1 Introduction

Energy, water and food form a nexus at the heart of sustainable development. Demand for all three is increasing rapidly with agriculture being the largest consumer of the world’s freshwater resources. Sustainable management of energy, water and food is necessary to balance the needs of people, nature and the economy [1]. In rural areas, where the farms are relatively close to a distribution network, the irrigation process is carried over by induction motor-driven water pumps powered from the radial distribution network. This can result in increasing the distribution network’s active-power loss and creating several voltage sag problems [2,3,4].

Management of water resources as well as the energy required for irrigation is crucial. Several researchers employed [5,6,7,8,9,10,11,12] PV-pumped systems to provide a different and more sustainable source of energy for irrigation purposes. Waleed et al. presented [5] a wireless-controlled PV-pumping system. A conceptual model of the wireless-controlled PV-pumping system was presented [5], which was aimed at reducing the quantity of irrigation water to tackle the water scarcity problem. They made use of soil and climate data to optimize the amount of water needed to avoid over-filling or under-filling of the land by irrigation water. While there were no results reported in [5], the presented conceptual model of the system could in theory reduce the required quantity of irrigation water via the inclusion of the soil and climate data into consideration. On the other hand, Chatterjee et al. took [6] a similar approach on designing a PV-based irrigation system with IoT control and also considered the soil and climate data. The proposed system was studied using HOMER software. The researchers considered [6] batteries as their storage device. The reported results showed that the proposed PV irrigation system managed to provide the irrigation water required by the crop with 100% renewable factor (all the energy was generated using renewable PV resources). Also, the economic aspect of the system was taken into consideration, and the payback period of the system was about 7 years for the 20-year project lifetime. To further validate the feasibility of the simulated system, a small-scale laboratory model was built [6]. The experimental set-up results showed that the proposed system should keep the soil moisture above the required limit for the crop to grow. Jahanfar et al., Zamanlou et al. and Soenen et al. considered [7,8,9] the use of a water tank as a storage system and compared its feasibility with the conventional batteries as a storage system. The problem of energy storage elements in solar water pumps was addressed [7], where they impose the greater cost in the PV irrigation system preventing the farmers from replacing their existing systems. A comparison was made [7] between battery storage and water tank storage for solar pumping systems using computer software. The aim was to find the more economically feasible solution to encourage the farmers to change the conventional grid-connected or diesel-powered irrigation systems to PV irrigation. Of course, it would be much cheaper if there was no need for a storage system; therefore, a case with no storage system was also considered. It was reported that the system operating with no storage provided the cheapest option where the total system cost over the total project lifetime was $30,000. However, operating a PV irrigation system with no storage system was reported to be not reliable and thus dismissed [7]. The obtained results also indicated that the cost for a water tank storage ($56,000) was much cheaper than the cost of a battery storage ($72,800) [7]. Although the water tank storage required building and installing a large capacity water tank, due to its longer life span and minimal maintenances when compared with that of battery storage, the water tank storage is much cheaper over the total project lifetime. A solar water pumping system provided with either a battery storage or a water tank storage was designed [8] to ensure high reliability of meeting the crop water demand. The system was simulated and studied using HOMER software. It was reported that the battery storage is approximately cheaper than the water tank storage by 20% [8]. However, the researchers only considered [8] the initial cost of the system and did not include the total cost calculated over the whole project lifetime. Therefore, the reported results before [7] presented a more realistic and reliable outcome of comparing between a water tank storage and a battery storage as it considered the lifetime cost of the irrigation system. Another comparison between water tank storage and battery storage for PV water pumping was attempted [9]. Other researchers looked [9] for the optimum size of a PV irrigation system with a battery storage and a PV irrigation system with a water tank storage to give the minimal cost for each system. It was concluded that the water tank storage has a much higher initial cost than that of batteries. However, the total system cost over the project’s life time is lesser for water tank storage ($24,100) due to its minimal replacement, operation and maintenance cost as compared with the batteries ($31,100). The reported results conformed [9] to the previous reported results [7, 8], where it was concluded that the water tank storage is a better option than batteries economically when calculated over the total project lifetime. Additionally, both scenarios (PV irrigation with water tank storage and PV irrigation with batteries) were implemented experimentally by a case study of an existing PV-pumped system in Africa [9]. Both the systems were simultaneously studied experimentally over a month and an additional interesting outcome was reported. In addition to the positive economic impact of the first system (PV irrigation with water tank storage) over the second system (PV irrigation with batteries), the second system was more stressful for the motor pump than the first system. In the presented case study, the first system required a maximum of 8 starts-stops of the motor with an average of 4 starts-stops per day. On the other hand, the second system required a maximum of 107 starts-stops per day with an average of 84 starts-stops per day. Of course, the more starts-stops the motor operate per day, the more damage that could be inflicted to the motor and shortening its expected lifetime. Therefore, the water storage tank was reported experimentally and by simulation to be a better solution for PV irrigation than batteries. Santra et al. evaluated [10] the performance of a stand-alone PV irrigation system with no storage systems at all. The absence of a storage device was justified by the researcher according to the fact that crops do not require water during every hour of the day, and therefore, irrigation of the crops during the sun hours could be sufficient if calculated accurately. The method of irrigation was taken into account on studying the system [10]. A micro-irrigation technique (drip irrigation) was considered in order to save water and deliver water directly to the plant. A small-sized solar PV irrigation system (1 hp induction motor) was studied experimentally and by simulation. The aim of the steady was to study the feasibility of a small motor to operate mini sprinklers to provide sufficient water to the plant irrigated in an area of 320 m2. Also, the economic and environmental aspects of the system were taken into consideration to show the benefits of utilizing a PV irrigation system over grid-connected or diesel-powered irrigation systems. The obtained results showed the small 1 hp pump could operate the mini sprinklers and provide enough water pressure to water the plant. The system showed a payback period of 6 years over a total project lifetime of 25 years, which indicated a significant economic impact on using PV irrigation system. Furthermore, the carbon footprint of the system was significantly reduced when compared with grid-connected or diesel-powered irrigation systems. The carbon footprint of the PV irrigation system calculated per hectare irrigated by the system was 0.009 kg CO2-eq. On the other hand, the carbon footprint of a grid-connected irrigation system and diesel-powered irrigation system calculated per hectare irrigated by the system was 1.124 kg CO2-eq and 0.381 kg CO2-eq, respectively. Almutairi et al. compared [11] between normal surface irrigation and micro-drip irrigation. The comparison confirmed that surface irrigation has a lower cost than drip irrigation. However, drip irrigation can save water up to 15% when compared with surface irrigation. Dursun et al. optimized [12] the soil moisture sensor placement to improve the performance of a PV-powered drip irrigation system. The obtained results conformed to the previously discussed results in literature highlighting the importance of PV irrigation and drip irrigation.

