Tolesa and Workneh (2017) investigated the effect of IAC+EC on the quality of stored tomato fruit over a 30-day period compared to ambient conditions. The tomatoes stored under IAC+EC showed higher firmness and hue angle, maintained lower concentration of sugars, had lower physiological weight loss and improved shelf-life and marketability as compared to ambient conditions. Sibanda and Workneh (2020) tested the performance of such a system and it provided storage conditions of temperature of 15.7 to 16.4°C and relative humidity of 89.6 to 93.8%. Both studies provided the efficacy of the system except the energy requirements. Incorporation of an IHE requires energy input that grid electricity could provide. However, in remote off-grid areas, renewable energy sources like solar can be utilised through use of stand-alone or hybrid solar photovoltaic (SPV) systems as they generate power with less maintenance and operational costs (Khatib et al., 2016). The integration of IAC+EC with solar energy in sub-Saharan Africa (SSA) where the average solar radiation is 4.5 to 6.5 kWh.m^-2 for 6 to 7 h a day can provide a cooling facility to SSF in areas with no access to grid electricity (Saxena et al., 2013). There is no information on energy requirements for such a system derived from actual performance data.

The sizing of stand-alone SPV systems depends on the efficiency of modules and other factors, such as ambient temperature, the quality and quantity of solar irradiance available in the location (Almarshoud, 2016). The efficiency of solar energy conversion depends on whether the module is monocrystalline, polycrystalline or thin-films type (Huang et al., 2013). Monocrystalline modules have the highest energy conversion efficiency; polycrystalline is in between, whilst thin-films are both least expensive and efficient in comparison (Bai et al., 2016). For solar arrays to produce maximum power output, they must be at   an   optimal   tilt   angle   to   trap  maximum  radiation  (Tripathy et al., 2017). The optimal tilt angle depends on the season and the latitude of the area (Kaddoura et al., 2016). For higher power output, incorporation of solar trackers allows automatic adjusting of the collector tilt angle to follow the sun’s change in elevation during the day (Pedro et al., 2016). In this study for cost reduction purposes, the optimum tilt angle was determined from historical data as provided by Schulze et al. (1999). The aggregate sun-oriented radiation received at a given geographical location varies depending on the length of the insolation on a specific day, the power of sunlight-based vitality, and the day or time of the year (Safa et al., 2016; Tripathy et al., 2017).

There is limited literature on sizing SPV systems for cooling; however, there are several published papers on utilisation of solar energy for water pumping applications for irrigation. There are various methods for sizing the SPV array for water pumping, some are simple, others in between while others are complex. One such simple method considers the monthly average of daily solar insolation to determine the associated peak sun hours (Hankins, 2010). Almarshoud (2016) summarized other simpler methods from literature on practical methods for SPV array sizing based on the peak sun hours. Other research work used the monthly average of daily global solar radiation and sunshine hours to size the SPV array for each month of the year, and then selected the highest values to increase the reliability in low radiation months. There are methods that are not simple but in between in sizing the SPV system required to operate a water pump like use of monthly average solar radiation on tilted surfaces of the worst month of the year (Munzer et al., 2013). Some authors have used the worst monthly average of solar radiation in addition to the maximum monthly water demand for sizing (Abidin and Yesilata, 2004). In other instances, the worst daily collectable solar energy in addition to the system efficiency has been applied (Campana et al., 2013). Complex methods include development of simple algorithms based on the monthly average of global horizontal irradiance only, while more complex approaches use dynamic programming to size the SPV array based on the average of solar energy available per day, taking into account all relevant elements of SPV pumping system (Zvonimir and Margeta, 2007; Kaldellis et al., 2007). Eltawil and Samuel (2007) sized a vapour-compression refrigeration system for storage of potatoes by estimating the battery capacity, the average load energy and the ampere-hour required per day, from the load profile. Instead of using the peak rated power of the modules at standard operating conditions they used several complex multiple regression equations to predict the SPV panel power output and its temperature in addition to energy consumption.

