An Electromagnetic Wind Energy Harvester Based on Rotational Magnet Pole-Pairs for Autonomous IoT Applications

Abstract

In this paper, we report a wind energy harvesting system for Internet of Things (IoT)-based environment monitoring (e.g., temperature and humidity, etc.) for potential agricultural applications. A wind-driven electromagnetic energy harvester using rotational magnet pole-pairs (rotor) with a back-iron shield was designed, analyzed, fabricated, and characterized. Our analysis (via finite element method magnetic simulations) shows that a back-iron shield enhances the magnetic flux density on the front side of a rotor where the series connected coils interact and convert the captured mechanical energy (wind energy) into electrical energy by means of electromagnetic induction. A prototype energy harvester was fabricated and tested under various wind speeds. A custom power management circuit was also designed, manufactured, and successfully implemented in real-time environmental monitoring. The experimental results show that the harvester can generate a maximum average power of 1.02 mW and maximum power efficiency of 73% (with power management circuit) while operated at 4.5 m/s wind speed. The system-level demonstration shows that this wind-driven energy harvesting system is capable of powering a commercial wireless sensor that transmits temperature and humidity data to a smartphone for more than 200 min after charging its battery for only 10 min. The experimental results indicate that the proposed wind-driven energy harvesting system can potentially be implemented in energetically autonomous IoT for smart agriculture applications.

1. Introduction

With the tremendous progress of microelectronic industries over the past several decades, uses of microelectronic devices and sensors are being increasing widely used in different fields, such as wireless sensor networks (WSNs) and the Internet of Things (IoT), to collect various information and to monitor environment conditions [1,2]. Their uses in smart-farming/agriculture are also increasing day by day. Recently, the world government summit launched a report entitled ‘Agriculture 4.0—The Future of Farming Technology’ to increase the production of food (by expanding smart technologies in agriculture) by 70% by 2050 [3]. Supplying electrical power to WSNs and IoT devices in remote areas is a critical challenge for the development of smart technologies in agriculture. The required operational power for these devices ranges from a few microwatts to several watt levels which are typically powered by conventional electrochemical batteries [4]. However, these batteries have a short operational lifespan and are sometimes difficult to replace [5,6]. Additionally, the disposal of hazardous materials in most of those batteries exerts a potential threat to the environment [7]. Therefore, there is an increasing demand for alternative power sources to continuously supply clean and sustainable energy to these WSNs and IoT devices. Kinetic energy in the form of vibration/motion [8], thermal energy in the form of heat [9], solar energy in the form of light [10], etc., are promising alternatives and can be harvested to meet these needs.

Wind energy, as a kind of widespread kinetic energy, can be harvested to generate electricity for powering WSNs and IoT devices located in remote areas [11]. Complex designs, large volumes, high manufacturing and maintenance costs, and high electromagnetic interferences and drag ratios of conventional wind turbines mostly restrict their application to remotely operated WSNs and IoT devices [12]. Their turbines cannot be driven simply by smaller wind speed; however, a wind turbine that starts turning at a small wind speed (cut-in wind speed) and can produce a useful amount of power is highly desired. Therefore, many researchers are interested in miniaturizing wind-based generators for self-powered WSNs and IoT devices in practical applications.

Over the years, different wind energy harvesting methods have been reported such as fluttering, galloping, vortex-induced vibration, and wake-galloping [13,14,15,16]. Each method converts wind energy into the vibration of an elastic structure which is then converted to electrical energy via electromechanical transducers, e.g., electromagnetic, piezoelectric, and triboelectric. For instance, Quy et al. [17] reported a fluid-induced flutter-powered energy harvester by fabricating two electromagnetic micro-generators while testing in both a wind tunnel and real conditions. A theoretical model based on galloping oscillations was developed and experimentally validated with an electromagnetic energy harvester by Ali et al. [18]. Kumar et al. [19] investigated the vortex-induced vibration of a flexible cantilevered flapper placed in the wake of a rigid circular cylinder to harvest wind energy by piezoelectric means. Usman et al. [20] recently reported a reliable method of harnessing naturally available wind energy by a wake-galloping-based novel piezoelectric energy harvester. Note that the above-mentioned wind energy harvesters were based on vibration/motion; however, rotational wind energy harvesters using rotor-based transducers have also been reported [21,22,23,24,25].

