Version of Record: https://www.sciencedirect.

com/science/article/pii/S0038092X18311459
Manuscript_a236de0d904047bb71db13c5165ff3f3

Optimization and assessment of floating and floating-

tracking PV systems integrated in on- and off-grid hybrid

energy systems

Pietro Elia Campanaa,b, Louise Wästhagea, Worrada Nookueaa, Yuting Tanb, Jinyue Yana,b

a
School of Business, Society and Engineering, Future Energy Center, Mälardalen University,

72123, Västerås, Sweden

b
Department of Chemical Engineering, KTH - Royal Institute of Technology, 10044,

Stockholm, Sweden

Abstract

Considering the targets of Thailand in terms of renewable energy exploitation and

decarbonization of the shrimp farming sector, this work evaluates several scenarios for

optimal integration of the hybrid renewable energy systems into a representative shrimp farm.

In particular, floating and floating-tracking PV systems are considered as alternatives for the

exploitation of solar energy to meet the shrimp farm electricity demand.

By developing a dynamic techno-economic simulation and optimization model, the following

renewable energy systems have been evaluated: PV and wind based hybrid energy systems,

off-grid and on-grid PV based hybrid energy systems, ground mounted and floating PV based

hybrid energy systems, and floating and floating-tracking PV based hybrid energy systems.

From a water-energy nexus viewpoint, floating PV systems have shown significant impacts

on the reduction of evaporation losses, even if the energy savings for water pumping are

moderate due to the low hydraulic head. Nevertheless, the study on the synergies between

water for food and power production has highlighted that the integration of floating PV

© 2018 published by Elsevier. This manuscript is made available under the Elsevier user license
https://www.elsevier.com/open-access/userlicense/1.0/
represents a key solution for reducing the environmental impacts of shrimp farming. For the

selected location, the results have shown that PV systems represent the best renewable

solution for integration into a hybrid energy system due to the abundance of solar energy

resources as compared to the moderate wind resources. The integration of PV systems in off-

grid configurations allows to reach high renewable reliabilities up to 40 % by reducing the

levelized cost of electricity. Higher renewable reliabilities can only be achieved by integrating

energy storage solutions but leading to higher levelized cost of electricity. Although the

floating-tracking PV systems show higher investment costs as compared to the reference

floating PV systems, both solutions show similar competiveness for reliabilities up to 45 %

due to the higher electricity production of the floating-tracking PV systems. The higher

electricity production from the floating-tracking PV systems lead to a better competitiveness

for reliabilities higher than 90 % due to lower capacity requirements for the storage systems.

1 Introduction

Two concerns growing along with the increase in energy demand are energy security and

global warming. Renewable energy alternatives have been emphasized within many different

business areas all around the world due to their capability of energy security enhancement and

greenhouse gas mitigation in comparisons with conventional fossil fuels [1]. However, the

major barrier of renewable alternatives is related to their high investment costs, while the

fossil fuels cost remain lower. Therefore, not only the environmental perspective but also the

economic profitability and the social wellbeing are needed to be considered for the evaluation

of suitable alternatives [2].

As with most of the Southeast Asian counties, Thailand is experiencing a rapid energy

consumption increase due to population growth. The country has almost tripled the energy

consumption within a period of 20 years [3]. Currently, the country is the top two largest
energy consumer in the Association of Southeast Asian Nations (ASEAN) [4].

Correspondingly, the greenhouse gas emission has almost doubled from the emission levels in

1990 since the majority of energy sources are fossil fuels (natural gas, oil, and coal).

Moreover, the projections for the Thailand energy sector show that by 2035, the estimated

imported fossil fuels will reach up to 90%. Therefore, energy security is a priority for the

Government of Thailand. The overall goal of the national plans set by the Department of

Alternative Energy Development and Efficiency is to reach 25% renewable energy of the total

energy consumption within the country by 2030 [5]. Hence, the installed renewable energy

systems must increase rapidly in the near future to reach the national target, since the

electrical demand will almost double by 2030 [4].

Among the ASEAN countries, Thailand has one of the highest solar capacity potential, which

makes the solar power one of the best renewable energy. The average solar irradiation in

Thailand is 18.0 MJ/m2/day [1, 5]. However, only 7.2% of the produced electricity are from

the renewable energies in 2015 [6]. The feed-in-tariffs (FIT) policy has been implemented

since 2007 by the Government of Thailand to encourage and support the private sector

producers. The FIT programme covers the electricity produced from biomass, biogas, solar,

wind, hydro and waste. The current average electricity price in Thailand is 0.1 US$/kWh. The

FIT varies between 0.09 – 0.21 US$/kWh, depending on the technology. For ground mounted

PV systems serving the agricultural or shrimp-farming sector, the FIT is approximately 0.16

US$/kWh [7].

Thailand is one of the main aquaculture producer worldwide with 0.9 million tonnes in 2015

[8]. In 2016, the country was the sixth biggest producer of shrimps and prawns worldwide

with almost 330,000 t [9]. One of the major challenges of shrimp cultivation is the energy

demand. Shrimp farming needs 24 hours surface and bottom aeration. Intensive pumping is

also needed to transfer large volumes of water from the sea to the water treatment ponds, the
nursery ponds, the grow-out ponds, and for the slurry waste discharge. The intensive energy

consumption does not only raise the operating cost but it is also associated with greenhouse

gas emissions. Most of the farms are connected to the electrical grid. However, there is still a

small percentage of shrimp farmers who still rely on diesel generators. The monthly

electricity consumption of a shrimp farm was reported in the work carried out by Nookuea et

al. [10]. A 6,400 m2 single shrimp farm required in average 15-20 MWh/month mainly for

aeration and pumping. By installing solar PV systems to meet the required electricity

consumption, the shrimp farms can become more energy self-sufficient, generate lower

environmental impacts, and have better economic outcome due to the competitive levelized

cost of electricity (LCOE). However, due to the low efficiency of solar PV modules, PV

systems require large installation areas. This limitation can be overcome by the

implementation of floating PV systems on the available water surface such as the surface of

the water treatment pond.

As compared to the previous study carried out by Nookuea et al. [10], this work aims to study

the optimization of floating PV systems for shrimp farm cultivation in Thailand from an

environmental, technical and economic viewpoint. At first, a short review of the cooling

effects of floating PV systems is presented. The results of the review are then used in a

techno-economic optimization model to study the effects of cooling on the optimization of

hybrid energy systems for meeting the intensive energy requirements of a representative

shrimp farm. This is done to also study how renewable energies can improve the sustainability

of shrimp farming and thus support the shrimps market. The water-energy nexus aspects of

floating PV systems are also investigated with special consideration to the reduction of

evaporation losses and mutual use of water for power and food production, and wastewater

treatment. This study also evaluates the advantages of integrating one-axis floating-tracking

PV systems by providing the cost assessment for an unconfined floating structure. Different
scenarios regarding the implementation of renewables into shrimp farming have been

considered in this work, including ground mounted PV systems and wind turbine hybrid

energy systems with storage, fixed floating and floating-tracking PV systems, and off- and on-

grid solutions.

2 Floating PV systems literature review

In recent years, floating PV systems have gained more attention from both a research point of

view and a market perspective due to the direct and indirect benefits related to their

installation [11]. At the end of 2016, the existing worldwide floating PV installed capacity

was more than 94 MWp. Japan represented 60% of the world installed capacity with 56 MWp,

followed by China (20 MWp), UK (10 MWp) and South Korea (6 MWp) [12, 13]. In 2018, the

floating PV capacity reached 211 MWp, only considering the top 70 floating PV installations

worldwide as shown in Figure 1 [14]. Japan has continued to lead the floating PV market

thanks to the high FIT and lack of suitable land for large-scale PV systems.

Figure 1: Installed capacity of the top 70 floating PV systems per country in 2018

(MWp) [14].

Floating PV systems are marked out by a higher electrical efficiency due to the cooling effects

of the water body on the solar panels. This increases the annual electricity production in
combination also with an increased albedo and thus reflected solar radiation. Several recent

studies have proved that PV systems installed on water bodies can have higher conversion

efficiency as compared to ground mounted PV systems due to the cooling effects of water.

