Solar photovoltaic-integrated energy storage system with a power electronic interface for operating a brushless DC drive-coupled agricultural load

1.1 Key contributions of this article

The following are the key aspects to which this article contributes:

• The presented system is utilized for parallel power management of a pre-installed solar PV system and an energy storage device employed for continuous operation.

• The described technique is especially applicable to those agricultural applications that require less power, such as grinders, chaff cutters, paddy threshers, winnowing fans, etc.

• The feasibility of the topology that has been implemented has been assessed based on a performance study in terms of various levels of solar radiation and power outages on the systems.

• The key goal of the implemented structure is to attain satisfactory performance while lowering costs and enhancing control over tuning between the available sources.

• The Maximum Power Point Tracking (MPPT) system aims to maximize the power output of a solar PV module by responding to fluctuating environmental and weather conditions to achieve both high accuracy and speed.

• Solar energy conversion is accomplished through the utilization of a DC–DC power converter that features magnetic components and semiconductor switches. However, the converter illustrates low current stress, minimal switching loss, and exceptional dynamic response.

• From a futuristic perspective, the adoption of Internet of Things (IoT) technologies could significantly enhance the efficiency of solar power generation and facilitate the maintenance of solar power plants. The following can be achieved by multiple pathways such as automated cooling, self-cleaning, defect detection, and tracking through data of temperature, irradiation, dust levels, peak power, short-circuit current density, and open-circuit voltage over cloud-based databases.

1.2 Organization of this article

This article is organized as follows: Section 2 describes the configuration of the solar-driven BLDC system with a battery as an energy storage device for uninterrupted operation in varying solar radiation and power failure situations. The design procedure for different components of the energy storage-incorporated BLDC system is demonstrated in Section 3. Section 4 deals with the control strategy employed for solar MPPT and for both the BLDC drive and with BDC for maintaining a constant DC link voltage. The simulation analysis of the proposed system under different irradiance conditions is shown in Section 5. Section 6 comes with the experimental validation of the system when subjected to different perturbations. Further, a comparative analysis of proposed systems with other existing works is carried out in Section 7. Finally, conclusive remarks are highlighted in Section 8.

3.1 PV array modelling

A solar PV system consists of PV cells, which is the building block of the system for generating electricity. The individual building blocks are commonly known as the single-diode model, which governs the power generation in a PV system. Figure 2 represents an equivalent single-diode PV cell structure made up of semiconductor material that works on the PV effect for electricity generation. The solar radiations energize the Si-based material for developing the potential difference across the PN junction of the PV cell (Narendra et al. 2020, Kishor and Patel 2022b, Cordeiro et al. 2020).

Figure 2

Single-diode equivalent PV circuit.

The output current of the PV equivalent circuit is proffer in the given equation as:

Thus, the total output current of the PV array combining numerous PV cells and modules:

where

where “ I ph ” is the photocurrent, “ I sat . ” is the reverse saturation current of the diode, “ q ” is the electron charge (1.602 × 10⁻¹⁹ C), “ a ” is the ideality factor of PV diode, “ k B ” is the Boltzmann constant (1.38 × 10⁻²³ J), “ T pv ” is the panel temperature, “ R se & R sh ” is the series and parallel resistors, and N ll & N sr is the number of parallel and series panels connected to form an array, respectively.

The physical changes of the parameters are shown in Figure 3 when the PV array is formed by parallel connections of panels. The characteristics of a constructed solar PV array are shown in the above figures under standard test conditions given in Table 1. The open-circuit voltage, short-circuit current, and maximum power are shown in the portrayed figures.

Figure 3

Standard SPV characteristics with variable irradiance.

