Design and Application of a Standalone Hybrid Wind–Solar ... - MDPI

applied sciences Article

Design and Application of a Standalone Hybrid Wind–Solar System for Automatic Observation Systems Used in the Polar Region Guangyu Zuo 1,2 , Yinke Dou 1, *, Xiaomin Chang 3 and Yan Chen 1 1 2 3

*

College of Electrical and Power Engineering, Taiyuan University of Technology, Taiyuan 030024, China; [email protected] (G.Z.); [email protected] (Y.C.) SOA Key Laboratory for Polar Science, Polar Research Institute of China, Shanghai 200136, China College of Water Resources Science and Engineering, Taiyuan University of Technology, Taiyuan 030024, China; [email protected] Correspondence: [email protected]; Tel.: +86-139-3464-6229

Received: 8 November 2018; Accepted: 21 November 2018; Published: 24 November 2018







Abstract: Continuous power supply for unmanned and automatic observation systems without suitable energy-storage capabilities in the polar regions is an urgent problem and challenge. However, few power-supply systems can stably operate over the long term in extreme environments, despite excellent performance under normal environments. In this study, a standalone hybrid wind–solar system is proposed, based on operation analysis of the observing system in the Arctic Ocean, the polar environments, and renewable-energy distribution in the polar regions. Energy-storage technology suitable for cold regions is introduced to support the standalone hybrid wind–solar system. Mathematical models of the power system at low temperature are also proposed. The low-temperature performance and characteristics of lead–acid battery are comprehensively elucidated, and a dedicated charging strategy is developed. A hybrid wind–solar charging circuit is developed using a solar charging circuit, a wind turbine charging circuit, a driver circuit, a detection circuit, an analog-to-digital converter (ADC) circuit, and an auxiliary circuit. The low temperature stability of charging circuit is test from −50 ◦ C to 30 ◦ C. Temperature correction algorithm is designed to improve the efficiency of the power supply system. The power generation energy of the power system was simulated based on the monthly average renewable energy data of Zhongshan Station. A case study was applied to examine the technical feasibility of the power system in Antarctica. The five-month application results indicate that the power system based on renewable energy can maintain stable performance and provide sufficient power for the observing system in low ambient temperatures. Therefore, this power system is an ideal solution to achieve an environmentally friendly and reliable energy supply in the polar regions. Keywords: hybrid wind–solar system; low-temperature energy storage; circuit design; application

1. Introduction The extent and thickness of Arctic sea ice have dramatically declined in recent decades, which is considered one of the most impactful changes to Earth’s surface [1,2]. The conditions of Arctic sea ice cover can be effectively measured using various means. The mass balance of sea ice has been recorded with noncontact instruments, such as upward-looking sonar and electromagnetic induction devices (EM-31) on a helicopter or ship [3–5], and contact instruments, such as ice-based buoys. Various ice-based buoys have been deployed for field observation of sea ice in the Arctic Ocean and Antarctica. Ice mass-balance buoys (IMBs) [6] and sea-ice mass-balance buoys (SIMBs) [7] enable the automatic

Appl. Sci. 2018, 8, 2376; doi:10.3390/app8122376

www.mdpi.com/journal/applsci

Appl. Sci. 2018, 8, 2376

2 of 23

observation of sea-ice thickness, snow depth, and sea-ice temperature profiles. For different depths of sea ice, the spectral intensity of solar radiation can be effectively measured using a fiber optic spectrometry system [8], and the design of the corresponding driver circuit was reported. All ice-based buoys are designed to achieve automatic and continuous observation of sea ice during annual cycles. However, very few buoys have achieved observation of sea ice for more than one year in the polar regions. The possible reasons for premature failure of buoys include: (1) decay of sea ice, (2) running out of power, (3) wild-animal damage, and (4) failure of the instrument itself. The robustness of observation equipment can be improved to protect against destruction by wildlife. In addition, the stability of the instrument has been greatly improved in recent years through continuous field tests and laboratory tests. In the Chinese National Arctic Research Expedition (CHINARE) from 2003 to 2017, the impact of premature sea-ice melting was overcome by locating the instruments in areas with suitable ice thickness and floe location [9,10]. The power-supply problem is the most potential challenge for observation equipment used in the polar region to work for a long time. Due to the special geographical location of the Arctic Ocean and Antarctica, the design of the power system used in the polar regions has special demands. Burning fossil fuels in the polar regions causes environmental pollution and is therefore restricted. Thus, new demands for clean and sustainable energy such as wind, solar, hydro, and ocean energy have been created. In the polar regions, renewable energy systems (RESs) based on wind and solar energy are considered a clean, inexhaustible, and environmentally friendly method. Solar- and wind-energy resources usually have fluctuations; thus, a hybrid wind–solar system is developed to solve the problem of energy fluctuations in the polar regions. Some work has presented research on hybrid wind–solar systems. For instance, based on the long-term recorded data of wind speed and irradiance, a calculation methodology of a PV array for a standalone hybrid wind/PV system was developed to improve the ability to find the optimum size of the PV array [11]. A report by Reference [12] indicates methodologies to model components and designs of hybrid renewable energy systems, and evaluates them. A case study of isolated site Potou in the northern coast of Senegal also described methodology based on a multiobjective genetic algorithm to realize the optimization of a hybrid solar–wind–battery system, and indicates the influence of load profiles on optimal configuration [13]. More accurate mathematical models for characterizing the PV module, wind generator, and battery were used to achieve the configurations of various types and capacities of system devices by changing the type and size of the devices’ systems [14]. A hybrid system was proposed and applied in many cities in Iraq, which can be considered a renewable resource of power generation. The simulation calculation of this system was done by software based on local meteorological data and the sizes of PV and wind turbines. The results illustrate that it is useful to utilize the solar and wind energy to generate power in harsh environments [15]. Wind and solar energy are unpredictable and weather-dependent in the polar regions. An energy-storage system (ESS) is necessary as a reliable back-up to the observing system in order to store excess energy and then release the power to guarantee periods when a net load occurs. Some integrated wind–PV–battery hybrid systems and power-management strategies were proposed to offer a lower cost and higher efficiency [16]. Isolated power systems built by hybridizing renewable-power sources (wind and solar) in Indian used a battery as appropriate energy storage [17]. Studies [18,19] have developed an energy-management strategy and new photovoltaic solar structure for a renewable-energy system for rural communities to meet load demands. The observation equipment used in cold regions often uses batteries as power storage. However, batteries are sensitive to low temperatures, which can decrease their capacity. Therefore, a hybrid wind- and solar-energy system is also temperature-dependent. In this study, a new observing system was designed to observe multiple parameters of polar environments based on analysis of the previous deployment of the observing system in the Arctic Ocean. Mathematical models of the PV array, wind turbine, and low-temperature characteristics of a PV panel are proposed. A lead–acid battery was selected for the standalone hybrid wind–solar system. Since battery capacity varies at different ambient temperatures, battery capacities must be calibrated

Appl. Sci. 2018, 8, 2376

Appl. Sci. 2018, 8, x FOR PEER REVIEW

3 of 23

3 of 23

based on charging and discharging experiments at low temperatures. Charging strategy for a battery in for a environments battery in polar also designed. A charging hybrid wind–solar charging circuit was polar wasenvironments also designed.was A hybrid wind–solar circuit was developed to provide developed to provide a reliable power supply to the observing system by adapting to the a reliable power supply to the observing system by adapting to the low ambient temperature inlow the ambient temperature in the Arctic Oceanwind–solar and Antarctica. Thiscircuit hybridiswind–solar charging circuit is Arctic Ocean and Antarctica. This hybrid charging composed of a solar charging composed of a solar charging chargingcircuit, circuit,a detection a wind-turbine charging circuit, a converter detectionADC circuit, an circuit, a wind-turbine circuit, an analog-to-digital circuit, analog-to-digital converter ADC circuit, and an auxiliary circuit [20]. We have comprehensively and an auxiliary circuit [20]. We have comprehensively evaluated the temperature dependence of the ◦ C to 30 ◦ C.laboratory evaluated the temperature dependence of− the with experiments from –50 °C to 30 circuit with laboratory experiments from 50 circuit The monthly average energy production °C. The monthly average energy production from the PV panel and the wind turbine from the PV panel and the wind turbine were simulated using monthly average meteorological were data. simulated using monthly average meteorological data. A case studyfield of the operationalinresults of the A case study of the operational results of the power system during observations Zhongshan power system during field observations in Zhongshan Station, East Antarctica was of examined to Station, East Antarctica was examined to evaluate the technical feasibility and stability the hybrid evaluate the technical feasibility and stability of the hybrid wind–solar system. wind–solar system. In these these contexts, contexts, this this paper paper focused focused on on exploring exploring aa standalone standalone hybrid hybrid wind–solar wind–solar system system for In for automatic observation used in in the the polar polar regions. regions. Section Section 22 describes describes the the automatic observation systems systems designed designed to to be be used considerations of the observing observing system system and and the the power power system. system. Section results of of the the considerations of the Section 33 gives gives the the results experiment on wind–solar experiment on the the low-temperature low-temperature characteristics characteristics of of batteries. batteries. The The design design of of the the hybrid hybrid wind–solar charging circuit is given in Section 4, and the performance of the circuit is evaluated in charging circuit is given in Section 4, and the performance of the circuit is evaluated in Section Section 5. 5. The results of the simulation of monthly average energy production from the PV panel and wind The results of the simulation of monthly average energy production from the PV panel and the the wind turbine, and and aa case casestudy studyof ofaafield fieldexperiment experimentininZhongshan ZhongshanStation, Station, East Antarctica presented turbine, East Antarctica areare presented in in Section 6. The conclusions are presented in final section. Section 6. The conclusions are presented in final section. 2. Observing-System and Power-System Considerations 2.1. Deployment of of Observing Observing System System in in the the Arctic 2.1. Deployment Arctic Ocean Ocean As As shown shown in in Figure Figure 1a, 1a, an an observing observing system system was was deployed deployed on on Arctic Arctic sea sea ice ice on on 99 August August 2016 2016 during the 2016 2016 Chinese Arctic Research This observing system was was designed during the Chinese National National Arctic Research Expedition. Expedition. This observing system designed by Institute of of China China (PRIC) (PRIC) and and Taiyuan Taiyuan University University of of Technology Technology (TYUT), (TYUT), by the the Polar Polar Research Research Institute which included an acoustic sounder (SR50A, Campbell, Camden, NJ, USA) and an underwater which included an acoustic sounder (SR50A, Campbell, Camden, NJ, USA) and an underwater sonar (PSA-916, Teledyne Teledyne Benthos, North Falmouth, Falmouth, MA, MA, USA) USA) to to measure snow depth depth and and ice sonar (PSA-916, Benthos, North measure snow ice thickness, a 4.5 m long thermistor string, a temperature sensor (HMP155A, Vaisala, Vantaa, Finland), thickness, a 4.5 m long thermistor string, a temperature sensor (HMP155A, Vaisala, Vantaa, Finland), a battery unit, unit, aa data a battery data logger logger developed developed by by TYUT, TYUT, and and an an Iridium Iridium system. system. The The power power of of this this system system was only dependent dependent on on the the battery battery unit unit (13.5 (13.5 V, V, 75 Blizzard weather weather occurred occurred three three days days after was only 75 Ah). Ah). Blizzard after the observation system was deployed (Figure 1b); these storms occur frequently in the Arctic Ocean. the observation system was deployed (Figure 1b); these storms occur frequently in the Arctic Ocean. In wewe cancan see that in which observing system wassystem deployed dramatically In Figure Figure1b, 1b, see the thatsituation the situation inthe which the observing washad deployed had changed, increasing the possibility of equipment damage. dramatically changed, increasing the possibility of equipment damage.

Figure 1. deployed in the Arctic Ocean during the 2016 National Arctic 1. (a) (a)Observing Observingsystem system deployed in the Arctic Ocean during the Chinese 2016 Chinese National Research Expedition. (b) Scene observing-system deployment. Arctic Research Expedition. (b) after Scenethree afterdays threeofdays of observing-system deployment.