Distributed generation (DG) using PV panels to reduce the distribution systems’ active-power loss is widely used and studied in the literature [13,14,15,16]. Radosavljević et al. and Hussain et al. looked [13, 14] for the optimal location and sizing of a renewable DG resources to reduce the active power loss in a radial distribution system. Combination of two optimization algorithms (particle swarm optimization and gravitational search algorithm) was aimed [13] to achieve minimum power loss on a radial distribution system by optimal location and sizing of DG resources. The IEEE 69-bus system was studied by computer simulation using MATLAB software. The economic impact of integration of the radial distribution system with DG resources on formulating the objective function to maximize the profit for the DG holder was considered [13]. The original system without connection of any DG resources suffered a total power loss of 224.946 kW. On connecting PV DG resources, the total power loss was reduced to 83.2 kW. This corresponded to a reduction in power loss by 63% when compared with no connection of any DG resources. The system was studied over a period of 10 years taking into account the price of purchased energy from the grid and the price of the power loss. On using PV DG resources, the net profit generated over the project life time [13] from reducing the system active-power loss and the purchased energy amounted to 2.879 M$. On the other hand, other researchers only considered [14] the power loss reduction on connecting DG resources to a radial distribution system. A 11-kV radial distribution system was studied using ETAP software. Three different levels of DG resources penetration were considered, namely 25%, 50%, and 75%, of the rated load supplied by the radial system. Before the installation of any DG resources, the system had a 544-kW active-power loss. On connecting DG resources with value of 25, 50 and 75% of the rated supplied load, the active-power loss was reduced to 378, 257 and 185 kW, respectively. The obtained results conformed [14] to the results obtained before [13], supporting the same conclusion that integration of DG resources with a radial distribution system managed to reduce the system’s active-power loss. Purlu et al. considered [15] both the active-power loss reduction and voltage improvement of a distribution system on connecting DG resources. The IEEE 33-bus system was considered [15] and studied using MATLAB software. It was aimed to find the optimal placement and sizing of PV DG resources along the system to achieve minimum active-power losses and voltage deviation. Two optimization algorithms were implemented (genetic algorithm and particle swarm optimization) to achieve the aims of the study. The system had a 202-kW power loss without the connection of any DG resources. On connecting maximum level penetration of PV DG resources without causing severe over voltage, the power loss was reduced to 96 kW. This corresponded to a 52.5% reduction in power loss when compared to the system with no DG resources. The maximum voltage on connecting the PV DG resources was 1.05 pu and remained within the acceptable limits of the voltage. Radosavljevic et al. looked [16] for the optimal location of PV DG resources to reduce the system active-power loss. The obtained results conformed to the results presented in the discussed literature [13,14,15].

From the previously discussed literature, the utilization of PV DG resources in order to improve the electrical performance of a distribution system was widely explored and discussed in the literature. Also, PV irrigation was also studied and the economic impacts of a PV irrigation system over conventional grid-connected or diesel-powered irrigation systems was reported. However, the integration of the PV DG resources with a radial distribution system to power an irrigation system in order to improve the irrigation system while simultaneously improving the electrical distribution system’s performance is yet to be studied.

This research presents a techno-economic study of a real radial distribution network in Egypt (Karot distribution line) feeding several water pumps. The system’s existing status is expressed by the distribution line’s active-power loss, the used water quantity and the total system cost and is considered a base case for the present study. Four scenarios are proposed for improving the system’s operation and compared against the present case study of Karot line (existing scenario). The main contribution points of the present research can be summarized as follows:

  1. 1.

    A techno-economic study seeking determination of the optimal minimal cost as well as minimal active power loss and water quantity without violating the crop water-demand.

  2. 2.

    Effect of utilizing an energy demand-side management technique on the operation of irrigation pumps for reducing the system cost and the distribution system’s active-power loss.

  3. 3.

    Study of stand-alone PV panels as well as grid-connected PV panels integration for further improvement of the system performance.

  4. 4.

    Impact of modifying the irrigation methodology on the performance of the system.

The rest of this research is organized as follows: The methodology is presented in Sect. 2 where the system configuration with the proposed scenarios along with the required software for system analysis is described. The obtained results and discussions are presented in Sect. 3. Finally, Sect. 4 concludes the main research findings.

2 Methodology

2.1 System configuration

The Karot distribution line is a radial distribution line extended along Al Quseyya town area in Assiut Governorate in Egypt. A schematic of the Karot distribution line is shown in Fig. 1. The line is fed from a 66/11 kV substation and supplies various loads via 47 step-down distribution transformers of voltage rating 11/0.38 kV. The first 31 transformers are used to supply residential loads, while each of the last 16 is used to feed a 30-hp water pump motor [17]. The 16 irrigated areas all have an approximate area of 20 acres and the main crop being planted is wheat having a planting season of 130 days. The irrigation method used by the local farmers at Karot distribution line is surface irrigation, and the existing irrigation schedule is quoted by surveying the local farmers. During irrigation season, the line suffers from increase in active power loss and several voltage sag problems [17, 18]. This is mainly due to the radial nature of Karot line and the motors connection at the far end of the line.

Fig. 1
figure 1

Schematic of Karot distribution line [17]

Full size image

2.2 Load flow analysis of the electric system

The electric system of Karot distribution line is analyzed using MATLAB software. The backward–forward sweep method is an iterative method utilized to calculate the nodes’ voltages, branches’ currents and power flow for radial networks [19]. The backward–forward sweep inputs include the loads’ apparent power in pu (Sℓoad), transformers’ impedances in pu (Ztr) and cables’ or overhead lines’ impedances in pu (Zcℓ).

First, all low-voltage load-buses are assumed to have voltage Vℓ equal 1.0 pu. Then, one can easily calculate the branch currents in each section using the following equations:

$$ i_{\ell } \left( j \right) = \left( {\frac{{S_{{{\text{load}}}} \left( j \right)}}{{V_{\ell } \left( j \right)}}} \right)^{*} ,\;\;j \in \left\{ {1,2, \ldots N} \right\} $$
(1)
$$ i_{{\text{h}}} \left( j \right) = i_{\ell } \left( j \right),\;\;j \in \left\{ {1,2, \ldots N} \right\} $$
(2)
$$ i_{c\ell } \left( j \right) = \mathop \sum \limits_{k = j}^{N} i_{{\text{h}}} \left( k \right),\;\;j \in \left\{ {1,2, \ldots N} \right\} $$
(3)

where N is the number of low-voltage load-buses, iℓ is the current in the transformer low-voltage side in pu, ih is the current in the transformer high-voltage side in pu and icℓ is the current in the cables or overhead lines in pu.