3.2 Boost converter modelling

The PV system has to work in the maximum power point (MPP) area for proper operation of the BLDC drive, for which a DC–DC power conversion device has to be used. Here, a DC–DC boost converter is placed across the PV array using the P&O algorithm for tracking maximum power (Errouha et al. 2019, Shongwe and Hanif 2015, Batzelis 2019, Kishor et al. 2021). The requirements for a basic boost converter are inductor (L), capacitor (C), and duty cycle (D) for operating the semiconductor switch (Q) to maintain MPP and DC-link voltage. The DC–DC boost schematic diagram in Figure 4a and b shows the voltage across the diode during different phases of the boost converter's switching cycle.

Figure 4

(a) Schematic of DC–DC boost converter, (b) voltage stress across the switch and diode.

When the value of the inductor has the proper inductance, it is capable of handlining the peak and RMS values of currents over a wide range of input and output voltages (V _in & V _out), specifically taking the frequency of the circuit into consideration. The maximum permissible ripple current (ΔIL) is 30% of the maximum value of load current (IL) used in the converter at the least duty cycle. If f _s/w is the converter's switching frequency, to determine the value of the ideal inductor value should be:

where V out is the desired output voltage (in Volts), D is the duty cycle (ratio of ON time to total switching period, typically between 0 and 1),   I load is the load current (in Amp), and f sw is the switching frequency of the converter (in Hz).

The capacitors placed at the output terminal are seen as filters in a boost regulator to overcome output voltage ripples and high RMS current stress. The maximum allowable voltage ripples (ΔV _L) are dependent on the minimum value of output capacitance at a steady state in the boost converter. At given limits of output ripples (ΔV _L), the expression for calculating the minimum value of capacitance required for the converter is given as follows (Table 2):

The semiconductor component is the main switching element of a DC–DC power conversion device, which can withstand the maximum possible voltage and current dynamics. In this converter, a low-resistance N-channel MOSFET with a switching characteristic of [T _on = 14 ns, T _off = 50 ns] is used for operation. The electrical characteristics are listed in Table 3.

3.3 ESS modelling

The ampere rating of any battery calculation depends on continuous working hours, efficiency, maximum allowable depth of discharge (DoD), and state of charge (SoC). The state of charge is a percentage that tells you the level of battery charge relative to its full capacity and is expressed in an integer range from 0 to 100%. Depth of discharge is the inverse of state of charge, so a DoD value of 100% means the battery is empty, while 0% indicates that it is fully charged. SoC is usually used when talking about the current battery state, while DoD is more often used when talking about the battery’s history and how it degrades over repeated use. The following formula shows the relationship between SoC and DoD.

The battery bank is sized to have enough power to last for days where there is less sun, called the days of autonomy. The depth of discharge for deep-cycle batteries is typically 80%, but if you are using this design, it is taken to be 65%. The temperature correction factor is different at high or low temperatures. As an example, at low temperatures, battery efficiency decreases, so the temperature correction factor would be 0.9. Battery life is supposed to be 10 years, according to the manufacturer’s specifications. The required battery capacity in A h depends on the load given by:

Here, D s is the battery storage or autonomy days, η W is the connecting wire efficiency, η Bat . is the battery efficiency, and η temp . is the temperature correction factor for the battery used.

Batteries connected in parallel:

Batteries connected in series:

where B sel . is the selected battery capacity (A h), V nom _ Sys is the nominal system voltage, and V nom _ Bat is the battery nominal voltage.

Therefore, total series and parallel batteries connected in the battery bank:

3.4 Bidirectional converter modelling

The BDC enables to control the power flow from PV to battery and vice versa by adjusting the voltage VPV of the PV output at the MPP during normal operation. By using the above function, BDC can adjust the battery’s charging and discharging modes, which are dependent on the difference between the load and PV power (Wu 2022). The following figure shows the schematic design of a BDC, which has the ability to change the level of voltage between the input and output. It operates in two different modes: buck mode and boost mode. Buck mode operates whenever the Ppv power is sufficient to drive the load itself; this takes place when the switch Sw_1 is ON and the battery starts charging. When the power from PV is not able to drive the load demand, the boost mode gets in position by operating the switch Sw_2. This makes the DC link voltage be maintained at the constant mentioned voltage (Jadhav et al. 2018, Iqbal and Islam 2017). The swapping of power takes place by providing reasonable switching signals to operate switches present in the BDC for transient conditions (Figure 5).

