Energy Conversion and Management 260 (2022) 115598

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman

Floating solar PV to reduce water evaporation in water stressed regions and

powering water pumping: Case study Jordan
Lewis W. Farrar a, AbuBakr S. Bahaj a, *, Patrick James a, Arif Anwar a, b, Nafn Amdar c
a
Energy and Climate Change Division, Sustainable Energy Research Group, (www.energy.soton.ac.uk), Faculty of Engineering and Physical Sciences, University of
Southampton, Southampton SO16 7QC, UK
b
International Water Management Institute (IWMI), 12km Multan Road Chowk, Thokar Niaz Baig, Lahore 5370, Pakistan
c
International Water Management Institute – IWMI MENA, Amman, Jordan

A R T I C L E I N F O A B S T R A C T

Keywords: Water resources are essential for human consumption and food production. The extraction and delivery of water
Floating solar PV resources are highly dependent on energy. Hence water, energy and food security are inextricably linked, and
Water evaporation reduction this nexus constitutes a major global societal challenge. Furthermore, globally, irrigation constitutes around 70%
Irrigation
of our freshwater resources, rising to 90% in developing countries. There are over 300 million drinking water and
Water-energy-food nexus
Water scarcity
irrigation ponds globally where 90% of the world’s standing irrigation water resides. There is a need to conserve
such resources, considering more than two thirds of the world’s population are currently experiencing water
stress. Hence, this work tackles the conservation of such resources addressing two important issues related to
energy and water, thereby addressing elements of the UN Sustainable Development Goals. Its considered
approach is the use of floating solar photovoltaic (FPV) technology implemented on irrigation reservoirs to
conserve water by reducing evaporation losses whilst providing sustainable electricity at enhanced yield that can
be utilised locally. For the study, we selected an arid and water stressed region of Jordan where real-world water
and energy consumption data were available. Various floating PV (FPV) system configurations were modelled for
installation on an irrigation reservoir where currently no FPV exists. A fixed tilt 300 kWp FPV system was found
to be the optimum design in terms of water savings, energy yield, economics, and reductions in CO2 emissions.
Standard floating PV was deemed the preferred option compared to ground-mounted PV and FPV with tracking
and/or active cooling. System payback period for the recommended design was 8.4 years with an annual
greenhouse gas emission reduction of ~ 141TCO2. For the considered site, around 12,700 m3 of water can be
saved annually or 42% savings when compared to the uncovered reservoir. This research has wider applicability
to other arid regions such as Africa, Middle East, and the Indian Subcontinent.

water conservation and availability as well as energy and thus consti­

1. Introduction tuting a major challenge for global society [3]. A recent study indicated
that globally irrigation is the largest consumer of fresh water repre­
Water scarcity has become a major issue during the 21st century, senting around 70% of our water resources. In developing countries 90%
with more than two thirds of the world’s population under water stress of available water resources are used for irrigation as compared with
for at least one month of the year and half a billion people experiencing 60% in developed countries [4]. Hence, in a developing country context,
severe water shortages daily [1]. Furthermore, water resources are not any approaches that can conserve and extend water availability, espe­
only needed for human consumption but also for agriculture and hence cially in hot, arid, and semi-arid climates will be highly appropriate not
food. In many regions around the world extraction and delivery of water only from the point of view of food security but also in terms of eco­
resources is highly dependent on energy. Hence the Food and Agricul­ nomic development.
ture Organisation of the United Nations (FAO), has established the Water ponds has been used for centuries for both drinking water and
“water, energy and food security nexus” inextricable linkage meaning crop irrigation. Globally, there are estimated 277,400,000 small scale
actions in any one particular area often can have effects in one or both of (<10,000 m2) and 24,120,000 large scale (10,000–100,000 m2) irriga­
the other areas [2]. Hence, guaranteeing food supply is interlinked to tion ponds, representing more than 90% of the world’s standing

* Corresponding author.
E-mail address: [email protected] (A.S. Bahaj).

https://doi.org/10.1016/j.enconman.2022.115598
Received 11 January 2022; Received in revised form 5 April 2022; Accepted 6 April 2022
Available online 18 April 2022
0196-8904/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

Nomenclature Subscripts
a Actual
A Footprint of PV installation, m2 b Beam
CFt Cashflow in year t, $USD d Diffuse
D Vapour pressure deficit, kPa Free Free water surface
e Vapour pressure, kPa LF Large footprint Floating Photovoltaic
E Open-water evaporation, mm/day Max Maximum
f Inflation rate Min Minimum
fu Wind function n Net
Gi Carbon dioxide equivalent emissions, kgCO2e/kWh s Saturation
GPV Embodied greenhouse gases emissions, kgCO2e/m2 SF Small Footprint Floating Photovoltaic
h Height above ground, m T Total
i Real discount rate W Water
i’ Nominal discount rate
Kd Diffuse fraction of the global horizontal irradiation Acronyms
Kt Cloudiness index CAPEX Capital Expenditure
LW Long wavelength radiation, MJ/m2/day FAO Food and Agriculture Organisation for the United Nations
n Project lifetime, years FIT Feed-in Tariff
R Solar radiation at the surface FPV Floating Photovoltaics
SW Short wavelength radiation, MJ/m2/day GHG Greenhouse Gas
T Temperature, ◦ C HAT Horizontal Axis Tracking
U Wind speed, m/s HDPE High Density Polyethylene
HOMER Hybrid Optimisation of Multiple Energy Resources
Greek letters LCOE Levelised Cost of Electricity
α Absorptivity of atmosphere NOCT Nominal Operating Cell Temperature
β Energy losses associated with electricity transmission and NPC Net Present Cost
distribution NPV Net Present Value
Γ Psychrometric coefficient, kPa/◦ C OPEX Operational Expenditure
Δ Slope of the saturation vapour pressure curve, kPa/◦ C SPIS Solar-Powered irrigation system
λ Latent heat of vapourisation, MJ/kg VAT Vertical Axis Tracking
σ Stefan-Boltzmann Constant, W/m2K4

economic, environmental, social and technical benefits for utilising

Table 1 floating PV on irrigation ponds for the purposes of water conservation
FPV Advantages and drawbacks compared to ground-mounted PV [7 10 11 12]. and pumping are unknown. To our knowledge practical examples of
FPV advantages FPV challenges
such utilisation has not been considered in the literature. Hence this
body of work is unique as it models floating PV technology for a specific
• Higher conversion efficiencies due to • Higher initial investment,
irrigation reservoir in an arid region for the dual purpose of reducing
the cooling nature of water and in operation & maintenance costs.
many cases the absence of dust. • Uncertainty of long-term environ­ water evaporation and providing low carbon electricity at enhanced
• FPV installations mental impacts. yield that can be utilised locally for groundwater pumping. The work
can reduce surface water • Electrical safety challenges when also compares two categories of fixed tilt floating to ground-mounted PV
evaporation. Particular importance in building and operating PV in water. designs coupled with tracking and active cooling systems. Furthermore,
arid regions. • Metallic structures are more prone to
• FPV requires no land, so does not corrosion, hence FPV has a shorter
to address this critical knowledge gap, we quantified the evaporation
compete with other land-users such as lifetime than ground-mounted PV. savings, emissions, and economics based on real-world data of a reser­
agriculture, mining or tourism. • Lack of separate regulations for voir in an arid and water stressed region of Jordan. The outcomes of this
• FPV limits algae growth thus permitting and licensing FPV projects. case study are important due to the nature of such regions where water
improving water quality.
stress is likely to be exasperated by the changing climate. Hence, the
• Risk of theft and vandalism is
reduced. work provides opportunities to be exploited not only regionally but also
globally where over 300 million water ponds exit. In essence, the
approach presented here can be generalised to provide global assess­
irrigation water [5]. The use of these ponds is expanding considerably. ment under different weather conditions.
An approach to the challenges of the energy–water–food nexus
particularly for water conservation and energy, is the use of solar
photovoltaic (PV) modules (panels) to cover water bodies such as the 1.1. Floating solar PV
ponds mentioned above. This results in multiple benefits for both water
conservation and energy delivery from a particular site. Placing solar PV Solar photovoltaics systems that float on a body of water were first
panels over water ponds using, for example, floating solar systems not developed in the 2000–2010 decade, with the first small scale system
only conserves water by reducing evaporation losses through effects on built in Japan in 2007, and the first commercial system was built in
incident solar radiation and surface wind speed, but enhances the energy California in 2008 [7]. It was not until 2013 that installations larger than
yield (hence economics) of the PV systems through the cooling effect 1 MWp were developed. By the end of 2019, the total installed capacity
[6]. An additional benefit of locating solar PV systems over such ponds of FPV exceeded 2 GWp [8]. By 2030, there will be an estimated 62 GWp
reduces the need for land which can exploited for other benefits. of FPV globally. Today, Asia dominates the FPV market, with 87% of
Despite the potential advantages mentioned above, the overall global capacity is situated [9]. To date, there are no FPV installations in
Jordan.

