Scientists in Russia have analyzed the most important technological advances achieved for solar water pumps over the past decades and have indicated the roadmap that future research should follow to expand their use and application. DECEMBER 2, 2020 EMILIANO BELLINI PV Mag
A solar-powered water pump in the United States. Image: US Department of Agriculture
A group of scientists from Russia’s Federal Scientific Agroengineering Center VIM (FSAC VIM) has provided a retrospective analysis of the most important research conducted on photovoltaic water pumping systems (PVWPS) over the past 55 years and has analyzed the most relevant issues and opportunities in their development, operation, and optimization.
“The [use of] PVWPS is an attractive alternative for irrigation and drinking-water supply in urban and rural regions of the countries [with] huge … solar energy [potential], where a considerable part of rural population lives in remote areas, such as India, China, other countries of Asia and Africa, in sunny, under-populated and mountainous areas of Russia, and other countries,” the researchers stated. “Unlike conventional water pumps, correctly-designed and sized PVWPS are capable [of providing] essential long-term savings.” https://www.pv-magazine-india.com/2020/12/02/the-unstoppable-trajectory-of-photovoltaic-water-pumps
EURACA, 13 to 16 April 2020, Athens: Review of photovoltaic water pumping system research
Under a Creative Commons license open access
Abstract: Currently a considerable and the most economically justified potential of making the solar energy conversion more efficient and making the solar energy more attractive for consumers involves the coordination of the entire equipment from generation to consumption while considering all this equipment as a unified system and taking the degree of satisfaction of the consumer needs as the measure of efficiency. A photovoltaic water pumping system (PVWPS) is the first and one of few types of ground photovoltaic systems where the consumption equipment was always considered from the onset as part of the system. So a retrospective analysis of PVWPS research is of particular interest.
This article contains the PVWPS research analysis beginning with the first research conducted in 1964–1966 to date. Discussed herein are the main problems that arose in the PVWPS development, operation, and optimization as well as options of their solution. Various areas of PVWPS improvement research were analyzed, including research involving photovoltaic array structure and photovoltaic modules interconnection; development and improvement of special structures of any components (pumps and invertors, etc.); improvement of photovoltaic array and electric motor interaction, electric motor and pump interaction against the backdrop of irradiance variation and in conjunction with consumption optimization and control improvement, to start with maximum power point tracing.
It would make sense to aim further research involving cost reduction, productivity gain, and PVWPS service life extension at the expansion of the fields of PVWPS application, improvement of individual PVWPS components, and a more efficient use of water that is drawn. One of the prospective lines of PVWPS improvement and upgrading is to include them in the system of end user consumption equipment and assess the system efficiency not at the pump output and not in the point of end consumption equipment but after the water consumption. Studying PVWPS with the account of the fact that, in real conditions, the average output power of photovoltaic array depends not only on irradiance but also on the combined effect of all environmental conditions (climatic parameters, current state of atmosphere and air, surrounding objects, etc.) is of great importance, today.
Photovoltaic water pumping systemEfficiencyPerformance analysisOptimization of components and controlsWater supplyIrrigation
AC alternating current
CP centrifugal pump
DC direct current
G irradiance (W/m2)
Gmin minimum irradiance level within the operating range (W/m2)
I motor-pump current (A) Iopt PV array current in maximum power point (in optimal point) (A)
MC mechanic characteristics with the reflecting surface
MPPT maximum power point tracing
PV array power (W)
Pmaxarray maximum PV array power (W)
Pmax STC maximum PV array power at STC (W)PV
PVWPS photovoltaic water pumping system
Q CP flow rate (m3/h)
R multiplier resistance (Ω)
STC Standard Test Conditions
V motor-pump voltage (V)
Vopt PV array voltage in maximum power point (in optimal point) (V)
WG working gear
Today an increase in photovoltaic water pumping system (PVWPS) efficiency using the inherent properties of photovoltaic equipment (varying the design and manufacturing technology of photovoltaic (PV) modules and PV cells) in many instances is technologically inexpedient and unprofitable for production in bulk. The PV modules and cells efficiency has reached such a level where the costs of introducing new, more efficient devices into production are often not sufficiently justified.
A considerable and more economically justified potential of making the solar energy conversion more efficient and also making the solar energy more attractive for consumers involves the coordination of the entire equipment from generation to consumption while considering all this equipment as a unified system and taking the degree of satisfaction of the consumer needs as the measure of efficiency (Izmailov et al. 2019 ). The more so as it is accepted for quite some time now that an efficient PV system, by virtue of its peculiarities, should comprise consumption equipment even if implicitly (as represented by all characteristics and parameters of components and detailed description).
Such an approach is in line with the general trends of increasing efficiency of any energy supply systems. The PV systems are the most critical case of energy supply systems with the most obvious necessity of having regard to all the particularities of energy conversion from generation to consumption. And a PVWPS is the first and one of few types of ground PV systems where the consumption equipment was always explicitly considered as part of the system (Izmailov et al. 2018 ).
PVWPS are supplied in their complete set of equipment in which photoelectric part and PV equipment (electric pump assembly) are sized to fit each other. In optimal design and manufacture conditions, this concept shall be applied to any type of PV systems dedicated not only for energy generation into power networks, in order to insure the maximum effectiveness of their application.
Many-year studies of PVWPS were focused mainly on achieving coordinated interactions between various components of photoelectric part and PV equipment of systems. At the same time, such studies for PV systems used in a majority of other applications, on a level of generic solutions in the field of compatibility between photoelectric part and PV equipment have not been intensive and significant.
In this connection, expanding the experience of PVWPS development onto PV systems of other types is of considerable interest. Detailed analysis of experience obtained in the course of PVWPS development is of great importance for designers and manufacturers of PV systems, as well as for enterprises operating such systems.
Therefore, a wide-scale use of PVWPS is rather desirable since they have a number of substantial advantages compared to other water supply systems, particularly in remote areas and locations with high solar energy potential. Solar water pump systems require minimum of maintenance and do not consume organic fuels. Unlike conventional water pumps, correctly designed and sized PVWPS are capable to provide essential long-term savings.
Taking this into account, a retrospective analysis of PVWPS research and improvement stages in terms of a consistent enhancement of components, improvement of interaction between the photovoltaic part of the systems and consumption equipment in combination with consumption accounting and optimization is of particular interest.
2. Photovoltaic water pumping system research
The research involving PVWPS began in 1964 (Lidorenko et al. 1965 ; Tarnizhevsky and Rodichev, 1966 ; Tarnizhevsky and Rodichev, 1968 ). In the first phase, several options of power supply to a motor-pump from a PV array were considered, including those with or without application of a buffer battery of any capacity (Belenov, and Tarnizhevsky, 1969 ). The analysis was performed while observing in all of the options the equality of average daily power consumption by the pump’s electrical motor with a 9-hour PV array generation per day; minimum difference between the actually generated PV array power Parray and maximum PV array power Pmaxarray at the minimum irradiance level within the operating range Gmin; and provided that equally uninterrupted water supply to the consumer is guaranteed. The options were compared by PV array rated power (maximum PV array power at Standard Test Conditions (STC)); by its relative cost and generated electrical energy utilization factor; by battery capacity and water tank volume; by the number and complexity of automatic control system components; and by the pump electric motor start-up conditions at Gmin.
The comparisons showed that the lowest relative PV array cost and the highest probability of commercial application is provided by the option at which Parray=Pmaxarray with G=Gmin and the battery is only used for power supply to the automatic control system. Hence the simplest schematic block diagram without a buffer battery was taken as a basis of further research: PV array–DC motor–working gear (WG). A centrifugal pump (CP) or centrifugal-vortex pump was used as WG the PV array and electric motor powers being commensurable. It was supposed that in the most commercial option the ratio of PV array rated power to the pump electric motors rated power should be equal to 1.2–1.25. The subsequent construction and operation of PVWPS having the same structure confirmed the results obtained in various countries. For example, in Akker and Lipp. (2004)  CPs with electric motor of rated power 1450 W were directly powered by a PV array of rated power 1800 W (the ratio being 1.24).