This paper proposes a simple and accurate approach for sizing the SPV array using the actual power requirements of the electrical accessories. When an IAC+EC  system   starts   operating  accessories  like  the water pump and heat exchanger will draw high current so, the SPV array must be able to provide the required current. The proposed approach will, integrate the simple approaches mentioned earlier; use the solar radiation data measured in Pietermaritzburg (PMB) over 50 years by Schulze et al. (1999) and the actual solar radiation measured in this experiment for the month of June and September. The accurate sizing of the SPV array is essential, because under sizing will make the IAC+EC system unable to power the electrical accessories resulting in failure to provide optimum storage conditions for fresh produce. On the other hand, over sizing of SPV array leads to unnecessary cost incurred in acquisition of additional modules and batteries (Almarshoud, 2016). Specifically, this study will provide actual data on the performance of SPV in powering a 3.8-ton (53 m³) sized storage chamber for tomatoes. The main objective of this study is to design, assemble and evaluate the performance of SPV-battery based IAC+EC system.

 

Experimental study

To determine the actual power requirements to operate an IAC+EC system, a case study for Pietermaritzburg, in South Africa was chosen. The IAC+EC system was used for cold storage of tomatoes over a period of 28 days in the summer month of September. The SPV system to power the IAC+EC system consisted of solar panels, solar charge controller and inverter, as well as a battery bank facility. This system was constructed and assembled on site at Ukulinga research station at the University of KwaZulu Natal, in PMB. The site is located at 30°24’S, 29°24’E at an altitude of 721 m. The SPV system and a battery bank facility provided electrical energy to the IHE, centrifugal pump and fans to facilitate airflow across the storage chamber of the IAC+EC.

Indirect air-cooling combined with evaporative cooling system set up

The IAC+EC consisted of a storage chamber, IHE, multiple charcoal cooling pads, buried water tank, a pump and two fans as shown in schematic diagram of Figure 1. The storage chamber had white double-jacket walls and a roof of 1 mm zintec (mild steel) both on the outside and inside. The floor was made of concrete mortar. The inner dimensions of the unit were 2340 mm high × 5880 mm long × 3880 mm wide giving a storage chamber volume of about 53-m³ with a holding capacity of 3.8 tonnes of tomatoes. The cooler had a 60 mm zinc wall thickness with 58 mm polyurethane insulation in between the zintec layers. The IAC+EC system design specifications provided environmental conditions of dry bulb temperature of 14 to 20°C, relative humidity of 89 to 94% and cooling efficiency of 88 to 96% depending on the time of the day.

Solar photovoltaic system set up

Nine solar panels were connected in three series-three strings arrangement in conjunction with a with a three-string series 48 V battery bank facility (batteries of 230 AH) were used to power the electrical appliances. The characteristics and dimensions of the modules are shown in Table 1 while Figure 2 shows the solar panels and battery connections. 

The solar array system was designed to provide enough power to operate electrical appliances of the IAC+EC system. These included a 1730 W Lytron indirect heat exchanger (M14-120) with 33 W fan (OW354) to facilitate airflow across, a 290 W second fan (6/P3HL/25/PA) ventilating the storage chamber and 260 W centrifugal water pump (Pedrollo SPVm 55) to reticulate water through the cooling pads. From 08.00 to 17.00 h, the SPV system powered the electrical appliances and thereafter the system was powered by battery bank facility. The hybrid SPV system was optimized by considering the number and sizes of modules as well as batteries required and balancing that with the system voltage and current. A number of combinations were considered as recommended by Goel and Sharma (2017). From these permutations, a three-series-three-string connection was chosen as it gave the highest output of 3503.8 W and did not overload the available solar charge controller.

The battery capacity was determined with reference to the electrical appliances specifications for the daily watt-hours at 50% discharge using a 48-V system and the available battery in the market, which was a 230 AH with a 90% efficiency. The number of batteries required to operate the IAC+EC system with 3.8 tons of tomatoes was determined as twelve. The total load from the electrical appliances was determined as 2343 W and the allowable battery discharge was 28116 W and such a system would produce 4196.4 W. h^-1 if the sunshine hours are 6.7 to cool 3.8 tons of tomatoes.