2
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

As shown in Table 1, floating photovoltaics (FPV) have many ad­ covering the entire 4490 m2 reservoir. A unique feature of their design
vantages and drawbacks compared to ground-mounted PV installations was the flexible couplings between each platform which allowed the
[7 10 11 12]. system to adapt to different water levels. Over the test period, it was
There is a wealth of floating solar designs, thus indicating the nascent estimated that 5,000 m3 of water was saved from reducing evaporation,
stage of the industry. However, researchers tend to classify them in three while producing 425 MWh of electricity annually. Rosa-Clot and Tina
categories [13]: [22] performed simulations of FPV plants on wastewater basins in Bo­
livar, South Australia. Fixed and vertical axis tracking (VAT) systems,
1. Free standing (Type 1): Modules are high above the water surface with and without cooling, were compared based on energy performance
with excellent convective cooling and a minimal footprint on the and water saving potential. It was concluded that a fixed system would
water. However, this type does not reduce surface water evaporation allow greater coverage of the basins thus producing more energy and
significantly. saving more water. Water cooling enhanced the energy yield by 10%.
2. Small footprint on water (Type 2): Similar to Type 1 but modules The authors quote that for every MWp of installed PV, between 15,000
are closer to the water surface and relatively small water surface and 25,000 m3 of water would be saved. Other studies, investigating
coverage. FPV and evaporation reduction include Durkovíc and Durišíc [23],
3. Large footprint on water (Type 3): Water surface almost entirely Melvin [24], and Taboada et al. [25]. Other studies include those by
blocked by floating structure with low convective cooling capability Zhou et al. [26], which addressed the assessment for long term com­
compared with Types 1 and 2. Usually consist of high-density Poly­ plementary operation between floating photovoltaic power and hydro­
ethylene (HDPE) rafts or pontoons. This type is likely to exhibit power generation linked to average annual food production and Qasem
excellent water evaporation reduction capability. Abdelal [27] which considered lab-scale, very small area (4 m2) FPV
which is more concerned with key water quality parameters such as
Due to the similarities between Type 1 and Type 2, these can be nitrate and chlorophyll concentrations. The majority of studies relied on
grouped together into one category. Hence, these will be referred to as simulations of FPV plants, and there is a gap in knowledge for fully-
Small Footprint structures herein. As indicated above, it is apparent that fledged large-scale commercial projects and their role in reducing
a trade-off exists between evaporation reduction due to the high foot­ evaporation. Except for the work by Santafé et al. [20], there are no
print on water and the increase in electricity production from the examples augmenting FPV with agricultural irrigation systems at scale
convective cooling effect. The Small Footprint FPV has good convective in the literature.
cooling effect and a poor evaporation reduction, whilst Large Footprint Given that FPV installations are not as prevalent as ground-mounted
FPV is more than likely to have the opposite in both categories i.e., PV, there is a degree of uncertainty surrounding cost data, both in terms
poor convective cooling effect and a good evaporation reduction [13]. of capital (CAPEX) and operating (OPEX) expenditure. Site location can
Novel FPV technologies, aimed at improving the overall energy also influence costs of FPV significantly with factors such as bathymetry,
capture, include tracking (vertical, horizontal and dual-axis), active water-level variation and wind characteristics all playing a role. Here,
cooling mechanisms in the form of a water veil or sprinklers, concen­ the literature surrounding the economics of FPV systems is discussed. At
trated FPV and submerged FPV (where the modules sit slightly below the present the CAPEX of FPV is greater than that of ground-mounted PV
surface of the water). Some installations incorporate multiple such due to the floating support structure, the associated mooring and
facets. All of these technologies incur higher investment and mainte­ anchoring system and the underwater DC cabling. Reindl et al. [7] states
nance costs compared to the standard fixed-tilt FPV design. that the total CAPEX for FPV projects in 2018 was $800–1200 per kWp,
One of the main attractions of floating PV is the increased efficiency having fallen from $2000–3000 per kWp in 2015–2016. This is broadly
of the PV modules due to the evaporative cooling effect of the water consistent with other literature sources. For instance, Rosa-Clot [22]
body. To quantify the cooling effect the heat loss coefficient or U-value is quotes $1100/kWp for FPV plant in Australia while Goswami et al. [28]
used. H. Liu et al. [13] found that a Small Footprint structure has an suggest a 10 MW FPV plant in India would cost $940/kWp. As with other
average U-value of 46 W/m2K, whereas a Large Footprint type has a U- renewable energy technologies, economies of scale are likely to bring
value of 31 W/m2K. By comparison, ground-mounted or rooftop PV the cost of FPV down significantly over the coming years, especially the
systems have a U-value of 15–29 W/m2K depending on the degree of cost of the floating support structures. In terms of operation and main­
ventilation. Oliveira-Pinto and Stokkermans [14] used Liu’s U-values to tenance costs, Reindl [7] suggests a value of $11/kWp/year, exactly the
simulate the energy production of FPV systems in three locations using same as for ground-mounted PV. Goswami [28], on the other hand,
PVSyst® [15]. It was found that a Large Footprint FPV installation has suggests $19.7/kWp/year. For FPV systems incorporating tracking and/
an increase in production of 0.31–0.46% while a Small Footprint or cooling technology both Giuseppe [29] and Durkovíc [23] propose
installation has an increase in production of 1.81–2.59% compared to a OPEX costs in the region of $20/kWp/year. The Levelised Cost of
ground-mounted PV system. In another study performed by Yadav et al. Electricity (LCOE) for FPV installed in a desert climate was estimated to
[16] where the power of a system based on a single module of capacity be around $54 per MWh [7]. A corresponding ground-mounted PV
250 W FPV installed on a high density HDPE floats (large footprint ty­ system has an LCOE 8–9% lower than that of FPV. This difference will
pology) was compared to a 250 W ground-mounted system. They found likely be narrowed as the FPV technology scales up.
the power of the FPV system was 2.24% higher than the ground- Solar-powered irrigation systems (SPIS) are also a technology
mounted system. Simulations by L. Liu et al. [17] show similar results, becoming ever more common. A number of studies have shown the
where a power gain of 1.58–2.00% was presented. It is noted that these benefits of deploying this technology in Jordan [30 31]. The growth of
increases in performance are less than other literature sources which both FPV and SPIS is likely to be accelerated by the reduction in cost of
suggest very optimistic power gains, ca. 10% compared to ground- photovoltaic systems. However, to our knowledge, there have been no
mounted PV [18 19]. studies examining the potential to combine FPV and SPIS in a Middle
Over the last decade several studies have reported the beneficial East context. This is at the core of this research which aims to investigate
effect FPV can have on reducing evaporation from water bodies. This is and design an FPV-SPIS system on a known site in Jordan where there is
achieved by (a) the floating structure provides shading and hence re­ a requirement for such a combination. The work considers two distinct
duces the incident solar radiation on the water surface and (b) the free classes of FPV installations to be compared and assessed, namely, Large
surface of water is covered and is subject less to the effects of wind. Footprint FPV and Small Footprint FPV. The economics and environ­
Santafé et al. published the earliest articles discussing FPV and evapo­ mental implications of the scheme in a Jordanian context will be
ration [20 21]. Initially a 20 kWp prototype was installed on an irriga­ assessed. The work has wider applicability where remote, enhanced
tion reservoir prior to their consideration of a 300 kWp of FPV system sustainable power and water evaporation reduction are needed.

3
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

Fig. 1. Overhead Image of Farm ‘A’ with irrigation reservoir highlighted [41].

