See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/258022869

Experimental Ocean Thermal Energy Conversion (OTEC) project on the

Reunion Island

Conference Paper · January 2012

DOI: 10.13140/RG.2.1.2926.1926

CITATIONS READS

5 1,893

3 authors, including:

Frantz Sinama Franck Lucas

University of La Réunion University of La Réunion
35 PUBLICATIONS   299 CITATIONS    88 PUBLICATIONS   590 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

MéGaPICS View project

Chaleur Solaire pour l'Industrie en Région Tropicale (CSIRT) View project

All content following this page was uploaded by Frantz Sinama on 25 March 2015.

The user has requested enhancement of the downloaded file.

 4th International Conference on Ocean Energy (ICOE 2012), 17th October, Dublin

Experimental Ocean Thermal Energy Conversion (OTEC)

project on the Reunion Island
A. Journoud1,2, F. Sinama2 and F. Lucas2
1
DCNS Nantes Indret, Indret
44620 La Montagne,
2
PIMENT laboratory, 40 avenue de Soweto
97410 Saint Pierre, Reunion Island

Abstract These assessments have lead governments to encourage

Ocean Thermal Energy Conversion (OTEC) is a projects for developing renewable energies. The purpose is to
process able to produce base electricity through the use resources available on earth in large quantities and have
temperature differences existing between warm seawater the capacity to replenish itself, to produce electricity in a
at the surface and cold depth seawater. “green” fashion: wind power, solar power, water power etc.
Up to now, very few experiments have been conducted This type of energy can be called infinite energy [3].
and oil prices fluctuations have often prevented results One of these projects is the GERRI program on the
exploitation. Recently, two factors have re-launched Reunion Island. It encourages the decrease of electricity
researches: a need for the production of a renewable consumption and greenhouse gas emission. The Reunion
energy ensuring the stability of the electric grid and the Island is no exception. Indeed, in the number of inhabitants
increase in oil prices. The purpose of this experiment is to increases too (1.55% increase per year) [4].
demonstrate the sustainability of an offshore 10 MW The objective of the politic based on renewable energies
power plant. Some issues first need to be examined: heat for the Reunion Island is to be energetically independent.
exchanges, control strategies, cycle risks, working fluids, This island is located north of the tropic of Capricorn, so it
and thermodynamic cycles. becomes the best candidate to use the Ocean Thermal Energy
Conversion.
This paper presents an experimental prototype as a Indeed, this process produces electricity from the
first steppingstone of an “Eliminating Risks Program” temperature difference between warm water at the surface of
and as a preliminary installation to the setting up of the the ocean, and cold depth seawater. One of the advantages of
first full scale OTEC power plant. The prototype is a this energy is that it might provide base-load electricity. The
reproduction on a reduced scale of a demonstration plant thermal resources of the ocean ensure an all day availability of
producing an electric power of 10 MW. To reduce costs the power source and with only modest variations from
and hazards of offshore installations (waves, cyclones, summer to winter [5].
maintenance, etc.), decision has been made to build an Before building an OTEC power plant on the Reunion
onshore plant (settled on the Réunion island), creating Island some matters need clarification.
artificial heat sources with a heat pump. The first step is to study heat exchanges, cycle risks, working
In the first part of the paper, the principle of OTEC fluids, thermodynamic cycle etc. To that purpose, an onshore
will be explained. Then, the prototype detailed: cycle, prototype has been built. It functions with artificial heat
main components (the electricity production is directly resources and without alternator. Therefore the prototype is
related to the quality of the heat exchanges, so some not connected to the sea and to the grid.
different heat exchanger technologies are developed), The second step will be the construction of an offshore 10
working fluids, heat sources, and control strategies. MW OTEC power plant. It will cover the electricity
Finally, results of an experimental parametric analysis consumption of roughly 30 000 homes. Control strategies,
will be presented, highlighting the major importance of maintenance issues, water pipes and pumps etc. will be
heat exchanges, within the evaporator. Results show, for clarified.
instance, that a vertical plate-type evaporator show This paper presents the onshore experimental set up and
better capacities than a shell and tube one. the experimentation to compare two evaporator technologies
and performances. Indeed, as Sinama et al. [6] have
established, the most important component in the OTEC cycle
1. INTRODUCTION is the evaporator. The challenge here is to generate the highest
Nowadays, the expansion of renewable energies can be heat exchange with the lowest quantities of working fluid (for
explained through international facts: indeed, between 1970 security and cost reasons).
and 2004, greenhouse gas emission increased by 70%, and
fossil energy prices increased by 50% [1]. Furthermore, the 2. THE PRINCIPLE OF OTEC
number of inhabitants on earth increased: in the beginning of
2012 to roughly 7 billion. Moreover, the standards of living The principle of the Ocean Thermal Energy Conversion is
improved with phone, car and computer advances, so did the to convert into electricity the differences in temperature
need for electricity [2]. between surface seawater and the depth ocean seawater. Some
cycles can be used to that purpose: Rankine, Uehara, Kalina.

