International Journal of Energy and Environmental Engineering (2021) 12:175–190 https://doi.org/10.1007/s40095-020-00364-y ORIGINAL RESEARCH Optimal bed thickness and effective size for improving wastewater quality for irrigation Godwin King‑Nyamador1 · Peace Korshiwor Amoatey1  · Samuel Amoah1 · Benny Adu‑Ampong2 Received: 6 August 2020 / Accepted: 24 October 2020 / Published online: 19 November 2020 © Islamic Azad University 2020 Abstract With the increased use of wastewater for irrigation, there is the need to reduce the contaminant levels in wastewater. The slow sand filtration (SSF) is one such method that can be used to improve wastewater quality. However, the treatment quality depends among other factors on the depth of sand bed and the effective size. Acquiring sand of a particular effective size is becoming increasing difficulty and, therefore, this study sought to investigate over a specified area, the optimal depth and effective size that will be able to get rid of contaminants in wastewater. In separate experiments, three depths (30 cm, 40 cm and 50 cm) and two effective sizes (0.27 mm and 0.45 mm) were set up to investigate their effectiveness in removing Faecal coliform, E. coli and heavy metals (Pb, Cu and Fe) for wastewater from a peri-urban drain used for irrigating vegetables. Results showed that a minimum sand bed thickness of 40 cm and an effective size of up to 0.45 mm reduced the contami- nants tested significantly, wastewater from the drain can be treated. It must be mentioned that the finer sand (0.27 mm) had a slightly better removal efficiency. This implies that the extra cost of acquiring sand of relatively smaller effective size and a higher bed depth with the aim of improving wastewater quality can be saved. Further investigations are being carried out on the combined effects of the optimal sand bed depth and effective size. Keywords Bed depth · Effective size · Kawukudi · Slow sand filter · Wastewater irrigation Introduction hospitals), industrial effluents (where present) and stormwa- ter from runoff. The composition of wastewater is dependent Water, the most common liquid on earth covering about on the uses to which the water was put and may contain a three-quarters of the earth’s surface is essential for the sur- variety of pathogens and other contaminants such as bacte- vival of all living things. Despite its valuableness to life, it ria, viruses, protozoa, helminth’s eggs, heavy metals, organic is increasingly becoming a scarce resource in many arid and and inorganic chemicals [4–7]. semi-arid countries [1]. It is used by humans in many ways The rapid growth in the world’s population, industriali- such as drinking, domestic use, industrial use and agricul- sation, urbanisation and changing lifestyle are some major tural irrigation. Agriculture is the largest consumer of the factors that result in increased volume of waste generation, earth’s fresh water of about 42% [2]. pressure on limited water resources and increased waste- Wastewater is any water used either domestically or water production [8–10]. As competition for freshwater industrially and contains waste materials or substances. It increases in urban and semi-urban centres, irrigated agri- was defined by Raschid-Sally and Jayakody [3] as a com- culture tends to suffer significantly as irrigation has been bination of domestic effluents (consisting of black and grey the largest consumer of water [2, 11]. As a result, many water), effluents from commercial establishments (like countries (mostly in arid and semi-arid regions) have turned to alternative water sources (with the potential of support- Peace Korshiwor Amoatey ing agricultural use) to supplement the limited freshwater * pkamoatey@ug.edu.gh; peacekorsh@yahoo.co.uk sources and, to raise both agricultural productivity and life standards of the rural poor [12]. In Pakistan, it was reported 1 Department of Agricultural Engineering, University that 80% of farmers use untreated wastewater for irrigation of Ghana, P. O. Box LG 77, Legon, Accra, Ghana [1]. 2 Department of Agricultural Engineering, KNUST, Kumasi, Ghana Vol.:(012 3456789) 1 76 International Journal of Energy and Environmental Engineering (2021) 12:175–190 The story is not too different in Ghana as small farms In the design of the SSF, certain design parameters/char- in urban and peri-urban areas of Accra use wastewater acteristics (such as depth of sand bed, effective size, uni- from drains to irrigate crops. It was reported that 70% of formity coefficient, hydraulic retention time and hydraulic farmers in Kumasi use polluted streams for irrigation [13]. loading rate of the filter media) are crucial for the effective Qadir et al. [14] reported that the increasing use of waste- performance of the filter. Barrett et al. [27] wrote on the water for irrigation is basically as a result of the scarcity design of the SSF and reported that the major component of water, increase in the cost of fertilisers and the fear of of the filter is the sand bed through which the water flows. food shortage crisis hitting the world. According to the Work has been done by a few researchers on filter design World Health Organization, more than 10% of the world’s characteristics of SSF for improving wastewater quality population consumes wastewater-irrigated food products for irrigation yielded good results and are summarised in [15]. This practice has some benefits and several other Table 1 [23, 29–32]. implications as far as soil properties and human health is From Table 1, it can be seen that SSFs in both experimen- concerned [13, 16]. tal and field work have been designed with a depth of up to Vegetables irrigated with wastewater have been found 1.2 m and effective size of up to 0.45. It is probable that bed to contain residues as Lead, Copper, Mercury, Iron, Zinc, depth significantly improves treatment quality; however, it is Nickel, Cobalt and Arsenic which causes carcinogenesis, quite expensive to procure even for experimental work and cell damage and loss of cellular functions [1, 13, 17–19]. even more