The solar modules were dusted and dirt removed from the surface following which they were installed away from trees and buildings on a fixed rectangular metal manual tilt-frame and mounted facing south on an inclined angle of tilt = -15° (Sun et al., 2016; Ronoh, 2017). In order to optimize solar radiation, the tilt angle varied at ±15° to the latitude of the area. In the design four different tilts: horizontal (0°) and tilt angle -14.6006, -29.6006 and 44.6006° were considered and from calculations the highest insolation was obtained at tilt angle of -14.60°. Therefore, a tilt angle of tilt = -15° was chosen to determine the optimal power and energy output. The DC power generated from the SPV modules was transmitted to the solar charge controller prior to charging the solar batteries and thereafter the inverter converted the generated DC power to AC power. Figure 3 is a schematic layout of the solar system and shows how the rest of the components were connected. The solar charge controller (60A, 145 VDC, SANTAKUPS PC16-6015F) ensured constant voltage and current to the load from the batteries according to Deveci et al. (2015). The chosen sinewave inverter (125A, 5 kW) matched the system in terms of voltage input, AC power output, frequency and voltage regulation (Chandel et al., 2015). Twelve fully charged 230 AH batteries, which were arranged as a three-string series 48V system, were used to start up the SPV system. These batteries also temporarily stored energy generated by solar panels for overnight use.

Determination of theoretical design power and energy

Monitoring solar radiation data for PMB took place during the year 2018. However, for this study, only data for June and September was used since PMB receives the least amount of solar radiation in June while September was the month of the experiment. Data recording occurred at the research site using South African Weather Services – Agricultural Research Council (SAWS-ARC) weather station based at the research site. The solar radiation data was used to determine the actual performance of the SPV system. The solar radiation values recorded by Schulze et al. (1999) over 50 years’ for PMB were extracted to obtain values for solar radiation received at 80% probability of exceedance at different tilt angles. These values were used to determine the theoretical power. The theoretical  power   represents   the   power   available   to   run  the electrical components of the IAC+EC system at the design stage. Equations 1 and 2 were used to determine the output power and energy based on the average theoretical design solar radiation for the two months at different tilt angles. The power output from Equation 1 was used to size the solar modules.

where P_out = average monthly power output (W), η_panel = overall SPV module efficiency (=0.1522); N_panels = number of SPV modules, A_panel = area of the module, G = solar radiation (W.m^-2).

where E_produced = energy produce on a day length D₁ (Wh. m⁻²) and D₁ = average monthly day length (h).

The theoretical power out of the SPV is the input power of the solar charge controller.  Equation 3 determines this power by incorporating the efficiency of the charge controller.

where P_out = power output from controller (W), Æž_{c =} efficiency  of  the charge controller from the supplier (90%) and P_in = power input to the charge controller.

The output power from the charge controller is the input power to the inverter.  The output power of the inverter was calculated by incorporating the efficiency of the inverter. The power output from Equation 4 should be higher than the power requirements of electrical appliances.

where P_out = power output from inverter (W); È _I = efficiency of the inverter from the supplier (90%) and P_in = power input to the inverter.

Performance evaluation

Measurement of parameters

On the days of the experiment (1 to 30 September 2018), the solar modules supplied the energy requirements during the day from 08.00 to 17.00 h. Thereafter, the battery bank supplied energy until 22.00 h when the system was switched off until 08.00 h of the following day as ambient temperature had fallen between 18 and 20°C. Fresh produce like tomatoes can tolerate such ambient temperatures for a short time. Therefore, there was no need for further cooling overnight time as recommended by Punja et al. (2016). The solar irradiance and ambient temperature for the period of the experiment was obtained at SAWS-ARC weather station located at the Ukulinga Research Station in PMB. The values were used to determine solar irradiance power using equation 4. The other meteorological parameter data like the temperature of the solar panels was measured using an infrared thermometer with K thermocouple (Fluke 63). Thermocouples connected to data loggers measured the SPV module temperatures at hour intervals to determine the influence of module temperature on the performance of solar systems (Sun et al., 2016).