2. Case study approach Water from the well is stored in an irrigation reservoir of area 13,500 m2
and depth 7 m. A 37 kW booster pump from the reservoir to the irri­
One of the worst affected countries is the Kingdom of Jordan. Ac­ gation pipelines is also installed. At present, the two AC pumps are
cording to the World Resource Institute, Jordan is ranked fifth place in powered by grid electricity at a price of $0.12 per kWh.
terms of countries suffering from water scarcity issues [32], with an Fig. 1, below, depicts an overhead image of Farm ‘A’.
annual water resource is 145 m3 per capita [33], far lower than the
United Nation’s 500 m3 threshold that indicates absolute water scarcity. 3. Methodology
This problem is getting worse due to a growing population and the
impact of climate change. Climate change is leading to unpredictable This section describes the methodology undertaken in this research.
rainfall while increasing temperatures are causing surface water evap­ It involves a PV system design using HOMER Pro®, estimations of sur­
oration and a more arid landscape. For Jordan, more than 90% of annual face water evaporation rates, economics, and estimations of carbon di­
precipitation is lost through evaporation [34]. Hence, finding ways to oxide equivalent emissions.
reduce surface water evaporation can become a major step towards
improving the nation’s ground and surface water resource.
Jordan also lacks natural resources but is blessed with a lot of sun­ 3.1. System design
shine. It relies heavily on imported fossil fuels to meet its electricity
demands (ca. 90%) costing 10% of the country’s GDP [35]. This not only HOMER Pro® microgrid software was used to design the PV system
poses an energy security problem but also makes the country vulnerable capable of satisfying the well and booster pumps at Farm ‘A’. HOMER
to fluctuations in fuel prices. Recently there has been a push to install (Hybrid Optimisation Model for Multiple Energy Resources) is the global
solar photovoltaics (PV), with more than 800 MW solar PV capacity standard for designing and optimising microgrids. It works by choosing
currently installed and a goal of 2.2 GW by 2021 is in place [36]. This is the system with the lowest total Net Present Cost (NPC) from a set of
part of a wider goal to reach 10% of the country’s electricity needs from variables and constraints.
renewables by 2020 [37]. Jordan has the potential to become a sus­ The total NPC (or life-cycle cost) of a system is defined as the present
tainable energy hub of the Middle East as it is one of the most value of all costs the system incurs over the project lifetime, minus the
economical and politically stable nations in the region. Jordan’s PV present value of all revenues it makes over the project lifetime [42]. It is
industry has significant room for growth, especially considering its solar calculated using Discounted Cash Flow (DCF) analysis. The costs include
resource which averages at 5.6 kWh/m2/day and 310 sunny days per installation costs, operation and maintenance costs, and replacement
year [35]. In Jordan, a 100% tax exemption for ten years for renewable fees. Revenues, on the other hand, include sales of electricity to the grid
energy investments in certain regions of the country is in place, partic­ and the salvage value once the project is completed. The NPC is a
ularly those where socio-economic developments are required [35]. convenient method to compare different options, with the lowest NPC
Another financial incentive was the 0.17 $/kWh feed-in-tariff (FIT) from scenario being the most economically favourable one. NPC can also be
2012 [38] which was revised downwards to $0.148 per kWh for 2016 referred to as Net Present Value (NPV) which is identical except multi­
and 2017 [39]. It is due to be revised yet again to $0.0705 per kWh plied by negative one. For financial analysts, NPC or NPV is the most
thereafter [40]. common tool when comparing the feasibility of different projects.
The aim of this research is to investigate and design an FPV-SPIS Net Present Cost is presented mathematically in Equation (1) below:
system on a known site in the Jordan where there is a requirement for ∑n
CFt
such a combination. The case study site is ‘Farm A’ [41], a 54 ha farm NPC = t (1)
t=1 (1 + i)
located in the Mafraq governate of Northern Jordan where irrigation is
supported by pumped ground water into a large reservoir. The four Where CFt is the cashflow in year t i.e. the sum of all costs minus the
systems selected for modelling and comparison are given in Section 3.2 sum of all revenues, n is the project lifetime and i is the real discount rate
with definitions explained under Section 1.1. The system design is based (or interest rate), calculated as:
on real operation data from the farm and will have the dual purpose of
i’ − f
saving water evaporation from an irrigation pond and providing elec­ i= (2)
1+f
tricity to two pumps for irrigation.
Established in 2003, the farm cultivates four crops, namely olives, Where i’ is the nominal discount rate and f is the expected inflation
stone fruit, pomegranates, and grapes. Water for irrigation is provided rate [43]. Hence, the time value of money is incorporated into Net
from groundwater using a submerged 83 kW pump in a private well. Present Cost.

4
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

3.2. PV systems analyses

A scenarios analyses approach was undertaken to determine and

understand the PV system requirements for the farm based on electrical
demand of the farm and solar resource. Initially, a business-as-usual
scenario using grid electricity to power both pumps (i.e., no PV) was
modelled to determine the NPC of the required system at Farm ‘A’ where
no changes are made to the irrigation system. In this scenario, grid
electricity was purchased at a rate of $0.12/kWh. As such, there is no
capital expenditure associated with this scenario and operational costs
arise purely from purchasing grid electricity.
Next, the FPV scenarios were modelled with input parameters shown
in Table 2. This was done for both a small footprint and large footprint
type floating structure. Finally, a system with ground-mounted PV was
Fig. 2. HOMER Pro® system schematic. simulated. In all the scenarios incorporating PV, grid electricity
providing backup power was modelled. Any excess energy captured
from the PV systems was sold to the grid at $0.0705/kWh.
PV system analyses using HOMER Pro® were undertaken for the
2500
following scenarios:
Average Daily Electrical Demand,

2000
1. Reservoir with no PV; power from Jordanian national grid.
2. Small Footprint FPV.
1500
3. Large Footprint FPV.
4. Ground-mounted PV.
kWh/d

1000

500
Table 2
0 Parameters used as inputs to simulation model for all PV scenarios.
Load
Well Pump Average Load (kWh/d) 121.6

Booster Pump Well Pump Deferrable Load Storage Capacity (kWh) 10,000
Booster Pump Average Load (kWh/d) 851.9
Figure 3 - Average daily electrical demand for Random day-to-day variability for booster pump 5%
Random hourly variability for booster pump 5%
each pump across the year [36]
PV system
Grid Interconnection Charge $100 per kWp [46]
Fig. 3. Average daily electrical demand for each pump across the year [36]. Efficiency 18% [47]
Derating Factor 80%
Temperature Effects on Power − 0.4 %/◦ C [47]
The system consists of six elements – PV array, DC to AC converter, Annual Performance Degradation 0.7 %/year [48]
grid electricity, booster pump load, well pump load and the irrigation Inverter
reservoir. Fig. 2 shows the HOMER Pro® schematic of the system. A CAPEX Cost Curve generated from [49]
Replacement Cost 90% of CAPEX
detailed list of specifications for the system design can be found in Ap­
OPEX 5% of CAPEX
pendix A. Efficiency 95%
From analysis of the energy consumption used in the farm, it is Lifetime 15 years
estimated that for every m3 of water pumped by the irrigation system, Grid Tariffs
1.55 kWh of electrical power was required. Fig. 3 shows the average Purchasing Price $0.12/kWh [41]
Selling Price $0.0705/kWh [40]
daily electrical demand of the well and booster pump in each month of Financials [50,51]
the year based on farm operational data [44]. Resource data for global Nominal Discount Rate 8%
horizontal irradiation and average daily ambient temperature were Expected Inflation Rate 2%
imported to HOMER Pro® from NASA Surface Meteorology and Solar Project Lifetime 20 years
Max. Annual Capacity Shortage 2%
Energy Database [45] as shown in Fig. 4.

30 10
Irriadiation, kWh/m2/day

8
Global Horizontal
Temperature,◦C
Avergae Daily

20
6

4
10
2

0 0
Jan Feb Mar Apr May June Jul Aug Sept Oct Nov Dec
GHI Temperature

Fig. 4. Average Global Horizontal Irradiation and ambient temperatures at 32.5◦ N, 36.5◦ E for each month of the year.

5
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

Table 3 3.3. Estimation of surface water evaporation

Comparison between HOMER Pro® inputs for FPV and ground-mounted PV.
Floating PV Ground- Notes In this section, the methods used to estimate the surface water
(Scenario 2 mounted PV evaporation with and without floating PV are described.
& 3) (Scenario 4)

PV CAPEX ($USD/ 1000 900 CAPEX of FPV tends to be 3.3.1. Method for open water surface
kWp) $100/kWp more than Water evaporation from the irrigation reservoir was estimated using
ground-mounted PV [7] the well-established Penman Equation (3), [53].
PV OPEX ($USD/ 11 11 [7]
6.43 (fu )D
kWp/year) E = Δ+γΔ
. Rλn + Δ+γ
γ
. λ ( SEQ Equation \* ARABIC 3).
Nominal 40 for Small 48 [47] To account for
Where E is the open-water evaporation in mm/day, Rn is the net
Operating Cell Footprint evaporative cooling effect
Temperature [13] radiation at the surface in MJ/m2/day, Δ is the slope of the saturation
(◦ C) 44 for Large vapour pressure curve in kPa/◦ C, γ is the psychrometric coefficient in
Footprint kPa/◦ C, λ is the latent heat of vapourisation in MJ/kg, fU is the wind
[13]
function and D is the vapour pressure deficit in kPa.
Ground 6% [13] 20% 20% is the default value
Reflectance for HOMER Pro® for
The term, Rn, in Equation (3) was calculated using the following
ground-mounted PV. equations:

Panel Inclination 10 32 32◦ is the optimum tilt SWn = (1 − α)(Rd + Rb ) (4)