challenging for a particular effective size. Since Therefore, low technology, relatively low-cost treatment, these parameters together influence retention time, it is pru- with ease of operation and minimal maintenance options dent to optimise bed depth and effective size over a specified are being considered to improve the quality of wastewater. area that will effectively treat wastewater for irrigation [30]. One such method of wastewater treatment is the slow The reported finding agrees to the set objectives of the sand filter method (SSF) [16]. The SSF is a simple waste- work in the sense that the results reveal the suitability and water treatment technology that can be used to reduce the effectiveness of the slow sand bed in the removal of sus- pollutant load of wastewater. It has been in use for centu- pended particles in wastewater as well as reducing pH, ries for the treatment of water for domestic use [20, 21]. electrical conductivity and biological contaminant, E coli. Slow sand filtration involves allowing water to slowly pass [23, 32, 37]. However, the above-mentioned studies did not through a bed of sand or other porous material for treatment. emphasise the role of uniformity coefficient, hydraulic reten- It involves the use of both physical and biological pathogen tion time and hydraulic loading rate which are influenced control mechanisms [20, 22, 23]. by effective sizes and the depth of the sand bed to improve It has over the years been considered as probably the sim- the wastewater quality to be used for irrigation purposes. plest, effective and low-cost water treatment process used In addition, most studies did not consider the removal of in many developing countries [24]. It requires few technical heavy metals. components for construction, natural materials for the fil- Our work situates in this context and seeks to optimise trate and has no chemicals added. It is being used in waste- over a specified area the depth of sand bed that would sig- water treatment in many developing countries for irrigation nificantly improve wastewater quality for irrigation. It also purposes mainly because of its efficient performance in the seeks to investigate over a specified area and depth, the removal of pathogens and chemicals in wastewater when effect of varying effective size (d10) in improving the phys- used [25]. ico-chemical as well as biological quality of wastewater for The SSF comprises of the supernatant water (provides the irrigation. hydraulic pressure that pushes water through the sand filter below), biofilm layer (called Schmutzdecke, serves as the Depth of filter media main source of biological pathogen control in the sand filter [26]), bed of graded sand (which provides a medium for Though Barrett et al. [27] reported that the contaminant physical filtration to occur and microorganisms to grow) on removal efficiency of the sand bed depends more upon the layers of gravels (allows for free drainage of water from the maturity of the “Schmutzdecke” than upon its depth, other sand bed to the outlet and also prevents sand from clogging researchers have affirmed that the depth of the sand media the outlet pipes or sand from leaving the filter tank [27]. The through which water passes is crucial to the water treatment filter media is enclosed with openings at both ends allow- efficiency [23, 24, 36, 38]. The biological activity of the ing water to flow in and out (under gravity). The filtration “Schmutzdecke” is enhanced as the bed thickness increases. process—a form of natural, biological water treatment—is The microorganisms and other suspended particles in the used in the removal of solids, precipitates, turbidity (mud- wastewater travel through the depth of the sand media to diness) and in some cases, bacterial particles that produce ensure a higher removal efficiency of contaminants at higher bad taste and odour [28]. sand depths. 1 3 International Journal of Energy and Environmental Engineering (2021) 12:175–190 177 Table 1 SSF design parameters. Modified after Pyper and Logsdon [33] Design Bed depth Effective Filtration Supported Supernatant Retention Scale of Parameters Removal criteria (m) size (mm) rate ( m3/h/ bed (m) water (m) time work tested efficiency m2) (%) Ten States 0.8 0.3–0.45 0.08–0.24 0.4–0.6 0.9 – Field – – Standards USA [61] Huisman and 1.2 0.15–0.35 0.1–0.4 Not reported 1–1.5 – Experimen- – – Wood [30] tal Visscher 0.9 0.15–0.30 0.1–0.2 0.3–0.5 1 – Field et al. [62] Bagundol 0.3–0.9 0.16–0.30 0.2–0.4 0.1–0.3 Not reported Not reported Experimen- E. coli 95.4 et al. [23] tal Turbidity 99.6 Thomas and – 0.3–3.0 1.38 × 10–3 Not reported Not reported 12 m 21 s Experimen- Turbidity 86 Kani [32] tal pH 13.3 Electrical 77 conductiv- ity Muhammad 0.73 0.20 0.1 0.06 1.4 Not reported Experimen- Faecal coli- 99.6 et al. [24] tal form 0.35 Turbidity 96.5 Colour 95.1 0.40 0.45 Faecal coli- 98.4 form Turbidity 87.5 colour 72 Troyan and 1.07 0.25–0.35 0.2 – – – – Viruses 99.8 Hansen Coliform 99 [34] Turbidity 40 Poynter and 0.60 0.2 Polio virus 1 99.99 Slade [35] Van Dijk and 0.6–1.4 0.15–0.35 0.1–0.2 0.4–0.7 1–1.5 – Field Pathogenic 99–99.9 Oomen bacterial [36] Muhammad et al. [24] confirmed that most of the bac- Effective size (D10) teriological purification of wastewater occurs within the top 400 mm of sand bed. Increasing sand bed depth causes Effective size (D10) is the “diameter in the particle-size dis- increases in the total surface area of the sand grains and tribution curve corresponding to 10% finer” [40]. In SSF, ultimately the total adsorption capacity throughout the the smaller effective sizes of the filter media produce good depth of the sand bed in the treatment of wastewater by quality effluent [30, 31]. This is confirmed by Van Der Hoek the SSF. This enhances mechanical entrapment of smaller et al. [41] who explored the performance of two grain sizes particles, including viruses, colloidal matter, pathogenic with effective sizes 0.19 mm and 0.25 mm, and Rolland et al. bacteria, E. coli, turbidity and colour removal efficien- [42] also reported that the biological