numerical model 40 m).
The three-dimensional current circulation in the Bangka strait
is simulated using the present model with a 174 x 318 finite
The vertical component of the velocity w at the difference mesh of equal Δx = Δy = 60 m. The numerical
new time level can be discretized from the continuity as in solutions have been achieved using four vertical layers and an
(1) becomes: integration time Δt = 1 sec, and inlet volume transports at
Δn n+1 −Δ n n+1
sections A and B (see Fig. 4) are 0.025 Sv, 0.1 Sv, 0.3 Sv and
wi j kn, ,+1 +1/ 2 = wi j kn, ,+1 −1/ 2 − zi+1/ 2, ,j kui+1/ 2, ,j k zi−1/ 2, ,j kui−1/ 2, ,j k 0.5 Sv.
3

Δx (11)
2.5

Δzi jn, +1/ 2,kvi jn, ++11/ 2,k −Δzi jn, −1/ 2,kvi jn, +−11/ 2,k 2

−
Δy 1.5

1
Where, k=m, m+1...M, and the no-flux condition across the
bottom boundary is assumed by taking wi j mn, ,+1 −1/2 = 0. 0.5

0
 0 24 48 72 96 120 144 168 192 216 240 264 288 312 336 360 384 408 432 456 480 504 528 552 576 600 624 648 672 696 720 744 768 792 816 840 864 888

Time (hour)
Discharge variable Δt 1 sec
Fig. 3. Tides measurement results of the Bangka strait Twater 20 C Tair 29 C

Fig. 3 shows a result of the tide measurements in The distributions of the available power per m2 (kinetic
the Bangka strait from January 16 to February 21, 2016. energy) when low tide currents (3D-simulation) shown in Fig. 6.
The types of tides are mixed type, in particular Discharge influence to the available power is very big where ever
semidiurnal type and diurnal type was only occurred on greater of discharge then ever greater also power availability like
February 9. The tidal range variations were taken place in 2D-simulation. At discharge of 0.025 Sv (a) shows that there
between 0.3 m to 1.9 m. The tidal period variations were are about 1.5-5 W/m2 available in around section A (see Fig. 4),
between 10 h to 20 h. The maximum and minimum of 2.4 whereas 5-350 W/ m2 at 0.1 Sv (b), 0.510 kW/ m2 at 0.3 Sv (c)
m and 0.3 m respectively that oscillates at datum line of and at 0.5 Sv (d) available of 1-45 kW/m2 which is maximum
1.2 m. The tides data were obtained by direct monitoring discharge.
of water level (1 h intervals) using a tide gage. The first Also, when high tide currents in Fig. 7, we found
day measurement on January 16, 2016 was started at around section A where the power availabilities per m2 are
01.00 am until 00.00 pm. Measurements of tide on maximal. Generally, there are about 2-9 W/m2 at 0.025 Sv (a),
second days until days 37th were performed as the first 5-550 W/m2 at 0.1 Sv (b), 0.5-16 kW/m2 at 0.3 Sv (c) and 1-77
day measurement. The results of tide measurement on kW/m2 at 0.5 Sv (d) power availabilities per m2 in the Bangka
January 18, 2016 which lower low water at level 0.3 m strait which the values are bigger than in Fig. 6. We also can see
and highest high water at level 2 m, while higher low that the two when low and high tide currents where can be
water at 0.6 m and lower high water at 1.8 m. The concluded that biggest values are at section A.
variation tidal range was obtained at
1.2 m to 1.7 m. The minimum tidal range was occurred at
03.00 am to 08.00 am at level 1.2 m and a maximum that Philippines Pacific Ocean
occurred at 02.00 pm to 08.00 pm at level 1.7 m. The
Sulawesi
variation tidal period was taken place between 11 to 12
Sea
hours. The minimum wave period was occurred during 11
Sulawesi Island
hours at 03.00 am to 02.00 pm and a maximum was
occurred during 12 hours at 08.00 am to 08.00 pm. I N D O N E S I A
In the 3D-simulations, we also have made two
types of simulations with four variation of discharge. The
Indian Ocean
first type, we also have conducted when low tide current Australia
where each simulation has considerate with constant
discharge inside. In second type, when high tide currents
with same condition discharge as in first simulations. Sulawesi Pacific Ocean
Parameter of entry discharge, we also have made varies Sea Talisei
from 0.025 Sv to 0.5 Sv with classifications are 0.025 Sv, Island
0.1 Sv, 0.3 Sv and 0.5 Sv (1 Sv = 1x10 6 m3/s). For the other C
parameter, we can see in table 1. Measurement results in
the area of numerical such as temperatures of sea water B
(Twater) and the air above the surface (T air) of 20 C and 29 C Gangga Bangka
respectively. Island Island
Fig. 5 illustrates the bathymetry of the Bangka
strait used for numerical simulation. The water depth A (a)
distributions show the complex areas where maximum D Bangka Strait
depth of 69 m (between Bangka island and Likupang
town). Sulawesi Islan
d * Likupang
t
TABLE I. NUMERICAL PARAMETER FOR 3D-SIMULATION
Fig. 4. Location of the Bangka strait in Indonesia and numerical area
Parameter Value Parameter Value
g 9.81 m s-2 ρseawater 1024 kg/m3
Cz 48 Δx 60 m
τo 2 days Δy 60 m
τi 1 day Δz 20 m
 Fig. 5. Bathymetry of the Bangka strait

a b

c
d

Fig. 6. Simulated (3D) distributions of the available power per m2 at seawater column of 20 m when low tide currents at (a) discharge 0.025 Sv, (b) discharge 0.1
Sv, (c) discharge 0.3 Sv and (d) discharge 0.5 Sv.
 a
b

c d

Fig. 7. Simulated (3D) distributions of the available power per m 2 at seawater column of 20 m when high tide currents at (a) discharge 0.025 Sv, (b) discharge 0.1
Sv, (c) discharge 0.3 Sv and (d) discharge 0.5 Sv.
 The results showed that the numerical program will be a product in analyzing potential kinetic energy
as the prime mover of turbines for marine current power plant in the Bangka strait.

VI. CONCLUSIONS
A numerical semi-implicit finite difference models for the study marine currents in the Bangka Strait
has been presented. The numerical program will be a product in analyzing potential kinetic energy as the prime
mover of turbines for marine current power plant in the Bangka strait. When low tide currents, available from
0.5 W/ m2 until 45 kW/m2 and from 0.5 W/m2 until 77 kW/m2 at high tide currents. The values obtained by
calculations will be enabling to choose a suitable place for installing the turbines adapted well for a future
undersea electricity power plant in the Bangka strait.

Acknowledgment
The authors wish to express their appreciation to Kemristekdikti of the Republic of Indonesia which
had financed all of the research activities, and Rector of Manado State University, Indonesia who has proposed
research grant.

References
[1] P. Broomans, “Numerical accuracy in solution of the shallow-water equations,” M.S. thesis, TU Delft & WL, Delft Hydraulics,
2003.
[2] V. Casulli and R.T. Cheng, “Semi-implicit finite difference methods for three-dimensional shallow water flow,” International
Journal for Numerical Methods in Fluids, vol. 15, pp. 629-648, 1992.
[3] V. Casulli and R.A. Walters, “An unstructured grid, three-dimensional model based on the shallow water equations,” International
Journal for Numerical Methods in Fluids, vol.32, pp. 331-348, 2000.
[4] L. Cea at al., “Numerical modelling of tidal flows in complex estuaries including turbulence: An unstructured finite volume solver
and experimental validation,” International Journal for Numerical Methods in Engineering, vol. 67, pp. 1909-1932, 2006.
[5] X. Chen, “A free-surface correction method for simulating shallow water flows,” Journal of Computational Physics, vol. 189, pp.
557-578, 2003.
[6] J.M. Hervouet, Hydrodynamics of free surface flows: Modelling with the finite element method, John Wiley & Sons, Ltd., Englang:
cop, ISBN 978-0-470-03558-0 (HB), 2007, pp. xiv-341.
[7] C. Rodriguez at al., “A numerical model for shallow-water flows: dynamics of the eddy shedding,” WSEAS Transactions on
Environment and Development, vol. 1, pp. 280-287, 2005.
[8] P.K. Stansby, “Semi-implicit finite volume shallow-water flow and solute transport solver with k-ε turbulence model,” International
Journal for Numerical Methods in Fluids, vol. 25, pp. 285-313, 1997.
[9] P.K. Stansby, “A mixing-length model for shallow turbulent wakes,” Journal of Fluid Mechanics, vol. 495, pp. 369-384, 2003.
[10] A.M. Treguier et al. “An eddy-permitting model of the Atlantic circulation: Evaluating open boundary condition,” J. Geophy. Res.
Oceans, 106 (C10): 22115-22129, pp. 1-23, 2001.
[11] A.R. Zarrati and Y.C. Jin, “Development of a generalized multi-layer model for 3-D simulation of free surface flows,” Int. J. Numer.
Meth. Fluids, vol. 46, pp. 1049-1067, 2004.
[12] P.L. Frankael, “Power from Marine Currents, Proceedings of the Institution of Mechanical Engineers,” Part A: J. Power and Energy,
vol. 216, No. 1, pp. 1-14, 2002.
[13] A.S.Bahaj, and L.E. Myers, “Fundamentals Applicable to the Utilisation of Marine Current Turbines for Energy Production,”
Renewable Energy, vol. 28, pp. 2205-2211, 2003.
[14] BC Hydro. (2016, January 20). Green Energy Study for British Columbia-Phase 2- Mainland Tidal Current Energy [Online].
Available: http://www.llbe.leg.bc.ca/public/PubDocs/bcdocs/357590/environment3
928.pdf
[15] L. Myers and A.S. Bahaj, “Simulated Electrical Power Potential Harnessed by Marine Current Turbine Arrays in the Alderney Race,”
Renewable Energy, vol. 30, pp. 1713-1731, 2005.
[16] (2016, January 18). Ocean Energy Technology, Priority Projects for Development of the New and Renewable Energy in China
[Online]. Available: http://www.newenergy.org.cn
[17] K. Thomas, “Low Speed Energy Conversion from Marine Currents,” Ph.D. dissertation, Acta Universitatis Upsaliensis, Digital
Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology, Uppsala, ISSN 1651-6214, ISBN
978-91-5547063-0, 383, pp. 68, 2007.
[18] (2016, January 25). Properties for seawater [Online]. Available:
http://www.seafriends.org.nz/oceano/oceans2.htm
 A Numerical Model of Seawater Volume and Velocity
Dynamic for Marine Currents Power Plant in the Bangka
Strait, North
Sulawesi, Indonesia
P T D Rompas*1, H Taunaumang2, and F J Sangari3
1
Departemen Pendidikan Teknologi Informasi dan Komunikasi, Universitas Negeri
Manado, Tondano 95618, Indonesia
2
Departemen Fisika, Universitas Negeri Manado, Tondano 95618, Indonesia
3
Departemen Pendidikan Teknik Elektro, Universitas Negeri Manado, Tondano 95618,
Indonesia

*parabelemrompas@unima.ac.id

Abstract. One of equipment as prime movers in the marine current power plant is turbine.
Marine current turbines require a data of marine currents velocity in its design. The objective
of this study was to get the velocities distribution of marine currents in the Bangka strait. The
method used survey, observation, and measurement in the Bangka strait. The data of seawater
density conducted measurement in the Bangka strait. The data of width and depth of the strait
collected from the map of Bangka strait and its depth of the sea. Problem solving of the study
used a numerical model. The velocities distribution of marine current obtained from a
numerical model in the form of numerical program. The results showed that the velocities
distribution at seawater column when low and high tide currents which the maximum
happened at 0.1 Sv were 0-0.9 and 0-1.0 m/s respectively, while at 0.3 Sv were 0-2.7 and 0-
3.0 m/s respectively. The results will be a product in analyzing the potential kinetic energy
that used to design profile of the turbines as prime mover for marine currents power plant in
the Bangka strait, North Sulawesi, Indonesia.

1. Introduction
The Indonesian government through the President and Vice President seriously encourage increased
electricity infrastructure in Indonesia [1]. It implemented with the issuance of Presidential Decree (Decree) No.
4 2016 on accelerating the development of electricity infrastructure. Earlier, the President had inaugurated the
35 thousand MW to Indonesia. Ministry Coordinator (CMEA) of Economic Affairs conducts socialization on
follow-up to Presidential Decree no. 4 2016. The socialization intended to provide insight for agencies, as well
as follow the implementation of efforts to accelerate the development of electricity infrastructure with emphasis
on the use of New and Renewable Energy (EBT) to support efforts to reduce greenhouse gas emissions.
A numerical model to determine the velocity distribution of of marine currents in some areas in
Indonesia has conducted by the Indonesian Hydrodynamics Laboratory BPPT and The Society of
Naval Architects of Japan [2]. Similarly, Rompas and Manongko [3] have done it in the study on the
distributions of marine current velocity in the Strait of Bunaken, North Sulawesi, Indonesia.

Content from this work may be used under the terms of theCreativeCommonsAttribution 3.0 licence. Any
further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and
DOI.
Published under licence by IOP Publishing Ltd
One component of marine current electricity power plant is the turbine. Making the turbine was badly in
need marine current velocity data. It obtained by a study on numerical model of seawater volume and velocity
dynamics.
The objective of this study was to get the velocities distribution of marine current in the Bangka strait, North
Sulawesi, Indonesia by a numerical model.
2. Governing Equations
Fundamental mathematics equation that used in the numerical model is the conservation of energy
equation. It would express the variations in temperature, especially in account dissipation by friction, will
ignore and temperature will later appear as a tracer only liable for the effects of buoyancy. Conservative of the
fluid mass based on the following equation [4]
1 (U )
U U)
t div ( ) g F
(1) where be
the density of the fluid, and U the velocity vector, whose components are U, V, W. is

the tensor operator “nabla”. g F is external forces where g is the constant gravitational acceleration
and the other forces (Coriolis acceleration, etc.)
In this study, based on the decomposition of preceding Reynolds and under the assumptions of hydrostatic
pressure, then conservative of the fluid mass become Realized Average Navier-Stokes (RANS) equations [5].
The RANS equations were as a basic for the formulation of numerical model.

2.1.The boundary condition.

The domain of the Bangka strait is more complex of forms to free surface flows. Therefore, it needs
limited. Some types of boundary conditions are required such as the boundary conditions at the bottom, the
surface of the water, the wall, and open boundary. The first, at the bottom only horizontal velocity that could
considered with used a Chezy formula [6]. At the surface, we used principally two conditions, the first is wind
shear stresses in x-direction and the second is wind shear stresses in y-direction [7]. At the wall, we used the
impermeable condition [4]. Finally, a condition of radiation and adaptive boundary condition that developed by
Treguier et al. [8] used at open boundary.

2.2.The turbulence model.

