Integrating Rainwater Harvesting and Solar Energy Systems for Sustainable Water and Energy Management in Low Rainfall Agricultural Region: A Case Study from Gönyeli, Northern Cyprus

Abstract

The primary objective of this study is to assess the techno-economic feasibility of an innovative solar energy generation system with a rainwater collection feature to generate electrical energy and meet irrigation needs in agriculture. The proposed system is designed for an agricultural area (Gonyeli, North Cyprus) with high solar potential and limited rainfall. In the present study, global rainfall datasets are utilized to assess the potential of rainwater harvesting at the selected site. Due to the lack of the measured rainfall data at the selected site, the accuracy of rainfall of nine global reanalysis and analysis datasets (CHIRPS, CFSR, ERA5-LAND, ERA5, ERA5-AG, MERRA2, NOAA CPC CMORPH, NOAA CPC DAILY GLOBAL, and TerraClimate) are evaluated by using data from ground-based observations collected from the Meteorological Department located in Lefkoşa, Northern Cyprus from 1981 to 2023. The results demonstrate that ERA5 outperformed the other datasets, yielding a high R-squared value along with a low mean absolute error (MAE) and root mean square error (RMSE). Based on the best dataset, the potential of the rainwater harvesting system is estimated by analyzing the monthly and seasonal rainfall patterns utilizing 65 different probability distribution functions for the first time. Three goodness-of-fit tests are utilized to identify the best-fit probability distribution. The results show that the Johnson and Wakeby SB distributions outperform the other models in terms of fitting accuracy. Additionally, the results indicate that the rainwater harvesting system could supply between 31% and 38% of the building’s annual irrigation water demand (204 m³/year) based on average daily rainfall and between 285% and 346% based on maximum daily rainfall. Accordingly, the system might be able to collect a lot more water than is needed for irrigation, possibly producing an excess that could be stored for non-potable uses during periods of heavy rainfall. Furthermore, the techno-economic feasibility of the proposed system is evaluated using RETScreen software (version 9.1, 2023). The results show that household energy needs can be met by the proposed photovoltaic system, and the excess energy is transferred to the grid. Furthermore, the cash flow indicates that the investor can expect a return on investment from the proposed PV system within 2.4 years. Consequently, the findings demonstrate the significance of this system for promoting resource sustainability and climate change adaptation. Besides, the developed system can also help reduce environmental impact and enhance resilience in areas that rely on water and electricity.

1. Introduction

1.1. Background

Water scarcity has become a major issue worldwide, particularly in semi-arid and arid regions, due to population expansion and climate change [1]. Furthermore, high temperatures and insufficient rainfall are expected to exacerbate water shortages in the near future as a result of climate change [2]. According to Shemer et al. [3], more than 40% of the world’s population experiences water scarcity, and more than 700 million people lack access to clean drinking water. Besides, water scarcity affects 80% of the world’s population, according to the fifth evaluation from the Intergovernmental Panel on Climate Change (IPCC) [4]. Additionally, 1.8 billion people are predicted to reside in regions with severe water scarcity by 2025 [5]. By 2050, two-thirds of the world’s population is expected to be impacted by water stress [6]. Therefore, there are significant challenges to social and economic growth due to the serious scarcity of water, especially in arid and semi-arid regions [6]. According to Lepcha et al. [6] and Ritchie and Roser [7], agriculture consumes the most water worldwide (69%), followed by industry (23%) and domestic consumption (8%). The world’s freshwater supply is continually declining as a result of rising urbanization, rapid industrialization, and overexploitation of groundwater. Freshwater is a vital natural resource for the existence and growth of all living organisms around the globe. Thus, effective techniques and the management of water resources are necessary for sustainability [8,9]. The Sustainable Development Goals (SDGs) and sustainability strategies need to be integrated. Velis et al. [10] demonstrated that policies that prioritize the environment should meet both the SDGs and current water demands without endangering the needs of future generations.

Besides, energy scarcity is a significant issue that impacts sustainable development in regions with arid or semi-arid climates [11]. Energy demand management has recently been recognized in both developed and developing countries as a significant global concern [12]. The traditional strategy for supplying the energy demands of urbanization and industrialization has been to increase the consumption of fossil fuels. Fossil fuel combustion produces toxic NO_X, SO_X, and CO_X gases that are harmful to both the environment and human health [13]. Fuel demand has increased due to recent population growth. Concerning levels of environmental contamination have been brought on by the rising use of fossil fuels [14,15]. The primary source of climate change, which is harmful to the environment, is CO₂ emissions from the burning of fossil fuels. Therefore, reducing the amount of CO₂ and other harmful gases in the atmosphere is necessary for addressing climate change. Increasing the usage of renewable energy, such as solar energy, as a power source is the most efficient approach to achieve this [16]. Renewable energy has the potential to help with ongoing challenges with energy security and cost; however, its current generating capacity remains significantly below that needed to reach net-zero goals by 2050, according to Belaïd et al. [17]. In 2021, renewable energy made up just 29% of the world’s electrical generation, with solar and wind accounting for 10% of the entire electricity mix [17]. Additionally, the percentage of renewable energy in the world’s total electricity generation should rise from 29% in 2020 to over 60% in 2030 and over 90% in 2050 under a net-zero CO₂ emissions route [18].

1.2. Rainwater Harvesting (RWH) System

Rainwater harvesting (RWH) is arguably the most traditional method of providing water [19]. According to Rahman [19] and Zhang et al. [20], RWH is an important tool for meeting the growing demand for water and addressing climate change and variability. RWH is a technique for generating, collecting, storing, and preserving local surface runoff for future usage [21,22]. In addition to representing one of the options for water resources, RWH can lower the price and utilization of water that has been treated [23,24]. The RWH system collects rainfall from natural land surfaces or impervious surfaces such as rooftops, terraces, courtyards, and road surfaces and stores it in tanks, cisterns, and underground dams [25,26].

RWH can be described by numerous techniques for collecting rainfall and distributing it as treated or untreated water that can be suitable for potable or non-potable applications [27]. This can vary from small-scale collection in externally mounted tanks for individual untreated garden supply or treated household consumption to neighborhood-scale collection, treatment, and pressurized distribution systems [27,28,29]. Recently, rooftop RWH has become more significant, particularly in urban and rural regions, due to water scarcity, deteriorating infrastructure, and rising water delivery costs [30,31]. Water that has been gathered and preserved by RWH systems can be utilized for both drinking and non-drinking purposes [32]. Moreover, according to previous studies [32,33,34,35,36,37,38,39,40,41], the additional advantages of RWH systems include lowering water stress, reducing the need for chemicals to treat rainwater for non-potable uses, reducing soil erosion, preventing floods in urban basins, lowering runoff peak flow, extending the lifespan of centralized water distribution infrastructure due to lower demand, increasing the flexibility and robustness of water supply systems, and lowering the amount of electricity utilized by sewage and water supply systems. Additionally, RWH is becoming more popular as a sustainable water alternative to supplement some freshwater requirements, thereby mitigating water scarcity problems to a certain extent, according to Nandi and Gonela [42].

1.3. Solar Energy

Generally, developing and developed countries have accelerated the transition from fossil fuels to renewable energy sources to encourage sustainability and minimize the serious impacts of climate change. Renewable energy, such as solar energy, is the key element of sustainable, environmentally friendly, and cost-effective electricity generation. One of the important natural resources for decarbonization and long-term, sustainable growth of human society is solar energy [43].

Solar energy is abundant, free, and non-polluting. Solar has the greatest potential to produce inexpensive electrical power compared to other renewable energy sources [44,45]. Moreover, PV generation costs have been decreasing quickly as a result of ongoing developments in technology [46]. Solar energy is a clean energy source that is widely available and has been utilized for a variety of purposes [47]. For instance, innovative solar-powered water treatment technologies, such as solar disinfection and solar-powered desalination systems, are crucial for addressing water scarcity in arid regions and remote areas [48,49]. Besides, solar-powered irrigation systems provide an environmentally friendly method of optimizing agricultural water usage. Also, solar pumps for groundwater extraction offer a practical substitute for remote agricultural regions, reducing operating expenses and enhancing food security. Therefore, solar-powered water pumping systems have emerged as an affordable alternative to diesel-powered water pumps, particularly in remote regions, as a result of the rising cost of fossil fuels [50]. Additionally, solar home systems raise living standards in rural areas by supplying consistent electricity for small appliances [51].

Consequently, countries can become more robust to water and energy shortages by using solar energy. Solar systems promote sustainable development by minimizing greenhouse gas emissions and reducing the consumption of fossil fuels (limited resources).

1.4. Importance of the Study

Based on the above literature review, the combination of solar energy and other sustainable methods provides a potential solution for minimizing the effects of climate change and achieving sustainable development goals (SDGs). Also, combining renewable energy, such as solar energy, with rainwater collection systems can be an alternative solution to address these issues (water and energy scarcity) for achieving SDGs. According to the authors’ review of the current study, few studies [52,53,54,55,56] have been designed to evaluate the economic viability of the integration of renewable energy systems (solar or wind) with the RWH system.

Table 1. Summary of previous studies related to the hybrid renewable energy–RWH system.