removal potential of cies. This happens when the bacteria become entrapped in sand media with effective sizes of 0.33 mm and 0.8 mm the spaces between the sand grains or become attached to produced a good effluent quality. each other or the bacteria may die due to scarcity of food Results from experiment conducted by Muhammad et al. and oxygen depletion [23, 34–36, 39]. [24] to measure the significance of increasing effective sizes from 0.20 to 0.35 mm to 0.45 mm on the performance 1 3 1 78 International Journal of Energy and Environmental Engineering (2021) 12:175–190 of slow sand filter showed good effluent quality with the The filter sand needs to be fine enough to provide sufficient smaller effective size giving slightly better results. Langen- filtration of suspended materials of concern. Long HRT bach [43] agreed with the conclusion of Muhammad et al. ensures significant contact of organic materials, biofilm for- [24] after his experiments showed no significant effluent mation increase the degree of water treatment. The hydraulic quality for effective sizes in the range of 0.25–0.80 mm. It retention time of most SSF systems ranges from 5 to 6 h is clear from the studies that media with smaller effective [49]. size are finer and can better trap particles in the wastewater Studies have shown that increasing the bed thickness thereby, resulting in better effluent quality. increases the hydraulic retention time and the infiltration It must be reiterated that effective size influences uni- rate resulting in higher degree of the treated water. Missimer formity coefficient, hydraulic retention time as well as infil- et al. [49] listed the hydraulic retention time from a set of tration rate. These are further discussed below. slow sand filters with differing bed thicknesses and infiltra- tion rate as seen in Table 2. Uniformity coefficient (C ) From Table 2, it is clear that bed thickness and infiltra-u tion time are major factors that affect the quality of water The C tells the homogeneity of the grain size distribution. treated. Thus, increased hydraulic retention time tends to u It is calculated by finding the ratio of 60% finer and 10% increase the degree of water treated. From the discussions, finer [40]. The uniformity coefficient plays a vital role in so far considering bed thickness and effective size, we can the removal of pollutants and prevents clogging. Huisman see that there is cost involved in attaining better effluent and Wood [30] recommended that the sand media should be quality results with respect to the design of SSF that is why fairly uniform with C between 1 and 3. Studies conducted this study seeks to optimise bed thickness as well as the u on the influence of uniformity coefficient (C ) on effluent effective size that can still clean pathogens as well heavy u quality considering the removal efficiency of colour, turbid- metals from wastewater. ity, heavy metal removal (total iron and manganese), faecal coliform and heterotrophic bacteria colonies confirmed that uniformity coefficient (Cu) higher than the World Health Materials and methods Organization (WHO)’s recommendation would lead to a lower filter run time [43, 44]. This research was carried out in two stages: design and construction of the slow sand bed filter which involved the selection of sand, gravels and stones, sieve analysis, and Hydraulic retention time (HRT)/residence time the design and set up of the SSF. The water quality analy- sis involved obtaining the water sample and testing for the The hydraulic retention time (HRT) is also referred to as the microbiological and heavy metal parameters. The study site hydraulic residence time and describes how long it takes and the statistical method used in analysing the results are raw water to travel through sand media during filtration, and further discussed. it is defined as the measure of how much water is moving through the filter over a certain amount of time. The reten- Study site description tion time is weighted by the volume of water stored within the filter [45]. In other words, it is the theoretical time the The study site, Kawukudi, is a developing area in the Accra wastewater will take to journey through the entire depth of Metropolis located in the Greater Accra Region of Ghana. It the filter media. is located south of Ghana to the Gulf of Guinea (5° 35′ 43″ Mathematically, V Table 2 Hydraulic retention time as a function of infiltration rate and HRT = , (1) Q bed thickness [49] Infiltration rate Infiltration rate Bed thickness Hydraulic (m/h) (m/d) (m) retention time where (h) V is volume of filter tank ( m3) Q is the flow rate of influent ( m3/s). 0.05 1.2 0.9 18 An important aspect of the SSF is the quality of the fil- 0.1 2.4 1.0 10 trate and the hydraulic retention time (HRT) of the water 0.2 4.8 1.25 6.25 within the filter media. The type of sand (coarse or fine), 0.3 7.2 1.3 4.3 how long filter is use bed depth and effective size are impor- 0.4 9.6 1.4 3.5 tant factors that directly affects the retention time [46–48]. 0.5 12.0 1.5 3.0 1 3 International Journal of Energy and Environmental Engineering (2021) 12:175–190 179 North, 0° 11′ 21″ East). The site, a peri-urban environment Sand, gravels and boulders was chosen due to the presence of small-scale vegetable farms. The primary and main source of water used in the The choice of sand, gravels and boulders affects the removal irrigation of these farms is the wastewater from the drain performance of pollutants and purification efficiency of the that passes the area. The wastewater in this drain comes from filter [50]. The coarse sand or gravel layers which support domestic and industrial sources. the sand media layer must be of adequate grain size to pre- vent migration of the sand through them. Larger grain sizes allow for faster movement of wastewater through the sand Materials and experimental setups media and the more wastewater can be filtered [46]. For an SSF, an adequate grain size of the filtration media The research was conducted in two experimental setups and should be between 0.15 and 