Equations (1–3) are used to calculate the branches’ currents starting from the last node and back toward the first node or slack bus. After acquiring values for all currents in each line section, the forward sweep utilizes the current values to calculate the voltage drops across the radial line toward the last node using the following equations:

$$ V_{{\text{h}}} \left( j \right) = V_{{\text{h}}} \left( {j - 1} \right) - i_{c\ell } \left( j \right)*Z_{c\ell } \left( j \right),\;\;j \in \left\{ {1,2, \ldots N} \right\} $$
(4)
$$ V_{\ell } \left( j \right) = V_{{\text{h}}} \left( j \right) - i_{{\text{h}}} \left( j \right)*Z_{{{\text{tr}}}} \left( j \right),\;\;j \in \left\{ {1,2, \ldots N} \right\} $$
(5)

where Vh is the voltage of the high-voltage buses in pu and Vℓ is the voltage of the low-voltage load-buses in pu. The process is then iterated until the variation in voltage in the last two iterations reaches a predefined error tolerance. After acquiring both the nodes’ voltages and the branches’ currents, one can simply calculate the power flow and losses of the system using the following equations:

$$ S_{c\ell 1} \left( j \right) = V_{{\text{h}}} \left( {j - 1} \right) * i_{c\ell }^{*} \left( j \right) ,\;\,j \in \left\{ {1,2, \ldots N} \right\} $$
(6)
$$ S_{c\ell 2} \left( j \right) = V_{{\text{h}}} \left( j \right) * i_{c\ell }^{*} \left( j \right) ,\;\;j \in \left\{ {1,2, \ldots N} \right\} $$
(7)
$$ S_{h} \left( j \right) = V_{{\text{h}}} \left( j \right) * i_{{\text{h}}}^{*} \left( j \right),\;\;j \in \left\{ {1,2, \ldots N} \right\} $$
(8)
$$ S_{\ell } \left( j \right) = V_{\ell } \left( j \right) * i_{\ell }^{*} \left( j \right),\;\;j \in \left\{ {1,2, \ldots N} \right\} $$
(9)
$$ S_{{{\text{loss}}}} = \mathop \sum \limits_{j = 1}^{N} \left( {S_{c\ell 1} \left( j \right) - S_{c\ell 2} \left( j \right)} \right) + \mathop \sum \limits_{j = 1}^{N} \left( {S_{{\text{h}}} \left( j \right) - S_{\ell } \left( j \right)} \right) = P_{{{\text{loss}}}} + jQ_{{{\text{loss}}}} $$
(10)

where Scℓ1 is the apparent power flow at the beginning of each cable/line in pu, Scℓ2 is the apparent power flow at the ending of each cable/line in pu, Sh is the apparent power flow at the high-voltage side of the transformers in pu, Sℓ is the apparent power flow at the low-voltage side of the transformers in pu, Sloss is the total system apparent power loss in pu, Ploss is the total system active-power loss in pu which include losses in transformers, cables and/or overhead lines and Qloss is the total system reactive-power loss in pu. Figure 2 shows a simple illustration of the backward–forward sweep method, and a flow chart for the method’s calculations is given in Fig. 20 in Appendix 1.

Fig. 2
figure 2

Backward–forward sweep method for radial networks

Full size image

2.3 Water demand for irrigation load management

For calculation of the water demand for irrigation load, the relation between water flow rate from the pump and the input power to the pump driving-motor is required and can be realized using the following equations [20]:

$$ P_{{{\text{in}}}} \left( t \right) = 2.72*Q_{{\text{w}}} \left( t \right)*H_{{{\text{TE}}}} \left( t \right)*\eta_{{{\text{mp}}}} $$
(11)
$$ H_{{{\text{TE}}}} \left( t \right) = H_{{{\text{OT}}}} + H_{{{\text{ST}}}} + (H_{{{\text{DT}}}} - H_{{{\text{ST}}}} ) * \frac{{Q_{{\text{w}}} \left( t \right)}}{{Q_{\max } }} + H_{{\text{F}}} $$
(12)

where Pin is the input power to the pump driving-motor in kW, Qw is the flow rate of pumped water in m3/h, HTE is the total head from the borehole in m, ηmp is the motor/pump efficiency, HOT is the vertical head from the water outlet to the ground level in m, HST is the static level of groundwater in m, HDT is the dynamic level of water in borehole in m, Qmax is the maximum daily discharge capacity of the well in m3/h and HF is the head friction loss in m.

2.4 Cost analysis

The total system cost is an important factor for evaluating the performance of the system. The system cost can be calculated as follows [21]:

$$ {\text{NPC}}_{{{\text{com}}}} = C_{{{\text{init}},{\text{com}}}} { } + { }C_{{{\text{rep}},{\text{com}}}} + { }C_{{{\text{o}}\& {\text{m}},{\text{com}}}} $$
(13)
$$ {\text{C}}_{{{\text{init}},{\text{com}}}} = N_{{{\text{com}}}} {* }P_{{{\text{nom}},{\text{com}}}} {* }C_{{{\text{com}}}} $$
(14)
$$ C_{{{\text{rep}},{\text{com}}}} = C_{{{\text{init}},{\text{com}}}} {* }\mathop \sum \limits_{k = 1}^{{k_{{{\text{rep}}}} }} \frac{1}{{\left( {1 + {\text{int}}} \right)^{{{\text{LF}}_{{{\text{com}}}} { * }k}} }} $$
(15)
$$ {\text{NPC}}_{{{\text{sys}},{\text{tot}}}} { } = \mathop \sum \limits_{{{\text{com}}}} {\text{NPC}}_{{{\text{com}}}} $$
(16)
$$ C_{{{\text{sys}},{\text{anu}}}} = {\text{NPC}}_{{{\text{sys}},{\text{tot}}}} {\text{* CRF}} + E_{{{\text{grid}},{\text{bought}}}} {\text{* PWF}} - E_{{{\text{grid}},{\text{sold}}}} {\text{* PWF}} $$
(17)
$$ {\text{CRF }} = { }\frac{{{\text{int}}\left( {1 + {\text{int}}} \right)^{{{\text{LF}}_{{{\text{proj}}}} }} }}{{\left( {1 + {\text{int}}} \right)^{{{\text{LF}}_{{{\text{proj}}}} }} - 1}} $$
(18)
$$ {\text{PWF}} = \left( {\frac{{1 + {\text{inf}}}}{{1 + {\text{int}}}}} \right){*}\left( {\frac{{1 - \left( {\frac{{1 + {\text{inf}}}}{{1 + {\text{int}}}}} \right)^{{{\text{LF}}_{{{\text{proj}}}} }} }}{{1 - \left( {\frac{{1 + {\text{inf}}}}{{1 + {\text{int}}}}} \right)}}} \right) $$
(19)

