The Kumasi metropolis was chosen because it is one of only five (Accra, Tema, Kumasi, Tamale, Takoradi) metropolis in Ghana with a well organised system for FS collection and disposal. This allowed for proper collection and identification of FS source. The sampling of FS was carried out in close collaboration with the Waste Management Department of the Kumasi Metropolitan Assembly (KMA).

Although toilet facilities in Ghana are classified as water closets, Kumasi Ventilated Improved Pit latrines (KVIP), bucket/pan or public toilets ( Agyei et al., 2011; Ghana Statistical Service, 2012), this work classified the FS used under three main categories based on the storage facilities from which they were taken: pit latrine (KVIP and bucket/pan), private septage (water closets) and public septage (public toilets) ( Appiah-Effah et al., 2014).

A total sample size of 90, comprising 30 of each category of FS was taken based on the sample size adopted by Klingel et al. (2002). Samples for analysis were taken from vacuum trucks at the waste disposal site of the KMA’s Oti sanitary landfill site, Dompoase in the Ashanti Region, at the point of discharge for 10 continuous weeks. Sampling was carried out from the 16 ^th of November 2011 – 18 ^th of January 2012.

The 43 communities from whom samples were collected are detailed in Dataset 1– Dataset 3.

The method of sampling was also based on previous recommendations ( Klingel et al., 2002). Three-point sampling was employed, which involved taking samples at specific times during the discharge of the FS: (a) immediately after the discharge commenced, (b) half way through the discharge, and (c) just when the tank was almost empty. 15L of sample (a full bucket) was taken from each truck, these samples were homogenized by mixing together thoroughly and a 1L sample taken from this was used as the final sample. Samples were analysed on the same day of collection, or alternately stored at -20°C and analysed within a week of collection.

Analyses were conducted according to standardized and well documented methods and protocols. Total chemical oxygen demand (COD), ( APHA— AWWA-WEF, 1998) pH (US EPA Method 150.1), total solids/moisture content (NREL/TP-510-42621 – method), and nitrogen (US EPA Method 351.1), phosphorous (US EPA Method 365.3) and lipid content by extraction using petroleum ether and 20g of dry FS ( Horwitz, 1980) were determined using analytical grade chemicals. The pH was taken on site with a KEDIDA CT 6023 digital pH probe, while the other parameters were analysed in the lab.

Means, standard deviations and T- Test were conducted for all the results, using Microsoft® Excel 2010. Results are presented as the mean ± error.

TS is a measure of the residue remaining after a waste water sample has been evaporated and dried at a specified temperature, mostly from 103 to 105°C ( Metcalf, 2003). In the present study, the TS content in all three categories of FS was found to be very low, and this invariably meant the moisture content was high in all three types of FS ( Table 1). The highest TS content was obtained in the pit latrine with the lowest found in private septage Private septage has less solids and more moisture because most of this FS type emanates from the water closet, where a substantial amount of water is used to flush the faecal matter ( Agyei et al., 2011; Cofie et al., 2006; Kuffour et al., 2013). Even though traditionally, pit latrines do not use water closet facilities, during empting of the latrines, water is pumped into the pit to allow for easy suction. This introduces large quantities of water into the FS. In most cases, FS from sources other than pit latrines are also added during collection, thus diluting the original content ( Doku, 2002; Strande & Brdjanovic, 2014). The various types of FS were all found to be statistically different (pit vs private, p = 0.00; pit vs public, p = 0.00; private vs public, p = 0.01). These results are similar to those obtained by other researchers ( Doku, 2002; Heinss et al., 1998; Strande & Brdjanovic, 2014; Strauss et al., 2000).

Generally, the pH of the FS from all three sources was found to be near-neutral to slightly alkaline and were similar ( Table 2). Torondel (2010) in a review of the literature on sanitation waste characteristics identified the pH of FS to be basic, ranging from a pH of 7.1 -9. Kuffour et al. (2009) also working on FS in Kumasi observed a pH of 7.77 ± 0.13, which are both similar to the results obtained in this study. In contrast, Appiah-Effah et al. (2014), working on FS in Kumasi, obtained a pH of 6.7 in peri-urban areas, but 7.3 in rural areas.