Reference	Year	Description of the Study	A Schematic of the Proposed System	Main Findings
[52]	2011	Present the techno-economic feasibility of a wind–solar hybrid renewable energy generation system with a rainwater collection feature for electrical energy generation.		A building’s energy consumption can be mainly fulfilled by a developed system, which helps to make the building independent (or partially independent) from the urban electricity grid.
[53]	2012	Design a 3-in-1 wind–solar hybrid renewable energy and rainwater harvester.		The Power-Augmentation-Guide-Vane improved wind turbine performance by 73.2% at low wind speed (3 m/s). Combining solar panels with rainwater collection improved resource efficiency and energy capture.
[54]	2016	Present a technical feasibility study of a hybrid solar–wind–rain eco-roof system with natural ventilation and skylight for electrical energy generation and saving.		The hybrid eco-roof technology increased wind turbine efficiency by implementing a V-shaped roof that accelerated airflow via the Venturi effect. An automated water-based system enhanced solar PV performance by cooling and cleaning. Skylighting and integrated ventilation improved the performance of sustainable buildings, reduced the demand for artificial lighting, and improved interior comfort.
[55]	2019	Conduct a technical, environmental, and economic feasibility study for a hybrid renewable energy harvester system for residential applications.		The developed system can help partially meet a building’s energy needs. Besides, wind turbine performance is significantly enhanced by the hybrid energy harvesting system’s integration of a V-shaped roof guiding vane, which speeds up wind flow.
[56]	2021	Propose a hybrid wind and rainfall energy harvesting system that would be mounted on the sea-crossing bridge’s pipe to capture both types of energy.		By combining piezoelectric and electromagnetic phenomena, the hybrid wind and rainwater energy harvesting achieves efficient dual energy harvesting. The S-rotor and water wheel operate independently, enabling real-time wind energy capture and delayed rainfall energy generation, allowing the sea-crossing bridge’s sensors to operate continuously under a range of weather conditions.

lists previous studies related to the hybrid renewable energy–RWH system. In contrast to previous studies (Table 1), this study does not include wind turbines, but rather focuses solely on the integration of solar and rainwater systems for agricultural households. There is a clear gap in the literature, because no previous work has addressed this particular mix designed to meet the needs of farming households.

The primary objective of this paper is to carry out a technical, environmental, and financial feasibility analysis for a patented hybrid solar–RWH system for irrigation purposes. The proposed system is designed for farming families to increase water and energy self-sufficiency. This system can be installed on the rooftops of residential buildings to save rainwater and generate electricity from renewable sources. In light of the aforementioned, the current study emphasizes three significant issues.

• First, to design an RWH system to meet North Cyprus’s rainfall characteristics. RWH was designed to be utilized for irrigation. The RWH design is based on historical daily rainfall data from 1981 to 2023. Obtaining detailed and suitable rainfall data to have a good design of the RWH system is a challenge in developing countries like Cyprus. Therefore, nine satellite precipitation products were selected for this study’s evaluation of their accuracy in comparison to ground-based gauge rainfall data. Based on the best dataset, finding the distribution function that best fits rainfall data is crucial to designing a successful rainwater harvesting system. It enables accurate projections of both typical and atypical rainfall occurrences, which helps with overflow management and tank size selection. By simulating the frequency of rainfall, the system may be adapted to local patterns, ensuring sufficient capacity for heavy rain while reducing costs. This statistical knowledge supports a reliable, sustainable design that maximizes water availability and system resilience. According to the authors’ review, only a few scientific studies [57,58,59,60,61] have statistically examined the rainfall time series in several Cyprus regions. Michaelides et al. [57] utilized a gamma distribution function to investigate the characteristics of Cyprus’s annual rainfall frequency distribution. Stamatatou et al. [58] examined the characteristics of the annual maximum rainfall depth and storm duration at the Limassol station in Cyprus employing a variety of distribution functions, including generalized extreme value (GEV), Gumbel, and generalized Pareto, gamma, exponential, and log-normal. The findings showed that the generalized extreme value distribution was chosen to describe the storm duration and yearly maximum rainfall depth at the chosen station. Zaifoglu et al. [59] conducted a regional frequency analysis of the annual maximum daily precipitation in Northern Cyprus using traditional cluster analysis and time series clustering techniques. Generalized logistic, generalized normal, and Pearson Type III distributions were determined to be the best fits for the various Northern Cyprus subregions. Besides, in the Güzelyurt region of Northern Cyprus, Kassem et al. [60] determined the best distribution model for monthly and total rainfall. The findings demonstrated that the Nakagami and Wakeby distributions provided the best fit for the actual total and monthly rainfall data. Furthermore, the best-fit probability distribution for monthly rainfall in seven Northern Cyprus locations was identified by Kassem et al. [61]. The findings show that the Gumbel Max best distribution and the generalized extreme value distribution performed effectively in examining the characteristics of average rainfall data. Therefore, 65 frequency distributions and three goodness-of-fit test statistics were applied in this study to determine the best-fit probability distributions in the case of average and maximum daily rainfall.
• Second, to evaluate the potential of water savings from the use of the solar system for irrigation purposes and generating electricity for the building.
• Third, to evaluate the technical performance, environmental impact, and economic feasibility of patented hybrid systems using RETScreen Expert software (version 9.1, 2023) to provide a comprehensive understanding of the system’s performance and financial feasibility.

2. Materials and Methods

2.1. Study Area

Cyprus is an island in the Mediterranean Sea, with a total land area of 9251 km² and a coastline of 1364 km² [62]. Cyprus is an island that is mostly dominated by the Troodos and Kyrenia Mountains. About half of the island is covered by the 1952-m-tall Troodos Mountain in the west and south [63]. In comparison to Troodos, Kyrenia Mountain is smaller in size and has a narrower shape, since it extends along the northern coastline [64]. Due to its limited surface water resources, Cyprus relies primarily on groundwater as its main source for domestic supply, industries, and crop production [65,66]. Cyprus has a Mediterranean climate, which is defined by hot, dry summers and cool, rainy winters. Rainfall primarily occurs from October through April [67]. The highest rainfall intensity was recorded on the Besparmak Mountain and the North Sea side [68]. Moreover, the average temperature and relative humidity are within the range of 12.9–21.3 °C and 50–70%, respectively [69]. Besides, the global horizontal irradiation and wind speed vary from 4.80 kWh/m² to 5.44 kWh/m² and from 2.0 m/s to 9.3 m/s, respectively, according to the global atlas map.

The study’s chosen location is a household-farm unit situated in Gönyeli, Lefkoşa, Northern Cyprus. The monthly variation in electricity consumption and cost for one year (2024) is illustrated in Figure 1. It was found that the average electricity consumption of the chosen location is 664 kWh, ranging from 562 kWh (March) to 741 kWh (August). Furthermore, according to the data, the cost of electricity for energy demand ranges from USD 140 to 185. It should be noted that electricity consumption and costs were collected from the bills provided by the electricity company.

The farm cultivates water-demanding summer crops, including tomatoes, peppers, and cucumbers, on an area that is about 1000 m² in size. Particularly during Northern Cyprus’s hottest summer months of July and August, when evapotranspiration is at its maximum, they need heavy and frequent irrigation. In this study, the amount of water and cost were obtained from the monthly bills issued by the local water supplier. Therefore, the monthly water requirements for the selected site were determined for May to October 2024, as shown in Figure 2. They are particularly important during these months because the family farm relies on water for irrigation. Figure 2 illustrates the gradual increase in irrigation level from May to a peak of over 57,000 L in July and August, after which it drops to roughly 12,000 L in October. The pattern correctly depicts the crop’s growth cycle and the climate season, with water requirement rising during the droughty summer months when the crop is fully developed. In addition to vegetable crops, there are also fruit trees such as grapevines and orange, lemon, and olive trees, which increase the need for water, especially when they are in fruit. These fruit trees and summer crops are grown to form a diverse farming system that requires proper water management. In an effort to increase productivity and decrease waste, the irrigation system utilized is probably a drip or hose system that offers spot watering. Additionally, a tiny chicken house is located on the farm, indicating a mixed-use, integrated farming operation. This offers benefits in dung for fertilizer and increases a sustainable approach through the inclusion of crop agriculture, fruit development, and chicken farming.

2.2. Dataset

2.2.1. Ground-Based Gauge Rainfall Data

The monthly measurement of rainfall data was collected from the Meteorological Department located in Lefkoşa, Northern Cyprus, from 1981 to 2023. The rain gauge was used to measure the rainfall at a height of 0.3 m above the ground level. The specifications of the rain gauge are listed in

Table 2. Specifications of the rain gauge.

Property	Description/Value
Sensor/transducer type	Tipping bucket/reed switch
Precipitation type	Liquid
Accuracy	±2%
Sensitivity	0.2 mm
Closure time	<100 ms (for 0.2 mm of rain)
Capacity	Unlimited
Funnel diameter	225 mm
Standard	400 cm2
With the expander unit	1000 cm2
Max. current rating	500 mA
Breakdown voltage	400 VDC
Capacity open contacts	0.2 pF
Life (operations)	108 closures
Material	Non-corrosive aluminum alloy LM25
Dimensions	390 (h) × 300 (Ø) mm
Weight	2.5 kg
Temperature range (operating)	0$\pm$85 °C

.