0.35 mm [30]. However, studies the aforementioned two stages were undertaken for each by Muhammad et al. [24] and Anggraini [51] revealed that setup. The experimental setups were done at the premises sand with effective size up to 0.45 mm can produce good of the Agricultural Engineering Workshop of the Univer- effluent quality. Thus, the sand media was sorted and meas- sity of Ghana. The first setup consists of a 250-L drum, dif- ured using a stack of mechanical sieves to obtain effective fuser bowl, Polyvinyl Chloride (PVC) pipe outlet, tap, fine sizes in the range of 0.15–0.45 mm. and coarse aggregates (with boulders ranging between 13.4 and 18.9 mm and gravels, 2.37–5.48 mm in diameter). The Grain size distribution of sand dimensions of the drum were 915-mm tall, 540 mm in diam- eter and can hold a total volume of 0.21 m3 of water. This test was performed to determine the distribution (per- The second setup was made up of three 1000-L tanks centage of different grain sizes) of large sized particles and of equal area, Polyvinyl Chloride (PVC) pipe outlet, tap, fine particles of the sand. The distribution of the different fine and coarse aggregates. The dimensions of the tank were grain sizes affects the rate of infiltration of the influent in the 109 mm by 94 mm in area with a height of 97 mm and can sand media as well as the hydraulic loading rates of the sys- hold a total volume of 1.02 m3 of water. The grain aggre- tem. The method used is the ASTM D 422—Standard Test gates ranged from boulders of 60 mm in diameter (at the Method for Particle-Size Analysis of Soils. The equipment base) to fine sand of different effective sizes. The coarse used were a mass balance, set of clean sieves, mechanical aggregates (consisting of the boulders and gravels) were sieve shaker and a cleaning brush. obtained from a construction site. The boulders form the underdrain of the SSF with the gravels on top of it. The fine Investigating filter bed thickness aggregate (sand), obtained from the beach was poured on the gravels to form the fine media, a very important part of This setup (Fig. 1) was designed using the 250-L plastic the SSF setup. drum. A hole was created at the base of the drum using a Fig. 1 The slow sand filter arrangement 1 3 1 80 International Journal of Energy and Environmental Engineering (2021) 12:175–190 hand drill, and the edges filed to the required diameter. A The washed gravels were used as the support of the sand white thread socket was fitted into the hole created at the media. Stones with sizes between 13.4 and 18.9 mm (serv- base. The socket was connected to a T-Pipe which also con- ing as the boulders) were placed at the bottom of the drum nects to 2 other T-Pipes and elbows at the base of the water to a depth of 15 cm. Right on top of it was placed another drum as shown in (Fig. 2). Care was taken to allow for easy batch of gravels with sizes between 2.37 and 5.48 mm to a removal of the water from the drum and for easy regulation depth of 15 cm. This gravel support holds the sand media of the filtrate flow rate. The diffuser bowl, wide enough to in place to prevent loss of the sand grains and choking of cover the top surface of the drum was placed on the opened the outlet pipes. top of the drum to serve as inlet control. Fig. 2 Arrangement of pipes in the tank Fig. 3 Arrangement of boulders, gravel and sand in the filter 1 3 International Journal of Energy and Environmental Engineering (2021) 12:175–190 181 The washed sand was then poured into the drum to a filtrate was collected to be tested for the parameters being varied depth of 30 cm, 40 cm and 50 cm on top of the gravel investigated. support (Fig. 3). The filter operated by pouring the wastewa- ter into the drum (with material set up) through the diffuser Investigating effective size bowl for the three different depths and flow-through rates per square metre (m2) of the filter media. This setup, three tanks, the first setup consists of fine sand A diffuser was designed to ensure even distribution of with d10 value of 0.45 mm while the second setup has sand the influent (wastewater) onto the filter bed. It also helps to with d 10 of 0.27 mm (Fig. 5). The boulders were first washed trap the large solid particles in the wastewater (Fig. 4). For and placed into the tanks A and B followed by the gravels, each of the depths investigated, a minimum of 21 days was to serve as the sand support. Washed sand formed the upper allowed for the Schmutzdecke to develop and mature, and for layers of both setups. The third tank was an empty tank with the water particles to travel through the sand bed before the no boulders, gravels and sand aggregates. Fig. 4 SSF wastewater filtration process Fig. 5 Slow sand filter arrangement for investigating effective size ( D10) 1 3 182 International Journal of Energy and Environmental Engineering (2021) 12:175–190 All three tanks had PVC pipes fixed to convey the water Investigating optimal depth from the underdrain into the outlet. The perforated pipes also help to reduce the loss of sand from the system. A stopper There was one independent variable (depth of sand media) was provided at the base to allow for easy removal of the with three levels (30 cm, 40 cm and 50 cm) and one depend- water from the tank (Fig. 6). The stopper also regulated the ent variable (% Removal) at three levels of filtration rates flowrate of effluent water. (215 L/h m2, 399 L/h m2 and 664 L/h m2) for each depth of sand media. The following hypothesis was formulated: Wastewater sampling and test H0: There are no statistically significant differ- The wastewater samples were collected in two sterilised ences in the percentage removal of pollutants at the 500-mL test bottles from a drain at Kawukudi. H NO three different depths of sand bed. Mathematically, 3 was added to each sample and immediately, the samples H0 = U30 = U40 = U50. were placed in a cooler with iced blocks to maintain the H1: There are statistically significant differences temperature at 4 °C before transporting to the Ecology between the percentage removal of pollutants