Yang et al. [26] reported a rotational piezoelectric wind energy harvester by utilizing ball-impact-induced resonance in order to enhance the mechanical (in this case wind) to electrical energy transformation. A low-speed broadband rotational energy harvesting methodology using magnetically coupled piezoelectric beams and frequency up-conversion was investigated by Fu and Yeatman [27] and demonstrated a wind energy harvester with a self-adjusting function for increased air speed range [28]. Han et al. [29] developed, studied, and tested three-dimensional (3D)-printed rotational air-driven electromagnetic energy harvesters with simple structures, low construction costs, and fast prototyping. Wu and Lee [30] introduced a miniature windmill-structured electromagnetic energy harvester for the wireless monitoring of forest fires by analyzing various blade types to achieve the highest aerodynamic efficiency for this windmill-structure. Ahmed et al. [31] introduced a farm structure concept that included wind-driven triboelectric nanogenerators for large-scale energy harvesting based on the freestanding mode between two disks with patterns of micro-sized circular sectors. Recently, Wang et al. [32] developed a gravity triboelectric nanogenerator to convert wind energy into gravitational potential energy that was then transformed into steady electric energy. Besides these single-transducer-based wind energy harvesting approaches, researchers also reported multi-transducer-based approaches by hybridizing piezoelectric and electromagnetic [33], piezoelectric and triboelectric [34], electromagnetic and triboelectric [35], and even piezoelectric, electromagnetic, and triboelectric transducers [25].

Till now, most rotational wind energy harvesters have been realized based on piezoelectric and triboelectric effects. According to our knowledge, a lot of research has been reported based on vibrational electromagnetic wind energy harvesting, but no research has been reported by using rotational electromagnetic wind energy harvesting in recent years. In 2006, Weimer et al. [28] presented remote-area rotational electromagnetic wind energy harvesting for low-power autonomous sensors, with a focus on an anemometer-based solution. Recently, rotational hybridized wind energy harvesters have become of interest where an electromagnetic effect is combined with other energy harvesting effects (piezoelectric and triboelectric); however, these require complex power management electronics for practical applications [25]. Nevertheless, rotational electromagnetic energy harvesting is a promising method for wind energy harvesting because of its simple design and large output power, low maintenance cost, and the fact that it does not rely on breakthroughs of materials, generates a large harvesting current, is suitable for a large number of low-frequency vibrations, and offers stable performance for long time periods.

In this work, we demonstrated a magnetic pole-pair-based electromagnetic rotational wind energy harvester to harvest energy from the ambient environment efficiently. Here, an array of magnetic pole-pairs work as a rotor and a back-iron shield is used on one side of the rotor to increase the magnetic field density on the other side where an array of series-connected coils is placed. Finite element magnetic method (FEMM) simulations were performed to determine the magnetic fields and to optimize the design. Subsequently, a prototype device was fabricated and characterized. It generates 1.02 mW average power from 4.5 ms⁻² wind speed with an optimum load resistance of 200 Ω. A custom-designed power management circuit with a power conversion efficiency of ~ 70% was manufactured and utilized for a system-level demonstration of the energy harvester. In this case, a self-powered wireless environmental (temperature and humidity) monitoring system was demonstrated for application in smart agriculture that exhibits the capability and application scenario for wind energy harvesting.

2. Harvester Structure and Its Working Principle

To convert the wind energy (mechanical energy) coupled by the wind turbine into electrical energy, an electromagnetic transducer was incorporated within the rotor. The schematic structure of the proposed electromagnetic wind energy harvester (EMWEH) using magnet pole-pairs is depicted in Figure 1. It consists of a rotating disk-based structure containing six NdFeB (N52) magnet pole-pairs with a back-iron shield and twelve series-connected copper coils fixed in the lower part of the housing by using a printed circuit board (PCB). The back-iron shield on the rotor increases the magnetic flux densities on the opposite side where the coils are placed. The harvester also contains a fin at the back of the housing structure that acts as a direction controller to keep the turbine against the direction of wind flow. The entire structure is attached to a support stand with a bearing. By the influence of external wind energy (coupled via the turbine blades), the rotor (i.e., the magnet pole-pairs) starts rotating, and the relative motion between the magnet pole-pairs and the series-connected coils creates a change in the magnetic flux densities which, in turn, generates electromotive force (emf) voltage across the coil terminals by Faraday’s law of electromagnetic induction. When a load resistance is connected, current flows through the circuit and power is delivered to the load.