The turbulent model for 2D used dept-average formulation from Stansby [9] as defined follow:

l 2

t h4 u 2 2 vy 2 vx uy 2 u hf 2 1/ 2 (2)

2
where the friction velocity u f b / with ρ is water density and bed shear stress b bx

2
by where ( bx , by ) Cf (u v u, ) 2 v2 with C f 0.0559 Re h
0.25
is friction coefficient of the

Blasius formula where the depth Reynolds number Reh ( u 2 v h2 ) / , and is the Elder constant
about 0.067 which are the depth-averaged vertical mixing, and the horizontal mixing length lh= h where is
a boundary layer constant ( =0.09), h is water depth and =lh/lv is the constant from result comparison with
experiment.
3. Numerical Approach
The numerical equations that used in this study were results developing from mathematical equations.
Also, from the results of modification of the YAXUM/3D model that used by Ramirez et al. [10]. In the
numerical model, we used semi-implicit finite difference method for the numerical solution of the
threedimensional Equation 1 in the computation of shallow water flows [6].
The velocities in x, y, and z directions, we used the equations which a general semi-implicit discretization of
the momentum equations and we can written in the more compact matrix form as [6]

Ain 1/2, j Uin 11/2, j Gin 1/2, j g t in 11, j i jn, 1 ΔZin 1/2, j

x (3)
 A Vi jn, 1/2i jn, 11/2 Gi jn, 1/2 g t i jn, 11 i jn, 1 ΔZi jn, 1/2

y (4)

4. Method

4.1.The domain presentation of study

The geographical location of the Bangka Strait is from 125 ° 04'40 "E to 125 ° 11'18" E and from 1 °
41'25 "N to 1 ° 44'03" N which consist of islands of Talise, Kinabuhutan, Ganges, Tindila, Lehaga, and
Sulawesi. In the east, there are the Maluku Sea and Pacific oceans, and to the west is the Sulawesi Sea (see
figure 1). In addition, there are two of the current circulations in the Bangka strait i.e. low tide currents and high
tide currents.

Bangka Strait

Figure 1. Location of numerical study (Bangka strait, North Sulawesi, Indonesia)

4.2.Methods of study
The methods used survey, observation, and measurement in the Bangka strait. The data of seawater
density conducted measurement in the Bangka strait. The data of width and depth in the strait collected from the
map of Bangka strait and its depth of the sea. Problem solving of the study used a numerical model. The
solution of a numerical model we take the case in the Bangka strait for calculating the velocities of u , v and w
respectively, we can explain step by step as the first is the beginning of computation with start. Then, the read
data that using the all of parameters in the numerical equations and time of calculating until time maximum for
doing iteration. Next step, generation of the mesh such as horizontal and vertical meshes and then continue to
the process generating the index such as generate the layers of vertical axis (depth) and boundary layers. The
next process, it makes the initial conditions of velocities and seawater surface elevation. Then, step in the start
of iteration process until maximum iteration that the beginning with the process of boundary conditions in the
Bangka strait. Next, calculate advections in u andv , which are the processes for calculating advections of
horizontal velocities. Then, calculate surface elevations with a linear five-diagonal system. The next calculation
is the calculation process of velocities in horizontal direction. Finally, the process to execute determination
when the iteration has been greater than maximum iteration, if no then process will be continue to calculate
again, and if yes then calculation to finish.

Table 1. Numerical Parameter for 2D-Simulations

Parameter Value Parameter Value
G 9.81 m s-2 ρseawater 1024 kg/m3
Cz 48 Δx 60 m

Τo 2 days Δy 60 m

τi 1 day Δt 1 sec

Discharge variable

In table 1, there are two discharges that calculated with classifications are 0.1 Sv and 0.3 Sv. τo and τi are
relaxation timescales at outflow and inflow conditions respectively [8]. Cz is Chezy coefficient and ρseawater is
density of seawater. Δx, Δy, and Δt are space step in x direction, space step in y direction, and time step
respectively.

5. Results and Discussion

Figure 2-5 showed the result of modelling and numerical simulation. We can see that in the forms of
simulations i.e. 2D-simulations when low tide currents and high tide currents. Figure 2 shows distributions of
velocities and current threads at seawater column when low tide currents. Generally, seawater enters from
section A and B where current flows from section A and go to section D and a small part flow to section C
which previous rotate form two eddies like elliptic diameter at centre between Gangga and Bangka Islands. On
the other side, current enter in section B flow to section C which previous form eddy in north area near Talise
Island and a small part flow to section D which previous form small maelstrom like diameter between Gangga
and Bangka islands. On the contrary, when high tide current (see Figure 3), current enters from section D go to
section A and a small part flow to section B which previous happened eddy at center east area near Bangka
Island whereas current from section C go to section B which previous form eddy like elliptic diameter at north
area near Talise Island and a small part of the other current go to south side of Gangga Island at section A.
We can see in figure 4 which maximal velocities predominated at around section A when low and high tide
currents at 0.3 Sv of 1.5-2.7 m/s (figure 4.b) and 1.2-3.0 m/s (figure 5.b) respectively. When high tide currents,
the volume of seawater passing through the area was so large which is the result of a combination of direction
section A and D (see also figure 3). Whereas when low tide current (see bottom centre area in figure 4), it was
the depth that only of 5 m. The values shown that currents flow when high tide currents greater when low tide
currents. Figure 5 shows distribution of marine current velocities at seawater column at discharges of 0.1 and
0.3 Sv respectively when high tide currents. The currents were so strong in the top of section C by 3 m/s. It is
because not only so large volumes of seawater but also shallow sea of 6 m. In the centre of area, we can see that
the average of marine currents velocity by 2.7 m/s.
Figure 2. Simulations (2D) of marine current velocities and current threads at seawater column when
low tide currents at discharges of 0.1 Sv (a) and 0.3 Sv (b)

Figure 3. Simulations (2D) of marine current velocities and current threads at seawater column when
high tide currents at discharges of 0.1 Sv (a) and 0.3 Sv (b)

a b
Figure 4. Distribution of marine current velocities at seawater column when low tide currents at
discharges of 0.1 Sv (a) and 0.3 Sv (b)
 a b

Figure 5. Distribution of marine current velocities at seawater column when high tide currents at
discharges of 0.1 Sv (a) and 0.3 Sv (b)

The results in the figures 4 and 5 especially in the centre area of strait that the values of velocity can be used
to design profile of the marine current turbine. Besides, this area is suited installed turbines for power plant
which the ideal locations for power plant installation of the current energy have velocities of current two
directions (minimum bidirectional) of 2.5 m/s or more, one way is minimum 1.2-1.5 m/s. The deepness not less
than 15 m and the most at 50 m, the construction near the beach so that energy can be supplied at low cost, the
area is spacious enough for more than one turbine installation, and no the area of sea transport and fishing [11].

6. Conclusions
We successfully obtained the velocities distribution of marine currents at seawater column in the
Bangka strait, North Sulawesi, Indonesia by the numerical model. The velocities distribution at seawater
column when low and high tide currents which the maximum happened at 0.1 Sv were 0-0.9 and 0-1.0 m/s
respectively, while at 0.3 Sv were 0-2.7 and 0-3.0 m/s respectively. The results will be a product in analyzing
the potential kinetic energy, which can used to design profile of turbines for marine currents power plant in the
Bangka strait North Sulawesi, Indonesia.

Acknowledgments
The authors would like to thank profusely to DRPM, Kementerian Riset, Teknologi, dan Pendidikan
Tinggi Republik Indonesia who has given funding all of the research activities, and to Rector of Universitas
Negeri Manado, Indonesia who has agreed our research.

References
[1] Murdifi A 2016 The 35,000 MW definitely step project powered by presidential decree No. 4 2016
(http://www.pln.co.id/blog/langkah-pasti-proyek-35-000-mw-didukung-oleh-perpresno-4-
tahun-2016/) access on 12 May 2016
[2] Erwandi 2016 Sources of energy current: alternatives to fuel oil, sustainable and renewable
(http://www.energi.lipi.go.id/utama.cgi?cetakartikel&1125749769) access on 12 May 2016
[3] Rompas P T D and Manongko J D I 2016 Numerical simulation of marine currents in the Bunaken
Strait, North Sulawesi, Indonesia Int. Conf. on Innovation and Vocational Education IOP Conf.
Series: Materials Science and Engineering 128 1-7
[4] Hervouet J M 2007 Hydrodynamics of free surface flows: Modelling with the finite element method.
(England: John Willey & Sons, Ltd.) pp. xiv-341
[5] Broomans P 2003 Numerical Accuracy in Solution of the Shallow-Water Equations (Master thesis,
TU Delft & WL, Delft Hydraulics)
[6] Casulli V and Cheng R T 1992 Semi-implicit finite difference methods for three-dimensional shallow
water flow Int. J. for Numerical Methods in Fluids 15 629-648
[7] Chen X 2003 A free-surface correction method for simulating shallow water flows J. Comput.
Phys. 189 557-578
[8] Treguier A M, Barnier B, and De Miranda A P 2001 An eddy-permitting model of the Atlantic
circulation: Evaluating open boundary condition J. Geophy. Res. Oceans 106 (C10) (2211522129) 1-
23
[9] Stansby P K 2006 Limitations of depth-averaged modeling for shallow wakes J. of Hydraulic
Engineering 132 (7) 737-740
[10] Ramirez H, Barrios H, Rodriguez C, and Couder C 2005 Baroclinic mathematical modeling of fresh
water plumes in the interaction river-sea Int. J. of Numerical Analysis and Modeling 2 1-14
Fraenkel P L 2002 Power from marine currents, proceedings of the institution of mechanical
engineers: part A J. Power and Energy 216 (1) 1-1
A Model of Small Capacity Power Plant in Tateli Village,
North Sulawesi

F J Sangari*, P T D Rompas
Universitas Negeri Manado, Tondano 95618, Indonesia

*ferry_sangari@yahoo.com

Abstract. The electricity supply in North Sulawesi is still very limited so ubiquitous electric
current outage. It makes rural communities have problems in life because most uses electrical
energy. One of the solutions is a model of power plants to supply electricity in Tateli village,
Minahasa, North Sulawesi, Indonesia. The objective of this research is to get the model that
generate electrical energy for household needs through power plant that using a model of
Picohydro with cross flow turbine in Tateli village. The method used the study of literature,
survey the construction site of the power plant and the characteristics of the location being a
place of research, analysis of hydropower ability and analyzing costs of power plant. The result
showed that the design model of cross flow turbines used in pico-hydro hydropower
installations is connected to a generator to produce electrical energy maximum of 3.29 kW for
household needs. This analyze will be propose to local government of Minahasa, North
Sulawesi, Indonesia to be followed.

1. Introduction
North Sulawesi region more specifically in the villages of Minahasa district has a mountainous
topography and has many rivers which is a potential source of enormous energy for power plants which, when
carefully planned can overcome the problem of electric energy crisis. Problems that have long and at this time
every day power outage for about 2-3 hours a day. However, the electricity crisis was not so much solved using
the integral energy source potential flow of river water in the area of North Sulawesi. There are still many
villages far from urban areas still do not have adequate power supply [1]. In anticipation of that, it is necessary to
build small-scale power plants (1 kW - 5 kW).
Figure 1 shows that the electrical energy production by 2013 in the province of North Sulawesi using water
power is still very little 9.02% [2]. It shows that the construction of the hydroelectric power plant is still very
much needed in the area of North Sulawesi. Shortage of electricity in rural areas is very likely to occur because it
is far from the urban and the power grid, but did not rule urban areas are also experiencing the same thing. In
fact, many cities and districts that rely on diesel and hard to come by when the oil, resulting in a power outage in
rotation may even are expanded. One solution is emerging development in Indonesia at this time is to find a way
out through the construction of power plants micro scale with the power source stream flow and more
dependable again when rural communities require the construction of power plants as small as possible, namely
less than 5 kW and that can be realized through pico-hydro power plant (PLTPH).
During this time there is a kind of consensus that development PLTPH must have a double impact, not only
to improve the provision and equitable distribution of power supply, especially in rural areas but also makes the
vehicle to improve the ability of local industry to address the development PLTPH starting from feasibility
studies, planning, manufacturing machinery and equipment, to the realization

Content from this work may be used under the terms of theCreativeCommonsAttribution 3.0 licence. Any
further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and
DOI.
Published under licence by IOP Publishing Ltd
of the installation. In addition PLTPH development pattern aligned with the level of presence in the
form of appropriate technology in rural areas. Rural technology in irrigation development people get almost the
same pattern with the construction PLTPH, just need refinement for power knows no season. PLTPH itself is an
intermediate technology that has been applied since long ago and is expected to have a positive impact or benefit
for the creativity and dynamism of society in lifestyle in order to improve the welfare of rural communities. It is
becoming urgency in this study in addition to the efficiency and effectiveness of the construction of power
plants, especially in the energy forces turn turbines and planning / selection of appropriate turbine to turn a
generator needs to be analyzed.
 Figure 1. Production of electric energy in North Sulawesi (2013)

Based on the analysis of the energy force of the waterfalls and design analysis turbines, all of which were
made through experiments in the laboratory and in the field (in the village where the installation location) to the
mounting PLTPH and generate electrical energy production is less than 5 kW is an invention/innovation targeted
in research this. PLTPH development is to be realized in the villages Minahasa North Sulawesi province are in
accordance with the roadmap of the Indonesian government as shown in figure 2 [3]. In the period 2015-2025,
the government has planned the role of government research and development (R & D) that update data on
potential PLTPH in the area and manufacturing feasibility study PLTPH, development turbines PLTPH the
efficient use of technology crosflow, propeller, and kaplan and system development capacity of 750 kW and the
role of manufacturing industries seek turbine development results. Based on the role of government and industry
in the 20152025 year, the study is in conformity with the government's plan 2015-2025 year on R & D activities
are potential PLTPH map updates and the total development of the system. Was further strengthened by the
government of North Sulawesi province that integrate and provide services to the activity centers in the region of
North Sulawesi province with the publication of the North Sulawesi provincial regulations No. 1 Year 2014 on
Spatial Planning of North Sulawesi province from 2014 to 2034 Year Article 22 (a) and 23 (a) that the
government regulates the energy network system includes one electric generating system in Minahasa [4].
For the creation of the PLTPH development, it previously had to be done: first, about the theoretical studies
as the basis for analysis through literature; second, conduct survey research sites (including survey the village is
selected villages that have the potential to obtain characteristics of the villages and communities for the purpose
of character education for these communities through socialization to shape the character of the community so
that there is a sense of belonging and accept, help build and maintain the power plant in village) in the form of:
discharge of river water, river water dam layout, long conductive flow of water to the location of the water falls,
the water level fell, and location of turbines; Third, experiments in laboratory experiments in the form of high
variation of water falling; Last, economic analysis and supply to the local government Minahasa district for
expansion into other villages potentially built PLTPH.
Pico-hydro is hydro power with a maximum electrical output of 5 kW [5] [6] [7]. The system is beneficial in
terms of cost and simplicity from a different approach in the design, planning and installation from the power
applied to the water is greater. The latest innovations in pico-hydro technology have made an economic
powerhouse even in some of the poorest places in the world and can be accessed. It is also a versatile resource.
AC electricity possible can be produced from a standard electrical equipment to be used. Common examples of
devices that can be powered by pico-hydro are light bulbs, radio and television.
Typically, pico-hydro power systems are found in rural areas or hilly. Figure 3 shows an example of a
typical application of pico-hydro hydro system in hilly areas [7][8]. This system will operate using a container of
water on which a few meters from the ground. From the reservoir, the water flows downhill through the piping
system and it allows the water to turn turbines. Thus, the turbine will rotate the alternator to produce electricity.
However, this study was conducted to demonstrate the potential of consuming water that is distributed to homes
in rural areas as an alternative renewable energy source. The flow of water in the pipe has the potential and
kinetic energy will be converted into the potential energy of motion of the turbine which then into electrical
energy in generators.
 Figure 2. Roadmap of PLTPH development in Indonesia

Technically, pico-hydro has three main components: water (as a source of energy), turbine and generator [9]
[10]. Pico-hydro get energy from the flow of water that has a certain altitude difference. Basically, pico-hydro
utilized the potential energy of water falling (head). The higher the water falls, the greater the potential energy of
water that can be converted into electrical energy. In addition to geographical factors (layout of the river), the
height of falling water can also be obtained by stemming the flow of water so that the water level is high. Air
flowed through a pipe plant rapidly into the house in general was built on the banks of the river to drive turbines
or waterwheels pico-hydro. Mechanical energy derived from rotation of the turbine shaft is converted into
electrical energy by a generator.
The objectives of this research is to get the model of efficient and effective from cross flow turbine models
for installation in PLTPH which generates electrical energy production of less than 5 kW for a household in the
village Tateli Minahasa North Sulawesi.
1.1. Impulse Turbine
The principle works: water flowing perpendicular to the turbine shaft through the rapidly incoming pipe
at high speed and push the rotor blades of the turbine, causing the turbine rotor rotation [9] [10] [11]. The
pressure drop in the water flow in the nozzle and the turbine wheel operates at atmospheric pressure. The
examples of Cross flow Impulse Turbines (figure 4) are a Pelton wheel, wheel Turgo and crossflow turbine
(Banki-Michell). Impulse turbines generally operate best with medium or high head above 10 m.

