Low-cost onsite selection of suitable unplanted sand drying bed structure for faecal sludge treatment in less developed countries: A case study of Yaounde in Cameroon Open Access

As the world population is growing very fast, sanitation problems are becoming more and more critical, especially in towns of less developed countries. Worldwide, according to UNICEF-OMS (2015), more than 2.4 billion people do not have access to safe sanitation, and 90% of wastewater is rejected without proper treatment in nature, causing serious environmental and health challenges. This population is anticipated to increase to 5 billion by 2030 (Strande 2014), leading to a critical faecal pollution crisis. In urban centres the majority of houses are served by onsite sanitation systems such as septic tanks and pit toilets. The faecal sludges (FSs) collected from these systems are usually discharged untreated into urban and peri-urban environments, posing great risks to water resources and public health (Ingallinella et al. 2002). In fact, FS management is a growing challenge in urban areas due to rapid urbanization, population growth, and poor FS treatment facilities, especially in Africa (Manga et al. 2016). The cost of constructing and maintaining existing efficient technologies is also prohibitive.

In Cameroon, there is no complete wastewater management chain and some of the existing localised treatment plants which were built around 1970 (in Yaounde and Douala) are obsolete and many of them are no more used. Thus, nature is the main end point of all FS collected from septic tanks and pit latrines all around towns (MINEE 2011). In Yaounde sludge collected from domestic, hostels public toilets, and industries are all discharged in Nomayos, a peri-urban populated neighbourhood in Yaounde. Due to the high environmental and health risks, Cameroon Government with support from the World Bank through the Yaounde Sanitation project phase 2 (PADY II) constructed the first large sludge treatment plant in 2021 at Etoa Ahala in the Yaounde vicinity. This was inaugurated on September 4, 2021. This system includes a primary treatment with two sets of sand drying beds (30 cm thick) loaded with 300 L/m² septage. This plant receives close to 260 m³/day septage. To align with the current loading rate in that plant, three bed structures were tested to assess the system's performance and select the suitable bed structure: (i) 20 cm of gravel and 20 cm of sand; (ii) 20 cm of gravel and 30 cm of sand, and (iii) 30 cm of gravel and 20 cm of sand. Despite many unplanted sand drying beds in the literature, no standard is set (Kouawa 2016) and little focus is placed on the sand filtering thickness. Most of the systems are designed to aim at enhancing nutrient recovery in the resulting dry solids, contaminant load removal in percolate, and shortening the dewatering time (Manga et al. 2016). Evaluating sand drying bed performance requires assessing several factors: dewatering time, contaminant load removal efficiency, solids generation rate, nutrient content, and helminth eggs viability in the dried sludge (Manga et al. 2016).

There are several parameters affecting the duration and efficiency of the dewatering process in sand drying beds. Bassan et al. (2014) indicate climate, properties of the sludge, loading rate, thickness of the sludge layer, and surface area of the drying bed. Besides, other advantages of using drying beds as a treatment technique are the low cost, low energy consumption, low chemical consumption, and low requirements of maintenance (Lindberg & Rost 2018). Indeed, onsite FS treatment plants generally include solid–liquid separation by settling/thickening processes, sludge dewatering and drying lagoons/beds, stabilization ponds, and co-composting with refuse (Strauss et al. 1997). The selection of one of these methods depends on several technical, social, and economic criteria. This article focuses on ten technical criteria and the cost of the testing equipment.

The selection of the suitable drying bed structure was based on 10 criteria providing higher organic, mineral, and bacteriological pollution removal, higher drainage capacity, and mass reduction (Table 1). This study focuses on parameters easy to determine in the context of less developed countries (see Table 1) where access to well-equipped laboratories, remoteness from experimental sites, and trained personnel are challenging. Indeed selection of adequate testing parameters depends on the end use of the results. This study considers only six out of 12 commonly tested parameters for physico-chemical WWTP performance (temperature, hydrogen potential (pH), COD, 5-day biochemical oxygen demand (BOD₅), total suspended solids (SS), phosphate, total nitrogen, total phosphorus, electrical conductivity (EC), nitrate, amonia nitrogen, orthophosphate) and two out of five parameters for microbiological analysis (E. coli, faecal coliform, faecal streptococci, helminth eggs, Ascaris eggs) according to several authors (Kuffour et al. 2009; Olufunke & Koné 2009; Hijosa-Valsero et al. 2012; Seck 2014; Defo et al. 2015; Kouawa 2016; Manga et al. 2016; Gnagne et al. 2019; Tagba 2019). As heavy metals can also be found in sludges (Al-Muzaini 2003), one out of 15 metallic elements which seem to be easily determined were tested: lead.

The study also considers drainage capacity in relation to the bed structure and bed material as it has been previously shown by Seck (2014) and Kuffour et al. (2009) that these have a significant influence on the drying bed drainage capacity.