Figure 5

BDC/battery structural design for uninterrupted power supply.

For smooth operation of BDC, the control circuit of the BDC works in two operating regions, as shown in Table 3. In addition to the above factors L and C selection also plays a vital role while in operation. The modelling of the inductor and capacitance values for a proper BDC is as follows:

3.5 BLDC-motor with electronic commutation modelling

For solar PV applications, BLDC motors are widely accepted over other conventional motors, i.e. induction, DC, and special motors. BLDC motors are taking over the traditional AC induction and DC motors due to their higher efficiency, minimal maintenance, reliability, and extended lifetime. They have a longer lifespan than traditional motors, as they do not have brushes that can wear out over time. This means they require less maintenance and have a longer operating life. BLDC motor drives can provide precise speed control over a wide range of speeds, making them ideal for applications that require precise control over motor speed. BLDC does not depend on the mechanical system (brushes) to control the current to the rotor for developing torque to drive the load (Darcy Gnana Jegha et al. 2020, Shanmugam et al. 2019, Jena 2019). Due to the absence of brushes, the BLDC is compact in size and weight with the same ratings as traditional motors. It makes to provide a high torque-to-weight ratio which almost reduces the energy consumption by 40% on average conventional motors, defining itself as an energy-efficient prime. With solar PV being an unreliable provider of consistent power, batteries have been ruled out as a potential alternative. To ensure reliable delivery of electricity, ESD is now being used as the other source of supply for driving the BLDC drive (Kumar et al. 2022, Kumar 2019) (Figure 6).

Figure 6

Schematic view of BLDC motor control.

Operating a BLDC motor uses a VSI for power to commutate the rotor by providing a switching logical pulse from the absolute positioning. The BLDC drive is also branded as an electronic commutated motor. The VS inverter basically works on the semi-conductor driver circuit, which provides six signals as shown in Table 4.

From the above figure, three stator windings and permanent magnets on the rotor make up the BLDC motor. Due to the high resistivity of both the stainless-steel retaining sleeves and the magnet, rotor-induced currents can be disregarded, and no damper windings need to be modelled.

The voltage equations of three windings can be written as follows:

where V a , V b , and V c are stator winding phase voltages; R a , R b ,  and  R c are the stator resistances per phase; i a , i b , and i c are the stator phase (or line) currents; L a , L b , and L c are the stator inductances per phase; L b a , L c b , and L c a are the mutual inductances between phases a, b, and c; e a , e b , and e c are the corresponding phase-back electromotive forces.

Let R a = R b = R c = R , L a = L b = L c = L , and L b a = L c b = L c a = M . The latter can be written as:

The back emf’s e a , e b , and e c are functions of " K e " is the back-EMF constant in V/rad/s, " θ e " is the electrical rotor angle in electrical degrees, and " ω m " is the mechanical rotor speed in rad/s and can be expressed as:

Due to their trapezoidal shape, the induced back emf’s lack sharp corners. The flux linkage derivatives produce the electromagnetic fields, which are a continuous function. The flux density function is smooth with no sharp edges, thanks to fringes. Thus, electromagnetic toque in BLDC is defined as:

Table 1 illustrates the motor specifications employed for this proposed system to operate the BLDC drive agriculture unit. The SPV array output power is optimized by the P&O MPPT technique. The design and control of the proposed system are elaborated as follows.

4.1 Proposed MPPT control for PV system

For harvesting maximum energy from the sun, the PV system here makes use of the P&O MPPT algorithm for tracking maximum output. For achieving MPPT, a DC–DC boost conversion system is interfaced in the mid-section of power flow from the source (i.e. PV system) to load. The algorithm adjusts the duty ratio ( D Boost ) of the DC converter used for operations by tracking the MPPs. Even though the solar insolation and cell temperature remain constant, the P&O algorithm continuously perturbs the system's operating point, forcing the PV array terminal voltage to vary about the MPP value. Identical current and power oscillations consequently take place when the perturbation frequency is low enough for the system to attain a steady state before the next perturbation; these typically oscillate between three levels.