The observing system observed the entire ice season for approximately 11 months. As shown in Figure 2, air temperature and power voltage from October 2016 to July 2017 are presented, which covered the period of the polar night. The sampling interval of air temperature and voltage was one hour.

Appl. Sci. 2018, 8, 2376

4 of 23

The observing system observed the entire ice season for approximately 11 months. As shown in Figure 2, air temperature and power voltage from October 2016 to July 2017 are presented, which Appl. Sci. 2018, the 8, x FOR PEERofREVIEW 4 of 23 covered period the polar night. The sampling interval of air temperature and voltage was one hour.

Figure 2. Air temperature power voltage of the observing system from October 2016 to July 2017. Figure 2. Air temperature andand power voltage of the observing system from October 2016 to July 2017.

temperature monitored observing system in the entire season included TheThe air air temperature monitored by by thethe observing system in the entire ice ice season included oneone peak (Figure 2). Beginning with the deployment of the observing system, air temperature continued peak (Figure 2). Beginning with the deployment of the observing system, air temperature continued to decrease until 18 February 2017. Subsequently, temperature continued to increase until 5 July to decrease until 18 February 2017. Subsequently, air air temperature continued to increase until 5 July 2017. Power voltagewas wasgreatly greatlyaffected affected by by air changes were relatively severe 2017. Power voltage air temperature, temperature,and anddaily daily changes were relatively and showed an oscillating downward trend. Low temperature was the main influencing factor of the severe and showed an oscillating downward trend. Low temperature was the main influencing power supply of the observing Although power voltage stayed above 13 V, which can factor of the power supply of thesystem. observing system. Although power voltage stayed above 13ensure V, the normal operation of the observation system, the continuous decline of power and the violent which can ensure the normal operation of the observation system, the continuous decline of power voltage could bring great hazards to the normaltooperation of operation the equipment. andfluctuation the violentoffluctuation of voltage couldpotential bring great potential hazards the normal of Thus, designing and developing energy-harvesting systems at low temperatures is vital for long-term the equipment. Thus, designing and developing energy-harvesting systems at low temperatures is observations of the polar regions. vital for long-term observations of the polar regions. 2.2. Environmental Analysis 2.2. Environmental Analysis A power system of the special observation equipment for the polar regions should fully consider A power system of the special observation equipment for the polar regions should fully the characteristics of the polar natural environment. China’s Zhongshan Station was selected as the consider the characteristics of the polar natural environment. China’s Zhongshan Station was one of the deployment locations of the observing system in Antarctica, and its monthly meteorological selected as the one of the deployment locations of the observing system in Antarctica, and its data (69.37◦ S, 76.38◦ E) in Antarctica in recent years are shown in Figure 3, including monthly average monthly meteorological data (69.37° S, 76.38° E) in Antarctica in recent years are shown in Figure 3, wind speed, radiation, day length, and air temperature. The meteorological data in Figure 3 were including monthly average wind speed, radiation, day length, and air temperature. The obtained from NASA’s atmospheric science data center. In Figure 3a, we can see that the distribution meteorological data in Figure 3 were obtained from NASA’s atmospheric science data center. In of wind speed was approximately parabolic, and monthly average wind speed for the whole year Figure 3a, we can see that the distribution of wind speed was approximately parabolic, and was greater than 7.50 m/s. The maximum monthly average wind speed (9.88 m/s) was in June. monthly average wind speed for the whole year was greater than 7.50 m/s. The maximum monthly The minimum monthly average wind speed (7.50 m/s) was in January. The monthly mean wind speed average wind speed (9.88 m/s) was in June. The minimum monthly average wind speed (7.50 m/s) was consistent with seasonal changes in the southern hemisphere. The average monthly wind speed was in January. The monthly mean wind speed was consistent with seasonal changes in the reached 9.02 m/s. Thus, wind resources around Zhongshan Station were relatively rich to establish southern hemisphere. The average monthly wind speed reached 9.02 m/s. Thus, wind resources power-supply systems for wind-power generation. Figure 3b shows the monthly average radiation around Zhongshan Station were relatively rich to establish power-supply systems for wind-power near Zhongshan Station. Radiation decreased from 6.05 KWh/m2 /day in January to 0.05, 0.00, and generation. Figure 3b shows the monthly average radiation near Zhongshan Station. Radiation 0.01 KWh/m2 /day in May, June, and July, respectively. After that, radiation rose from August until decreased from 6.05 KWh/m2/day in January to 0.05, 20.00, and 0.01 KWh/m2/day in May, June, and December, reaching maximum value (6.69 KWh/m /day) in one year. The average monthly radiation July, respectively. After that, radiation rose from August until December, reaching maximum value 2 /day. The results of monthly average day length are shown in Figure 3c. Similar was 2.49 KWh/m 2 (6.69 KWh/m /day) in one year. The average monthly radiation was 2.49 KWh/m2/day. The results to the results of the radiation shown in Figure 3b, day length decreased from January (24 h) to the of monthly average day length are shown in Figure 3c. Similar to the results of the radiation shown minimum value of 0 h in June. Then, day length increased from July until December, reaching the in Figure 3b, day length decreased from January (24 h) to the minimum value of 0 h in June. Then, maximum value of 24 h. As can be seen from Figure 1c, Zhongshan Station was in the polar day from day length increased from July until December, reaching the maximum value of 24 h. As can be December to January of the following year. In June and July, Zhongshan Station was in the polar night, seen from Figure 1c, Zhongshan Station was in the polar day from December to January of the and solar-radiation energy was equal to zero. Average monthly day length was 12.2 h. Figure 3d following year. In June and July, Zhongshan Station was in the polar night, and solar-radiation shows the monthly average air temperature near Zhongshan Station. The maximum monthly mean energy was equal to zero. Average monthly day length was 12.2 h. Figure 3d shows the monthly average air temperature near Zhongshan Station. The maximum monthly mean temperature was –7.13 °C in January. The minimum monthly mean temperature was –27.3 °C in July. Average monthly air temperature was –19.1 °C.

Appl. Sci. 2018, 8, 2376

5 of 23

Appl. Sci.8, 2018, 8, x PEER FOR PEER REVIEW Appl. Sci. 2018, x FOR REVIEW ◦C

temperature was −7.13 in January. The minimum monthly mean temperature was −27.3 Average monthly air temperature was −19.1 ◦ C.

◦C

5 of523of 23

in July.

Figure 3. Monthly meteorological data of Zhongshan Station, Antarctica. (a–d) Monthly wind speed, Figure 3. Monthly meteorological data of Zhongshan Station, Antarctica. (a–d) Monthly wind speed, solar radiation, day length, and air temperature, respectively. solar radiation, day length, and air temperature, respectively.

Figure 3. Monthly meteorological data of Zhongshan Station, Antarctica. (a–d) Monthly wind speed, Solar power is the most widely used field power supply in the polar regions, but the polar night solar Solar radiation, day length, air temperature, respectively. power is the mostand widely used field power supply in the polar regions, but the polar night limits the time of its application. Although the polar night is only two months, this period has the limits the time of its application. Although the polar night is only two months, this period has the lowest temperature of the year, which often leads to the obstruction of power-supply systems. When Solar is the of most used fieldleads power supply in theof polar regions, but the polar night lowestpower temperature the widely year, which often to the obstruction power-supply systems. When solar energy is minimal in June or July, this happens when wind energy is the strongest, so a hybrid isits minimal in June Although or July, this happens energy the strongest, so a hybrid limitssolar the energy time ofpower-supply application. polarwhen nightwind is only twoismonths, this period has the wind–solar system could bethe established. wind–solar power-supply system could be established. lowest temperature of the year, whichour often leads to system the obstruction of power-supply systems. When We also planned to deploy observing in the Arctic Ocean to observe sea ice We also planned in to deploy our observing system in the Arctic Ocean to observe sea ice parameters. solar energy is minimal June or July, this happens when wind energy is the strongest, so a hybrid parameters. The drift trajectory (Figure 4a) of an Ice Mass-Balance (IMB) buoy and the observed The driftpower-supply trajectory (Figure 4a) ofcould an Ice Mass-Balance (IMB) buoy and the observed temperature data wind–solar system established. temperature data (Figure 4b) from 10be September 2012 to 16 January 2014 are shown in Figure 4. (Figure 4b) from 10 September 2012 to 16 January 2014 are shown in Figure 4. Figure 4b shows that Figure 4b planned shows thattomaximum temperature wassystem 5.93 °C on temperature was ice We also deploy our observing in July the 2013. ArcticMinimum Ocean to observe sea maximum temperature was 5.93 ◦ C on July 2013. Minimum temperature was −45.06 ◦ C on February –45.06 °C on drift February 2013. Average was Mass-Balance –15.34 °C. parameters. The trajectory (Figuretemperature 4a) of an Ice (IMB) buoy and the observed 2013. Average temperature was −15.34 ◦ C.

temperature data (Figure 4b) from 10 September 2012 to 16 January 2014 are shown in Figure 4. Figure 4b shows that maximum temperature was 5.93 °C on July 2013. Minimum temperature was –45.06 °C on February 2013. Average temperature was –15.34 °C.

Figure 4. (a) Drift trajectory (green line) of an Ice Mass-Balance buoy from 10 September 2012 to 16 January 2014 in the Arctic Ocean. (b) Air temperature measured by the buoy.

Figureon 4. (a) Drift trajectory (green line)data of aninIce Mass-Balance buoythe from 10 September 2012 to 16 Based analysis of meteorological the polar regions, operating temperature range January 2014 in the Arctic Ocean. (b) Air temperature measured by the buoy. ◦ of the power-supply system in this study should be −50–30 C. The design of the observing system and the power-supply system should consider the effects of low temperatures and other Based on analysis of meteorological data in the polar regions, the operating temperature range meteorological elements. of the power-supply system in this study should be –50–30 °C. The design of the observing system and the power-supply consider the effectsbuoy of from low 10 temperatures and toother Figure 4. (a) Drift trajectorysystem (green should line) of an Ice Mass-Balance September 2012 16 meteorological elements. January 2014 in the Arctic Ocean. (b) Air temperature measured by the buoy.

Based on analysis of meteorological data in the polar regions, the operating temperature range of the power-supply system in this study should be –50–30 °C. The design of the observing system

Appl. Sci. 2018, 8, 2376

6 of 23

2.3. New Observing System and Power System Based on the analysis of Sections 2.1 and 2.2, we designed a new observation system and power system, and adjusted the observation parameters of the sensor. The accuracy and temperature range of the sensor were fully considered. Table 1 summarizes the various sensors used in our observing system on the Arctic and Antarctic ice cover. The observing system of ice mass balance consisted of a wind-speed and a direction sensor, a temperature and humidity sensor (HMP155A, Vaisala, Vantaa, Finland), an atmospheric-pressure sensor (CS106/PTB110, Vaisala, Vantaa, Finland), and a 10 m long thermistor string and two spectroradiometers (TriOS, TriOS Optical Sensors, Oldenburg, Germany). The wind-speed and direction sensor was developed by TYUT, designed to overcome the effects of frozen environments by relying on self-heating strings around the sensor. The 10 m long thermistor string was also designed by TYUT and successfully observed the temperature profiles of shallow ice cover. Table 1. Sensor information of the observing system. Sensor Name Wind speed Wind direction Air temperature Humidity Atmospheric pressure Thermistor string Spectroradiometer

Performance Temperature range: −60–30 ◦ C Accuracy: 0.5 m/s Temperature range: −60–30 ◦ C Accuracy: 0.3◦ Temperature range: −60–10 ◦ C Accuracy: 0.1 ◦ C Temperature range: −60–10 ◦ C Accuracy: 2%RH Temperature range: −60–30 ◦ C Accuracy: 0.6 hPa Temperature range: −60–30 ◦ C Accuracy: 0.1 ◦ C Temperature range: −60–30 ◦ C Accuracy: 5%

Sensor Model TYUT TYUT HMP155A, Vaisala, Finland HMP155A, Vaisala, Finland CS106/PTB110, Vaisala, Finland TYUT TriOS, TriOS Optical Sensors, Germany

As shown in Figure 5, the power-supply system of the observing system of the ice mass balance involved in this study is equipped with a power generator that includes the PV array and small wind turbine (SWT), an energy-storage system (a low-temperature battery pack), an end user (load), and a control station (hybrid wind-solar charging circuit). The whole power system is a standalone system. This power-supply system was specifically designed for our small ice mass-balance station. In the design process of the hybrid wind–solar charging circuit, three subcircuits were developed: (1) SWT charging circuit: A three-phase rectifier circuit is required to convert the three-phase alternating current generated by the small wind turbine into a stable direct current. After the rectified and filtered direct current passes through the chopper circuit, it becomes a stable direct current and supplies power to the load or the battery. (2) Solar charging circuit: It is necessary to design a suitable DC chopper circuit and implement the control strategy for the solar charging circuit by adjusting the conversion circuit. (3) Detection circuit and ADC circuits: These subcircuits are mainly used to monitor solar-battery voltage and current, SWT voltage and current, battery voltage, and charging current. Battery capacity generally decreases at lower ambient temperatures. Therefore, the capacity of battery at low-temperature ranges (−50–30 ◦ C) should be completely understood to make the observing system work properly and stably. The performance of the ADC, detection, and auxiliary circuits at low temperatures should be considered, which may affect the charging strategy of the charging circuit.