2.2.2. Satellite-Based Rainfall Data

In Northern Cyprus, the number of stations and data is limited. This is considered a major problem for working with the observed rainfall records. Thus, it is necessary to evaluate satellite precipitation products (SPPs) with local rain gauge data before utilizing these products as a source for understanding the rainfall distribution in Northern Cyprus. SPPs are based on data derived from various sensors and satellites [70,71,72]. They can also include other data sources, such as ground radar, gauge networks, or forecasts from models or reanalysis [64,65,66]. Therefore, nine SPPs were selected for this study’s evaluation of their accuracy in comparison to ground-based gauge rainfall data because of their high spatial resolution, coverage domain, and availability periods, as shown in

Table 3. Summary of specifications of satellite precipitation products.

Product	Description/Full Name of the Dataset	Resolution	Period
ERA5	Fifth-generation reanalysis product of the European Centre for Medium-Range Weather Forecasts	0.05°/1 d	1979–present
ERA5-Land	ERA5-Land has been produced by replaying the land component of the ECMWF ERA5 climate reanalysis	0.125° $\times$ 0.125°	1963–present
ERA5 -Ag	Agriculture-specific dataset of the ECMWF ERA5	0.1°$\times$0.1°	1979–present
MERRA-2	Second-generation Modern-ERA Retrospective Analysis for Research and Applications	0.5° $\times$ 0.625°	1980–present
TerraClimate	Global gridded dataset of meteorological and water balance for global terrestrial surfaces	0.042$°\times$0.042°	1958–present
CHIRPS	A new land-only climatic database for precipitation, obtained through the Climate Hazards Group of the University of California at Santa Barbara (UCSB)	0.05°$\times$0.05°	1981–present
CFSR	NCEP (NOAA NWS National Centers for Environmental Prediction) Climate Forecast System Reanalysis dataset	1/5°	1979–present
CPC-CMORPH	The Satellite Precipitation—CMORPH Climate Data Record (CDR) consists of satellite precipitation estimates that have been bias-corrected and reprocessed using the Climate Prediction Center (CPC) Morphing Technique (MORPH) to form a global, high-resolution precipitation analysis	0.5°$\times$0.5°	1998–present
CPC-UPP	This data set is part of the products suite from the Climate Prediction Center (CPC) Unified Precipitation Project (UPP) that is underway at NOAA CPC	0.5°$\times$0.5°	1979–present

.

It should be noted that the study used 43 years (1981–2023) of monthly rainfall measurements from the Meteorological Department of Lefkoşa, North Cyprus, as a reference dataset. The accuracy and reliability of satellite-based sources were evaluated by comparing measured data with these databases to determine the most reliable and representative satellite source. The selected satellite dataset was then used to estimate the potential of the rainwater harvesting system and ensure water availability.

Various statistical metrics, including Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Coefficient of Determination (R²), are used to evaluate the performance of these datasets.

$$R^2=1-{\displaystyle \frac{\sum _{i=1}^n{\left(a_{a,i}-a_{p,i}\right)}^2}{\sum _{i=1}^n{\left(a_{p,i}-a_{a,ave}\right)}^2}}$$

(1)

$$RMSE=\sqrt{{\displaystyle \frac{1}{n}}\sum _{i=1}^n{\left(a_{a,i}-a_{p,i}\right)}^2}$$

(2)

$$MAE={\displaystyle \frac{1}{n}}\sum _{i=1}^n\left|a_{a,i}-a_{p,i}\right|$$

(3)

where$n$ is the number of data, $a_{p,i}$ is the predicted value, $a_{a,i}$ is the actual value, $a_{a, ave}$ is the average actual value, and i is the number of input variables.

2.2.3. Solar Radiation Dataset

Designing solar energy systems in a place lacking measuring data is challenging. Additionally, Northern Cyprus has an insufficient number of stations for measuring the data. Moreover, according to the literature [73,74,75], NASA, the leader in space technology, offers satellite and model-based products through its network of satellite systems, which are needed to investigate climate dynamics, improve energy efficiency, and meet agricultural needs. The POWER (Prediction of Worldwide Energy Resources) Project of the NASA Earth Science Research Program may have enabled this. This global grid provides solar and meteorological data from 1981 onwards and covers regions lacking surface weather observations with a geographical resolution of 0.5° × 0.5° [73,74,75]. Several studies have utilized the NASA database to assess the potential for solar energy in various locations [76,77,78]. Furthermore, previous studies [79,80,81] concluded that there is a strong correlation between the NASA database and observed data (global solar irradiation). Bhatia also argued that by disregarding the impact of cloud absorption, solar radiation can be presumed to be constant across wide regions [82]. Therefore, the NASA POWER dataset was used to gather monthly solar irradiation for the selected location for the period 1982–2022.

2.3. System Design

2.3.1. Shading Distance Calculations

The PV arrays’ output can be decreased by shading. PV arrays should be positioned in an area that lacks shade for that purpose. Nonetheless, grid-connected systems are typically found in cities, and the modules are typically mounted on roofs, where some shade is occasionally unavoidable. To position as many panels as possible in a specific area and to optimize the solar installation’s performance. Orientation and row spacing, often known as shading distance, are two primary elements that are frequently taken into account.

To compute the shading distance, as shown in Figure 3, Equation (4) is used [83,84].

$$X=Y{\displaystyle \frac{cos\gamma }{tan\alpha }}$$

(4)

$$\alpha =arcsin\left(sin\left(\delta \right)·sin\left(\varphi \right)+cos\left(\delta \right)·cos\left(\varphi \right)·cos\left(\omega \right)\right)$$

(5)

$$\delta =23.44°·sin\left({\displaystyle \frac{360}{365}}·\left(n+10\right)\right)$$

(6)

$$\omega =15\left(12-LAT\right)$$

(7)

$$\begin{array}{ll}LAT=standard & tim \left(clock time\right)\\ & \pm 4\left(standard time longitude-longitude of location\right)\\ & +EOT\end{array}$$

(8)

$$\begin{gathered} EOT=229.18\left(0.000075+0.001868·cos\left({\displaystyle \frac{360·\left(n-1\right)}{365}}\right)-0.032077 \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}·sin\left({\displaystyle \frac{360·\left(n-1\right)}{365}}\right)-0.014615·cos2\left({\displaystyle \frac{360·\left(n-1\right)}{365}}\right) \\ -0.04089·sin2\left({\displaystyle \frac{360·\left(n-1\right)}{365}}\right)\right) \end{gathered}$$

(9)

where $X$ is the shading distance [m], $Y$ is the tilted PV panel height [m], $\gamma$ is the azimuth angle [$°$],$\alpha$ is the solar altitude at cerain solar time [$°$],$\delta$ is the solar declintion, $\varphi$ is the latitude of the location, $\omega$ is the solar hour angle, and $n$ is the day of the year.

It should be noted that the first correction for $LAT$ applies to the eastern hemisphere with a negative sign, while the western hemisphere applies to the positive sign.

Considering that Jinko Solar’s mono-Si JKM360M-66H module, with dimensions of 1841 × 1002 × 30 mm, an efficiency of 19.52%, and capacity of 360 W, is an efficient PV module that is currently on the market, it was chosen for the proposed system. The specifications of the selected module are listed in

Table 4. Specifications of the selected module.

Specification	Value
Mechanical Characteristics
Module type	JKM360M-66H
Cell type	Mono PERC 158.75 × 158.75mm
Number of cells	132 (6 × 22)
Weight	20.0 kg
Electrical Characteristics	STC	NOCT
Maximum power (Pmax)	360 Wp	268 Wp
Maximum power voltage (Vmp)	36.97 V	33.78 V
Maximum power current (Imp)	9.74 A	7.93 A
Open-circuit voltage (Voc)	43.58 V	41.04 V
Short-circuit current (Isc)	10.48 A	8.46 A
Module efficiency STC (%)	19.52%
Operating temperature (°C)	−40 °C~+85 °C
Temperature coefficients of Pmax	−0.35%/°C
Temperature coefficients of Voc	−0.28%/°C
Temperature coefficients of Isc	0.048%/°C
Nominal operating cell temperature (NOCT)	45 ± 2 °C

STC: standard test conditions; NOCT: nominal operating cell temperature.

.

Using Equations (2)–(6), the mean solar altitude at the selected locations was found to be 47.16° at solar noon. The values of shading distance with various tilt angles are listed in

Table 5. Values of shading distance with various tilt angles.

Tilt Angle [°]	Shading Distance [M]
10	0.3
20	0.6
30	0.9
40	1.1
50	1.3
60	1.5
70	1.6
80	1.7

.

2.3.2. Description of System

Figure 4 displays a clever dual-use system that combines solar energy generation and rainwater collection on the rooftop of a residential building. In design D#1, the panels are mounted on an elevated platform and organized in rows with a slope. In order to get the most solar energy during the day, this arrangement is oriented to face the sun. To increase the overall effectiveness of the system, spaces are also provided between the rows so that the adjacent panels cannot shadow them. As an alternative, the second configuration (D#2) has a design in which the panels create a continuous layer without any interstices, bringing the entire roof surface to a slightly sloping condition. Theoretically, this method can accommodate more panels and produce more energy overall, but it may lead to less-than-ideal tilt angles, which could lower each panel’s performance in comparison to a tilted array.

Utilizing photovoltaic (PV) panels to generate electricity in conjunction with a rainwater collection system effectively utilizes renewable resources. With reduced energy use and the ability to collect water for irrigation, this technology will contribute to thermal gain, enhance indoor comfort, and promote sustainability. Solar panels are mounted on the rooftop of a residential building to gather sunlight and generate electricity. The system is grid-connected, meaning that the solar-generated electricity is directly used by the household and irrigation equipment, and excess energy is routed back into the public electrical grid. When solar power is not enough, electricity is drawn from the grid. This approach eliminates the requirement for on-site electrical energy accumulation and simplifies the system architecture. In addition to generating power, the panels serve as a roof covering, providing shade and reducing heat gain in the summer to improve the indoor environment. The structure of the system distributes rainwater flowing on the solar panels into particular channels for collection.