The research Belenov (1967) , Belenov and Voronetsky (1968) , Belenov and Voronetsky (1971) , Belenov and Shepovalova (2017)  showed that the power supply to the electric motor from the PV array of a commensurable power entails a substantial change in the parameters of mechanic characteristics (MC) n=f(T), i.e. dependences of the electric motor shaft speed on developed torque in all types of direct current (DC) motors: an independent, parallel or series excitation of magnetic field. Furthermore, the research showed that the MC parameters are influenced by the PV array rated power and, therefore, irradiance and its variation during daylight hours, as well as the connection circuits of PV modules.
As presented in Chilikin and Sandler (1981)  in conditions of stable electric grid voltage supply natural MCs of one and the same DC motor operating in modes of independent or parallel excitation are practically identical. They are absolutely linear, they exhibit insignificant rotation speed reduction with the rising DC motor output torque and they can be characterized by relatively high value of starting torque that 5 to 10 times exceeds normal values. The natural MC of serially excited DC motor, according to Chilikin and Sandler (1981)  data, has a hyperbolic shape with abnormally high rotation speed corresponding to operation conditions when the load on shaft is lower than 15% to its rated value. It also has a very high value of starting torque (higher than that of DC motor with independent and parallel excitation).
As carried out by research Belenov (1967) , Belenov and Voronetsky (1968) , Belenov and Voronetsky (1971) , Belenov and Shepovalova (2017)  these typical characteristics change significantly when DC motor is supplied from PV array of comparable power. Particularly, starting torque ratio deceases to values 0.3 to 1.5 compared to those corresponding to optimal DC motor’s operation mode depending on excitation method. The characteristics of DC motors with parallel and independent excitation appear to have completely different shapes and they become non-linear and smoother ones. DC motors with series excitation can operate in full-load loss mode while for torques exceeding their nominal values this dependence loses its hyperbolic character, and rotation speed falls abruptly up to zero. Fig. 1 shows the families of natural MC for three DC motors of practically equal rated power with different excitation systems (i.e. of series, independent and parallel types) fed from directly connected by the same PV array having rated power equal to 280 W, for solar radiation intensity G = 800 ± 10 W/m2.
Substantial change in shape of curves and varying parameters of DC motors’ mechanic characteristics for DC motors fed from PV array of comparable power was conditioned by:
– low starting current ratio (1.15 to 1.35) of electric motor armature defined by PV array short-circuit current in relation to its value for nominal operation mode,
– abrupt decrease of voltage across windings of electric motor with parallel excitation during startup,
– nearly linear dependence of short-circuit current and optimal current of PV array on the irradiance,
– variability of these PV array currents depending on incoming radiation parameters and on temperature of PV cells.
The MC research was model-based and carried out by solving nonlinear equation systems and using a graphic analytic method. The obtained results were affirmed experimentally with the electric motor powered by the PV array in natural sunlight conditions in Lidorenko et al. (1966) . When performing research, two types of specially developed samples of electromagnetic torque meter brakes were used. As a PV array, a photovoltaic installation with parabolic concentrators was used (concentration ratio 8.5; PmaxSTC=500 W; actually measured power level Parray=420 W with a direct irradiance of 800 W/m2 and optimal voltage of 34 V; open circuit voltage: 45.8 V; short-circuit current: 16.1 A, 6 strings of 7 PV modules in each).
Analysis of the PV array–electric motor–WG system for actual operation environment and conditions was carried out based on family of curves of MC for various values of irradiance (G = var), various values of PV array rated power and different variants of PV modules interconnection in PV array. These families of mechanic characteristic curves for G = var actualizes in real conditions following the daily cycle of irradiance, change in ambient temperature and wind conditions. With rising or decreasing solar radiation intensity, values of starting and optimal torque increase or decrease, proportionally. In these conditions, values of DC motor’s idle speed (i.e. shaft speed) without any external load do not vary at all (for parallel excitation) or vary insignificantly (for independent and series excitation). Some results of experimental research into the families of MC of motors with series, independent and parallel excitation (which are the same as in Fig. 1) are shown in Fig. 2 (Belenov and Shepovalova, 2017 ).
The subsequent research into the interaction of commutator DC motors and PV arrays, when they are directly connected, completely confirmed the results of our theoretic approaches, calculated data, and experiments (Appelbaum and Bany, 1979 ; Appelbaum, 1981 ; Roger, 1979 ; Singh and Wahi, 1981 ; Hsiao and Blevins, 1984 ; Abete et al. ).
Autonomously operated system comprising PV array, DC motor and WG is characterized by instability of maximal possible generated power Pmaxarray (for constant G value) and power Parray actually generated in these conditions depending on, first of all, radiation intensity varying during solar time. Real shape and parameters of DC motors’ mechanic characteristics, shape and parameters of the load characteristics of the WG (pump), as well as an eventual variation of water lift height and water consumption in the period of system operation significantly affect the values of Parray. So dependences Pmaxarray=Iopt×Vopt=f (G) and Parray=Iarray×Varray=φ (G) are not identical. For a wide range of irradiance variation, difference ΔP=Pmaxarray−Parray characterizing utilization rate of PV array’s rated power may attain rather high values, in the self-adaptive PV array–electric motors–WG system, in spite that for calculated irradiance level parameters values of Pmaxarray and Parray are practically equal to each other
This is thrown into sharp relief in Fig. 3 where the family of PV array current–voltage curves is shown within the limits of irradiance variation from 200 W/m2 to 1000 W/m2. The maximum power points Parray=Pmaxarray are dotted and two load characteristics I=ψ1(V) and I=ψ2(V) are shown in respect of different electric motors and pumps. The load characteristic No. 1 I=ψ1(V) meets the condition of Parray=Pmaxarray at G=200 W/m2 and the calculated point for the load characteristic No. 2 I=ψ2(V) is the equation of Parray=Pmaxarray at G=1000 W/m2 (Standard Test Conditions).
The analysis of the characteristics in Fig. 3 demonstrated that in a stand-alone and self-adaptive PV array–electric motors–WG system there is no possibility of fully using the PV array power except in calculated points. Owing to the natural daily variation of irradiance, the utilization factor Pmax STC and, therefore, PV cells efficiency decreases, at first inconsiderably and then substantially. Thus, for the load characteristic 1, at G=600 W/m2 the utilization factor of PV array power is only 0.59, while at G=1000 W/m2 it decreases to 0.48. For the load characteristic 2, when the irradiance goes down to 600 W/m2, the PV array rated power falls to 0.75 and at 400 W/m2 it falls to 0.26. The PV cells efficiency diminishes in the same proportion.
No PV arrays were used in the researched stand-alone and self-adaptive systems with full power throughout the light day and the high values of PV cells efficiency, as a matter of fact, were solely utilized during a small portion of light day periods. So for the purposes of more efficiently using the PV array power, increasing the daily energy generation, and, accordingly, the system’s rate of water output, it was proposed to forcibly control the value of Parray at irradiance variable bringing it up, as much as possible, to the value of Pmaxarray. As a result, a number of alternate technical solutions arose to eliminate or reduce the impact of this unfavorable effect.
In late 1960s it was proposed to employ a step (discrete) control of a stand-alone PV array–electric motors–WG system for securing its operation in near-optimum modes. Particularly, in Belenov and Tarnizhevsky (1969)  three ways were outlined and developed in general terms:
– series–parallel automatic switching of PV modules;
– use of PV array fractional load, i.e. connection PV array with two or more different power consumers;
– forced variation of the pump flow rate (and, hence, consumed power Parray) through a change of the pump impellers’ rotation speed by under-fluxing the exciting winding of power motor.
The method of discrete series–parallel switching of PV modules can be implemented if they are in sufficient quantity and there are no unused PV modules when the required PV modules connection circuits are created. The sufficient quantity of artificially created circuits may be 3 or 4 subject to the width of the operating range of irradiance variation. Efficiency of this optimization method for a stand-alone PV array–electric motors–WG system is illustrated in Fig. 4, Fig. 5. Fig. 4 shows three families of PV array current–voltage curves with three circuits of series–parallel PV modules connection (4×6,3×8,2×12) and at irradiance variation 250 to 950 W/m2. It also includes the load characteristic of the consumer that is motor-pump: I = ψ(V). The boundary values of irradiance at which circuit switching is required were determined based on equality of the power outputs of the two adjacent connection circuits. The calculations demonstrated that for the chosen PV array option assembled of 24 PV modules the conditions of a connection circuit switching should correspond to the data in Table 1.