Criteria thresholds were based on Benin and French treatment performance for suspended matter, turbidity, BOD₅, COD, helminth reduction, and E. coli reduction as indicated by Kerekou et al. (2001) and INERIS (2007). For nitrate and lead, Cameroon norms were used. Concerning other thresholds, they were defined after consultations with national experts in sludge treatment from academia and practitioners.

According to Heinss et al. (1998), filtration and drainage processes in sand drying reduce 50–80% of sludge initial volume and the siccity of pre-digested sludge in septic tanks from cities ranges between 18 and 70%. As siccity is more complex to determine on the field with generally limited access to adequate equipment, the mass reduction was chosen as the key selection criteria with a critical value of 80%.

Fresh experimental FS was found to be low concentration sludge that does not fit Cameroon's norms for safe discharge in the natural environment. Besides, all percolates seem to comply with these norms for almost all standards apart from COD (Table 2). Percolates are of very low dissolved oxygen content which is bad for aquatic life since it would reduce available oxygen for life development. This is in line with conclusions from Olufunke & Koné (2009) who showed that percolates from sand drying are generally of high salinity and low oxygen.

The parasitic constitution of these sludges is of higher concentration compared to sludge tested by Tagba (2019) for helminth but very low for E. coli. Faecal coliform was not analysed in this study but general performances on sand drying bed range from 2 to 4 log reduction according to Tagba (2019) and Gnagne et al. (2019).

Table 2 shows that Cameroon norms cover most of the analysed physico-chemical parameters except dissolved oxygen, TDS, colour, and turbidity. Sludge studied shows characteristics close to those studied by Koné et al. (2016) for DCO, dissolved oxygen, SS, nitrates, and pH value.

It is observed that sand filtration reduces the acidity of percolates with a reduction of the organic content and it captures some acidic elements in the biosolid like lead. Indeed, filtration significantly reduces all parameters except dissolved oxygen which is increased, leading to basic low oxygen moderately saline discharge. As a consequence, it is recommended to always proceed to additional treatment before discharging them into water courses or ponds (Olufunke & Koné 2009).

The drying process contributes to mass and volume reduction in proportions which vary with bed structure. The mass reduction recorded in Table 3 is between 79 and 82% while volume reduction ranges between 60 and 70%. The mass reduction is within the range of 80–82% for common dewatering techniques for digested sludge according to Stefanakis et al. (2014). In fact, regardless of the organic charge, dried sludge siccity is globally higher than 80% while using unplanted WWTP (Tagba 2019).

Volume reduction too is within the normal range indicated by Heinss et al. (1998) who reported that there is a change in volume of 50–80% using sand drying for sludge dewatering through filtration and drainage. According to Olufunke & Koné (2009), it can even rise up to 90%. Comparison between the different bed's structure for this study shows that the G20S30 bed has the greater mass and volume reduction, 82 and 70% respectively.

Drainage-specific flows recorded on experimental drying beds vary from 0.022 to 0.031 L/s m² (Figure 5). These are two to three times greater than 0.009 L/s m² found by Tagba (2019) in sokode (Togo) over 30 cm of gravel (5–10 mm), 40 cm of gravel (20–40 mm), and 10 cm of concrete slab. It is also close to the theoretical drainage capacity suggested by Nielsen (2003) on planted drying beds constituted of 15 cm of sand, 30–45 cm of gravel, and 10 cm of sandy loam soils. These values are better than drainage recorded by Nielsen (2003) for a continuous feeding mode (0.005–0.018 L/s m²), thus reducing the gravel layer with finer porosity reduces drainage capacity, while reducing 10 cm of sand thickness leads to an increase of 41% of drainage capacity. Then the bed G20S20 has the highest drainage capacity while G20S30 has the lowest. As a consequence, reducing drying bed thickness increases drainage capacity as previously found by Manga et al. (2016) and Tchobanoglous et al. (2003).

Considering the fact that Cameroon does not have clear standards on physico-chemical treatment performances for sludge, the study relied on Benin standards. According to Table 3, all drying beds reduce appearance and organic as well as ionic concentration by more than 80%, even though there is an increase in pH (12–16%) and dissolved oxygen (87–148%). Conductivity and lead reduction have lower values with the greatest values recorded for beds having 20 cm of gravel. As far as SS, COD and BOD₅ are concerned, all drying beds comply with Benin standards. Regarding overall performance, drying beds having 20 cm of gravel seems to be more efficient (G20S30 and G20S20).

Sand drying performances comply with Benin and French norms for SS, BOD₅, COD, and turbidity (70–90% removal required). Other performance parameters were not compared due to the absence of a clear performance standard.

The selection is based on 10 criteria listed in Table 1 with scorecards determined based on literature review and expert consultation. Based on these criteria beds, G20S30 and G20S20 comply with seven of the 10 selection criteria (Table 4). These criteria do not consider other important parameters like arsenic, mercury, zinc, total solids, and organic matter which were analysed by Al-Muzaini (2003) for Jahra treatment plant since adequate laboratories are difficult to find and the cost of analysing samples are high.