As shown in Figure 4, the rainwater collected in square channels at the base of the PV modules emerges from the solar panels. These channels are designed to efficiently gather and channel water so that it flows toward a central outlet without overflowing. The channels help reduce water waste by channeling all collected rainwater into a connected piping system. Moreover, a dual-purpose assembly framework supports both PV modules and rainfall collection channels. In order to promote effective water movement, the height of this assembly structure changes by 5 cm, providing a slight slope that focuses water flow on a single spot. This smooth incline minimizes water stagnation and increases collection efficiency by concentrating water from all the channels into one single pipe.

Rainwater collected is transferred via a series of PVC pipes to the storage tank of the system. This water storage tank ensures the efficient and secure storage of the collected rainwater for future use. In regions where water availability varies, this storage device becomes essential for water conservation initiatives, helping to reduce dependency on outside water sources. Besides, rainwater can be utilized to irrigate gardens, plants, and crops through the use of PVC pipes that connect the storage tank to an irrigation system. By using less potable water for irrigation, this method promotes sustainable water management, particularly in areas with limited water resources.

This integrated system provides numerous benefits to residential buildings. Solar panels reduce electricity use, while the rainwater collection system ensures a consistent supply of water for irrigation needs. The system uses less water, reduces heat gain, and produces renewable energy simultaneously, thus providing an attractive option for both urban and rural settings. Figure 5 shows the advantages and disadvantages of both designs (D#1 and D#2).

2.4. Analyzing the Distribution of Rainfall

Accurate modeling and evaluation of the empirical rainfall distributions are crucial initial steps toward obtaining extensive knowledge of rainfall patterns and distribution. For this purpose, probability distribution functions (PDFs) can be used and applied to study the distribution of rainfall at different locations around the globe [85,86]. PDFs can be classified as either discrete or continuous distributions, contingent upon whether they define probabilities associated with discrete variables or continuous variables [87]. In addition, continuous distribution functions consist of four classifications: bounded, unbounded, non-negative, and advanced [87]. Numerous scientific studies have been conducted worldwide on the selection of probability distributions in rainfall frequency analyses [88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118]. For instance, Parchure and Gedam [85] determined the best-fit probability distribution for extreme storms using two-parameter and three-parameter distribution functions. The results indicated that a three-parameter generalized extreme value was considered the best distribution function for studying the extreme storm series in the Mumbai region, India. Yuan et al. [111] utilized the Expanded Automated Meteorological Data Acquisition System (EA) weather data of 20 years (1981–2000) to analyze the annual maximum hourly rainfall characteristics for 15 locations in Japan. The results showed that the Log-Pearson type 3 distribution provided the best fit to the actual data for most locations in Japan. Kassem and Gökçekuş [116] utilized four distribution functions to analyze the characteristics of rainfall in the Beirut region, Lebanon, using daily rainfall of 25 years (1991–2015). The results showed that Gumbel maximum and logistic distributions were able to provide the best fit to the actual data for the selected region.

Based on the above, in the current analysis, in order to determine the optimum statistical distribution, which best describes the characteristics of rainfall of the studied site, a comprehensive range of probability distribution functions, including 65 models covering all continuous types and discrete types of PDFs outlined in the flow diagram of Figure 6, are evaluated.

In this study, the Kolmogorov–Smirnov (K–S), Anderson–Darling (A–D), and chi-squared (C–s) tests are employed for evaluating the performance of the selected distributions. The mathematical expressions for these tests are provided below.

Kolmogorov–Smirnov (K–S)

$$D=\underset{1\le i\le n}{\max }\left(F\left(x_i\right)-{\displaystyle \frac{i-1}{n}}, {\displaystyle \frac{i}{n}}-F\left(x_i\right)\right)$$

(10)

$$F_n\left(x\right)={\displaystyle \frac{1}{n}}\times \left(Number of observation\le x\right)$$

(11)

Anderson–Darling (A–D)

$$A^2=-n-{\displaystyle \frac{1}{n}}\sum _{i=1}^n\left(2i-1\right)\times \left[ln{F}_X\left(x_i\right)+ln\left(1-F_X\left(x_{n-i+1}\right)\right)\right]$$

(12)

where $F_X\left(x_i\right)$ is the cumulative distribution function of the proposed distribution at $x_i$, for i = 1, 2, …, n.

Chi-squared (C-s)

$${\mathcal{X}}^2=\sum _{i=1}^k{\displaystyle \frac{{\left(O_i-E_i\right)}^2}{E_i}}$$

(13)

$$E_i=F\left(x_2\right)-F\left(x_1\right)$$

(14)

where${ O}_i$ is the observed frequency for bin i, $F$ is the cumulative distribution function of the probability distribution being tested, and $x_1$, $x_2$ are the limits for bin i.

Furthermore, one of the simplest and most effective methods for choosing the best-fit model is to employ a graphical test. The P-P and Q-Q plots of the observed and estimated values are displayed in this study.

2.5. Assessment of Rainwater Harvesting Potential

The potential volume of rainwater ($PVR$) available for harvesting is calculated using Equation (15) [119].

$$PVR={\displaystyle \frac{R\times \left(A_{PV }+A_{chennel}\right)\times C}{1000}}$$

(15)

$$A_{PV }=Number of module\times Araa of one module$$

(16)

$$\begin{gathered} A_{chennel}=Width of the squared channel\times Height of the channel \\ \times Length of the channel \end{gathered}$$

(17)

where $R$ is the amount of rainfall, $A$ is the rooftop area, and $C$ is the runoff coefficient.$C$is considered to be 0.85, since the PV panels are well-designed to optimize runoff.

Moreover, the potential volume of harvested rainfall is then divided by the yearly domestic demand using the following formula to determine the potential saving percentage (PSP), as given below [119].

$$PSP={\displaystyle \frac{PVR}{total annual water used }}$$

(18)

2.6. Economic Model

The performance of renewable power plants can be estimated using a variety of simulation tools, such as the RETScreen simulation tool. Natural Resources Canada developed the RETScreen, which is a widely used tool for evaluating the feasibility of different renewable energy systems [120]. Assessing the risk, expenses, economic feasibility, emissions reduction, energy production, and savings of various renewable energy options constitutes the goal. For a specific region, the RETScreen retrieves meteorological data from long-term monthly average meteorological data in the National Aeronautics and Space Administration (NASA) database [121]. It is an effective and rapid method to evaluate the renewable energy system’s energy efficiency and environmental impact reduction. Therefore, this study utilizes RETScreen software to evaluate the technical, financial, and environmental aspects of the proposed systems. Financial analysis and evaluation are essential to provide investors, policymakers, and the government with a precise understanding of the feasibility of grid-connected solar projects. In order to assess the profitability of the proposed PV power plant and the returns on investment, this study employed a variety of important economic metrics, including net present value (NPV), levelized cost of energy (LCOE), simple payback (SP), equity payback (EP), annual life cycle savings (ALCS), annual greenhouse gas emissions (A-GHG), the cost of reducing GHG emissions (GHG-E-RC), and benefit–cost ratio (BCR). The mathematical expressions of the capacity factor (CF) and important economic metrics are given below.

$$CF={\displaystyle \frac{P_{out}}{P\times 8760}}$$

(19)

$$NPV=\sum _{n=0}^N{\displaystyle \frac{C_n}{{\left(1+r\right)}^n}}$$

(20)

$$LCOE={\displaystyle \frac{sum of costs over the lifetime}{ of electricity generated}}$$

(21)

$$SP={\displaystyle \frac{C-IG}{\left(C_{ener}+C_{capa}+C_{RE}+C_{GHG}\right)-\left(C_{o\&M}+C_{fuel}\right)}}$$

(22)

$$EP=\sum _{n=0}^N{C}_n$$

(23)

$$ALCS={\displaystyle \frac{NPV}{\frac{1}{r}\left(1-\frac{1}{{\left(1+r\right)}^N}\right)}}$$

(24)

$$B-C={\displaystyle \frac{NPV+\left(1-f_d\right)C}{\left(1-f_d\right)C}}$$

(25)

$$\begin{gathered} A-GHG=\left[\left(Base case GHG emission factor\right) \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}-\left(Proposed case GHG emission factor\right)\right] \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\times End use energy delivered \end{gathered}$$

(26)

$$\mathrm{G}\mathrm{H}\mathrm{G}-\mathrm{E}-\mathrm{R}\mathrm{C}={\displaystyle \frac{ALCS}{{∆}_{GHG}}}$$

(27)

where $P_{out}$ is energy generated per year,$P$ is installed capacity, N is the project life in years, $C_n$ is the after-tax cash flow in year n, and r is the discount rate. C is the total initial cost of the project, $f_d$ is the debt ratio, B is the total benefit of the project, $IG$ is the incentives and grants, and $C_{ener}$ is the annual energy savings or income. $C_{capa}$ is the annual capacity savings or income, $C_{RE}$ is the annual renewable energy (RE) production credit income, $C_{GHG}$ is the GHG reduction income, and $C_{o\&M}$ is the yearly operation and maintenance costs incurred by the clean energy project. $C_{fuel}$ is the annual cost of fuel, which is zero for renewable projects, and ${∆}_{GHG}$ is the annual GHG emission reduction.