where NPCcom is the net present cost of each system component, Cinit,com is the initial cost of each system component, Crep,com is the replacement cost of each system component, Co&m,com is the operation and maintenance cost of each system component, Ncom is the number of units of each system component, Pnom,com is the rated power of a single unit of each system component in (kW), Ccom is the cost of system component in ($/kW), int is the interest rate, LFcom is the component lifetime, NPCsys,tot is the total net present cost of the system, Csys,anu is the annualized system cost over the total system life time, CRF is the capital recovery factor, Egrid,bought is the required energy from the grid to operate the motors, PWF is the present worth factor, Egrid,bought is the energy injected from the PV panels back into the grid, LFproj is the project lifetime which is assumed to be 20 years the same as that of PV panels and inf is the inflation rate.

The technical and economic parameters of the system’s components are given in Table 1.

Table 1 Technical and economic parameters of the system’s components
Full size table

2.5 Proposed scenarios for improving the system operation

The currently existing scenario of operation of the investigated distribution system is analyzed along with four proposed scenarios to improve the system’s steady-state performance. In the existing scenario, the motors are operated from the investigated distribution system. The operation hours of the motors during the day are quoted from the farmers. The performance of the investigated distribution system in the existing scenario of operation is evaluated and taken as a base case to be compared against the proposed scenarios. The proposed scenarios include:

2.5.1 Scenario 1: Modification of surface-irrigation schedule

Optimization of the existing surface-irrigation schedule is achieved using CROPWAT computer program. It is a decision support tool developed by the Land and Water Development Division of the Food and Agriculture organization (FAO). CROPWAT calculates the irrigation load water requirement based on climate, crop and soil data [25, 26]. Furthermore, the program provides the optimum daily water demand for surface irrigation.

2.5.2 Scenario 2: Energy demand-side management

Time of use tariff method is an energy demand-side management technique implemented to shift the motors’ operating time toward light-load hours in order to improve the system’s steady-state performance. The price of energy is higher during peak-load hours than that of light-load hours making it more favorable for the farmers to irrigate during light-load hours. Peak-load tariff is applied for 4 h during the peak-load period as dictated by the Ministry of Electricity and Renewable Energy [23].

For this scenario, the optimum quantity of daily required water for surface irrigation is obtained from CROPWAT software the same as that of Scenario 1. Binary optimization [27, 28] is utilized to acquire the best hourly scheduling for the irrigation to further reduce the total system active-power loss and the total system cost. The decision variable is the motors’ state represented by a binary vector (Msmxn), and each value of this vector can either be on (1) or off (0) to express the binary nature of the optimization. The number of motors for irrigation (= 16) is represented by (m) (m = 1, 2…, 16), and the day hours are represented by (n) (n = 1, 2…, 24). For example, motor number 3 operating at hour 5, the value of element Ms3,5 is equal to 1. The objective function (ObjfA) is to minimize the seasonal system cost (Csys,seas) and the total system active-power loss (Ploss) and being added together with weights K1 and K2; both being assumed equal to 0.5.

$$ {\text{Obj}}f_{A} = {\text{min}}\left( {K_{1} *C_{{{\text{sys}},{\text{anu}}}} + { }K_{2} * P_{{{\text{loss}}}} } \right) $$
(20)

Substituting for the daily water quantity obtained from CROPWAT (Qw) in Eq. (11), the daily required power to meet this water demand is determined. With the knowledge of the daily required power and the irrigation motor’s rated power, one can easily calculate the daily required operation hours (tref) for each motor to meet the water demand. The optimization solution must satisfy the following equality constraint:

$$ \mathop \sum \limits_{n = 1}^{24} M\left( {m,n} \right) = t_{{{\text{ref}}}} ,\;\;m \in \left\{ {1,2, \ldots 16} \right\} $$
(21)

This constraint ensures that each motor operates for the exact required hours of operation to pump enough water for irrigation purpose. The built-in MATLAB genetic algorithm function is utilized, and the lower and upper limits of the decision variable (M) are set to be an integer of values zero or one only to meet the binary nature of the problem.

2.5.3 Scenario 3: PV resources integration

PV panels are sized for each motor and is located close to it; hence, no active power losses in transformer or cables and overhead lines feeding the motors. PV panels output power can be easily determined as follows [21]:

$$ P_{{{\text{pv}}}} \left( {\text{t}} \right) = P_{{{\text{pv}},{\text{rated}}}} {*}\frac{G\left( t \right)}{{G_{{{\text{ref}}}} }}{*}\left[ {1 + \beta_{{{\text{ref}}}} {*}\left( {T_{{\text{c}}} \left( t \right) - T_{{{\text{ref}}}} } \right)} \right] $$
(22)
$$ T_{{\text{c}}} \left( t \right) = T_{{\text{a}}} \left( t \right) + \frac{{{\text{NOCT }} - { }20}}{800}*G\left( t \right) $$
(23)

where Ppv is the PV panels output power in kW, Ppv,rated is the rated PV panels power in kW, G is the hourly solar irradiation in W/m2, Gref is the reference solar irradiation and is equal to 1000 W/m2, βref is the PV panels power thermal coefficient and is equal to − 3.7*10–3 / °C, Tc is the cell temperature, Tref is the reference temperature and equals 25 °C, Ta is the ambient temperature in °C and NOCT is the nominal operating cell temperature in °C.

2.5.3.1 Sub-scenario 3.1: stand-alone PV resources

The motors are operated from PV resources and disconnected from the investigated distribution system.

2.5.3.2 Sub-scenario 3.2: stand-alone PV resources with water tank storage

A water tank storage is introduced along with the use of PV resources to provide a different source of irrigation water. For Sub-scenarios 3.1 and 3.2, the motors receive power only from the PV panels and are completely disconnected from the distribution line. The PV panels size for both sub-scenarios and water tank size for Sub-scenario 3.2 only is determined using the built-in MATLAB genetic algorithm (GA) optimization function. Both the PV panels size and the water tank size can vary continuously between upper and lower limits. This is why the optimization strategy is considered a continuous one. The objective function (ObjfB) is to minimize the annualized system cost (Csys,anu). The constraint imposed on the optimization problem is the water shortage (Qshortage) to ensure that the irrigation system meets successfully the crop water requirement. Water shortage is represented as an equality constraint where Qshortage must equal zero in any possible solution.

