Figure 1 shows the schematic and constructed views of the 5-tonne capacity SBHD at the site. The SBHD is based on a greenhouse structure design with an overall dimension of 10.7 × 6.5 × 3.3 m. It is partitioned into three internal drying sections i.e., the Right Section (RS), Left Section (LS) and a Middle Section (MS). Each of these stations have four levels of drying shelves/racks. The drying shelves are arranged from bottom to top in order of Level 1 (L1), L2, L3 and L4with corresponding heights (0.6 m, 1.2 m, 1.8m and 2.4 m) from the ground, respectively. The dryer is coupled with a biomass burner enclosed with a crossflow type heat exchanger to raise the temperature of ambient air that is blown into the drying chamber using a 0.374-kW axial-flow fan that draws in drying air through the heat exchangers and passes it through the drying beds. The SBHD was test-run during the minor maize production season under load conditions during the month of January 2017 at the installed location.

To assess the performance of the SBHD, white shelled maize at 22.8% moisture content (wet basis) was obtained from a local farmer and dried in the dryer and in the open sun. Drying occurred using experimental cages (0.3 m × 0.3 m × 0.05 m). Each cage was filled with maize seed samples weighing 1kg and were replicated three times at each level within the dryer. A control set-up of equal maize seeds (1 kg), in triplicate was also dried in the open sun as practiced by most rural farmers in the area. The experimental trays containing the maize samples were taken out every hour and weighed to monitor the moisture reduction in the maize samples. The moisture content at the i^th time during the drying period was calculated using Equation (1) until the required final moisture content was attained.

Moisture content (%); M C i = ( 1 − W o 1 − M o W i ) × 100 (1)

MC_i = moisture content at i^th time (%);

M_o = initial moisture content in decimal;

W_o = initial weight of samples in gram (g);

W_i = weight of samples at i^th time in grams (g).

The drying process started at 09:50 am and continued for a period of seven (7) hours. During this period, TinyTagTGP-4017 data loggers (Gemini Data Loggers, Chichester, UK) were installed near the experimental cages to monitor and register variations in temperature and relative humidity in the drying environment. Five data loggers were positioned, as shown in Figure 2, at each of the four levels and the average used to represent the drying conditions at each level. Drying conditions in the dryer were compared with the ambient conditions.

The germination test was carried out in accordance with the criteria established in the Rules for testing seed [14]. Four sub-samples each of 100-seedsfor each drying level and a similar set-up for grains dried in the open sun were randomly selected and planted in germination trays filled with soil fetched from a river bank. The setup was kept under room temperature conditions. Emergence counts started from the 5^th day after planting and continued on the 6^thand 7^th days. The percentage of normal seedlings was calculated eight days after the test was set up. The percentage germinated seeds were calculated using Equation (2) as reported by [15].

%Germination = number of seeds germinated number of seeds set for germination × 100 % (2)

Maize grains were selected by random sampling from each station per layer of dried grains. For each sample, 100 seeds free from insect attack were counted and analysed for stress cracks. The samples were placed on a light box and checked for single, double, multiple or no cracks. The stress crack index (SCI) for each sample was determined using Equation (3) as proposed by [16].

SCI = ( 1 × singlecrack ) + ( 3 × doublecrack ) + ( 5 × multiplecrack ) + ⋯ (3)

The experiment was conducted in a complete randomized design and data obtained were subjected to analysis of variance (ANOVA) using GenStat statistical

software version 12 at a significance level of 5% (p ≤ 0.05). Results were presented in tables and graphs using Microsoft Excel using the mean values obtained.