Despite relatively low performance on helminth and E. coli reduction, sand drying beds are suitable for pre-digested domestic FS treatment. Among tested sand gravel drying beds, according to the scoring G20S30 and G20S20 are better but the most suitable is then G20S20 with a higher specific drainage flow (41% greater than G20S30) and top values for seven out of 10 criteria. It is observed that G20S20 does perform within the threshold but is less performant than G20S30 and G30S20 for E. coli removal and volume reduction. Globally, these selected sand bed structures have comparable performance as 30 cm thick multi-soil layer (MSL30) tested by Latrach et al. (2014) the only difference lies in bacteriological performance which is higher for MSL30.

The performance registered within this experiment gives an idea of the practicability of the method and some indicative results. Further research with long-term testing, diverse wastewater sources, quality, and unloading would help to better appreciate the usefulness of both the approach and sand drying technology. According to Hijosa-Valsero et al. (2012), there is a significant decrease in unplanted wastewater treatment plant performance throughout time for all performance parameters. Thus, their performance, even if tested at the micro or mesoscale, should not rely only on the first month's efficiency.

Despite no replication being made to assess the influence of multiple charges or clogging dynamics these drying beds can be considered as appropriate passive technology for digested FS treatment in less developed countries. Indeed, treatment of low-concentration sludge with unplanted drying beds seems to significantly reduce acidity, conductivity, colour, turbidity, nitrates, BOD₅, and SS in proportions that are in accordance with discharge standards in force in Cameroon (Lako et al. 2021). However, further treatment is required if by-products (dry matter and percolates) are to be re-used for agriculture. Lavrnić & Mancini (2016) also concluded from an experiment in south Europe that in general, constructed wetlands like drying beds have trouble reaching the strictest standards, especially regarding microbiological parameters, but their effluents can still be suitable for reuse in areas that do not require water of the highest quality.

Despite this, constructed drying beds are still popular because of their reliability, ease of use, and low cost (Dharmappa et al. 1997). Latrach et al. (2014) even recommend a multi-soil-layering (MSL) system for decentralized sewage treatment in small communities.

Regarding the financial aspect, which is critical for adoption, a good indicator may be the cost per unit volume of sludge treated or the cost per unit area constructed. In this case, the tested drying bed cost 1 USD/L and 391 USD/m², which are affordable costs for onsite experiments. Indeed Gnagne et al. (2019) proved that results from the mesoscale (1–50 m²) are still coherent with results from the micro-scale (less than 1 m²). This is very interesting since the cost of testing appropriate bed structure and the material will drastically reduce when using micro-scale cheap prototypes.

From these, it is clear that the selection of appropriate low- to high-concentration FS drying bed structure for onsite experiments is possible using low footprint, low-cost prototypes with a simple test of several onsite and laboratory parameters. The evaluation frame proposed in this paper with these 10 criteria helps to decide on the most suitable bed structure based on drainage capacity, organic and ion pollution removal efficiency as well as bacteria removal efficiency for key indicators. Additional criteria could be identified and included in this methodology depending on the intended reuse of percolates or biosolids. For example, Manga et al. (2016) also considered dewatering time, solids generation rate, nutrient content, and helminth eggs’ viability in the dried sludge. Whatever the goal, the focus should be the selection of a bed structure providing at least a 70% score. Since loading rates and sludge quality or origin could influence the result of tests it is recommended to consider loading rates between 200 and 400 L/m² day for such experiments.

This paper aimed at demonstrating a simple decision support method for the selection of a suitable unplanted FS drying bed based on 10 guiding technical criteria and simple field and laboratory tests. From this article, it is indicated that the drying bed structure (material and respective thickness) is very important to consider while designing testing schemes. The criteria proposed in this paper are easily determined from field measurements and experimentation is short (14 days for dewatering and 8 days for laboratory analysis). It seems then really practical with low-cost small units to rapidly test drying bed structures and identify the most suitable for the intended use.

Domestic pre-digested FS from households in Yaounde analysed as a case study to support decision on the drying bed structure helped to identify bed structures of 20 cm of gravel and 20–30 cm of sand as suitable for the reduction of most harmful parameters. So, the best structure is the one having the greatest drainage capacity, best organic and ionic contaminant load removal efficiency, and highest solid matter reduction performance. According to these criteria, G20S20 is preferred. Other criteria could be added to the selection grid depending on the end use of treatment by-products. For future studies, it is advisable to experiment sludge from other origins or concentrations and also conduct replications to analyse the change in performance and the clogging dynamics. Variable climate conditions could also be an interesting influential factor to assess.

The authors wish to acknowledge the authorization of the Ministry of Urban Development and Housing for this research to be carried out as well as the University of Dschang for technical support and Water For Life Cameroon for sponsoring.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.