COD is a measure of the degree of reduction (or electron content) of the organic material in wastewater ( Metcalf, 2003). In this study, the COD of the FS was found to be generally very high. Pit latrine FS was found to have the highest COD, which was almost twice as high ( p = 0.02) as the level in public septage FS, whose COD level was also almost three times higher ( p = 0.00) than that of private septage FS ( Table 3). Strauss et al. (2000) found that the COD of septage in Accra, Ghana, to be about 7,800mg/L and that of public toilet FS to be about 49,000mg/ L. Kuffour et al. (2009) also observed a highly variable COD in FS sampled in Kumasi (50320 ± 28780mg/L), akin to that observed in this study. The very high error margins indicate the high variability of COD in the various samples regardless of the source.

The nitrogen content in the FS collected ranged from 649.40 ± 484.52mg/L in private septage to 4479.03 ± 2323.77mg/L in pit latrine ( Table 4). The various sources of FS were found to be statistically different (pit vs private, p = 0.00; pit vs public, p = 0.00; private vs public, p = 0.01). Kuffour et al. (2009) obtained 3580mg/L of nitrogen in public toilet sludge from Kumasi. Strauss et al. (2000) working on FS in Bangkok obtained 830 mg/L of nitrogen in septage, while Strande & Brdjanovic (2014) reported that the typical nitrogen content in public septage is around 3,400mg/L, which are all similar to that obtained in this research. Nitrogen in FS is an important resource because it can be used as fertilizer for plant growth ( Chandran, 2014; Strande & Brdjanovic, 2014).

Chandran (2014) suggests that global phosphorus deposits are expected to be depleted rapidly, unless replenished or recovered; thus making the recovery of phosphorous from FS an interesting prospect. For example, recovered phosphorous can typically be used as fertilizer supplements ( de-Bashan & Bashan, 2007; de-Bashan & Bashan, 2004; Gaterell et al., 2000; Kuffour et al., 2013). Phosphorous content in this study ranged from 137.92 ± 139.0 in private septage, which is statistically similar ( p = 0.05) to that in public septage (228.68 ± 185.26). Both are statistically lower (pit vs private, p = 0.00; pit vs public, p = 0.00) than that of pit latrine (521.07 ± 207.58) ( Table 5). This might be so because according to Kuffour et al. (2013) phosphorous is equally distributed between urine and faeces, which implies that with a lot more of water used to flush faeces into soakaway pits for the private and public septage some of the phosphorous could be leached into the ground, while the pit latrine which has less water would retain more of its phosphorous.

Lipids are not a typical parameter characterised in FS. However, researchers looking at biodiesel production or lipids as an end use product of FS and sewage sludge have found the need to determine the lipid content ( Angerbauer et al., 2008; Kargbo, 2010; Réveillé et al., 2003). Though the percentage of lipids extracted were all statistically similar (pit vs private, p = 0.51; pit vs public, p = 0.86; private vs public, p = 0.67), it must be noted that the lipids were extracted from the dried FS and the values reported are therefore normalized to a dry solids basis. Therefore, taking into account the different solids content of the different FS sources, 1000 g of FS will consequently yield 4.38 g lipids in the case of pit latrines, which is statistically higher ( p = 0.02) than that of public septage (1.78 g), and further higher ( p = 0.00) than private septage (0.77 g) ( Table 6). Chaggu (2004) reported that there is approximately between 4–6g of lipids in fresh human faeces. However, since the method of extraction or the initial mass of FS used is not stated, it is difficult to compare the yield of lipids obtained with that obtained in this study. Typically, lipid content may degrade with the age of the faecal matter into less complex molecules ( Torondel, 2010), and as the age of FS in Ghana can span between days in the case of public septage to a couple of years in the case of pit and private septage ( Agyei et al., 2011; Kuffour et al., 2009) obtaining a lower lipid content was therefore not surprising.