A flow chart for the optimization problem for Sub-scenario 3.2 is shown in Fig. 21 in Appendix 2. The water storage tank is used to store water in days where the PV panels output power is greater than the required power to irrigate the land and to replenish the shortage in water (represented by a variable called Qdiff) over the planting season in 130 days when the PV panels power is not sufficient to provide the required water requirement.

$$ {\text{Obj}}f_{B} = {\text{min}}\left( {C_{{{\text{sys}},{\text{anu}}}} } \right) $$
(24)
$$ \mathop \sum \limits_{t = 1}^{3120} Q_{{{\text{shortage}}}} \left( t \right) = 0 $$
(25)
2.5.3.3 Sub-scenario 3.3: grid-connected PV resources

PV resources are connected to the investigated distribution system at the motors’ buses and the motors receive power from both the distribution system as well as the PV panels. Figure 22 in Appendix 3 shows a flow chart describing the scheduling process of Sub-scenario 3.3. The grid acts as a storage to provide the motors with the required power during the days when the PV panels power is not sufficient. During days where PV panels power is greater than the required power for irrigation, the surplus power from the PV panels is injected back to the grid.

The built-in MATLAB multi-objective genetic algorithm function is utilized. In this scenario, the objective function (ObjfC) is to minimize both ƒ2, the total net present cost of the system (NPCsys,tot) and ƒ2, the total system active-power loss (Ploss). The possible solutions are presented by drawing the Pareto front. The optimum solution is selected by determining the utopia point, which is the point with the smallest Euclidean distance (d) from the minimum of the two functions [29,30,31,32]. Figure 3 shows a simplified chart presenting the Pareto front and the utopia point.

$$ {\text{Obj}}f_{C} = {\text{min}}\left( {{\text{f}}_{1} ,{\text{f}}_{2} } \right) $$
(26)
$$ {\text{f}}_{1} = {\text{NPC}}_{{{\text{sys}},{\text{tot}}}} $$
(27)
$$ {\text{f}}_{2} = {\text{P}}_{{{\text{loss}}}} $$
(28)
Fig. 3
figure 3

Pareto front and utopia point for multi-objective optimization

Full size image

It is important to point out that during the sub-scenarios of PV-irrigation, the output power of the PV panels is fluctuating, and therefore, a lower limit is imposed on the output power of PV panels (Pmin) to ensure no pump operation with insufficient power. For a 30-hp pump operating at a total head of 70 m, a minimum input power Pmin of 5 kW is requested [33] following the manufacturers’ recommendation. If the PV panels output power (Ppv) is less than Pmin, the motors are not operated from the PV panels during this particular hour.

2.5.4 Scenario 4: drip irrigation

Studying a different irrigation technique is necessary to evaluate how the performance of the system is influenced by the irrigation technique. The optimum required daily quantity of irrigation water for drip irrigation is obtained from CROPWAT software. The drip irrigation is studied with the use of PV resources following the same three sub-scenarios as those of Scenario 3, namely

  1. 1.

    Sub-scenario 4.1: Drip irrigation using stand-alone PV resources

  2. 2.

    Sub-Scenario 4.2: Drip irrigation using stand-alone PV resources with water tank storage

  3. 3.

    Sub-Scenario 4.3: Drip irrigation using grid-connected PV resources

Figure 4 represents a system chart summarizing the proposed techno-economic scenarios to improve the system’s operational performance.

Fig. 4
figure 4

System chart representing the proposed scenarios

Full size image

3 Results and discussion

The present results are based on data collected through an interview with farmers on site and are presented in Table 2.

Table 2 Data used for the proposed scenario calculations
Full size table

The motor/pump efficiency (ηmp) is obtained from the pump datasheet [33], and it has a range from 45 to 55%. In the present research, it is taken to equal 50%. The existing electrical load profile of Karot distribution line is measured on site.

3.1 Existing scenario results

The existing irrigation schedule for the surface irrigation scheme in Karot line is studied and taken as the base case for comparison against other proposed scenarios. The farmers follow a 7-day cycle for irrigation. Every motor is responsible for the irrigation of 20 acres over 5 days and the farmers operate the motors starting from 5:00 a.m. for 12 h daily to irrigate 4 acres per day. Due to the lack of proper estimation of the crop water requirement, the farmers apply the same irrigation cycle for the whole planting season of 130 days. Using Eq. (11) and substituting for the motor rated power in the input power to the pump driving-motor (Pin), one can determine the hourly water flow rate (Qw). Since the motors operate 12 h daily in the existing scenario, the daily water quantity is determined, and since each motor irrigates 4 acres per day, the daily water quantity per acre can be calculated. The existing irrigation schedule represented by the daily water quantity per acre and the daily motor operation hours is shown in Fig. 5. Figure 6 shows the existing scenario’s hourly load profile where the motor load is mostly during the peak-load hours. For this scenario, the total required water quantity per acre over the whole planting season is obtained by summing up the daily water quantity required per acre as presented in Fig. 5a to reach about 3100 m3/acre. The total system active-power loss (Ploss) as obtained from Eq. (10) is equal to 133,683 kW. The annualized system cost (Csys,anu) obtained from Eq. (17) is equal to $595,311. Each motor is operated for a total of 1080 h over the whole planting season which is the total sum of the daily motor operation hours as presented in Fig. 5b. These values are taken as reference for comparison against the proposed scenarios results.

Fig. 5
figure 5

Existing irrigation schedule: a water quantity per acre and b motor operation hours

Full size image
Fig. 6
figure 6

Hourly load profile for the existing scenario

Full size image

3.2 Results of Scenario 1

In this scenario, irrigation of 4 acres per day per motor is kept the same as scheduled in the existing scenario. Climate, crop and soil data are used with the CROPWAT software to obtain the optimum daily quantity of water for surface irrigation. With the known optimum daily quantity of water for surface irrigation and the determined hourly water flow rate (Qw) of the pump in the same way as in the existing scenario, one can determine the daily motor operation hours. The irrigation schedule represented by the daily water quantity per acre and the daily motor operation hours is presented in Fig. 7. On inspecting Fig. 7a, water quantity demand by irrigation increases with the crop growth stage, while it is constant in the existing scenario as shown in Fig. 5a. The total required water quantity per acre over the whole planting season is obtained by summing up the daily water quantity required per acre as presented in Fig. 7a to reach about 1944 m3/acre with a percentage reduction of 37.29% when compared with the existing scenario. Each motor is operated for a total of 585 h over the whole planting season, which is the total sum of the daily motor operation hours shown in Fig. 7b.