A positive influence of using different PV module connection circuits on the PVWPS productivity was demonstrated in other studies such as Belenov and Aliev (1987) , Belenov and Shkatov (1989) , Matam (2018) . However the application of the circuit switching method was associated with a considerably higher consumption of copper conductors, use of supplementary interposing relays with power-plant contact units, growth of ohmic losses in wires and contact joints. The PV module circuit switch control turned out to be relatively expensive and with regard to PVWPS for low-debit wells the water output gain was not justified by purchase, adjustment, maintenance, and repair costs for such a device as a whole. Thereafter in Askarov et al. (1983)  it was proposed to substitute some contact joints by diodes functioning as switches upon current reversal which made the method more reliable by reducing the number of contacts but did not curtail the copper conductor consumption.
Table 1. Calculation results for switching moment of PV module connection circuits.
|Initial connection circuits (number of parallel strings × number of series|
connected modules in a string)
|6 × 4||4 × 6||3 × 8|
|Finite connection circuits (number of parallel strings × number of series|
connected modules in a string)
|4 × 6||3 × 8||2 × 12|
|Expected level of irradiance at the moment of switch (W/m2)||202||390||618|
Beginning in the mid-1970s several patents for PV module switching circuits were obtained with the aim of power take-off optimization, a number of articles on series–parallel PV module connection circuits were published (e.g. Turfler et al. (1980) ) but these solutions scarcely found commercial application, inter alia, due to substantially more sophisticated connection circuits, a higher consumption of copper conducting material, and a higher ohmic losses in wires and contacts.
The use of fractional load for optimizing the power take-off mode in a stand-alone PV array–electric motors–WG system showed that 2 consumers would be enough in such systems with a 1:2 power proportion and actuation algorithm based on irradiance increase 0.5; 1.0; 1.5 (Grigoyan, 1969 ). During research in the period of service testing at the watering point of Ovez-Shikh (Turkmenistan), two PV array–electric motors–WG system units were in operation drawing salt water from the well and pumping sweet water into the impounding basin (Lidorenko. 1969 ). The input power Parray was generally close to Pmaxarray. Considered in Gasque et al. (2020)  is a typical example of optimization through the distribution of PV array generated power in a pumping system composed of two equal power pumps operating in parallel. Nevertheless the connection of more than one consumers of different or equal power to the same PV array not always may be translated into practice.
The method of PV array–electric motors–WG system optimization through a forced change of the power motor shaft rotation speed driven by weakening the magnetic flux created by its exciting winding requires to use neither supplementary wires in PV array connection circuits nor another electric motors–WG unit. This technique is based on the quadratic dependence of the CP shaft input power on the pump impellers’ rotation speed with a relatively low value of its breakaway torque. The implementation of this method is simple and does not require any material additional costs because the speed of a DC motor of any kind of electromagnetic excitement may be easily controlled by means of magnetic flux weakening.
These optimization methods have been developed in Sharma et al. (2017) , Mahmoud et al. (2019) , Talbi et al. (2018) , Ksentini and Azzag (2019) , Xie and Deng (2006) , Cbergui and Benaissa (2015) .
Studies Chilikin, and Sandler (1981)  have shown that the DC motor speed control by magnetic flux weakening is the most energy saving and may be quite simply implemented in independent, parallel or series excited electric motors. Such control reliably ensures control limits in relation to the magnitude of velocity with a rated magnetic flux of 1.5:1 to 4:1 for an independent or parallel excited motors and to 2:1 for a series excited motors. Therefore, where necessary, the CP flow rate and, accordingly, power consumed from the PV array Parray may be forcibly increased 1.5 times at least.
While conducting experiments with a Kama centrifugal–whirling pump and DC motor SL-570 (see MC in Fig. 1, Fig. 2) in independent and parallel excitement modes, a magnetic flux weakening was achieved by introducing multiplier resistance R into the motor excitation winding circuit (Belenov, 2000 ). In the experimental PVarray–electric motors–CP system the dependences Pmaxarray=f(G) and Parray=φ(G) were researched at R = varand irradiance variation within 450–900 W/m2. The analysis of the obtained dependencies confirmed efficiency of the proposed input power control method. The experiment allowed to determine the values of multiplier resistance that secures, for a specific PV array–electric motors–CP system with a specific type of motors, the PV array operation near maximum power point and relevant parameters at the moment of switching from one multiplier resistance value to another. It enabled to appraise the resulting PV array efficiency growth and increase in the system’s rate of water output.
In Shermazanyan et al. (1970)  the adopted method of power control in a PV array–electric motors–CP system was implemented on the basis of the developed four-stage discrete speed controller of a tracking type ensuring a variation of the R value in the irradiance value function. Individual PV cells in short-circuit conditions were used as primary sensors of the controller. They were located in conditions being similar, by external influence, to the PV array operation conditions. Consequently, the sensors current values were virtually proportional to the PV array energetic capabilities.
In on-site conditions (Gelendzhik, Russia), the quantitative efficiency evaluation of automatic control of CP flow rate Q and increase in power take-off from the PV array by the electric motors–CP aggregate using the proposed technique was carried out at irradiance 450 to 805 W/m2 owing to which only two (of four) controller stages actuated (Belenov, 2000 ). The daily G, Q and Pmaxarray variation diagrams related to the parallel excited electric motors experiment are represented in Fig. 6. Dotted are the modes that could take place without control and with a one-stage actuation of the controller. The experimental data analysis showed that even with a two-stage control the PVWPS capacity by daily water supply volume was up 15% and the CP hourly flow rate was up 9%–19%. Thus, at 10 a.m. and 03 p.m. the water supply rose by 9.1%–10%, while at 11 a.m.–02 p.m. by 18–19.4%. The graphic analytic evaluation of the maximally possible control efficiency with irradiance variation in a wider range of 420–950 W/m2 and calculations showed that the daily energy generation and water supply increment in a controlled system could be 18–29%. Further research in this area led to the development of a host of devices used to regulate the efficiency of photovoltaic water pumping systems fitted out with commutator motors of all the three types of excitation, in particular Belenov et al. (2018) .
The method of raising the PVWPS performance by automatic discrete control of the shaft rotation speed in the electric motor–CP set within the irradiance variation function is quite efficient and simple to implement. However it was not translated into practice due to the then existing high mortality of the DC commutator motor’s brush assembly when used in outdoor apparatuses, as well as in connection with the change-over from DC motor-driven pumps to alternating current (AC) vibration pumps, more convenient in service.
To enhance the utilization efficiency of the PV array power even more, it was proposed in the research Zinger, and Braunstein (1981)  to combine a PV module connection circuit switch called ‘direct-current transformer’ (DCT) with magnetic flux variations of the electric motor whose excitation winding is connected to the PV array in parallel with the DCT. With small mechanical load changes, the PV array and electric motor operation was coordinated by relevant variation of the excitation winding magnetic flux. In case of considerable load variations, including a change-over from the start-up mode to a normal one, a DCT with a two or three-stage switching system is put into action. In the research author’s opinion, the proposed method is expedient to apply in water-pumping installations powered by PV arrays. Such coordination between the PV array optimal mode and load allows to prevent up to 50% loss of energy which the PV array could potentially generate during the daily irradiance and load variation.
One of the options of increasing electric energy generation in the PVWPS’s PV array–electric motor system by weakening the motor magnetic flux is considered in Aidan et al. (1983)  with regard to refrigeration units powered by PV arrays. The results of experimental corroboration of this technical solution carried out using a 750 W electric motor powered by PV array over an area of 7.5 m2 are represented in Fig. 7. To implement this method, it was proposed to build two supplementary (compound) magnetizing windings in the electric motors; the windings would be connected in series to the motor armature circuit and their magnetic flux in steady-state operation would creep counter-currently with respect to the main magnetic flux. When irradiance grew, the armature current stepped up and, hence, back induction of these supplementary excitation windings increased. So this weakened the total magnetic field, picked up the compressor’s electric motor speed, and increased the power take-off from the PV array. During the motor start-up and acceleration, these supplementary (compound) electric motors would temporarily actuate according to the main magnetic flow to facilitate a more reliable electric motors starting at low levels of irradiance, and only after full electric motor acceleration, on speed sensor signal, they would switch over to the proposed speed and power control mode (counter-currently in relation to the main magnetic flux) securing a smooth speed variation as a feedback link in accordance with the required trend.