1.2. Reaction Turbine

Reaction turbines operate under pressure inside the stator (casing). When water passing through the
stator in the direction of the turbine shaft helical, causing a whirlpool. The flow was then directed by the blades
of the turbine wheel. The angular momentum of the forces in the water rotates the turbine wheel. In contrast to
the impulse turbine, the water pressure drops in the stator and the turbine wheel. Examples of a reaction turbine
(figure 5) is Propeller (propeller), Kaplan, and Francis, Screw and kinetic turbines water (used to lower head is
less than 5 m). Reaction turbines often have houses and geometry turbine blades of a complex which makes it
more difficult to process large-scale production of the smallest in the settings in developing countries. However,
the reaction turbine can perform well even in low head distance of less than 10 m, thus making it more desirable
since the low head of water resources are more accessible and closer location. Water turbines can be classified in
two categories namely [11] [12]:
 Figure 3. Examples of application of pico-hydro power systems in rural areas
Turbin power (Pt) defined [9] [10]:
= (1) where ρ is water density (1000
3 2 3
kg/m ), g is gravity (m /s), H is head, Q is water flow (m /s), and ɳ is turbine efficiency (normally 70-80 %
depend on turbine type).
Calculation of Electric Power and Energy:
Power turbine shaft:
= 9.81 (2)
Power is transmitted to the generator:
= 9.81 (3)
Power generated generator:
,
= 9.81 (4)
Where ɳ = turbine efficiency (0.74 and 0.75 for the cross flow turbine to turbine axial), ɳbelt =
transmission efficiency, 0.98 for a flat belt, 0.95 for the V belt, ɳgen = generator efficiency

Figure 4. T ype of crossflow i mpulse turbine

 Figure 5. Type of axial reaction turbine

This generator power generated will be distributed to users. In planning the required amount of power at the
load center should be under the power capacity is raised, so that the power supply voltage is stable and the
system becomes more reliable (long).
The water flowing with capacity and a certain height distributed to the house installation (casing). At home
the turbine, the water plant will pound turbine, turbine ascertained in this case will receive the water energy and
convert it into mechanical energy in the form of turbine shaft rotation. The rotating shaft is then connected to the
generator by using the tire/belt. Of the generator will produce electricity that will go into the control system
before the electrical current supplied to homes or other purposes (load). That briefly the process of pico-hydro,
alter energy flow and water level into electrical energy [10].
2. Approach method
The method used literature study, site survey research (including survey the village in order to obtain
the characteristics of villages and rural communities Tateli for the purpose of character education for these
communities in receiving, helped build and maintain the power plant), analysis beginning on the ability of
electric power, and discussions with local authorities (see figure 6). Conducting preparatory activities before
carrying out such research; reflecting the result of socialization, preparation of materials and experimental tools,
setting work schedules. Data collection in the village Tateli form: discharge of river water, water velocity,
channel length conductor, high waterfalls, and location of the turbines. Methods of direct observation in the field
through measures such as speed streams and cross-sectional area perpendicular to the water flow of the river to
get water discharge flowing river as initial data in the analysis of the ability of river water, and then to analyze
the electrical energy taken preliminary data height measurements falling water (planned 1.34 m) including
measuring the distance from the dam to the water fall. Technique of direct measurement with the following
procedure: first measure the water velocity and the second measuring cross-sectional area perpendicular to the
flow of river water in order to get water discharge (cross-sectional area multiplied by the speed of the water,
m3/s), and the last measure the height of falling water to get the length of the aqueduct of dam water to the
waterfall.
 Figure 6. Location of research in village Tateli, Minahasa
Analyze and calculate a preliminary survey in the form of the findings of the theory through literature, site
survey and collection of field data and then analyze the debit mainstay and planning ability hydroelectric
developed in the planning and continued in the field with a pass calculation electrical energy through analysis of
turbine and generator, voltage, and power generated. Finally, proceed with analyzing the PLTPH development
costs.

3. Results and discussion

The analysis results of water capability that obtained from basis of the gross calculation before analysis
further can be seen in table 1. The water discharge (Q) is planned to enter the pipe rapidly with water fall head of
1.34 m and if we calculate minimum water flow of 20% of the water flow was then obtained
91.6 l litre/s. Generates power without taking into account the total efficiency of 5.12 kW and if we take
into account the total efficiency of 0.643, the installed power of 3.29 kW.

Table 1. Results of water and power capability analysis

Q (m3/s) Hbruto (m) Hlosses (m) Heff (m) P (kW) P’ (kW)

0.458 1.34 0.134 1.2 5.12 3.29

Table 2 shows the results of the analysis of electrical energy which is calculated based on the total efficiency,
the force of gravity, high falls and the actual effective water obtained from the difference between the height of
fall slop and total loss of height of falling water (0.1 m previously planned 0.134 m). The total energy obtained
during a year high real effective water fall of 1.2 m was 29.335 MWh with installed power of 3.29 kW. Total
electrical energy obtained within a year of 29,335 kWh. If we calculate the value of selling electricity to PLN by
calculating the total cost of expenditure per year of IDR 25 million, the value of the electricity sold at IDR
852/kWh.
All of the results are showed that the electrical energy needs can be fulfiled not only for the village Tateli but
also other villages nearby. So as to realize it, then the results of this study will be proposed to local governments
Minahasa to be realized in order to meet the needs of not only domestic but also for street lighting and other
facilities available in the village Tateli.

Table 2. Results of electrical energy analysis

∑η H eff ( m)
3
Q80 ( m / s)
3
Q90 ( m / s)
3
Q100 ( m / s) ∑ E (MWh )

0.643 1.2 0.458 0.407 0.254 29.335

4. Conclusions
The ability of hydroelectric power of 5.12 kW is installed or the electrical power produced by the
capability of the water. The effective head of 1.2 m with the generated power of 3.29 kW and water flow of
0.458 m3/s. The total energy is obtained within a year is 29.335 MWh. The electrical needs could be satisfied in
Tateli. This model will be proposed to local government of Minahasa, North Sulawesi, Indonesia to be followed.

5. Acknowledgments
The Authors gave the highest appreciation to DRPM Kemenristekdikti Jakarta Republic of Indonesia
that has funded all of these activities. Also, thank to head of Tateli village, Minahasa, North Sulawesi, Indonesia
who had been allowed for implementing this study.

References
[1] Harian Metro 30 September 2014 Pemadaman Meresahkan, PLN Mengaku Kekurangan Pasokan
Listrik. http://www.harianmetro.co.id/index.php/2013-02-02-05-25-09/minsel-
mitra/19104pemadaman-meresahkan-pln-mengaku-kekurangan-pasokan-listrik#.VTJcuPCPxVc
Accessed on February 15 2016
[2] PT PLN (Persero) 2014 Statistik PLN 2013 Sekretariat Perusahaan PT PLN (Persero) ISSN 0852–
8179 No. Publikasi: 02601-140722 http://www.pln.co.id/dataweb/STAT/STAT2013IND.pdf Accessed
on January 23 2015
[3] Indonesia 2005-2025 Buku Putih. 2006. Penelitian, Pengembangan dan Penerapan Ilmu Pengetahuan
dan Teknologi Bidang Sumber Energi Baru dan Terbarukan untuk Mendukung
Keamanan Ketersediaan Energi Tahun 2025, Kementerian Negara Riset dan Teknologi Republik
Indonesia. Jakarta
[4] Peraturan Daerah Provinsi Sulawesi Utara No. 1 Tahun 2014 tentang Rencana Tata Ruang Wilayah
Provinsi Sulawesi Utara Tahun 2014-2034
[5] Basar M F., Sapiee R, Rahman S, Hamdan Z, Borhan S and Sopian K 2014 Advances in Environmental Biology 8 14
147-151
[6] Martin S and Sharma A B 2014 International Journal of Advanced Research in Computer Engineering & Technology
(IJARCET) 3 6 2121-2126
[7] Zainuddin H, Yahaya M S, Lazi J M, Basar M F M and Ibrahim Z 2009 World Academy of Science, Engineering and
Technology 59 154-159
[8] Maher P and Smith N 2001 Pico Hidro Potencia para Aldeas: Un Manual Práctico para Instalaciones de hasta 5 kW
en Terrenos de Pendientes Fuertes. Ed. 2.0 http://www.riaed.net/IMG/pdf/Pico_Hidro_-
_Potencia_para_Aldeas_1.pdf. Accessed on March 5 2015
[9] Kapoor R 2013 International Journal of Scientific Research 2 9 159-160
[10] Nimje A A and Dhanjode G 2015 IOSR Journal of Environmental Science, Toxicology and Food Technology (IOSR-
JESTFT) 9 1 Ver. III 59-67
[11] At-Tasneem M A, Azam W M and Jamaludin U 2014 Word Applied Sciences Journal (Innovation Challenges in
Multidiciplinary Research & Practice) 30 420-423
[12] Sangal S, Garg A and Kumar D 2013 International Journal of Emerging Technology and Advanced Engineerng 3 3
424-430
Study on the Seawater Surface Elevation through Numerical
Modeling Approach in Gulf of Manado
*
P T D Rompas , J D I Manongko
Universitas Negeri Manado, Tondano 95618, Indonesia

*Corresponding author: parabelemrompas@unima.ac.id

Abstract. This paper presents the seawater surface elevation in Gulf of Manado with
numerical approach. The RANS shallow water flow equation is used to obtain a numerical
equation of seawater surface elevation with a semi-implicit approximation approach in which
the pressure distribution in the vertical layer of seawater is assumed hydrostatic. The results
found that the high seawater level is not more than 1.81 m upper seawater level and 1.30 m
under seawater level respectively. The results can be used as a recommendation to predict the
condition of sea waves in the Gulf of Manado.

1. Introduction
The horizontal and vertical motion of seawater occurs caused by the wind stress at the sea surface. If
surface seawater is a divergence, water at the bottom will rise to surface and this situation will happened
upwelling; conversely, where it is a convergence and will happened downwelling. For condition where a system
of cyclonic wind on surface seawaters where the average movement of the wind-driven layer is to the right of
the wind which causing divergence of surface water and upwelling, and the sea-surface is lowered and the
thermocline is raised, it is called Ekman pumping. In condition inverse correspond in anticyclonic wind is
convergence and sinking (downwelling), it causes the seasurface to slope upwards the middle of the gyre [1]. It
will produce a sea wave called seawater surface elevation. Transport at the position of Gulf of Manado is
convergence.

A numerical modeling can predict seawater surface elevation. It conducted by [2-5] in the Straits of
Bunaken and Bangka, North Sulawesi, Indonesia. Chen [6] has discussed the simulation seawater surface
elevation from shallow water flows with a free surface correction method. Rodrigues-Cuevas et al [7] have
proposed the distribution of seawater surface elevation with respect to a reference level (m) by study on
modeling shallow water wakes using a hybrid turbulence model. Claeyssen et al [8] have conducted calculating
of seawater surface elevation by using the semi-lagrangian method and they compute the non-linear response
of the full model due to a shear stress that comes from the action of the wind at the ocean surface. Sirjacobs et
al [9] have simulated seawater surface elevation in the variations of typical seasonal corresponding to the
period 1956-1960 from the surface wind stress in the Aral sea. Also, Balasubramanya et al [10] have
implemented fully-implicit time integration schemes in a version of parallel ocean program for calculation of
seawater surface elevation by its simulation in the North Atlantic.

The objective of this study was to get distribution of seawater surface elevation by using numerical
modeling in the Gulf of Manado, North Sulawesi, Indonesia.

2. The Free Surface Navier-Stokes Equations

The conservation of energy equation, which would express the variations in temperature, particularly taking
into account dissipation by friction, will be put aside and temperature will later appear as a tracer solely
responsible for the effects of buoyancy. Let be the density of the fluid, and the velocity vector, whose

components are U, V, W. Conservative of the fluid mass contained in a domain is expressed as [11] d

d 0 (1) dt
 where Ω is the domain of study (see figure 1) and t is time.
y
z

free surface

bottom

Ω2 D
x

Figure 1. Modeling domain

If we assume hydrostatic pressure, and by using the decomposition of preceding Reynolds, the
realized average Navier-Stokes equations are written [11, 12]:
Continuity equation

u v w
0 (2)
x y z
Momentum equation

u u u u

u v w g div eff grad u fcorv (3)

t x y z x

v v v v

u v w g div eff grad v fcoru (4)

t x y z y
Free surface equation

t x hudz y hvdz 0 (5)

where eff is an effective diffusion taking of account turbulent viscosity and dispersion, eff t.
This effective diffusion is given by means of a model of turbulence adapted to the problem considers.
Conservative of the fluid mass becomes Realized Average Navier-Stokes (RANS) shallow water flow
equations (as in equations (2) to (5)) which were as a basic for the computation of numerical modeling.

3. Numerical Modeling Approach

The numerical modeling equations used diffusion step [11] with a general semi-implicit discretization of
the momentum equations in equations (3) and (4) to get the velocities while the seawater surface elevation
used equation (5). The velocities in x, y, and z directions, we can written in the more compact matrix form as in
[2-5]. For determine the seawater surface elevation (as in equation (5)) in form numerical modeling, we can be
written in the matrix notation form [2, 3]
 ΔZ U ΔZ Un 1 x (6)

in, j1 in, j t i 1/ 2, j T in 11/ 2, j i 1/ 2, j T i 1/ 2, j

ΔZ V ΔZ Vin 1/ 2

t i, j 1/ 2 T in, j 11/ 2 i, j 1/ 2 T , j 1

4. Methods
The Gulf of Manado is located between the Pacific Ocean and the Sulawesi Sea (Celebes Sea) whose
2
area is approximately 300 km (Figure 2), with an average width about 2.2 km and down to 79 meters deep
(figure 3).

The three-dimensional current circulation in the Gulf of Manado is simulated using the present model
with a 174 x 318 finite difference mesh of equal Δx = Δy = 7 m. The numerical solutions have been achieved
using four vertical layers and an integration time Δt = 0.4 sec, and inlet volume transports at sections Singkil
3
river, North, and South of Manado bay (figures 2 and 3) are 0.1 Sv (100000 m /s). In addition, there are two of
the current circulations in the Gulf of Manado i.e. low tide currents and high tide currents.