at the Laboratory of the University of Ghana for the testing of three different depths of sand bed. Mathematically, the parameters. The wastewater samples were tested for H1 ≠ U30 ≠ U40 ≠ U50, microbiological and heavy metal parameters for both influ- ent and effluent. The following parameters were tested for microbiological parameters (E. coli and Total coliform) where Ux represents the mean percentage removals at the and heavy metals (Pb, Cu and Fe). In the detection of the selected depths (30 cm, 40 cm and 50 cm). microbiological parameters in the wastewater sample, the Membrane Filter (MF) (CFU/100 mL)—Standard methods Investigating effective size for determining heavy metals—Perkin Elmer PIN Accle 900 T Graphite Atomic Absorption Spectrophotometer There was one independent variable (effective size) and (AAS) was used. one dependent variable (% Removal). There were two d 10 values and two tests were carried for each d10 (one before Statistical method and analysis the Schmutzdecke was formed and the second test after the Schmutzdecke was formed). The following hypothesis was The statistical method used in analysing the results of this formulated: experimental study was one-way ANOVA. The level of sig- H0: There are no statistically significant differences in nificance (α) was set at 0.05. The analysis was done using the percentage removal of contaminants for d10 values Microsoft Excel. in the experiment. Mathematically, H0 = UA = UB = UC. Fig. 6 The wastewater filtration setup at the Agricultural Engineering Workshop 1 3 International Journal of Energy and Environmental Engineering (2021) 12:175–190 183 H1: There are statistically significant differences Biological analysis of wastewater samples at 40‑cm sand between the percentage removal of contaminants for bed thickness d10 values in the experiment. Mathematically, H1: at least one of the means is different. The influent coliform counts per 100  mL recorded also exceeded the WHO [52] permissible limit where U represents the mean removal of contaminants for of ≤ 1.0 × 10 3 cfu/100 mL for irrigation. After filtration at y tanks A, B and C. 40-cm bed thickness, the results show significant reduction To ensure a 95% confidence, an α value of 0.05 is chosen. in the coliform counts in the wastewater with the system Given that two effective sizes were tested at three different recording as high as 100% removal efficiency for both Total depths, the degrees of freedom (df) becomes (2,3). There- coliform and E. coli. It was observed that at filtration rates fore, the decision criteria given statistically as f critical of of 215 L/h m 2, 399 L/h m2 and 664 L/h m2, there were sig- 9.55 were obtained from statistical table. With an f value nificantly high percentage removals of 99.9–100% for Total computed from data of df (2,3) and at an α value 0.05. This coliform and E. coli. These percentage removals were sig- means that the null hypothesis is rejected if f obtained from nificantly high compared to the 98.4% reported by Muham- computation exceeds f . mad et al. [24] at the 0.4-m depth of sand bed. This is in crit conformity with the coliform percentage removal of the sand bed depth at 40 cm as reported by Bagundol et al. [23]. Results and discussions Biological analysis of wastewater samples at 50‑cm sand Grain size distribution for optimal bed thickness bed thickness and effective size (D10) Again, the influent coliform counts per 100 mL exceed the The results of the grain size distribution of the sand filter WHO [52] permissible limit of ≤ 1.0 × 103 for irrigation. media for optimal bed thickness were carried out to deter- After filtration, at rates of 399 L/h m2 and 664 L/h m2 the mine, the effective size D10 was found to be 0.26 mm and effluent recorded significant reduction in the both Total coli- D60 was found to be 0.67 mm. This gave a uniformity coeffi- form and E. coli counts in the wastewater with the system cient, Cu computed to be 2.58 which meets the requirements recording between 98.8 and 100% removal efficiency. At the for SSF as recommended by WHO. various filtration rates, there were significant removal effi- ciencies recorded for both parameters. These results agree Water quality analysis for optimal bed thickness with the reports that coliform removal efficiencies increase with increasing depth of the sand bed [23, 32]. Pollutant concentrations were measured for three filtration rates and the results are briefly presented relative to bed thickness. Heavy metal analysis of wastewater samples at 30‑cm sand bed thickness Biological analysis of wastewater samples at 30‑cm sand bed thickness Trace metal concentrations found in the influent were lower than the WHO [52] permissible limits of 5 mg/L (Pb and Fe) The number of coliform units per 100 mL in the influent and 0.20 mg/L (Cu) for irrigating crops. A 100% removal wastewater obtained exceed the WHO [52] permissible limit efficiency was recorded for Lead and Copper at filtration rate of ≤ 1.0 × 103 cfu/100 mL for wastewater irrigation of crops. of 215 L/h m2. At a filtration rate of 399 L/h m 2, the percent- However, after treatment, the results show there was a 100% age removals of 83.7%, 78.3% and 98.5% were recorded for removal efficiency of the Total coliform at filtration rates of Pb, Fe and Cu, respectively. A low percentage of 34.9% was 215 L/h m2 and 399 L/h m2. The E. coli removal efficiency recorded for Fe at the filtration rate of 215 L/h m 2 for the was 100% at the filtration rate of 399 L/h m2. These removal depth of 30 cm. The values at 664 L/h m2 were, however, efficiencies agree with the study conducted by Bagundol below detectable limits (BDL). et al. [23] who reported a 100% removal efficiency for 30-cm depth of sand filter at flow-through rates of 200 L/h m2 and 400 L/h m2. In addition, the results of the study by Mwabi Heavy metal analysis of wastewater samples at 40‑cm sand et al. [53] also recorded a percentage removal of 99–100% of bed thickness coliform bacteria from the raw water sample tested. The results show significant reduction in the heavy metals