Fig. 7
figure 7

Irrigation schedule for Scenario 1 (a) water quantity per acre and (b) motor operation hours

Full size image

Because the motors operate for less total number of hours when compared with the existing scenario (585 h against 1080 h), the total system active-power loss obtained from Eq. (10) is reduced to 119,882 kW with a percentage reduction of 10.32% when compared with the existing scenario. The system annualized cost obtained from Eq. (17) is reduced to $301,687 with a percentage reduction of 49.3% when compared with the existing scenario. On careful inspection of Fig. 7b, there exists several days by the end of the planting season where the motors operate for most of the day hours including those of the peak load. This is associated with excessive increase in the total system active-power loss and over suction of the underground well, which may eventually lead to drying up the well. Subsequently, this scenario calls for further improvement.

3.3 Results of scenario 2

Time of use tariff is utilized to make it more favorable for the farmers to operate the motors outside the peak-load hours where the energy price will be lower on operation away from the peak-load hours. The peak-load tariff is applied only for 4 h during the peak-load period as dictated by the Ministry of Electricity and Renewable Energy [21]. The 4 h are selected from 9:00 in the morning to noon time where the load is at its peak as shown in Fig. 6. Careful examination of Fig. 7a shows that there exist several days where no irrigation is required. The days when irrigation is required are 35 days represented by 7 periods each of 5 days length, meaning that irrigation takes place only 27% of the whole planting season. This allows extension of each of the 7 irrigation periods shown in Fig. 7a to exceed 5 days while maintaining the same total required water quantity per acre and the same total motor operation hours over the whole planting season as Scenario 1 to ensure the crop gets its required water quantity. The 7 irrigation periods are extended in Fig. 8a to be 10 days for each period instead of 5 days as in Scenario 1. The total required water quantity per acre over the whole planting season is obtained by summing up the daily water quantity required per acre as presented in Fig. 8a and found equal to about 1944 m3/acre, which is exactly the same as that of Scenario 1. Each motor operates for a total of 585 h over the whole planting season being equal to the total sum of the daily motor operation hours shown in Fig. 8b and is also the same as that of Scenario 1. However, since the duration of each irrigation periods is doubled (from 5 to 10 days each), the system now irrigates 2 acres per day instead of irrigating 4 acres in Scenario 1. This significantly decreases the motor operation hours per day as shown in Fig. 7b as the maximum daily operation hours for each motor does not exceed 12 h in any given day over the whole planting season, while in Fig. 7, the operation hours exceed 23 h. On inspecting Fig. 8b, there exist 5 unique irrigation periods where the motors are required to operate for 3, 5, 9, 10 and 12 h daily. Applying binary optimization discussed in Sect. 2.5.2, one can obtain the hourly load profile for the 5 irrigation periods as shown in Fig. 9. The resulting binary optimization solution presented in Figs. 9a, b, c, d and e shifts the motor operation hours away from the peak hours.

Fig. 8
figure 8

Irrigation schedule for Scenario 2: a water quantity per acre and b motor operation hours

Full size image
Fig. 9
figure 9

Hourly load profile for Scenario 2: a 3-h period, b 5-h period, c 9-h period, d 10-h period, e 12-h period

Full size image

This leads to a decrease in the total system active-power loss obtained from Eq. (10) and the system annualized cost obtained from Eq. (17) by 12% to reach 117,590 kW and 54.7% to reach $269,538, respectively, when compared with the existing scenario. It is worth pointing out that the objective function of the binary optimization was initially formulated to minimize the system annualized cost only. In some runs, simultaneous operation of all motors is observed during the same hour. These runs manage to successfully minimize the system annualized cost (Csys,anu) being represented only by the cost of purchased energy from the grid as expressed in Eq. (17) since there are no other components like PV panels, inverter or a water tank in this scenario. Any solution where the motors operate away from the 4 h when the peak-load tariff is applied would minimize the optimization problem’s objective function (Csys,anu). However, the simultaneous operation of all motors during the same hour leads to system overloading with a subsequent increase of the total system active-power loss. This calls for including the total system active-power loss (Ploss) in the formulation of the objective function to be minimized as discussed in Sect. 2.5.2. Load factor is defined [34] as equal to the ratio between the average load (Pℓavg) and the peak load (Pℓpeak). It is generally best to acquire a load factor of 1 where the load profile remains flat line with no variations. Inclusion of the total system active-power loss in the formulation of the objective function helps to improve the load factor of the system. Figure 9e is compared against the existing scenario’s hourly load profile because both figures have the same daily hours of operation per motor (12 h). The load factor for the existing scenario is 0.69 against 0.83 for Scenario 2 with an increase of the load factor by 20.23%.

3.4 Results of Sub-scenario 3.1

The maximum operation hours for each motor presented in Fig. 8b do not exceed 12 h in any given day, a condition which can be met using only PV panels whose power generation continues for up to 12 h (from sun rise to sun set) during the planting season’s days. Figure 10 shows the hourly PV panels data over the 113rd day of the planting season represented by the solar irradiance and the output PV panels power against the motor’s required power. The 113rd day is specially taken into consideration as it represents one of the days as the daily motor operation hours reach 12 h as shown in Fig. 8b. The results from the optimization problem discussed in Sect. 2.5.3.2 give an optimum size of PV panels of 434.84 kW per motor. In this scenario, the motors are completely disconnected from the distribution line. Evaluating the distribution line’s power flow using backward–forward sweep method discussed in Sect. 2.2 but without the motor loads shows that the total system active-power loss is 106,714 kW. This corresponds to a reduction in the total system active-power loss by 20.17% when compared with the existing scenario. However, the system annualized cost obtained from Eq. (17) is increased to $1,030,835 with a percentage increase of 73.16% when compared with the existing scenario. One major drawback of this scenario is the excess energy. The total energy generated over the whole planting season from the 16 PV panels; each rated 434.84 kW for each motor as discussed in Sect. 2.5.3 equals 4742.6 MWh. This total generated energy is calculated using Eq. (22) to determine the PV panels output power. Because this power is calculated hourly, the energy is equal to the power. Knowing the total motor operation hours over the whole planting season being equal to 585 h and with the knowledge of the motors rated power, the total quantity of energy required by all 16 motors during the whole planting season is only 265.6 MWh. Therefore, the excess energy in this scenario equals 4477 MWh. The excess energy as a percentage of the corresponding total generated energy from PV panels is equal to 94.4%. The presence of large excess energy is due to the fact that irrigation is not required daily in a surface irrigation scheme as depicted by Fig. 8a while PV panels generate power on a daily basis. Observing Fig. 10b, it shows that the PV panels are excessively sized in this scenario. At hour equals to 17 when the solar irradiance is the lowest, the PV panels are excessively sized to generate enough power to meet the motor’s rated power. This is due to the fact that there are no other sources of energy for the motors in this scenario and no storage system, so the PV panels alone must satisfy the motor’s power requirement. The excessive sizing of PV panels significantly increases the system annualized cost and the excess energy from the PV panels. Subsequently, this scenario calls for another further improvement.