In early 1980s in Roger et al. (1980)  for reducing the difference ΔP=Pmaxarray−Parray, it was proposed to install in a stand-alone PV array–load system, between PV modules and load, special means of an electronic adaptor so that the PV array could always run in the maximum power point bringing all the usable electric energy into load. A circuit breaker was chosen as a matching device base. Two types of breakers were studied and tested: a series one and a parallel one. Series breakers were used for a maximum power of 1.5 kW, while parallel breakers for the power range of 1 to 10 kW. The efficiency of a series matching device was 0.9 for 80 W load and 0.98 for 300 W or higher load; transistors were used in its technology. In case of a parallel breaker, two technologies were implemented: transistor and thyristor for each of which efficiency amounted on average to 0.93 with load power 1–5 kW.
Each matching device, irrespective of its type, should satisfy the following conditions: be powered by a PV array only; couple the PV array to its optimal impedance (i.e. overall pure resistance and reactance), whatever is the level of radiance; ensure high reliability and need no periodic tune-up; respond to the thermal bias of the PV module current–voltage curves caused by PV cells heating. The operating principle of the matching device was as follows. A power capacitor forming a part of the device charged until the PV array voltage achieved the value V1 and discharged when the voltage reached the value V2 the values V1 and V2 being in vicinity of the optimal value Vopt. As the value Vopt could change when the PV modules temperature varied, a platinum temperature sensor was employed with a feedback circuit for automatic adjustment of the values V1 and V2. Block diagram for parallel adaptor is shown in Fig. 8.
The series matching device had a very high efficiency and was suitable for powers below 1.5 kW. It prevented a loss of portion of the obtained power in poorly coordinated systems, such as water pumping systems, refrigerator compressors, and resistance loads. On the other hand, the interest in parallel matching devices involved the possibility of using high-voltage loads under a low-voltage connection circuit of PV modules that would reduce, firstly, ohmic losses in the motors and interconnecting wires and, secondly, would enable to obsolete high-voltage connection circuits of PV modules.
The further development of this technical solution resulted in the creation of individual devices Maximum Power Point Tracing (MPPT) controllers based on improved and more reliable electronic components. The research of recent years involving the use of MPPT controllers in PVWPS was devoted to the improvement of control algorithm, application of various up-to-date programs, extension of relations, and optimization of influencing parameters accounting (Rahrah et al. 2015 ; Khatib et al. 2017 ; Khazane and Tissir, 2018 ; Ehssein et al. 2018 ; Rafika et al. 2018 ; Yahyaoui et al. 2015 . They are widely applied in the systems fitted out with DC and AC motors and PVWPS circuits containing intermediate batteries or energy-intensive power capacitors (Das and Mandal, 2018 ).
A direct connection of a PV array and a DC motor is the simplest, cheapest, and most reliable technique of converting solar energy into mechanical one. That is why it is just this option which was chosen in the initial phases of developing PVWPS. The AC motors of the then existing motor-pumps were replaced with commutator DC motors suitable by power and speed the latter being generally installed on the surface for easier access to the electric motors terminals and brush assembly. Subsequently special pumps were developed with a submersible DC or AC motors, centrifugal-type floating motor-pumps with a DC motors, etc.
When creating a stand-alone PVWPS, for the purpose of minimizing the cost, the water delivery height was taken as a maximum of 30 m with the maximal daily supply volume of 7.5 m3 which met the requirements of the bulk of primary PVWPS consumers. To minimize the preliminary designing costs, series-produced pumps were used. As basic ones, series-produced Malysh-type vibration pumps were chosen which were capable of providing for the required water delivery heights and volumes and were easy to operate. Main efforts were made to develop an invertor converting the direct current generated by PV array into alternating current, suitable for powering the vibration pump, and matching the PV array parameters with the pump parameters, as well as develop PV modules for implementing the necessary connection circuits best suited, by current and voltage values, to operate as a specific packaged system ‘invertor–pump.’ For this purpose, the PV array parameters (power, current, voltage) were supposed to be differentiated subject to the range of daily water supply volumes (0.5 m3, 1.5 m3, 6 m3) and water lift heights (10 m, 20 m, 25 m, 30 m).
Beginning in late 1970s PVWPS were developed, manufactured, and put into operation in the Kazakh, Turkmen, and Armenian Soviet Socialist Republics (Bazarov et al. 1977 ). They contained a special invertor worked out for converting direct current into alternating current and configured to being powered by a high-voltage connection circuit of PV modules with the use of some PV modules for invertor power feeding. As a powering source, a PV array was employed consisting of 24 or 36 units of 10 W that created at the invertor inlet an operating voltage of about 200 V and auxiliary voltage of 20 V for powering the invertor circuit. With powering by a PV array, the invertor–pump set, when it was tested in Kazakhstan at a dug well with a static water level of 12 m, went into operation at G > 680 W/m2 and went out at G < 345 W/m2, while at G=800 W/m2 it ensured a pumping flow rate of 800 l/h. During the tests in Erevan, the same invertor–pump set powered by the same PV array, with G = 915÷950 W/m2 and a maximum water head of 15 m, ensured the water supply of 500–600 l/h. Nevertheless these invertor circuits required to single out a separate source from several PV modules with a 20 V electric tension for powering the circuit which thereafter led to their abandonment in terms of production and use.
Within the same period (since July 1979), the implementation of Project ‘Small-Scale Solar-Powered Pumping System’ started on the initiative of UN humanitarian departments and the World Bank (UNDP Project GLO/80/003, 1983 ). On the second stage of Project (from April 1981 to March 1983) comparative test were carried out followed by the evaluation of operability and effectiveness of 12 anticipatorily selected PVWPS produced by various manufacturers, and a general economical characterization of PVWPS was made. All selected PVWPS were divided into 3 categories according to their hydraulic parameters:
– category A applies to systems with 60 m3/day flow rate, for 2 m lift (water head), designed mainly for irrigation applications,
– category B applies to systems with 60 m3/day flow rate, for 7 m water head, designed for application in irrigation or/and water supply installations,
– category C applies to systems with 20 m3/day flow rate, for 20 m water head, designed for water supply of dwellings and farms.
All comparative tests were performed in conditions of sunny weather characterized by the aggregate solar radiation incoming onto horizontal surface 5 kWh/m2 per day. It made it possible to compare systems with high reliability based on their output parameters of water supply. Four installations were tested in each category. Operation in conditions of daily solar radiation of 4 kWh/m2 was permitted with 70% reduction of performance. In the course of tests, efficiency had to be not less than 75% of its design value, for water head variations within the range 75% to 150%.
One of the systems in category A failed to perform its functions (Manegon manufacturer, USA). Besides, only six installations of twelve (2 samples in each category) had complete set of equipment, i.e. were equipped with PV arrays. Therefore, in category A, only products of Solar Electric International (Malta) and installation by TPK (Canada) passed the tests. Systems of the first type were equipped with PV modules manufactured by Solar Power (USA), electric motor by AEG (Germany), floating CP by KSB (Germany) while the other system had PV modules and water pump of original manufacture and electric motor by Boston Gear (Canada) These two systems had daily performance 90 m3 and 16 m3 for PV arrays rated power 350 W and 210 W, respectively. In category B, two systems passed the complete test program. One was manufactured by AEG and equipped with original PV modules, electric motor produced by Engel and floating CP by Loewe. This system had daily performance 73 m3, for 614 W rated power of PV array. The other system manufactured by KSB having PV array with rated power 480 W had electric motor by AEG and original floating CP of type Aquasol-50M. It was able to supply 48 m3 water per day.
In category C, PVWPS manufactured by Grundfos demonstrated good performance. I was equipped with three-phase AC motor, inverter with controlled frequency and multi-stage rotary pump of original design. This system was capable to supply 33 m3 per day, for rated power of PV array of 840 W. Among systems that have passed the tests was that by Vmlamb (USA) with PV modules manufactured by Arco Solar having aggregate rated power 555 W. It was equipped with Honeywell DC motor and pump jack by Baker. Its daily performance was about 15 m3 to 16 m3.
Generally, average daily efficiency of PVWPSs that have passed the tests was within the range from 0.6% to 3.8%, for specified values of water head. According to these tests, the best results in category A demonstrated the system by Solar Electrical International with 2.3% efficiency, while in category B those were products by KSB (3.4%) and AEG (2.8%). In category C, the highest efficiency was obtained for systems manufactured by Grundfos (3.8%) and Vmlamb (2.5%). Of all of the 12 systems selected for tests only that by Grundfos included AC electric motor and DC-AC inverter with variable AC frequency. All other systems were equipped with DC motor connected to PV array either directly or via maximum power point tracking blok.