The mechanical system of the proposed EMWEH represents a rotational single-degree-of-freedom (SDOF) system excited by wind energy where the rotor is aligned parallel to the wind turbine using a shaft and rotates in accordance with the wind turbine. Therefore, the rotation speed of the rotor is the same as the turbine rotation speed (both are rotating together). The aerodynamic model of the turbine/rotor under no-load electric condition is then given by [36].

$$J\ddot{\theta }+c\dot{\theta }=\frac{P_{in}}{\dot{\theta }}$$

(1)

where $J$ is the inertia of the wind turbine/rotor and $c$ is the mechanical damping coefficient. $\dot{\theta }$ and $\ddot{\theta }$ are the mechanical angular velocity and angular acceleration of the turbine/rotor, respectively. $P_{in}$ is the input mechanical power produced by the turbine blades due to the moving air, which is given by [36].

$$P_{in}=\frac{1}{2}\rho A{v}^3\chi$$

(2)

where $\rho$ is the density of air, $A$ is the area swept by the turbine blades, $v$ is the wind speed, and $\chi$ is the power coefficient of the turbine/rotor. It is to be noted that the theoretical maximum value of $\chi$ is 0.59 for high-speed, two-blade turbines, and between 0.2 and 0.4 for slow speed turbines with more blades [36]. Now, the amount of emf voltage induced in the series coils due to the relative motion between the magnet pole-pairs embedded in the rotor and the coils is:

$$V_{emf}=-N\frac{d}{dt}\left[{{\displaystyle \int }}^{ }\overrightarrow{B}·d\overrightarrow{A}\right]=-NBl\dot{\theta }$$

(3)

where $N$ is the total number of coil turns, ${{\displaystyle \int }}^{ }\overrightarrow{B}·d\overrightarrow{A}$ indicates the magnetic flux through the differential element area $dA$ of the magnet-coil assembly, $B$ is the magnetic flux density, and $l$ is the coil length across the magnetic flux lines. The relative angular velocity $\dot{\theta }$ between the magnet pole-pairs (i.e., the rotor) and the coils can be determined by numerically solving the governing equation of the turbine/rotor motion driven by the wind, presented in (1). Finally, the amount of power delivered to a resistive load $R_l$ connected to the output ports of the EMWEH (neglecting the coil inductance as its impedance is significantly smaller at low frequencies, i.e., below 1 kHz) is:

$$P=\frac{1}{T}\underset{0}{\overset{T}{{\displaystyle \int }}}\frac{{\left(V_{emf}\right)}^2}{R_l}dt$$

(4)

In order to optimize the back-iron shield thickness to maximize the magnetic flux density of the magnet pole-pairs, we performed a 2D axis-symmetric simulation using finite element method magnetic (FEMM) software, as illustrated in Figure 2. Figure 2a shows the distribution of the magnetic flux lines of a portion of the magnet pole-pairs with no back-iron shield. In this case, the average flux density at 1 mm distance (where the coil is located) is ~0.51 T. As we see in Figure 2b, this value is increased by 15% when a 1 mm thick back-iron shield is attached behind the magnet pole-pairs. Figure 2c shows how the value of average flux density (at 1 mm distance from the magnet) changes with the change in the thickness of the back-iron shield. The plot indicates that the average flux density increases with the thickness of up to 1.1 mm and then saturates. Since the increased thickness of the back-iron shield can increase the mass of the whole prototype, therefore, a 1 mm thick back-iron shield was selected that gave a near-maximum flux density value. To maximize the flux linkage through all the layers of the coil windings, the smallest possible air gap (<1 mm) between the magnet pole-pairs and the coil array was maintained during prototype assembly (discussed in the next section).