Figure 2. Gulf of Manado, North Sulawesi, Indonesia

 Figure 3. Bathymetry of Manado Gulf

The three-dimensional semi-implicit numerical method used in this study. The solution of a numerical
model in calculating the velocities of u , v and w respectively, we can explain step by step as follows:
- start is the beginning of numerical computation.
- read data is the process to read all data that using the all of parameters in equations 2 to 5 and table 1
which needed in calculation and also read data of variables: time of calculating maximum for doing
iteration.
- generation of the mesh is the process to make the computational meshes which using Arakawa C-grid
[13] and notations (spatial discretization).
- generation of the index is the process to make the index and generate the layers of vertical axis
(depth) and generate index of boundary layers as denote calculating in the meshes.
- initial conditions are the process to make the initial conditions of velocities and seawater surface
elevation, and also to make the coordinates for the result simulations in the tecplot program.
- the quantity of calculation with iteration do-process until maximum iteration.
- boundary conditions are the process which give boundary conditions of calculation domain in the Gulf
of Manado. Also, generation and determination about boundaries of velocities when enter and exit the
Gulf of Manado bases data of investigation results.
- calculate advections in u and v are the process for calculation advection of u and v .
- turbulence model is the process calculating the turbulence of a mixing-length model [7] (for three-
dimensional calculation). In this process also calculated diffusion step.
- calculate components of velocities (u , v and w ) are the process for calculating of horizontal velocities
u and v (equations 4 and 5) with a linear three-diagonal.
- calculate seawater surface elevations are the process for calculating equation 6 which is values of the
seawater surface elevation with a linear five-diagonal system.
- print results are the process for print results calculation that we necessary.
- T > Tmax is the process to execute determination when the iteration has been greater than maximum
iteration, if no then process will be go to the quantity of calculation for continue to calculate again, and
if yes then it go to “finish”.
 - finish is the process where the calculation is stop.

Table 1. Parameter value for 3D-numerical simulations

Parameter Value Parameter Value
-2 1024
g 9.81 m s ρs 3
kg/m
Cz 48 Δx 7m

τo 2 days Δy 7m

τi 1 day Δz 1m

discharge 0.1 Sv Δt 0.5 sec

6 3
Table 1 shows a discharge is 0.1 Sv (0.1 x 10 m /s). τo and τi are relaxation timescales at outflow and
inflow conditions respectively [14]. Cz is Chezy coefficient and ρs is density of seawater. Δx, Δy, and Δt are
space step in x direction, space step in y direction, and time step respectively.

Figure 4. Simulations of seawater surface elevation

distributions when low tide currents at discharge of 0.1 Sv
5. Results and Discussion
The numerical modeling results we can see in figures 4 and 5 which are in the forms of simulations i.e.
3D-simulations (showed in 2D) when low tide currents and high tide currents. Figure 4 shows distributions of
seawater surface elevation when low tide currents. High seawater level when low tide currents is biggest
around Southside which the maximum of 75 cm and Northside locations of 130 cm under seawater level.
Whereas when high tide currents, there is 181 cm of maximum high seawater level in Northside and the
minimum of 117 cm under seawater level in Southside (figure 5). The distributions of seawater surface
elevation when low tide currents in Eastside are 2 cm, Westside around 28-79 cm, and Centerside around 28-
53 cm respectively under seawater level. Whereas when high tide currents in Eastside same as Westside
around 32-107 cm, and Centerside around 70 cm respectively upper seawater level.
 Figure 5. Simulations of seawater surface
elevation distributions when high tide currents at discharge
of 0.1 Sv

The high seawater level occurs when high tide currents greater than low tide currents. It caused by the
influence of velocities of seawater when high tide currents greater than when low tide currents [3]. The other
effect is the horizontal and vertical movement of seawater results in the wind stress at the sea-surface when
high tide currents greater than when low tide currents [1, 10].

6. Conclusions
The seawater surface elevation by numerical modeling approach in the Gulf of Manado, North
Sulawesi, Indonesia has been investigation. The distributions of high seawater level were 1.81 m upper
seawater level and 1.30 m under seawater level respectively. The results can be used as a recommendation to
predict the condition of sea waves in the gulf of Manado.
References
[1] Broomans P 2003 Numerical Accuracy in Solution of the Shallow-Water Equations (Master thesis, TU
Delft & WL, Delft Hydraulics)
[2] Rompas P T D and Manongko J D I 2016 Numerical simulation of marine currents in the
Bunaken Strait, North Sulawesi, Indonesia Int. Conf. on Innovation and Vocational Education IOP
Conf. Series: Materials Science and Engineering 128(012003) 1-7
[3] Rompas P T D, Sangari F J, and Tanunaumang H 2017 Study on marine current with approach of a
numerical model for marine current power plant (PLTAL) in the Bangka Strait North Sulawesi
Proceedings-2016 International Seminar on Application of Technology for Information and
Communication, ISEMANTIC 2016. IEEE 1 104-110.
[4] Rompas P T D, Taunaumang H, and Sangari F J 2017 A numerical design of marine current for
predicting velocity and kinetic energy Indonesian Journal of Electrical Engineering and Computer
Science 5(2) 401-409.
[5] Rompas P T D, Taunaumang H, and Sangari F J 2017 A numerical model of seawater volume and
velocity dynamic for marine currents power plant in the Bangka Strait, North Sulawesi, Indonesia IOP
Conference Series: Materials Science and Engineering 180(012100) 1-7.
[6] Chen X 2003 A free-surface correction method for simulating shallow water flows J. Comput. Phys. 189
557-578
[7] Rodriguez-Cuevas C, Couder-Castaneda C, Flores-Mendez E, Herrera-Diaz I E, and CisnerosAlmazan
R 2014 Modelling shallow water wakes using a hybrid turbulence Journal of Applied Mathematics 1 1-10.
[8] Claeyssen J R, Garibotti C, and Vielmo S 2006 The free surface of a coupled ocean-atmosphere model
due to forcing effects Mathematics and Computer in Simulation 73 114-124.
[9] Sirjacobs D, Gregoire M, Delhez E, and Nihoul J C J 2004 Influence of the aral sea negative water
balance on its seasonal circulation patterns: use of a 3D hydrodynamic model Journal of Marine System
47 51-66.
[10] Balasubramanya T N, Taylor M, and Lorenz J 2006 Ocean modelling for climate studies: Eliminasi short
time scales in long-term, high resolution studies of ocean circulation Mathematical and Computer
Modelling 44 870-886.
[11] Hervouet J M 2007 Hydrodynamics of free surface flows: Modelling with the finite element method.
(England: John Willey & Sons, Ltd.) xiv-341.
[12] Casulli V and Cheng R T 1992 Semi-implicit finite difference methods for three-dimensional shallow
water flow Int. J. for Numerical Methods in Fluids 15 629-648.
[13] Arakawa A and Lamb V R 1977 Computational design of the basic dynamical processes of the UCLA
general circulation model Methods of Computational Physics 17 173-265.
[14] Treguier A M, Barnier B, and De Miranda A P 2001 An eddy-permitting model of the Atlantic circulation:
Evaluating open boundary condition J. Geophy. Res. Oceans 106 (C10) (2211522129) 1-23.

Acknowledgments
The authors are grateful for DRPM, Ministry of Research, Technology, and Higher Education of the
Republic of Indonesia who has given funding all of the research activities, and also thank the Rector of Manado
State University, Indonesia who has supported this research.

Study on the Seawater Surface Elevation through Numerical

Modeling Approach in Gulf of Manado
*
P T D Rompas , J D I Manongko
Universitas Negeri Manado, Tondano 95618, Indonesia

*Corresponding author: parabelemrompas@unima.ac.id

 Abstract. This paper presents the seawater surface elevation in Gulf of Manado with
numerical approach. The RANS shallow water flow equation is used to obtain a numerical
equation of seawater surface elevation with a semi-implicit approximation approach in which
the pressure distribution in the vertical layer of seawater is assumed hydrostatic. The results
found that the high seawater level is not more than 1.81 m upper seawater level and 1.30 m
under seawater level respectively. The results can be used as a recommendation to predict the
condition of sea waves in the Gulf of Manado.

1. Introduction
The horizontal and vertical motion of seawater occurs caused by the wind stress at the sea surface. If
surface seawater is a divergence, water at the bottom will rise to surface and this situation will happened
upwelling; conversely, where it is a convergence and will happened downwelling. For condition where a system
of cyclonic wind on surface seawaters where the average movement of the wind-driven layer is to the right of
the wind which causing divergence of surface water and upwelling, and the sea-surface is lowered and the
thermocline is raised, it is called Ekman pumping. In condition inverse correspond in anticyclonic wind is
convergence and sinking (downwelling), it causes the seasurface to slope upwards the middle of the gyre [1]. It
will produce a sea wave called seawater surface elevation. Transport at the position of Gulf of Manado is
convergence.

A numerical modeling can predict seawater surface elevation. It conducted by [2-5] in the Straits of
Bunaken and Bangka, North Sulawesi, Indonesia. Chen [6] has discussed the simulation seawater surface
elevation from shallow water flows with a free surface correction method. Rodrigues-Cuevas et al [7] have
proposed the distribution of seawater surface elevation with respect to a reference level (m) by study on
modeling shallow water wakes using a hybrid turbulence model. Claeyssen et al [8] have conducted calculating
of seawater surface elevation by using the semi-lagrangian method and they compute the non-linear response
of the full model due to a shear stress that comes from the action of the wind at the ocean surface. Sirjacobs et
al [9] have simulated seawater surface elevation in the variations of typical seasonal corresponding to the
period 1956-1960 from the surface wind stress in the Aral sea. Also, Balasubramanya et al [10] have
implemented fully-implicit time integration schemes in a version of parallel ocean program for calculation of
seawater surface elevation by its simulation in the North Atlantic.

The objective of this study was to get distribution of seawater surface elevation by using numerical
modeling in the Gulf of Manado, North Sulawesi, Indonesia.

2. The Free Surface Navier-Stokes Equations

The conservation of energy equation, which would express the variations in temperature, particularly taking
into account dissipation by friction, will be put aside and temperature will later appear as a tracer solely
responsible for the effects of buoyancy. Let be the density of the fluid, and the velocity vector, whose

components are U, V, W. Conservative of the fluid mass contained in a domain is expressed as [11] d

d 0 (1) dt
where Ω is the domain of study (see figure 1) and t is time.
 y
z

free surface

bottom

Ω2 D
x

Figure 1. Modeling domain

If we assume hydrostatic pressure, and by using the decomposition of preceding Reynolds, the
realized average Navier-Stokes equations are written [11, 12]:
Continuity equation

u v w
0 (2)
x y z
Momentum equation

u u u u

u v w g div eff grad u fcorv (3)

t x y z x

v v v v

u v w g div eff grad v fcoru (4)

t x y z y
Free surface equation

t x hudz y hvdz 0 (5)

where eff is an effective diffusion taking of account turbulent viscosity and dispersion, eff t.
This effective diffusion is given by means of a model of turbulence adapted to the problem considers.
Conservative of the fluid mass becomes Realized Average Navier-Stokes (RANS) shallow water flow
equations (as in equations (2) to (5)) which were as a basic for the computation of numerical modeling.

3. Numerical Modeling Approach

The numerical modeling equations used diffusion step [11] with a general semi-implicit discretization of
the momentum equations in equations (3) and (4) to get the velocities while the seawater surface elevation
used equation (5). The velocities in x, y, and z directions, we can written in the more compact matrix form as in
[2-5]. For determine the seawater surface elevation (as in equation (5)) in form numerical modeling, we can be
written in the matrix notation form [2, 3]
 ΔZ U ΔZ Un 1 x (6)

in, j1 in, j t i 1/ 2, j T in 11/ 2, j i 1/ 2, j T i 1/ 2, j

ΔZ V ΔZ Vin 1/ 2

t i, j 1/ 2 T in, j 11/ 2 i, j 1/ 2 T , j 1

4. Methods
The Gulf of Manado is located between the Pacific Ocean and the Sulawesi Sea (Celebes Sea) whose
2
area is approximately 300 km (Figure 2), with an average width about 2.2 km and down to 79 meters deep
(figure 3).

The three-dimensional current circulation in the Gulf of Manado is simulated using the present model
with a 174 x 318 finite difference mesh of equal Δx = Δy = 7 m. The numerical solutions have been achieved
using four vertical layers and an integration time Δt = 0.4 sec, and inlet volume transports at sections Singkil
3
river, North, and South of Manado bay (figures 2 and 3) are 0.1 Sv (100000 m /s). In addition, there are two of
the current circulations in the Gulf of Manado i.e. low tide currents and high tide currents.

Figure 2. Gulf of Manado, North Sulawesi, Indonesia

 Figure 3. Bathymetry of Manado Gulf

The three-dimensional semi-implicit numerical method used in this study. The solution of a numerical
model in calculating the velocities of u , v and w respectively, we can explain step by step as follows:
- start is the beginning of numerical computation.
- read data is the process to read all data that using the all of parameters in equations 2 to 5 and table 1
which needed in calculation and also read data of variables: time of calculating maximum for doing
iteration.
- generation of the mesh is the process to make the computational meshes which using Arakawa C-grid
[13] and notations (spatial discretization).
- generation of the index is the process to make the index and generate the layers of vertical axis
(depth) and generate index of boundary layers as denote calculating in the meshes.
- initial conditions are the process to make the initial conditions of velocities and seawater surface
elevation, and also to make the coordinates for the result simulations in the tecplot program.
- the quantity of calculation with iteration do-process until maximum iteration.
- boundary conditions are the process which give boundary conditions of calculation domain in the Gulf
of Manado. Also, generation and determination about boundaries of velocities when enter and exit the
Gulf of Manado bases data of investigation results.
- calculate advections in u and v are the process for calculation advection of u and v .
- turbulence model is the process calculating the turbulence of a mixing-length model [7] (for three-
dimensional calculation). In this process also calculated diffusion step.
- calculate components of velocities (u , v and w ) are the process for calculating of horizontal velocities
u and v (equations 4 and 5) with a linear three-diagonal.
- calculate seawater surface elevations are the process for calculating equation 6 which is values of the
seawater surface elevation with a linear five-diagonal system.
- print results are the process for print results calculation that we necessary.
- T > Tmax is the process to execute determination when the iteration has been greater than maximum
iteration, if no then process will be go to the quantity of calculation for continue to calculate again, and
if yes then it go to “finish”.
 - finish is the process where the calculation is stop.

Table 1. Parameter value for 3D-numerical simulations

Parameter Value Parameter Value
-2 1024
g 9.81 m s ρs 3
kg/m
Cz 48 Δx 7m

τo 2 days Δy 7m

τi 1 day Δz 1m

discharge 0.1 Sv Δt 0.5 sec

6 3
Table 1 shows a discharge is 0.1 Sv (0.1 x 10 m /s). τo and τi are relaxation timescales at outflow and
inflow conditions respectively [14]. Cz is Chezy coefficient and ρs is density of seawater. Δx, Δy, and Δt are
space step in x direction, space step in y direction, and time step respectively.