It was found out, in the course of tests that operational energy efficiency of subsystem ‘motor-pump’ and the effectiveness of factual use of PV array maximum power by this subsystem differ for different PVWPS architectures and for various water heads (McNelis, 1987 ). Floating and submersible units use most effectively the PV array maximum power. Such assemblies comprise DC motor (either with or without brushes) installed exteriorly and CP. Their mean value of this parameter, for water head 7 m, was 40% while the maximum value was as high as 60%. Other PVWPS architectures (for other values of water heads) were, in general, characterized by lower effectiveness with its mean values 30%, 35% and 40%, and maximum values 40%, 45% and 50%. As far as energy efficiency of ‘motor-pump’ subsystems is concerned, the best indicators were obtained for water heads 50 m and higher (but not exceeding 100 m). Their average daily efficiency was 35%, while its maximum value was 45%. For other water heads, average efficiency values were in the range from 25% to 32% while the best results were 30% to 42%.
Since 1985, two-year field test of PVWPSs have been carried out in various areas of Jordan, in the frames of the program on reliable round-the-clock water supply for habitants of deserted areas with the use of PV batteries (Odeh and Mahmoud, 1995 ). One of the principal goals of this program was to demonstrate ability of PVWPS to ensure high operability and reliability in environmental conditions of deserts compared to conventional water-pumping systems equipped with diesel engine drive. Measurements were performed automatically with the use of dedicated data samplers having the functions of statistical processing and storage of sampled data in data bases for further use and analysis. As an example, the following information on water pump station Shara El-Hasa may be of particular interest: PV array rated power having the surface area of PV modules 45.36 m2 was 6300 W. Its daily performance was 47 m3, for average water head 81.3 m. This performance value was achieved for average daily insolation 6.2 kWh/m2 during Spring–Autumn season of the year. The mean daily value of energy efficiency varied from 2.45%, in September, to 3.7%, in January.
In 1986, Electrotechnical department of State New Mexico (USA) initiated and carried out a survey by questioning the users of 111 photovoltaic water-pumping systems installed during the period of 1981–1985 of which74 stations had DC motors while the rest of them were equipped with AC electric motors operating from 60 Hz power supply (Schaefer, 1986 ). The PV arrays maximum power was in the range 110 W to 2080 W. This study had shown that 86.5% of installations retained their operability. 15 systems that were not operable of which 6 stations had failed converters, 4 stations failed to operate due to the damage of pump and 5 stations had problems that were not related to PVWPS. It was found out that AC PVWPSs were not so reliable compared to those powered from DC supply in which electric motor is connected to PV array either directly or via converter. Major problems were associated with charging control of battery used to power electric motor in cloudy weather (24%), with inverter operation (17%), with electric motor failure (11%) and with mechanical problems in pump (11%). Among all end users participated in this survey, 84% characterized PVWPS as a reliable system while only 13% gave a negative feedback. In 94% of all cases, systems were installed and commissioned by manufacturer representatives having adequate qualification. Similarly to research made in , the most effective were DC PVWPSs without buffer battery. Such systems included DC-AC converter, when it was necessary.
Fig. 9 presents application fields for various types of pumps having various types of electric motor for PVWPS depending on the required water head and daily supply rate by PDA Center Sandia National Laboratories (1987) . Presented diagram is a result of summarizing research and operation of PVWPSs, first of all, those with especially designed pumps (such as electric pump with submersible AC or DC electric motor, floating motor-pump of centrifugal type with DC electric motor, etc.). In Fig. 9, symbol ‘○’ denotes possible application fields of stand-alone PVWPSs with vibration pumps of type ‘Malysh’.
When analyzing the diagrams in Fig. 9, what calls attention to itself is the extensive area of application (both by water head and productivity) of rotor pumps integral with submersible AC motors or forming part of submersible alternating-current motor-pump sets. Grundfos are the pioneer in using AC motors in PVWPS completed with rotor pumps; this firm as far back as 1979–1983 presented such system with home-built pumps for comparative testing (UNDP Project GLO/80/003, 1983 ). The high metrics of this firm’s stand-alone PVWPS are also due to reliability of new invertors suitable for regulating the alternating current frequency as a command function of MPPT in order to increase the pump flow rate through a change of the electric motor shaft’s and pump impellers’ rotation speed subject to the level of irradiance and PV module temperature .
Inverter with AC frequency control  was used in the course of research Maranhão et al. (2016) , where a calibrated PV cell and a sensor module installed on the same plane of PV array were applied as an auxiliary device, along with MPPT. As a result, daily performance has increased from 21% to 23% compared to fixed PVWPS.
Analysis of technical, energy-related and hydraulic characteristics of various types of PVWPS by Metlov (1986)  has shown that the average daily rated output power of PV array required for lifting 1 m3 water to the height of 1 m3, during daylight hours, is about 2 W independently of the type of pump equipment. Besides, it was found out, in the course of practical operation of PVWPSs, that each 100 W of PV array output power can save from 120 kg to 150 kg of liquid fuel.
The use of high-efficiency PV modules, including bifacial modules and modules based on the high-voltage PV cells, expands the possibility of the increase of pumping systems efficiency (Shepovalova, 2018 ; Arbuzov et al. 2017 ; Shepovalova, 2017 ). The use of modules based on high-voltage cells also significantly increases the pumping systems’ compactness.
Inverter for vibration electric pump without separate power supply was developed in late 1980-s at Kharkov Polytechnic Institute based on the technical design specification by VIESH (Belenov and Eresko, 2002 ). One other specific feature of this inverter was that it generated 220 V (r.m.s.) series of pulses with 50 Hz frequency that were applied to the winding of vibration electric pump. Number of pulses in series and their repetition rate increased with PV array output power. Therefore, performance of pump grew automatically with solar radiation intensity and, vice versa, it fell with decreasing illumination. In case of power supply from PV array having rated power 330 W, the unit comprising inverter an vibration pump had the aggregate energy efficiency from 3.3% to 4.0%, 3.5% to 4.5% and 4% to 5%, for H=10m, 20 m and 30 m, respectively. For G in the range 790 W/m2 to 800 W/m2, system performance attained 600 l/h, 490 l/h and 360 l/h, for water head 10 m, 20 m and 30 m, respectively. Electronic module of such inverter was mounted into a special cylindrical casing designed for installing in the upper part of water well cap with the aim to stabilize environmental temperature thus ensuring safe operation of condensers and of the entire electronic assembly.
Transition to a wide-scale manufacturing of PV modules on the basis of PV cells with wafer diameter of 100 mm and larger (or pseudo-quadratic wafers with the size of side of 100 mm, 125 mm, 150 mm and larger) that, correspondently, resulted in increasing the PV modules rated power made it difficult to manufacture solar batteries with output voltage of about 160 V to 200 V having rated power less than 300 W by interconnecting large PV sells. That is why it was necessary to design inverters of advanced electronic circuitry and input voltage 24 V to 27 V avoiding application of additional specialized power supply, for inverter. Block diagram of one of such inverters Loskutov inverter I–200) is shown in Fig. 10. Voltage stabilizer is used to power pulse generator, control unit for switches VS1, VS2 and voltage comparing circuit. Voltage converter serves to raise the voltage to 200 V for charging storage condenser C2. Pulse generator forms impulses with frequency 100 Hz that are directed into the control unit of switches. The latter, in its turn, gives signals for alternative opening of switches VS1 and VS2 with frequency 50 Hz. Optothyristors play the role of switches VS1 and VS2. They provide galvanic isolation of the high-voltage output circuits of electronic assembly thus ensuring operation reliability and safety.
Condenser C3 and inductance of vibration pump winding form an oscillating circuit tuned to this frequency (50 Hz). Voltage comparing circuit initiates signal into the control unit of switches so that the latter generates impulses on condenser C1 of upper limit (30 V). Comparing circuit initiates the signal to stop generating these impulses when voltage falls to its lower limit (25 V). In this way, pulse transformation takes place whose repetition period depends on voltage and current in the output of PV array that are functions of solar radiation and temperature of PV cells. Threshold voltage levels of 30 V and 25 V correspond to the limits of maximum power modes of PV array in the range of solar radiation variations from 350 W/m2 to 850 W/m2, for PV module temperature in the range of 20 °C to 55 °C.