The step size must also be large enough to prevent noise from interfering with and to enable the perturbation to result in a discernible change in array output power. The performance of the MPPT method is significantly impacted by noise, particularly at low step sizes where the system's response to noise is comparable to the MPPT perturbations. The perturbation is held in the same direction if ΔP is positive, which indicates that the algorithm is getting close to the maximum panel power. If the output power diminishes, however, the perturbation must be applied in the other direction.

Figure 7 shows that the MPP is attained by the expression ΔP = 0 in below displayed P&O algorithm’s flowchart.

Figure 7

Conventional P&O algorithm flowchart.

Evidently, the perturbation step size affects whether the MPP is reached since a large step size may result in a fast-tracking response but also high steady-state oscillation amplitudes. On the other hand, if the step size is small, tracking will be slower, and there will still be a small oscillation. The fundamental drawback of the P&O method is continuous oscillation around the MPP. Unfortunately, it is a cause for energy losses and an undesirable phenomenon termed MPPT drift, as described in Figure 8, where two power–voltage curves are presented at varying irradiation levels. The typical operation sequence is 1-2-3-2-1 indefinitely under steady-state circumstances with constant irradiation. Let us assume, however, that the irradiance increases sharply when the operational point shifts from 2 to 3. The operational point thus shifts to 4. When the algorithm detects a rise in PV power, the next perturbation is applied in the same direction, resulting in the culminating operation point being 5, which results in instability and power losses. The next section introduces the modifications to the standard P&O algorithm that aim to counteract the MPPT drift phenomena and fast convergence. By introducing an additional measurement in the course of the MPPT period, the proposed technique improves the assessment of the power fluctuation in contrast to the standard P&O.

Figure 8

Drift occurrence while irradiance variation in P&O algorithm.

Therefore, in this approach, the power variation brought on by the perturbation imposed and the one carried on by the change in irradiance may be isolated to some extent. The addition calculation is accomplished in the execution part of Figure 9 by inserting a decision step that investigates whether the criterion ΔP and ΔV, which resemble the MPP area for that irradiance value, is zero or not. Now, two positions [(Vx, Px), (Vx − 1, Px − 1)], which are spaced apart by a signal step change (ΔD), establish the area of convergence. Therefore, the new ΔD _New is 90% less than the previous one for searching in this area. The new duty cycle ΔD _New is described as follows:

and

Figure 9

The implemented P&O algorithm MPPT.

The algorithm's step size plays an important role in making it adaptable for approaching the MPP. The duty cycle is multiplied by a correction factor shown in equation (18) that is directly proportional to the PV module's voltage and inversely proportional to its current when the algorithm is approaching the MPP to the right. In a similar manner, when the algorithm is approaching the MPP to the left, the coefficient is directly proportional to the current and inversely proportional to the voltage. The methodology helps to monitor the MPP more accurately since the step size drops as it approaches the MPP, as depicted in the above flowchart in Figure 9.

4.1.1 BDC and ESS control

For uninterruptable power supply to the BLDC drive for continuous operation, a battery bank unit with suitable power flow management is described in the below section. This is performed by means of systematic control of the charging and discharging of batteries used during operation. The operational control for the battery is shown in Figure 10, which states the power flow, i.e. charging and discharging procedures during the operational events. The energy storage device is designed in such a manner that if the battery SoC is greater than the reference SoC, it gets discharged and charged whenever the SoC gets less than the reference SoC till the maximum SoC. Therefore, battery charging can be adjusted according to the energy management requirements and strategy. The SoC of the battery is made to operate at 40%, which can be modified depending on the mentioned data sheet of the used battery.

Figure 10

Schematic diagram of BDC control strategy for maintaining constant DC link.