Appl. Sci. 2018, 8, 2376

7 of 23

Appl. Sci. 2018, 8, x FOR PEER REVIEW

7 of 23

Figure 5. Schematic of the power-supply system.

2.4. Mathematical Modeling of the decreases Power System Battery capacity generally at lower ambient temperatures. Therefore, the capacity of battery at low-temperature ranges (–50–30 °C) should be completely understood to make the 2.4.1. PV Array observing system work properly and stably. The performance of the ADC, detection, and auxiliary In this study, polycrystalline PVbe panels were designed and affect assembled by TYUT. In the Arctic circuits at low temperatures should considered, which may the charging strategy of the Ocean and Antarctica, all PV panels were designed to be positioned in a fixed direction parallel to charging circuit. the ground, which can be considered an optimum angle for PV panels installed in the polar regions. 2.4. Mathematical Modeling of TYUT the Power The key specifications of the PV System panel are presented in Table 2. 2.4.1. PV Array

Table 2. Key specifications of the TYUT PV panel.

In this study, polycrystalline PV panels were designed and assembled by TYUT. In the Arctic Characteristics Value Ocean and Antarctica, all PV panels were designed in a fixed direction parallel to Open circuit voltage (Vocto) be positioned 21.6 V the ground, which can be considered an optimum angle Optimum operating voltage (Vmpfor ) PV panels 18 V installed in the polar regions. Short circuit (Isc ) 0.916 The key specifications of the TYUT PV panelcurrent are presented in Table 2. A Optimum operating current (Imp ) 0.833 A 15 W Maximum power at STC 1 (Pmax ) Table 2. Key specifications of the TYUT PV panel. 1

Standard test conditions.

Characteristics Value Open circuit voltage (V oc ) 21.6 V The relationship between output of solar power and radiation intensity based on an individual Optimum operating voltage 18 V PV cell can be obtained from the Equation (1):(Vmp) Short circuit current (Isc) 0.916 A = P − P T − T · 0.4% · Rβ 1 β 2Aβ 3 (1) [ ( ) ] OptimumEoperating current (I mp ) 0.833 max max stc solar 1 Maximum power at STC (Pmax) 15 W where Esolar is the output of solar power (Wh/day); Pmax is the maximum power at standard test 1 Standard test conditions. conditions (15 W); T is the ambient temperature (◦ C); Tstc is the ambient temperature at standard test 2 /day); β is the soiling losses factor conditions (25 ◦ C); R isbetween the average radiation (KWh/m The relationship output of solarintensity power and radiation intensity based on an individual 1 0.97; β2 can is the point coefficient PV cell benon-MPPT obtained from the Equation0.96; (1): β3 is antireverse diode coefficient 0.98.

2.4.2. Wind Turbine

Esolar =  Pmax − Pmax (T − Tstc )  0.4%   R 1 2 3

(1)

The wind turbine from TYUTpower was employed theis power system of this study. The key where Esolar is the output of solar (Wh/day); in Pmax the maximum power at standard test specifications of the TYUT wind turbine are presented in Table 3. conditions (15 W); T is the ambient temperature (°C); Tstc is the ambient temperature at standard test conditions (25 °C); R is the average radiation intensity (KWh/m2/day); β1 is the soiling losses factor 0.97; β2 is the non-MPPT point coefficient 0.96; β3 is antireverse diode coefficient 0.98. 2.4.2. Wind Turbine

Appl. Sci. 2018, 8, 2376

8 of 23

Table 3. Key specifications of the TYUT wind turbine. Characteristics

Value

Rated power Peak power Cut-in speed Cut-off speed Temperature range

10 W (1 min average at 10 m/s) 12 W 3 m/s 40 m/s −60–10 ◦ C

Different wind-turbine types put out different power based on their power-curve characteristics. Through a comprehensive literature review, a model used to describe the performance is proposed as follows [21]:   E = Ea + Eb (vc ≤ v ≤ v F )   wind  3 (2) Ea = ∑ PR t vvR (vc ≤ v ≤ v R )    E = P t(v ≤ v ≤ v ) R F R∑ b where Ewind is the output of wind power (Wh/day), which consists of Ea and Eb ; vc is the cut-in wind speed (3 m/s); vR is the rated wind speed (10 m/s); vF is the cut-off wind speed (40 m/s); v is the wind speed; PR is the rated electrical power (10 W), which is average energy at the wind speed of 10 m/s for one minute; and t is the time (hours). 2.4.3. Low-Temperature Characteristics of PV Panel The PV panel is the major power-generation unit of the power system in this study. The environmental characteristics of the installation location of the power-supply system are described in Section 2.1. We do not understand the power-generation characteristics of photovoltaic panels at low temperatures. Therefore, the characteristics of PV panels at low temperatures require a comprehensive study. The key parameters of the PV panel have been described as follows:  I = Isc (1 − σ1 (exp(σ2 · V )) − 1)    
 V V−ocV   oc mp  σ = I − I /I sc mp sc  1     σ2 = (1/Voc )ln[(1 + σ1 )/σ1 ] isc ( R, T ) = Isc · RRr · [1 + a( T − Tr )]     voc ( R, T ) = Voc · ln[e + b( R − Rr )] · [1 − c( T − Tr )]     i  mp ( R, T ) = Imp · RRr · [1 + a( T − Tr )]   vmp ( R, T ) = Vmp · ln[e + b( R − Rr )] · [1 − c( T − Tr )]

(3)

where Isc is the short circuit current (A); Voc is the open circuit voltage (V); Imp is the optimum operating current (A); Vmp is the optimum operating voltage (V). Isc , Voc , Imp and Vmp can be found in Table 2. σ1 and σ2 are the are parameters for the convenience of describing the model; isc (R, T) is the short circuit current (A) at radiation intensity R (Wh/m2 /day) and ambient temperature T (◦ C); voc (R, T) is the open circuit voltage (V) at the same conditions as isc (R, T); imp (R, T) and vmp (R, T) are the optimum operating current (A) and voltage (V) at radiation intensity R (Wh/m2 /day) and ambient temperature T (◦ C), respectively; Rr is the reference value of 1000 W/m2 ; Tr is the reference temperature of 25 ◦ C; a is 0.0025; b is 0.0005; c is 0.00288. Based on the established PV panel parameter model as shown in Equation (3), the temperature was set 0 ◦ C. Simultaneously, radiation intensities of 1000, 800, 600, and 400 W/m2 were selected for the PV panel. Then, the output characteristics were tested (Figure 6). Figure 6a shows the open circuit voltage and short circuit current of the PV panel. At the same temperature, as radiation intensity decreased, the open circuit voltage and short circuit current of the PV panel would gradually decrease. Figure 6b shows the output power of the PV panel. The output power of the PV panel also increased as the radiation intensity increased. When radiation intensity was kept constant at

Appl. Sci. 2018, 8, 2376

9 of 23

1000 W/m2 and ambient temperature was raised from −50 ◦ C to −20 ◦ C, the output characteristics of the PV panel were as shown in Figure 6c,d. Figure 6c shows that the open circuit voltage of the PV Appl. Sci. 2018, 8, x as FOR PEER REVIEW decreased, but the short circuit current decreased. Figure 6d shows 9 of 23 panel increased the temperature that the maximum power point of the PV panel increased with decreasing temperature. Although the temperature T (°C), Rr isreduced, the reference value output of 1000power W/m2point ; Tr is the reference short circuit current of respectively; the PV panel was the maximum increased in the temperature of 25 °C; a is 0.0025; b is 0.0005; c is 0.00288. low-temperature environment.

Figure Figure6.6. (a,b) (a,b) Output Output characteristics characteristics of of PV PV panel panel under under different different radiation radiation intensities intensities at at the the same same ◦ C). (c,d) Output characteristics of PV panel under 1000 W/m2 at different temperatures. temperature temperature(0 (0 °C). (c,d) Output characteristics of PV panel under 1000 W/m2 at different

temperatures.

3. Experiment on Low-Temperature Battery Characteristics Based on the established PV panelbattery parameter shown in Equation (3), the As shown in Figure 5, a lead–acid wasmodel used asasan energy-storage device fortemperature the powerwas setsystem 0 °C. Simultaneously, of 1000, 800, 600, 400observing W/m2 were selected for supply of the observingradiation system. intensities The lead–acid batteries usedand in our system were the PV panel. Then, the output characteristics were tested (Figure 6). Figure 6a shows the open developed by Taiyuan University of Technology. We used Pb as the negative active material of lead– circuit voltagePbO and short circuit current of the PV as chemical radiation acid batteries, the positive active material, andpanel. dilutedAtH2the SO4same as thetemperature, electrolyte. The 2 as intensity decreased, the open circuit voltage and short circuit current of the PV panel would reaction principle is shown in Equation (4). gradually decrease. Figure 6b shows the output power of the PV panel. The output power of the PV Pb Oradiation ↔ 2H2 O (4) panel also increased Pas increased.2PWhen intensity was kept 2 + 2H2 SOintensity 4 b +the b SO4 +radiation Charge−−− Discharge constant at 1000 W/m2 and ambient temperature was raised from –50 °C to –20 °C, the output characteristics of the PV paneliswere as shown in Figure reaction, 6c,d. Figure 6c shows the open When a lead–acid battery subjected to a discharge the Pb on the that cathode plate, circuit PbO2 voltage of the PV panel increased as the temperature decreased, but the short circuit current on the anode plate, and the diluted H2 SO4 in the surrounding area chemically react, consuming diluted decreased. Figure 6d shows that the maximum power point of the PV panel increased with H 2 SO4 in the electrolyte, and simultaneously forming compounds PbSO4 and H2 O. After the discharge decreasing Although the short consumed, circuit current of the PV the panel was reduced, the behavior, H2temperature. SO4 in the electrolyte is gradually and the longer discharge time of the maximum output power point increased in the low-temperature environment. battery is, the lower the concentration of H2 SO4 in the electrolyte will be.

When the lead–acid battery is charged, H2 SO4 , Pb, and PbO2 are formed on the cathode and 3. Experiment on Low-Temperature anode plates. Then, the concentration Battery of H2 SOCharacteristics 4 in the electrolyte gradually recovers. As the charging reaction H2 SO slowlywas returns its an concentration at thedevice beginning of As continues, shown in the Figure 5,4 aconcentration lead–acid battery usedto as energy-storage for the the discharge. At this time, the active substances of the cathode and anode electrodes in the lead–acid power-supply system of the observing system. The lead–acid batteries used in our observing system battery can be considered to have recovered their activity. When of the cathode anode were developed by Taiyuan University of Technology. We usedthe Pb PbSO as the4negative active and material of plates is completely and PbO battery isand considered and the process 2 , thematerial, lead–acid batteries, reduced PbO2 as to thePbpositive active diluted full H2SO 4 as thecharging electrolyte. The ischemical complete. reaction principle is shown in Equation (4).