3. Prototype and Experimental Setup

We fabricated, characterized, and demonstrated a prototype of the proposed EMWEH to drive a self-powered wireless smart agriculture monitoring sensor (temperature and humidity). The best suitable magnetic structure was chosen using FEMM analysis. Figure 3a shows the components of the EMWEH before assembly. It has a rotating disk-based structure (rotor) which is composed of a 3D-printed magnet-carriage to hold six NdFeB (N52) magnet pole-pairs and a back-iron shield (made of 1010 stainless steel) glued behind the magnet pole-pairs of the rotor. The magnets were unwaveringly attached to the circular carriage. Twelve self-supported coils (350 turns each) were wound using 44 AWG laminated copper wire and connected in series with the help of a PCB interconnect that worked as the coil carriage. It is to be noted that the magnet pole-pairs placed in the rotor had an opposite polarity; that is why the adjacent coils were placed in anti-mount orientations. The coil carriage was fixed to one side of a T-shaped pipe. To maintain the rotation of the turbine smooth, a shaft made of wood (Ø8 × 150 mm) was installed using two front-facing bearings mounted inside the pipe. The magnet carriage was fixed with the shaft, perfectly aligned with the coil carriage, and the turbine was attached to the shaft-end (near the magnet carriage). As wind can flow from any direction, a fin structure was integrated at the end of the T-shape pipe (opposite side of the turbine) that acted as a direction controller in order to align the EMWEH across the wind flow direction. One more bearing was installed at the bottom leg of the T-shaped pipe to ensure the rotation of the entire device for alignment with the wind flow direction. Finally, the fabricated EMWEH was assembled on a portable base stand, as shown in Figure 3b. Focusing on practical applications, the transducer of the EMWEH was enclosed by a thin plastic frame and sealed with commercial glue to protect the device from external environmental hazards such as dust, water splashes, and strong wind. The geometric parameter of each components used to fabricate the EMWEH prototype is given in

Table 1. Geometric parameter of each component of the EMWEH prototype.

Component	Parameter	Value
Rotor	Magnet dimension	Ø8 mm × 5 mm
	Carriage dimension	Ø44 mm × 6 mm
	Back-iron thickness	1 mm
Coil	Coil inner diameter	0.5 mm
	Coil outer diameter	8 mm
	Coil thickness	1 mm
	Number of turns (each)	350
	Coil resistance (each)	17 Ω
	PCB thickness	1 mm
Bearing	Inner diameter	8 mm
	Outer diameter	22 mm
	Height	7 mm
Turbine	Diameter	230 mm
	Number of blades	5

.

The fabricated and assembled EMWEH was first characterized at various wind speeds in a laboratory setup. Figure 4 shows the schematic and photograph of the experimental setup. The wind was generated and controlled by a traditional electric fan. The wind speed could be continuously varied from 0 to 4.5 m/s, and the speed was measured using a digital anemometer (BT-100-APP). To observe and record the output performance of the EMWEH, the transducer output was connected to a USB-powered Digilent Analog Discovery 2 (DAD) data acquisition system. A computer interface was used to record and store the measured data from the DAD. Later, the system-level performance of the EMWEH was demonstrated by using a custom power management circuit (PMC) and low-power sensor electronics, which is discussed in the following section.

4. Results and Discussion

4.1. Transducer Outputs

The voltages generated by the EMWEH prototype under various input (wind speed) and load conditions were observed and recorded for further analysis. At first, we studied the variation of open-circuit voltage outputs of the harvester at different wind speeds ranging from 2 to 4.5 m/s, as shown in Figure 5a. As seen in the figure, the open-circuit voltage is proportional to the wind speed that increases (peak-to-peak voltage from 0.89 V to 2.75 V) linearly with the increase in the wind speed. Figure 5b illustrates the time histories (waveforms) of the generated voltage for wind speeds from 2 to 4.5 m/s. It is clear that the amplitude of the generated voltage increases as the relative velocity between the magnet pole-pairs and the coils (in other words, the rotational frequency) is increased at higher wind speeds. More peaks are observed as the wind speed increases which, in turn, increases the energy conversion. We determined the frequency of the generated voltage waveforms by using the First Fourier Transform (FFT). The dominant response frequency increases from 30 Hz to 87 Hz when the wind speed increases from 2 to 4.5 m/s.