Fig. 10
figure 10

Hourly PV panels data during the 113rd day of the planting season: a solar irradiance and b PV panels output power and motor’s required power

Full size image

3.5 Results of Sub-scenario 3.2

In order to attempt to solve the problem of excessive sizing of PV panels and the excess energy problem associated with Sub-scenario 3.1, a water tank storage is considered in Sub-scenario 3.2. On days without irrigation, the generated energy from PV panels is utilized to charge the water tank and store excess water in it. During the days when the PV panels power is not sufficient to meet the crop water requirement, the water tank storage acts as an alternate source of water to meet the crop water requirement. Figure 11 shows the water quantity in the water tank over the planting season demonstrating the charge and discharge of the water tank. The optimum PV panels size and the water tank size obtained from the optimization problem discussed in Sect. 2.5.3.2 equals 45.93 kW and 1666 m3 per motor, respectively. The PV panels size is significantly reduced by 89.44% when compared with Sub-scenario 3.1. The system annualized cost obtained from Eq. (17) is reduced to $159,489 with a percentage reduction of 73.21% when compared with the existing scenario. The total system active-power loss is the same as that of Sub-scenario 3.1 since the motors in both sub-scenarios are completely disconnected from the distribution line and the active power loss is caused by the residential loads only. The total energy generated from the 16 PV panels; each rated 45.93 kW over the whole planting season equals 489.6 MWh obtained the same way as in Sub-scenario 3.1. The total quantity of energy required by all 16 motors during the whole planting season is 265.6 MWh the same as that of Sub-scenario 3.1 since the irrigation methodology is the same in both sub-scenarios. Therefore, the excess energy for this scenario equals 224 MWh. The excess energy in this scenario equals 45.75% of the corresponding total generated energy which representing a percentage reduction of 51.54% when compared with Sub-scenario 3.1. This scenario provides a better annualized system cost and a more reasonable PV panels size when compared with Sub-scenario 3.1, but still faces the problem of unutilized excess energy with a percentage excess energy that reaches almost 46%. Subsequently, utilization of this excess energy is due.

Fig. 11
figure 11

Water quantity in the water tank for Sub-scenario 3.2

Full size image

3.6 Results of Sub-scenario 3.3

To utilize the excess energy presented in Sub-scenarios 3.1 and 3.2, a grid-connected scheme is proposed in Sub-scenario 3.3 to sell the excess energy to the grid. In this scenario, the water tank storage is not required since the grid will act as a reserve for the excess energy from PV panels and as an alternate source to operate the motors when no sufficient power is generated from the PV panels. The resultant Pareto front of the multi-objective optimization problem along with the optimization solution’s utopia point discussed in Sect. 2.5.3.3 are shown in Fig. 12. The function ƒ1 and ƒ2 represent total system net present cost (NPCsys,tot) and the total system active-power loss (Pℓoss), respectively. The minimum point of the two functions is (103,040, 165,994). Calculating the Euclidean distance for the Pareto front points gives a minimum distance equal to 1598.6 at point (104,328.2, 166,940.6). The optimized PV panels size to achieve this utopia point equals 43.1 kW per motor with a total system net present cost (ƒ1) of $166,941 and a total system active-power loss (ƒ2) of 104,328 kW. This corresponds to a reduction in the total system active-power loss by 21.96% when compared with the existing scenario. The system annualized cost (Csys,anu) obtained from Eq. (17) is reduced to $41,401.2 with a percentage reduction of 93% when compared with the existing scenario. Figure 13 shows the sold and bought energy from the grid over the whole planting season. Bought energy when the power flows from the grid to the motors is presented by blue lines, while sold energy when the power flows from the PV panels to the grid is presented by green lines.

Fig. 12
figure 12

Pareto front for Sub-scenario 3.3 optimization problem

Full size image
Fig. 13
figure 13

Sold/bought energy to/from the grid in Sub-scenario 3.3

Full size image

3.7 Results of scenario 4

Scenario 4 introduces modification of the irrigation methodology to drip irrigation rather than surface irrigation. Climate, crop and soil data are used with the CROPWAT software to obtain the optimum daily quantity of water for drip irrigation. Using Eq. (11) and substituting for the motor rated power in the input power to the pump driving-motor (Pin), one can determine the hourly water flow rate (Qw). As the optimum daily quantity of water for drip irrigation is acquired from CROPWAT, one can easily determine the daily motor operation hours. The irrigation schedule represented by the daily water quantity per acre and the daily motor operation hours is presented in Fig. 14.

Fig. 14
figure 14

Irrigation schedule for Scenario 4: a water quantity per acre and b motor operation hours

Full size image

The total required water quantity per acre over the whole planting season is obtained by summing up the daily water quantity required per acre as presented in Fig. 14a to reach about 1609 m3/acre with a percentage reduction of 48% when compared with the existing scenario. The total required water quantity per acre is also reduced on using drip irrigation by 17.23% when compared with scenarios 1, 2 and 3 that utilizes surface irrigation. Each motor is operated for a total of 509 h over the whole planting season, which is the total sum of the daily motor operation hours as shown in Fig. 14b.

3.8 Results of Sub-scenario 4.1

The maximum operation hours for each motor presented in Fig. 14b does not exceed 7 h in any given day, a condition that can be met using only PV panels. Figure 15 shows the PV panels output power during the 113rd day of the planting season against the motor’s required power in that day. The 113rd day is again specially taken into consideration as it represents one of the days where maximum motor operation hours reach 7 h as shown in Fig. 14b and also to be the same day discussed in Sub-scenario 3.1 for comparison purpose. The optimum PV panels size obtained from the optimization problem discussed in Sect. 2.5.3.2 is 88.83 kW per motor with a percentage reduction of 79.6% when compared with Sub-scenario 3.1. It is clear that modifying the irrigation methodology from surface to drip solves the problem of excessive sizing of PV panels in Sub-scenario 3.1. This is because drip irrigation exists daily as depicted by Fig. 14a opposing to that of surface irrigation shown in Fig. 8a.