(◦ ) angle for Farm ‘A’. FPV
√̅̅̅̅̅
inclination typically LWn = σ Tw4 (0.56 − 0.0092 ea )(0.1 + 0.9C) (5)
between 10◦ & 15◦ . Once
initial modelling was
Rn = SWn + LWn (6)
complete the FPV
inclination angle was 2
Where Rd is the diffuse irradiation in MJ/m /day, Rb is the direct
optimised based on NPC
beam radiation in MJ/m2/day, α is the absorptivity of the Earth’s at­
and renewable energy
yield. mosphere, taken to be 0.3, SWn is the net incoming short wavelength
radiation in MJ/m2/day, LWn is the net incoming long wavelength ra­
diation in MJ/m2/day, σ is the Stefan-Boltzmann constant, 5.67x10-8 W/
m2/◦ C4 and C is the cloudiness index. For simplification, the summation
Table 4
Cost comparison between advanced energy capture technologies for FPV.
of Rd and Rb in Equation (4) was taken to be equivalent to the global
horizontal irradiation, RGHI. This assumption has limited impacts on the
Horizontal Vertical Active HAT + VAT +
results of the calculation [54].
Axis Tracking Axis Cooling Cooling Cooling
Tracking
3.3.2. Evaporation estimation incorporating FPV
CAPEX 1200 1200 1063 1263 1263
($USD)
To incorporate FPV into the evaporation model, the method
OPEX 20 20 25 25 25 described by Scavo et al. was deployed [55]. Below is the consideration
($USD/ for undertaken for the two floating PV cases.
year) A) Small Footprint FPV:
Small Footprint FPV systems are not completely in contact with the
A comparison between the floating PV and ground-mounted PV water surface, so they protect the water surface from direct beam radi­
specifications are shown in Table 3. ation but not diffuse radiation. Hence, Equation (4) is modified to only
The impact of incorporating advanced technologies to improve the include the diffuse component of the solar radiation as shown in Equa­
energy capture of FPV were then assessed. This was done for both small tion (7).
footprint and large footprint FPV. The technologies assessed were: SWnSF = (1 − α)Rd (7)
2
• Horizontal axis tracking (HAT) Rd is the diffuse horizontal radiation in MJ/m /day, which is
• Vertical axis tracking (VAT) determined as from Equation (8).
• Active cooling mechanism Rd = Kd RGHI (8)
• HAT and active cooling
• VAT and active cooling Where Kd is the diffuse fraction of the global horizontal radiation,
RGHI. It is calculated from the following empirical Equation for Amman,
FPV systems which use tracking technology have a higher capital Jordan [56] which uses the cloudiness index, C defined in Equation (9).
investment and higher operation and maintenance costs, but improved Kd = 0.847 − 0.985C (9)
energy capture. Hence, it was assumed the CAPEX and OPEX for a sys­
tem incorporating single axis tracking technology $1,200 per kWp and The net longwave radiation is calculated using Equation (10).
$20/kWp/year respectively. √̅̅̅̅̅
LWnSF = σTw4 (0.56 − 0.0092 ea )(0.1 + 0.9 × 0.3) (10)
As with tracking technology, the use of active cooling technologies
incurs additional CAPEX and OPEX requirements. Dizier [52] estimates To determine the evaporation rate from the entire irrigation reser­
that deploying a sprinkling type cooling system adds $63/kWp of CAPEX voir, EFPV, we incorporate the fraction of the water’s surface covered by
and the FPV with cooling has an OPEX of $25/kWp/year while the FPV structure, x, determined from Equation (11).
improving the energy capture of the system by 6.6%. The cost specifi­ EFPV = Efree (1 − x) + xEnSF (11)
cation comparison for these options is summarised in Table 4. Sub­
merged PV, FPV with concentrators and bi-facial FPV were not analysed Where EFree is the evaporation from the free surface not covered by
as part of this study due to the nascent stage of these technologies. the FPV structure.
B) Large Footprint FPV.
For Large Footprint FPV systems, negligible solar radiation hits the

6
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

Net Present Cost (1000 $USD)

500

400

300

200

100

0
No PV Small Footprint FPV Large Footprint FPV Land-based PV

Fig. 5. NPC comparison between No PV, small footprint FPV, large footprint FPV and ground-mounted PV.

250
Annualised GHG Emissions

200
(kgCO2e/year)

150

100

50

0
No PV Small Footprint FPV Large Footprint FPV Land-based PV

Fig. 6. Annual GHG emission comparison between no PV, small footprint FPV, large footprint FPV and ground-mounted PV.

surface of reservoir. As a result, the equations are modified to:

∑
n

SWnLF = 0 (12) GT = A*GPV + Gi *Eg *(1 + β) (17)

i

Where A is the footprint of the FPV installation in m2, Eg is the annual

√̅̅̅̅̅
LWnLF = σTw4 (0.56 − 0.0092 ea )(0.1) (13)
grid electricity purchased from the grid in kWh, GPV is the embodied
Rn = Rfree (1 − x) + xRLF (14) greenhouse gases emissions of the PV system in kgCO2e/m2. For a
ground-mounted PV system, the embodied carbon emissions are 115
EFPV = (1 − x)Efree ’ (15) kgCO2e/m2 [20]. For floating PV, Santafé et al. [20] estimates embodied
emissions of 137.73 kgCO2e/m2. To be on the conservative side, a value
Where RFree is the net radiation on a free water surface and EFree’ is
of 150 kgCO2e/m2 was used for this analysis.
the evaporation rate when Rn is calculated using Equation (3).

4. Results
3.4. Estimation of greenhouse gas emissions

This section provides a summary of the results for the various sce­
The total greenhouse gas emissions, GT, for the ‘no PV’ or business-
narios modelled and their implications to water evaporation, economics
as-usual scenario was estimated using Equation (16).
and emissions.
∑
n
GT = Gi *ES *(1 + β) (16)
i
4.1. Comparison of modelled PV systems

Where Gi is the carbon dioxide equivalent emissions per unit grid Based on the input data, including operational data for the farm, the
electricity produced in Jordan in kgCO2e/kWh in year i, ES is the annual simulations indicated that a 300 kWp PV system and 200 kW inverter as
energy consumption by the system in year i, β is the energy losses the cost optimum solution for the FPV scenarios. To make a fair com­
associated with electricity transmission and distribution taken to be 12% parison between FPV and ground-mounted PV, the same PV and inverter
[57], n is the project lifetime taken to be 20 years. capacity were chosen for the ground-mounted PV scenario. Figs. 5-7
While the FPV system itself does not emit any greenhouse gases, below compare the four scenarios based on NPC, annual evaporation,
there are emissions associated with manufacturing, transportation, and and annualised greenhouse gas emissions. A complete list of the overall
installation. For the PV scenarios, some grid electricity had to be pur­ modelling results can be found in Appendix A and recommended system
chased to provide power during high demands or during periods of low designs are in Appendix B.
solar resource. The greenhouse gas emissions for the PV scenarios were Fig. 5 shows that large footprint FPV is the most economically
estimated using Equation (17). attractive option. Meanwhile the business-as-usual scenario is the least
attractive from an NPC standpoint. Fig. 7 shows the Large Footprint FPV
reduces surface water evaporation dramatically while the Small

7
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

2500

Annual Evaporation from

2000

Reservoir (mm/year) 1500

1000

500

0
No PV Small Footprint FPV Large Footprint FPV Land-based PV

Fig. 7. Annual evaporation comparison between no PV, small footprint FPV, large footprint FPV and ground-mounted PV.

520

500
NPC in $1000 USD

480

460

440

420

400

380
Fixed Horizontal Vertical Tracking Active Horizontal Vertical Tracking +
Tracking Cooling Tracking + Cooling
Free Standing Large Footprint Cooling

Fig. 8. Net Present Cost comparison between fixed FPV and FPV with advanced energy capture technologies.

Renewable Energy Fraction in %

442 75.5

440
75.4
NPC in $1000 USD

438

436 75.3

434 75.2
432
75.1
430

428 75
5 7 9 11 13 15 17 19 21 23 25
Panel Inclination in degrees
NPC Renewable Energy Fraction

Fig. 9. Impact of panel inclination on NPC and renewable energy fraction.

Footprint FPV installation only reduces evaporation by around 200 mm The ground-mounted PV option has an NPC competitive with the two
per year. Fig. 6 shows all PV options reduce greenhouse gas emissions FPV options. On the basis that it has higher annualised GHG emissions
substantially, with the Large Footprint FPV reducing emissions the most and does not reduce evaporation from the reservoir, it was deemed
compared to ground-mounted PV and Small Footprint FPV. The Large unfavourable compared to the FPV scenarios.
Footprint design was estimated to save 141 tCO2e per year compared to The best option is the Large Footprint FPV since it saves a large
the ‘No PV’ scenario. amount of water from evaporation (Fig. 7) and has the lowest GHG
From the results shown in the Figs. 5 – 7 it is evident that the ‘No PV’ emissions (Fig. 6). Its NPC is also slightly lower than the small footprint
or business-as-usual scenario is the least attractive option given it has type FPV option (Fig. 5). The Large Footprint design was estimated to
the highest NPC, highest GHG emissions, and highest annual evapora­ have a simple payback of 8.4 years.
tion from the irrigation reservoir. Its only advantage is the zero-capital
expenditure compared to the other options.

8
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

10

Avergae daily evaporation 8

(mm/day)
6

0
Jan Feb Mar Apr May June Jul Aug Sept Oct Nov Dec

No PV Large Footprint Small Footprint

Fig. 10. Average daily evaporation in each month of year for the No PV and FPV scenarios.

4
Evaporation in mm/day

3
Saving in Water

0
Jan Feb Mar Apr May June Jul Aug Sept Oct Nov Dec

Large footprint Small Footprint

Fig. 11. Average daily savings in evaporation in each month of the year for large footprint and small footprint FPV.

Table 5
Enhanced energy capture of FPV compared to ground-mounted PV model.
FPV Operating Cell Temperature in ◦ C Annual electricity production in MWh Annual electricity production in MWh for % increase in energy
Typology (reduced from the 48 ◦ C standard) for 300 kWp FPV System equivalent 300 kWp ground-mounted System production

Large 44 462 456 1.3%

Footprint
Small 40 467 456 2.4%
Footprint

the requirements of Farm A’s water pumping.