1
 4th International Conference on Ocean Energy (ICOE 2012), 17th October, Dublin

The first one is the most common cycle used because of its tightness of the pipe’s process division, individual equipment
simplicity and its proven results. protection etc.
Furthermore, in terms of explosiveness, ammonia is safe
2.1. Rankine enough: the interval of explosiveness is from 16 to 28% of the
The Rankine cycle is a thermal machine which produces room volume and its self-flammability temperature is very
electricity from a hot and a cold source. This cycle is high, 651°C [11].
illustrated on Figure 1 and the (P;h) diagram is drawn on Despite of theses risks are known and controlled, if the
Figure 2. quantity of ammonia is reduced, risks and cost will be too.
In number 1, the ammonia is in its liquid phase under a
low pressure. The pump supplies the working fluid to the 2.3. Heat sources
evaporator where it is heated and vaporized (2) by the warm The temperature differential must be of at least 20°C and
seawater from the sea surface. The generated high pressure stable enough. This is the reason why the tropical zone is
vapour flows into the turbine (3) and its enthalpy is considered as the best geographic location on earth to develop
converted into work and then in electricity with the the OTEC plant. Indeed, all requirements are met: 4.9°C cold
alternator. The low pressure vapour exits the turbine and is seawater can be found at 1000 m deep and a temperature over
led to the condenser where it is liquefied (4) by the depth 25°C at the surface. This area represents 1/3 of the ocean's
seawater. The liquid in the condenser is pumped back to the surface, and could produce 10 million MWe [12].
upper heat exchanger and a new cycle begins [7].
3. EXPERIMENTAL INSTALLATION
An experimental installation has been built by a French
firm: DCNS. It has been decided to realise an onshore
prototype to avoid sea related issues such as anchorage,
floating platforms, 1000 m long flexible to pump cold water
etc.
Concerning its power, it is a 1/200 scale prototype in
comparison to the first 10MW OTEC plant. It is designed to
produce the equivalent of 15 kW. But for scale and cost
reason, no alternator has been installed.
Besides, for scale reasons, this onshore prototype doesn’t
produce electricity.
Cold and hot sources are simulated by a heat pump, so that
the hot water temperature can be adjusted to simulate the hot
seawater temperature round the year (from 23 to 28°C).
Figure 1: Schematic Rankine cycle Concerning the cold water temperature, it is stabilized to
around 5°C.
To bring the prototype as close as possible to the 10MW
OTEC plant, it is necessary that they have the same working
pressure.
The prototype has been designed for having stable regime.

3.1. Ammonia loop and its main components

On Annex 1, the ammonia loop is drawn.
The ammonia loop has been designed to extrapolate results on
this loop to the first OTEC power plant ammonia loop.
The components are going to be presented in the
circulating flow way.
The PAF is the pump which makes liquid ammonia flow
Figure 2: P;h diagram of the Rankine cycle with ammonia into the pipes.
Then ammonia is vaporized in the evaporator.
2.2. Working fluid The pseudo-turbine replaces the turbine. It reproduces the
The working fluid used here is ammonia. This choice has same thermodynamic effect as a turbine (isentropic
been made for different reasons: first of all, it is a fluid expansion).
naturally present in the environment and its evaporating
pressure is low in comparison to other refrigerating fluids
[8], then its latent heat is the highest after water’s (907.7
kJ/kg [9]), for one power ammonia has the highest
performance [10], moreover its molar mass is one of the
lowest [9] which leads to a decrease of the pipe diameter
[10].
However, ammonia presents some risks. A vapour leak from
1700 ppm could lead to respiratory damages in the human
body [11], and a liquid leak if there are eye or skin contact,
could lead to cold burns [8].
The advantage is that all these risks are well known and
a series of security measures have been taken: ammonia
detectors, ventilation, ammonia training for staff, high Figure 3: p-h diagram "as a guide"

2
 4th International Conference on Ocean Energy (ICOE 2012), 17th October, Dublin

shell-and-tube evaporator, and in Table 1, detailed

Thereby, an expansion valve and a desuperheater have information are presented.
been installed. In Figure 5, this is explained. Before the
turbine, the working fluid is in its vapour form (after the
evaporator) at high pressure. The “isentropic expansion” line
represents the path that follows the ammonia with a real
turbine. So the working fluid is expanded as an isenthalpic
process through the expansion valve. Then, the temperature
of the vapour decreases when injecting cold liquid ammonia
through the desuperheater.