Choi studied the performance of three floating PV systems on a reservoir in South Korea [15].

The 2.4 kWp floating PV system reached an average conversion efficiency 7.6% higher than a

reference ground mounted PV system. The 100 kWp floating PV system reached a conversion

efficiency of 17.6%, 13.5% higher compared to the ground mounted PV system. The 500 kWp

floating PV system reached an efficiency 10.3% higher than the reference PV system. Majid

et al. carried out an experimental study by using a 80 Wp floating PV system installed in a

pond simulator, which showed an increase of the power gain from 5.9% up to 15.5% as

compared to the reference ground mounted PV system [16]. A summary of the cooling effects

on the floating PV systems efficiency is presented in Table 1. In most of the works on the

cooling effects of floating PV systems, very limited attention was given to those effects on the

system optimization. In this study, we have focused on how the cooling effects can influence

the system optimization.

Table 1: Summary of the cooling effects on the efficiency of floating PV systems reported in

previous studies.

Increased efficiency due to cooling effects compared to a reference ground mounted PV system Reference

7.6 %, 13.5% and 10.3% [15]

15.5% (increment of energy gain compared to a normal PV during a two hour experiment) [16]

9% [17]

1.58-2.00 % [18]

0.79%. [19]

2.82 %-14.58 % [20]

From a water-energy nexus perspective, floating PV systems decrease evaporation rates from

water bodies. In hot and dry climates, also considering the future adverse effects of climate

changes on temperature and precipitation patterns, evaporation losses from water bodies

represent a threat for the water resources management, in particular for irrigation purposes as

well as for power production. Rosa-Clot et al. investigated the integration of floating PV

systems in wastewater treatment plant in Australia, focusing mainly on the avoided

evaporation losses [21]. The authors concluded that, due to the cooling effect of the waste

water, floating PV system can reduce evaporation losses ranging from 15,000 to 25,000

m3/MWp with further several electricity generation benefits as compared to ground mounted

PV systems. Helfer et al. estimated that 40% of the reservoir water storage capacity is lost

through evaporation in Australia [22], and climate change projections on rising temperature

could led to 15% higher evaporation losses. Bengoechea et al. estimated that evaporation

losses in agricultural water reservoirs can reach up to 17% in Spain [23]. Santafé et al.

calculated a water loss reduction of 25% after the installation of a floating PV system on a

water irrigation reservoir in Spain [24]. Sahu et al. concluded that floating PV systems can

prevent water storage capacity losses to up to 33% on natural lakes and ponds, and up to 50%

on human-made water storage facilities [11]. A water-food-energy nexus approach was used

by Pringle et al. in describing the concept of aquavoltaics that is the combination of floating

PV systems with aquaculture [25]. The authors reviewed several applications in which the

synergies between floating PV systems and aquaculture can be strengthened for water, energy

and food security. From an environmental point of view, beside reducing CO2 emissions as

for other PV systems, floating PV systems can also decrease the algae growth and reduce the

erosion of the bank by minimizing the negative effects of the waves. From an economic point

of view, floating PV systems are also attractive because they do not require land for the
installation, especially in areas where the cost of land represents an important item cost for the

entire project investment. The environmental and economic aspects related to floating PV

systems are particularly important for the shrimp farming sector because they represent a

sustainable solution to the environmental pollution created by shrimp farms and the high cost

of land [10, 26]. Most of the published studies have described the energy-water nexus concept

and investigated only one part of the energy-water nexus. In this study, we have highlighted

the relationships between the water losses and energy requirements for shrimp farms.

Moreover, we have also highlighted the relationships among the combined use of water for

food production, power production and wastewater treatment. These represent key issues for

reducing the environmental impacts of shrimp farms both from a greenhouse gases point of

view and from a water pollution point of view.

Similar to other solar tracking systems, floating-tracking PV systems can achieve higher

electricity production as compared to fixed systems. The tracking system follows the sun

patterns to minimize the angle of incidence and thus maximize the direct fraction of the solar

radiation hitting the PV surface. Contrary to from ground based PV tracking systems,

floating- tracking PV systems have greater rotation capacities [27]. One of the first floating-

tracking PV system was installed in Italy in 2010, followed by the project at Lake Colignola

in Italy in 2011 [28]. Cazzaniga et al. recently presented a review on the performances and

design solutions of floating PV systems [29]. The authors provided a comparison between

submerged and floating PV systems, and analysed different supporting structure designs,

including floating-tracking PV system, and cooling techniques. The authors reported the

performances analysis of two floating PV systems operating in Italy. The authors also

presented an interesting floating-tracking solution without confinement. This cost effective

solution consisted of connecting the centre of the floating platform to a submersible concrete

anchor through a mooring chain. The rotation was guaranteed through a sun-tracking
algorithm that drives submerged propellers. The solution was implemented and tested for a

pilot floating PV systems in Italy. A schematic diagram of a floating-tracking solution without

confinement is presented in Figure 2.

Figure 2: Schematic diagram of a floating-tracking PV system solution.

Several other studies have focused on the detailed structural design of floating PV systems

[30, 31] and floating-tracking PV system structures [32, 33]. Nevertheless, few studies have

focused on the economic aspects related to the implementation of floating-tracking structures.

From an energy system point of view, in a recent study conducted by Silverio et al. [34], the

authors studied the coordinated operation of hydroelectric and floating PV power plants in the

São Francisco River basin. The authors concluded that the integration of floating PV systems

could increase the capacity factor of hydroelectric power plants by 17.3% with the potential

effects of replacing thermal power plants. Trapani and Millar studied the integration of off-

shore floating thin-film PV systems in Malta [35] with special consideration to optimal

capacity designs and CO2 emission reductions. An interesting energy system aspect addressed

by the authors was the study on the maximum capacity of floating PV systems allowed by the

island power system if no interconnection with Sicily was finalized. Cazzaniga et al. have also

studied the integration of compressed air energy storage into floating PV systems and

wastewater treatment plants to explore the synergies and develop novel system concepts [36,
21]. In recent years, there have been several studies on hybrid energy systems for both off-

grid and on-grid applications, of which most of the works focused on the optimal design of

the system components [37]. Shi et al. used a preference-inspired coevolutionary algorithm to

solve a multi-objective optimization problem for a hybrid PV-wind-diesel-battery system to

minimize the annual cost of the system, the loss of power supply probability and emissions

[38]. The model predictive control showed better performances as compared to open-loop

control system. Kaabeche et al. adopted the Firefly Algorithm for the optimization of a hybrid

PV-wind-diesel-battery system [39]. The authors found that the proposed algorithm had better

performances as compared to other well-known algorithms such as Accelerated Particle

Swarm Optimization, Generalized Evolutionary Walk and Bat in solving the optimization

problem. Tazvinga et al. focused on developing an energy dispatch model using model

predictive control techniques in a hybrid PV-wind-diesel-battery system [40]. The authors

showed the robustness of the model in dealing with disturbances. Few studies have

investigated the optimal capacity selection of the system components during the optimization

process. Using a multi-objective optimization algorithm, in this work we have analysed the

mutual relationships between objectives and the selected capacities (decisional variables) of

the system components. In particular, we have analysed how different energy management

strategies, both for on- and off-grid hybrid energy systems, and the effects of cooling can

affect the selection of decisional variables, in particular PV and battery capacities.