Table 6 At present, the additional costs of these technologies in terms of
Water saved in terms of m3/MWh, m3/MWp and percentage savings compared CAPEX and OPEX outweigh the advanced energy capturing properties. If
to literature.
the costs of tracking and cooling decline, then these technologies may
300 kW Large 300 kW Small Published become competitive in the future. There is also the added practical
Footprint FPV Footprint FPV Literature problem associated with maintenance. Farm ‘A’ is in a rural location so
Data
access to suitably qualified maintenance staff is problematic. Tracking
Water saved (m3/MWh) 27.5 3.2 0.8–30 and active cooling both require a more intensive maintenance schedule
Water saved (m3/MWp) 42,000 7,500 1,200–60,000
compared to standard FPV. Hence, this further justifies why standard
Percentage saving 75% 43% 60–90%
compared to equivalent fixed tilt FPV without active cooling is preferable for Farm ‘A’.
area of free surface
Percentage saving 42% 8.3% –
compared to entire 4.3. Impact of PV module inclination
reservoir uncovered
Compared to all other scenarios, including business-as-usual,
ground-mounted PV, Small Footprint FPV and FPV with tracking and/
4.2. Impact of using advanced energy capture technologies or cooling, a fixed tilt FPV with a Large Footprint type structure was
deemed best across all three criteria as discussed in Sections 4.1 - 4.2.
Based on HOMER Pro® the best system design chosen to simulate The impact of altering the panel inclination angle on the Net Present
advanced technologies, such as tracking and cooling, has a PV capacity Cost and fraction of renewable energy consumed by the project was
of 200 kWp and a 150 kW inverter. investigated next as shown in Fig. 9. It is clear from the results that the
Fig. 8 compares the NPC of these scenarios to the standard fixed tilt optimum tilt angle in terms of NPC is 20◦ . However, commercial FPV
FPV scenarios. The figure shows that a standard fixed tilt FPV without installations rarely exceed an angle of 15◦ so that they can withstand
active cooling system is superior in terms of its NPC than FPV with wind loads. A tilt of 15◦ corresponded to the highest renewable energy
tracking and/or active cooling when designing a FPV solution to satisfy fraction, hence reducing greenhouse gas emissions. On these bases, an

9
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

inclination angle of 15◦ was selected for the recommended system 5.3. Limitations of FPV in Jordan
design.
The significant up-front cost of floating PV is one of the main
drawbacks for implementing the technology in Jordan. The proposed
4.4. Evaporation and cost savings system has a capital expenditure of $USD 344,000. This level of in­
vestment is unviable unless the farm owner has access to finance or can
Based on the model, described by Scavo et al. [55], to estimate the receive support through a water-energy-food saving scheme that may be
evaporation rates with and without floating PV, it is apparent that the required in the future.
Large Footprint FPV structure reduces evaporation the most. Fig. 10 Another drawback is maintenance, although Jordan has capable
shows the average daily evaporation in each month of year for the ‘No engineering capacity, rural locations may present a challenge to provide
PV’ and FPV scenarios. Fig. 11 shows the average daily savings in urgent support when needed. Nevertheless Farm ‘A’ is accessible by road
evaporation in each month of the year for Large Footprint and Small and is 1 h and 15-minute drive from the capital city, Amman, so
Footprint FPV. As can be seen from the figures, the Large Footprint PV accessing maintenance should not be problematic.
provides the maximum reduction in evaporation and hence water sav­
ings as compared with the other scenarios. 5.4. Recommendations to increase uptake of FPV in Jordan
The marked difference in evaporation rates between the Large
Footprint and Small Footprint FPV is due to the fact that the former The Feed-in Tariff (FIT) for the project was taken to be $0.0705 per
blocks almost all incoming short wavelength radiation while the latter kWh, the most up to date value for grid-connected PV projects in Jordan,
blocks only direct beam radiation while diffuse radiation is still sub­ which is less than half what it was in 2012 ($0.148 per kWh). It is un­
jected to the water surface. clear whether the FIT will reduce again but should the value change, the
In addition, the water savings from deploying FPV will also result in a modelling and the Net Present Cost of the FPV system would need to be
reduced electrical demand since less water is required to be pumped to updated to reflect this.
the irrigation reservoir. There is, hence, a cost saving associated with the Other countries have specific FITs for floating PV such as Vietnam
water saving for each type of FPV design. For the Large Footprint design, where the FPV Feed-in Tariff is 8.5% higher than that of ground-
this cost saving is estimated to be $1,360 per year while the Small mounted PV [60]. Taiwan, has a generous FIT for FPV, set at $0.14/
Footprint FPV saves around $280 per year. This cost saving was esti­ kWh in 2019 compared to $0.13/kWh for ground-mounted PV [61]. It is
mated using Equation (18) with the values for the large footprint design therefore, recommended that Jordan adopts a similar policy whereby
given as an example. the financial incentives for FPV are greater than ground-mounted PV to
encourage uptake over the coming years.
kWh m3 $68.1
CostSaving = 1.55 × 12, 700 × = $1, 360peryear (18) Another financial incentive that could be applicable for FPV in Jor­
m3 year MWh
dan include a Renewable Energy Certificates (REC) scheme whereby
Where 1.55 kWh/m3 is the energy requirement of the two water FPV has a higher weighting than other forms of electricity generation
pumps mentioned earlier, $68.1/MWh is the Levelised Cost of Electricity such as the scheme in South Korea [62]. At present, Jordan does not
(LCOE). have such as scheme, but it is being considered for the future [63].
Should Jordan want to deploy multi-MW FPV projects, it is recom­
5. Discussion mended that the tendering process is separated from ground-mounted
PV as is the case in Taiwan and India [7]. Jordan has several large
5.1. Enhanced energy capture due to evaporative cooling effect water bodies, such as King Talal Dam and Sharhabeel Dam which may be
suitable for housing large scale FPV installations. Deployment of large-
As stated in Table 3, section 3.2, the evaporative cooling effect of scale FPV would improve Jordan’s energy security, by reducing its
water on the PV modules was modelled by reducing the nominal oper­ reliance on fossil fuel imports from neighbouring countries.
ating cell temperature from the standard 48 ◦ C to values given in In essence, this study has wider applicability, not just at Farm ‘A’, but
Table 3. To check this assumption was correct, the annual electricity at many such farms in Jordan where irrigation reservoirs are present.
production from the PV was compared as shown in Table 5. This is Where farms are not connected to the national grid, floating PV would
broadly consistent with the findings by Oliviera-Pinto [14], Yadav [58] be a particularly attractive option for irrigation pumping compared to
and L. Liu [17]. Hence, the assumptions were deemed appropriate. pumps powered by diesel generators.

5.5. Global recommendation for utilising FPV on irrigation reservoirs

5.2. Water evaporation reductions
Water resources are essential for both drinking water and crop irri­
The model used to estimate the evaporation under the ‘No PV’ sce­ gation, most of this is derived from freshwater reservoirs or ponds.
nario yielded a total of 2235 mm/year (see Appendix C). This is Globally, these reservoirs represent a very large surface area for
consistent with literature values which stated that, in Northwest Jordan, deployment of floating PV reducing the impact on valuable land whilst
the annual surface water evaporation is 1900 mm [59]. providing clean power for pumping at enhanced energy yield. In a
In terms of the evaporation under the FPV scenarios, the Large development setting and in addition to their role in providing water,
Footprint FPV saves 42% of water compared to an uncovered reservoir these reservoirs present a resource that can be exploited for power
whereas the small footprint FPV saves 8% of water compared to an generation whilst reducing evaporation of scares water. The generated
uncovered reservoir. These values are consistent with literature, hence power could be used for the required pumping but also as an anchor load
the model used can be deemed appropriate. for remote communities’ power supply systems.
The water savings in terms of m3/MWh and m3/MWp are shown in In developing countries where 90% of irrigation stems from such
Table 6 below. Hence, the model used is consistent with the literature. reservoirs, which are mainly located in remote areas where the national