concentration in the filtered water with as high as 98.3%, 1 3 184 International Journal of Energy and Environmental Engineering (2021) 12:175–190 86.3% and 100% removal efficiencies for Pb, Fe and Cu, results are comparably low compared to 97.9–99.9% respectively. The percent removal efficiency of Cu metal was reported by Zhang et al. [56]. significantly high at this bed thickness recording more than In addition, the average removal efficiencies of 90% removal in concentration. The percentage removal effi- 61.2–98.8% for Pb show an improvement upon the 31–61% ciencies of this bed thickness is significantly high compared reported by Zhang et al. [56] but Mbir and Tetteh-Narh [54] to the 31.9% (Cu) recorded in Mbir and Tetteh-Narh [54]. reported 100% removal efficiency when activated charcoal These results indicate the effectiveness of the sand bed of layer was used. Further, the average removal efficiency of depth 40 cm in the efficient removal of these trace elements 66.2–85.3% recorded for Cu also shows an improvement in the influent wastewater. The removal of Fe at high filtra- upon 32.8% reported by Mbir and Tetteh-Narh [54]. The tion rate of 664 L/h/m2 was, however, as low as 46.69% low removal efficiencies may be due to the fact that the representing the lowest removal efficiency of the heavy sand bed thickness of 0.30 m was small and the sand media metals at this depth. This is consistent with theory because, was unable to adhesively bind the trace metals to itself due as the filtration rate increases, the removal efficiency also to higher the concentrations in the water sample than the decreases. media can remove. But generally, Fe is best removed from water by oxidation of F e2+ to Fe3+ which precipitates. Fig. 7 Heavy metal analysis of wastewater samples at 50‑cm sand shows a graph of average removal efficiencies at different bed thickness bed thickness. There were significant reductions in the heavy metal con- Water quality analysis for effective sizes centration in the effluent with as high as 100% removal effi- ciencies for all three metals analysed. At a filtration rate of For each of the effective sizes 0.45 mm and 0.27 mm, the 664 L/h m2, Fe recorded as low as 53.81% at 50-cm bed wastewater samples were tested before and after the forma- thickness. However, according to Khatri et al. [55], at 60 cm tion of the Schmutzdecke at the flowrate of 0.105 m 3/h and of sand bed thickness, there was 90.2% removal efficiency. conventional hydraulic loading rate of 0.1 m/h. Average percentage removal of contaminants Biological analysis of wastewater sample The average percentage removal of Total coliform, E. coli, before Schmutzdecke formation Lead, Iron and Copper for the different sand bed thick- ness of 0.30 m, 0.40 m and 0.50 m is presented in Table 3. Results before the formation of the Schmutzdecke showed According to Bagundol et al. [23] and Thomas and Kani that the influent concentration of Total coliform counts of 3 [32], the coliform removal efficiency increases as the depth 1.73 × 10  cfu/100 mL exceeded WHO [52] permissible limit 3 of the sand bed is increased. The increasing high removal of 1 × 10  cfu/100 mL while the E. coli recorded a lower 2 efficiencies recorded may be due to the maturing of the count of 3.3 × 10  cfu/100 mL relative to the WHO permis- Schmutzdecke thus increasing the biological activity of the sible limit. Upon treatment, there were significant reductions filter is enhanced with increasing filter depths. Hence, the in Total coliform and E. coli counts with removal efficiencies microorganisms and other particles have to travel longer dis- of 90.1% and 87.9%, respectively, for 0.45 mm effective size tance through the sand. Some of these microorganisms gets sand media. The setup with sand media of effective size of trapped between sand grains, while others are adhesively 0.27 mm recorded the highest removal efficiencies of 98.3% held onto the sand grains. Thus, a higher removal efficiency and 96.7% for Total coliform and E. coli, respectively. The is expected at deeper depths of the sand bed [24, 32, 38]. The control setup also showed significant removal efficiencies high percentage removals of the Total coliform and E. coli of 71.1% and 93.3% removal efficiencies for Total coliform from the wastewater means reduced risk of coliform would and E. coli, respectively. be found in the effluent irrigated vegetable after harvest. The removal efficiency of the heavy metal concentration can be attributed to the adsorption of the heavy metals into Table 3 Percentage removal of contaminants at different thickness of sand bed the Schmutzdecke and adhesion to the sand grains in the sand media [56]. The average removal efficiencies of 37.7–84.1% Bed thick- Average percentage removal of the sand filter for Fe confirms the results of Mbir and ness (m) Total coliform E. coli Pb Fe Cu Tetteh-Narh [54] who reported a 38.1% removal efficiency and Khatri et al. [55] who reported 90.15% removal effi- 0.30 66.67 33.33 61.23 37.73 66.17 ciency at 0.60 m of sand bed thickness. However, these 0.40 99.97 99.99 74.40 70.69 96.67 0.50 100 99.54 98.80 84.07 85.33 1 3 International Journal of Energy and Environmental Engineering (2021) 12:175–190 185 Fig. 7 Average removal effi- ciencies of Total coliform, E. Removal efficiency (%) at different bed thickness (m) coli, Pb, Fe and Cu at different bed thickness 110 100 90 80 70 60 50 40 30 0.3 0.4 0.5 Bed Thickness (m) T.C E. coli Pb Fe Cu Biological analysis of wastewater sample Fe can be attributed to adsorption to the container and pre- after Schmutzdecke formation cipitation due to reactions between the wastewater and the atmospheric gases [58]. There were, however, no traces of There was a significant reduction in the coliform counts Cu in the influent and effluent water samples. in wastewater with the system recording as high as 100% removal efficiency each for Total coliform and E. coli for d10 Removal of contaminants of 0.27 mm. The filter with d10 of 0.45 mm also recorded a 100% removal of E. coli and 95% removal of Total coliform. The