Drying occurs as a result of the vaporization of liquid by supplying heat to wet feedstock. Drying is categorised into: direct (convection), indirect or contact (conduction), radiant (radiation) and dielectric or microwave (radio frequency) drying ( McCabe et al., 1993; Parikh, 2014). Heat and mass transfer are the dominant mechanisms of the drying process, as heat is transmitted to the product in order to evaporate liquid, and mass is transferred as a vapour into the surrounding gas. The factors that affect heat and mass transfer control the drying rate ( Parikh, 2014).

Mass-transfer-dependent dryers are direct dryers and essentially pass a gas through the solids to facilitate the removal of moisture. These dryers are often called adiabatic dryers. Heat-transfer-dependent dryers are indirect dryers and remove moisture by vaporizing the water or liquid using heat. These dryers are therefore non-adiabatic dryers and limited by heat transfer, and depend on the thermal conductivity for their design. Many non-adiabatic designs are based solely on the consideration of heat transfer analysis alone ( McCabe et al., 1993).

Similar research of organic material done in other investigations, suggests that there is strong dependence of thermal conductivity on the moisture content of the material. A study conducted by Nayyeri et al. (2009) on the thermal properties of dairy cattle manure suggests that the effect of moisture content on the thermal conductivity of cattle manure is even greater than the effect of temperature. However, data on the thermal conductivity of human waste are not very common and motivated this study.

For this study, the COD, volatile solids and moisture content were measured to identify their correlation with the thermal conductivity of faeces. COD refers to the amount of oxygen required by a portion of sample such that it is completely oxidised, and is calculated using the Closed Reflux Titrametric Method. COD is thus indicative of the biodegradable material in a sample of material. Moisture makes up approximately 79.2% ( Almeida et al., 1999) of fresh faeces. A proportion of the remaining solids (84.4% - quoted by Lopez Zavala et al., 2002) are made up of volatile solids, solids which ignite upon high temperatures.

Figure 1(a) illustrates a linear model of thermal conductivity and moisture content that was fitted to the experimental data obtained from cattle manure and in Figure 1 (b) four parameters model was fitted to the data and the experimental curves obtained over the 3 temperatures (9, 24 & 39°C). From the study it was observed that the temperature did not have significant effect on the thermal conductivity of separated manure solids over the 30ºC range investigated.

The study was carried out in 2013 whereby individual samples of fresh faeces were collected at the Pollution Research Group Laboratory facilities at the University of KwaZulu-Natal (Durban, South Africa) based on anonymous donations and stored in a cold room at 4°C. The individual stool samples were homogenised together using a blender. The entire mixed sample was transferred into a bucket and closed with a lid. Holes were drilled into the side to allow movement of oxygen to the surface such that aerobic degradation could occur. The sample with the faecal mix was then stored in the laboratory for a one-week period at a room temperature of 24-25°C. The sample was sealed so to hinder accelerated moisture content loss as a result of unlimited air supply.

The C-Therm TCi Thermal Conductivity Analyser was used for this study. It has an accuracy of 5% and a precision of 1%, and is able to measure substances of temperatures between -50°C and 200°C. It can also measure a wide range of thermal conductivities, varying from 0 to 120 W/m.K, see Figure 2 below. Samples of 1.88 mL were placed in the sampling sensor, as indicated by the standard operating procedure. C-Therm TCi 2.3 software was used, and the thermal conductivity was directly measured. The software has the ability to take multiple measurements on a single sample and samples were analysed in triplicate.

Three samples of 10 g each were used to analyse the moisture content of the faeces in triplicate. The samples were placed in an oven (Gallenkamp) at 105°C and dried for 24 hours. The mass evaporating and remaining represent the moisture content and total solids respectively. The samples remaining were then placed in a furnace (Furnace E160) at 550°C for 2 hours, after which the mass was measured. The mass loss on ignition was taken to be the volatile solids, and therefore the difference in mass from before and after being put in the furnace was measured. Moisture content and volatile solids was calculated using the following equations:

MC = m crucible + m faeces − m exit   oven m faeces ( 1 )

Where:

MC = Moisture Content

m _crucible = mass of a crucible

m _faeces = mass of faeces

m _{exit oven} = mass of crucible & feaces after drying in the oven

VS = m exit   oven − m exit furnace m faeces ( 2 )

Where:

VS = Volatile solids

m _faeces = mass of faeces

m _{exit furnace} = mass of faeces & crucible exiting the furnace

m _{exit oven} = mass of faeces & crucible exiting the oven

The COD was calculated using the Standard Closed Reflux, Titrametric Method. A sample with mass of 1 g was dissolved in 1 litre of distilled water. Triplicate samples of 5 mL were measured from the mixed solution. The samples were digested for 2 hours in an acidic dichromate solution in the Ethos One High Performance Microwave Digester. The digested mixture was then titrated with ferrous ammonium sulphate (FAS). The COD was then calculated from the standard equations:

C O D = ( B l a n k − T i t r a t i o n ) × m o l a r i t y o f   F A S × 8000 S a m p l e ( m L ) ( 3 )

C O D ( g O 2 / g   s a m p l e ) = C O D ( m g   O 2 / L ) D i l u t i o n   f a c t o r × 1000 ( 4 )

Where:

FAS = ferrous ammonium sulphate

mg O ₂/L = milli grams of Oxygen per litre

gO ₂/g sample = grams of oxygen per gram of sample

Averages and standard deviations were calculated using the inbuilt functions of Microsoft Excel. Confidence interval testing was done for part 2, for data relating the thermal conductivity to the moisture content.