Figure 4. Simulations of seawater surface elevation

distributions when low tide currents at discharge of 0.1 Sv
5. Results and Discussion
The numerical modeling results we can see in figures 4 and 5 which are in the forms of simulations i.e.
3D-simulations (showed in 2D) when low tide currents and high tide currents. Figure 4 shows distributions of
seawater surface elevation when low tide currents. High seawater level when low tide currents is biggest
around Southside which the maximum of 75 cm and Northside locations of 130 cm under seawater level.
Whereas when high tide currents, there is 181 cm of maximum high seawater level in Northside and the
minimum of 117 cm under seawater level in Southside (figure 5). The distributions of seawater surface
elevation when low tide currents in Eastside are 2 cm, Westside around 28-79 cm, and Centerside around 28-
53 cm respectively under seawater level. Whereas when high tide currents in Eastside same as Westside
around 32-107 cm, and Centerside around 70 cm respectively upper seawater level.
 Figure 5. Simulations of seawater surface
elevation distributions when high tide currents at discharge
of 0.1 Sv

The high seawater level occurs when high tide currents greater than low tide currents. It caused by the
influence of velocities of seawater when high tide currents greater than when low tide currents [3]. The other
effect is the horizontal and vertical movement of seawater results in the wind stress at the sea-surface when
high tide currents greater than when low tide currents [1, 10].

6. Conclusions
The seawater surface elevation by numerical modeling approach in the Gulf of Manado, North
Sulawesi, Indonesia has been investigation. The distributions of high seawater level were 1.81 m upper
seawater level and 1.30 m under seawater level respectively. The results can be used as a recommendation to
predict the condition of sea waves in the gulf of Manado.
References
[15] Broomans P 2003 Numerical Accuracy in Solution of the Shallow-Water Equations (Master thesis, TU
Delft & WL, Delft Hydraulics)
[16] Rompas P T D and Manongko J D I 2016 Numerical simulation of marine currents in the
Bunaken Strait, North Sulawesi, Indonesia Int. Conf. on Innovation and Vocational Education IOP
Conf. Series: Materials Science and Engineering 128(012003) 1-7
[17] Rompas P T D, Sangari F J, and Tanunaumang H 2017 Study on marine current with approach of a
numerical model for marine current power plant (PLTAL) in the Bangka Strait North Sulawesi
Proceedings-2016 International Seminar on Application of Technology for Information and
Communication, ISEMANTIC 2016. IEEE 1 104-110.
[18] Rompas P T D, Taunaumang H, and Sangari F J 2017 A numerical design of marine current for
predicting velocity and kinetic energy Indonesian Journal of Electrical Engineering and Computer
Science 5(2) 401-409.
[19] Rompas P T D, Taunaumang H, and Sangari F J 2017 A numerical model of seawater volume and
velocity dynamic for marine currents power plant in the Bangka Strait, North Sulawesi, Indonesia IOP
Conference Series: Materials Science and Engineering 180(012100) 1-7.
[20] Chen X 2003 A free-surface correction method for simulating shallow water flows J. Comput. Phys. 189
557-578
[21] Rodriguez-Cuevas C, Couder-Castaneda C, Flores-Mendez E, Herrera-Diaz I E, and CisnerosAlmazan
R 2014 Modelling shallow water wakes using a hybrid turbulence Journal of Applied Mathematics 1 1-10.
[22] Claeyssen J R, Garibotti C, and Vielmo S 2006 The free surface of a coupled ocean-atmosphere model
due to forcing effects Mathematics and Computer in Simulation 73 114-124.
[23] Sirjacobs D, Gregoire M, Delhez E, and Nihoul J C J 2004 Influence of the aral sea negative water
balance on its seasonal circulation patterns: use of a 3D hydrodynamic model Journal of Marine System
47 51-66.
[24] Balasubramanya T N, Taylor M, and Lorenz J 2006 Ocean modelling for climate studies: Eliminasi short
time scales in long-term, high resolution studies of ocean circulation Mathematical and Computer
Modelling 44 870-886.
[25] Hervouet J M 2007 Hydrodynamics of free surface flows: Modelling with the finite element method.
(England: John Willey & Sons, Ltd.) xiv-341.
[26] Casulli V and Cheng R T 1992 Semi-implicit finite difference methods for three-dimensional shallow
water flow Int. J. for Numerical Methods in Fluids 15 629-648.
[27] Arakawa A and Lamb V R 1977 Computational design of the basic dynamical processes of the UCLA
general circulation model Methods of Computational Physics 17 173-265.
[28] Treguier A M, Barnier B, and De Miranda A P 2001 An eddy-permitting model of the Atlantic circulation:
Evaluating open boundary condition J. Geophy. Res. Oceans 106 (C10) (2211522129) 1-23.

Acknowledgments
The authors are grateful for DRPM, Ministry of Research, Technology, and Higher Education of the
Republic of Indonesia who has given funding all of the research activities, and also thank the Rector of Manado
State University, Indonesia who has supported this research.
Validation of a Numerical Program for Analyzing Kinetic
Energy Potential in the Bangka Strait, North Sulawesi,
Indonesia

P T D Rompas*, H Taunaumang and F J Sangari

Universitas Negeri Manado, Tondano, 95618, Sulawesi Utara, Indonesia

*parabelemrompas@unima.ac.id

Abstract. The paper presents validation of the numerical program that computes the
distribution of marine current velocities in the Bangka strait and the kinetic energy potential in
the form the distributions of available power per area in the Bangka strait. The numerical
program used the RANS model where the pressure distribution in the vertical assumed to be
hydrostatic. The 2D and 3D numerical program results compared with the measurement results
that are observation results to the moment conditions of low and high tide currents. It found no
different significant between the numerical results and the measurement results. There are
0.97-2.2 kW/m2 the kinetic energy potential in the form the distributions of available
power per area in the Bangka strait when low tide currents, whereas when high tide currents of
1.02-2.1 kW/m2. The results show that to be enabling the installation of marine current turbines
for construction of power plant in the Bangka strait, North Sulawesi, Indonesia.

1. Introduction
The potential of kinetic energy in Bangka Strait, North Sulawesi, Indonesia can analyzed by approach
of a numerical program that analyzes the velocity of the ocean currents necessary to calculate the kinetic energy
of ocean currents.
The study of ocean currents through numerical modeling has performed [1]. Who implicated the numerical
model in San Francisco Bay California and the Lagoon of Venice, Italy by using semiimplicit finite difference
method for 3D shallow water flow, Rompas and Manongko [2] simulated the marine currents in the Bunaken
strait, North Sulawesi, Indonesia with the numerical model, Rompas et al [3-5] studied the marine currents in the
Bangka strait, North Sulawesi, Indonesia with the modelling and numerical simulation which predicted the
velocities and the kinetic energies but the numerical program has not been validated, Rodrigues-Cuevas et al [6]
researched a numerical model with difference turbulence models, O’Donncha et al [7] were applied the model-
3D of hydro-environmental code to a designated tidal energy test site on the East Coast of the United States,
Gonzales-Gorbena et al [8] optimized the hydrokinetic turbine array layouts by surrogate modelling, Zangiabadi
et al [9, 10] developed the a CFD model by using the bathymetry of a potential tidal stream turbines deployment
site and presented the tidal stream turbines for tidal power production. Martins et al [11] used 3D modelling in
the Sado estuary by using a new generic vertical discretization approach. Luquet et al [12] tested design and
model of an optimized ducted marine current turbine. The 3D hydrodynamic model used [13] for knowing
influence of the Aral Sea negative water balance on its seasonal circulation patterns, whereas [14] studied
climate with ocean modelling for eliminated short time

Content from this work may be used under the terms of theCreativeCommonsAttribution 3.0 licence. Any further
distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd
scales in long-term and high resolution of ocean circulation. The simulation of electrical power
potential in the Alderney Race has been successfully which conducted by Myers and Bahaj [15] for marine
current turbine arrays. On the other hand, Thiebot et al [16] developed the effects of large arrays of tidal turbines
with dept-average Actuator Disks by modelling. Pinon et al [17] predicted wake of marine current turbines with
a particle method by using the numerical simulation. Whereas, Elhanafi et al [18] used the numerical simulation
with CFD to analyzing the offshore stationaryfloating oscillating water column-wave energy converter. Also, Ho
and Riddette [19] applied CFD to evaluate hydraulic performance of spillways in Australia.
The purpose of numerical program validation is to obtain a numerical program that can generate data in
accordance with the measurement results [20]. Validation done by comparing the numerical programs with the
data of measurement results and there are some researcher have conducted it. Maters et al [21] have validated a
numerical program to analyzing tidal stream turbine. The study of numerical and experimental has conducted by
Mycek et al [22] for analyzing interaction between two marine current turbines.
This paper studied on the validation of the numerical program computation results to the measurement
results of the velocities of marine current when the low tide and high tide in the Bangka Strait. The objectives of
the study are getting validation results of the numerical program computation result to measurement result and
analyzing the available power potential of marine current kinetic energy in the Bangka strait, North Sulawesi,
Indonesia.

2. A Numerical Program

2.1 A three-dimensional semi-implicit numerical method

The three-dimensional semi-implicit numerical method used by Casulli & Cheng [1] and Chen [23] in the
computation of shallow water flows. The method the finite differences is used for simplicity of its
implementation. The basic idea consists with [24]:
• To reduce the field of resolution of the differential equation to a limited field.
• To define a grid (or grid) finished points of this field.
• To approach the derivative this appears in the equation using a development of Taylor around the points of
the grid. For the points located at the edges of the field of calculation, we will write the boundary conditions
in an exact or possibly approximate way.
Concerning the first and second phase, there are no general methods [25]. The determination of the field of
calculation depends on the problem to approach. For the third stage, it is enough to recall that if f is a sufficiently
regular function of the variable reality x in a vicinity of the point x0, then we have the development of following
Taylor [25]:

f (x) n (x x )k f (k ) k (!x0) (x

x0)n 1 f (n 1)[ (xn (11)! )x0] (1)

0
k0
(k)
Where f indicates the derivative of f of order k and θ {0,1}. The order to which the development is
truncated gives the space order of approximation of the quantity f.

2.1.1. Fractional steps method. The basis algorithm consists of three fractional steps. We write [25]: f fn 1

fD fD fC fC fn (2)

t t
This leads to a successive resolution of three steps: An
advection step:

fC f n
+ advection terms = 0 ( f U,V )
(3) t A diffusion step:
fD fC
+ diffusion terms = source terms ( f U,V ) (5) t A
pressure-continuity step:
 fn 1
fD
+ pressure terms = 0 (6)
t

This last step includes also the resolution of the continuity equation div(U) 0 that helps to deduce the
water depth h and the vertical component of velocity W [25]. Each of these steps will be resolved in detail one
after the other. Prior to this, however, discretization in space and building up the mesh to specified. 2.1.2. Spatial
discretization. The spatial discretization selected by a discretization with the differences finished in
parallelepiped elements [1]. The parallelepiped makes it possible to build a threedimensional grid starting from a
two-dimensional grid. It is enough to net in rectangles the twodimensional field, then to duplicate this grid on the
vertical. It is possible to make with the same grid a calculation in 2D and 3D [26]. The grid makes it possible to
discretize the physical field in a whole of material points to which we will apply the finite difference method. For
the calculation of the variables, those are in a fixed mesh with different positions. Into particulate, the scalar sizes
cantered.
An Arakawa C-grid [26] is used. For the calculation of the variables, those are in a fixed mesh in different
positions (staggered concealment), so much on the horizontal levels as on the vertical levels. The velocities
defined on the edge of the mesh; we guessed virtual meshes to write the limiting conditions with the walls; we
decorate the free surface with the grid.

2.1.3. Advection step. In advection step we are using Eulerian-Lagrangian discretization of convective and
viscous terms. This discretization is one of the major difficulties in the numerical treatment of the shallow water
equations. Consider then the following convection-diffusion equation in three space dimensions [1]:

Cu C v C w C x2C2 y2C2 z
eff Cz (7)

t x y z

Where μ and νeff are non-negative diffusion coefficients and for the time being, the convective velocities u , v
and w are assumed to be constants.
The equation (7) can solve numerically in a variety of ways. A convenient semi-implicit finite difference
method, whose stability does not depend upon the vertical eddy diffusivity, is obtained by discretizing the
convective terms by explicit upwind finite differences, the horizontal eddy diffusivity by explicit central
differences and the vertical eddy diffusivity term by an implicit finite difference.

For non-negative u , v and w the resulting finite difference equation is:

Cin, j,1k Cin, j,k Cin, j,k Cin 1, j,k Cin, j,kCin, j 1,k Cin, j,k 1

u v w t x y z

Cin 1, j,k 2Cxin,2j,k Cin 1, j,k Cin, j 1,k 2Cyin,2j,k Cin, j 1,k (8)

Cin, j,1k 1 Cin, j,1k Cin, j,1k Cin, j,1k 1

 z

zi, j,k

For every i and j this method requires the solution of a symmetric, positive definite, tri-diagonal system. The
necessary and sufficient stability condition of scheme equation (8) is

u v w 1 1
t x y z 2 x2 y2 (9)

In convection-dominated problems, the stability condition equation (9) is not very restrictive. This method,
however, is only first-order-accurate in space and the truncation error is in the form of a diffusion term. This
artificial viscosity is directionally dependent. Hence, in convection-dominated problems, not only the artificial
viscosity will prevail over the physical viscosity, but also drastically different numerical predictions can obtained
simply because of different spatial orientations of the computational grid.
In order to improve the stability and accuracy of an explicit finite difference method, consider again equation
(7) in the Lagrangian form

2 2
C C C C

t x2 y2 z eff z (10)

Where the substantial derivative d/dt indicates that the time rate of change is calculated along the streak line
defined by dx/dt = u , dy/dt = v , dz/dt = w (11)

A natural semi-implicit discretization of equation (10) is simply given by

Cin, j,1k 1 Cin, j,1k Cin, j,1k Cin, j,1k 1

Cin, j,1k Cin a1, j b,k d eff k 1/ 2 zi, j,k 1/ 2 eff k 1/ 2 zi, j,k 1/ 2

(12)
t zi, j,k

C C C C C C
in a 1, j b,k d 2 in a, j 2b,k d in a 1, j b,k d in a, j b 1,k d 2 y
in a, j 2b,k d in b, j b 1,k d

Where a=u Δt/Δx, b=v Δt/Δy and d= w Δt/Δz are the grid Courant Numbers.

Then Cin a, j b,k d in equation (12) approximated by Casulli and Cheng [1] become:

Cin a, j b,k d (1 r){(1 p)[(1 q)Cin l, j m,k n qCin l, j m,k n ]

 p[(1 q)Cin l 1, j m,k n qCin l 1, j m 1,k n ]}

r{(1 p)[(1 q)Cin l, j m,k n 1 qCin l, j m 1,k n 1 ]

p[(1 q)Cin l 1, j m,k n 1 qCin l 1, j m 1,k n 1 ]} (13)

The stability condition for the scheme equation (12) as follow [1]
1

1 1
t 2 x2 y2 (14)

Which is much less restrictive than equation (9). Clearly, when μ=0, this scheme becomes unconditionally
stable.

2.1.4. Diffusion step. A general semi-implicit discretization of the momentum equations can write into form as
[1]:

uin 11/ 2, j,k (Fu)in 1/ 2, j,k g t in 11, j in, j1

x
uin 11/ 2, j,k 1 uin 11/ 2, j,k uin 11/ 2, j,k uin 11/ 2, j,k 1

t, (15)
z
i 1/ 2, j,k

vin, j 11/ 2,k (Fv)in, j 1/ 2,k g t in, j 11 in, j1

vin, j 11/ 2,k 1 vin, j 11/ 2,k vin, j 11/ 2,k vin, j 11/ 2,k 1

t (16)
z
i, j 1/ 2,k

Where zi 1/ 2, j,k and zi, j 1/ 2,k are in general the thickness of the kth water layer more simply denoted by

zk .