The battery is treated as a simple DC voltage source with series negligible internal resistance (R _int.). The selected battery model is a R _int. Equivalent, which is widely used for dynamic studies due to its reasonable performance with simple configuration and low count of parameters. The battery state of charge and DC voltage are inversely proportional and can be calculated according to Coulomb’s method:

(19) SoC t = SoC t − 1 + ∫ t t − 1 η I Bat . 3 , 600 C Bat . ,

where t is the time in hours, I Bat . is the battery current in Ampere’s, C Bat . is the battery capacity in A h, and η is the Coulomb’s efficiency in %.

To achieve the above control for maintaining a constant DC voltage across the DC link, a PI-control BDC converter is employed in parallel with the PV source. The PI operates the two switches present in BDC by continuously observing the reference voltage of the DC link voltage across the DC link ( V r e f . = 48 V o l t s ) and battery reference output current ( I r e f _ B a t = 22 A m p s ). The PI determines the values of gains in terms of proportionality and integrality, i.e. K p & K i , respectively.

From Figure 10, switching a PI (Proportional-Integral) controller in a bidirectional converter system can ensure that the converter can provide power in the event that the PV system fails. The PI controller is used to control the power flow between the PV system and the battery through BDC switches, as shown above. When the PV system fails to provide power, the PI controller can be configured to switch the bidirectional converter into power supply mode, allowing it to provide power to the load which is accomplished by V _ref , V _{DC_link} voltages and battery current (I _Bat.).

The process of switching the PI controller involves changing the controller's setpoint. In normal operation, the setpoint is set to zero, which means that the bidirectional converter operates as a power flow controller, directing power from the PV system to the ESS when it is producing excess power and from the ESS to the load when the PV system is not producing enough power. When the PV system fails, the setpoint can be set to a positive value, which signals the bidirectional converter to operate as a power supply, providing power to the load. In addition to switching the PI controller, other measures may also be necessary to ensure smooth and reliable operation of the bidirectional converter in power supply mode. These measures may include adjusting the voltage and current limits of the converter, implementing voltage regulation, and monitoring the health of the batteries. In the end, switching the PI controller in a bidirectional converter system can provide a backup power supply in the event that the PV system fails. This can ensure continuity of power to the load and help maintain the stability of the operation of the BLDC drive for processing the mechanical loads.

4.2 Implemented control for optimal operation

The below flowchart describes the power management between a solar PV system and an energy storage device aided by BDC to drive BLDC-driven agricultural mechanical units. The PV power and load power are measured and compared so that the PV itself can drive the load or not. If PV power is greater than the load requirement, the power is transferred to both the BLDC drive and the battery for charging until it reaches rated capacity. When the PV fails to drive the load, then power transfer takes place from the energy storage device for its operation (Figure 11).

Figure 11

Flowchart describing the power management between solar PV and ESD to drive BLDC motor.

5.1 PV performance for different insolation levels

The performance of a PV system can be evaluated by studying its response to changes in the amount of sunlight it receives, or solar irradiance. By comparing the hardware results seen in Figure 13 to the expected performance under STC conditions, it is possible to determine the efficiency and performance of the system. The results indicate that the system is operating more efficiently than expected or that the cells are producing more power to their rated capacity for the time period, t0 – t1 s. Under slow irradiance change, the solar irradiance is gradually increased or decreased over a period of time, as shown in Figure 12 from t1 to t2 time period. This test is used to observe the response of the system's maximum power point tracker (MPPT) and to determine the voltage-current characteristics of the PV cells. The results of this test evaluate the system's efficiency and its ability to track the MPP under gradually varying conditions. In contrast, a step irradiance change involves rapidly changing the solar irradiance from one level to another, such as going from full sun to shade or vice versa. This test is used to evaluate the system's ability to quickly respond to changes in solar conditions and to track the MPP under rapidly changing conditions from t2 to t3 time period. The results of this test can be used to evaluate the system's dynamic performance and to identify any potential issues with the MPPT algorithm or system components.