results show an improvement in the water quality after The control recorded a 100% removal of E. coli and 75% treatment using the slow sand filter. It was observed that the removal of Total coliform. effluent quality attains higher percentage removal of heavy metals after the formation of the Schmutzdecke. This high- Heavy metal analysis of wastewater sample lights the importance of the adsorption that takes place in before Schmutzdecke formation the Schmutzdecke. In addition, it affords the opportunity for microbes to feed on the waste in the wastewater, thereby The initial concentration of heavy metals in the wastewa- reducing its concentration. ter sample was 0.34 mg/L for Pb, 0.04 mg/L for Cu and 3.24 mg/L for Fe, respectively. However, it was noted that Removal efficiency of sand filter with effective size these concentrations were below WHO [52] permissible lim- of 0.45 mm its of 0.2 mg/L for Cu and 5 mg/L for Pb and Fe. There were no traces of Pb in the effluent for all three There was significant removal of contaminants in the waste- setups representing a 100% removal efficiency. For both water after filtration with sand media of d10 = 0.45 mm effective sizes, Cu was completely removed with the control before and after the formation of the Schmutzdecke. In the showing a 43.90% removal. The percentage removal of Fe removal of E. coli, the removal efficiencies of 90.1% (before was 24.28%, 37.94% and 23.29% for d 0.45 mm, 0.27 mm formation of Schmutzdecke) and 95% (after formation of 10 and control experiment, respectively. The low percentage Schmutzdecke) of the filter were consistent with the removal removals may be due to the fact that Fe is better removed efficiencies of 96.7–99% reported by Muhammad et al. [24] from water by means of oxidation [57]. for sand media with d10 = 0.45 mm. The removal efficien- cies of Total coliform for before and after the formation of Heavy metal analysis of wastewater sample Schmutzdecke at the same effective size were 87.9% and after Schmutzdecke formation 100%. This was also consistent with 96.4–98.6% reported despite the slight reduction. This confirms the feasibility After the Schmutzdecke formation, there was a 100% of removing coliforms in wastewater using sand media of removal of Pb and Fe for the selected effective sizes while d10 = 0.45 mm. The performance of the filter with sand the control experiment recorded a 100% removal of Pb and media of d 10 = 0.45 mm in the removal of Pb, Cu and Fe was 34.59% removal for Fe. This low percentage removal for significant and can be attributed to the adhesion onto sand 1 3 Removal Efficiency (%) 186 International Journal of Energy and Environmental Engineering (2021) 12:175–190 grains and adsorption into the Schmutzdecke. The removal Analysis of bed thickness data efficiencies of 100% recorded is in line with the 99.6% (Cu) and 100% (Pb) reported by Muhammad [59] despite using An analysis of the results showed that the effect of the sand media of d10 = 0.32 mm. selected sand bed thicknesses on the efficient removal of col- iforms from the wastewater sample was significant for Total coliform and E. coli (see Tables 4 and 5 in appendix). In Removal efficiency of sand filter with effective size addition, the analysis of the results of heavy metal removal of 0.27 mm showed that the effect of these sand bed thicknesses on the The E. coli removal increased from 96.7% (before Schmutz- efficient removal of Pb, Fe and Cu from the water were also decke formation) to 100% (after Schmutzdecke formation) significant (see Tables 6, 7 and 8 in appendix). Therefore, with sand media of d = 0.27 mm. These removal efficiencies the null hypothesis is rejected. This means that the removal 10 were consistent with the 100% reported by Bagundol et al. of contaminants from the wastewater is increased as the sand [23] at d10 between 0.16 mm and 0.30 mm. From a removal bed thickness is increased. efficiency of 98.3% (before Schmutzdecke formation), the Total coliform removal increased significantly to 100% (after Analysis of effective size data Schmutzdecke formation). The overall coliform removal from the water was observed to recur in the filter after the matu- The ANOVA results show that there was significant dif- ration of the Schmutzdecke. The significantly high removal ference in the removal efficiency of E. coli between efficiencies recorded have been found to verify the 99.3% and d10 = 0.45 mm and 0.27 mm (see Table 9 in appendix). 99.6% reported by Muhammad et al. [24] for the sand media Hence, the null hypothesis is rejected. However, the results with d10 of 0.35 mm and 0.20 mm, respectively. The filter still obtained (see Tables 10, 11, 12 and 13 in appendix) for achieved significantly high removal efficiency levels for the Total coliform, Cu, Pb and Fe show that there were no heavy metals (Pb, Cu and Fe) investigated before and after significant differences in the removal efficiencies between the Schmutzdecke was formed. The results of contaminant sand media of d10 = 0.45 mm and 0.27 mm. For Total coli- removal with sand media of d = 0.27 mm were slightly better form, Cu, Pb and Fe, we fail to reject the null hypothesis 10 than sand media with d = 0.45 mm. which means that there are no significant differences in the 10 removal efficiency of both sand media of d10 = 0.45 mm and 0.27. Control experiment The control setup also recorded some reduction in the con- Conclusion taminants level. The reduction in biological parameters in the control shows that external factors in the environment The treatment of wastewater to be used for irrigation pur- also play a role in the removal of pollutants in wastewater. poses using a relatively low-tech slow sand filter is feasi- The reduction in coliform counts can be as a result of micro- ble and can be adopted by small-scale vegetable farmers bial degradation that took place during the residence time. whose main source of water for irrigating their crops is the The reduction in heavy metal contaminants concentration wastewater from the drains. This is because the efficiency [100% (Pb), 