Fig. 15
figure 15

Hourly PV panels output power and motor’s required power during the 113rd day of the planting season

Full size image

The system annualized cost obtained from Eq. (17) is $581,765 and is equal to about 97.7% of the existing system cost. The total energy generated over the whole planting season from the 16 PV panels; each rated 88.83 kW for each motor equals 948.7 MWh. This total generated energy is calculated using Eq. (22) determining the PV panels output power. As this power is calculated hourly, the energy is equal to the power. Knowing both the total motor operation hours over the whole planting season as equal to 509 h and the motors rated power, the total quantity of energy required for all 16 motors during the whole planting season is 231.1 MWh. Therefore, the excess energy in this scenario equals 717.6 MWh, which represents 75.64% of the corresponding total generated energy. While the percentage excess energy in this scenario is reduced by 19.87% when compared with Sub-scenario 3.1, it’s still a considerably large value and must be reduced or utilized.

3.9 Results of Sub-scenario 4.2

A water tank storage is introduced with the PV panels. The optimum PV panels size and water tank size obtained from the optimization problem discussed in Sect. 2.5.3.2 are 38.27 kW and 632 m3 per motor, respectively. The system annualized cost obtained from Eq. (17) is reduced to $477,399 with a percentage reduction of 19.8% when compared with the existing scenario. However, the system annualized cost of Sub-scenario 3.2 is less than that of Sub-scenario 4.2 by 66.6%. This is mainly due to the components’ as well as operational and maintenance cost of the drip system when compared with that of surface irrigation as elaborated in Sect. 2.4. On the other hand, drip irrigation provides a better saving in water when compared with surface irrigation by up to 17.23% as discussed in Scenario 4. The total energy generated over the whole planting season from the 16 PV panels; each rated 38.27 kW for each motor is equal to 392.1 MWh being determined following the same way as described in Sub-scenario 4.1. The total quantity of energy required by all 16 motors during the whole planting season is 231.1 MWh the same as that of Sub-scenario 4.1 since the irrigation methodology is the same in both sub-scenarios. Therefore, the excess energy in this scenario is equal to 161 MWh. The excess energy as a percentage of the corresponding total generated energy from PV panels is equal to 41.1% in this scenario corresponding to a percentage reduction of 45.66% when compared with Sub-scenario 4.1 and a percentage reduction of 10.16% when compared with Sub-scenario 3.2. Figure 16 shows the water quantity in the water tank over the planting season demonstrating the charge and discharge of the water tank for Sub-scenario 4.2.

Fig. 16
figure 16

Water quantity in the water tank for Sub-scenario 4.2

Full size image

3.9.1 Results of Sub-scenario 4.3

Grid-connected scheme is proposed to utilize the evaluated excess energy in Sub-scenarios 4.1 and 4.2. The resultant Pareto front of the multi-objective optimization problem along with the optimization solution’s utopia point discussed in Sect. 2.5.3.3 is shown in Fig. 17. The function ƒ1 represents total system net present cost (NPCsys,tot) and the function ƒ2 represents the total system active-power loss (Ploss). The minimum point of the two functions is (99,762, 463,570). Calculating the Euclidean distance for the Pareto front points gives a minimum distance equal to 339.4 at point (100,092, 463,649.3). The optimized PV panels size to achieve this utopia point equals 33.5 kW per motor with total system net present cost (ƒ1) of $463,649 and total system active-power loss (ƒ2) of 100,092 kW. This corresponds to a reduction in the total system active-power loss by 25.13% when compared with the existing scenario. The system annualized cost obtained from Eq. (17) is reduced to $285,892 with a reduction percentage of 51.98% when compared with the existing scenario. Figure 18 shows the sold and bought energy from the grid over the whole planting season.

Fig. 17
figure 17

Pareto front for Sub-scenario 4.3 optimization problem

Full size image
Fig. 18
figure 18

Sold/bought energy to/from the grid in Sub-scenario 4.3

Full size image

From the previously discussed scenarios, the employment of a water tank storage with a stand-alone PV panels system in Sub-scenarios 3.2 and 4.2 significantly improves the system operational performance when compared with stand-alone PV panels without storage in Sub-scenarios 3.1 and 4.1. Sub-scenarios 3.3 and 4.3 provide the optimal minimal system annualized cost, total system active-power loss and water quantity when compared with the other scenarios. Sub-scenario 3.3 gives a lower system annualized cost when compared with Sub-scenario 4.3. On the other hand, Sub-scenario 4.3 gives a much more saving in total required water quantity per acre with further reduction in the total system active-power loss than Sub-scenario 3.3. Table 3 gives a summary of the obtained results. Figure 19 shows a comparison between the obtained steady-state proposed scenarios results. It is concluded from Fig. 19 that if the goal is to minimize the annualized system cost which is generally the goal for the farmers, then Sub-scenario 3.3 gives the minimum cost. On the other hand, if the goal is to minimize the water quantity for irrigation and the total system active-power loss which is generally the goal for the government, then Sub-scenario 4.3 should be implemented.

Table 3 Summary of the obtained results
Full size table
Fig. 19
figure 19

Comparison between the proposed scenarios obtained results

Full size image

4 Conclusions

Optimal operation of pump-based irrigation system powered from a real distribution network in Egypt (Karot line) is studied, and 4 different scenarios are proposed to improve the system performance. The obtained results conclude the following:

This research managed to decrease (1) the system annualized cost, (2) the distribution system’s total active-power loss and (3) the water quantity used while simultaneously meeting the crop water requirement. Accurate calculation of crop water requirement using climate, soil and crop data facilitates efficient use of water with water saving up to 37.29% when compared with the existing scenario of the Karot line. Applying a demand-side management technique (time of use tariff) on the operation of irrigation pumps helps in reducing the system annualized cost and the total system active-power loss by up to 54.73 and 12.04%, respectively, when compared with the existing scenario of the Karot line. Use of stand-alone PV panels without water tank storage was found applicable for drip irrigation & not for surface irrigation. While surface irrigation provides a lower cost when compared with drip irrigation, drip irrigation saves water by up to 17.23% when compared with surface irrigation. Use of grid-connected PV panels with surface irrigation provides the minimum possible system annualized cost where it is reduced by 93% when compared with the existing scenario of the Karot line. Use of grid-connected PV panels with drip irrigation gives the minimum possible total system active-power loss and irrigation water quantity used where they are reduced by 25.13 and 48.1%, respectively, when compared with the existing scenario of Karot line.

To extend this research for future work, it is possible to consider the following as a future research work:

Study the effect of applying demand-side management techniques to the residential loads. Integrating internet of things (IoT) with the irrigation system to achieve smart irrigation. Study the potential of a grid-connected hybrid renewable energy system rather than PV alone for irrigation.