Increasing population equals increase in agricultural product consumption for which continuous food production is not a viable option. Solar drying, on the other hand is a promising method to preserve agricultural products for longer durations. This thesis focuses on calculating the predictability of the independent factors and a comprehensive risk assessment to improve the performance of Solar Hybrid Kiln. Biochar samples with different moisture content were selected for 3 tests. Principal component analysis and multiple linear regression analysis were conducted on the gathered data using Minitab 18R platform. Risk response plans associated with the kiln were provided through failure mode effects analysis. Results exhibited 3 significant principal components and reliable prediction model limits were obtained for both training and testing datasets. A total of 41 risks were identified and risk response plans were proposed for them. These results can be further used to increase the efficiency of biochar drying processes.

Thesis (MSc (Forest and Wood Science))--University of Stellenbosch, 2011. / ENGLISH ABSTRACT: The Saasveld solar kiln, valued at US$ 10,000 (R 70,000), is a greenhouse-type solar kiln requiring no additional heating. A unique auto-regulated process of natural dehumidification at night controls the humidity, and uses no venting system, nor any human/controller intervention. The natural dehumidification is achieved by circulating the humid kiln air through an air jacket on the periphery of the solar kiln. Diurnal temperature fluctuations allow condensation in the air jacket when air is cooled to below its dew-point temperature. The process is simple as colder night temperatures are a certainty. Almost no research has been done on similar kilns. The purpose of this thesis was to study the operation and performance of a solar drying kiln to dry high density hardwoods utilising natural dehumidification. The kiln load consisted of 20.1m 3 of a high density hardwood, Eucalyptus diversicolor, with air-dry density of 893kg/m 3 . The ZA Dry Q drying quality assessment indicated outstanding moisture distribution with little drying stress, a few surface checks and no internal checks, collapse, short bow or surface discolouration. The results conformed to the ZA Dry Q softwood appearance grade specification, except for end checks – remarkable for the drying of any hardwood. The final moisture content (MC) distribution was extremely tight with MC average of 11.5 per cent (%), standard deviation of 0.97% and the moisture gradient averaged 1.49%, standard deviation of 0.5%. The MC and MG were equally distributed in all three geometric directions throughout the load. The timber dried in 130 days. The kiln operation was analysed by temperature and humidity parameters at five positions inside and outside the kiln while monitoring the MC. The extent of moisture evaporation from the timber and the dehumidification of air were derived from this data. The results showed that the average of the daily kiln temperature, independent from the external temperature, increased from 18.0°C initially to 25.7°C. A maximum temperature of 38.1°C was recorded at the end of drying. Cell collapse did not occur under these low temperatures. Evaporation occurred generally from 07h30 to 16h45. The rate reduced at lower moisture contents. There was no effect of stack width. Natural dehumidification by condensation happened generally from 17h00 to 07h30. The equilibrium moisture content (EMC) decreased automatically from an initial daily average of 18.1% (variation 5%) to 8% (variation 2%) finally. It was similar to a smoothed T3C2–schedule. The mild drying conditions, with EMCs above 7.3%, ensured minimal moisture gradients and drying stresses. This reduced surface and internal checks. Nocturnal equalising and conditioning, comparable to intermittent schedules, resulted from: no evaporation, improved diffusion of heated timber and adsorption of moisture on the wood surface. The kiln protected the timber in