Number Passes Length Tubes’ Exchange Ammo-

of tubes number (mm) materials surface nia
(m²) volume
(L)
280 4 tube 3050 Titanium 39.6 17
passes bare
tubes
Table 1: Condenser information
Figure 5: Flooded shell-and-tube evaporator

Number Passes Length Tubes’ Exchange Am-

of tubes number (mm) materials surface monia
(m²) volume
(L)
259 2 tube 4911 Titanium 64.8 270
passes bare
tubes
Table 2: Flooded evaporator specification
• Plate evaporator is a vertical aluminium brazed plate
evaporator. Figure 4 illustrate of this evaporator. It is
composed of a series of aluminium plates grazed together.
Then between each plate, another aluminium wave is
brazed. The working fluid flows through the small tubes
and water through the large one. Please note that the
aluminium wave is hollow in order to ease fluid flow and
heat exchange. Finally, ammonia and hot water enter from
the bottom of the evaporator and are injected at the same
Figure 4: Shell-and-tube condenser time between the aluminium plates. Please note that
ammonia is evaporated vertically so it is hard to know
Ammonia vapours are condensated into the condenser. exactly the liquid quantity. The Table 2 details information
The condenser installed on the onshore prototype is a shell- on the aluminium brazed plate evaporator.
and-tube exchanger. Water circulates through the tubes, and
ammonia around in the grille. The working fluid in its
vapour form arrives on top of the grille. At tubes’ touch, the
vapour condensates and flows at the bottom in the grille. It is
paramount for the performance of the condenser not to have
the tubes flooded. In Figure 6, there is the representation of a
condenser used on the onshore prototype. In Table 3 gives
some information about the condenser.
Finally, ammonia is pumped again through the PAF
pump.
3.2. Evaporators Figure 6: Aluminium brazed plate evaporator top view [13]
After the turbine, the evaporator is the most important
For the plate evaporator tests, a separator has been added
component in electricity production [6]. Indeed, the higher
(annexe 3) because the outlet quality of vapours
the output pressure, the higher is the electricity production.
is unsure.
This is the reason why two different evaporators are tested
on this onshore prototype: a flooded shell-and-tube
evaporator and an aluminium plated evaporator. High Plates’ Exchange Ammonia
• Flooded shell-and-tube evaporator is the most widely used (mm) materials surface volume
type of heat exchanger for industrial evaporator (m²) (L)
applications [7]. Water flows through the tubes and the 2978 Alumi- 69 37
working fluid arrives on the bottom of the grille and nium
floods the tubes. The nucleation appeared around the tubes Table 3 : Plate evaporator information
surface and vapour is evacuated through the top of the
grille. In Figure 3 there is a cross-section of the flooded

3
 4th International Conference on Ocean Energy (ICOE 2012), 17th October, Dublin

3.3. Water loops The thermal surface power of the plate evaporator is upper
The two water loops of the prototype are very different than the flooded evaporator one. Indeed, at the warm water
from the OTEC plants. Indeed, a heat pump produces heat temperature of 28 °C, the thermal surface power of the plate
sources, the pipes’ geometry, pumps etc. They have been evaporator is 4.1 % higher.
designed to reproduce the ocean temperature only. The mass Even if the plate evaporator exchanges more thermal power
flow-rate is calculated following the similitude between the than the flooded one, another important parameter to compare
ammonia loop of the prototype and the ammonia loop of the is the outlet evaporator pressure. Indeed, this value is the inlet
OTEC plant. Therefore, none of the results from water loops data for the turbine. The more it is high, the more the
will be significant for the Rankine cycle. alternator produces electricity.
The components and principle of cold and hot water 940
Flooded evaporator
loops are similar. Annex 2 presents the supervision image of 920
the hot water. 900 Plate evaporator

Pressure (kPa)
The temperature requirement is given in the box entitled 880 Linéaire (Flooded
“heating pump”. Water circulates in the heating pump thanks evaporator)
860 Linéaire (Plate evaporator)
to the pump “EPEC2”, and thanks to the “EPEC 1” in the 840
evaporator. The separating bottle delivers a constant water 820
flow in the heat pump, while controlling the water flow in 800
the evaporator. 780
760
4. TESTS AND RESULTS 23 24 25 26
Temperature (°C)
27 28