Measurement of module current and voltage at different points

For the SPV electrical system, there were four positions (Figure 4) identified  to  evaluate  the performance of the solar array system. A digital multi-meter (Fluke 381) measured both open circuit voltage and current under these different positions. Ohms law (Equation 5) was used to determine power:

where V = voltage (v) and I = current (A).

The test procedures to be followed were:

(The power output tests were done by measuring both the voltage and  current  at  different  points  and  these  values  were   used  to calculate the power output using the Ohm’s Law.

The test procedures followed were:

(a) Measurements at position 1 of the system (exit point of solar panels [1a] and the input side of the solar charge controller [1b]). The voltage and current measured at this point were used to calculate solar modules power output and was compared with the theoretical calculation of the power output from the solar modules;

(b) Position 2 measures both voltage and current at the exit of the charge controller and the input of the inverter.

(c) Position 3 read voltage and current to and from the batteries, and

(d) Position 4a read current and voltage between the inverter and heat exchanger, pump and fans.

Measurements at this point provide how much power the appliances draw. Position 4b reads the current and the voltage drawn by electrical appliances.

 

Theoretical power and energy

The performance of SPV systems depends on the tilt angle and orientation of the array. In studying the effect of insolation on modules a solar tracking device helps in adjusting the position of the solar panels so that the highest possible energy output obtains compared to a fixed SPV system. This necessitates that installations of the modules be at an optimal tilt angle that maximizes the solar radiation captured by SPV panels. Solar radiation data for PMB from Schulze et al. (1999) and from SAWS-ARC weather station was collated at tilt positions +15° for June and -15° September to determine probability of exceedance which information was used to calculate theoretical power and energy.  It is important to base the design of a SPV system on a higher chance that the expected solar radiation will occur or exceeded in the given period and therefore a probability of exceedance is considered. The solar radiation data at 80% probability of exceedance at different tilt angles data for PMB was considered and only data for maximum theoretical output was extracted. Table 2 summarises the data of the probability of exceedance of monthly solar radiation for June and September. From Table 2 at 20% of the time in each month there was a higher radiation received in PMB than in 50 and 80% of the time. As the exceedance probability increased, the amount of radiation received decreased.  Relatively lower percentages were recorded at high irradiance  levels  and  the  converse  is  true. The high irradiance levels, are associated with a direct beam component, that is spread more widely with very small individual frequency percentages. For the purpose of calculation, 80% exceedance probability was used, as the values are closer to reality as is possible. From Table 2, the average optimal solar radiation received at 80% probability of exceedance in PMB in June and September were 522.81 W.m^-2 at tilt = +15° and 1047.49 W.m^-2 at tilt = -15°, respectively.  The optimal power for the months of June and September in PMB are 1468 and 2 941.7 W, respectively and from Equation 2, this translated to optimum energy output of 628.5 and 1068.1 Wh. m⁻² for the two months, respectively.

The theoretical power and energy were low in June because solar insolation levels were low. To generate adequate energy under such circumstances would require more solar modules and this would increase the cost of installation. The sizing of stand-alone SPV considers meeting electrical loads requirements with lowest average daily solar insolation on the array surface, which usually occur during winter months. However, the temperatures are also generally low in winter (June), and the maximum temperatures are 16 to 20°C in PMB. Under such conditions for tomatoes and many tropical and sub-tropical FV in SSA, either no cooling or minimal cooling will be required during short periods as alluded to by Punja et al. (2016). To ensure optimization of the solar insolation, a switch could be incorporated to the system coupling the electrical load (pump, fans and heat exchanger) to the SPV array system. This will allow the SPV system to switch off when the battery bank facility is fully charged. Optimising the system is important, as the costs of installation are reduced to ensure the IAC+EC systems translate to a low cost cooling technology (Chandel et al., 2015; Goel and Sharma, 2017).  