The inverter allows the system to operate in lower irradiance conditions. The use of the I-200 inverter makes it possible to increase the duration of the system average daily operation by an average of 1.5 times and consequently to increase the average daily output by about 15%–20% depending on weather conditions.
Application of the designed inverters made it possible to use low-potential solar energy G ≤ 300 W/m2, and to attain high effectiveness in normal solar radiation conditions. Therefore, PV array operation mode was close to its optimum. It means that PV array output power was always in the vicinity of its maximum point. Field tests of PVWPS comprising inverter, given in Fig. 10, vibration pump ‘Malysh’ and PV array with maximum output power 120 W carried out Tveryanovich et al. (1994)  in the period from May to October showed the following results:
– water lift to the height 10 m and 20 m began when G = 120÷140 W/m2 and G = 140÷180 W/m2; respectively.
– for G ≥ 800 W/m2, performance was from 818 l/h to 857 l/h and from 620 l/h to 645 l/h, for H=10 m and H=20 m, respectively.
– in conditions of variable cloud when solar radiation intensity, in midday hours, varied from 650 W/m2 to 750 W/m2, daily average performance was 1.8 m3 and 1.57 m3, for H=10 m and H=20 m, respectively.
– in cloudy whether, PVWPS operated, as well, but with lower effectively, and its daily average performance was in the range from 50 l to 100 l, for water head 10 m.
Initially, PVWPSs were studied only in stationary operation mode. At the same time, studies of transient processes occurring in the course of PVWPS operation are of essential importance. Such processes may take place whilestarting the system ‘electric motor–WG’ powered from PV array of comparable power, in conditions of fast changing of solar radiation, after system shutout. It happens, for example, when the water storage tank is full. Studying transient processes is also of great practical interest in terms of reducing energy loss and preventing negative effect of step-wise parameter change on PVWPS components. Partially these issues have been considered by Vissarionov et al. (1994) , and the results of this work were used to develop calculation methods for transient phenomena in PVWPS. Research was made with the application of mathematical simulation of dynamical systems. The principal results of this research were checked with the help of mathematical model and in the course of field tests with the use of physical analog. Development of computer facilities and software made it possible to study transient processes in greater depth. (Maranhão et al. 2016 ; Mokhlis et al. 2017 ; Ammar et al. 2017 ; Dridi et al. 2017 ).
By 2011, the experience of PVWPS operation and results of their studying were sufficient for establishing generic requirements for their designing and output parameters definition. Results of research and the specifics of system design were taken in consideration, generalized and systematized in the frames of standard IEC 62253:2011 ‘Photovoltaic pumping systems. Design qualification and performance measurements’ issued by the International Electrotechnical Commission and in GOST R 57903–2017 ‘Photovoltaic systems. Stand-alone pumping systems forwater supply. Performance determination. Selection and verification’ developed on the basis of IEC 62253.
The major scope of research of the last years in the field of PVWPS belong to scientific-application studies dedicated to the specifics of local use of PVWPSs in a number of countries (Rahrah et al. 2015 ; Kazem et al. 2017 ; Togola, 2003 , Abdellakh, 2006 ; Valer et al. 2016 ; Ibrakhim et al. 2016 ). Otherwise, they a related to the application of more advanced components equipment.
In recent years, considerable part of studies deals with problems of practical use of solar energy for irrigation (Valer et al. 2016 ; Belenov et al. 2016 ; Chandel et al. 2015 ; Shindel and Wandre, 2015 ; Sobor, 2017 ; Khaled et al. 2015 ; Reca et al. 2016 ; Brahmi et al. 2018 ). The advantages of solar energy application in water supply systems for irrigation are associated with the seasonal correlation between the annual maximum of solar energy income and that of water demand characterized by low probability of cloudy weather, during spring-summer period. The main problems of irrigation PVWPS: (1) this PVWPS needs to take account of the fact that demand for irrigation system water will vary throughout the year. Peak demand during the irrigation system seasons is often more than twice the average demand. This means that PVWPS for irrigation are under-utilized for most of the year; (2) attention should be paid to the system of irrigation water distribution and application to the crops; (3) irrigation PVWPS should minimize water losses, without imposing significant additional head on the irrigation pumping system and be of low cost. One of the most perspective technologies ofPVWPS application may be their use in combination with drip lines for watering various arable crops along ribs. Such systems comprise flat piping made of polymers. The main advantageous feature of such systems is that water head not exceeding 1 m to 3 m is required for water feed from open irrigation network. Besides, plants are protected against thermal shocks since water in such systems is supplied into ribs between the rows, and system may operate during daylight hours. In order to reduce energy consumption, application of low-pressure water pipes is advisable in combination with pump units having axial flow wheels.
This article contains the PVWPS research analysis beginning with the first research conducted in 1964–1966 to date. The research and practice has affirmed the energetic and economic expediency of PVWPS application for local water supply to any industrial and domestic consumers, especially in rural regions away from power lines. The using PVWPS is an attractive alternative for irrigation and drinking-water supply in urban and rural regions of the countries having huge potential of solar energy where a considerable part of rural population lives in remote areas, such as India, China, other countries of Asia and Africa, in sunny under-populated and mountainous areas of Russia, and other countries.
Application of PVWPS, in areas of the Arctic where solar radiation is rather high, during a certain period of the year, is not studied, so far. Low temperatures, high transparency of atmosphere, clean air and radiation reflection from snow and ice may contribute considerably to the efficiency of photoelectric modules. However, these areas are under-populated and difficult to access.
Various areas of PVWPS improvement research were analyzed, including research involving PV array structure and PV modules interconnection; development and improvement of special structures of any components (pumps and invertors, etc.); improvement of PV array and electric motor interaction, electric motor and pump interaction against the backdrop of irradiance variation and in conjunction with consumption optimization and control improvement, to start with MPPT.
It would make sense to aim further research involving cost reduction, productivity gain, and PVWPS service life extension at the expansion of the fields of PVWPS application, improvement of individual PVWPS components (including the use of highly-efficient photovoltaic modules, such as bifacial PV modules, and reduction in their performance deterioration), and a more efficient use of water that is drawn. One of the prospective lines of PVWPS improvement and upgrading is to include them in the system of end user consumption equipment and assess the system efficiency not at the pump output and not in the point of end consumption equipment but after the water consumption. Studying PVWPS with the account of the fact that, in real conditions, the average output power of PV array depends not only on irradiance but also on the combined effect of all environmental conditions (climatic parameters, current state of atmosphere and air, surrounding objects, etc.) is of great importance, today.