4.1. Tests Figure 8: Evolution of the outlet ammonia pressure for both flooded
and plate evaporators
A test is defined by the value of several parameters: hot Figure 8 details the outlet ammonia pressure of the two
water temperature, cold water temperature, hot water mass evaporators depending on the hot water.
flow rate, cold water temperature and eventually ammonia The two evaporators have an evaporation pressure similar.
mass flow rate.
For the study of the prototype using flooded evaporator, As shown above, the plate evaporator exchanges more
four points have been tested. Five points have been tested for thermal power but with the same outlet pressure. So, it would
the plate evaporator. The warm water temperature ranges have been interesting to study the global electricity production
from 23 to 28°C, cold temperature is at 4.9 °C. Flow rates (Pbrute) minus the electricity consumption for the auxiliaries
are calculated to keep similitude between the prototype and (ammonia pump, cold water and hot water pumps) : defined as
the first OTEC power plant. the net electric power.
4.2. Results and discussion Pnet = Pbrute − PNH 3 pump − Photwaterpump − Pcoldwaterpump (4.3)
All data used to calculate the thermal power have been As it is written above, a pseudo-turbine replace the turbine
averaged on a stable period of a minimum of 20 minutes. in the Rankine cycle. The thermal power in the pseudo-turbine
is calculated from the enthalpy difference between the point 3
The thermal power reflects the energetic performance of and 4 (Figure 1). But, to calculate an electrical power, these
the evaporator. The two evaporators have not the same hypothesis must be taken: a 100% turbine and alternator
exchange surface. So, to compare their energetic efficiency.
performance, it is necessary to compare the thermal power First, the thermal power which is lost in the pseudo-
divided by their exchange surface: the thermal surface turbine is calculated as follow (see Figure 1):
power. Pturbine = m& ft (h3 − h4 ) (4.4)
In an evaporator, the refrigerant begins to warm up, and
when it reaches its saturation temperature, it is vaporized. Secondly, the ammonia pump consumption is calculated
15
from the constructor consumption curves.
As it is detailed in the paragraph 3.3, water loops have not
14.5
been designed to extrapolate water results on the OTEC plant.
Surface thermal power (kW/m²)

14 So, water loops and ammonia loop (for the prototype) does not
13.5
Flooded evaporator have the same scale. The two water pump consumption can be
calculated, but including them in the 4.3 equation is
13 Plate Evaporator
meaningless.
12.5
Linéaire (Plate Pnet’ is the electrical power produced minus the ammonia
12 Evaporator) pump consumption: Pnet ' = Pbrute − PNH 3 pump (4.5)
Linéaire (Flooded
11.5 evaporator) Figure 9 compares Pnet’divided by the exchange surface
11 for each evaporator (as in the study of Avery and Wu [6]).
23 24 25 26
Temperature (°C)
27 28 29 According to Figure 9, the plate evaporator produces more
Figure 7: Evolution of the thermal surface power in function of the electricity per m² than the flooded one. Indeed, for a hot water
warm water temperature temperature of 28°C, the alternator in the plate evaporator
So, the thermal power is the addition of the reheat power version produces 26% more electricity than the flooded
and the evaporation power. Here is the equation: version.
Pth = Preheat + Pevap (4.1)
Pth = m& wCp w (Tew − Tsw ) + m& ft (hsft − heft ) (4.2)
Figure 7 shows the thermal surface power of the two
types of evaporator in function of the hot water temperature.

4
 4th International Conference on Ocean Energy (ICOE 2012), 17th October, Dublin

The most important difference between the two

0.14 evaporators is the quantity of ammonia. As written above, the
Surface net power (kW/m²)