The theoretical power output for the month of September is very significant for design as this month is the beginning of summer and higher solar radiation is received in subsequent months until April of the following year when temperatures begin to fall. In subsequent months, the theoretical power output is higher as the area receives more solar irradiation and this coincides with higher cooling loads as the ambient temperature is also relatively higher. This is the reason why most of the large-scale SPV systems are built in arid and semi-arid areas, where the solar insolation levels are high (Sayyah et al., 2014). However, caution has to be taken as high ambient  temperature  affects  performance  of   the  SPV system due to high cell temperature (Ronoh, 2017). 

From Equation 3, the theoretical output power from the charge controller was:

in September. From Equation 4, the optimum output power of the inverter is in September. This means that the theoretical power available to run the electrical components at 80% probability of exceedance in September in PMB was 2382.8 W.

Variation of insolation with ambient temperature

Ambient air temperature and solar radiation outside the IAC+EC system around the SPV system was studied in the month of September 2018. It was observed that ambient temperatures and solar irradiance were low in the morning and increased from 08.00 h to between 12.00 to 14.00 h and thereafter decreased towards 18.00 h (Figure 5). Ambient temperature increased due to increasing incident solar radiation from morning until afternoon 13.00 to 14.00 h and then decreased from then onwards towards sunset as also confirmed by Madhava et al. (2017). The average insolation values rose from 293.4 W.m^-2 at 08.00 h in the morning to 1 059.6 W.m^-2 at mid-day. Eltawil and Samuel (2007) observed a similar trend. At any location like PMB, the length of the path the radiation takes from source  to  ground  level  varies  with time of the day as the spectrum of the radiation changes through each day because of the changing absorption and scattering path length (Ronoh, 2017). Figure 5 relates to data obtained on clear days of September where the solar insolation increased from early morning to a peak at midday and then decreases to zero at night. The peak is achieved at midday as the sun is overhead and its path length is shortened. At midday, less solar radiation is scattered or absorbed by atmospheric mediums, and more radiation that is direct reaches the modules compared to any other time of the day and Olomiyesan et al. (2015) complements these results. The highest average solar irradiance received in September 2018 of 1 059.6 W.m^-2 was slightly higher than the average value of 1 047.9 W.m^-2 received over 50 years at 80% exceedance for the same month. This implies that the average insolation received in this month should produce enough power and energy for the designed electrical appliances.

Solar photovoltaic module power and solar irradiance power

The SPV module power (P_module) and solar irradiance power (P_irridiance) were studied in the month of September on clear and, sunny days selected (11 days) for the experiment. The P_module was obtained by measuring voltage and current between the solar modules and solar charge controller while the P_irridance values were obtained by converting solar irradiance values in Figure 5 to power by  using  Equation  1.  Figure  6  shows  the  variation  of P_module (W) and the P_irridance (W) during the period of study from 08.00 to 18.00 h. The P_module increased with the P_irridance to a peak between 12.00 and 14.00 h and decreased thereafter as the solar radiation intensity decreased. The results from the present study agree with findings of Charfi et al. (2018) in Tunisia who obtained similar trends. The P_irridance received and P_module output had very similar trends with the maximum and minimum values at the same hours during the period of the experiment. This shows that the amount of electricity generated by SPV system is largely depended on the availability of the solar energy at a particular location as corroborated by Li et al. (2005) and Chaabane et al. (2019). The design power which is the maximum average solar power received in PMB over 50 years in the month of September at 80% probability of exceedance is 2941.9 W (Table 2) while the peak P_irridance in the month of September 2018 was 2975.4 W (Figure 6).  Therefore, P_irridance was just 1% higher the design power. This shows the importance of basing the design on higher percentage of probability of exceedance.