Izmailov A.Yu., Lobachevsky Ya.P., Shepovalova O.V.Complex energy supply systems for individual sitesEnergy Procedia, 157 (2019), pp. 1445-1455, 10.1016/j.egypro.2018.11.309ArticleDownload PDFView Record in ScopusGoogle ScholarIzmailov A.Yu., Lobachevsky Ya.P., Shepovalova O.V.Comparison and selection of off-grid PV systemsAIP Conf Proc, 1968 (2018) (2018), Article 030001, 10.1063/1.5039188CrossRefView Record in ScopusGoogle ScholarLidorenko N.S., Nabiullin F.Kh., Tarnizhevsky B.V.Experimental solar power plant [for powering electric pump]Geliotekhnika, 3 (1965)Google ScholarTarnizhevsky B.V., Rodichev B.Ya.Test results for solar energy installations with photoelectric converters [for powering electric pump]Geliotekhnika, 2 (1966)Google ScholarTarnizhevsky B.V., Rodichev B.Ya.Characteristics of water lifting system powered from solar energy plantsGeliotekhnika, 3 (1968)Google ScholarBelenov A.T., Tarnizhevsky B.V.Selecting optimal diagrams for solar generator connection to electric drivesGeliotechnika, 3 (1969), pp. 6-11View Record in ScopusGoogle ScholarAkker J., Ju Lipp.The power of human unity: Renewable energy in AurovilleLipp, Ju REFOCUS (2004), p. 28, 10.1016/S1471-0846(04)00139-8Google ScholarA.T. Belenov, Some characteristics of electric drives of solar water pump unit. In: Abstracts of scientific-methodical conf. on problem ‘Designing of systems of electric drives for agricultural machinery, Zernograd, 1967, pp. 114–5.Google ScholarBelenov A.T., Voronetsky B.B.Mechanical characteristics of DC motors powered from solar photovoltaic generatorElectrotechnics, 8 (1968)Google ScholarBelenov A.T., Voronetsky B.B.Mechanical characteristics of electric motors for agricultural drives powered from photovoltaic generator of comparable capacityAutomated electric drives in national economy. proc. all-union conf. on automated electric drives, vol. 4, Energiya Publ., Moscow (1971), pp. 173-175View Record in ScopusGoogle ScholarBelenov A.T., Shepovalova O.V.Investigation of DC motors mechanical characteristics with powered by comparable capacity PV arrayEnergy Procedia, 119 (2017), pp. 990-994, 10.1016/j.egypro.2017.07.132Google ScholarChilikin M.G., Sandler A.S.Geberal course of electric driveEnergoizdat Publ, Moscow (1981)Google ScholarLidorenko N.S., Nabiullin F.Kh., et al.Design of solar power units with photovoltaic convertersGeliotekhnika, 2 (1966)Google ScholarAppelbaum J., Bany J.Performance analysis of DC motor photovoltaic converter system – I. Separately excited motorSol Energy, 22 (5) (1979), pp. 439-445ArticleDownload PDFView Record in ScopusGoogle ScholarAppelbaum J., Bany J.Performance analysis of d.c. motor photovoltaic converter system – II. Series and shunt excited motorSol Energy, 27 (5) (1981), pp. 421-431ArticleDownload PDFView Record in ScopusGoogle ScholarRoger J.A.Theory of the direct coupling between d.c. motors and photovoltaic solar arraysSol Energy, 23 (3) (1979), pp. 193-198ArticleDownload PDFView Record in ScopusGoogle ScholarSingh S.N., Wahi A.R.Analytical model of powered separate excited d.c. motorJ Inst Eng (India) Elec Eng Div, 62 (1) (1981), pp. 17-21View Record in ScopusGoogle ScholarHsiao Y.R., Blevins B.A.Direct coupling of photovoltaic power source to water pumping systemSol Energy, 32 (4) (1984), p. 489ArticleDownload PDFView Record in ScopusGoogle ScholarA. Abete, E. Barbisio, R. Tommasini, A graphic-analytical method to study the direct coupling photovoltaic generators – DC motors. In: 10-th E.C. Photovoltaic Solar Energy Conf. 8-12 April 1991, Lisbon, Portugal, pp. 1037–41.Google ScholarRoger J.A., Pivot J., Gucher P.Optimization of the p.V. Array-load energy transfer by means of an electronic adaptatorRevue de Phys Appl, 15 (1980), pp. 603-607View Record in ScopusGoogle ScholarBelenov A.T., Tarnizhevsky B.V.Optimization of electric energy generation modes in stand-alone system comprising photovoltaic generator and DC motorCollection of reports of all-union conf. on application of solar energy (yerevan), section 2., VNIIT Publ., Moscow (1969), pp. 151-166View Record in ScopusGoogle ScholarBelenov A.T., Metlov G.N.Solar photoelectric water-pump systems(second ed.), VIESH, Moscow (2014)Google ScholarBelenov A.T., Aliev R.K.Some ways to improve energy production by photovoltaic array in installations of agricultural applicationGeliotekhnika, 6 (1987)Google ScholarBelenov A.T., Shkatov I.V.Effect of switching circuits of photoelectric batteries on performance of water pumping systemNTB on electrification in agriculture, Vol. 2, No. 65, VIESH, Moscow (1989)Google ScholarMatam M., Barry V.R., Govind A.R.Optimized re configurable PV array based PV water-pumping systemSol Energy, 170 (2018), pp. 1063-1073, 10.1016/j.solener.2018.05.046ArticleDownload PDFView Record in ScopusGoogle ScholarAskarov S.M., Grigoriev V.A., Irodionov A.E., et al.Device for powering loads with DC current. certificate of authorship USSR no. SU 1124854(1983)Google ScholarR.M. Turfler, T.G. Zanbarski, C.B. Rogers, Technique for aggregating cells in series and parallel. In: Proc. of the 14th IEEE photovolt. spec. conference, vol. 4,1980, pp. 518–22.Google ScholarGrigoyan R.C.Rational use of solar photoelectric conversion unit for power supply of several consumersReport selection of all-union conference on application of solar energy (Yerevan), VNIIT Publ., Moscow (1969), pp. 167-182View Record in ScopusGoogle ScholarLidorenko N.S.Current state of research in direct conversion of solar energy into electricityGeliotekhnika, 6 (1969)Google ScholarGasque M., Gonzalez-Altozano P., García-Marí E.Optimization of the distribution of power from a PV generation between two pumps working in parallelSol Energy, 198 (2020), pp. 324-334, 10.1016/j.solener.2020.01.013ArticleDownload PDFView Record in ScopusGoogle ScholarSharma U., Singh B., Jain C.Vienna converter fed two stage grid connected photovoltaic pumping systemProc. of the 2017 7th international conference on power systems (ICPS), Pune (2017), pp. 236-241, 10.1109/ICPES.2017.8387299View Record in ScopusGoogle ScholarMohamed El Mamy Mohamed Mahmoud, Moubarrack J.Y., Ehssein S., Yahya A.M., Mahmoud A.K., Youm I.Performance optimization and modelization of a photovoltaic pumping systemAm J Energy Eng, 7 (1) (2019), pp. 1-7, 10.11648/j.ajee.20190701.11View Record in ScopusGoogle ScholarTalbi B., Krim F., Rekioua T., Mekhilef S., Laib A., Belaout A.A high-performance control scheme for photovoltaic pumping system under sudden irradiance and load changesSol Energy, 159 (2018), pp. 353-368, 10.1016/j.solener.2017.11.009ArticleDownload PDFView Record in ScopusGoogle ScholarKsentini A., Azzag B., Bensalem A.Sizing and optimisation of a photovoltaic pumping systemInt J Energy Technol Policy, 15 (1) (2019), pp. 71-85, 10.1504/IJETP.2019.096619View Record in ScopusGoogle ScholarXie L., Deng W.Research on water output optimization of photovoltaic pumping systemEnergy Convers Manage, 47 (11–12) (2006), pp. 1449-1463View Record in ScopusGoogle ScholarCbergui M.I., Benaissa M.O.Strategy photovoltaic pumping system in scattered areaProc. of the 4th international conference on renewable energy research and applications, Palermo, Italy (2015), pp. 283-286Google ScholarBelenov A.T.Method of improvement of performance of photovoltaic water pumping unitGeliotekhnika, 1 (2000)Google ScholarShermazanyan Y.T., Tarnizhevsky B.V., Belenov A.T., et al.Power sensor of a solar power plant. author’s certificate of the USSR no. SU 259259(1970)Google ScholarBelenov A.T., Gusarov V.A., Shepovalova O.V.Device for improvement of effectiveness of electric energy use in DC electric drives. patent RF no. RU 2673460(2018)Google ScholarZinger Z., Braunstein A.Optimum operation of a combined system of a solar cell array and DC motorIEEE Trans Power App Syst, 100 (3) (1981), pp. 1193-1197CrossRefView Record in ScopusGoogle ScholarM. Aidan, Y. Hamam, J.P. Poirier, F. Rocaries, Optimal coupling of photovoltaic solar panels