3 System description

There are three main phases for the cultivation of shrimps, which are hatchery, nursery and

grow-out phase. Typically, the hatchery phase is performed by the hatchery farms. The

shrimp larvae are then sold to the grow-out farms for nursery and grow-out phase. For all

phases, the most energy intensive operations are the aeration and water pumping. Shrimp

farms need a secure and sufficient power supply, which is not subject to lengthy power
failures. Moreover, an on-site emergency generator is of critical importance to ensure that the

key operations can continue functioning during power blackouts. Among these three phases,

the grow-out phase is the most energy intensive. For the Pacific white shrimp, the nursery and

grow-out periods last around 90 days. The shrimp larvae are first reared in the indoor nursery

tank with a bottom aeration system. This phase requires around 20-30 days before the juvenile

shrimps are transferred to the full-size grow-out pond. In the grow-out pond, both surface and

bottom aeration are required 24 hours a day to maintain dissolved oxygen levels sufficiently

high. The wastewater from the shrimp grow-out pond can be fed to the nearby fishpond for

the preliminary treatment. Detailed descriptions of the shrimp farm used in this work and the

corresponding renewable configurations that can be undertaken to increase the energy

sustainability of the process are presented in Figure 3. The shrimp farm is located in Nakhon

Si Thammarat (8.4 N, 100.0 E). To avoid negative effects on the shrimp farm cultivation,

only the water treatment pond has been considered as a potential site for the floating PV

installation. The pond area is 3,600 m2.

Figure 3: Shrimp farm layout, main loads and suggested renewable energies integration.
The required electrical load was calculated based on a system with two grow-out ponds

operating partly in parallel (one nursery pond, and one water treatment pond). Most of the

load is from the motors used to drive the propellers for surface aeration and the pumping

system for filling the water in the ponds. The overall system consumes more electricity during

the night-time due to the decline of oxygen, and also because the pond is filled at night. The

load for the bottom aeration for both the grow-out and the nursery ponds represents a minimal

part of the load. By including the wastewater treatment load, the energy consumption profile

of the entire system changes significantly. The load for the wastewater treatment was

calculated based on the weight of produced shrimp as reported in Sun [41]. The monthly

electricity consumption profile is depicted in Figure 4.

Figure 4: Monthly electricity consumption profile of the shrimp farm with and without

wastewater treatment.

4 Methodology

4.1 Evaporation
The hourly evaporation from shallow water bodies has been calculated from the reference

evapotranspiration ET0 using the Penman-Monteith equation [42]:

 0.408 − + −
= 1
+ 1 + 0.34

Where, Kw is the open water coefficient, Rn is the hourly net radiation at the water surface

(MJ/m2), G is the hourly soil heat flux density (MJ/m2), Ta is the mean hourly air temperature

(°C), ∆ is the saturation slope of vapor pressure curve at Ta (kPa/ °C), γ is the psychrometric

constant (kPa/°C), es is saturation vapour pressure (kPa), ea is the average hourly actual

vapour pressure (kPa), and u2 is the average hourly wind speed (m/s). Kw has been assumed

equal to 1.05 as suggested by Allen et al. [42]. The hourly evaporation pattern for the selected

site has been calculated using Equation 1 with the hourly climatic parameters taken from the

global meteorological database Meteonorm [43]. The evaporation negatively affects the daily

energy requirements for water pumping in two different ways: reducing the downstream water

head, thus increasing the total pumping head, and increasing the water volume losses. The

resulting daily electricity demand ∆Eel,i (kWh) to maintain stable the water level in the ponds

of the shrimp farm is given by the following equation:

0.0027∆' ,$ ∆ ()$
∆ = 2
!",$
*+

Where, 0.0027 is a conversion factor between Joule and kWh considering the density of water

(1000 kg/m3) and the gravity acceleration (9.8 m/s2), ∆Vw,i is the daily water volume lost due

to evaporation, ∆TDHi is the daily variation of the total dynamic head (m), mainly due to the

variation of the geodetic head ∆Hg (m), and ηps is the efficiency of the pumping system (%). A

schematic diagram of the hydraulic heads and related variations due to evaporation losses is

depicted in Figure 5 considering the natural canal, the water treatment pond and the shrimp

grow-out pond.
 Figure 5: Hydraulic scheme of the natural canal, water treatment pond and shrimp grow-out

pond with the related change of the geodetic heads due to evaporation.

4.2 Simulation models

This section shortly describes the energy system models used in this work, in particular the

solar PV system, the wind turbine, the battery, and the diesel generator used as back-up in off-

grid systems.

4.2.1 PV model

The PV model calculates the solar energy converted into electricity through PV modules. The

global solar radiation has been calculated considering the beam, diffuse and reflected

radiation using the methods described in Duffie and Beckman [44]. For the fixed PV systems,

the angle of incidence has been calculated by using following equation [44]:

,-./ = .01/2 .013 ,-. − + ,-./2 ,-.3 3

Where, θz is the zenith angle (°), β is the tilt angle of the module (°), γs is the solar azimuth

angle (°), and γ is the surface azimuth angle (°). For PV systems tracking the sun while

rotating around a vertical axis with a fixed tilt angle, the angle of incidence θ is given by the

following equation [44]:

,-. / = ,-./2 ,-. 3 + .01/2 .01 3 4

The efficiency of the PV system ηPV (%) has been calculated using the following equation

[44]:

9
*45 = *45,6 81 + −
7
*45 6 7

9 : + ;<=>7 ?@A − 20
+ B1 − *45,6 7 CD EF 5
*45,6 7 : + ;< 800

Where, ηPV,STC is the standard test condition (STC) efficiency of the PV module (%), µ is the

temperature coefficient of the power output (%/°C), Ta is the ambient temperature (°C), TSTC is

the STC temperature (25˚C), vNOCT is the wind speed at the nominal operating cell

temperature (NOCT), v is the actual wind speed (m/s), a and b are coefficients, Gt is the

global incident radiation on the PV array (W/m2), and Ψ is a correction factor to take into

account the cooling effect. The potential cooling effect of solar floating PV systems on the PV

efficiency refers to the previous theoretical and experimental studies summarized in Table 1.

The reference PV module is Yingli YL250P-29b [45].

4.2.2 Wind turbine

The wind turbine power output PWT (kW) has been modeled with the following set of

equations that describes the characteristic wind turbine power curve [46]:

0 N < N$ :1P N > NR

M
K1
HI = 2 STA+ N N$ ≤ N ≤ NV 6
LH NV ≤ N ≤ NR
K I ,V
J

Where, ν is the actual wind speed (m/s), νi, νr and ν0 are the cut-in, rated and cut-out

characteristic speeds of the wind power curve (m/s), ρ is the air density (kg/m3), A is the rotor

area (m2), Cp the power coefficient and PWT,r is the wind turbine rated power (kW). The values

assumed for νi, νr, ν0, A, and Cp are 3.5 m/s, 13.5 m/s, 25 m/s, 572 m2, and 0.46 respectively.

Those values refer to the wind turbine ACSA A27/225 as reported in Carrillo et al. [47].
4.2.3 Diesel generator

In case of off-grid power systems, the back-up power is provided by a diesel generator. The

fuel consumption of the diesel generator has been calculated with the following equation [48]:

XAY$! !" = Z ∗ H\] + 3 ∗ H\],V 7

Where, α and β are the experimental coefficients of the fuel consumption curve, 0.246 and

0.08145 respectively [48]. PDG is the power output from the diesel generator to cover the load

(kW), and PDG,r is the rated power output of the diesel generator, assumed 25% higher than

the maximum power requirement (kW).

4.2.4 Battery

The battery energy capacity (BEC) (kWh) has been calculated by using an energy balance

model described by the following equation:

P ^ A
= − ∗*− 8
P_ $ `aE "`

Where, dt is the simulation time step (1 h), Eloss represents the energy lost due to the self-

discharge rate (kWh), η is the efficiency of the charging-discharging process (%), Ein and Eout

are the energy injected and extracted from the battery during the time step (kWh). The hourly

self-discharge rate and the battery charging-discharging efficiency have been assumed equal

to 0.02% and 85% respectively [49]. The BEC varies between an upper and lower value,

BECMAX and BECMIN. If the battery reach its maximum energy capacity, the power surplus is

exported to the grid or dumped. Otherwise, if the battery is at its lowest energy capacity, the

load is covered by the electrical grid for on-grid systems or by the diesel generator for off-grid

systems.