10
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

grid supply is either not available or is weak and unreliable. Hence, it technology should be considered, such as introducing a separate FIT for
can be envisaged that the FPV system could also represent the power FPV compared to ground-mounted PV.
plant for a community mini grid supplying energy to services such as In the approach presented to alleviate some of the challenges of the
schools, health clinics, other surrounding buildings [64–67] Further­ energy–water–food nexus, particularly for water conservation and the
more, it is clear from our analyses that the presented approach can displacement of fossil fuel utilisation, it is clear from our results that the
alleviate some of the challenges of the energy–water–food nexus, use of floating photovoltaics (FPV) to cover irrigation reservoirs pro­
particularly for water conservation and the displacement of fossil fuel vides such benefits. Such systems do not only conserve water by
utilisation. Hence, the local and global benefits from deploying FPV on reducing evaporation losses but results in enhancement of the energy
reservoirs are vast not only from the point of view displacing carbon but yield (hence economics) of the deployed PV systems. As indicated
also in addressing such a nexus whilst providing a nucleus for devel­ earlier, in developing countries 90% of irrigation use water from such
opment in remote off grid areas. reservoirs, some of these are in remote off-grid regions. Hence deploy­
Our recommendation, which is part of our future work, is to map ment of FPV will also aid in providing energy access not only for water
these reservoirs and their surrounding demographics in a selected pumping but also for community utilisation. Hence, the global benefits
representative sample of developing countries to scope the potential for from such deployment are huge, and through the presented case study,
community power generation and how this can be linked to reducing this work provides key knowledge to support such needed development
water evaporation through the installation of PV on these reservoirs. The including those represented in the water-energy-food nexus.
work will also need to address policy and economic implications taking This work has shown that floating PV can provide both economic and
into account the added benefit of power supply to the community. The environmental benefits when used on irrigation reservoirs not only in
work will culminate in providing a road map for such deployment the case study country but also beyond. There is no body of knowledge
backed up with evidence for policy makers and financial institutions, that is available to mobilise these ponds for power generation as well as
such as the World Bank. conserving water through FPV and this work is providing initial seminal
evidence of such potential. Furthermore, the presented research is
6. Conclusions generalisable which can address some of the global challenges faced in
the water-energy-food nexus. The novel combination presented here in
Water resources are central to human and are essential for biodi­ not only important from the point of view of sustainable power gener­
versity. Water, energy, and food security are inextricably linked and, as ations but is also linked to water scarcity and the energy-water-food
such, this nexus is a major global societal challenge. There are over 300 nexus. Water scarcity is becoming worse due to climate change and
million drinking water and irrigation ponds (reservoirs) globally where the proposed multifunctional solution of floating solar PV mitigates such
90% of the world’s standing irrigation water resides and their use is impacts. To our knowledge no previous studies have been conducted to
expanding considerably. Hence, this work tackles two important issues investigate this combination whilst providing impactful results in terms
related to energy and water in such reservoirs, thereby addressing ele­ of scientific outcome, evidence, and direction to policy as well as
ments of the UN Sustainable Development Goals. sustainability.
Floating solar photovoltaics (FPV) is a developing technology nor­
mally generating electricity from systems deployed on water bodies thus Declaration of Competing Interest
eliminating the need for utilising valuable land. The feasibility of a
floating photovoltaic system, that can be deployed on irrigation reser­ The authors declare that they have no known competing financial
voirs was investigated. The case study selected was Jordan, a country interests or personal relationships that could have appeared to influence
with a harsh environment and scarce water resources. The systems the work reported in this paper.
investigated have the dual purpose of (a) providing power to two irri­
gation pumps at the case study farm and (b) reducing water evaporation Acknowledgments and funding
from the reservoir. Three main criteria were used to establish the best
option: water savings, economics, and greenhouse gas emissions. An This work forms part of the activities of the Energy and Climate
optimisation tool was used to select the best system design based on Net Change Division and the Sustainable Energy Research Group in the
Present Cost. Water evaporation rates were estimated using the Penman Faculty of Engineering and Physical Sciences at the University of
Equation. The method outlined by Scavo et al. [55] was deployed, to Southampton (www.energy.soton.ac.uk). In addition to this main
account for the presence of FPV, in the evaporation rate modelling. source, support for the work is also derived from EPSRC grant EP/
It was found that a 300 kWp large footprint type FPV structure with a R030391/1 Fortis Unum: Clustering Mini Grid Networks to Widen Energy
tilt of 15◦ connected to a 200 kW inverter was the best solution across all Access and Enhance Utility Network Resilience; and Innovate UK Project
three criteria. Augmenting advanced energy capture technologies to the Number: 40582 Demand Side Renewable Agriculture Business Led
system, such as tracking and cooling, were deemed economically Enterprise.
unfavourable by comparison to the fixed FPV options. Hence, a standard
fixed tilt design without active cooling was adopted. The proposed FPV Appendix A:. Modelling results
design was also deemed favourable economically compared to an
equivalent ground-mounted PV system. The overall modelling results are summarised in Tables A.1-A.3,
In terms of water evaporation benefits, the FPV installation was below, for the four main scenarios (No PV, small footprint FPV, large
estimated to save 12,700 m3 of water per year. This is equivalent to a footprint FPV and ground-mounted PV) as well as the best scenario from
saving of 42% of water as compared to the business-as-usual scenario the analysis of FPV incorporating advanced energy capture technolo­
(without PV cover). The financial impact of reducing evaporation was gies, namely vertical axis tracking with large footprint FPV. The rec­
estimated to be a saving of $1,360 per year for the farm. ommended system design, which proved best across all three criteria
The proposed system was estimated to save 141 tonnes of carbon (economics, water savings and greenhouse gas emissions), is highlighted
dioxide equivalent emissions per year. A simple payback time of 8.4 in green. The decision for the recommended system design was based on
years and an NPC $170,000 less than the current regime of pumping the data presented in Figs. 5-8 in Sections 4.1 and 4.2.
irrigation water was identified for the best solution. Our results show
that FPV on agricultural irrigation reservoirs is a suitable technology
that can contribute to the goal of increasing the water conservation in
Jordan. As such, policies to encourage widespread uptake of the

11
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

Table A1
Overall Modelling Results: System Design.
No PV Fixed Tilt Small Footprint Fixed Tilt Large Footprint Vertical Axis Tracking Large Fixed Tilt Ground-
(Business as FPV FPV Footprint FPV mounted PV
Usual)

PV Capacity(kW) – 300 300 200 300

PV Capacity Factor (%) – 17.4 17.6 19.9 17.6
Annual PV Generation – 456 462 349 462
(MWh)
Footprint on Water (m2) – 3000 4285 2860 –
Inverter Capacity (kW) – 200 200 150 200
Inverter Capacity Factor – 24.6 24.8 25.2 24.7
(%)
Grid Purchases (MWh) 369 142 130 133 151
Grid Sales (MWh) – 207 213 111 216

Table A2
Overall Modelling Results: Economics.
No PV Fixed Tilt Small Footprint Fixed Tilt Large Footprint Vertical Axis Tracking Large Fixed Tilt Ground-
(Business as FPV FPV Footprint FPV mounted PV
Usual)

NPC ($USD) 600,000 492,000 431,000 457,000 439,000

LCOE ($/MWh) 140 71.0 68.1 87.3 66.9
CAPEX ($USD) 0 344,000 344,000 270,000 314,000
OPEX ($/kWp/year) 53,800 12,100 9900 18,000 12,800
Simple Payback Time (Years) – 8.8 8.4 8.0 8.0
Discounted Payback Time – 12.5 11.75 11.25 11.25
(Years)

Table A3
Overall Modelling Results: Environmental Implications.
No PV Fixed Tilt Small Footprint Fixed Tilt Large Footprint Vertical Axis Tracking Large Fixed Tilt Ground-
(Business as FPV FPV Footprint FPV mounted PV
Usual)

Absolute Evaporation – 2490 12,700 11,200 –

Reduction (m3)
Percentage Reduction (%) – 8.3 42.1 37.0 –
Value of Water Saved ($USD/ – 274 1360 1520 –
year)
GHG Emissions (tCO2e/year) 235.8 102.7 95.0 106.2 107.7

Appendix B:. Energy balance for recommended system design December to March when the electrical demand from the two pumps is
and cash flow analysis low but there is still a reasonable solar resource. This excess electricity is
sold to the grid. Fig. B.5 shows export and import form the grid to the
The recommended system design is the 300 kW fixed tilt large selected PV system.
footprint floating PV design, as informed by Figs. 5-8 in Sections 4.1 and Fig. B.6 and Fig. B.7 show the high capital expenditure of $344,000
4.2. Here, plots depicting key attributes of this system are presented in in year 0, which mainly derives from the floating PV system (modules,
Figs. B.1-B.4. moorings, floating structure etc). As the years progress, the costs
It is apparent from Fig. B.1 and Fig. B.2 that the maximum and most incurred are mainly associated with operating expenses i.e. purchasing
consistent power output from the PV and inverter occurs during the grid electricity to provide backup power. In year 15, the inverter needs
summer months while the lowest and least reliable power output occurs replaced, hence the added cost in that year. In the final year, when the
in the winter months. Fig. B.3 clearly shows that backup power is needed system is decommissioned, the positive cash flow in this year arises from
during the night time of summer months where there is no solar resource the salvage value of the PV system and inverter. Fig. B.8, below, shows
yet a high demand to power the two pumps. Fig. B.4 shows that excess the cumulative cash flow for the recommended system design:
electricity is produced from the PV during the daylight hours from

12
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

Fig. B1. Hourly power output from 300 kW large footprint FPV array during the first year of the project.

Fig. B2. Hourly inverter output from 200 kW inverter during the first year of the project.

Fig. B3. Hourly electricity purchased from grid to provide backup power for the pumps during the first year of the project.

Fig. B4. Hourly electricity sold to grid during periods of excess electricity production by the FPV array during the first year of the project.

13
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

Fig. B5. Average monthly electrical production per month from FPV array (orange) and average electricity purchased per month from the grid (green).

Fig. B6. Annual discounted cash flow on cost type basis over project lifetime.

Fig. B7. Annual discounted cash flow on system component basis over project lifetime.

14
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

Fig. B8. Cumulative cash flow for recommended system design showing both discounted and nominal cash flows.