The finite difference operator F in equations (15) and (16) can define as [1]:

(Fu)in 11/ 2, j,k uin 1/ 2 a 1, j b,k d

 uin 1/ 2 a 1, j b,k d 2uin 1/ 2 a, j b,k d uin 1/ 2 a 1, j b,k d

2 x

u u u
in 1/ 2 a, j b 1,k d 2 in 1/ 2 a,2j b,k d in 1/ 2 a, j b 1,k d

y
n
fcor tv i 1/ 2 a, j b,k d , (17)

(Fv)in, j1 1/ 2,k vin a, j 1/ 2 b,k d

v v
vin a 1, j 1/ 2 b,k d 2 in a, j 1/ 2 b,k d in a 1, j 1/ 2 b,k d

t x2

v v v
in a, j 1/ 2 b 1,k d 2 in a, j 1/22 b,k d in a, j 1/ 2 b 1,k d

y
fcor
The boundary conditions at the free surface and at the sediment-water interface are following [1]:

uin 11/ 2, j,M 1 uin 11/ 2, j,M w vin, j 11/ 2,M 1 vin, j 11/ 2,M w

z z
(19)
i 1/ 2, j,M ,
1/ 2
i, j 1/ 2,M 1/ 2

u n 1u n 1g i 1/ 2, j,m i 1/ 2,
j,m 1 uin 1/ 2, j,m 2

m 1/ 2

vin 1/ 2, j,m 2 n 1

z u
i 1/ 2, j,m 1/ 2 2 i 1/ 2, j,m , (
Cz 20)
v n 1 v n 1 g i, j 1/ 2,m i, j 1/
2,m 1 uin, j 1/ 2,m 2 vin, j 1/
2,m
m 1/ 2

2
 z
i, j 1/ 2,m 1/ 2
M 1/

2 x M 1/ 2 y

n 1
v

C
z2
i, j 1/ 2,m

The equations (15) and (16) with the respective boundary conditions equations (19) and (20), we can write in
the compact matrix form as [1, 5]:

Uin 11/ 2, j Gin

n 1/ 2, j g t in 11, j in, j1 ΔZin 1/ 2, j

(21) Ai 1/ 2, j
x

Ain, j 1/ 2 Vin, j 11/ 2 Gin, j 1/ 2 g t in, j 11 in, j1 ΔZin, j 1/ 2

(22)
y

Where U, V, ΔZ, G and A are defined as:

u v z
in 11/ 2, j,M in, j 11/ 2,M M

uin 11/ 2, j,M 1 vin, j 11/ 2,M 1 zM 1

n 1 uin 11/ 2, j,M 2 ,

Ui 1/ 2, j

: :
:
uin 11/ 2, j,m

vin, j 11/ 2,m zm

 zM (Fu)in 1/ 2, j,M zM (Fv)in, j 1/ 2,M

t xw t yw

zM 1(Fu)in 1/ 2, j,M 1 zM 1(Fv)in, j 1/ 2,M

Gin 1/ 2, j zM 2 (Fu)in 1/ 2, j,M Gin, j 1/ 2 zM 2 (Fv)in, j 1/

2 2,M 2

: :

zm (Fu)in 1/ 2, j,m zm (Fv)in, j 1/ 2,m

 ,

( eff )M 1/ 2 t ( eff )M 1/ 2 t

zM 0

zM 1/ 2 zM 1/ 2

( eff )M 1/ 2 t( eff )M 1/ 2 t( eff )M 3/ 2 t ( eff )M 3/ 2 t

A zM 1/ 2 zM zM 1/ 2 zM 3/ 2 zM 3/ 2

: : :

:
( eff )m 1/ 2 t( eff )m 1/ 2 t

g t u2 v2

0 zm 1/ 2 zm zm 1/ 2

Cz2

Equations (21) and (22) are linear tri-diagonal systems which are coupled to the seawater surface elevation

n 1
( ) at time (tn 1).

n 1
2.1.5. Pressure-continuity step. For determine i , j and for numerical stability, the new velocity field has to satisfy

for each i,j the finite difference analogue of the seawater surface elevation equation:

x
in, j1 in, j t k M m zi 1/ 2, j,k n 1 zi u
1/ 2, j,k in 11/ 2, j,k

ui 1/ 2, j,k
(23)

y
t

k M m n 1 zi, j v
1/ 2,k in, j 11/ 2,k

zi, j v
1/ 2,k i, j 1/ 2,k
 Or we can be written in the compact matrix form:

in, j1 in, j t ΔZi 1/ 2, j T Uin 11/ 2, j ΔZi 1/ 2, j T Uin 11/ 2, j

x
(24)

t ΔZi, j 1/ 2 T Vin, j 11/ 2 ΔZ

T Vin, j 11/ 2

i, j 1/ 2

Since A is positive definite, A-1 is also positive definite and therefore (ΔZ)TA-1ΔZ is a non-negative number.
Hence equation (25) constitutes a linear five-diagonal system of equations for in, j1 which is symmetric and strictly
diagonally dominant with positive elements on the main diagonal and negative ones elsewhere. Thus the system is
positive definite and has a unique solution. In practice, this fivediagonal system can be solved very efficiently by a
conjugate gradient method. Once the new free surface location has been determined, equations (21) and (22) are
readily applicable to yield the new velocities u , v at time tn 1.

( Z)

in, j,1 g xt 22 T A 1 Z in 1/ 2, j

in 11, j in, j1 ( Z)TA 1 Z in 1/ 2, j in, j1 in 11, j

g t 2 ( Z)TA 1 Z in, j 1/ 2 in, j 11 in, j1 ( Z)TA 1 Z in, j 1/ 2

in, j1 in, j 11

y2

in, j t ΔZ T A 1G in 1/ 2, j ΔZ T A 1G in 1/ 2, j
 x

t T 1 G in, j 1/ 2 ΔZ T A 1G in, j 1/ 2

ΔZ A
y
(25)

Finally, the vertical component of the velocity w at the new time level can discretized from the continuity
equation becomes:

z u z u
n 1 win, j ,1k 1/ 2 in 1/ 2, j,k in 11/ 2, j,k in 1/ 2, j,k in 11/ 2, j,k

wi, j,k 1/ 2

n vin, j 11/ 2,k zin, j v

1/ 2,k in, j 11/ 2,k (26) zi, j 1/ 2,k

Where k=m,m+1,...,M, and the no-flux condition across the bottom boundary is assumed by taking win, j ,1m 1/ 2

0.

2.2 The available power of marine current and the boundary conditions
The available kinetic energy in this study is the available power per m2 (kW/m2). We used the available power that
is equation of the marine current power in the Bangka strait from [2, 5]:

P (vin, 1 3
j ,k ) 10
3
(27)
A
Where P is the marine current power in the Bangka strait in kW/m2, Ek is kinetic energy, dt is time,

n 1 u 2 v2 w 2 is velocity resultant with u 1(uin, j ,1k

and vi, j,k

2

1 n 1 vin, j 11,k ) and w 1(win, j ,1k win, j ) are scalars, respectivel

,1k 1

v (vi, j,k
2 2
We used the boundary conditions for the simulations of model 2D and 3D such as:
• On the free surface the effect of the wind, supposed negligible which is not taken into account,
 U
which translated by 0.
n
• At the bottom, the coefficient of friction (law of Chezy) is Cz = 48.
• At open boundary condition, we can radiation equation from Treguier et al [27].

The values of parameter in the computational of numerical program, we used:

• The domain of computation is discretized of 55332 elements (for 2D) where 174 elements in direction x and
318 elements in direction y.
• Each element of horizontal is 60 m x 60 m. In 3D, there are 221328 elements where 174 elements in x-direction,
318 elements in y-direction and 4 elements in z-direction.
• Each element is parallelepiped dimensions of 60 m x 60 m x 20 m.
• The flow rate is 0.3 Sv (300000 m3/s where 1 Sv = 1 x 106 m3/s [28]) with two conditions of marine currents
e.g. when low and high tide.

3. Results and Discussion

The comparisons of numerical model results and measurement results have been obtained by specify the
velocity measurement results at area inlet and outlet as boundary conditions. Then, the numerical model results (2D
and 3D) are studied and compared mainly focused at points location of P1 to P9 (see Figure 1). Finally, the
numerical model results for the study on marine currents in the Bangka strait has been obtained both in 2D and 3D
with four flow rates variations which are as the inlet boundary conditions.

Figure 1. The locations of measurement.

3.1. The comparisons of the numerical program results and the measurement
results
Figure 2 shows the comparisons of the minimum velocities from the measurement (observation) results and the
numerical program (2D and 3D model) results when low tide currents. The velocities in locations of P3 and P8 are
come near with results of observation but the other locations are rather different. The difference values of velocities
for both 2D and 3D to observed results in P3 are 0.1 m/s and its directions are 11 0 (Figure 3), whereas in P8 of 0.02
m/s for 3D and there is not for 2D and only 2 0 for both 2D and 3D respectively. On the contrary, in P1, there is 0.34
m/s for 2D and 0.38 m/s for 3D, whereas its directions for both 2D and 3D of 18 0 respectively.

bs

Figure 2. Current velocities at the minimum velocities boundary condition when low tide
currents.

The current directions from the numerical model results are not so differ with the observation results mainly P4
to P8 (Figure 3) and also points of P5 to P7 in Figure 5 with boundary conditions of the maximum velocities
whereas the other locations are rather different. The biggest differences are in locations of P1, P2, P5, and P9. If we
see in Figure 4, the velocities at P5 to P7 are not so differ between observation results and numerical model with the
boundary conditions. Whereas velocities from the calculation results of 2D and 3D models are not so differ.

bser

OI

Figure 3. Current directions at the minimum velocities boundary condition when low tide currents.

The difference values of velocities for both 2D and 3D to observed results in P5 are 0.04 m/s (Figure 4) and its
directions are 30 in Figure 5, then in P7 of 0.04 m/s for 2D and 0.02 m/s for 2D and there are not differ with its
directions for both 2D and 3D respectively. On the contrary, in P1, there is 0.55 m/s for 2D and 0.58 m/s for 3D,
whereas its directions for both 2D and 3D of 200 respectively.

bs

Figure 4. Current velocities at the maximum velocities boundary condition when low tide currents.

bse

OI

Figure 5. Current directions at the maximum velocities boundary condition when low tide currents.

Figure 6 shows comparisons of the minimum velocities from observation results with the numerical model
results when high tide currents. At point P4 where the velocity of 2D and 3D models result compared to the
observation result is almost same and also at points of P4 to P6 in Figure 7. Whereas current directions rather near
between models (2D and 3D) and observation in Figure 8 except at points P1 to P3.
 b

Figure 6. Current velocities at the minimum velocities boundary condition when high tide currents.

Ban
gka Island

0
 Figure 7. Distributions of the available power per m2 at seawater column of 20 m when low tide currents at flow
rate of 0.3 Sv in the Bangka strait.

Figure 8. Distributions of the available power per m2 at seawater column of 20 m when high tide currents at flow
rate of 0.3 Sv in the Bangka strait.
3.2. The kinetic energy potential in the Bangka strait
The kinetic energy potential in this study that generated in form the available power of marine current in
the Bangka strait have been analyzed. We can see the distributions of the available power per unit of area (kW/m2)
at seawater column of 20 m in Figures 7 and 8 which described the potential in the Bangka strait when low and high
tide currents at flow rate of 0.3 Sv (1 Sv = 1 x 10 6 m3/s [27]).
Figure 7 shows distributions of available power per m2 at seawater column of 20 m when low tide currents at of
0.3 Sv in the Bangka strait. There is 2.22 kW/m2 of the available power at P4 and its value is biggest compared at
the other points. It caused by nearby the point is existed a manger with deepness of ~5 m. On the contrary, the value
of available power per m2 is smallest at P1 (0.97 kW/m2). We can also see that available power per m2 in South area
of Bangka island (in the enter channel) where around 3-5 kW/m2 bigger than the other area in around that of 1.5
kW/m2. Also, in West area, especially at center area where power availabilities around 2-7 kW/m2. Whereas in
North and South area where available power per m2 still less unless near point of P9 about 2-3 kW /m2. If we see in
Westside of P6 and P7 where there are power availabilities biggest around 9-10 kW/m2. That thing caused by
existence of manger and average depth in the place of ~5 m [3, 5].
The distributions of the available power per m2 at seawater column of 20 m when high tide currents at flow rate
of 0.3 Sv in the Bangka strait showed in Figure 8. The available power per m 2 counted of 1.02 kW/m2 at point of P1
and that value is smallest if compared with the other points. On the contrary, there is 2.1 kW/m 2 available of the
biggest available power per m2 at P8 which has the biggest velocity.

4. Conclusions
The validations of the numerical program through the comparisons of the numerical computation results
and the measurement results have been analyzed and the results are not so far different. The kinetic energy potential
in the form the distributions of available power per area in the Bangka strait when low tide currents of 0.97-2.2
kW/m2, whereas when high tide currents of 1.02-2.1 kW/m2. The results show that to be enabling the installation of
marine current turbines for construction of power plant in the Bangka strait, North Sulawesi, Indonesia.
 Acknowledgments
The authors express gratitude to DRPM, Ministry of Research, Technology, and Higher Education of the
Republic of Indonesia that had finance all of the research activities, and to Rector of Universitas Negeri Manado,
Indonesia through the head of research institution who has approved this research.