Figure 13

Experimentally investigated out power waveforms of SPV system for varying solar irradiances.

Both slow irradiances change and step irradiance change tests are important for understanding the performance of a PV system and for optimizing its design and operation. When referring to PV performance under a rapid loss of sunlight or zero insolation. This can occur, for example, during a cloud passing over the PV panels or during a sudden power outage. The PV system is equipped with a DC–DC converter, and as a consequence, it will immediately shut down to generate power to prevent back-feeding from the system. When a system includes energy storage, the stored energy can be used to continue providing power to the system. In conclusion, the performance of a PV system under sudden zero insolation levels can be characterized during the time period from t3 to t4, which rapidly decreases the power generation and gets shut down as per the components and design of the system.

5.2 DC–DC boost and BDC performance under different conditions

A DC boost converter with a battery-operated bidirectional converter is used in the proposed architecture for the SPV system. The experimental results given in Figure 14 show a DC–DC boost converter with a battery-operated bidirectional converter that affords to convert the PV voltage to the required DC link voltage level to drive the BLDC drive for processing the mechanical units. The waveform depicts information about the performance and efficiency of the device, as there are no voltage spikes, overshoots, and undershoots to achieve the targeted voltage for the DC link voltage.

Figure 14

Experimentally investigated output voltage waveforms for SPV and DC–DC boost converter.

As shown from the above result in Figure 15, it is impossible to identify any issues with the BDC converter. The waveform illustrates the optimized performance of the converter by adjusting the duty cycle of the pulse width modulation (PWM) signal and the control parameters of the converter. The battery management for charge and discharge events is well controlled by the PI controller battery life and prevents damage to the battery while switching periods. These results show the optimized converter's operation, improve battery management, and ensure the reliable and efficient operation of SPV systems (Figure 16).

Figure 15

Experimentally investigated waveforms for ESD for maintaining a constant DC Link voltage.

Figure 16

Experimentally investigated waveforms for speed vs torque of BLDC motor.

5.3 BLDC-driven mechanical load units

Figure 17 shows a three-phase BLDC drive with a six-step commutation pattern. Operating under a constant load of 2 Nm and a speed of 3,000 rpm, the amplitude and frequency of the waveforms are maintained constant throughout the operation. However, the power outage does not create a big difference in speed and torque, which can be observed in a magnified view from the above figure. Overall, the waveforms for a BLDC drive operating under a constant load of 2 Nm and a speed of 3,000 rpm with variable input power are complex and define that the control strategy used by the drive is up to the mark. The BLDC drive controller is able to generate the commutation to maintain the desired speed and torque.

Figure 17

Experimentally investigated waveforms for BLDC motor 3-ϕ stator currents.

In order to produce a rotating magnetic field in the air gap, the stator current should be periodic in nature for creating a torque as output at the rotor shaft. Thus, the three phases of stator winding are energized by VSI, which produces a rotating magnetic field to drag the rotor towards itself. The above waveform in Figure 17 shows that they are periodic with the proper level of stator current as specified per motor design, the load, and the desired speed and torque. The motor's behaviour under different operating conditions requires a proper control of the stator current through the PWM technique to enhance the motor's performance and efficiency.

Figure 18 validates the performance of a Brushless DC (BLDC) motor whose back EMF waveform of a BLDC motor has a characteristic zero-crossing pattern that can be seen above as the motor accelerates to 3,000 rpm. The zero-crossing waveform of the BLDC motor's back EMF is a critical parameter that implies control and operation of the motor, permitting precise commutation and optimal performance. The pulses are used to determine the position of the rotor and to switch the motor's phases at the proper period for optimal performance and efficiency, which is observed in subpart (b) of the above figure.

Figure 18

Experimentally investigated waveforms: (a) BLDC motor 3-ϕ back EMF’s and (b) view of 3-ϕ back EMFs separately showing zero-crossing detection.