Appendix C:. Evaporation rate data

Table C1
Evaporation data with and without FPV for each month of the year.
Evaporation (mm)
Scenario Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total

No PV 85 88 141 176 257 287 298 282 220 180 130 91 2235
Large footprint FPV 49 51 81 102 148 166 173 163 127 104 76 53 1293
Small footprint FPV 78 89 127 163 230 265 266 254 204 165 126 85 2051

References Renewable and Sustainable Energy Reviews, vol. 81. Elsevier Ltd, pp. 1730–1741,
Jan. 01, 2018. doi: 10.1016/j.rser.2017.05.269.
[13] H. Liu, V. Krishna, J. Lun Leung, T. Reindl, and L. Zhao, “Field experience and
[1] Mekonnen MM, Hoekstra AY. Sustainability: Four billion people facing severe
performance analysis of floating PV technologies in the tropics,” Progress in
water scarcity. Sci Adv Feb. 2016;2. https://doi.org/10.1126/sciadv.1500323.
Photovoltaics: Research and Applications, vol. 26, no. 12, pp. 957–967, Dec. 2018,
[2] Food and Agriculture Organisation of the United Nations, “The Water-Energy-Food
doi: 10.1002/pip.3039.
Nexus: A new approach in support of food security and sustainable agriculture,”
[14] Oliveira-Pinto S, Stokkermans J. Assessment of the potential of different floating
2014. https://www.fao.org/documents/card/en/c/182bf36b-87fa-4ea5-b898
solar technologies – Overview and analysis of different case studies. Energy
-06c89c88f241/ (accessed Jan. 01, 2022).
Convers Manage May 2020;211:112747. https://doi.org/10.1016/j.
[3] UN Water, “UN World Water Development Report 2021,” 2021. https://www.unw
enconman.2020.112747.
ater.org/publications/un-world-water-development-report-2021/ (accessed Jan.
[15] “PVsyst - Photovoltiac Software,” 2021. https://www.pvsyst.com/ (accessed Jan.
01, 2022).
09, 2021).
[4] O. Adeyemi, I. Grove, S. Peets, and T. Norton, “Advanced Monitoring and
[16] N. Yadav, M. Gupta, and K. Sudhakar, “Energy assessment of floating photovoltaic
Management Systems for Improving Sustainability in Precision Irrigation,”
system,” in International Conference on Electrical Power and Energy Systems, ICEPES
Sustainability 2017, Vol. 9, Page 353, vol. 9, no. 3, p. 353, Feb. 2017, doi: 10.3390/
2016, May 2017, pp. 264–269. doi: 10.1109/ICEPES.2016.7915941.
SU9030353.
[17] Liu L, Wang Q, Lin H, Li H, Sun Q, Wennersten R. Power Generation Efficiency and
[5] J. F. Velasco-Muñoz, J. A. Aznar-Sánchez, L. J. Belmonte-Ureña, and I. M. Román-
Prospects of Floating Photovoltaic Systems. Energy Procedia May 2017;105:
Sánchez, “Sustainable Water Use in Agriculture: A Review of Worldwide Research,”
1136–42. https://doi.org/10.1016/j.egypro.2017.03.483.
Sustainability 2018, Vol. 10, Page 1084, vol. 10, no. 4, p. 1084, Apr. 2018, doi:
[18] W. Charles Lawrence Kamuyu, J. Lim, C. Won, and H. Ahn, “Prediction Model of
10.3390/SU10041084.
Photovoltaic Module Temperature for Power Performance of Floating PVs,”
[6] Liu H, Kumar A, Reindl T. “The Dawn of Floating Solar—Technology, Benefits, and
Energies (Basel), vol. 11, no. 2, p. 447, Feb. 2018, doi: 10.3390/en11020447.
Challenges”, in Lecture Notes in Civil Engineering. Springer 2020;41:373–83. https://
[19] Majid ZAA, Ruslan MH, Sopian K, Othman MY, Azmi MSM. Study on performance
doi.org/10.1007/978-981-13-8743-2_21.
of 80 watt floating photovoltaic panel. Journal of Mechanical Engineering and
[7] T. Reindl et al., “Where Sun Meets Water FLOATING SOLAR MARKET REPORT,”
Sciences Dec. 2014;7(1):1150–6. https://doi.org/10.15282/jmes.7.2014.14.0112.
2019. https://esmap.org/where_sun_meets_water_floating_solar_market_report
[20] M. Redón Santafé, J. B. Torregrosa Soler, F. J. Sánchez Romero, P. S. Ferrer Gisbert,
(accessed Apr. 10, 2020).
J. J. Ferrán Gozálvez, and C. M. Ferrer Gisbert, “Theoretical and experimental
[8] Intersolar Europe, “Floating PV: On the Rise in Europe.” https://www.intersolar.
analysis of a floating photovoltaic cover for water irrigation reservoirs,” Energy,
de/market-trends/floating-pv-europe (accessed Jan. 01, 2022).
vol. 67, pp. 246–255, Apr. 2014, doi: 10.1016/j.energy.2014.01.083.
[9] M. Cox, “The State of Floating Solar: Bigger Projects, Climbing Capacity, New
[21] Ferrer-Gisbert C, Ferrán-Gozálvez JJ, Redón-Santafé M, Ferrer-Gisbert P, Sánchez-
Markets,” Greentech Media, 2019. https://www.greentechmedia.com/articles/
Romero FJ, Torregrosa-Soler JB. A new photovoltaic floating cover system for
read/the-state-of-floating-solar-bigger-projects-and-climbing-capacity (accessed
water reservoirs. Renewable Energy Dec. 2013;60:63–70. https://doi.org/
Apr. 24, 2020).
10.1016/j.renene.2013.04.007.
[10] Friel D, Whittaker M, Doran T, Howlin WJ. A review of floating photovoltaic design
[22] Rosa-Clot M, Tina GM, Nizetic S. Floating photovoltaic plants and wastewater
concepts and installed variations. 4th International Conference on Offshore
basins: An Australian project. Energy Procedia Oct. 2017;134:664–74. https://doi.
Renewable Energy, ASRANet Ltd. 2019.
org/10.1016/j.egypro.2017.09.585.
[11] A. Sahu, N. Yadav, and K. Sudhakar, “Floating photovoltaic power plant: A
[23] Durkovíc V, Durišíc Ž. Analysis of the potential for use of floating PV power plant
review,” Renewable and Sustainable Energy Reviews, vol. 66. Elsevier Ltd, pp.
on the skadar lake for electricity supply of aluminium plant in montenegro.
815–824, Dec. 01, 2016. doi: 10.1016/j.rser.2016.08.051.
Energies (Basel) 2017;10(10):1–23. https://doi.org/10.3390/en10101505.
[12] R. Cazzaniga, M. Cicu, M. Rosa-Clot, P. Rosa-Clot, G. M. Tina, and C. Ventura,
[24] Melvin GKX. Experimental Study Of The Effect Of Floating Solar Panels On
“Floating photovoltaic plants: Performance analysis and design solutions,”
Reducing Evaporation In Singapore Reservoirs. (Accessed Apr 2015;5:2022).