Moreover,

Table 6. Economic and financial parameters used in this study.

Parameter	Unit	Value
PV module cost	USD/Watt	0.30
The lifetime of the PV module	Year	25
Cost of the inverter	USD	2600
Cost of plastic tank	USD	792
Support structure for solar panel	USD/kW	150
Rainwater collection channel	USD/m2	25
Miscellaneous/contingency fund	% of the total initial cost	3
Installation and spare parts	% of the total initial cost	8.6
The lifetime of the inverter	Year	10
Feasibility study, development, and engineering cost	% of the total initial cost	0.6
Inverter replacement periodic cost	Every ten years	Equal to the inverter’s cost
Inflation rate	%	8
Discount rate	%	6
Project life	Year	25
Energy cost increase rate	%	5
Reinvestment rate	%	9
Debt ratio	%	70
Debt interest rate	%	0
Debt term	Year	20

summarizes the economic parameters that were used in the RETscreen financial study. Table 6 also displays the cost of the developed system. It should be mentioned that the system’s financial and economic parameters are based on current market data and are in line with the cost prices found in the literature.

3. Results and Discussion

3.1. Comparison of Nine Satellite Rainfall Products

The monthly gauge observation from Lefkoşa, Northern Cyprus, was used to assess the accuracy of the nine global precipitation datasets (see Table 3). Data regarding gauge observations were gathered between 1981 and 2023.

Table 7. Performance of global precipitation datasets.

Dataset	Monthly			Seasonally
	R-Squared	RMSE [mm]	MAE [mm]	R-Squared	RMSE [mm]	MAE [mm]
CHIRPS	0.91	19.06	11.90	0.99	47.72	33.62
CFSR	0.89	19.67	16.57	0.97	56.49	49.71
ERA5 LAND	0.93	16.56	13.72	0.99	46.95	41.16
ERA5	0.95	3.81	3.23	1.00	2.87	2.60
ERA5-Ag	0.94	6.50	4.89	0.99	15.75	14.66
MERRA2	0.91	8.22	6.28	0.96	20.03	16.54
CPC-CMORPH	0.81	14.70	12.36	0.90	40.92	37.08
CPC-UPP	0.74	11.02	9.01	0.79	28.20	23.49
TerraClimate	0.92	13.74	8.89	0.94	36.59	24.06

displays the statistical metrics employed for evaluating the performance of global precipitation datasets. The R² values were found to be high, with monthly rainfall datasets ranging from 0.74 to 0.95 and seasonal rainfall datasets ranging from 0.79 to 1.00. R2’s highest and lowest values were identified for ERA5 and CPC-UPP, respectively. According to Santhi et al. [122] and Van et al. [123], an R-squared value greater than 0.5 is acceptable. All rainfall products demonstrated excellent performance and accuracy for monthly and seasonal rainfall analysis based on R² values. Moreover, the results indicate that the EAR5 produced the lowest values of RMSE (3.81 mm) and MAE (3.23 mm), followed by ERA5-Ag (RMSE = 6.50 mm and MAE = 4.89 mm) based on the monthly rainfall dataset. Besides, the seasonal analysis showed that EAR5 had better performance compared to the other datasets.

According to the results, ERA5 performed the best compared to the other datasets. Previous scientific studies [124,125,126,127] support the results presented in this study. For instance, Kokkalis et al. [124] found that ERA5 is a promising alternative data source, especially in areas with limited weather station coverage, as it has demonstrated high predictive ability under both frontal and convective precipitation conditions. Salih et al. [126] concluded that ERA5 was highly accurate in predicting the Tensift basin’s extreme precipitation. Hassler and Lauer [127] concluded that ERA5 has a good correlation with observed rainfall data in the South Asian and Central European monsoon regions, but underestimates very low rainfall rates in the tropics. According to Tarek et al. [128] and Lavers et al. [129], ERA5’s high geographical resolution of 24 km and daily interval temporal precision allow it to better capture localized precipitation events than poorer datasets. This resolution is particularly important for accurately simulating convective rainfall, as it is often missed by lower-resolution models. According to previous studies [130,131,132,133], high-resolution datasets are necessary for more accurately capturing extreme rainfall occurrences, which are critical for hydrological applications and agricultural planning. ERA5 is based on atmospheric physics, modeling precipitation processes using fluid dynamics and thermodynamics. This model’s resilience is demonstrated by its capacity to capture the basic mechanics of rainfall generation. Besides, Dee et al. [134] pointed out that reanalysis models that effectively reproduce atmospheric processes yield more accurate precipitation outputs. This physical basis ensures that rainfall estimates are statistically correlated with observations and are based on sound meteorological principles.

Furthermore, the monthly and seasonal variations in rainfall for the observations and global precipitation datasets for the study’s chosen location are depicted in Figure 7 and Figure 8, respectively. It can be seen that ERA5 and ERA5-Ag rainfall are generally similar to the observation data.

3.2. Characteristics of Rainfall in the Selected Location

As mentioned above, the number of stations and data are limited. This is considered a major problem for working with the observed rainfall records in Northern Cyprus. Based on the results (Section 3.1), EAR5 has given the highest value of R-squared and the lowest value of RMSE and MAE. Therefore, ERA5 reanalysis was employed in this study and proposed as an alternative dataset for developing RWH systems at the selected location. The data was gathered for the period 2000–2023. The average and maximum daily variations in rainfall are illustrated in Figure 9 and Figure 10, respectively. It was found that the maximum and lowest daily rainfall values, 4.35 mm and 0.00035 mm, respectively, occurred in December and August (see Figure 9). Moreover, it was found that the highest and lowest daily rainfall occurred in October 2012 and August 2004, with values of 48.89 mm and 0.0061 mm, respectively, as shown in Figure 10. It should be noted that the maximum value of rainfall for each month between 2000 and 2023 was utilized as the parameter for analyzing the location’s extreme rainfall estimation. The amount of maximum daily/monthly rainfall influences design features such as tank size, filtering needs, and so on. Besides, the maximum monthly rainfall data over long periods helps ensure that storage is appropriate for peak usage [135]. Imteaz et al. [135] concluded that size based on extremes enhances water availability during dry periods.

Additionally, as illustrated in Figure 11, the highest monthly rainfall value was recorded in January (67.76 mm for average monthly and 483.43 mm for maximum monthly rainfall data), while the lowest value was recorded in July (1.37 mm and 24.67 mm for average and maximum monthly rainfall, respectively).

3.2.1. Selecting the Best-Fit Results for Average Daily and Seasonally Rainfall

The maximum likelihood approach was utilized to compute the distribution parameters based on the mean daily rainfall. Anderson-Darling (A-D), chi-squared (C-s), and Kolmogorov-Smirnov (K-S) tests were used to determine which of the 65 distribution functions for the chosen location had the best distribution.

Table 8. The best distribution function for average daily rainfall (Rank #1).