The temperature profile recorded during the drying of the maize grains for seed in the SBHD is presented in Figure 3. Drying air temperature at level 4, increased during the first four hours, from 41.1˚C to a maximum of 58.9˚C by noon and reduced to 43.2˚C at the end of the drying process. Over the drying period of 7 hours, a mean temperature of 52.8˚C ± 5.4˚C was recorded at L4. Similar variations in temperature were observed for L3, L2 and L1 in the dryer with mean temperatures of 49.8˚C ± 5.6˚C, 45.4˚C ± 5˚C and 44.4˚C ± 4.6˚C, respectively. Comparably, the overall mean temperature inside the dryer (L1-L4) of 48.1˚C was higher than the ambient temperature by 12˚C. This stratification was attributed to the solar insolation during the time of the experiment as suggested by [17]. Additionally, hot air from the furnace rises due to forced convection with the heat accumulating at the upmost levels of the dryer leading to higher temperatures at L4.

As the drying temperatures increased, the moisture content of the maize grains reduced. Maize grains at 22.8% moisture content were dried to an overall final average moisture content of 12.8% ± 0.2% (wb) within 7 hrs. However, samples at L4 and L3 reached average moisture content of 13% within 5 and 6hrs of drying, respectively, while samples at L1 and L2 reached final moisture content of 12.7% by the 7^th hour. Comparably, samples dried in the sun reduced from 22.8% to 12.9% in 7hrs. The result showed that overall drying rates of 1.3%/hr were achieved in the dryer compared to 1.2%/hr for samples dried in the open sun. The high moisture reduction rate achieved in the dryer could be attributed

to the additional heat from the biomass furnace attached to the dryer which facilitated the faster drying of grains to the required moisture content. Figure 4 shows the variation of moisture content of maize samples dried at different levels in the SBHD as compared to the samples dried in the open sun.

Drying maize using the SBHD did not significantly reduce the viability of the kernels. However, there was a 9% reduction in overall germination potential of grains dried in the SHBD compared to grains dried in the open sun. From the temperature effect, it was observed that, due to the increase in temperature of drying air at the upper level (L4), the percentage of kernel viability decreased compared to the lower levels (Table 1). Germination rate of grains dried at various levels with mean temperatures of 52.8˚C, 49.8˚C, 45.4˚C and 44.4˚C were 44%, 61%, 64%, and 77%, respectively (Table 1). In comparison, the mean temperature and germination rate for maize dried in the open sun was 36˚C and 71%, respectively.

Percent stress-cracked kernels in different categories for maize grains dried at the various levels in the SBHD are presented in Table 2. It was observed that there was much variation among the different stress-crack categories for grains

*Mean values followed by same letter in a column are not significantly different (P > 0.05).

dried at the different levels in the dryer. In all categories, grains dried at level 4 (L4), had higher stress crack values compared to grains dried at L3, L2, L1 and the open sun. Among the stress-crack categories, the percentage multiple stress cracks were the lowest at all the levels followed by the doubles and the singles in that order. To establish variations in drying temperature effect on the quality of the grains, the SCI was determined for grains dried at the various levels. It was revealed that, temperature had an effect on the final quality of the dried maize in terms of the kernels susceptibility to damage or breakage. As shown in Table 2, the highest SCI of 160 was recorded for Level 4 followed by L3, L2, L1 and the open sun with SCI of 70, 24, 22 and 14, respectively. The outlier of the SCI at L4 can be attributed to the higher temperatures recorded at the level and indicates a high susceptibility to damage or breakage of grains at L4.

The results revealed that drying temperatures in the SBHD increased from bottom (L1) across the upper levels (L2, L3 and L4) of the drying shelfs/racks. A mean drying temperature of 44.4˚C was recorded at L1 while a mean temperature of 52.8˚C was recorded at L4. Comparably, the mean open-sun drying temperature of 35.99˚C was recorded during the same period. The high temperatures recorded in the dryer compared to the open-sun temperature could be attributed to the heat build-up in the dryer due to the transparent cover material used to construct the SBHD. As reported by [10], the transparent cover material allows direct solar insolation to occur and can prevent heat from escaping by acting as a heat trap for infrared (thermal) radiation thereby confining the heated air. The increasing trend in temperature from L1 to L4 could also be attributed to the air density differences due the buoyancy effect which caused stratification of hot and cooler air. The closeness of the top drying shelves to the roof where the hot air is accumulated beneath and recirculated in the dryer may also have contributed to the increasing trend. Similar results were obtained by [18] during maize drying in a solar cabinet dryer. The burning of maize cobs in the biomass furnace also contributed to the high temperature trend in the dryer compared to the ambient temperature agreeing with the findings of [13].