4.3 Optimization
The optimization has been carried out with the open-source code OptiCE written in Matlab®

environment [50]. The model uses genetic algorithm (GA) as optimization method to solve
single or multi-objective optimization problems. In this study, the code is set to minimize the

LCOE, while maximizing the energy system reliability (REL) (%). The LCOE has been

defined from the life cycle cost (LCC) (US$) calculation and life cycle electricity production

as follows:

bAA
bA@ = 9
∑=id
deV fgh
!", d Y f

Where, Eel,n is the annual electricity production (kWh), r is the degradation rate (%) and d is

the real discount rate (%). Eel,n for off-grid systems is calculated as sum of the net electricity

produced by the renewable energies (defined as difference between electricity produced and

electricity dumped) and diesel generator. This is to avoid overdesign of the renewables

capacities and thus overproduction. The real discount rate d used in the calculation of the LCC

has been calculated with the following equation:

P′ − l
P= 10
1+l

Where, d´ is the discount rate (%) and f is the inflation rate (%). The reliability REL (%) has

been defined as follows:

)mn= op
b= ∗ 100 11
8760

Where, HSL is the number of hours the renewable energies supported by the energy storage

can cover the load during the year (8760 h). The calculation of the land required for the

installation of the PV system, and thus the potential installed peak power, has been performed

considering the land packing factor (PF), defined as the ratio between PV area and land area

required for the installation of the PV system and given by the following equation [51]:
ed
HX = qcos 3 + cos γ } 12
$ u
vwx yRez{

The tilt angle β has been assumed equal to 8° (corresponding to the latitude of the shrimp

farm). Equation 12 has been applied at noon of the winter solstice [51]. The area calculated by
using Equation 12 has been increased of 2 m2/kWp for the secondary floating platform for

maintenance. This corresponds to a required installation area of 10 m2/kWp. Santafé et al.

reported similar values for Spain [24]. The optimization problem is based on six decisional

variables: tilt angle (°), azimuth angle (°), PV capacity (kWp), wind tower height (m), wind

power capacity (kWr), and battery energy capacity (kWh). The lower and upper bounds of the

decisional variables are summarized in Table 2. The upper bound for the PV power capacity

has been set to 200 kWp based on the water surface available for the floating PV system in the

water treatment pond. Accordingly, the upper bound of the wind power capacity has been set

equal to 200 kWr.

The main input economic data for the optimization are summarized in Table 3, while the

specific breakdown costs for the different PV systems configurations are given in Table 4.

Table 2: Decisional variables upper and lower bounds.

Decisional variables Lower bound Upper bound

Tilt angle (ᴼ) 0 30
Azimuth angle (ᴼ) -10 10
PV capacity (kWp) 0 200
Wind tower height (m) 0 80
Wind power capacity (kWr) 0 200
Battery capacity (kWh) 0 4000
 Table 3: Economic input data for the optimization.

Economic parameter Value Reference/Comment

Specific cost of ground mounted PV system (US$/kW) 2,000 See Table 4
Specific cost of floating PV system (US$/kW) 2,350 See Table 4
Specific cost of floating-tracking PV system (US$/kW) 2,410 See Table 4
Specific cost of wind turbine (US$/kW) 1,700 [52]
Specific cost of battery (US$/kWh) 500 [53]
[54]
Specific cost of diesel generator (US$/kW) 1000
[55]
Specific cost of diesel (US$/l) 0.85
Specific cost of electricity bought from national grid (US$/kWh) 0.1 [56]
Specific cost of electricity surplus sold to national grid with FIT (US$/kWh) 0.17 [56]
Based on the system´s component with
Project lifetime (years) 25
longest lifetime
PV system lifetime (years) 25 [57]
Wind turbine lifetime (years) 25 [57]
[53]
Battery lifetime (years) 10
Inverter lifetime (years) 10 [58]
Refer to a prime-power, liquid-cooled
Diesel generator lifetime (years) 5
diesel. Adapted from HOMER [59]
Tax rate (%) 20 [60]
Discount rate (%) 2.1 [60] 10 years average
Inflation rate (%) 1.9 [60] 10 years average
Maintenance rate of PV system (%) 2 [54]
Maintenance rate of wind turbine (%) 2 [54]
Maintenance rate of battery (%) 2 [53]
Maintenance rate of diesel generator (%) 2 [54]
 Table 4: Specific breakdown costs for the reference PV systems configurations.

Component Ground mounted PV Floating PV Floating-tracking PV

PV module (US$/kWp) 700 [61] 700 [61]

700 [61]
Inverter (US$/kWp) 150 [61] 150 [61] 150 [61]

Other hardware costs (including racking

150 [61] 150 [61] 150 [61]
and wiring) (US$/kWp)

Land (US$/m2) 50 [61, 62] - -

Floating structure, mooring and

- 85 [24] 85 [24]
anchoring (US$/m2)

Tracking system (US$/kWp) - - 12,000 (see Appendix A)

Installation labour (US$/kWp)

300 [61] 300 [61] 300 [61]
Profit (US$/kWp)
150 [61] 150 [61] 150 [61]
Others soft costs (including contracting,
50 [61] 50 [61] 50 [61]
permitting and financing) (US$/kWp)

1,500 (without land)

Investment cost (US$/kWp) 2,350 2,410
2000 (with land at 50 US$/m2)

As regards the PV module costs, the 2015 National Survey Report of PV Power Applications

in Thailand reported an average price of about 0.7 US$/Wp, with the lowest and highest equal

to 0.6 and 0.8 US$/Wp [61]. In this study, we used 0.7 US$/Wp as a reference specific cost.

To avoid site-specific results, sensitivity analyses have been carried out considering the

variation of the cost for the land and the floating structure, especially in section 5.2.3 when

comparing ground mounted and floating PV systems. In the period 2012-2015, the specific

cost of the land in Nakhon Si Thammarat province varied between 15 US$/m2 up to 625

US$/m2 [62]. In the period 2016-2018, according to the Treasury Department's, the average

price of the land in Nakhon Si Thammarat province has varied between 2.3 to 77.5 US$/m2
[63]. The high variation of the cost of land for shrimp farming is due to several factors, such

as the distance from the sea, main road, water canal and electrical grid, and the possibility to

use the land also for other purposes. The high cost of land is also connected to the high

income of shrimp farming and related speculation [26]. A reference cost of 50 US$/m2 was

used in this study. The sensitivity analyses have been performed by varying the land cost from

20 to 200 US$/m2. A further scenario concerning the effect of the land rental instead of land

ownership on the economy of PV systems has been considered. The annual land rental has

been estimated as 5% of the land value. The cost variation for the floating structure has been

assumed equal to ± 50% of the cost reported by Santafé et al. [24]. As concerns the floating-

tracking PV system structure, the solution described in Cazzaniga et al. [29] and depicted in

Figure 2 has been selected in this study. This solution is cost-effective and easy to implement

in an extremely confined water body, as the wastewater treatment pond of the shrimp farm.

The calculations of the floating-tracking platform are described more in details in the

Appendix.

4.4 Scenarios definition

A summary of the scenarios (S1-4) with the related hybrid energy system components taken

into consideration in the simulations and optimizations is given in Table 5. Scenario S1

concerns the integration of an off-grid wind-diesel-battery system (WT + BATT + DG) in the

representative shrimp farm. Scenario S2 considers the integration of an off-grid PV-diesel-

battery system (PV + BATT + DG) to cover the electricity requirements of the shrimp farm.

A sub-scenario for S2 has been also considered and it refers to a grid-connected hybrid energy

system. For this sub-scenario, the optimization objective of maximizing the reliability REL

(%) is changed with the objective of maximizing the renewable penetration (RP) (%) by

assuming that the electrical grid has a reliability of 100%. In scenario S3, the ground mounted

PV system of scenario S2 is replaced with a fixed floating PV system. The last scenario
analyses the integration of one-axis floating-tracking PV systems into the off-grid hybrid

energy system. While analysing those scenarios, several sensitivity analyses have been

conducted. These sensitivity analyses have mainly concerned the impact of the cooling effect

and components´ investment costs on the system optimization.

Table 5: Summary of the investigated scenarios.