Month	DF	Test	Classification	Parameter
Jan	Johnson SB	KS	Continuous-Bounded	γ = 0.18753 δ = 0.92797 λ = 4.6823 ξ = 0.03519
	Johnson SB	AD	Continuous-Bounded	γ = 0.18753 δ = 0.92797 λ = 4.6823 ξ = 0.03519
	Normal	C-s	Continuous-Unbounded	σ = 1.0229 μ = 2.1858
Feb	Dagum (4P)	KS	Continuous-Non-negative	k = 0.30979 α = 5.685 β = 1.9334 γ = 0.32815
	Dagum (4P)	AD	Continuous-Non-negative	k = 0.30979 α = 5.685 β = 1.9334 γ = 0.32815
	Levy	C-s	Continuous-Non-negative	σ = 1.4103
Mar	Log-Logistic (3P)	KS	Continuous-Non-negative	α = 2.3899 β = 0.90586 γ = 0.04141
	Wakeby	AD	Continuous-Advanced	α = 1.6205 β = 3.0465 γ = 0.51091 δ = 0.19898 ξ = 0.13324
	Log-Logistic (3P)	C-s	Continuous-Non-negative	α = 2.3899 β = 0.90586 γ = 0.04141
Apr	Inv. Gaussian (3P)	KS	Continuous-Non-negative	λ = 3.6049 μ = 1.1232 γ = −0.1355
	Wakeby	AD	Continuous-Advanced	α = 8.3601 β = 26.197 γ = 0.75358 δ = −0.10774 ξ = 0
	Weibull (3P)	C-s	Continuous-Non-negative	α = 1.3339 β = 0.88001 γ = 0.17673
May	Wakeby	KS	Continuous-Advanced	α = 1.4478 β = 0.37189 γ = 0 δ = 0 ξ = −0.01538
	Wakeby	AD	Continuous-Advanced	α = 1.4478 β = 0.37189 γ = 0 δ = 0 ξ = −0.01538
	Johnson SB	C-s	Continuous-Bounded	γ = 0.91051 δ = 0.89552 λ = 3.9419 ξ = −0.16837
Jun	Johnson SB	KS	Continuous-Bounded	γ = 1.572 δ = 0.91146 λ = 1.534 ξ = −0.02936
	Wakeby	AD	Continuous-Advanced	α = 0.29902 β = 0.10448 γ = 0 δ = 0 ξ = 9.5636 × 10−4
	Erlang (3P)	C-s	Continuous-Non-negative	m = 1 β = 0.25206 γ = 0.0135
Jul	Log-Pearson 3	KS	Continuous-Advanced	α = 18.887 β = −0.35822 γ = 2.7703
	Phased Bi-Weibull	AD	Continuous-Advanced	α1 = 0.99972 β1 = 0.01554 γ1 = 0 α2 = 0.67204 β2 = 0.03706 γ2 = 0.00261
	Log-Logistic	C-s	Continuous-Non-negative	α = 1.0305 β = 0.01698
Aug	Log-Pearson 3	KS	Continuous-Advanced	α = 2.2362 β = −1.0863 γ = −0.67154
	Dagum	AD	Continuous-Non-negative	k = 0.21083 α = 3.154 β = 0.18395
	Gen. Logistic	C-s	Continuous-Advanced	k = 0.33186 σ = 0.04097 μ = 0.06905
Sep	Lognormal	KS	Continuous-Non-negative	σ = 0.79939 μ = −1.057
	Lognormal (3P)	AD	Continuous-Non-negative	σ = 0.77923 μ = −1.0317 γ = −0.00649
	Wakeby	C-s	Continuous-Advanced	α = 2.4861 β = 24.442 γ = 0.32909 δ = 0.11924 ξ = 0
Oct	Johnson SB	KS	Continuous-Bounded	γ = 0.42553 δ = 0.64897 λ = 2.7761 ξ = 0.09971
	Johnson SB	AD	Continuous-Bounded	γ = 0.42553 δ = 0.64897 λ = 2.7761 ξ = 0.09971
	Weibull	C-s	Continuous-Non-negative	α = 1.505 β = 1.2768
Nov	Burr (4P)	KS	Continuous-Non-negative	k = 410.35 α = 2.2025 β = 19.348 γ = 0.10019
	Gen. Extreme Value	AD	Continuous-Advanced	k = −0.16419 σ = 0.51505 μ = 0.99221
	Dagum	C-s	Continuous-Non-negative	k = 0.30935 α = 6.5299 β = 1.6691
Dec	Wakeby	KS	Continuous-Non-negative	α = 15.39 β = 14.166 γ = 0.89436 δ = −0.07454 ξ = 0.03718
	Wakeby	AD	Continuous-Advanced	α = 15.39 β = 14.166 γ = 0.89436 δ = −0.07454 ξ = 0.03718
	Wakeby	C-s	Continuous-Advanced	α = 15.39 β = 14.166 γ = 0.89436 δ = −0.07454 ξ = 0.03718

and

Table 9. The best distribution function for seasonally daily rainfall (Rank #1).

Season	DF	Test	Classification	Parameter
Winter	Error	KS	Continuous-Unbounded	k = 4.0395 σ = 1.511 μ = 5.6338
	Johnson SB	AD	Continuous-Bounded	γ = 0.23375 δ = 1.004 λ = 7.3287 ξ = 2.3215
	Gen. Extreme Value	C-s	Continuous-Advanced	k = −0.20079 σ = 1.4778 μ = 5.03
Spring	Johnson SB	KS	Continuous-Bounded	γ = 2.0689 δ = 2.1565 λ = 15.213 ξ = −1.1827
	Johnson SB	AD	Continuous-Bounded	γ = 2.0689 δ = 2.1565 λ = 15.213 ξ = −1.1827
	Triangular	C-s	Continuous-Bounded	m = 1.9374 a = 0.47754 b = 7.2821
Summer	Log-Pearson 3	KS	Continuous-Advanced	α = 20.475 β = −0.18999 γ = 2.6709
	Wakeby	AD	Continuous-Advanced	α = 0.47551 β = 0.27013 γ = 0 δ = 0 ξ = 0.02749
	Fatigue Life (3P)	C-s	Continuous-Non-negative	α = 0.77395 β = 0.33182 γ = −0.03038
Autumn	Wakeby	KS	Continuous-Advanced	α = 3.2617 β = 1.0513 γ = 0.07273 δ = 0.53543 ξ = 1.0616
	Wakeby	AD	Continuous-Advanced	α = 3.2617 β = 1.0513 γ = 0.07273 δ = 0.53543 ξ = 1.0616
	Pearson 6 (4P)	C-s	Continuous-Non-negative	α1 = 6.2001 α2 = 133.47 β = 56.203 γ = 0.17101

list the best distribution function for average and seasonally daily rainfall, respectively, based on the K-S, A-D, and C-s tests. The best-fit result is defined as the smallest goodness-of-fit result among all the probability distributions developed.

It was found that the Johnson SB distribution was selected as the best fit in January based on both the Anderson-Darling (AD) and Kolmogorov-Smirnov (KS) tests, classifying it as a continuous-bounded distribution, while the normal distribution was the best fit according to the chi-squared (CS) test, classifying it as a continuous-unbounded distribution. When non-negative distributions were seen in February and March, it was found that the Dagum (4P) and Log-Logistic (3P) distributions suited the data the best. Additionally, the Wakeby distribution, which falls under the continuous-advanced category, was frequently suitable, particularly from April to December. It was often mixed with other deliveries, depending on the month. Moreover, the summer season was shown to be a suitable time for continuous distribution. The KS test indicated that the Log-Pearson 3 distribution was continuous and that it performed best in July and August. For July, the Phased Bi-Weibull distribution was a particularly complex choice. In several months (May, June, and December), the Wakeby distribution was especially robust throughout the year, periodically matching with advanced rainfall trends. Additionally, the Burr (4P), general extreme value, and Weibull distributions appeared in the months that followed, including October and November, in response to the fluctuation of rainfall.

Figure 12 illustrates the frequency histograms and probability plots of seasonal rainfall, respectively.

3.2.2. Selecting the Best-Fit Results for Maximum Daily and Seasonal Rainfall

The distribution parameters based on the mean daily rainfall were calculated using the maximum likelihood method. Using the A-D, C-s, and K-S tests, the most suitable distribution was identified among the 65 distribution functions for the selected site. Based on the K-S, A-D, and C-s tests, the optimal distribution function for average and seasonal daily rainfall is listed in

Table 10. The best distribution function for maximum daily rainfall (Rank #1).

Month	DF	Test	Classification	Parameter
Jan	Dagum	KS	Continuous-Non-negative	k = 0.47664 α = 4.5829 β = 19.002
	Johnson SB	AD	Continuous-Bounded	γ = 1.9767 δ = 1.599 λ = 72.121 ξ = −1.8612
	Log-Pearson 3	C-s	Continuous-Advanced	α = 17.449 β = −0.13141 γ = 4.9063
Feb	Gen. Logistic	KS	Continuous-Advanced	k = 0.25388 σ = 4.0656 μ = 13.288
	Burr (4P)	AD	Continuous-Non-negative	k = 0.95673 α = 3.9551 β = 15.408 γ = −2.3151
	Gen. Logistic	C-s	Continuous-Advanced	k = 0.25388 σ = 4.0656 μ = 13.288
Mar	Gen. Logistic	KS	Continuous-Advanced	k = 0.31917 σ = 3.2124 μ = 8.5246
	Wakeby	AD	Continuous-Advanced	α = 18.346 β = 3.4325 γ = 3.4914 δ = 0.32051 ξ = 1.1552
	Frechet (3P)	C-s	Continuous-Non-negative/Unbounded	α = 4.4227 β = 18.71 γ = −11.865
Apr	Gamma (3P)	KS	Continuous-Non-negative	α = 1.9461 β = 3.9671 γ = 1.0613
	Triangular	AD	Continuous-Bounded	m = 2.6375 a = 1.0479 b = 22.451
	Log-Logistic (3P)	C-s	Continuous-Non-negative	α = 2.7315 β = 7.6946 γ = −0.19962
May	Johnson SB	KS	Continuous-Bounded	γ = 0.95882 δ = 0.76586 λ = 44.446 ξ = −0.82892
	Johnson SB	AD	Continuous-Bounded	γ = 0.95882 δ = 0.76586 λ = 44.446 ξ = −0.82892
	Levy (2P)	C-s	Continuous-Non-negative	σ = 3.7662 γ = 0.15548
Jun	Log-Pearson 3	KS	Continuous-Advanced	α = 12.186 β = −0.32963 γ = 4.922
	Log-Pearson 3	AD	Continuous-Advanced	α = 12.186 β = −0.32963 γ = 4.922
	Frechet (3P)	C-s	Continuous-Non-negative/Unbounded	α = 2.0452 β = 3.8148 γ = −1.9001
Jul	Pareto 2	KS	Continuous-Non-negative	α = 2.1788 β = 1.0299
	Burr	AD	Continuous-Non-negative	k = 38.604 α = 0.76293 β = 78.059
	Lognormal	C-s	Continuous-Non-negative	σ = 1.5801 μ = −1.1549
Aug	Wakeby	KS	Continuous-Advanced	α = 2.5662 β = 1.3486 γ = 0.25904 δ = 0.41054 ξ = −0.19362
	Wakeby	AD	Continuous-Advanced	α = 2.5662 β = 1.3486 γ = 0.25904 δ = 0.41054 ξ = −0.19362
	Gen. Pareto	C-s	Continuous-Non-negative/Unbounded	k = −0.3155 σ = 1.8361 μ = −0.05725
Sep	Gen. Gamma (4P)	KS	Continuous-Non-negative	k = 0.81775 α = 0.92604 β = 5.0817 γ = 1.1591
	Gen. Pareto	AD	Continuous-Non-negative/Unbounded	k = 0.25254 σ = 3.6722 μ = 1.0147
	Wakeby	C-s	Continuous-Advanced	α = 51.126 β = 44.732 γ = 3.489 δ = 0.27459 ξ = 0
Oct	Phased Bi-Weibull	KS	Continuous-Advanced	α1 = 1.0577 β1 = 13.578 γ1 = 1 α2 = 1.7341 β2 = 13.127 γ2 = 12.453
	Gen. Pareto	AD	Continuous-Non-negative/Unbounded	k = −0.11597 σ = 11.494 μ = 1.3564
	Reciprocal	C-s	Continuous-Bounded	a = 1.8756 b = 48.89
Nov	Wakeby	KS	Continuous-Advanced	α = 51.208 β = 5.7024 γ = 5.5576 δ = −0.04698 ξ = −0.29588
	Wakeby	AD	Continuous-Advanced	α = 51.208 β = 5.7024 γ = 5.5576 δ = −0.04698 ξ = −0.29588
	Error function	C-s	Continuous-Unbounded	k = 1.8886 σ = 6.3089 μ = 12.653
Dec	Wakeby	KS	Continuous-Advanced	α = 174.22 β = 23.352 γ = 7.8323 δ = 0.00332 ξ = 0
	Wakeby	AD	Continuous-Advanced	α = 174.22 β = 23.352 γ = 7.8323 δ = 0.00332 ξ = 0
	Nakagami	C-s	Continuous-Non-negative	m = 0.85455 Ω = 284.19