Pb + PbO2 + 2H 2 SO4 ⎯⎯⎯⎯⎯⎯ → 2Pb SO4 + 2H 2O Ch arg e −−− Disch arg e

(4)

When a lead–acid battery is subjected to a discharge reaction, the Pb on the cathode plate, PbO2 on the anode plate, and the diluted H2SO4 in the surrounding area chemically react, consuming diluted H2SO4 in the electrolyte, and simultaneously forming compounds PbSO4 and H2O. After the discharge behavior, H2SO4 in the electrolyte is gradually consumed, and the longer the discharge time of the battery is, the lower the concentration of H2SO4 in the electrolyte will be. When the lead–acid battery is charged, H2SO4, Pb, and PbO2 are formed on the cathode and anode plates. Then, the concentration of H2SO4 in the electrolyte gradually recovers. As the charging

the discharge. At this time, the active substances of the cathode and anode electrodes in the lead–acid battery can be considered to have recovered their activity. When the PbSO4 of the cathode and anode plates is completely reduced to Pb and PbO2, the battery is considered full and the charging Appl. Sci. 2018,process 8, 2376 is complete. 10 of 23 3.1. Low-Temperature Calibration Experiment of the Battery Capacity 3.1. Low-Temperature Calibration Experiment of the Battery Capacity The activity of the electrolyte of the lead–acid battery is affected by low temperatures, which Thethe activity of the electrolyte lead–acid battery is temperatures affected by low temperatures, which affect capacity of the battery. of A the battery-capacity at low experiment was designed affect the capacity of the battery. A battery-capacity at low temperatures experiment was designed and implemented. and implemented. In the experiment, to calibrate battery capacity, we placed the self-developed lead–acid battery In experiment, calibrate batterytest capacity, we placed self-developed lead–acid battery into the a GDJS series to low-temperature chamber, whichthecan provide stable environmental into a GDJS series low-temperature test chamber, which can provide stable environmental temperature. temperature. The temperature range for the experiment was from –50 °C to 0 °C. The ideal capacity ◦ C to 0 ◦ C. The ideal capacity of the battery The range the experiment wasa from −50temperature oftemperature the battery is 13.6for V and 75 Ah under normal environment. We discharged the is battery 13.6 V and 75 Ah under a normal temperature environment. We battery at 7.5 A at 7.5 A at different ambient temperatures (–50–0 °C) to discharged determine the battery capacity. The ◦ C) to determine battery capacity. The voltages of battery at voltages differentof ambient temperatures ( − 50–0 battery were measured by an oscilloscope (MSO70404C, Tektronix, Beaverton, OR, USA), were by ancut-off oscilloscope Tektronix, OR, USA), temperature and the discharge andmeasured the discharge voltage(MSO70404C, was 10.5 V. The intervalBeaverton, for the experimental change ◦ C. cut-off voltage was 10.5 V. The interval for the experimental temperature change was set to 10 was set to 10 °C. At each temperature in the experiment, the low-temperature test chamber Atmaintained each temperature in the experiment, low-temperature test chamber maintained temperature the temperature at whichthe battery voltage reached the discharge cut-offthe voltage. We took at the which battery voltage reached the discharge cut-off voltage. We took the average values of the average values of the voltage to minimize statistical error and uncertainty. voltageThe to minimize statistical error and uncertainty. experiment results are shown in Figure 7 and demonstrate that, when a lead–acid battery is The experiment arecurrent shown in in Figure 7 and demonstrate that, when lead–acid battery discharged with a results constant a low-temperature environment, the aoverall trend of the is discharge dischargedprocess with a is constant current in a low-temperature environment, the overall trend of similar. As temperature decreased, the discharge time of the batterythe also discharge process is similar. As temperature decreased, thethat discharge time of the battery decreased at a constant current. This finding indicates the discharge capacity ofalso the decreased battery was at weakening, a constant current. This finding the discharge capacity of the battery was weakening, simultaneously, theindicates amountthat of discharged electricity decreased. Lead–acid batteries simultaneously, the amount of discharged electricity decreased. Lead–acid batteries have a capacity have a capacity of 75.6 (100%), 65 (85.9%), 51.25 (67.8%), 37.5 (49.6%), 28.75 (38%), and 22.5ofAh ◦ C, 75.6 (100%), (85.9%), (67.8%), 37.5 (38%), and 22.5 Ah (29.7%) at 0 ◦ C,was −10higher (29.7%) at 065°C, –10 °C,51.25 –20 °C, –30 °C, –40(49.6%), °C, –50 28.75 °C, respectively. When the temperature ◦ ◦ ◦ ◦ ◦ −20 30battery C, −40capacity C, −50remained C, respectively. When the temperature was higher than 0 C, battery thanC,0 − °C, nearly constant. capacity remained nearly constant.

Figure 7. 7. Discharge curves of of thethe lead–acid battery at low temperatures. Figure Discharge curves lead–acid battery at low temperatures.

3.2. Low-Temperature Charging Experiment of Lead–Acid Battery 3.2. Low-Temperature Charging Experiment of Lead–Acid Battery Low ambient temperatures affect the charging capacity of a lead–acid battery. Similar to the Low ambient temperatures affect the charging capacity of a lead–acid battery. Similar to the experiment to calibrate battery capacity, in the charging experiment, we put the self-developed experiment to calibrate battery capacity, in the charging experiment, we put the self-developed lead–acid battery into a low–temperature test chamber, and a two-stage charging mode was adopted. lead–acid battery into a low–temperature test chamber, and a two-stage charging mode was In the first stage, we charged the lead–acid battery with a constant current, and at the second stage adopted. In the first stage, we charged the lead–acid battery with a constant current, and at the with constant voltage charging. The temperature range for the experiment was −50–0 ◦ C. The voltage second stage with constant voltage charging. The temperature range for the experiment was –50–0 for constant voltage charging was selected to be 14.6 V. The currents of constant current charging were °C. The voltage for constant voltage charging was selected to be 14.6 V. The currents of constant selected as 1, 2, and 3 A, respectively, performed at each temperature. The interval of the experimental current charging were selected as 1, 2, and 3 A, respectively, performed at each temperature. The temperature change was 10 ◦ C. Each process of charging was repeated, and the voltage–time curves interval of the experimental temperature change was 10 °C. Each process of charging was repeated, were obtained at the different temperatures. Three experimental sets of curves are shown in Figure 8. and the voltage–time curves were obtained at the different temperatures. Three experimental sets of curves are shown in Figure 8.

Appl. Sci. 2018, 8, x FOR PEER REVIEW Appl. Sci. 2018, 8, 2376

11 of 23 11 of 23

Figure 8. (a–c) Charging curve for the lead–acid battery at temperatures of −10, −20, and −30 ◦ C under the charging currents of 1, 2, and 3 A, respectively. Figure 8. (a–c) Charging curve for the lead–acid battery at temperatures of –10, –20, and –30 °C under the charging currents of 1, 2, and 3 A, respectively.

As shown in Figure 8, we found that, at the same temperature, the capacity of the battery differed due to different charging currents. The battery-capacity calculation results of the charging experiment As shown in Figure 8, we found that, at the same temperature, the capacity of the battery are shown in Table 4. differed due to different charging currents. The battery-capacity calculation results of the charging experiment are shown Table 4. Table in 4. Battery-capacity calculation results of the charging experiment. Temperature (◦ C) Charging Current (A) Capacity (Ah) Table 4. Battery-capacity calculation results of theBattery charging experiment.

−10 ◦ C

1A

40.0

Temperature (°C) ◦ Charging Current (A) Battery Capacity (Ah) −10 C 2A 45.4 –10 °C −10 ◦ C 13AA 40.0 47.1 30.0 –10 °C −20 ◦ C 21AA 45.4 ◦C − 20 2 A 27.4 –10 °C 3A 47.1 −20 ◦ C 3A 25.0 –20 °C 1 A 30.0 −30 ◦ C 1A 16.7 –20 °C −30 ◦ C 22AA 27.4 16.5 ◦ –20 °C −30 C 33AA 25.0 16.2 –30 °C 1A 16.7 –30 °C 2A 16.5 The results in Table 4 demonstrate that the battery capacity at −10 ◦ C was 40.0 Ah, 45.4 Ah, and –30 °C 3A 16.2 47.1 Ah with charging currents of 1 A, 2 A, and 3 A, respectively. The battery capacity increased with the charging currents. However, when the experimental temperature was −20 ◦ C, the relationship The results in Table 4 demonstrate that the battery capacity at –10 °C was 40.0 Ah, 45.4 Ah, and between the battery capacity and charging current was different from that at −10 ◦ C. The battery 47.1 Ah with charging currents of 1 A, 2 A, and 3 A, respectively. The battery capacity increased with capacity at −20 ◦ C was 30.0 Ah, 27.4Ah, and 25 Ah with charging currents of 1 A, 2 A, and 3 A, the charging currents. However, when the experimental temperature was –20 °C, the relationship respectively. Hence, the battery capacity decreased with the increase of the charging currents from between the battery capacity and charging current was different from that at –10 °C. The battery 1 A to 3 A. This phenomenon occurred mainly because the drop in ambient temperature to a certain capacity at –20 °C was 30.0 Ah, 27.4Ah, and 25 Ah with charging currents of 1 A, 2 A, and 3 A, level decreased the activity of the active material on the two-electrode plates of the lead-acid battery respectively. Hence, the battery capacity decreased with the increase of the charging currents from 1 decreased, resulting in a slower chemical reaction rate. The battery capacity at −30 ◦ C was 16.7 Ah, A to 3 A. This phenomenon occurred mainly because the drop in ambient temperature to a certain 16.5 Ah, and 16.2 Ah with charging currents of 1 A, 2 A, and 3 A, respectively. The battery capacity level decreased the activity of the active material on the two-electrode plates of the lead-acid battery under different charging currents changed little, because the activity of the active material on the decreased, resulting in a slower chemical reaction rate. The battery capacity at –30 °C was 16.7 Ah, two-electrode plates was further reduced. 16.5 Ah, and 16.2 Ah with charging currents of 1 A, 2 A, and 3 A, respectively. The battery capacity under different charging currents changed little, because the activity of the active material on the 3.3. Charging Strategy two-electrode plates was further reduced. Based on the experimental results in Sections 3.1 and 3.2, the magnitude of the charging current would affect the actual capacity of the battery at different temperatures. The relationship between 3.3. Charging Strategy battery capacity and charging currents reversed between −10 ◦ C and −20 ◦ C. The critical ambient

Appl. Sci. 2018, 8, 2376

12 of 23

temperature of the relationship between battery capacity and charging currents was therefore tested experimentally. The temperature range for the experiment was from −10 ◦ C to 20 ◦ C. The interval Appl. Sci. 2018, 8, x FOR PEER REVIEW 12 of 23 of the experimental temperature was set as 1 ◦ C. The experimental method and other experimental Based on thethe experimental results in Sections conditions were same as those in Section 3.2. 3.1 and 3.2, the magnitude of the charging current wouldThe affect the actual results capacityshown of thein battery temperatures. between experimental Figureat9different demonstrate that whenThe the relationship ambient temperature battery and charging currents °Cthe and –20 °C.ofThe critical ambient was lesscapacity than a critical value, the batteryreversed capacity between increased–10 with increase the charging current temperature of the relationship betweenwas battery capacity and chargingofcurrents and decreased when the temperature greater. The temperature −16 ◦ Cwas wastherefore observedtested as the experimentally. The temperature range for the experiment was from –10 °C to 20 °C. The interval of critical level. Thus, a low-temperature charging strategy for lead-acid batteries is proposed. When the ◦ the experimental temperature was set ashas 1 °C. The experimental method and other experimental temperature is above −16 C, the battery a strong activity of active substances. At this time, a large conditions the same those in Section 3.2.temperature is lower than −16 ◦ C, a small current can current canwere be selected forascharging. When the be selected for charging.

