Design and Fabrication of a Control System for a Photovoltaic-Greenhouse Solar Dryer and Performance Evaluation of the Dryer

Abstract

Solar dryers have become important components for the delivery of dried agricultural products of high quality. They generally are designed to maximize the production of a heating system based on solar radiation. This study involved the design, building and testing of a control system to allow the monitoring of temperature, solar radiations and control of humidity in a Photovoltaic (PV)-Greenhouse Solar Dryer (PVGSD). This was accomplished by assembling a solar charge controller, three solar panels, a power inverter, and a battery system. An Arduino UNO Rev3, DHT22-sensor, solderless breadboard, Liquid Crystal Display (LCD) screen, jumper cables, Light-Emitting Diode (LED) red, Global System for Mobile (GSM) communication module, 10k Ω, 220 Ω and 1k Ω resistors and two 5V one channel relays formed the control system. These were assembled in a multi-functional workstation linked to the solar dryer. The GSM communication module was an innovation to allow connection to the internet over General Packet Radio Service (GPRS) network and send/receive SMS. It was used to control the fans remotely and further allowed logging of periodic sensor data (temperature, humidity, voltage and solar radiation values) by sending SMS to a phone number and server. Dryer and product characteristics were measured including temperature, humidity, solar radiation and moisture content. The temperature and humidity profile were monitored for 11 days in the empty PVGSD without the workstation and Open Sun Drying (OSD) showed that the PVGSD recorded the highest temperature of 69oC occurring between 12:00 and 14:00 hours GMT while the OSD recorded at 41.5oC. The highest relative humidity recorded in the PVGSD was 75.5% and 83% for OSD observed at night from 20:00 hours to 5:00 hours GMT. Two commodities, cassava (slices and chunks) and red pepper were dried to evaluate the dryer efficiency, and this was compared with OSD and existing greenhouse type solar dryer (GSD). The cassava slices dried faster than the cassava chunks in all the drying methods indicating that sample size influenced drying rate. PVGSD had a faster drying rate for cassava slices of 0.0732g/g.h compared to 0.04908g/g.h for GSD and 0.02074g/g.h for OSD, while drying rate of PVGSD, GSD and OSD for cassava chunks were 0.0457g/g.h, 0.0355g/g.h and 0.01667g/g.h respectively. Similarly, PVGSD was observed to obtain the highest rate of drying for red pepper of 0.097g/g.h compared to GSD of 0.094g/g.h and OSD of 0.047g/g.h. The drying method, time and sample size were found to have statistical significance (p<0.05) on temperature, moisture content, humidity and drying rate. The PVGSD dryer with the workstation was able to keep the humidity conditions in the dryer low to prevent moisture uptake during the night. The samples in this dryer showed consistent drop in moisture content throughout the drying period. To further evaluate the drying effect on the samples, laboratory analyses such as water activity, color profile and particle size determination were performed on the resultant dried cassava flour samples and red pepper (whole). The color profile of red pepper flour and cassava flour from the PVGSD was observed to be better in terms of the redness (a*) for red pepper, lightness (L*) for cassava, browning index (BI) and color change (ΔE). The color change of red pepper for PVGSD, GSD and OSD were 11.94, 16.43 and 25.25 respectively. The color change of cassava slices for PVGSD, GSD and OSD were 10.49, 10.84 and 11.59 respectively, while that for cassava chunks recorded at 10.60, 12.09 and 12.11 respectively for PVGSD, GSD and OSD. The SD50 values for flours from the slices were finer than those from the chunks. The SD50 values for cassava slices under PVGSD, GSD and OSD were 62 μm, 76 μm and 78 μm respectively, while SD50 values for cassava chunks under PVGSD, GSD and OSD recorded at 80 μm, 92 μm and 106 μm respectively. In general, cassava slices under the three drying methods demonstrated a better quality than the chunks in terms of color and particle size. All samples were dried to a water activity below 0.500 to prevent microbial spoilage. Final water activity for red pepper was 0.384, 0.388 and 0.414 respectively for PVGSD, GSD and OSD, while that for cassava slices and chunk for PVGSD was 0.361 and 0.415 respectively, GSD was 0.384 and 0.446 respectively and OSD recorded at 0.421 and 0.490 respectively. To further understand and describe the drying curves for the samples under the three drying methods, 12 Thin Layer Mathematical models were evaluated. The best model under this study for red pepper is the Midilli and Kucuk model under both PVGSD and GSD while Logarithmic model describes best under OSD. Approximation of diffusion model can be used for cassava slices under PVGSD and GSD, while Midilli and Kucuk model for OSD. These models were selected according to the lowest Root Mean Square Error (RMSE) and chi square (χ2) and highest correlation coefficient (R2).