and

Table 11. Best distribution function for seasonally maximum daily rainfall (Rank #1).

Season	DF	Test	Classification	Parameter
Winter	Wakeby	KS	Continuous-Advanced	α = 2105.9 β = 87.559 γ = 30.657 δ = −0.49652 ξ = 0
	Johnson SB	AD	Continuous-Bounded	γ = 0.59032 δ = 0.95127 λ = 71.693 ξ = 17.303
	Pearson 5	C-s	Continuous-Non-negative	α = 8.9676 β = 355.01
Spring	Wakeby	KS	Continuous-Advanced	α = 119.98 β = 13.908 γ = 17.447 δ = −0.1198 ξ = 6.8618
	Wakeby	AD	Continuous-Advanced	α = 119.98 β = 13.908 γ = 17.447 δ = −0.1198 ξ = 6.8618
	Wakeby	C-s	Continuous-Advanced	α = 119.98 β = 13.908 γ = 17.447 δ = −0.1198 ξ = 6.8618
Summer	Johnson SB	KS	Continuous-Bounded	γ = 0.77046 δ = 0.53354 λ = 17.666 ξ = 1.0915
	Wakeby	AD	Continuous-Advanced	α = 6.5362 β = 0.16191 γ = 0 δ = 0 ξ = 0.50184
	Frechet (3P)	C-s	Continuous-Non-negative/Unbounded	α = 2.7693 β = 7.4869 γ = −4.0114
Autumn	Burr	KS	Continuous-Non-negative	k = 1.082 α = 4.2353 β = 27.892
	Gen. Logistic	AD	Continuous-Advanced	k = 0.23139 σ = 6.4346 μ = 27.028
	Inv. Gaussian	C-s	Continuous-Non-negative	λ = 123.57 μ = 29.637

, respectively. The lowest goodness-of-fit result among all the probability distributions developed was referred to as the best-fit result. It was found that the Wakeby distribution promises frequently over several months and seasons; therefore, it appears to be an excellent fit for characterizing a range of rainfall data patterns. The Johnson SB, Log-Pearson 3, Frechet (3P), and General Logistic distributions further demonstrate their ability to capture various rainfall trends by appearing frequently. Additionally, the probability plots and frequency histograms of the maximum seasonally daily rainfall are shown in Figure 13. According to previous studies, these functions are often used due to their adaptability in fitting skewed and restricted data [136,137]. As a result, they are ideal for assessing rainfall, which often exhibits sharp seasonal variations [136,137].

3.3. Potential Rainwater Harvesting

The $PVR$ is determined by utilizing Equation (15). Figure 14 and Figure 15 show the potential water that could be harvested at the selected location based on the average and maximum daily rainfall with various tilt angles of PV panels, respectively. The annual rainwater that can be gathered from the rooftop of the selected building is shown in Figure 16 and Figure 17 based on the average and maximum daily rainfall with various tilt angles of PV panels, respectively. These figures provide the following results:

• It is found that the value of rainwater that can be collected from the roof of the selected building is 30–36 m³, 17–20m³, 2–3 m³, and 15–18m³ for winter, spring, summer, and autumn, respectively, based on the average daily data.
• Based on the maximum daily statistics, it is determined that the amount of rainwater that can be gathered from the rooftop of the building selected is 283–233 m³, 195–161 m³, 39–32 m³, and 190–156 m³ for winter, spring, summer, and fall, respectively.
• Besides, the highest of 36 m³ and 283 m³ of rainwater that can be collected from the roof of the selected buildings at the tilt angle of 10° is found in the winter season.
• Moreover, the annual rainwater that can be gathered from the rooftop is estimated to vary from 77 m³ to 63 m³, with an average of 71 m³, based on the average daily data and from 708 m³ to 582 m³, with an average of 649 m³, based on the maximum daily data.

Figure 14. Seasonal variation in potential water that could be harvested in the selected locations with various tilt angles based on average daily rainfall data.

Figure 14. Seasonal variation in potential water that could be harvested in the selected locations with various tilt angles based on average daily rainfall data.

Figure 15. Seasonal variation in potential water that could be harvested in the selected locations with various tilt angles based on the maximum daily rainfall data.

Figure 15. Seasonal variation in potential water that could be harvested in the selected locations with various tilt angles based on the maximum daily rainfall data.

Figure 16. Annual value of potential water that could be harvested at the selected location with various tilt angles based on average daily rainfall data.

Figure 16. Annual value of potential water that could be harvested at the selected location with various tilt angles based on average daily rainfall data.

Figure 17. Annual value of potential water that could be harvested at the selected location with various tilt angles based on the maximum daily rainfall data.

Figure 17. Annual value of potential water that could be harvested at the selected location with various tilt angles based on the maximum daily rainfall data.

Furthermore, as mentioned previously, the total amount of water needed for irrigation is estimated to be 204 m³/year for the selected building. Based on average versus maximum daily rainfall data, the results show considerable variation in the rainwater harvesting (RWH) system’s potential supply performance (PSP), particularly for residential buildings. The RWH system may supply 31% to 38% of the chosen building’s yearly irrigation water demand (204 m³/year), taking into account the average daily rainfall (Figure 18). This implies that the system contributes somewhat, but significantly, to the irrigation requirements under normal rainfall conditions. On the other hand, a significantly larger PSP, ranging from 285% to 346%, is seen when maximum daily rainfall data is used (Figure 19). This suggests that the RWH system may be able to gather far more water during periods of high rainfall than the yearly requirement for irrigation, possibly creating an excess that could be saved for later use or put to other non-potable uses.

These results demonstrate the need to install rainwater harvesting systems, especially in urban areas, where water resources are often needed for the irrigation of green spaces. By capturing rainfall, RWH systems can reduce dependency on conventional water sources and provide a sustainable alternative for meeting irrigation requirements. According to previous studies [138,139,140,141], the reviewed literature emphasizes the importance of RWH in urban settings in promoting sustainable water management and reducing reliance on municipal water supply. For instance, RWH can mitigate the negative impacts of urbanization on hydrology, particularly in regions with limited water supplies [138]. In Barcelona’s single-family and multi-family buildings, Domènech and Saurí [139] found that RWH was both economically viable and provided significant drinking water savings. Ndeketeya and Dundu [140] demonstrated that RWH can enhance water security and promote sustainable urban growth in South African cities. RWH is particularly helpful for low-demand applications and lowers the quantity of potable water utilized for non-drinking purposes, according to Fernandes et al. [141].

3.4. Techno-Economic Feasibility of the Proposed System

3.4.1. Optimum Tilt and Azimuth Angle of the Proposed System

To optimize their exposure to sunlight, the panels are arranged in rows, depending on the structure’s location. These angled rows allow for efficient solar absorption, which could lead to increased energy production. Therefore, RETScreen is used to estimate the optimal azimuth and tilt angle.

Moreover, it should be noted that as the tilt degree increases, the roof may accommodate additional solar panels, boosting the system’s energy-generating potential. However, the optimal tilt angle is determined based on the annual electricity exported to the grid. It was determined that the highest values of capacity factor (CF) and maximum amount of power were generated at an optimal tilt angle of 30° and 10°, respectively, when the azimuth angle was 0° (north-facing) and 180° (south-facing), as shown in Figure 20. The best tilt angle for generating the most electricity in east–west orientations was also 10° for azimuth angles of 90° (east-facing) and 270° (west-facing), as shown in Figure 20.

These results are supported by previous studies related to grid-connected PV systems. According to Imam and Al-Turki [142], a grid-connected photovoltaic system with a capacity of 12.25 kW had a CF of 22%. Kebede [143] found that the average CF value of 5 MW grid-connected PV installations was 19.8%. Furthermore, Mohammadi et al. [144] found that the CF value of a 5 MW grid-connected PV plant with varying sun-tracking varied between 17.54% and 27.42%. Based on the findings, it can be concluded that the value gained from the current study is consistent with the accepted values according to previous studies. Thus, it is technically sustainable to install grid-connected PV systems at the selected location.