Figure9.9.Discharge Dischargecurves curvesofofthe thelead-acid lead-acidbattery batteryatatlow lowtemperature. temperature. Figure

Is is experimental the solar charging current; the wind turbine charging IB is the charging current The results shownIwinis Figure 9 demonstrate that current; when the ambient temperature of battery, and I is equal to the sum of I and I ; T is the ambient temperature; T is the critical s w a c B was less than a critical value, the battery capacity increased with the increase of the charging current ◦ temperature above (−16 was C). The charging strategy of the hybrid wind-solar system and decreasedmentioned when the temperature greater. The temperature of –16 °C was observed as theis shown in Tables 5 and 6. critical level. Thus, a low-temperature charging strategy for lead-acid batteries is proposed. When the temperature isTable above5. –16 °C, the battery has a strong activity of active substances. At this time, a The charging strategy of the hybrid wind-solar system (Ta > Tc ). large current can be selected for charging. When the temperature is lower than –16 °C, a small PV Output Status 1 Wind Turbine Output Status 1 current can Environmental be selected forConditions charging. Is isPolar the day solar charging current; Iw is the wind turbine the charging current Wind energy available MPPTcharging current; IB isMPPT Polarand dayIB is Wind Not working of battery, equalenergy to theunavailable sum of Is and Iw; TMPPT a is the ambient temperature; Tc is the critical Polar night Wind energy available Not working MPPT temperature mentioned above (–16 °C). The charging strategy of the hybrid wind-solar system is Polar night Wind energy unavailable Not working Not working shown in Tables 5 and 6. 1 The output statuses of PV and wind turbine include Maximum Power Point Tracking (MPPT) mode and not working mode.

Table 5. The charging strategy of the hybrid wind-solar system (Ta > Tc).

1 Table 6. The charging strategyPV of the hybrid wind-solar system (Ta Imax Current-limiting available Not workingMPPT MPPT Polar day Wind energy available Is + Iw < Imax MPPT MPPT Polar night Wind energy unavailable Not working Not working

Polar day Wind energy available Is > Imax Current-limiting Unloading 1 The output statuses of PV and wind turbine include Maximum Power Point Tracking (MPPT) Polar day Wind energy unavailable Is > Imax Current-limiting Not working mode and not working Polar day Wind energymode. unavailable Is ≤ Imax MPPT Not working Polar night Wind energy available Iw > Imax Not working Current-limiting Polar Wind energy available Iw ≤ Imax for hybrid Not working MPPT If Tnight a>Tc, there are three operating conditions wind-solar system. Polar night Wind energy unavailable Is + Iw = 0 Not working Not working

Operating condition 1 (the1 first and second rows of Table 5): In the polar day, the PV always Imax is the lower limit of small charging current. works in MPPT mode. And if the value of wind speed is between Vc and VF, which means wind energy is available, the wind turbine also works in MPPT mode. Otherwise, wind energy cannot be utilized and the wind turbine does not work. Operating condition 2 (the third and fourth rows of Table 5): In the polar night, the PV cannot work. At this time, if the wind energy is available, the wind turbine will work in MPPT mode. If the wind speed is below 3 m/s or above 40 m/s, the wind turbine does not work.

Appl. Sci. 2018, 8, 2376

13 of 23

If Ta > Tc , there are three operating conditions for hybrid wind-solar system. Operating condition 1 (the first and second rows of Table 5): In the polar day, the PV always works in MPPT mode. And if the value of wind speed is between Vc and VF , which means wind energy is available, the wind turbine also works in MPPT mode. Otherwise, wind energy cannot be utilized and the wind turbine does not work. Operating condition 2 (the third and fourth rows of Table 5): In the polar night, the PV cannot work. At this time, if the wind energy is available, the wind turbine will work in MPPT mode. If the wind speed is below 3 m/s or above 40 m/s, the wind turbine does not work. Operating condition 3: During normal daytime, the PV works in MPPT mode. And if the value of wind speed is between Vc and VF , the wind turbine also works in MPPT mode. Otherwise, the wind turbine does not work. In the night, the PV cannot work. At this time, if the wind energy is available, the wind turbine will work in MPPT mode. If the wind speed is below 3 m/s or above 40 m/s, the wind turbine does not work. If Ta < Tc , a small current should be selected for battery charging. There are four operating conditions for hybrid wind-solar system. Operating condition 4 (the first, second and third rows of Table 6): When it is in the polar day and the wind speed is between Vc and VF , if Is + Iw > Imax , the PV will work in MPPT mode and the wind turbine will enter current-limiting mode; if Is + Iw < Imax , the PV and wind turbine will all work in MPPT mode. And if Is > Imax , the PV will work in the current-limiting mode and the wind turbine will work in uploading mode to protect the system. Operating condition 5 (the fourth and fifth rows of Table 6): When it is in the polar day and the wind speed is unavailable, the wind turbine cannot work. If Is > Imax , the PV will enter current-limiting mode; if Is ≤ Imax , the PV will enter MPPT mode. Operating condition 6 (the sixth and seventh rows of Table 6): When it is in the polar night and the wind speed is available, the PV cannot work. If Iw > Imax , the wind turbine will work in current-limiting mode; if Iw ≤ Imax , the wind turbine will work in MPPT mode. Operating condition 7 (the eighth row of Table 6): When it is in the polar night and the wind speed is unavailable, the PV and the wind turbine all cannot work and Is + Iw = 0. 4. Circuit Design 4.1. Charging Circuit In the process of circuit design, we considered the impact of extremely cold environments of the Polar Region. Thus, low temperature characteristics and power consumption of the circuit devices must be examined in detail. The charging circuit consists of four parts: (1) the solar and SWT charging circuit, (2) the driver circuit, (3) the detection circuit and (4) ADC circuit and auxiliary circuit. The solar and SWT charging circuit is shown in Figure 10. The solar charging circuit (Figure 10a) is a Buck circuit where C1 and C2 are electrolytic capacitors with values of 1000 µF; L1 is a toroidal inductor of 470 µH; the resistance values of R1 , R4 and R5 are both 2 KΩ; the resistance value of R2 is 1 KΩ; the resistance value of R3 is 10 KΩ; and D1 and D2 are dual high-voltage Schottky rectifiers (MBR20H100CT, Vishay Semiconductors, TX, USA). The MOSFETs IRF540NPbF (International Rectifier, El Segundo, CA, USA) are selected for charging by controlling the on-off state. The SWT charging circuit is shown in Figure 10b. The AC generated by the SWT is rectified to DC using 3SRB5016 (ASEMI, Ningbo, China), with an operating temperature range between −55 ◦ C and 150 ◦ C. An unloading circuit is designed in the SWT charging circuit to protect the SWT and charging circuit from damage when the wind speed in the polar region is too large and exceeds the limits of operation of SWT. C3 and C4 are electrolytic capacitors with values of 1000 µF, and L2 is a toroidal inductor of 470 µH. Under the condition that wind speed reaches the cut-in value, the MCU (MSP430F5438A, Texas Instruments, Dallas, TX, USA) can collect the voltage and current of the small

Appl. Sci. 2018, 8, 2376

14 of 23

WT via MPPT control strategy. MCU implements the MPPT strategy by changing the duty cycle of Appl.PWM. Sci. 2018, 8, x FOR PEER REVIEW 14 of 23 the

Figure 10. Schematic of the solar circuit (a) and SWT charging circuit (b).

4.2. Driver Circuit The solar charging circuit (Figure 10a) is a Buck circuit where C1 and C2 are electrolytic capacitors withcircuit valuesofofthe 1000 μF;and L1 isSWT a toroidal inductor the resistance values circuit of R 1, Ris4 The main solar charging circuitofis470 theμH; Buck circuit. The driver and2018, R5 in are KΩ; the resistance value of R 2 is 1 KΩ; the resistance value of R 3 is 10 KΩ; D1 of 23 Appl. shown Sci. 8,Figure xboth FOR 2PEER REVIEW 15 11 and was developed by the IR2104S (International Rectifier, El Segundo, CA,and USA). and D2 are dual high-voltage Schottky rectifiers (MBR20H100CT, Vishay Semiconductors, TX, USA). The MOSFETs IRF540NPbF (International Rectifier, El Segundo, CA, USA) are selected for charging by controlling the on-off state. The SWT charging circuit is shown in Figure 10b. The AC generated by the SWT is rectified to DC using 3SRB5016 (ASEMI, Ningbo, China), with an operating temperature range between –55 °C and 150 °C. An unloading circuit is designed in the SWT charging circuit to protect the SWT and charging circuit from damage when the wind speed in the polar region is too large and exceeds the limits of operation of SWT. C3 and C4 are electrolytic capacitors with values of 1000 μF, and L 2 is a toroidal inductor of 470 μH. Under the condition that wind speed reaches the cut-in value, the MCU (MSP430F5438A, Texas Instruments, Dallas, TX, USA) can collect the voltage and current of the small WT via MPPT control strategy. MCU implements the MPPT strategy by changing the duty cycle of the PWM. Figure11. 11.Schematic Schematic of of the Figure thedriver drivercircuit. circuit. 4.2. Driver C is aCircuit capacitor with value of 100 nF. D9 is a low power-consumption Schottky diode SK34 (Diodes Incorporated, Plano, TX, USA), and C6 is a 1 µF tantalum capacitor that can be considered as a The main circuit of the solar and SWT charging circuit is the Buck circuit. The driver circuit is bootstrap capacitor. Current limiting resistors R and R24necessary are designed to havemultiple a resistance 2 KΩ. In the solar and11 SWT charging it is to collect setsofof voltage shown in Figure andhybrid was developed bycircuit, the23IR2104S (International Rectifier, El Segundo, CA, The IN is connected to the PWM port of the MCU, and the LO has the same polarity as the pin IN, the and USA). current data, such as the voltage of the wind turbine, the voltage of the solar cell, and while the HO has the opposite polarity. The function of the drive circuit can be described as follows: voltage of current charging current of thediode SWT and the C5 the is a battery. capacitor The withcollected value of 100 nF. Dincludes 9 is a low the power-consumption Schottky SK34 After inputting the PWM at pin IN, (1) when the HO is low, the LO is high. At this time, the IRF540 charging current of the battery. TheUSA), MAX471ESA San Jose, CA, USA) is be selected as the (Diodes Incorporated, Plano, TX, and C6 is a(Maxim, 1 μF tantalum capacitor that can considered as core connected to R24 is turned on and grounded, and the VCC is charged to C6 through D9 , (2) When the a bootstrap capacitor.the Current limiting resistors and R24 areisdesigned to havebidirectional, a resistance of 2high-side KΩ. of the circuit during design process. TheR23MAX471 a complete, HO is high, the LO is low. At this time, the VCC and the voltage of C6 are output from the HO, thereby The IN is connected to the PWM port of the MCU, and the LO has the same polarity as the pin current-sense amplifier for portable battery monitoring, and operates at a temperatureIN, range controlling the turn-on and turn-off of the IRF540 connected to R23 . while–40 the°C HOand has85 the°C. opposite polarity. The function of the drive circuit can be described as follows: between After inputting the PWM at pin IN, (1) when the HO is low, the LO is high. At this time, the IRF540 connected to and R24 isAuxiliary turned onCircuit and grounded, and the VCC is charged to C6 through D9, (2) When the 4.4. ADC Circuit HO is high, the LO is low. At this time, the VCC and the voltage of C 6 are output from the HO, The ADS1100 (Texas Instruments, Dallas, USA) was selected thereby controlling the turn-on and turn-off ofTX, the IRF540 connected to Ras 23. the core of the ADC circuit. 5 4.3. Detection Circuit