Description Scenarios

S1 S2 S3 S4

Ground mounted PV - x - -

Floating PV - - x -

Floating-tracking PV - - - x

Wind turbine x - - -

Battery x x x x

5 Results and discussion

5.1 Evaporation losses and water-energy nexus aspects connected to

floating PV systems

The hourly trend of evaporation losses from the water body at the selected location during the

whole year is depicted in Figure 6. The total annual evaporation losses account for about

1,100 mm (i.e. 11,000 m3 of water losses each hectare of water body). Given the geometrical

dimension of the investigated water treatment pond for shrimp cultivation and assuming an

overall pumping efficiency of 60%, the annual water losses account for about 3800 m3 that

corresponds to about 100 kWh of electricity required to maintain the water level constant

during the year. According to Sahu et al., floating PV systems can prevent water storage

capacity losses up to 50% on human-made water storage facilities [11]. Thus, the floating PV

system and the floating-tracking PV system (assumed maximum 200 kWp to allow the normal

operation of the) can potentially avoid about 1050 m3 of water losses corresponding to about

25 kWh of electricity for water pumping. Although the avoided water losses are significant,
the electricity savings for water pumping and the related CO2 emissions are minimal mainly

due to the low pumping head. Although in the selected shrimp farm the head difference

between the natural canal and the water treatment pond has been assumed constant, in real

conditions the geodetic head increases due to the evaporation losses. This can affect the

energy consumption for water pumping in two different ways: due to the linear increase of the

total dynamic head (linear effect), and the different pump efficiency at different hydraulic

head operating conditions (non-linear effect). Although the water-energy nexus results for the

shrimp farm show that the energy savings due to the reduced evaporation are minimal, from a

regional or sector perspective they can be significant. Indeed, shrimp aquaculture involves

more than 20,000 shrimp farms across the country [64]. The use of water for the combined

production of food and electricity is a key aspect for the water-food-energy nexus that allows

reducing the environmental impact of shrimp farming. As seen in Figure 4, the annual

electricity consumption of the shrimp farm is around 230 MWh of which 45 MWh are due to

the wastewater treatment. The annual electricity production of the assumed 200 kWp PV

system can actually offset the annual electricity demand of the shrimp farm. This allows

producing a near carbon-free product with high economic value in the market chain due to

environmental labelling. It is worth noting that the surplus of electricity production can

generate further revenues in on-grid systems due to the possibility of electricity trading with

the national grid. In both on- and off-grid systems, the surplus of power can be dumped

through the aeration and related wastewater treatment processes to further improve the quality

of the effluent.
 Figure 6: Hourly evaporation from the water treatment pond (left) and monthly specific PV

electricity production (right).

5.2 Optimization
5.2.1 Hybrid energy system: wind turbine and ground mounted PV system (S1 and S2)

The relationship between the LCOE and the REL for two off-grid hybrid energy systems (PV

system, battery and diesel generator (PV + BATT + DG) and wind turbine, battery and diesel

generator (WT + BATT + DG)) is depicted in Figure 7. The relationship between LCOE and

REL represents the Pareto front of the optimization problem. By increasing the REL for the

PV + BATT + DG off-grid system, the LCOE initially decreases until 40% REL. Afterwards,

the LCOE starts to increase until 90%. To reach higher REL than 90%, the LCOE dramatically

increases. The optimization of the hybrid WT + BATT + DG system shows a rapid increase

of the LCOE at very low REL that never exceeds the 15%. This is due to the low wind speeds

(wind speeds at 10 m height) at selected location, as superimposed in Figure 7, and thus low

wind power output. The Pareto fronts shown in Figure 7 can be explained more in details by

analysing the trend of the PV and battery capacities selection against the REL during the

optimization process for the PV + BATT + DG off-grid system, as shown in Figure 8 (right).

Similarly, the trend of wind turbine and battery capacities for the WT + BATT + DG off-grid

system is shown in Figure 8 (left).

 Figure 7: Pareto front for the off-grid hybrid PV and WT energy systems.

As it can be seen, for the PV + BATT + DG off-grid system the REL initially increases due to

the increase in capacity of the PV system until 40% REL. Afterwards, higher REL are

achieved by increasing the battery capacity. As concerns the WT + BATT + DG off-grid

system, a REL of about 8% is achieved by selecting immediately the wind turbine capacity to

values closer to the upper boundary of 200 kWr. Further minor increase of the REL are

achieved by increasing the battery capacity up to the maximum values set in the optimization

problem. Due to the low/moderate winds (average annual wind speed at 10 m height lower

than 1 m/s), by increasing the wind turbine capacity or increasing the battery capacity to store

the energy surplus has an insignificant effect on the REL. Thus, most of the energy

requirements are fulfilled by the diesel generator.

From the spatial maps of the annual average solar irradiation and wind speed, shown in Figure

9, it can be clearly deducted that the results concerning the comparison between PV + BATT

+ DG and WT + BATT + DG off-grid systems have a national validity. Indeed, the average

wind speed map shows very low/moderate wind conditions in almost all the country. Similar

low/moderate wind speed conditions were reported in a previous study for the World Bank
focused on the Southeast Asia [65] and, more recently, in the critical work carried out by

Chingulpitak and Wongwises on the wind power sector in Thailand [66].

Figure 8: Relationship between PV (left), wind turbine (right) and battery capacities with the

reliability.

Figure 9: Spatial distribution of the average annual global solar irradiation and wind

speed in Thailand [67].

5.2.2 Off-grid and on-grid ground mounted PV system

The Pareto fronts for the PV + BATT + DG off-grid system, where the back-up energy is

provided by the DG, and for the on-grid system, where the back-up energy is provided by the

grid, are shown in Figure 10. The LCOE at parity of REL and RP is significantly lower for the

on-grid system, especially for REL and RP lower than 90-95%. The Pareto front point marked

out by the lowest REL for the off-grid system corresponds the LCOE of the electricity

produced through the diesel generator. Similarly, the Pareto front point marked out by the

lowest RP for the on-grid system corresponds to the electricity price bought from the grid. To

a diesel price of 0.85 US$/l corresponds a LCOE of 0.32 US$/kWh, while the LCOE of the

electricity taken from the grid is 0.1 US$/kWh. From Figure 10 it can also be seen that the

increase of PV electricity penetration leads to a decrease of the LCOE to RP of about 45%.

That means that the LCOE of the electricity produced through the PV system is lower than the

electricity price bought from the grid. By comparing Figures 8 (left) and 10 (right), it is

possible to compare the optimization and selection process. To avoid overproduction in off-

grid system, as soon as the PV peak capacity exceeds the maximum power consumption the

battery capacity starts to increase to avoid dumping power surplus (Figures 8 (left)). On the

other hand, in on-grid systems, the surplus of power production can be injected into the grid

producing revenues. In this case, the battery capacity starts to increase only when the PV

capacity reaches its upper bound of 200 kWp (Figures 10 (right)).

 Figure 10: Pareto fronts for the off-grid and on-grid hybrid energy system (left) and

relationship between PV and battery capacities with the renewable penetration for the on-grid

energy system (right).

5.2.3 Hybrid energy system: ground mounted PV system and floating PV system (S2 and S3)

The Pareto fronts of the ground mounted PV system and floating PV system are depicted in

Figure 11. A sensitivity analysis considering different potential cooling effects is also taken

into account. In particular, 2% and 11% increased efficiencies due to cooling effects (from

Table 1) are considered. The Pareto fronts for all the investigated cases overlap except for the

case FLOATING PV (COOLING 11%) + BATT + DG that at REL higher than 45% clearly

shows a better competitiveness in terms of LCOE and REL compared to the other cases.

Indeed, the higher efficiency of the floating PV system and thus the higher PV electricity

production reduces the requirements for the battery capacity leading to significant cost

reductions and thus lower LCOE. This can be clearly seen from Figure 12 where the PV and

battery capacities for the two cases are compared (the reference ground mounted PV + BATT

+ DG and the FLOATING PV (COOLING 11%) + BATT + DG). The higher electricity

production of the floating PV system due to the cooling effect contributes to lower the PV

capacity of the floating PV system compared to the ground PV system at parity of REL. The

graph concerning the relationship between optimal battery capacity and REL shows that the
energy storage is selected to achieve REL higher than 40%. It is interesting to note that

achieve REL higher than 90%, the storage capacity has to increase significantly from 0.7

MWh up to 2-4 MWh.