To examine the optimal performance, the rms (root mean square) voltage and the average power outputs of the EMWEH prototype were measured under various load resistances swept from 50 Ω to 1000 Ω with variable wind speeds ranging from 2 to 4.5 m/s. The rms voltages across various load resistances and the corresponding average power delivered are shown in Figure 6. From the plots, the optimum load resistance is found to be 200 Ω, which closely matches the resistance of the series connected coil resistance (204 Ω) of the series-connected coils of the transducer. The average power $P_{avg}$ was calculated from the experimentally obtained rms voltage $V_{rms}$ and the corresponding load resistance $R_l$ by using $P_{avg}=V_{rms}^2/R_l$. Both the rms voltage across optimum load resistances and the corresponding power values were increased (from 153 mV to 452 mV and from 126 µW to 1.02 mW, respectively) when the wind speed was increased from 2 to 4.5 m/s.

4.2. Power Management Circuit (PMC)

The EMWEH prototype generates a sinusoidal open-circuit voltage (peak to peak) on the scale of a few volts that changes with the wind speed. In order to supply a constant voltage to power any electronic load (e.g., WSN, IoT sensor, etc.), the ac output voltage waveform needs to be rectified and regulated to a stable dc voltage for direct use or to store in a secondary storage unit (e.g., a battery or super-capacitor). For that, we designed and manufactured a custom PMC on a compact two-layer printed circuit board (PCB), as shown in Figure 7. In the initial stage, a full-wave bridge rectifier with four Schottky diodes (PMEG2020AEA) converts the ac output of the EMWEH into a dc output. A BQ25570 nano-power boost charger and buck converter used in the second stage accepts the rectified dc voltage and stores the energy on an external storage element at a programmable voltage level. The buck converter of the BQ chip also regulates the voltage output to an external load. The programmed voltage for the storage unit was set to 4.2 V (standard for charging a lithium polymer (LiPo) battery [37]), and the output voltage was fixed at 3.4 V (required to operate the WSN).

The charging capability was investigated by connecting a capacitor at the battery terminals of the PMC, while the output of the EMWEH was supplied to the input of the PMC. As shown in Figure 8, the tests were performed with two different conditions: (i) vary the capacitance value by keeping the wind speed fixed at 4.5 m/s and (ii) vary the wind speed by keeping the capacitance fixed at 1000 µF. In both cases, no load was connected to the load terminals. As seen from the plots, the voltage curves across the capacitor are dependent on the performance of the BQ25570 chip. Due to the cold start process of the BQ25570 chip, initially (up to 1.5 V), the voltage curve was growing slowly. After the capacitor voltage reached 1.5 V, the voltage grew relatively faster. There was a programmed starting point of 3.3 V after which the PMC started outputting 3.4 V to the load terminals. Finally, the capacitor voltage reached 4.2 V and began cyclically charging and discharging around it.

The next characterization parameter of interest was the “efficiency” of the PMC. The efficiency $\eta$ is the ratio of maximum DC power $P_{DC}$ measured from the PMC to the maximum average ac power $P_{avg}$ the EMWEH could be delivered to its optimum load. In this case, DC voltages $V_{DC}$ across different resistive loads $R_L$ (swept from 1 kΩ to 60 kΩ) connected to the load terminals of the PMC were measured, while the EMWEH was operated at various wind speeds. The DC power values were then calculated by using $P_{DC}=V_{DC}^2/R_L$. Figure 9 shows the DC power versus resistive load for various wind speeds. The results show that the calculated maximum DC power values increase from 277 μW to 737 μW with the increase in the wind speed from 3 m/s to 4.5 m/s, respectively. However, the values of the optimum load resistance for various wind speeds are different and decrease with the increase in the wind speed. These numbers are in the range of tens of kΩ which are very much different to the optimum load resistance (200 Ω) of the EMWEH itself. This is not unexpected because when the wind speed increases, the current flow through the PMC also increases which changes (reduce) the internal impedance of the PMC. Now, the calculated power conversion efficiency of the PMC lies around 70% for the EMWEH operating under various wind speeds, as summarized in

Table 2. Summary of maximum AC power, DC power, and power efficiency for different wind speeds.

Parameter	3 m/s	3.5 m/s	4 m/s	4.5 m/s
Max. AC power (µW)	375	573	798	1022
Max. DC power (µW)	277	432	552	737
Max. efficiency (%)	73.8	75.4	69.2	72.1

.