0.12 plate one need only 37L versus 270L for the flooded one and
produces more net electricity power.
0.1 Nevertheless, the plate evaporator has some drawbacks:
0.08 the intern technology lead to increasing charge loss (water
pump consumption increase), fouling issue, and corrosion of
0.06 Flooded evaporator the material (aluminium), addition of a separator (space lost
Plate evaporator
0.04 Linéaire (Plate evaporator) and higher cost).
Linéaire (Flooded evaporator) All those drawbacks lead to investigate on water pumps
0.02 consumption (On the OTEC power plant, are the charge loss
23 24 25 26 27 28
Temperature (°C) generated by the plate evaporator insignificant in front of the
Figure 9: Evolution of the surface net power of the two evaporators
charge loss of the water pipes?) on other evaporator
depending on the hot water temperature technology (a falling film one will be tested), other
thermodynamic cycles etc.
In the future, the prototype can be used to validate the
Concerning the water pumps consumption, it depends on simulation model established by Sinama [6] with the Gibbs
few parameters: mass flow-rate, charge loss and pump method.
technology (efficiency). The goal is to compare the electrical
consumption of the pumps between the two evaporators. As
the cold water loop does not change between the two tests, 6. REFERENCES
the cold water pumps consumption does not change either. [1] GIEC, Changements climatiques, Rapport de synthèse,
Then, for the hot water loop, the evaporator changes only. 2007.
Moreover, the mass flow-rates for each points are similar
between the two evaporator. So, the comparison of the [2] Nihous G., An Order-of-magnitude Estimate of ocean
charge loss in the hot water loop between the two tests thermal energy conversion Resources, Journal of Energy
reflects the water pumps consumption between the two Resources Technology, 2005.
evaporators. And finally, between the two tests, the [3] www.ifpaenergyconference.com (visited on April 2012)
evaporator is the only thing which changes in the hot water
loop. Therefore, the comparison of the charge loss in the [4] Praene J.P., David M., Sinama F., Moreau D., Marc O.,
evaporator reflects the comparison between the water pumps Renewable energy: Progressing towards a net zero energy
consumption. Island, the case of Reunion Island, Appled energy, 2011.
[5] ABS Energy Report, The Ocean Energy Report, 2007.
[6] Martins M, Sinama F, Lucas F, Equivalent Gibbs systems
Charge loss (mbar)

Flooded evaporator
for modelling an onshore OTEC experimental plant in
Plate evaporator Reunion Island, International Journal of Energy
Linéaire (Flooded Research, 2011.
evaporator) [7] Avery H., Wu C., Renewable energy from the ocean, A
Linéaire (Plate
evaporator)
guide to OTEC, 1994.
[8] IIAR green paper, Ammonia: the natural refrigerant of
choice.
35 40 45 50 55
Mass flow rate (kg/s)
[9] Tchanche B et al, Fluid selection for a low-temperature
Figure 10 : Evolution of the water charge loss in the flooded solar organic Rankine cycle, Applied Thermal
evaporator and the plate one Engineering, 2009.
Figure 5 show the charge loss for the two evaporators.
But, the conclusion is: charge loss of the plate evaporator is [10] Pearson A., Refrigeration with ammonia, 2007.
higher than the charge loss for the flooded one. But, because
[11] GDF Suez, Formation ammoniac, 2011.
of the scale effects (paragraph 3.3) the question of “how
higher?” has no answer with the prototype. [12] Vega L. A., Economics of Ocean Thermal Energy
Conversion: an update, Offshore technology conference,
5. CONCLUSION Houston, Texas, 2010.
This paper has presented the onshore OTEC prototype [13] www.fivesgroup.com (visited on April 2012)
built by the firm DCNS. It has reminded the OTEC principle
and exposed the specificities of this prototype: it does not 7. NOMENCLATURE
produce electricity, and hot and cold water are produced by a
heat pump. Pth: Thermal power of the evaporator (kW)
Then, the different components have been shown, with a Preheat: Power generated by the reheat part of the evaporator
focus on the two evaporators tested (a flooded shell-and-tube (kW)
evaporator and a brazed plate evaporator). Pevap: Evaporation power of the evaporator (kW)
First, all tests permit to conclude that the plate Pturbine: Thermal power lost by the working fluid in the
evaporator exchange more heat per square meter than the turbine (kW)
flooded one with the same evaporation pressure. Then the m& w : Hot water mass flow rate (kg.s-1)
comparison of the net electrical power (calculated on the Cp w :
ammonia loop) between the two evaporators has been made, Thermal capacity at constant pressure of the water
leading to conclude that the plate one produces more (kJ.K-1)
electricity. Tew : Inlet water temperature (°C)

5
 4th International Conference on Ocean Energy (ICOE 2012), 17th October, Dublin

Tsw : Outlet water temperature (°C) heft : Inlet working fluid mass enthalpy (kJ.kg-1)
m& ft : hinlet: Working fluid mass enthalpy which arrives in the
Working fluid mass flow rate
turbine
(kg/s)
hinlet: Working fluid mass enthalpy which goes out of the
hsft : Outlet working fluid mass enthalpy (kJ.kg-1) turbine

6
4th International Conference on Ocean Energy (ICOE 2012), 17th October, Dublin

Annex 1

Figure 11: Ammonia cycle in the flooded evaporator version

Annex 2

Figure 12: Warm water cycle

7
 4th International Conference on Ocean Energy (ICOE 2012), 17th October, Dublin

Annex 3

Figure 13: Ammonia cycle in the plate evaporator version

View publication stats