References
[1] Casulli V and Cheng R T 1992 Semi-implicit finite difference methods for three-dimensional shallow water
flow International Journal for Numerical Methods in Fluids 15 629-648
[2] Rompas P T D and Manongko J D I 2016 Numerical simulation of marine currents in the Bunaken Strait,
North Sulawesi, Indonesia IOP Conf. Series: Materials Science and Engineering 128(012003) 1-7
[3] Rompas P T D, Taunaumang H, and Sangari F J 2017 A numerical design of marine current for predicting
velocity and kinetic energy Indonesian Journal of Electrical Engineering and Computer Science 5(2) 401-
409.
[4] Rompas P T D, Taunaumang H, and Sangari F J 2017 A numerical model of seawater volume and velocity
dynamic for marine currents power plant in the Bangka Strait, North Sulawesi, Indonesia IOP Conference
Series: Materials Science and Engineering 180(012100) 1-7.
[5] Rompas P T D, Sangari F J, and Tanunaumang H 2017 Study on marine current with approach of a
numerical model for marine current power plant (PLTAL) in the Bangka Strait North Sulawesi Proceedings-
2016 International Seminar on Application of Technology for Information and Communication, ISEMANTIC
2016. IEEE 1 104-110.
[6] Rodriguez-Cuevas C, Couder-Castaneda C, Flores-Mendez E, Herrera-Diaz I E, and CisnerosAlmazan R
2014 Modelling shallow water wakes using a hybrid turbulence Journal of Applied Mathematics 1 1-10.
[7] O’Donncha F, James S C, and Ragnoli E 2017 Modelling study of the effects of suspended aquaculture
installations on tidal stream generation in Cobscook Bay Renewable Energy 102 65-76
[8] Gonzalez-Gorbena E, Qassim R Y, and Rosman P C 2016 Optimisation of hydrokinetic turbine array layouts
via surrogate modelling Renewable Energy 93 45-57
[9] Zangiabadi E, Masters I, Williams A J, Croft T N, Malki R, Edmunds M, Mason-Jones A, and Horsfall I
2017 Computational prediction of pressure change in the vicinity of tidal stream turbines and the
consequences for fish survival rate Renawable Energy 101 1141-1156
[10] Zangiabadi E, Edmunds M, Fairley I A, Togneri M, Williams A J, Masters I, and Croft N 2015
Computational fluid dynamics and visualisation of coastal flows in tidal channels supporting ocean energy
development Energies 8 5997-6012
[11] Martins F, Leitao P, Silva A, and Neves R 2000 3D modelling in the Sado estuary using a new generic
vertical discretization approach Oceanologica Acta 24 (1) 1-12
[12] Luquet R, Bellevre D, Fréchou D, Perdon P, and Guinard P 2013 Design and model testing of an optimized
ducted marine current turbine International Journal of Marine Energy 2 61-80
[13] Sirjacobs D, Gregoire M, Delhez E, and Nihoul J C J 2004 Influence of the Aral sea negative water balance
on its seasonal circulation patterns: use of a 3D hydrodynamic model Journal of Marine System 47 51-66
[14] Balasubramanya T N, Taylor M, and Lorenz J 2006 Ocean modelling for climate studies: Eliminasi short
time scales in long-term, high resolution studies of ocean circulation Mathematical and Computer Modelling
44 870-886
[15] Myers L and Bahaj A S 2005 Simulated electrical power potential harnessed by marine current turbine
arrays in the Alderney Race Renewable Energy 30 1713-1731
[16] Thiebot J, Guillou S, and Nguyen V T 2016 Modelling the effects of large arrays of tidal turbines with dept-
averaged Actuator Disks Ocean Engineering 126 265-275
[17] Pinon G, Mycek P, Germain G, and Rivoalen E 2012 Numerical simulation of wake of marine current
turbines with a particle method Renewable Energy 46 111-126
[18] Elhanafi A, Fleming A, Macfarlane G, and Leong Z 2017 Numerical hydrodynamic analysis of
an offshore stationary-floating oscillating water column-wave energy converter using CFD
International Journal of Naval Architecture and Ocean Engineering 9 (1) 77-99
[19] Ho D K H and Riddette K M 2010 Application of computational fluid dynamics to evaluate hydraulic
performance of spillways in Australia Australian Journal of Civil Engineering 6 (1) 81-104
[20] Georgia Tech Research Corporation 2013 Assessment of energy
production potential from ocean currents along the United State
Coastline (Final project report September 15, 2013,
Award number: DE-EE0002661, [Online].
Available:
https://energy.gov/sites/prod/files/2013/12/f5/energy_production_ocean_currents_us_0.pdf
[Accessed: 25 March 2017])
[21] Masters I, Williams A, Croft T N, Togneri M, Edmunds M, Zangiabadi E, Fairley I, and Karunarathna H A
comparison of numerical modelling techniques for tidal stream turbine analysis Energies 8 7833-7853
[22] Mycek P, Pinon G, Germain G, and Rivoalen E 2013 Numerical and experimental study of the interaction
between two marine current turbines International Journal of Marine Energy 1 70-83
[23] Chen X 2003 A free-surface correction method for simulating shallow water flows J. Comput. Phys. 189
557-578
[24] Ferziger J H and Peric M 2002 Computational methods for fluid dynamics (Springer 3rd rev.ed., ISBN 978-
3-540-42074-3, XIV-423. 128 illus)
[25] Hervouet J M 2007 Hydrodynamics of free surface flows: Modelling with
the finite element method. (England: John Willey & Sons, Ltd.) xiv-341
[26] Arakawa A and Lamb V R 1977 Computational design of the basic dynamical processes of the UCLA
general circulation model Methods of Computational Physics 17 173-265
[27] Treguier A M, Barnier B, and De Miranda A P 2001 An eddy-permitting model of the Atlantic circulation:
Evaluating open boundary condition J. Geophy. Res. Oceans 106 (C10) (2211522129) 1-23
[28] Siedler G, Church J, Gould J 2001 Ocean Circulation and Climate, Observing and Modelling the Global
Ocean (Academic Press, San Diego, xx-715)
Performance of Savonius Blade Waterwheel with Variation
of Blade Number
1* 2
L Sule and P T D Rompas
1 Universitas Hasanuddin, Makassar, Sulawesi Selatan,
Indonesia
2
Universitas Negeri Manado, Tondano 95618, Sulawesi Utara, Indonesia

*luther.sule@yahoo.co.id

Abstract. The utilization of water energy source is mainly used as a provider of electrical
energy through hydroelectric power. The potential utilization of water flow energy is relatively
small. The objective of this study is to know the best blade of Savonius waterwheel with
various variables such as water discharge, blade number, and loading. The data used the
efficiency of waterwheel, variation of blade number, variable water discharge, and loading in
the shaft. The test results have shown that the performances of a top-water mill with the
semicircular curve where the variation in the number of blades are 4, 6, and 8 at discharge and
loading of 0.01587 m3/s and 1000 grams respectively were 9.945%, 13.929%, and 17.056%
respectively. The blades number of 8 obtained the greatest performance. The more number of
blades the greater the efficiency of the waterwheel Savonius.

1. Introduction
The condition of Indonesian topography which has many mountains and hills as well as the
stretch of river is almost found in every region [1]. Water energy is the energy that is suitable and the
most potential to be developed in Indonesia. The rate of growth of hydropower in Indonesia is very slow,
whereas the potential of Indonesia's hydro power is quite large reach 75.000 MW. Utilization through the
national electricity supply only reached 10.1% or 7,572 MW. Based on data from the Ministry of Energy
and Mineral Resources, the potential of hydropower energy is spread by 15,600 MW in Sumatra, 4,200
MW in Java, 21,600 MW in Kalimantan, 10,200 MW in Sulawesi, 620 MW in Bali-NTT-NTB, 430 MW in
Maluku and 22,350 MW in Papua. While in the world, the potential for water energy is estimated to reach
657 million HP or 489,924.8156 MW, but the utilization to 15%. The potential of water energy in each
continent is different [2] and we can be seen in the Table 1.
The greatest potential energy of water is in the African continent [3]. It can be seen with the river
Congo in Africa which became the largest potential energy of water in the world as well as several other
rivers that are used as a producer of electricity. Electric power is obtained from the conversion of
hydropower that turns the waterwheel or water turbine that utilizes the waterfall or stream in the river [4].
Fine the optimum performance of each specific amount of blade, i.e. 4, 6 and 8 blades on the water
(discharge) and the constant end equal head for Savonius model blade of capacity used for hydropower.
Determine the maximum power coefficient of the change in the number of Savonius model drops used for
hydropower [5]. As reference in the design of water wheel with the Savonius model savings benefits
research. The research results can be applied in the field of renewable energy conversion, especially in
the design of small-scale hydro power plant to micro [6].

Content from this work may be used under the terms of theCreativeCommonsAttribution 3.0 licence. Any
further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and
DOI.
Published under licence by IOP Publishing Ltd
 Table 1. Potential and developed hydropower in the world
Potential (in Percentage of Developed (in Percentage
Continent million horse power) total million horse power) of total

Africa 212 41.4 0.6 0.6

Asia 151 23.0 13.7 13.6
North America 87 13.2 41.1 40.8
Europe 69 10.5 40.8 40.5
South America 55 8.4 3.1 3.1
Oceania 23 3.5 1.4 1.4

2. Methods
This study used experiment in fluid machineries laboratory of Mechanical Engineering
Department, Universitas Hasanuddin with the installation as in Figure 1. The data retried by repeating the
test/retrieval procedure at least 5 repetitions for analysis. The data collection procedure/testing such as
we conducted by: the initial, check the state of the tool to be used on the waterwheel test and check the
valve is in good condition; connecting the pump with a power source; set the valve opening to determine
the discharge; allow the water to flow for 2 minutes to obtain stable flow conditions in the channel;
calculate the amount of first discharge by using bucket and stopwatch; install one of the waterwheels with
the number of 4 pieces of blade in position; record the water level and water temperature in the channel;
after that calculate the number of turns of the waterwheel using a tachometer without loading until loading
until the wheel cannot spin or stop; after that record the amount of rounds generated in each loading; after
obtaining the data from the waterwheel test the 4 pumps are turned off by disconnecting the electric
current source; and the finally, repeat procedure 2-10 for waterwheels 6 and 8 (Figure 2).

Figure 1. Testing installation

FRONTVIEW FRONTVIEW FRONTVIEW

SIDE VIEW SIDE VIEW SIDE VIEW

TOP VIEW TOP VIEW TOP VIEW

 Figure 2. Parameters of Blades (4, 6, and 8)

3 Results and Discussion

3.1. Relation of efficiency to load for 4 blades

The Figure 3 and Table 2 showed the relation of efficiency to the waterwheel load with variation of
3 3 3
discharge i.e. 0.01124 m /s (Q1), 0.01358 m /s (Q2), and 0.01587 m /s (Q3) with waterwheels of 4
3
blades. The maximum efficiencies occurred at 600 gram loading to discharge of 0.01124 m /s and
3 3
0.01358 m /s are 6.13% and 5.95% respectively. While, at discharge of 0.01587 m /s and loading of 1000
gram, the maximum efficiency is 9.66%.
3
In Figure 5, for 0.01124 m /s of discharge, it is seen that the efficiency increases from no load to 600
gram loading and decreases the efficiency of the waterwheel as the addition of loading, from load 600 to
1200 gram loading. The waterwheel will produce power because the wheel can offset the given torque
[7][8][9]. When a given torque is equal to zero, the waterwheel will not produce power because the
windmill will spin very quickly as a result of the absence of loading given so that no incubation occurs
[10][11]. Conversely, after reaching the critical point along with the increase in loading then braking occurs
in the pulley resulting in angular velocity and the power of the waterwheel decreases [12][13][14]. The
3
phenomenon that occurs for discharges of 0.01124 and 0.01358 m /s is caused by the same thing that
3
occurs at a discharge of 0.01587 m /s for the same blade.

20
10
y = -1E-05x2 + 0.0281x + 1.1745
0 R² = 0.9771
Q1
0 1000 2000 3000
-10
y = -1E-05x2 + 0.0219x + 0.6489 Q2
R² = 0.9745
-20
Q3
-30 y = -2E-05x2 + 0.0278x
R² = 1
-40
Load (gram)

Figure 3. Relation of efficiency to load with variation of discharge on waterwheel with 4 blades
3
In Figure 4 and Table 2, it is found that the discharge of 0.01587 m /s is the largest discharge between
the variations of the discharge used where the efficiency of 9.95% to loading of 100 gram. This is
because the amount of water flow that flows directly proportional to the water speed in the channel makes
the mass of the waterwheel lighter so that the resulting round is also large and requires a large load to
brake [15].

Table 2. η efficiency to load and discharge

Load Variation for 4 blades Variation for 6 blades Variation for 8 blades
(gram) η (Q1) η (Q2) η (Q3) η (Q1) η (Q2) η (Q3) η (Q1) η (Q2) η (Q3)
0 0 0 0 0 0 0 0
0 3,213 3,176 3,552 3,797 3,970 4,691 4,674 4,975
200 5,258 5,161 6,251 6,427 6,749 8,244 8,179 9,097
400 6,134 5,955 8,099 7,887 8,338 11,087 9,640 12,366
600 4,674 6,352 9,661 8,179 8,735 12,508 10,516 14,782
 800 2,921 5,955 9,946 7,303 8,933 13,503 10,224 16,346
1000 0 4,764 9,377 5,258 8,338 13,645 8,763 17,056
1200 4,169 7,956 4,089 6,948 13,929 6,134 16,914
1400 0 6,820 0 7,764 13,645 4,674 15,919
1600 5,115 3,573 12,792 0 15,351
1800 2,842 0 9,950 2,792
2000 0 7,814 9,381
2200 5,117 6,823
2400 0 3,696
2600 0
2800
Note: variation of discharge i.e. 0.01124 m3/s (Q1), 0.01358 m3/s (Q2), and 0.01587 m3/s (Q3

3.2. Relation of efficiency to load for 6 blades

The relation of efficiency to the waterwheel load with variation of debit i.e. 0,01124 m3 / s,
0,01358 m3 / s and 0,01587 m3 / s with waterwheels 6 blades is shown in Figure 5 and Table 3.

40

20
y = -1E-05x2 + 0.0355x + 0.9892
R² = 0.9912
0 Q1
0 1000 2000 3000
Q2
-20
y = -1E-05x2 + 0.0281x + 1.1745
R² = 0.9771 Q3
-40
y = -2E-05x2 + 0.0314x + 0.4134
R² = 0.9843
-60
Load (gram)

Figure 4. Relation of efficiency to load with variation of discharge on waterwheel with 4 blades
3
Based on Figure 4 and Table 3 were found that for discharge of 0.01124 m /s, the maximum
efficiency is 8.17920093% which occurs at 800 gram loading. While at discharge of 0.01358 m3 / s
3
maximum efficiency is obtained 8.93% with loading 1000 gram. While at 0.01587 m /s of discharge,
the maximum efficiency at loading of 1400 gram is 13.93%.
3
In Figure 5, for 0.01124 m /s of discharge, it was seen that the efficiency increased from no-load to
800 gram loading and decreased the efficiency of the waterwheel as the addition of loading, from a
load of 800 to 1600 gram loading. The waterwheel will produce power because the wheel can offset
the given torque [7][8][9]. When a given torque is equal to zero, the waterwheel will not produce
power because the windmill will spin very quickly as a result of the absence of loading given so that
no incubation occurs [16]. Conversely, after reaching the critical point along with the increase in
loading then braking occurs in the pulley resulting in angular velocity and the power of the waterwheel
3
decreases [17]. The phenomenon that occurs for discharges of 0.01124 and 0.01358 m /s is caused
3
by the same thing that occurs at a discharge of 0.01587 m /s for the same blade.
 3.3. Relation of efficiency to load for 8 blades
3
Based on Figure 5 and Table 4 it is found that for debit 0.01124 m /s maximum efficiency is
3
10.52% which occurs at 800 gram loading. While at debit 0.01358 m /s maximum efficiency is
3
obtained by 13.90% with loading 1000 gram. While at 0.01587 m /s of discharge, the maximum
efficiency at loading of 1200 gram is 17.06%.

20

10
y = -9E-06x2 + 0.0237x + 0.8052
R² = 0.9882
0
0 1000 2000 3000 Q1
-10
Q2
-20 Q3
y = -1E-05x2 + 0.0243x + 0.7673
R² = 0.9867
-30
y = -1E-05x2 + 0.0224x + 0.5577
R² = 0.982
-40
Load (gram)

2
Figure 5. Distributions of the available power per m at seawater column of 20 m when low tide
currents at flow rate of 0.3 Sv in the Bangka strait.
3
In Figure 6 for a 0.01124 m /s of discharge, it is seen that the efficiency increases from no-load to 800
gram loading and decreases the efficiency of the waterwheel as the addition of loading from a load of 800
to 1600 gram. The waterwheel will produce power because the wheel can offset the given torque
[7][8][9]. When a given torque is equal to zero, the waterwheel will not produce power because the
windmill will spin very quickly as a result of the absence of loading given so that no incubation occurs [16].
Conversely, after reaching the critical point along with the increase in loading then braking occurs in the
pulley resulting in angular velocity and the power of the waterwheel decreases [17][18][19]. The
3
phenomenon that occurs for discharges of 0.01124 and 0.01358 m /s caused by the same thing that
3
occurs at discharge (0.01587 m /s) for the same blade [20].