This circuit was assembled compatible with I2C serial interfaces, and has a simple circuit structure. The auxiliary circuit (Figure 12) contains two levels of voltage conversion: The primary and secondary voltage regulators. The primary voltage regulator (LT1129, Analog Devices Inc.,

Figure 11. Schematic of the driver circuit. Appl. Sci. 2018, 8, 2376

15 of 23

4.3. Detection Circuit

In the solar and SWT hybrid charging circuit, it is necessary to collect multiple sets of voltage 4.3. Detection Circuit and current data, such as the voltage of the wind turbine, the voltage of the solar cell, and the solar and SWT charging circuit, it is necessary to collectcurrent multipleofsets voltage andthe voltageInofthethe battery. Thehybrid collected current includes the charging theof SWT and current current data, such as the voltage of MAX471ESA the wind turbine, the voltage of the solar cell,isand the voltage charging of the battery. The (Maxim, San Jose, CA, USA) selected as the of core collected the The charging currentisofathe SWT and bidirectional, the charging current of of the thebattery. circuitThe during thecurrent designincludes process. MAX471 complete, high-side the battery. The MAX471ESA (Maxim, battery San Jose,monitoring, CA, USA) is selected as the core the circuit during current-sense amplifier for portable and operates at aof temperature range the design process. between –40 °C and 85The °C.MAX471 is a complete, bidirectional, high-side current-sense amplifier for portable battery monitoring, and operates at a temperature range between −40 ◦ C and 85 ◦ C. 4.4. ADC Circuit and Auxiliary Circuit 4.4. ADC Circuit and Auxiliary Circuit The ADS1100 (Texas Instruments, Dallas, TX, USA) was selected as the core of the ADC circuit. The ADS1100 (Texas Instruments, Dallas, TX, USA) was selected as the core of the ADC circuit. This circuit was assembled compatible with I22C serial interfaces, and has a simple circuit structure. This circuit was assembled compatible with I C serial interfaces, and has a simple circuit structure. The auxiliary circuit (Figure 12) contains two levels of voltage conversion: The primary and The auxiliary circuit (Figure 12) contains two levels of voltage conversion: The primary and secondary secondary voltage regulators. The primary voltage regulator (LT1129, DevicesMA, Inc., voltage regulators. The primary voltage regulator (LT1129, Analog Devices Analog Inc., Norwood, Norwood, MA, USA) 3.3of V at 12 V by of a plurality of battery The secondary USA) generates 3.3 generates V at supply 12 supply V by aof plurality battery units. The units. secondary voltage voltage regulator (LM2596, Texas Instruments, Dallas, USA) generates 5 V at of 12the V to regulator (LM2596, Texas Instruments, Dallas, TX, USA)TX, generates 5 V at a supply ofa12supply V to fulfill fulfill the requirements of the iridium circuit. requirements of the iridium circuit.

Figure12. 12. Schematic Schematic of the auxiliary Figure auxiliarycircuit. circuit.

5. Performance Evaluation of the Charging Circuit 5. Performance Evaluation of the Charging Circuit In the Arctic Ocean or Antarctica, low temperatures are the primary challenge for the power In the Arctic Ocean or Antarctica, low temperatures are the primary challenge for the power supply of unmanned equipment. Our charging circuit and charging strategy may have been affected supply of unmanned equipment. Our charging circuit and charging strategy may have been affected by instabilities of the components and modules. Although for the charging circuit components and bymodules instabilities of the components and modules. Although for the charging circuit components and were selected based on their good performance at low temperatures (−50–30 ◦ C), the potential modules were dependence selected based oncharging their good performance at low temperatures (–50–30 °C), the temperature of the circuit had to be assessed. The temperature dependence potential temperature dependence of the charging circuit had to be assessed. The temperature of the detection and ADC circuits, which are important for the charging strategy, were evaluated dependence of the detection and ADC circuits, which are important for the charging strategy, were and assessed. evaluated and assessed. 5.1. Temperature Dependence of the Detection Circuit We placed the detection circuit into the GDJS series high–low-temperature test chamber providing a stable environmental temperature over the range of −50 ◦ C to 30 ◦ C. The input voltage of the experiment was generated with programmable power-supply E3631A (Agilent, Santa Clara, CA, USA) to generate voltages from 2 V to 22 V. The output of the detection circuit was measured by an oscilloscope (MSO70404C, Tektronix, Beaverton, OR, USA). The interval of the experimental temperature was 20 ◦ C. The interval of the measured voltage was set to 1 V. At each temperature, the high–low-temperature test chamber maintained the temperature for 20 min, during which

Appl. Sci. 2018, 8, x FOR PEER REVIEW

16 of 23

5.1. Temperature Dependence of the Detection Circuit Sci. 2018, 8, 2376 of 23 WeAppl. placed the detection circuit into the GDJS series high–low-temperature test16 chamber providing a stable environmental temperature over the range of –50 °C to 30 °C. The input voltage of the experiment was generated with300 programmable E3631A (Agilent, Santa Clara, measurements were repeated times. We tookpower-supply the average values to minimize statistical error CA, USA) toand generate voltages from 2 V to 22 V. The output of the detection circuit was measured by an uncertainty. output voltage Tektronix, and average error results areOR, shown in Figure The resultsofdemonstrate that oscilloscopeThe (MSO70404C, Beaverton, USA). The13.interval the experimental the output of the detection circuit had a strong upward trend, proportional to the increase of input, temperature was 20 °C. The interval of the measured voltage was set to 1 V. At each temperature, the and the temperature dependence of the detection circuit was within a narrow range. The R2 of the high–low-temperature test chamber maintained the temperature for 20 min, during which output of the detection circuit was greater than 0.99. The maximum average error of the output voltage measurements were repeated 300 times. We took the average values to minimize statistical error and of the detection circuit was 0.017 V at 10 V, and the minimum average error of the output voltage uncertainty. was −0.027 V at 6 V. Therefore, the experimental results show that the detection circuit has excellent

stability at low temperatures, and can be used in the charging circuit for the polar regions.

Figure 13. Temperature dependence of the output of detection circuit and the average error. Figure 13. Temperature dependence of the output of detection circuit and the average error.

5.2. Temperature Dependence of the ADC Circuit

The output voltage and average error results are shown in Figure 13. The results demonstrate In the temperature-dependence experiment of the ADC circuit, we also placed the ADC circuit that the into output of the detection circuit had a strong trend, proportional to the increase of the high–low-temperature test chamber with theupward same temperature range. The input voltage of 2 input, and dependence of the detection was within narrow range. The the the ADCtemperature circuit was generated by a nominal 5 V voltagecircuit reference chip and aapotentiometer with an R of the output ofrange the detection was greater 0.99. in The maximum averagethat error of the output input from 1 to 3 circuit V in intervals of 0.5 V. than The results Figure 14a demonstrate the minimum ◦ C, maximum value was 3.308 V at −30 ◦ C, average value of the reference voltage was 3.302 V at − 50 voltage of the detection circuit was 0.017 V at 10 V, and the minimum average error of the output of the output V, and standard deviation was 0.0015show V. We that calculated the coefficient of has voltage value was –0.027 V at 6was V. 3.306 Therefore, the experimental results the detection circuit Appl. Sci. 2018, 8, x FOR PEER REVIEW variation of the reference voltage as 0.048%. The results in Figure 14b show that the five measured 17 of 23 excellent stability at low temperatures, and can be used in the charging◦circuit for the polar regions. ADC outputs can be considered stable in a temperature range of −50–30 C.

5.2. Temperature Dependence of the ADC Circuit In the temperature-dependence experiment of the ADC circuit, we also placed the ADC circuit into the high–low-temperature test chamber with the same temperature range. The input voltage of the ADC circuit was generated by a nominal 5 V voltage reference chip and a potentiometer with an input range from 1 to 3 V in intervals of 0.5 V. The results in Figure 14a demonstrate that the minimum value of the reference voltage was 3.302 V at –50 °C, maximum value was 3.308 V at –30 °C, average value of the output was 3.306 V, and standard deviation was 0.0015 V. We calculated the coefficient of variation of the reference voltage as 0.048%. The results in Figure 14b show that the five measured ADC outputs can be considered stable in a temperature range of –50–30 °C. Figure 14. Temperature dependence of the (a) reference voltage, and (b) output of the ADC circuit (the Figure 14. Temperature dependence of the (a) reference voltage, and (b) output of the ADC circuit measured and ideal voltage values of 1, 1.5, 2, 2.5, and 3 V). (the measured and ideal voltage values of 1, 1.5, 2, 2.5, and 3 V).

5.3. Temperature-Correction Algorithm for the Power System As mentioned above, the evaluation reveals that the low-temperature performance of the detection circuit maintained a strong stability at a temperature range of –50–30 °C. However, we still developed an approach to correct and minimize deviation, because the accuracy of the detection circuit directly affects the operation effect of the charging strategy. The relationship

Figure 14. Temperature dependence of the (a) reference voltage, and (b) output of the ADC circuit Appl.(the Sci.measured 2018, 8, 2376 and ideal voltage values of 1, 1.5, 2, 2.5, and 3 V).

17 of 23

5.3. Temperature-Correction Algorithm for the Power System 5.3. Temperature-Correction Algorithm for the Power System As mentioned above, the evaluation reveals that the low-temperature performance of the As mentioned above, the evaluation reveals that the low-temperature performance of the detection detection circuit maintained a strong stability at a temperature range◦ of –50–30 °C. However, we circuit maintained a strong stability at a temperature range of −50–30 C. However, we still developed still developed an approach to correct and minimize deviation, because the accuracy of the an approach to correct and minimize deviation, because the accuracy of the detection circuit directly detection circuit directly affects the operation effect of the charging strategy. The relationship affects the operation effect of the charging strategy. The relationship between output current of the between output current of the detection circuit ID(T) and corrected current Iout(T) could be detection circuit ID (T) and corrected current Iout (T) could be approximated as follows: approximated as follows:

) ID (3T )3 + b( T ) ID ( T )22 + c( T ) ID ( T ) + d( T ) I out (TIout T )a(IT ) =(Ta)(= D ( T ) + b ( T ) I D ( T ) + c ( T ) I D ( T ) + d (T )

(5) (5)

wherea(T), a(T),b(T), b(T),c(T), c(T),and andd(T) d(T)are arethe thecoefficients coefficientsofofEquation Equation(5). (5). where Weput putthe the whole whole detection circuit high–low-temperature testtest chamber over We circuit into intothe theGDJS GDJSseries series high–low-temperature chamber ◦ C and measured the accuracy of the output current. During the experiment, a range of − 50–30 over a range of –50–30 °C and measured the accuracy output current. During the experiment, ◦ C from −50 ◦ C to 30 ◦ C. We measured the theinterval intervalofofthe themeasured measuredtemperature temperaturewas wasset setasas5 5°C the from –50 °C to 30 °C. We measured the outputofofthe the detection circuit Oscilloscope results werewere compared with with the output output circuitby bythe theoscilloscope. oscilloscope. Oscilloscope results compared the current of the detection circuit. At each measured temperature, the high–low-temperature test chamber output current of the detection circuit. At each measured temperature, the high–low-temperature maintained the temperature for 20 min in measurement was repeated 300 times. We the test chamber maintained the temperature forwhich 20 min in which measurement was repeated 300took times. average values to minimize and uncertainty. The coefficients b(T), c(T), and d(T) We took the average values tostatistical minimizeerror statistical error and uncertainty. The a(T), coefficients a(T), b(T), could simultaneously calculated. calculated. c(T), andbed(T) could be simultaneously The results measured currents detection circuit and ideal currents shown Figure The results ofof measured currents ofof detection circuit and ideal currents areare shown in in Figure 15.15. Weobserve observe that the actual currents andand measured values remained in the minimal We thediscrepancy discrepancybetween between actual currents measured values remained in the range. The maximum and minimum errors were 0.0061 0.0015 respectively. minimal range. The maximum and minimum errors wereand 0.0061 andA, 0.0015 A, respectively.