3.4.2. Techno-Economic Feasibility of the Proposed System

Notably, the minimum global solar irradiation value of 3.28 kWh/m²/day was recorded in December (Figure 21), and the highest daily electricity usage for selected households was recorded in August, with a value of 24 kWh/day. Using Equation (28) [121], the size of the PV system is estimated to be 11 kW.

$$P_{max}={\displaystyle \frac{E_{AC}{P}_i}{G_{SR}{f}_{PV}{\eta }_{inv}}}$$

(28)

where $P_{max}$ is the maximum power,${ P}_i$ is the solar radiation at STC in kW/m², $G_{SR}$ is the global solar radiation (kWh/m²/d), $f_{PV}$ is the PV derating factor (0.8), $E_{AC}$ is the daily power consumption in kWh/d, and ${\eta }_{inv}$ is the inverter yield.

As previously mentioned, the optimum orientation angles (tilt and azimuth angles) are 30$°$ and 0$°$, respectively. Hence, the system capacity is estimated to be 19.08 kW for D#1 and 28.44 kW for D#2. It should be noted that mono-Si JKM360M-76 PV modules with an efficiency of 18.55% and a capacity of 360 W were selected for the proposed system. Moreover, a 10 kW FroniusSymo Advanced 10.0-3 208-240 V 3-Phase String Inverter was utilized in this study.

The monthly energy profile of system D#1 and D#2, which includes energy consumption, solar energy generation, and the energy surplus, is depicted in Figure 22. It is found that the highest and lowest energy production from the solar systems was recorded in July and December, respectively. This variation aligns with typical solar irradiance trends, where output peaks in summer due to longer days and high solar intensity. Every month, the PV system produces significantly more energy than is regularly used, which leads to a steady surplus.

When D#2 is used, for instance, the excess can range from 1656 kWh in December to as much as 3996 kWh in July. This cyclical overproduction demonstrates that the system can satisfy energy resilience and sustainability requirements by not only supplying on-site energy demands but also returning extra energy to the grid or storing it for later use. The significant surplus during the production months, especially March to September, indicates that there is a great deal of room for energy export or integration with other energy-demanding applications, such as electric vehicle charging, water heating, or battery storage. Additionally, the system’s ability to generate more than three times the necessary energy, even during December, when production is at its lowest, showcases the resilience and effectiveness of the installed PV capacity.

RETScreen software is used to calculate the economic feasibility indicators. Prior research has shown that NPV and payback period are crucial considerations when assessing economic feasibility [144,145,146]. Therefore,

Table 12. Economic performance and a net annual reduction in GHG emissions.

Economic Performance of D#1
NPV [USD]	EP [year]	SP [year]	LCOE [USD/kWh]	B-C	ALCS [USD/year]
31744	2.8	7.4	0.041	5.75	2483
Net annual reduction of GHG emissions of D#1
Net annual GHG emission reduction [tCO2]		Cars and light trucks are not used	People are reducing energy use by 20%		Hectares of forest absorbing carbon
19.4		3.6	19.4		1.8
Economic Performance of D#2
NPV [USD]	EP [year]	SP [year]	LCOE [USD/kWh]	B-C	ALCS [USD/year]
54827	1.8	5.0	0.028	9.10	4289
Net annual reduction of GHG emissions of D#2
Net annual GHG emission reduction [tCO2]		Cars and light trucks are not used	People are reducing energy use by 20%		Hectares of forest absorbing carbon
29.0		5.3	29.0		10.0

presents the most important findings on the economic performance of the 24 kW grid-connected photovoltaic system for north–south orientation with a tilt angle of 30°. Based on the NPV value and Refs. [144,145,146], the results show that the proposed plant is financially and economically feasible. Additionally, the cash flow shows that the investor could expect positive cash flow from the solar plant project starting in 1.8 years. Additionally, the feasibility of cash flows from the proposed system can be estimated using the benefit-cost ratio (B-C). A project is considered profitable when the value of B-C is more than one, according to Abdur-Rehman and Al-Sulaiman [147]. Thus, the B-C value (see Table 12) of all the proposed systems indicates the feasibility of the projects. Furthermore, as indicated in Table 12, the results reveal that utilizing the PV project greatly lowers GHG emissions.

4. Conclusions

Investigating alternative energy and water sources has always been an important issue in water resources and environmental management. Therefore, the current study investigated a novel 2-in-1 system that combines renewable solar energy and rainwater harvesting as an alternative solution to address energy and water challenges. This system is designed for family farms. It is installed on the roof of an existing farm building, where it generates electricity from solar energy and collects rainwater for irrigation. To this end, the study addressed the lack of locally observed rainfall data by comparing the available actual data with global satellite databases. The most precise satellite dataset was then utilized to estimate the potential of the rainwater harvesting system. The results demonstrate that the mean amount of rainwater gathered from the rooftop was 71 m³/year and 649 m³/day based on the average and maximum daily rainfall data. Furthermore, the results indicated that the rainwater harvesting system can meet between 31% and 38% of the building’s annual irrigation water demand (204 cubic meters per year) based on average daily rainfall, and between 285% and 346% based on maximum daily rainfall. Consequently, the system may be able to collect much larger quantities of water than is required for irrigation, potentially producing a surplus that can be stored for non-potable uses during periods of heavy rainfall.

Moreover, this study evaluated the feasibility of hybrid solar–RWH systems providing a sustainable alternative to conventional energy sources and water use for purposes other than drinking. It was found that the family’s energy requirements were met by the proposed PV system, and the remaining energy could be transmitted into the grid. Furthermore, the cash flow indicates that the investor can expect a return on investment from the proposed PV system after 2.4 years.

This study concluded that implementing rainwater harvesting in buildings has several advantages, such as lower monthly water bills and compatibility with green building concepts. Besides, it provides a significant reference for sustainable water and environmental management, and this approach can be efficiently applied, especially for a region with a water and electricity crisis like Northern Cyprus. In addition, the results illustrate the system’s significance in climate change adaptation and resource sustainability by showing how it can reduce environmental impact while increasing resilience in water-scarce and energy-dependent places.

5. Limitations and Future Work

Although this study’s primary goal was to evaluate the preliminary techno-economic feasibility of innovative hybrid systems (solar and rainfall harvesting), this research has some limitations that should be addressed in the future to ensure the practicality, safety, and performance of the rooftop solar and rainwater collection system.

First, high temperatures can have an impact on long-term energy yield, increase system degradation, and reduce solar panel efficiency, particularly in Northern Cyprus. The impact of thermal management on solar panel performance and the structural safety of the system under various weather conditions, including wind loads, should be discussed in the future.

Second, the reliability and efficiency of the system are significantly affected by the materials utilized. The system’s effectiveness, resilience, and cost may be improved by selecting the most suitable material. Therefore, the material selections should be examined in the future. To address these challenges, future studies will employ structural analysis and simulation tools such as ANSYS 2021 and SolidWorks 2021. ANSYS or SolidWorks’ finite element analysis (FEA) is a helpful tool for evaluating factors, including material fatigue, wind loads, and the effects of thermal expansion on PV mounting structures.

Third, the shading effects were not examined, since the primary goal of this study was to assess the system’s overall capacity for energy generation rather than to maximize the performance of individual panels. Furthermore, a prior study shows that shading has a significant impact on PV system efficiency [148,149]. Therefore, future studies should conduct an in-depth shade analysis using PVsyst software (7.4.8 2024) to increase the reliability of performance assessments. PVsyst software allows for more precise simulation of shading effects and facilitates the implementation of optimization techniques, such as power optimizers and improved panel arrangement [150,151,152]. This approach will address the study’s limitations and offer a more comprehensive understanding of how shading impacts system performance. Besides, the symmetrical installations might complicate the inverter design due to potential variations in panel performance. If panels with different roof orientations receive varied amounts of sunshine, this could result in inefficiencies. To minimize mismatched performance and enhance overall system efficiency, future research will address particular design issues, such as optimum inverter designs—particularly string inverters with numerous Maximum Power Point Trackers (MPPTs)—using PVsyst software based on previous studies [151,152].

Fourth, the effects of heavy rainfall were not investigated in this study. However, previous studies [153,154,155,156] have highlighted the importance of considering heavy rainfall conditions in the design and maintenance of rainwater harvesting systems. Moreover, the efficiency of solar panels can be reduced by heavy rain due to the limited sunlight, which causes moisture-related damage, such as corrosion and electrical issues [157,158,159,160]. Hence, future research should focus on developing robust system designs and adaptive strategies to improve resilience and ensure reliable performance under extreme conditions.

Finally, the practical performance of the proposed system will be assessed in future studies by means of experimental validation in real-world field environments. Among the important operational parameters that will be examined are water pumping rates, energy efficiency, and the effectiveness of system automation. Additionally, the collected rainwater’s quality will be evaluated to ensure that it meets relevant agricultural standards for irrigation. In order to improve the water’s quality, the project will also investigate the design and testing of basic water treatment parts, such filter units and first-flush systems.

In the end, this study provides a first exploration of the hybrid system’s viability and potential for solving important issues (water and energy). The limits indicated will be considered and resolved in future work to enhance the robustness and usability of the proposed system. These constraints include structural concerns, inverter design, cost optimization, and shade analysis.