As discussed in the methodology, section 4.3, the variation of the land cost in the study area is

marked out by a wide range depending on several factors. Thus, a comprehensive sensitivity

analysis has been carried out assuming different land costs, land ownerships schemes and

floating PV platform costs. The summary of the sensitivity analysis is presented in Figure 13

(left). The Pareto fronts market out by the lowest and highest LCOE at parity of REL are those

related to ground mounted PV systems with land leased or land at 2 US$/m2 and with land

bought at 200 US$/m2 respectively. The third most profitable solution to support the shrimp

farm operation is the floating PV system with a floating platform marked out by a specific

cost 50% less than the reference case. A more detailed view of the Pareto fronts between 30%

and 60% REL is given in Figure 13 (right).

Figure 11: Pareto front for the ground mounted and floating PV hybrid energy system.
Figure 12: Relationship between PV (left) and battery (right) capacities with the reliability for

ground mounted PV system and floating PV system with 11% of increased efficiency due to

cooling.

Figure 13: Pareto fronts generated form the sensitivity analysis of land prices, land ownership,

and floating PV platform (left) and detailed look for the reliabilities comprised between 30%

and 70% (right).

5.2.4 Hybrid energy system: floating PV system and floating-tracking PV (S3 and S4)

The Pareto fronts for the floating PV and floating-tracking PV systems integrated into an off-

grid hybrid energy system are depicted in Figure 14. The results do not consider any cooling
effect. The floating tracking (1 axis tracking) PV hybrid system allows producing an higher

amount of electricity compared to the reference floating PV system, about 5% more on annual

basis. Although the higher costs of the floating-tracking PV system compared to the reference

floating PV system, both solutions show similar Pareto front for REL up to 45%. Similar to

Figure 11, at very high REL the Pareto front of the floating tracking PV system shows higher

competitiveness compared to the reference floating PV system due to the high specific cost of

the energy storage system. At high REL, the increased electricity production of the PV system

becomes more important than the specific investments costs of the PV system since it allows

reducing the energy storage capacity. As highlighted by Durković and Đurišić [68], there is

limited amount of data and information concerning the economic aspects of floating-tracking

PV systems due to the few commercial installations worldwide. The cost estimation and

related optimization results provided in this study should thus serve as starting point for more

comprehensive evaluations of floating-tracking PV systems supported by data from

commercial plants.

Figure 14: Pareto front for the floating PV and floating-tracking PV hybrid energy system.
6 Conclusions

This works has evaluated different scenarios for the decarbonization of a representative

shrimp farm in Thailand. Technical, environmental and economic aspects have been

considered in this study. The water-energy nexus aspects related to the implementation of

floating PV systems in shrimp farms have also been discussed. A comprehensive dynamic

simulation and optimization model has been developed in Matlab® environment to analyze

the relationship between levelized cost of electricity and renewables reliability and

penetration for off- and on-grid energy systems respectively. The key results of this study are

the following:

• from a water point of view, floating PV system represents an important technological mean

to reduce evaporation from the ponds of the shrimp farm. Nevertheless, from an energy

point of view, the energy losses due to evaporation are minimum compared to the potential

energy conversion of the installed floating PV system. Significant energy consumption

reduction can be achieved considering larger areas or from a sector perspective. From a

water-food-energy nexus perspective, floating PV systems can be combined in the shrimp

farm sector to reduce the carbon footprint of shrimp and at the same time produce

electricity for supporting the wastewater treatment consumption;

• for the selected location, PV systems represent the best solution among the investigated

ones (PV system and wind turbine) to be integrated in hybrid off- and on-grid energy

systems. The optimization model shows that the weak wind resources lead to a maximum

renewable reliability of 15%, achieved mostly by implementing large scale batteries. From

a spatial analysis perspective, the results achieved for the selected locations have a national

validity due to the low/moderate wind conditions of Thailand;

• floating PV systems represent an interesting solution to increase the profitability of PV

installations, especially in locations marked out by a high cost of land. Low land prices or
 land leasing scheme represent the most profitable solutions for PV systems making them

more competitive than floating PV systems. The positive effects of cooling on the

efficiency of the PV modules show that higher renewable reliabilities can be achieved at a

lower levelized cost of electricity compared to ground based PV systems;

• although floating-tracking PV systems have higher specific investment costs, the higher

electricity production compared to fixed floating PV system make them competitive from a

levelized cost of electricity point of view, especially for reliabilities higher than 45%.

The investigations carried out in this work led to the following recommendations. The

integration of renewables in the shrimp farm sector should be mainly focused on PV systems,

unless specific measurements of the wind speed shows better performances for the installation

of wind turbines. The possibility of having access to the grid represents the most cost-

effective solution compared to off-grid systems due to the high price of diesel as back-up

power. The optimal choice between different PV solutions is mainly driven by the costs of the

land and floating platform. Thus, accurate estimation of those lasts two item costs should be

carried out before the investment. From a techno-economic point of view, floating and

floating-tracking PV systems can be competitive to ground based PV systems even at low

land prices but a detailed analysis of the cooling effects should be carried out. From an

economic viewpoint, the best solution is to design hybrid energy system for renewables

reliability/penetration of about 40% that guarantees the lowest levelized cost of electricity

both for off- and on-grid shrimp farms.

Acknowledgements

This work has received funding from the Swedish Knowledge Foundation Future Energy

Profile at Mälardalen University and the National High Technology Research and

Development Program (863 program) of China (No. 2015AA050403). This work has been
also supported by the Applied Energy Innovation Institute (AEii). The author Yuting Tan

acknowledges the financial support from China Scholarship Council (CSC).

Appendix A. Cost estimation for the floating-tracking PV system

The cost of the floating-tracking PV installation has been calculated assuming a tracking

platform without a confining structure as described in Cazzaniga et al. [29] and presented in

Figure 2. This floating-tracking PV configuration has been already implemented and tested in

Italy as reported in Cazzaniga et al. [29] and represent an ideal solution for the selected

shrimp farm due to its cost-effectiveness and the confined environment. The floating structure

can be realized through high-density polyethylene floating cubes with enhanced interlocking

systems. The structure can be afterwards reinforced by the supporting structure of the PV

modules. The main components of the installation are the following: the floating platform, the

concrete anchor, the solar tracker controller, and the propulsion system. The cost for the

floating PV platform, concrete anchor and the mooring system have been assumed equal to 85

US$/m2 as for traditional floating PV systems [24]. The main difference with traditional

floating PV systems is that the anchor system is installed at the bottom of the pond rather than

on the edge. The cost for the solar tracker controller that actuate at least four propulsion

systems has been taken from an e-commerce retailer and equal to 2,000 US$ [69]. The

propulsion system costs have been estimated from the propulsion system power Pp (W) given

by the following equation [70]:

H!~~ '
H+ = = T. 1
*+ *+

Where, Peff is the effective power (W), ηp is the efficiency of the propulsion system (%) (it

takes into account the efficiency of the motor, gearbox, shaft and propeller), RT is the total

resistance (N), and V is the speed (m/s). The total resistance typically depends on several

contributions. By assuming that the tracking system works at low speeds, the main
contributions to the total resistance is given by the viscous resistance RV and wind speed

resistance Rv. The viscous resistance RV can be calculated with the following equation [70,

71]:

1
= A5 S •' T. 2
5
2

Where, Cv is the coefficient of viscous resistance, ρw is the water density (kg/m3), S is the

wetted surface area (m2), and V is the speed (m/s). The tracking system speed has been

assumed equal to 1.0 m/s for the propulsion system capacity design. The coefficient of

viscous resistance is given by the following traditional equation of hydrodynamics [71]:

0.075
A5 = T. 3
[ logdR − 2]

Where, Rn is the Reynolds number. The wind resistance Rv is given by the drag equation:

1
= A ST< T. 4
„
2 Y

Where, Cd is the drag coefficient, ρa is the air density (kg/m3), A is the area it by the wind

(m2), and v is the wind speed (m/s). The drag coefficient has been assumed equal to 1.28 that

corresponds to the drag coefficient of a flat plate perpendicular to flow [72]. This represents

the most conservative scenario assuming that the tilt of the PV modules is equal to 90°. The

interference area with the wind has been assumed equal to the area of the PV modules row

along the circle radius. A wind speed of 7 m/s has been assumed as design wind speed based

on the strongest wind measured at the selected location (see Figure 7). The cost of the

propulsion systems for a 200 kWp floating-tracking PV system has been estimated form the

cost of an electric maritime propulsion system with energy storage depicted in Figure 15 [73].