to compression cold room. In: Proc. of the 6th miami int. conference alternative energy source. conf. pap, 1983, pp. 10–3.Google ScholarRahrah K., Rekioua D., Rekioua T., Seddik B.PV Pumping system in bejaia climate with battery storageInt J Hidrogen Energy, 4 (2015), pp. 1-11, 10.1016/J.IJHYDENE.2015.04.048CrossRefGoogle ScholarKhatib M.F., Shaaban S., Sebah M.I.A.A proposed advanced maximum power point tracking control for a PV-solar pump systemSol Energy, 158 (2017), pp. 321-331, 10.1016/j.solener.2017.09.051Google ScholarKhazane J., Tissir H.Achievement of MPPT by finite time convergence sliding mode control for PV pumping systemSol Energy, 166 (2018), pp. 13-20, 10.1016/j.solener.2018.03.026Google ScholarC.O. Ehssein, A. Ba, N.O. Dah, M.I. Lam, D. Amadou, A.E. Hassan, Monitoring a maximum power point tracking photovoltaic pumping system, In: Proceedings of the 9th international renewable energy congress (IREC), Hammamet, 2018.Google ScholarR. Idrissi, A. Abbou, M. Mohcine, M. Salimi, A comparative study of MPPT controllers for photovoltaic pumping system. In: Proceedings of the 9th international renewable energy congress (IREC), Hammamet, 018).Google ScholarYahyaoui I., Nafaa J., Charfi S., Chaabene M., Tadeo F.MPPT techniques for a photovoltaic pumping systemProceedings of the irec2015 the sixth international renewable energy congress, Sousse (2015), 10.1109/IREC.2015.7110905Google ScholarDas M., Mandal R.A comparative performance analysis of direct, with battery, supercapacitor and battery- supercapacitor enabled PV water pumping system using centrifugal pumpSol Energy, 171 (2018), pp. 302-309, 10.1016/j.solener.2018.06.06ArticleDownload PDFView Record in ScopusGoogle ScholarBazarov B.A., Shaymerdangulyev G., Strebkov D.S.Studies of photovoltaic water pumping complexAshkhabad. Publ. of Acad. of Sci, Turkmen SSR (1977), pp. 323-327View Record in ScopusGoogle ScholarUNDP Project GLO/80/003Main report: small-scale solar-powered pumping system: the technology, its economics and advancement. – London, Swindon and Reading, UK(1983)Google ScholarMcNelis B.Photovoltaic water pumpingPresentations of the photovoltaics: Investing in development conf. 1987, New Orleans Louis, SEIA, Arlington, VA (1987)Google ScholarOdeh I., Mahmoud M.Field results of photovoltaic water pumping systemsGeliotekhnika, 1–3 (1995), pp. 231-240Google ScholarJ.F. Schaefer, Photovoltaic well pumping in New Mexico, in Proceedings of the 21st intersoc. conwers. eng. conf. San Diego, Calif., vol. 2 , 1986, pp. 1269–74.Google ScholarPDA Center Sandia National LaboratoriesWater pumping: The solar alternativePresentations of the photovoltaics: Investing in development conf. May 1987, New Orleans Louis, SEIA, Arlington, VA (1987)ν-75-ν-85Google ScholarRSI Renewable solar inverter – three-phase pumps(2020)https://www.grundfos.com/products/find-product/rsi.html [Accessed 25 June 2020]Google ScholarMaranhão G.N.A., Brito A.U., Pinho J.T., Fonseca J.K.S., Leal A.M., Macêdo W.N.Experimental results of a fuzzy controlled variable-speed drive for photovoltaic pumping systems: A reviewIEEE Sens J, 16 (9) (2016), pp. 2854-2864, 10.1109/JSEN.2016.2524530CrossRefView Record in ScopusGoogle ScholarMetlov G.N.Photovoltaic water-pumping plantsWater Supply Sanit Tech, 7 (1986)Google ScholarShepovalova O.V.Planar high-voltage photoelectric module. Patent RF No. RU 2686449(2018)Google ScholarArbuzov Yu. D., Evdokimov V.M., Shepovalova O.V.Angle-depended photocurrent characteristics of upper-limit efficient cascade PV cells on the base of homogeneous semiconductorEnergy Procedia, 119 (2017), pp. 980-989, 10.1016/j.egypro.2017.07.131ArticleDownload PDFView Record in ScopusGoogle ScholarShepovalova O.V.Fabrication process study for matrix silicon solar cellsAIP Conf Proc, 1814 (2017) (2017), Article 020077, 10.1063/1.4976296CrossRefView Record in ScopusGoogle ScholarBelenov A.T., Eresko A.B.Solar photovoltaic water pumping unit SVU-6-10Geliotekhnika, 3 (2002)Google ScholarTveryanovich E.V., Belenov A.T., Kidyashev Yu. K.Solar photovoltaic installations of small capacityJ. Energeticheskoe stroitelstvo, 12 (1994)Google ScholarV.I. Vissarionov, R.M. Buranov, Methods of calc ulating transient processes in pumping system powered from solar battery. In: Current problems of non-traditional power industry: Abstracts of int. sci-technical conference, 1-2 Sept. 1994, Saint-Petersburg, 1994, p. 82.Google ScholarMokhlis M., Ferfra M., Chraygane M.Nonlinear control of a photovoltaic pumping system under partial shadingProc. of the 2017 international renewable and sustainable energy conference (IRSEC), Tangier (2017), 10.1109/IRSEC.2017.8477413Google ScholarAmmar H., Benbaha N., Boukebbous S.E.P & O control of a photovoltaic pumping system to efficiency improvement using PSIMProceedings of the 2017 international renewable and sustainable energy conference (IRSEC), Tangier (2017), 10.1109/IRSEC.2017.8477404Google ScholarDridi T., Aouani N., Mami A.V/f controlled photovoltaic pumping system under LPV modelProc. of the 2017 international conference on control, automation and diagnosis (ICCAD), Hammamet (2017), pp. 516-521, 10.1109/CADIAG.2017.8075712CrossRefView Record in ScopusGoogle ScholarKazem H.A., Waeli A.H.A., Chaichan M.T., Mamari A.S., Kabi A.H.Design, measurement and evaluation of photovoltaic pumping system for rural areas in OmanEnviron Dev Sustain, 19 (2017), pp. 1041-1053, 10.1007/s10668-016-9773-zCrossRefView Record in ScopusGoogle ScholarTogola I.Solar energy use for water supply of rural areas in East Africa[PhD. Thesis](2003)Google ScholarAbdellakh B.Use of solar photovoltaic units for powering stand-alone water-pumping installations in Morocco[Ph.D. Thesis](2006)Google ScholarValer L.R., Melendez T.A., Fedrizzi M.C., Zilles R., Moraes A.M.Variable-speed drives in photovoltaic pumping systems for irrigation in BrazilSustain Energy Technol Assess, 15 (2016), pp. 20-26, 10.1016/j.seta.2016.03.003ArticleDownload PDFView Record in ScopusGoogle ScholarIbrakhim I.A.Kh.Energy complex on the basis of solar converters for water supply of consumers in climatic conditions of Iraq[PhD. Thesis](2016)Google ScholarBelenov A.T., Korolev V.A., Metlov G.N., Sokolsky A.K.Optimal composition and parameters of mobile solar photovoltaic water-pumping plants for irrigationInnov Agricult, 5 (20) (2016), pp. 193-198View Record in ScopusGoogle ScholarChandel S.S., Naik M.N., Chandel R.Review of solar photovoltaic water pumping system technology for irrigation and community drinking water suppliesRenew Sustain Energy Rev, 40 (2015), pp. 1084-1099, 10.1016/j.rser.2015.04.083ArticleDownload PDFView Record in ScopusGoogle ScholarShindel V.B., Wandre S.S.Solar photovoltaic water pumping system for irrigation: A reviewAfr J Agricult Res, 10 (22) (2015), pp. 2267-2273, 10.5897/AJAR2015.9879Google ScholarSobor I.Photovoltaic pump system design for small irrigationProc. of the 2017 international conference on electromechanical and power systems (SIELMEN), Iasi (2017), pp. 315-320, 10.1109/SIELMEN.2017.8123339CrossRefView Record in ScopusGoogle ScholarKhaled A., Boubakeur H., Essounbouli N., Ait-cheikh S.M., Cheknane A.Improvement of a photovoltaic pumping system for irrigation of greenhouses: case study for laghouat, algeriaProceedings of the 3rd international conference on control, engineering & information technology (CEIT), Tlemcen (2015), pp. 1-7, 10.1109/CEIT.2015.7233006CrossRefGoogle ScholarReca J., Torrente C., López-Luque R., Martinez J.Feasibility of standalone direct pumping PV system for irrigation in mediterranean greenhousesRenew Energy, 85 (2016), pp. 1143-1154, 10.1016/j.renene.2015.07.056ArticleDownload PDFView Record in ScopusGoogle ScholarBrahmi A., Abounada A., Chbirik G., Amrani A.Design and optimal choice of a 1.5 kW photovoltaic pumping system for irrigation purposesAIP Conf Proc, 2056 (2018) (2018), Article 020004, 10.1063/1.5084977CrossRefView Record in ScopusGoogle ScholarView Abstract