4. Conclusions
The efficiency produced by the waterwheel Savonius blade is affected by the discharge and the
amount of the blade. In the waterwheel of 4 maximum efficiency blade obtained 9.95% with debit 0.01587
3
m /s at loading 1000 gram. While on the waterwheels 6 blades, the greatest efficiency of 13.93% is
3
obtained at discharge of 0.01587 m /s at 1400 gram loading. While on the waterwheels 8 blades, the
3
greatest efficiency of 17.06% is obtained at a discharge of 0.01587 m /s for with a loading of 1200 grams.
The maximum power of the wheel produced by the water wheel on the flat plate blade occurs at the
3
opening of the discharge III (Q3 = 0.01587 m /s) with the number of 8 blades of 2.403984 W. The
3
greatest efficiency was obtained at 0.01587 m /s discharge of 17.06% with 1200 gram loading. So the
best blade is a waterwheel with 8 blades.

References
[1] Patty O F 1995 Tenaga Air (Erlangga: Jakarta)
[2] http://www.yourarticlelibray.com/powers/distribution-of-hydro-power-in-the-word/25579/
[3] Robert H D 2002 Water Wheel, E-Maill : www.waterhistori.org
[4] Szytula A and Leciejewicz J 1989 Handbook on the Physics and Chemistry of Rare Earths 12 ed K
A Gschneidner Jr and L Erwin (Amsterdam: Elsevier) 133
[5] Nakajima M, Iio S and Ikeda T 2008 Performance of Savonius rotor for environmentally friendly
hydraulic turbine Journal of Fluid Science and Technology 3 3 420-429
[6] Kyozuka Y 2008 An experimental study on the Darrieus Savonius turbine for the tidal current power
generation Journal of Fluid Science and Technology 3 3 439-449 [7] Esty J 2005 Water wheels
(Peabody museum of natural history: Yale University)
[8] Samman M 2005 Rediscovering the waterweel “Noria Al-Muhammadiyah” (Asme Internasional
Histori Lanmark; Asme Northwest Houston Subsection Meeting)
[9] Kusnaidi 1999 Karakteristik Kincir Air (online) available at http://anonime.co.id
[10] Gerholdt M 2009 Modeling the chaotic waterwheel Department of Mathematics Graduate Student
Presentation, Spring
[11] Gerald M, Cristian W 2004 The breastshot waterwheel: design and model tests (Berlin, Germany:
Researcher, Leibniz-Institute of Freshwater Ecology and Inland Fisheries)
[12] Hansen R D 2002 Water wheel (online) available at www.waterhistory.org
[13] Pelz P F and Bearrf T 2006 Studies on the maximum harvestable hydropower in open
channelaplication to a water wheel and a turbine Technische Universitat Germany
[14] Radheika 2008 Kincir pusar untuk irigasi (online) available at http//radheika.com
[15] Tevata A and Inprasit C 2011 The effect of paddle number and immersed radius on water wheel
performance Energy Procedia 9 359-365
[16] Pujol T, Sola J, Montoro L and Pelegri M 2010 Hydraulic performance of an ancient Spanish
watermill Renewable Energy 35 387-396
[17] Ibrahim A G, Haron C C H and Ashari H C 2002 Traditional water wheels as a renewable rural
energy. Department of Mechanichanical and Material Engineering, Unversiti
Kebangsaan Malaysia, Bangi Darul Ehsan Selangtor 436000: Malaysia
[18] Pujol T and Montoro L 2010 High hydraulic performance in horizontal waterwheels Renewable
Energy 35 2543-2551
[19] Denny M 2004 The Efficiency of overshot and undershot waterwheels Eur.J.Phys. 25 193-202
[20] Choi Y D, Lim J I and Lee Y H 2008 Performance and internal flow characteristics of a cross-flow
hydro turbine by the shapes of nozzle and runner blade Journal of Fluid Science and Technology 3
3 398-409
 Model of Vocational Education Learning Format
to Comply
Hopefulness in Companies and Industries
Parabelem Tinno Dolf Rompas & Herry Sumual
Universitas Negeri Manado, Indonesia

____________________________________________________________________________________________________________________
_______________
ARTICLEINFO ABSTRACT
Article history: The companies and industries were very demanding manpower
Received 23 December 2016 not only must be prepared to work in accordance with their
Received in revised form 9
January 2017 competence but also that very important of them was the work
Accepted 21 January 2017 attitude should be good. In fact, currently it was not so because of
Available online 28 February 2017 any manpower that was when received as worker. The study
presents the concept of vocational education learning
format model that appropriate and comply hopefulness of
companies and
Keywords: industries. The method used the literature review, interview,
Vocational education learning creation of the competency-based learning program of work in
Learning format companies and industries. The results have shown that the model
Comply hopefulness of learning format was vary and depending on the characteristics of the subject
of course to be achieved and attitude learning applied in
every learning. Implementation of learning format model
should be supported by together from education leadership
and
Corresponding author: head of companies and industries in achieving the goals and
parabelemrompas@unima.ac.id benefits together.

Introduction

The vocational education is a higher education diploma program that prepares students to work with
particular expertise applied to the applied bachelor program (Lows of Indonesia Republic No. 12 2012). In
the explanation of Article 16 to paragraph one which states that the vocational education is education that
preparing student to become professional with skill or ability of high work. It has historically identified with
their internship in the workplace that the learning process conducted by observation, imitation and
personal correction, not by the application of a preposition common in classrooms and through textbooks
(Moodies, 2008). It is concern on the development of vocational education practical ability, practical
knowledge, and understanding of certain jobs (UNESCO Institute for Statistics 2010). Vocational
education study program of engineering faculty is the unit of education and learning activity that has
curriculum and specific learning method in one type of vocational education under clumps of engineering
science in the faculty of engineering (Regulation of Education and Culture Minister of Indonesia Republic
No. 49 2014).
Coherence and relevance of skill or ability to work from the process result of mechanical
engineering vocational education with the companies and industries were the major factor of priceless
human resources for a community that will promote economic growth and prosperity public (Arifin, 2014).
According to direct observation by interview to the directors in several companies and industries such as:
Automotive Machine Shop, PT. Telkom, Electronics, Tourism and Hospitality, Building Contractors, Food
and Beverage, and Coconut Oil Industry where there are the gap especially in terms of attitude between
the student competence (not only current practice or industrial internship but also already graduated and
worked) and the ability to work. The attitude shown by some student of field work practice (PKL) or
internship industry and graduate or manpower was still lacking so well thus the results of work that
obtained far from the expectation of com-panies and industries (actually expectation of companies and
industries was the work of student and manpower must be professional and can benefit their business).
Base on Presidential Decree No. 8 (2012) on Indonesian Qualification Framework (KKNI) that learning
achievement is the ability obtained by the internalization of knowledge, attitudes, skills, competencies,
and accumulation of work experience. It is useless for the companies and the industries if student of PKL
or industrial internship and graduate or manpower) have high of the ability to internalize the knowledge,
skills, competencies, work experience but very low of the ability of attitude (Rifandi, 2013). Also,
according to the results of directly interview to the heads of vocational education study program in FT-
UNIMA that attitude element included in the creation of learning outcome in the curriculum that created
by Decree of National Education Minister of Indonesia Republic No. 045 (2002) and created in 2003,
Laws of Indonesia Republic No. 12 (2012), Indonesian Presidential Regulation No. 8 (2012), and the draft
Regulation of Education and Culture Minister of Indonesia Republic on National Standard for Higher
Education (creation of the 2013 about revision of curriculum based on KKNI) (The Education and Culture
Ministry, General Directorate of Higher Education and the National Education Standards Agency 2013),
and the lectures had already applied.
Competency standards are minimum criteria concerning the qualifications of graduates capabilities
that include attitudes, knowledge, and skills are expressed in the formulation of graduate learning
outcomes (Regulation of Education and Culture Minister of Indonesia Republic No. 49 2014). The
attitudes are correct and cultured behavior as a result of internalizing and actualizing the values and
norms reflected in the spiritual and social life through a learning process, student work experience,
research and/or community service related learning. The attitude formulation can be added by
universities, proposed to the DirectorGeneral to set into the learning achievement of graduates, assessed
and determined by the DirectorGeneral as a kind of study program. The fulfillment of graduate learning
outcomes for each course the learning process must facilitated by forms of learning such as lectures,
responsiveness and tutorials, seminars, lab, studio practice, practice workshops, or field practice,
research, and community service. The elements of attitudes in the learning achievement of graduates
must become very important elements in the learning process through the container forms of learning,
but the problem is how the model of a learning program with elements of at-titude becomes a key
element in the learning process of each course in vocational technical school? And how do the
implementation?
The problem was particularly important for the companies and industries which really need student
that practice or industrial internship, and graduate or manpower has the attitude that suitable with minimal
standard of good when they were doing the job.
The aim of this study is to get the concept of vocational education learning format model that
appropriate and comply hopefulness of companies and industries.

Method

The beginning, we did library research, direct interview to the director in several enterprises fields of
automotive engineering workshop in the province of North Sulawesi such as Automotive Machine Shop
Company, PT. Telkom, Electronics, Tourism and Hospitality, Building Contractors, Food and Beverage,
and Coconut Oil Industry. Then, we conducted direct interview to the heads of the vocational education
study program in FT-UNIMA such as: Mechanical Engineering Education, Education Electrical
Engineering, Engineering Education Building, Education Information and Communication Technology,
and Education and Family Welfare. We analyzed the interview results, conducted decision-making, and
created a solution through the creation of form model for learning in all subjects. Finally, we planned the
implementation to subject of PKL/industrial apprenticeship in the Vocational Education Study Program
FT-UNIMA on Academic Year 2016-2017 second semester.
Results and Discussion

The result of analysis from the directly interview to the directors in several enterprises in the
province of North Sulawesi indicated that all the directors have the problems in the elements of attitude
as the main element that owned by the student and manpower who currently working was poor
performance. It was very interfere in achieving goal of corporate especially to produce the profits (Batalla-
Busquets & PachecoBernal 2013).
The attitude such as the behavior that not right in the job, for example the workers were often
violated the discipline of the company (in and out of working time was not timely, damage and not
keeping the infrastructure company, work arbitrarily, sometimes dishonest, lacking respect among fellow
workers, never willing to follow orders from seniors). Also, they had knowledge, skills, competencies, and
the accumulation of work experiences is a good average.
We made the model form of learning for next implementation which refer to a form of learning such
as lectures, responsiveness and tutorials, seminars, labors, studio practices, practice workshops, or field
practices, researches, and community services. Then, in the learning process, the selections of learning
very depend on the material characteristics learning each course.
Figure 1 shows the scenario of the learning process for one meeting on a subject which the
emphasis on the process of internalizing and actualizing values and norms are reflected in the spiritual
and social life, especially in the working environment in the enterprise. It contained the formulation of
attitudes according to attachment of Regulation of Education and Culture Minister of Indonesia Republic
No. 49 (2014) as: (1) be cautious to God Almighty and be able to demonstrate religious attitude; (2)
uphold the human values in the line of duty based on religion, morals, and ethics; (3) contribute to
improving the quality of life of society, nation, state, and the progress of civilization based on five
precepts; (4) acting as a citizen of pride and patriotism, nationalism and a sense of responsibility to the
state and nation; (5) respect for cultural diversity, views, religions, and beliefs, as well as the original
findings of an opinion or any other person; (6) cooperate and social sensitivity and concern for the
community and the environment; (7) obeying the law and discipline in the life of society and state; (8)
internalized the values, norms, and academic ethics; (9) show a responsible attitude to work
independently in the automotive field; and (10) internalize the spirit of independence, innova-tion, effort,
and entrepreneurship.
 Figure 1. The scenario of learning process for one meeting on a subject

In the preparation of semester learning plan (RPS) formulation attitude (in the application) are
translated into steps of learning activity (opening activity, main activity, and closing activity). The process
of drafting RPS referred to article 12, paragraph 3 of the Regulation of Minister of Education and Culture
of Indonesia Republic No. 49 (2014) and Indonesian Presidential Regulation No. 8 (2012). During the
process of learning activities conducted observation (monitoring) to student and lectures to measure the
achievement of attitude elements. At the end of the main activities, we evaluated by using the attitude
scale questionnaires. After learning of one meeting, we evaluated and reflected as the basis for
developing a learning plan of next meeting.
The minimum standard of attitude that expected from the results of the learning process is the
attitude that developed when someone worked in the companies and industries that provides a positive
evaluation of someone on the aspects of the work environment of them (Robbins & Judge, 2007). In a job
in companies and industries, the most important of the attitudes should be developed are job satisfaction,
commitment of organizations or enterprises, and job involvement, so that what is expected from the
enterprise to the problems of someone's attitude when doing the work can be solved and resolved
(Akhtar at al. 2016).
Implementation of learning form that involves companies and industries must be supported by
strong global leadership education and business entities in achieving a goal and the advantages of each,
so that the expectations of business and industry can be met and that the perpetrators of street vendors
and workers can be good at her job (Lamsa et al. 2008). If it conducted, then interactive cooperation
between academic activity in the FT-UNIMA with the activity of companies and industries has been good.

Conclusion
 The model forms of learning are varied and depending on the material characteristic of the subjects
to be achieved that base on the formulation of attitudes according to attachment of Regulation of Minister
of Education and Culture of Indonesia Republic No. 49 year 2014 about National Standard for Higher
Education which in the application of learning are elaborated into steps of learning activities. The results
of the implementation of the model forms of learning by focusing on the application of the attitude
learning in each subject can comply hopefulness for the companies and industries when student were on
PKL and graduate student as manpower are working.

References

Akhtar, Omair, Wheeler & Christian, A. 2016. Belief in the immutability of attitudes both increases
and decreases advocacy. Journal of Personality and Social Psychology 111(4): 475-492.

Arifin Z.N. 2014. Model Teaching Industri Politeknik Negeri Jakarta (Industry Based Vocational
Education System / IVE-PNJ System), (http://dikti.go.id/blog/2014/02/10/ model-teaching-industri-
politeknik-negerijakarta-industry-based-vocational-education-system-ive-pnj-system/), access on 10
Desember 2014.

Batalla-Busquets, J. & Pacheco-Bernal, C. 2013. On-the-job e-learning: workers’ attitudes and

perceptions. IRRODL The International Review of Research in Open and Distributed Learning 14(1): 40-
64.

Decree of Minister of National Education of Indonesia Republic No. 045 / U / 2002 Higher
Education Main Curriculum.

Indonesian Presidential Regulation No. 8. 2012. Indonesian Qualifications Framework.

Lamsa, A., Vehkapera, M., Puttonen, T. & Pesonen, H.L. 2008. Effect of Business education on
women and men students’ attitudes on corporate responsibility in society. Journal of Business Ethics
82(1): 45-48.

Lows of Indonesia Republic No. 12. 2012 Higher Education.

Ministry of Education and Culture Directorate General of Higher Education and the National Education
Standards Agency. 2013. The draft Regulation of Minister of Education and Culture of the Republic of
Indonesia about National Standard for Higher Education (SNPT).

Moodies, G. 2008. From Vocational to Higher Education: an International Perspective. London:

McGrawHill International (UK) Limited.

Regulation of Minister of Education and Culture of Indonesia Republic No. 49. 2014. National
Standards for Higher Education.

Rifandi, A. 2013. Mutu Pembelajaran dan Kompetensi Lulusan Diploma III Politeknik. Cakrawala
XXXII(1): 125-138.

Robbins & Judge. 2007. Perilaku Organisasi. Salemba Empat.

UNESCO Institute for Statistics. 2010. International Standard Clasification of Education (ISCED)
2011Draft, For Global Consultation June-October. p. 6, United Nations Educational, Scientific and
Cultural Organization.
View publication stats