A summary of the design parameters and costs for the propulsion system of the floating-

tracking PV system is given in Table 6. The corresponding specific cost of the tracking

system is 60 US$/kWp.
 Figure A.1: Maritime propulsion system costs [73].

Table A.1: Summary of the design parameters and cost for the propulsion system of the
floating-tracking PV system.

Parameter Value Reference/Comment

Power peak capacity (kWp) 200

Speed V (m/s) 1.0 Conservative design parameter

Drag coefficient Cd 1.28 Drag coefficient of a flat plate perpendicular to flow [72]

Wind speed v (m/s) 7 See Figure 7

Propulsion efficiency ηP (%) 65 [74, 75]

Propulsion system power Pp (kW) 4.84 kW Calculated with Equation 13

Cost of the propulsion system (kUS$) 10 Assuming to install two propulsion systems for a total

capacity of 10 kW for balancing the rotational

movement of the floating platform. Costs estimated from

 Figure A.1

References

[1] Chimres N, Wongwises S. Critical review of the current status of solar energy in Thailand.

Renew Sustain Energy Rev 2016;58:198–207. doi:10.1016/j.rser.2015.11.005.

[2] Kumar S. Assessment of renewables for energy security and carbon mitigation in

Southeast Asia: The case of Indonesia and Thailand. Appl Energy 2016;163:63–70.

doi:10.1016/j.apenergy.2015.11.019.

[3] Huenteler J, Niebuhr C, Schmidt TS. The effect of local and global learning on the cost of

renewable energy in developing countries. J Clean Prod 2016;128:6–21.

[4] Department of Alternative Energy Development and Efficiency. Energy in Thailand, Facts

and Figures. 2014.

[5] Department of Alternative Energy Development and Efficiency. Energy in Thailand: facts

and figures. 2017.

[6] Energy Policy and Planning Office. Energy Statistics. 2016.

[7] EGAT. Electrical Generating Authority of Thailand (EGAT) 2013.

https://www.egat.co.th/en/index.php?option=com_content&view=featured&Itemid=101.

[8] FAO Yearbook. Fishery and Aquaculture Statistics-2015. 2017

[9] FAO Fisheries and Aquaculture Department. Available at:

http://www.fao.org/fishery/topic/16140/en

[10] Nookuea W, Campana PE, Yan J. Evaluation of Solar PV and Wind Alternatives for Self

Renewable Energy Supply: Case Study of Shrimp Cultivation. Energy Procedia 2016;88:462–

9. doi:10.1016/j.egypro.2016.06.026.
[11] Sahu A, Yadav N, Sudhakar K. Floating PV power plant: A review. Renew Sustain

Energy Rev 2016;66:815–24. doi:10.1016/j.rser.2016.08.051.

[12] NREL. Floating Solar PVs Gaining Ground. Available at:

https://www.nrel.gov/technical-assistance/blog/posts/floating-solar-PVs-gaining-ground.html.

Accessed: 27th January 2018

[13] Solar Asset Management—North America. Floating solar plants: niche rising to the

surface? Available at: https://solarassetmanagement.us/news-source/floating-plants-article.

Accessed: 27th January 2018

[14] Solarplaza. Available at: https://www.solarplaza.com/channels/top-10s/11761/top-70-

floating-solar-pv-plants/. Accessed: 5th November 2018

[15] Choi YK. A study on power generation analysis of floating PV system considering

environmental impact. International Journal of Software Engineering and Its Applications,

8(1), (2014) 75-84

[16] Majid Z, Ruslan MH, Sopian K, Othman MY, Azmi M. (2014). Study on performance of

80 watt floating photovoltaic panel. Journal of Mechanical Engineering and Sciences, 7(1),

1150-1156

[17] Bahaidarah H, Subhan A, Gandhidasan P, Rehman S. (2013). Performance evaluation of

a PV (photovoltaic) module by back surface water cooling for hot climatic conditions.

Energy, 59, 445-453.

[18] Liu L, Wang Q, Lin H, Li H, Sun Q, Ronald Wennersten. Power Generation Efficiency

and Prospects of Floating Photovoltaic Systems. Energy Procedia 105 ( 2017 ) 1136 – 1142

[19] Yadav N, Gupta M, Sudhakar. Energy Assessment of Floating Photovoltaic System.

2016 International Conference on Electrical Power and Energy Systems (ICEPES)

Maulana Azad National Institute of Technology, Bhopal, India. Dec 14-16, 2016
[20] Azmi MSM, Othman MYH, Ruslan MHH, Sopian K, Majid ZAA. Study on Electrical

Power Output of Floating Photovoltaic and Conventional Photovoltaic. AIP Conference

Proceedings 1571, 95 (2013); doi: 10.1063/1.4858636

[21] Rosa-Clot M, Tina GM, Nizetic S. Floating photovoltaic plants and wastewater basins:

an Australian project. Energy Procedia 134 (2017) 664–674

[22] Helfer F, Lemckert C, Zhang H. Impacts of climate change on temperature and

evaporation from a large reservoir in Australia. Journal of hydrology, 475, (2012) 365-378

[23] Bengoechea JM, Pérez Cobos J, Pérez Parra J, López Segura JG. Evaluación de las

pérdidas de agua de riego en el Campo de Dalías. In: Symposium sobre el agua en Andalucía;

1991. Córdoba, España

[24] Santafé MR, Soler JBT, Romero FJS, Gisbert PSF, Gozálvez JJF, Gisbert CMF.

Theoretical and experimental analysis of a floating photovoltaic cover for water irrigation

reservoirs. Energy, 67, (2014) 246-255.

[25] Pringle AM, Handler R, Pearce J. (2017). Aquavoltaics: Synergies for dual use of water

area for solar photovoltaic electricity generation and aquaculture. Renewable and Sustainable

Energy Reviews, 80, 572-584. doi:10.1016/j.rser.2017.05.191

[26] Engle CR, Mcnevin A, Racine P, Boyd CE, Paungkaew D, Viriyatum R, Minh HN.

(2017). Economics of Sustainable Intensification of Aquaculture: Evidence from Shrimp

Farms in Vietnam and Thailand. Journal of the World Aquaculture Society, 48(2), 227-239.

doi:10.1111/jwas.12423

[27] Choi YK, Lee, NH, Lee AK, Kim KJ. A study on major design elements of tracking-type

floating photovoltaic systems. International Journal of Smart Grid and Clean Energy, 3(1),

(2014) 70-74.

[28] Scienza Industria Tecnologia. Available at:

http://www.scintec.it/ricerca/energia/ftcE.html. Accessed 19th August 2017

[29] Cazzaniga R, Cicu M, Rosa-Clot M, Rosa-Clot P, Tina GM, Ventura C. Floating PV

plants: Performance analysis and design solutions. Renewable and Sustainable Energy

Reviews 81 (2018) 1730–1741

[30] Kim S, Yoon S, Choi W. (2017). Design and Construction of 1 MW Class Floating PV

Generation Structural System Using FRP Members. Energies, 10(8), 1142.

doi:10.3390/en10081142

[31] Lee Y, Joo H, Yoon S. (2014). Design and installation of floating type photovoltaic

energy generation system using FRP members. Solar Energy, 108, 13-27.

doi:10.1016/j.solener.2014.06.033

[32] Choi Y, Kim I, Hong S, Lee H. (2014). A Study on Development of Azimuth Angle

Tracking Algorithm for Tracking-type Floating Photovoltaic System.

doi:10.14257/astl.2014.51.45

[33] Choi Y, Lee N, Lee A, Kim K. (2014). A study on major design elements of tracking-

type floating photovoltaic systems. International Journal of Smart Grid and Clean Energy,

3(1), 70-74. doi:10.12720/sgce.3.1.70-74

[34] Silvério NM, Barros RM, Filho GL, Redón-Santafé M, Santos IF, Valério VE. (2018).