extreme climatic conditions. It was found that the main circulating fan could be switched off from 17h00 to 07h30. A 30-minute manual venting at noon was permitted once fibre saturation point had been reached. Beneficiaries to this study can include hardwood processors who need to dry high density hardwoods within a reasonable time to an excellent drying quality. Processors in developing countries or who have little drying expertise could also benefit. / AFRIKAANSE OPSOMMING: Die Saasveld sondroër, met ‘n beraamde waarde van US$ 10,000 (R70,000), is ‘n groentetonneltipe sondroër met geen addisionele verhitting nie. ‘n Unieke auto-reguleerder beheer die humiditeit snags deur natuurlike dehumidifikasie. Dit gebruik geen ventileerder of enige menslike/kontroleerder inmenging nie. Die natuurlike dehumidifikasie word verkry deur die vogtige lug te sirkuleer deur ‘n lugkussing op die omtrek van die tonnel. Dag-en-nag temperatuur skommelings laat kondensasie toe binne-in die lugkussing sodra die lug benede die doupunttemperatuur afkoel. Dit is ‘n eenvoudige proses aangesien kouer nagte ‘n gegewe is. Byna geen navorsing is nog op soortgelyke sondroërs gedoen nie. Die doel van die tesis was om die werking en werkverrigting van ‘n sondroër wat hoë digtheid loofhout met behulp van natuurlike dehumidifikasie droog, te bestudeer. Die houtvrag het bestaan uit 20.1m 3 van hoë digtheid loofhout, Eucalyptus diversicolor, met ‘n gemiddelde lug-droë digtheid van 893 kg/m 3 . Die ZA Dry Q drogingskwaliteit toets het uitstaande resultate getoon. Daar was min drogingspannings, ‘n paar oppervlak krake en geen interne krake, sel-ineenstorting, kort-boog of oppervlak-verkleuring nie. Die resultate bevredig die ZA Dry Q naaldhout voorkomsgraad spesifikasie, behalwe vir end-krake – merkwaardig vir die droging van enige hoë digtheid loofhout. Die finale vogverspreiding was nou-verspreid met ‘n gemiddelde voggehalte van 11.5%, standaard afwyking van 0.97% en ‘n gemiddelde vog-gradiënt van ‘n 1.49%, standaard afwyking van 0.5%. Die voggehalte en vog-gradiënt was eweredig verspreid in al drie geometriese rigtings van die houtvrag. Die hout het in 130 dae gedroog. Die werking van die sondroër is geanaliseer deur temperatuur- en humiditeits-parameters by vyf posisies binne- en buite die droogkamer tegelykertyd met die voggehalte te monitor. Die omvang van vogverdamping uit die hout en dehumidifikasie van die lug is afgelei van hierdie data. Die gemiddelde daaglikse sondroër-temperatuur, wat onafhanklik van die buite temperatuur was, het verhoog van 18.0°C aanvanklik tot 25.7°C. Die maksimum temperatuur van 38.1°C is gemeet aan die end van droging. Sel-ineenstorting het nie voorgekom by hierdie lae temperature nie. Verdamping het algemeen voorgekom vanaf 07h30 tot 16h45, teen ‘n vertragende verdampingstempo by laer voggehaltes. Die stapelwydte-effek was van geen belang nie. Natuurlike dehumidifikasie deur kondensasie het gereeld voorgekom vanaf 17h00 tot 07h30. Die ewewigsvoggehalte (EVG) het outomaties verminder vanaf ‘n aanvanklike daaglikse gemiddelde van 18.1% (variasie van 5%) na 8% (variasie van 2%). Dit was soortgelyk aan ‘n gelykmatige T3C2 –drogingskedule. Die matige drogingskondisies, met EVG bokant 7.3%, het minimale vog-gradiënte en drogingspannings verseker wat oppervlak-krake en interne-krake verhoed het. Nagtelike houtvog egalisasie en kondisionering, vergelykbaar met puls-humiditeit skedules, was die gevolg van: geen verdamping, verbeterde diffusie in verhitte hout en adsorpsie van vog op die hout oppervlak. Die droogkamer het die hout beskerm teen buitengewone klimaatstoestande. Daar is gevind dat die hoof-waaier afgeskakel kan word vanaf 17h00 tot 07h30 en dat ‘n 30-minute nie-outomatiese ventilering smiddae kan plaasvind sodra veselversadigingspunt bereik is. Hierdie studie sal van nut wees vir die loofhout verwerkers wat hoë-digtheid loofhout moet droog binne ‘n redelike tydperk met ‘n uitstekende drogings-kwaliteit, asook houtverwerkers in opkomende lande of met gebrekkige drogingsvaardighede.