Figure 15. Temperature dependence of the current of the detection circuit. Figure 15. Temperature dependence of the current of the detection circuit.

The coefficients a(T), b(T), c(T), and d(T) decreased approximately in a parabolic manner with The coefficients a(T),(Figure b(T), c(T), decreased approximately in a − parabolic manner 7 at 30 ◦with increase of temperature 16). and The d(T) maximum and minimum a(T) were 3.4 × 10− C and –7 at 30 °C and−5 increase of temperature (Figure 16). The maximum and minimum a(T) were –3.4 × 10 − 7 ◦ −5.83 × 10 at −50 C, respectively. The maximum and minimum values of b(T) were 1.28 × 10 at −50 ◦ C and 7.47 × 10−7 at 30 ◦ C, respectively. The maximum and minimum values of c(T) were 0.0298 at −50 ◦ C and 0.0017 at 30 ◦ C, respectively. The maximum and minimum values of d(T) were 0.211 at −50 ◦ C and 0.012 at 30 ◦ C, respectively. The value of coefficients a(T), b(T), c(T), and d(T) at the nonmeasured temperatures could be predicted by interpolation.

–5.83 × 10–7 at –50 °C, respectively. The maximum and minimum values of b(T) were 1.28 × 10–5 at –50 °C and 7.47 × 10–7 at 30 °C, respectively. The maximum and minimum values of c(T) were 0.0298 at –50 °C and 0.0017 at 30 °C, respectively. The maximum and minimum values of d(T) were 0.211 at –50 °C and 0.012 at 30 °C, respectively. The value of coefficients a(T), b(T), c(T), and d(T) at the Appl. Sci. 2018, 8, 2376 18 of 23 nonmeasured temperatures could be predicted by interpolation.

Figure 16. (a–d) Temperature dependence of coefficients a, b, c, and d, respectively. Figure 16. (a–d) Temperature dependence of coefficients a, b, c, and d, respectively.

6. Results 6. Results 6.1. Monthly Average Energy Production from the PV Panel and the WT 6.1. Monthly Average Energy Production from the PV Panel and the WT The monthly average PV and WT power generation near Zhongshan Station and monthly average PV and WT power generation nearradiation-intensity, Zhongshan Stationwind-speed, and monthly averageThe loadmonthly power are presented in Figure 17. Monthly average average load power are presented in Figure 17. Monthly average radiation-intensity, and air-temperature data were obtained from NASA’s atmospheric science data center. wind-speed, It can be andthat air-temperature data were obtained from atmospheric science66.07 data KWh center. can be noted wind-power production dominated the NASA’s power supply and generated in It a year. thataverage wind-power production dominated power and generated 66.07 KWh in a Thenoted monthly maximum (minimum) value of the wind powersupply was 7.11 KWh (3.13 KWh) in August year. The monthly (minimum) valuebecause of wind was 7.11inKWh KWh) (January). Solar power average showed maximum strong seasonal fluctuations of power the polar night June (3.13 and July. August average (January). Solar power because of the at polar Theinmonthly solar power of showed the polarstrong day inseasonal Januaryfluctuations and December was larger 2.90night and in and July. The monthly average solar power of the polar solar day in January 3.21June KWh, respectively. The minimum value of the monthly average power in a and year December appeared inwas larger at 2.90 andwas 3.21during KWh, the respectively. minimum value of thewind monthly average solar power June and July, which polar nightThe in Antarctica. In contrast, energy was considered in significant a year appeared in June and July, during of thethe polar night ininAntarctica. contrast, to be from May to August. Thewhich powerwas generation PV and WT ZhongshanInStation wind energy was considered to be significant from May to August. The power generation of the PV showed good monthly characteristics. Monthly load demands of the observing system were basically WTThe in Zhongshan showed load gooddemands monthly characteristics. demands theand same. maximum Station and minimum were 3.11 KWhMonthly in June load and 3.56 KWhof inthe observing system basically theand same. The maximum and minimum load demands wereof3.11 January, March, May,were August, October, December, respectively. The minimum power output in June and 3.56 KWh in January, May, August, and December, respectively. theKWh power-supply system was 5.35 KWh inMarch, February, which canOctober, satisfy the power demands of the The minimum power output of the power-supply system was 5.35 KWh in February, which can observing system in normal operation. satisfy the power demands of the observing system in normal operation.

Appl. Sci. 2018, 8, x FOR PEER REVIEW

Appl. Sci. 2018, 8, x FOR PEER REVIEW Appl. Sci. 2018, 8, 2376

19 of 23

19 of 23 19 of 23

Figure 17. Monthly average PV and WT energy power production and load demand of the observing system in Zhongshan Station, Antarctica. Figure 17. Monthly average PV and WT energy power production and load demand of the observing

6.2.Figure Case Study in Antarctica Monthly average and WT energy power production and load demand of the system17. in Zhongshan Station, PV Antarctica.

observing system in Zhongshan Station, Antarctica. The new observing system was deployed in the station area of Zhongshan Station on 6 6.2. Case Study in Antarctica February 2017 during the 2016 Chinese National Antarctica Research Expedition. The observing 6.2. Case Study in Antarctica Theobserved new observing system was deployed station area data of Zhongshan Station on 6 February system the meteorological data and in icethe mass balance from 6 February 2017 to 9 July The new system was deployed inwere the also station area The of during Zhongshan Station on 6 2017 during theobserving 2016 Chinese National Antarctica Research Expedition. observing observed 2017. The hybrid solar and SWT charging circuit assembled the system operation of the February 2017 during theand 2016 National Research The The observing the meteorological data ice Chinese mass balance dataAntarctica from 6measured February 2017 to 9 July 2017. hybrid observing system. The air-temperature distribution byExpedition. the observing system in system observed the meteorological data and ice mass balance data from 6 February 2017 to 9 July solar and SWT charging circuit were also assembled during the operation of the observing system. Zhongshan Station from 6 February 2017 to 9 July 2017 is presented in Figure 18. The sampling 2017. The of hybrid solar and SWT charging were also assembled the the operation theair The air-temperature distribution measured by the observing insee Zhongshan Stationoffrom interval the air temperature was one circuit hour. In Figure 17, system we canduring that minimum observing The distribution measured the of observing system in 6temperature Februarysystem. 2017 to 9–41.375 July air-temperature 2017 presented in Figure 18. was The sampling interval of thepolar air temperature was °C ison 6 July 2017, which in the by period the night. The ◦ Zhongshan Station from 17, 6 February 2017 tothe 918July 2017 is in Figure 18. Theof was one hour. Figure we can see thaton minimum airpresented temperature was −41.375 Csampling on July maximum air In temperature was –4.625 °C February 2017, which was in the period the6polar ◦ C on interval of average the airintemperature one hour. Figure we can was see that the°C. air 2017, was the period ofwas the polar night. In The maximum air2017 temperature was −minimum 4.625 18 day. which The temperature from 6 February 2017 to 917,July –23.94 Temperature temperature was –41.375 onperiod 6 July 2017, which was in theand period of the polar night. ofThe February 2017, which was in the polar day. The average temperature from 6 February 2017 distribution was below 0°Cthe °C, in lineofwith our expectations the temperature range the maximum air temperature –4.625 °C ondistribution 18 Februarywas 2017, which was thewith period the polar to 9 July 2017 was −23.94 ◦was C. Temperature below 0 ◦ C, inin line our of expectations observing and power-supply systems. day. average temperature from 6 February 2017 to 9 Julysystems. 2017 was –23.94 °C. Temperature and The the temperature range of the observing and power-supply distribution was below 0 °C, in line with our expectations and the temperature range of the observing and power-supply systems.

Figure 18. Daily air temperature measured by the observing system from 6 February 2017 to 9 July Figure 18. Daily airStation, temperature measured by the observing system February 2017 in Zhongshan Antarctica. The area between the red lines from is the 6polar night.2017 to 9 July 2017 in Zhongshan Station, Antarctica. The area between the red lines is the polar night.

Figure 19 shows the daily radiation intensity and solar power from 6 February 2017 to 9 July 2017, Figure shows the dailyintensity radiation intensity solar power from 6 February 2017 to 9Then, July Figure 18.19 Daily airradiation temperature measured by the observing system 6 February to 919a). July respectively. Overall, continued toand decline from 6from February 2017 2017 (Figure 2 2017, respectively. Overall, radiation intensity continued to decline from 6 night. February 2017 (Figure 2017 in Zhongshan Station, Antarctica. Thewas area0 between red lines is the polar from 27 May to 9 July, radiation intensity W/m , the which was during the polar night. Although 2, which was during the polar night. 19a). Then, from 27 May to 9 July, radiation intensity was 0 W/m the observation system was initially deployed during the polar day, the intraday variation of radiation Figurewas 19 shows the daily radiation solar power from 6 February 2017 9 Julyof 2 in Although themore observation system wasintensity initially deployed during the polar theto intraday intensity severe. The minimum value ofand radiation intensity was 0 W/mday, the period 2017, respectively. Overall, radiation intensity to2 decline fromof6radiation February 2017 (Figure variation of Maximum radiation intensity was more severe. The minimum intensity was polar night. radiation intensity wascontinued 1073 W/m on 10value February 2017, which was in the0 2, which 2 in the 2 the 19a). Then, from 27 May to 9 July, radiation intensity was 0 W/m was during polar night. W/m period of polar night. Maximum radiation intensity was 1073 W/m on 10 February period of the polar day. The average radiation intensity from 6 February 2017 to 9 July 2017 was Although observation system was initially deployed during the polar day, the intraday 114 W/m2the . variation of radiation intensity was more severe. The minimum value of radiation intensity was 0 W/m2 in the period of polar night. Maximum radiation intensity was 1073 W/m2 on 10 February

Appl. Sci. 2018, 8, x FOR PEER REVIEW

20 of 23

Appl. Sci. 2018,was 8, x FOR PEERperiod REVIEWof the polar day. The average radiation intensity from 6 February 20 of 23 2017, which in the 2017 to 9 July 2017 was 114 W/m2. 2017, which was in the period of the polar day. The average radiation intensity from 6 February Appl. Sci.to 2018, 8, 2376 20 of 23 2017 9 July 2017 was 114 W/m2.

Figure 19. (a,b) Daily radiation intensity measured by the observing system, and solar power generated by the power system from 6 measured February by 2017 9 July 2017, respectively. between the Figure 19. (a,b) Daily radiation intensity thetoobserving system, and solar Area power generated Figure 19. (a,b) Daily radiation intensity measured by the observing system, and solar power red lines is the polar night. by the power system from 6 February 2017 to 9 July 2017, respectively. Area between the red lines is generated by the power system from 6 February 2017 to 9 July 2017, respectively. Area between the the polar night. lines is 19b, the polar InredFigure the night. trends of solar power and radiation intensity were considered to be