The peak P_module was 2 873.5 W, which was about 3.5% lower than the peak P_irridance of 2975.4 W during the period of the experiment. The difference between the P_module and the P_irridance is attributable to the efficiency of the SPV system that peaks at 14.9% (Figure 8) which value was lower than the rated solar panel efficiency of 15 to 19%.  The other contributors are environmental factors including module temperature, soiling material accumulating on the module surfaces, resistance in the wiring and connections and in some instances, modules of  the   same  type  have  slight  differences  in  electrical characteristics. Ghazi et al. (2014) mentioned that solar modules need regular cleaning as soiling is regarded as one of the significant contributors to reduction of the power output of SPV systems as it reduces the solar radiation reaching the surface of modules. When modules are soiled, the dust particles deposited on the surface absorb and scatter the incoming incident light and this might have contributed to the reduction of the P_module value (Sayyah et al., 2014). The peak P_module of 2873.5 W was 24% higher the design load for electrical appliances of 2343 W. However, it is important to note that the efficiencies of the controller and inverter could account for 24% more power generated by the modules.

Solar photovoltaic module parameters and ambient temperatures

The P_module output and module temperature increased with ambient temperature to about 32°C (Figure 7), which coincided with the highest ambient temperature at midday. The maximum P_module output occurred at 31 to 32°C ambient temperature (Figure 6) and the system was most efficient at this period (Figure 8).  Thereafter both power output and ambient temperature declined after midday as shown in Figure 5.  Ya’acob et al. (2014) made similar observations in Malaysia where they had the highest generated power data at midday with ambient temperature at 32.5 to 34.5°C. The P_module output increased as short circuit current increased with insolation due to the increase in the number of photons generating the current. Increased solar panel temperature also increases the kinetic energy of the photons resulting in increased current and module output. Figure 7 shows that P_module increased with module temperature until 25°C and decreased thereafter.

The use of SPV systems is increasing as installations costs are decreasing and the application is finding expression in remote and isolated communities and in new farming setting ups of SSF with no access to cooling facilities. Electricity supply is of great concern, as it is inadequate and in SSA, connection to the national grid for most SSF seems highly unlikely in the near future. This has turned interest to renewable energy sources like solar as a means of bridging the energy gap and providing environmentally friendly energy. In this study, a SPV system IAC+EC was evaluated based on actual performance. Furthermore, this experiment  explored  the possibility of integrating solar energy to power IAC+EC system targeting SSF in remote areas with no access to grid electricity.

Most of the literature does not give actual values of energy required by different cooling systems, but merely states which cooling systems are more energy intensive than others are. Energy required to operate modern cooling systems are greater than the energy required to operate IAC+EC system. The SPV systems used in the study supplied energy during the critical period of the day when temperatures were high from 08.00 to 22.00 h. To cool one ton of tomatoes using IAC+EC requires 1354.4 W. h^-1 and the batteries had to store 5146.6 W. h^-1 to provide energy for the 3.8-ton storage chamber to cool tomatoes from 17.00 to 22.00 h when the IAC+EC system was switched off. The efficiency of the solar panels was 14.7%. The energy to power an IAC+EC system relates to the size of the solar array system required to provide the energy and the cost of the system. The study concludes that combinations of the solar array system can power the cooling system at daytime during summer season and the excess energy can be stored in the battery to run the system for another five hours into the night. A bigger system is required to run all-night and in the near future, this will be possible, as the prices of modules continue to decline.

Therefore, where grid electricity or other commercial energy sources are unavailable and solar energy is available, IAC+EC system is a viable alternative to these more complex and costly modern day cooling systems. This shows that stand alone SPV systems have an expression in rural, dispersed and remote areas where grid electricity supply may not be readily accessible. Integrated solar and indirect evaporative cooling is an attractive alternative for SSF with no access to cooling technologies in developing countries, especially in Africa.

 

The authors have not declared any conflict of interests.

 

The authors are grateful to the Agricultural Research Council for funding and University of KwaZulu Natal, School of Engineering in Pietermaritzburg for allowing the research team to review the work done on evaporative cooling at the university research farm.