The confidence interval was done on the slope and intercept of the data.

Y = a + b X ( 7 )

Where Y is represented by the thermal conductivity and X is represented by the moisture content. a and b represent the y-intercept and slope of the best fit line respectively. Values for a and b using a best fit line was 0.041 and 0.561 respectively. The following are the equations used to calculate the confidence intervals.

a ± t ( α 2 , N − 2 ) ⋅ S b ⋅ ∑ ( X 2 ) N S X X ( 8 )

b ± t ( α 2 , N − 2 ) S b S X X ( 9 )

Where the following variables are defined and calculated using table in appendix with X ¯ = 0.42 and Y ¯ = 0.28, and t = 2.228

S X X = ∑ i ( X i − X ¯ ) 2 = 1.15 ( 10 )

S Y Y = ∑ i ( Y i − Y ¯ ) 2 = 0.38 ( 11 )

S b = S Y Y − b 2 S X X N − 2 = 0.049 ( 12 )

∑ ( X 2 ) = 2.82 ( 13 )

The confidence intervals for a and b were thus:

a : 0.043 ± 2.228 × 0.049 2.82 12 × 1.15 = 0.043 ± 0.050

b : 0 . 56 ± 2 . 228 0.049 1.15 = 0.56 ± 0.10

Figure 3 illustrates effect of storage time on the thermal conductivity of the faeces. Thermal conductivity varied slightly over the course of the experiment. The initial average thermal conductivity was 0.46 W/m.K, then increased to 0.48 W/m.K and remained constants as days progressed; however, day 6 attributed to variation of material within the storage bucket, with an average thermal conductivity of 0.33 W/m.K (which is considered as an outliner). It needs to be noted that the rate of degradation and dehydration may not have been uniform throughout the sample as it was observed that the surface material was dryer and harder than the wetter, softer material beneath, resulting in inconsistencies with analytical tests. In a study aimed at determining the thermal properties of composting bulking materials such as saw dust, soil compost blend, beef manure and turkey litter by Ahn et al. (2009), thermal conductivity ranged from 0.12–0.81 W/m°C. Therefore, the thermal conductivity obtained in this study fell within the range reported by Ahn et al. (2009).

Figure 4 illustrates the effect of storage time on the COD of the faeces. The COD did not present any significant trend over the period of investigation, and averaged at a value of 0.82 g O ₂/g dry sample, suggesting that the amount of oxidizable material remained fairly constant. There was no significant relationship to suggest that storage time has an effect on the COD.

Figure 5 illustrates average results of moisture content ( Figure 5a) and volatile solids ( Figure 5b) during the effect of storage time experiment. Over the same period of days, the moisture content decreased by approximately 16%, from the initial moisture content of 77% to 61%. This can be expected due to the concentration gradient existing between the sample and the surrounding air and because of this gradient, mass transfer of water from the sample to the air occurs ( McCabe et al., 1993).

The volatile solids resembled a trend inverse to that of the moisture content, consisting of (on average) 85% of the total solids, compared to 84.4% as quoted by Lopez Zavala et al. (2002). Two factors contributed to the volatile solids, (i) the distribution of particles within the homogenous mixture as it was difficult to completely homogenise a sludge mixture and (ii) the variation of degradation and dehydration of the faeces within the storage bucket. The rate of degradation and dehydration may not have been uniform throughout the sample as it was observed that the surface material was dryer and harder, as compared to the wetter, softer material beneath it.

For better understanding of the relationship between thermal conductivity and moisture content, the thermal conductivity of water (100% moisture content) and dry faeces (0% moisture content) were also measured. A simple model, using a composition-weighted average of the thermal conductivity of dry faeces and water was measured at 0.074 W/m.K and 0.61 W/m.K respectively, was proposed to predict the thermal conductivity of faeces ( Figure 6). A 95% confidence interval test was performed on both the slope and intercept of the proposed model, and it was observed that both were well within range. The slope of the proposed model (0.54) was within the confidence interval of [0.45: 0.66] and the intercept of 0.074 was within the calculated range of [-0.0069 : 0.093]. When plotting the upper and lower confidence lines, it was evident that the model fitted well within the 95% confidence of the linear line of best fit. Thermal conductivity of water measured precisely to a literature value of 0.61W/m.K ( Ramires et al., 1995). Nayyeri et al. (2009) also carried out a study on thermal properties of dairy cattle manure and observed a linear relationship between thermal conductivity and moisture content where by increasing trend in the thermal conductivity of dairy cattle manure was also observed with the increase in moisture content, which was the same trend as observed by Yang et al., 2002.