In Figure 19b, the trends of solar power and radiation intensity considered bepolar gradually gradually reduced. The minimum value of solar power was 0 KWhwere in the period of to the night In Figure 19b, the trends of solar power radiation intensity were considered to be reduced. minimum value of solar was 0 and KWh in February the period of the polarwas night (June and (June andThe July). Maximum solar powerpower was 4.2 KWh on 10 2017, which in the period gradually reduced. The minimum value of solar power was 0 KWh in the period of the polar night July). was 4.2 power KWh on 10 February which the KWh. periodThe of the polar of theMaximum polar day.solar The power average solar from February2017, to July 2017was wasin1.25 trend of (June and July). Maximum solar power was 4.2 KWh2017 on 10 February 2017,The which was the period day. The average solar power from to July was 1.25 KWh. trend ofin calculated calculated solar power based on theFebruary radiation-intensity data obtained in 2017 was consistent with of the polar day.on The average solar powerdata from February to July 2017 was 1.25 KWh. The trend of solar power based radiation-intensity obtained 2017 was consistent with theparticularity calculation the calculation using the average data from NASA, except forinsolar power in January. The calculated solar power based on the radiation-intensity data obtained in 2017 was consistent with using data from NASA, except for solar powermay in January. The particularity the installation of theaverage installation location of the observing system cause differences in the of calculated values the calculation using average NASA, exceptin forthe solar power in January. The particularity location the observing systemdata mayfrom cause differences calculated values between measured betweenof measured and average data. of the installation location of the observing system may cause differences in the calculated values and average data. The wind speed and monthly wind power are shown in Figure 20, respectively. As shown in between measured and data. The20a, wind andaverage monthly are shown in Figure 20, respectively. As shown in Figure thespeed minimum value of wind power speed was 0 m/s, and the maximum wind speed was 69.7 The wind speed and monthly wind power shown in maximum Figure 20, wind respectively. As69.7 shown Figure the minimum value of wind speed was60are m/s, and 2017 the speed was m/sin m/s on20a, 7 March 2017. Average wind speed from February to 9 July 2017 was 22.7 m/s (Figure Figure 20a, the Average minimum value of wind speed was 02017 m/s,toand the 2017 maximum wind speed was 69.7 on 7 March 2017. speed from February 9electricity July was 22.7 m/s (Figure 20a). 20a). The 93.5% of windwind energy can be 6used to generate using the power system m/s on 7 of March 2017. Average wind to speed from electricity 6 February 2017the to 9power July 2017 was(dark-blue 22.7 m/s (Figure The 93.5% wind system area (dark-blue area in energy Figure can 19a).beInused Figuregenerate 20b, the difference using between actual wind-power generation 20a). The 93.5% of wind energy can be used to generate electricity using the power system in Figure 19a). In Figure 20b, the difference between actual wind-power generation and calculated and calculated wind power was in a small range, except that the maximum wind power was 13.46 (dark-blue area in aFigure 19a). In except Figure that 20b,the themaximum difference wind between actual wind-power wind waswhich in small power KWh generation in April, KWh power in April, was range, about twice the calculated wind power. Sincewas the13.46 deployment of the and calculated wind power was in awind smallpower. range,Since except that the maximum power was 13.46 which was about twice calculated the thewind observation observation system wasthe until 9 July 2017, the wind energy ofdeployment the currentof month was much system smaller KWh in9April, which was twice thecurrent calculated wind Since the of the was Julypower 2017, the windabout energy of the month was power. much smaller thandeployment the wind power thanuntil the wind calculated by average data. observation system was calculated by average data.until 9 July 2017, the wind energy of the current month was much smaller than the wind power calculated by average data.

Figure 20. (a) Wind speed measured by the observing system, and (b) monthly wind power generated by the power system and calculated by the average data from 6 February 2017 to 9 July 2017, respectively. Area between the red lines is the polar night. The dark-blue area represents the available wind.

(June and July). Maximum solar power was 4.2 KWh on 10 February 2017, which was in the period of the polar day. The average solar power from February to July 2017 was 1.25 KWh. The trend of calculated solar power based on the radiation-intensity data obtained in 2017 was consistent with the calculation using average data from NASA, except for solar power in January. The particularity of the installation location of the observing system may cause differences in the calculated values Appl. Sci. 2018, 8, 2376 21 of 23 between measured and average data. The wind speed and monthly wind power are shown in Figure 20, respectively. As shown in Figure the minimum of wind speed 0 m/s, in and the maximum wind speed was 69.7 The20a, power supply of value the observing systemwas is shown Figure 21. The supply-voltage range m/s on 7 for March Average windtospeed from 6 February 201713 toto 9 July was 22.7inm/s (Figure required the 2017. observing system operate normally is from 20 V.2017 As shown Figure 21, 20a). The 93.5% of wind energy can be used to generate electricity using the power system the voltage of the output of the power system continued to decline from February to July. From (dark-blue area in Figure In Figure 20b,the thevoltage difference actual wind-power the date of deployment to19a). the end of April, was between a fluctuant downward trendgeneration owing to andsuperposition calculated wind power was in aenergy. small range, except that theintensity maximum wind power wasit 13.46 the of solar and wind After that, radiation was reduced until was 2 which wasduring aboutwhich twicewind the calculated Since the deployment of the 0KWh W/mininApril, the polar night, power waswind alone power. as a source of energy. In the operating observation system was system until 9 July 2017, the wind energy of was the current cycle of the observation in Antarctica, supply voltage always month higher was thanmuch 13 V, smaller that is, than the wind power calculated average the data. the power-supply system could by guarantee operation of the observation system in the range of ◦ −50–30 C.

Figure 21. Power voltage of the observing system. Area between the red lines is the polar night.

7. Conclusions In this study, we proposed a standalone hybrid wind–solar system for the polar regions. The operation of the observation system in the Arctic Ocean in 2016 and the meteorological environment of the deployment positions were analyzed. A new observation system and a power-supply system with a hybrid solar and wind charging circuit were designed to achieve long-term stable operation in the polar regions at a temperature range of −50–30 ◦ C. We presented a comprehensive understanding of the low-temperature performance and characteristics of batteries, including battery capacity at different temperatures, and the charging characteristics of the lead–acid battery. Based on the low-temperature characteristics of the battery, a charging strategy adapted for the battery was proposed. We evaluated the low-temperature performances of the charging circuit to implement charging strategy. The detection and ADC circuits of the charging circuit have excellent stabilities in the range of −50–30 ◦ C. The simulation calculation of power generation of the standalone hybrid wind–solar system was presented in this study. The results of the simulation calculation reveal that wind power had a relatively higher share of energy production than the PV. The observing system and the standalone hybrid wind–solar system were applied to the observation of ice-matter balance in Zhongshan Station, East Antarctica from February to July 2017. The power generation of the PV and wind turbine during the field experiment was calculated. The case study shows that the standalone hybrid wind–solar system can be proven to have high stability, and also indicates that our study on battery characteristics in low-temperature environments could greatly improve battery reliability. This paper demonstrates that the standalone hybrid solar–wind system is an ideal solution for automatic observation systems in the polar regions. In this study, the standalone hybrid solar–wind system does not emit any pollutants during its lifecycle. However, lead–acid batteries were used as energy-storage units for the power system. Although we designed a leak-proof container to place small batteries to avoid polluting the environment, more environmentally friendly energy storage unit needs to be adopted in future work. More renewable energy, such as wave energy, temperature-difference energy, and salinity-difference energy, in the polar region can be used to design power-supply systems with larger installed capacity.

Appl. Sci. 2018, 8, 2376

22 of 23

Author Contributions: conceptualization, G.Z.; methodology, G.Z.; software, G.Z.; validation, G.Z.; formal analysis, Y.D.; investigation, Y.C.; writing—original draft preparation, G.Z.; funding acquisition, Y.D., X.C., and Y.C. Funding: This research was funded by the National Natural Science Foundation of China, grant number 41606220, 41776199; the National Key Research and Development Program of China, grant number 2016YFC1402702, 2016YFC1400303; and the Natural Science Foundation of Shanxi, grant number 201701D121127. Acknowledgments: The authors would like to thank the Chinese National Arctic Research Expeditions and Chinese National Antarctica Research Expeditions for supporting the deployment of the observing and power-supply systems. Comments from the anonymous reviewers and the editor are also gratefully appreciated. Conflicts of Interest: The authors declare no conflict of interest.

References 1.

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

14. 15. 16.

17. 18.

Rothrock, D.A.; Percival, D.B.; Wensnahan, M. The decline in arctic sea-ice thickness: Separating the spatial, annual, and interannual variability in a quarter century of submarine data. J. Geophys. Res. Oceans 2008, 113, C05003. [CrossRef] Perovich, D.K.; Grenfell, T.C.; Light, B. Transpolar observations of the morphological properties of Arctic sea ice. J. Geophys. Res. 2009, 114, C00A04. [CrossRef] Haas, C.; Hendricks, S.; Eicken, H. Synoptic airborne thickness surveys reveal state of Arctic sea ice cover. Geophys. Res. Lett. 2010, 37. [CrossRef] Haas, C.; Pfaffling, A.; Hendricks, S. Reduced ice thickness in Arctic Transpolar Drift favours rapid ice retreat. Geophys. Res. Lett. 2008, 35, L17501. [CrossRef] Rack, W.; Haas, C.; Langhorne, P. Airborne thickness and freeboard measurements over the McMurdo Ice Shelf, Antarctica, and implications for ice density. J. Geophys. Res. Oceans 2013. [CrossRef] Richter-Menge, J.A.; Perovich, D.K.; Elder, B. Ice mass balance buoys: A tool for measuring and attributing changes in the thickness of the Arctic sea ice cover. Ann. Glaciol. 2006, 44, 205–210. [CrossRef] Jackson, K.; Wilkinson, J.; Maksym, T. A novel and low cost sea ice mass balance buoy. J. Atmos. Ocean. Technol. 2013, 30, 13825. [CrossRef] Wang, H.Z.; Chen, Y.; Song, H. A Fiber Optic Spectrometry System for Measuring Irradiance Distributions in Sea Ice Environment. Atmos. Ocean. Technol. 2014, 31, 2844. [CrossRef] Lei, R.B.; Li, N.; Heil, P. Multiyear sea ice thermal regimes and oceanic heat flux derived from an ice mass balance buoy in the Arctic Ocean. J. Geophys. Res. Oceans 2014, 119, 537–547. [CrossRef] Lei, R.B.; Cheng, B.; Heil, P. Seasonal and Interannual Variations of Sea Ice Mass Balance From the Central Arctic to the Greenland Sea. J. Geophys. Res. Oceans 2018. [CrossRef] Borowy, B.S.; Salameh, Z.M. Optimum photovoltaic array size for a hybrid wind/PV system. IEEE Trans. Energy Convers. 1994, 9, 482–488. [CrossRef] Deshmukh, M.K.; Deshmukh, S.S. Modeling of hybrid renewable energy systems. Renew. Sustain. Energy Rev. 2008, 12, 235–249. [CrossRef] Bilal, B.O.; Sambou, V.; Ndiaya, P.A. Optimal design of a hybrid solar–wind-battery system using the minimization of the annualized cost system and the minimization of the loss of power supply probability (LPSP). Renew. Energy 2010, 35, 2388–2390. [CrossRef] Diaf, S.; Diaf, D.; Belhamel, M. A methodology for optimal sizing of autonomous hybrid PV/wind system. Energy Policy 2007, 35, 1303–1307. [CrossRef] Dihrab, S.S.; Sopian, K. Electricity generation of hybrid PV/wind systems in Iraq. Renew. Energy 2010, 35, 1303–1307. [CrossRef] Ghoddami, H.; Delghavi, M.B.; Yazdain, A. An integrated wind-photovoltaic-battery system with reduced power-electronic interface and fast control for grid-tied and off-grid applications. Renew. Energy 2012, 45, 128–137. [CrossRef] Sreeraj, E.S.; Chatterjee, K.; Bandyopadhyay, S. Design of isolated renewable hybrid power systems. Sol. Energy 2010, 84, 1124–1136. [CrossRef] Gunasekaran, M.; Ismail, H.M.; Chokkalingam, B. Energy Management Strategy for Rural Communities’ DC Micro Grid Power System Structure with Maximum Penetration of Renewable Energy Sources. Appl. Sci. 2018, 8, 585–608. [CrossRef]

Appl. Sci. 2018, 8, 2376

19. 20. 21.

23 of 23

Vasiliev, M.; Alameh, K.; Nur-E-Alam, M. Spectrally-Selective Energy-Harvesting Solar Windows for Public Infrastructure Applications. Appl. Sci. 2018, 8, 849–864. [CrossRef] Kadir, A.E.; Hu, A.P. A Power Processing Circuit for Indoor Wi-Fi Energy Harvesting for Ultra-Low Power Wireless Sensors. Appl. Sci. 2017, 7, 827–845. [CrossRef] Ma, T.; Yang, Y.X.; Lu, L. Technical feasibility study on a standalone hybrid solar-wind system with pumped hydro storage for a remote island in Hong Kong. Renew. Energy 2014, 69, 7–15. [CrossRef] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).