15
L.W. Farrar et al. Energy Conversion and Management 260 (2022) 115598

https://cpb-us-w2.wpmucdn.com/blog.nus.edu.sg/dist/b/4438/files/2015/04/ [46] L. Bird, F. Flores, C. Volpi, K. Ardani, D. Manning, and R. Mcallister, “Review of
FYP-Final-report-14ijhop.pdf. Interconnection Practices and Costs in the Western States,” 2018.
[25] Taboada ME, Cáceres L, Graber TA, Galleguillos HR, Cabeza LF, Rojas R. Solar [47] WIOSUN, “Photovoltaic Module CM Series,” 2017. http://wiosun.co/wp
water heating system and photovoltaic floating cover to reduce evaporation: -content/uploads/2018/01/CM-290-300.pdf (accessed Jul. 31, 2020).
Experimental results and modeling. Renewable Energy May 2017;105:601–15. [48] J. K. Copper, K. Jongjenkit, and A. G. Bruce, “Calculation of PV System
https://doi.org/10.1016/j.renene.2016.12.094. Degradation Rates in a Hot Dry Climate,” Asia-Specific Research Conference, 2017.
[26] Zhou Y, et al. An advanced complementary scheme of floating photovoltaic and https://apvi.org.au/solar-research-conference/wp-content/uploads/2017/02/J-
hydropower generation flourishing water-food-energy nexus synergies. Appl Copper-K.-Jongjenkit-and-A-Bruce_Calculation-of-PV-System-Degradation-Rates-
Energy Oct. 2020;275:115389. https://doi.org/10.1016/J. in-a-Hot-Dry-Climate.pdf (accessed Jul. 31, 2020).
APENERGY.2020.115389. [49] Wind & Sun, “SolarEdge Inverters,” 2020. http://www.windandsun.co.
[27] Abdelal Q. Floating PV; an assessment of water quality and evaporation reduction uk/products/Inverters/SolarEdge/SolarEdge-Inverters#.XwBx_ChKhPY (accessed
in semi-arid regions. International Journal of Low-Carbon Technologies Sep. 2021; Jul. 04, 2020).
16(3):732–9. https://doi.org/10.1093/IJLCT/CTAB001. [50] Al-Ghussain L. Economic Assessment of PV Investments in Jordan. Innovative
[28] Goswami A, Sadhu P, Goswami U, Sadhu PK. Floating solar power plant for Energy & Research Nov. 2017;6(2):1–4. https://doi.org/10.4172/2576-
sustainable development: A techno-economic analysis. Environ Prog Sustainable 1463.1000159.
Energy 2019;38(6):Nov. https://doi.org/10.1002/ep.13268. [51] “Inflation, consumer prices (annual %) - Jordan | Data.” https://data.worldbank.
[29] M. Rosa_Clot and G. Tina, “Levelized Cost of Energy (LCOE) Analysis,” in Floating org/indicator/FP.CPI.TOTL.ZG?
PV Plants, Elsevier Inc., 2020, pp. 119–127. doi: 10.1016/B978-0-12-817061- end=2018&locations=JO&start=2018&view=map (accessed Apr. 05, 2022).
8.00010-5. [52] Dizier A. Techno-economic analysis of floating PV solar power plants using active
[30] C. Perakis, G. Kyriakarakos, N. B. Hani, S. Hammad, and M. Damasiotis, cooling technique: A case study for Taiwan. KTH School of Industrial Engineering
“Investigation of solar-powered drip irrigation: The case study of the Jordan and Management 2018.
Valley,” Research in Agricultural Engineering, vol. 63 (2017), no. No. 4, pp. 168–171, [53] Valiantzas JD. Simplified versions for the Penman evaporation equation using
Dec. 2017, doi: 10.17221/12/2016-RAE. routine weather data. J Hydrol Dec. 2006;331(3–4):690–702. https://doi.org/
[31] M. Al-Smairan, “Application of photovoltaic array for pumping water as an 10.1016/j.jhydrol.2006.06.012.
alternative to diesel engines in Jordan Badia, Tall Hassan station: Case study,” [54] P. Rosa-Clot, “FPV and Environmental Compatibility,” in Floating PV Plants,
Renewable and Sustainable Energy Reviews, vol. 16, no. 7. Pergamon, pp. Elsevier Inc., 2020, pp. 101–118. doi: 10.1016/b978-0-12-817061-8.00009-9.
4500–4507, Sep. 01, 2012. doi: 10.1016/j.rser.2012.04.033. [55] Scavo FB, Tina GM, Gagliano A. An assessment study of evaporation rate models on
[32] M. Wright, “Most Water-Stressed Countries In The World For 2019,” CEO World a water basin with floating photovoltaic plants. accessed Apr. 08, 2020 Int J Energy
Magazine, 2019. https://ceoworld.biz/2019/08/08/most-water-stressed-countries- Res 2020. https://onlinelibrary.wiley.com/doi/full/10.1002/er.5170.
in-the-world-for-2019/ (accessed Apr. 24, 2020). [56] Hijazin MI. The diffuse fraction of hourly solar radiation for Amman/Jordan.
[33] Yorke V. Politics matter: Jordan’s path to water security lies through political Renewable Energy Feb. 1998;13(2):249–53. https://doi.org/10.1016/S0960-1481
reforms and regional cooperation. (Accessed Apr 2013;05:2022). https://www.wti. (97)00082-7.
org/media/filer_public/00/98/00989164-e76d-4887-be98-2acd3b193c59/workin [57] World Bank, “Electric power transmission and distribution losses (% of output) -
g_paper_2013_19.pdf. Jordan,” 2018. https://data.worldbank.org/indicator/EG.ELC.LOSS.ZS?
[34] S. Nortcliff, G. Carr, R. B. Potter, and K. Darmame, “Jordan’s Water Resources: locations=JO (accessed Jul. 07, 2020).
Challenges for the Future,” 2008. https://www.reading.ac.uk/web/files/ [58] Neha Yadav and Manju Gupta, “Energy Assessment of FPV,” International
geographyandenvironmentalscience/GP185.pdf (accessed Apr. 05, 2022). Conference on Electrical Power and Energy Systems , 2016. https://ieeexplore.ieee.
[35] G. Abu-Rumman, A. I. Khdair, and S. I. Khdair, “Current status and future org/stamp/stamp.jsp?tp=&arnumber=7915941 (accessed Jun. 10, 2020).
investment potential in renewable energy in Jordan: An overview,” Heliyon, vol. 6, [59] A.-N. Fardous, M. Mudabber, M. Jitan, and R. Badwan, “Harnessing Salty Water To
no. 2, Feb. 2020, doi: 10.1016/j.heliyon.2020.e03346. Enhance Sustainable Livelihoods Of The Rural Poor In Four Countries In West Asia
[36] M. Willuhn, “Orange Jordan adds 37 MW of solar under ‘wheeling’ scheme,” PV And North Africa,” National Center For Agriculture Research And Technology Transfer,
Magazine International, 2019. https://www.pv-magazine.com/2019/08/01/ 2004. https://www.iwmi.cgiar.org/assessment/files_new/research_projects/ICBA
orange-jordan-adds-37-mw-of-solar-under-wheeling-scheme/ (accessed Apr. 10, NationalReport_Jordan.pdf (accessed Jul. 27, 2020).
2020). [60] B. Publicover, “Vietnam finally unveils new FITs for large-scale, rooftop, floating
[37] A. Jager-Waldau, “JRC Science for Policy Report: PV Status Report 2019,” 2019. PV,” PV Magazine International, 2020. https://www.pv-magazine.com/2020/04/
https://ec.europa.eu/jrc/en/publication/eur-scientific-and-technical-research- 07/vietnam-finally-unveils-new-fits-for-large-scale-rooftop-floating-pv/ (accessed
reports/pv-status-report-2019 (accessed Apr. 10, 2020). Jul. 27, 2020).
[38] T. Wenzel and J. Asen, “Market Info Jordan - Photovoltaics,” Deutsche Energie- [61] A. Bhambhani, “Lowered Solar FIT Rates Come Into Effect In Taiwan,” Taiyang
Agentur (DENA), 2014. https://www.dena.de/fileadmin/dena/Dokumente/Pdf/ News, 2020. http://taiyangnews.info/markets/lowered-solar-fit-rates-come-into-
3205_Market_Info_Jordan_Photovoltaic.pdf (accessed Jul. 07, 2020). effect-in-taiwan/ (accessed Jul. 27, 2020).
[39] L. Al-Ghussain, “Economic Assessment of PV Investments in Jordan,” Innovative [62] Korea Energy Agency, “Renewable Portfolio Standards (RPS),” 2015. https://www.
Energy & Research, vol. 6, no. 2, 2017. doi: 10.4172/2576-1463.1000159. energy.or.kr/renew_eng/new/standards.aspx (accessed Jul. 27, 2020).
[40] Dyak ATA, Abu-Lehyeh EO, Kiwan S. “Assessment of Implementing Jordan’s [63] EDAMA, “Recommendations for Energy Sector Strategy,” 2019. https://www.fe
Renewable Energy Plan on the Electricity Grid”, Jordan Journal of Mechanical and s-jordan.org/fileadmin/user_upload/events/Climate_and_Energy_Wuppertal_
Industrial. accessed Apr 10 Engineering 2020;11(2):2017. http://jjmie.hu.edu. COnference/Recommendations_for_Energy_Sector_Strategy.pdf (accessed Jul. 28,
jo/vol-11-2/JJMIE-65-16-01.pdf. 2020).
[41] A. Anwar and N. Amdar, “A Case Study of The cost and benefits of investing in [64] Vernet A, Khayesi JNO, George V, George G, Bahaj AS. How does energy matter?
Water Innovation Technology,” 2019. Rural electrification, entrepreneurship, and community development in Kenya.
[42] HOMER Energy, “Total Net Present Cost,” 2020. https://www.homerenergy.com/ Energy Policy Mar. 2019;126:88–98. https://doi.org/10.1016/J.
products/pro/docs/latest/total_net_present_cost.html (accessed Aug. 07, 2020). ENPOL.2018.11.012.
[43] H. Energy, “Real Discount Rate,” 2020. https://www.homerenergy.com/products/ [65] A. S. Bahaj et al., “The Impact of an Electrical Mini-grid on the Development of a
pro/docs/latest/real_discount_rate.html (accessed Aug. 08, 2020). Rural Community in Kenya,” 2019, doi: 10.3390/en12050778.
[44] N. Amdar, “Private Communication.” 2020. [66] A. S. Bahaj and P. A. B. James, “Electrical Minigrids for Development: Lessons
[45] Atmospheric Science Data Center and NASA, “ASDC | Processing, archiving, and From the Field”, doi: 10.1109/JPROC.2019.2924594.
distributing Earth science data at the NASA Langley Research Center.” https:// [67] Bahaj AS, et al. Pathways to Universal Electricity Access for Rural Communities in
eosweb.larc.nasa.gov/ (accessed Jul. 07, 2020). Africa. IOP Conference Series: Earth and Environmental Science Nov. 2020;588(2):
022047. https://doi.org/10.1088/1755-1315/588/2/022047.

16