Use of floating PV plants for coordinated operation with hydropower plants: Case study of the

hydroelectric plants of the São Francisco River basin. Energy Conversion and Management,

171, 339-349. doi:10.1016/j.enconman.2018.05.095

[35] Trapani K, Millar DL. (2013). Proposing offshore photovoltaic (PV) technology to the

energy mix of the Maltese islands. Energy Conversion and Management, 67, 18-26.

doi:10.1016/j.enconman.2012.10.022

[36] Cazzaniga R, Cicu M, Rosa-Clot M, Rosa-Clot P, Tina GM, Ventura C. Compressed air

energy storage integrated with floating PV plant. Journal of Energy Storage 13 (2017) 48–57
[37] Luna-Rubio R, Trejo-Perea M, Vargas-Vázquez D, Ríos-Moreno G. (2012). Optimal

sizing of renewable hybrids energy systems: A review of methodologies. Solar Energy, 86(4),

1077-1088. doi:10.1016/j.solener.2011.10.016

[38] Shi Z, Wang R, Zhang T. (2015). Multi-objective optimal design of hybrid renewable

energy systems using preference-inspired coevolutionary approach. Solar Energy, 118, 96-

106. doi:10.1016/j.solener.2015.03.052

[39] Kaabeche A, Diaf S, Ibtiouen R. (2017). Firefly-inspired algorithm for optimal sizing of

renewable hybrid system considering reliability criteria. Solar Energy, 155, 727-738.

doi:10.1016/j.solener.2017.06.070

[40] Tazvinga H, Zhu B, Xia X. (2014). Energy dispatch strategy for a photovoltaic–wind–

diesel–battery hybrid power system. Solar Energy, 108, 412-420.

doi:10.1016/j.solener.2014.07.02

[41] Sun W. Life cycle assessment of indoor recirculating shrimp aquaculture system. 2009.

Available at: https://deepblue.lib.umich.edu/handle/2027.42/63582. Accessed 19th August

2017

[42] Allen, RG, Pereira LS, Raes D, Smith M. Crop evapotranspiration. Guidelines for

computing crop water requirements. FAO (1998).

[43] Meteonorm. Available at: www.meteonorm.com. Accessed 19th August 2017

[44] Duffie JA, Beckman WA. Solar Engineering of Thermal Processes. Chichester: John

Wiley & Sons Ltd (2013).

[45] Yingli Solar. Available at: www.yinglisolar.com/us/. Accessed: 7th June 2018

[46] Lydia M, Kumar SS, Selvakumar AI, Kumar GEP. A comprehensive review on wind

turbine power curve modeling techniques. Renewable and Sustainable Energy Reviews 30

452–460 (2014)
[47] Carrillo, C., Montaño, A. O., Cidrás, J., & Díaz-Dorado, E. (2013). Review of power

curve modelling for wind turbines. Renewable and Sustainable Energy Reviews, 21, 572-581

[48] Ismail MS, Moghavvemi M, Mahlia TMI. Techno-economic analysis of an optimized

photovoltaic and diesel generator hybrid power system for remote houses in a tropical

climate. Energy Conversion and Management 69 163–173 (2013)

[49] Kanase-Patil, A. B., Saini, R. P., & Sharma, M. P. (2011). Sizing of integrated renewable

energy system based on load profiles and reliability index for the state of Uttarakhand in

India. Renewable Energy, 36(11), 2809-2821.

[50] OptiCE. Available at: www.optice.net. Accessed: 19th August 2018

[51] Martín-Chivelet N. Photovoltaic potential and land-use estimation methodology. Energy

94 (2016) 233e242

[52] IRENA. Renewable power generation costs in 2014. 2015. Available at:

http://www.irena.org/documentdownloads/publications/irena_re_power_costs_2014_report.pd

f. Accessed: 3rd March 2017

[53] Lazard, Lazard’s Levelized Cost of Storage Analysis, November 2017. Available at:

https://www.lazard.com/media/450338/lazard-levelized-cost-of-storage-version-30.pdf

[54] Merei G, Berger C, Sauer DU. Optimization of an off-grid hybrid PV–Wind–Diesel

system with different battery technologies using genetic algorithm. Solar Energy 97 (2013)

460–473

[55] Global Petrol Prices. Available at:

http://www.globalpetrolprices.com/Thailand/diesel_prices/. Accessed: 3rd March 2018

[56] IEA. Thailand Electricity Security Assessment 2016. 2017. Available at:

https://www.iea.org/publications/freepublications/publication/Partner_Country_Series_Thaila

nd_Electricity_Security_2016_.pdf. Accessed: 7th Nov 2017

[57] Khiareddine A, Salah CB, Rekioua D, Mimouni MF. Sizing methodology for hybrid PV

/wind/ hydrogen/battery integrated to energy management strategy for pumping system.

Energy 153 (2018) 743e762

[58] Kaabeche A, Ibtiouen R. Techno-economic optimization of hybrid

photovoltaic/wind/diesel/ battery generation in a stand-alone power system. Solar Energy 103

(2014) 171–182

[59] HOMER. Available

at:https://www.homerenergy.com/products/pro/docs/3.11/generator_lifetime.html. Accessed:

20th June 2018

[60] Trading Economics. Available at: http://sv.tradingeconomics.com/thailand/corporate-

tax-rate. Accessed: 7th Nov 2017

[61] Department of Alternative Energy Development and Efficiency (DEDE), Ministry of

Energy. 2015 National Survey Report of PV Power Applications in Thailand. Available at:

www.iea.org.

[62] OK Nation Blog. Available at:

http://oknation.nationtv.tv/blog/meeboo/2012/09/30/entry-2. Accessed: 7th Nov 2017

[63] Treasury Department. Available at:

http://www.treasury.go.th/main.php?filename=price_thing#. Accessed: 7th June 2018

[64] Portley N. Report on the Shrimp Sector: Asian Shrimp Trade and Sustainability.

Sustainable Fisheries Partnership. 2016

[65] TrueWind Solutions, LCC. Wind energy resource atlas of Southeast Asia. 2001

[66] Chingulpitak S, Wongwises S. Critical review of the current status of wind energy in

Thailand. Renewable and Sustainable Energy Reviews 31 (2014) 312–318

[67] Fick SE, Hijmans RJ. Worldclim 2: New 1-km spatial resolution climate surfaces for

global land areas. International Journal of Climatology (2017)

[68] Durković V, Đurišić Ž. (2017). Analysis of the Potential for Use of Floating PV Power

Plant on the Skadar Lake for Electricity Supply of Aluminium Plant in Montenegro. Energies,

10(10), 1505. doi:10.3390/en10101505

[69] Alibaba. Available at: www.alibaba.com. Accessed: 7th June 2018

[70] Molland, A. F., Turnock, S. R., & Hudson, D. A. (2017). Ship resistance and propulsion.

Cambridge university press

[71] Remmlinger U. (2014) How to Determine the Viscous Resistance of the Delft Systematic

Yacht Hull Series.

[72] Shape effects on drag. Available at: www.grc.nasa.gov/WWW/k-

12/airplane/shaped.html. Accessed: 7th June 2018

[73] OceanPlanet Energy. www.bruceschwab.com/. Accessed: 7th June 2018

[74] Symington WP, Belle A, Nguyen HD, Binns JR. (2014). Emerging technologies in

marine electric propulsion. Proceedings of the Institution of Mechanical Engineers, Part M:

Journal of Engineering for the Maritime Environment Vol 230, Issue 1, pp. 187 – 198,

https://doi.org/10.1177/1475090214558470

[75] Oladokun SO (2015) Study of Efficiency and Environmental Performance of Propeller. J

Coast Zone Manag 18: 400. doi: 10.4172/2473-3350.1000400