4.3. System-Level Demonstration

To demonstrate the application of the proposed EMWEH, an autonomous wireless smart-agriculture monitoring (AWSAM) system using wind energy was developed, which can transmit signals to a smartphone interface via Bluetooth. The complete architecture of the AWSAM system together with the EMWEH is presented in Figure 10a. The custom-developed PMC was used to maintain a 3.4 V DC voltage to power up a low-power wireless sensor devices (an environment sensor, a microcontroller unit, and a Bluetooth module). To monitor the environmental conditions, a commercial environmental sensor (Mi Temperature and Humidity Monitor 2) was used as a physical sensor node which can display (it has a 1.5-inch LCD display) and transmits the results to a remote device (e.g., smartphone) via Bluetooth. For better crop yielding from remote areas, smart-agriculture has become an utmost important tool. A conceptual illustration of the real-time monitoring of the environment for smart agriculture using a smartphone interface is shown in Figure 10b. The experimental setup of the AWSAM system with the EMWEH, the PMC, and the wireless sensor devices is depicted in Figure 10c. Using the PMC, the generated electrical energy from EMWEH is continuously stored in the storage element (20 mAh LiPo battery) until the power necessary to operate the wireless sensor devices (connected to the load terminal of the PMC) is obtained. Figure 11 illustrates the real-time charging and discharging characteristics of the battery during the operation of the wireless sensor devices. It was observed that the wireless sensor devices started running when the voltage level reached 3 V after charging the battery for 5.27 min and kept running as long as the EMWEH was operating under a wind speed of 4.5 m/s. When the battery was charged up to ~3.1 V in about 10 min, the EMWEH was turned off. The wireless sensor devices kept running and transmitting the data for more than 200 min when the battery voltage dropped to 2.7 V. Note that, to connect the wireless sensor devices with a smartphone via Bluetooth, it required some extra power (nearly 3.6 mW) that was supplied from the energy stored in the battery beforehand. However, once the sensor was connected to the smartphone, it consumed very low power to sense and transmit.

5. Conclusions

In this article, we designed, fabricated, and successfully demonstrated a magnetic pole-pair-based electromagnetic rotational wind energy harvester to harvest energy from the ambient environment for wireless sensor applications. The electromagnetic transducer of the harvester was optimized to maximize the magnetic flux linkage between the magnet pole-pairs and series-connected coils for higher performance. A fabricated prototype harvester was tested in the laboratory setup under controlled wind flow conditions. At 4.5 m/s wind speed, it can generate 2.75 V peak-to-peak open-circuit voltage. At the same operating condition, it is capable of delivering a maximum 1.02 mW average AC power to an optimum load resistance of 200 Ω and a maximum 737 µW DC power to a 40 kΩ resistive load via a custom-designed power management circuit. The power conversion efficiency of the power management circuit is around 70%. Combined with the PMC, the proposed wind energy harvester is capable of powering a remote smart-agricultural monitoring system. A real-time wireless environmental monitoring system enabled by the proposed wind-driven energy harvester was demonstrated that continuously transmits temperature and humidity data to a smartphone.

Table 3. Performance comparison among various similar wind energy harvesters.

Reference	Harvester Type	Transducer Type	Wind Speed (ms−1)	Transducer Swept Area (cm2)	Avg. Power (µW)	PD Per Swept Area (µW cm−2)	NPD (µW cm−2/ms−1)
Kwon [38]	Translational	PE	4.6	60	250	4.2	0.9
Hu [13]	Translational	PE	7	52	39	0.7	0.12
Zhang [39]	Rotational	PE	7	684	800	1.2	0.17
Iqbal [40]	Translational	EM-PE	6	11.3	11.4	1	0.17
Zhao [32]	Rotational	EM-PE	7	110	1218	11	1.6
Howey [20]	Rotational	EM	4	8	56	7	1.7
This work	Rotational	EM	2	15.2	122	8	4
			3.5		572	37.6	10.7
			4.5		1020	67.1	14.9

summarizes the performance comparison of various similar small-scale wind energy harvesters, and the proposed device offers the highest power density/normalized power density per transducer swept area. It also has a simple design and is easy to fabricate with lightweight and inexpensive materials. In addition to potential autonomous agricultural monitoring applications, the proposed energy harvesting system can be used for traffic monitoring, forest fire detection, subway tunnel monitoring, etc. However, there is room for further improvement of the harvester performance within the constrained volume by optimizing the turbine structure as well as the transducer unit (to maximize the wind coupling and power generation, respectively) which will be reported in future work.