As drying temperature increased from the bottom to upper drying shelves (L1 - L4) in the SBHD, grain moisture content decreased. Mean moisture content decrease steadily from 22.8% to 12.8% (wb) within 7 hours of drying. In a similar work by [13], maize cobs dried in a biomass natural convection dryer had its moisture content reduced to 14% between 5 and 6 hours of drying. Similar results were also obtained for the drying of beans as reported by [19], and coroba slices by [20]. The final moisture content of the dried maize grains was close to 13%. This according to [21] is appropriate for seed storage and may prevent mould growth and preserve seed quality in seed production.

The effect of drying conditions on grain quality was demonstrated using a damage index (DI) as reported by [22]. The results as presented in Table 2 showed that there was a correlation between kernel stress crack formation and drying temperature. An increase in drying temperature across the drying shelves resulted in an increase in stress cracks on maize kernels. This is corroborated by [23] who reported of increased damage index as drying temperature increased above 75˚C during drying of a corn variety named SC-10.

In this study, the drying temperature with the lowest stress crack index was 35.99˚C ± 3.2˚C for kernels dried in the open-sun. Comparatively, similar results were observed for kernels dried at the lower levels (L1 - L3) though there was an overall temperature increase in the dryer compared to the open-sun. The reduction in stress crack index at the lower levels could be attributed to lower temperatures and the overcast created by the drying shelf above, L4, which shielded the lower levels from direct solar insolation on the kernels. However, the percentage damage of kernels in the single, double and multiple cracking categories increased significantly for kernels dried at L4 (Table 2) with the highest recorded stress crack index of 160.The direct exposure of the kernels and the maximum grain temperatures reached in the dryer at the top trays (L4) may have contributed to this occurrence. Similar works by [24] and [25] showed similar results where an increase in drying temperature increased moisture gradient, creating internal tensions, cracks, breakages, and fractures in dried maize. Damage and fractures which results in cracks in the structure of agricultural products due the creation of internal stresses has also been reported by [26].

Germination of dried maize kernels was affected by drying temperature in the SBHD. Evaluation of maize germination performance established the negative effects of increase in drying temperature on the physiological quality of maize seeds. An increase in drying temperature from the bottom to the upper drying shelves (L1 - L4) resulted in decreases in kernel germination. Maize kernels dried at the upper tray (L4) of the SBHD where the highest mean temperature (52.8) and highest SCI (160) were recorded had the lowest kernel germination (44%). Similar observation was made by others in evaluation of sweet sorghum seeds in which a reduction in germination was recorded, especially at temperatures above 40˚C [27]. The low germination rate observed at L4 could be attributed to the high drying temperature, which resulted in the formation of stress cracks in the seed coat and microfissures in the cotyledons as suggested by [28] and [29].

Viability of maize kernels dried at the lower levels (L1 - L3) were not significantly affected by the drying air temperature. Kernels dried at L1 - L3 showed a germination range (61% - 77%), compared to kernels dried at the upper tray (44%). However, improvement in the drying process can help achieve the required 90% recommended by the Ghana Certified Seed Producers Association. Overall, lower drying temperatures at the lower levels (L1 - L3) resulted in increased germination by 33%. Similar observations were made by [13], who reported that, drying maize using a biomass-heated convection dryer did not significantly reduce germination potential. Also [19] reported that, grain dried using the natural solar convection dryer technology is also good for seed production.