43.90% (Cu) and 23.29% (Fe)] can be as a result of the slow sand filter to remove the parameters analysed; of adsorption to container, reactions that took place between E. coli, Total coliform, Pb, Fe and Cu varied and were the wastewater and the environment (atmospheric gases) significantly high at sand bed thicknesses of 40 cm and over time and precipitation [58, 60]. The concentration of 50 cm. In addition, there was a significantly high removal samples that remained below detectable levels implies that of the parameters for both sand media of effective sizes, the heavy metal was not introduced into the sample from 0.27 mm and 0.45 mm. However, the SSF with an effec- other external sources. tive size of 0.27 mm obtained slightly better results in the removal of the contaminants. It was established that with a Statistical analyses minimum sand bed thickness of 40 cm and sand media of effective sizes up to 0.45 mm, wastewater from the drain A one-way Analysis of Variance (ANOVA) was conducted can be treated for irrigation of vegetable crops. The impli- to compare the effect of the selected thickness (30 cm, 40 cm cation is that the extra cost of acquiring extra sand with the and 50 cm) of sand bed and effective sizes (0.45 mm and aim of improving wastewater quality can be saved. 0.27 mm) on the efficient removal of contaminants from the It must be indicated that investigation of effective sizes water. The complete table of ANOVA results can be found was subsequent to establishing that the thicker the bed in the appendix section. depth, the higher the removal efficiency of contaminants 1 3 International Journal of Energy and Environmental Engineering (2021) 12:175–190 187 and so both the experiments were conducted at differ- Appendix: ANOVA results ent times. The combined effects of sand bed depth and effective sizes are being investigated and this will be later See Tables 4, 5, 6, 7, 8, 9, 10, 11, 12, and 13. reported. Part A: Bed thickness. Part B: Effective size. Table 4 ANOVA results for Source of variation SS df MS F p value F E. coli crit ANOVA  Between groups 10750.67 1 10750.67 18.13 0.001 4.49  Within groups 9486.68 16 592.92  Total 20237.34 17 Table 5 ANOVA results for Source of variation SS df MS F p value F Total coliform crit ANOVA  Between groups 6369.44 1 6369.44 6.33 0.02 4.49  Within groups 16094.70 16 1005.92  Total 22464.14 17 Table 6 ANOVA results for Source of Variation SS df MS F p value F Lead crit ANOVA  Between groups 6546.73 1 6546.73 11.06 0.004 4.49  Within groups 9469.92 16 591.87  Total 16,016.66 17 Table 7 ANOVA results for Source of variation SS df MS F p value F Iron crit ANOVA  Between groups 2627.64 1 2627.64 4.50 0.05 4.49  Within groups 9351.97 16 584.49  Total 11979.61 17 Table 8 ANOVA results for Source of variation SS df MS F p value F Copper crit ANOVA  Between groups 8213.77 1 8213.77 14.51 0.002 4.49  Within groups 9054.88 16 565.93  Total 17268.65 17 1 3 1 88 International Journal of Energy and Environmental Engineering (2021) 12:175–190 Table 9 ANOVA results for Source of variation SS df MS F p value F E. coli crit ANOVA  Between groups 736.03 2 368.01 52.308 0.005 9.552  Within groups 21.11 3 7.04  Total 757.13 5 Table 10 ANOVA results for Source of variation SS df MS F p value F total coliform crit ANOVA  Between groups 19.69 2 9.84 0.292 0.766 9.552  Within groups 101.24 3 33.75  Total 120.92 5 Table 11 ANOVA results for Source of variation SS df MS F p value F copper crit ANOVA  Between groups 1049.07 2 524.535 0.144 0.872 9.552  Within groups 10963.61 3 3654.535  Total 12012.68 5 Table 12 ANOVA results for Source of variation SS df MS F p value F lead crit ANOVA  Between groups 0 2 0 65,535 9.552  Within groups 0 3 0  Total 0 5 Table 13 ANOVA results for Source of variation SS df MS F p value F Iron crit ANOVA  Between groups 1834.19 2 917.10 0.567 0.618 9.552  Within groups 4856.33 3 1618.78  Total 6690.52 5 References 4. FAO: Quality control of wastewater for irrigated crop production (Water reports-10) [Internet]. Westcot DW, editor. Rome (1997). 1. Khalil, S., Kakar, M.K.: Agricultural use of untreated urban https: //webcac he.google userc onten t.com/search ?q=cache: https: // wastewater in Pakistan. Asian J. Agric. Rural Dev. 1, 21–26 www.bvsde .ops-oms.org/bvsai r/e/repin dex/repi8 4/vleh/fullt ext/ (2011) acrob at/W5363 e031.pdf 2. FAO: Food and Agriculture Organization of the United Nations 5. Itchon, G.S., Gensch, R.: Pathogens and contaminants [Internet] (FAO) [Internet]. Water (2010). https ://www.fao.org/water /en/. (2011). https: //www.sswm.info/conten t/pathog ens-contam inant s. Accessed 17 Sep 2019 Accessed 4 Oct 2019 3. Raschid-sally, L., Jayakody, P.: Drivers and characteristics of 6. LeChevallier, M.W., Au, K.-K.: Water Treatment and Pathogen wastewater agriculture in developing countries: Colombo, Sri Control World Health Organization titles with IWA Publishing. Lanka (2008) 1 3 International Journal of Energy and Environmental Engineering (2021) 12:175–190 189 World Health Organization and International Water Association, rule-based modelling. Environ. Model Softw. 19(5), 507–515 Cornwall (2004) (2004) 7. Pandey, P.K., Kass, P.H., Soupir, M.L., Biswas, S., Singh, V.P.: 2 6. Mcnair, D.R., Sims, R.C., Sorensen, D.L., Hulbert, M., Mcnair, Contamination of water resources by pathogenic bacteria. MMB D.R., Sims, R.C., et  al.: Schmutzdecke characterization of Express 4, 51 (2014) clinoptilolite-amended slow sand filtration. J. Am. Water Work 8. Abegunrin, T.P., Awe, G.O., Idowu, D.O., Adejumobi, M.A.: Assoc. 79(12), 74–81 (1987) Impact of wastewater irrigation on soil physico-chemical proper- 27. Barrett, J.M., Bryck, J., Collins, M.R., Janonis, B.A., Logsdon, ties, growth and water use pattern of two indigenous vegetables G.S.: Manual of design for slow sand filtration. In: Hendricks, in southwest Nigeria. CATENA 139, 167–178 (2016). https: //doi. D. (ed.) Journal of American Water Works Association. AWWA org/10.1016/j.caten a.2015.12.014 Research Foundation and American Water Works Association, 9. Darvishi, H.H., Manshouri, M., Farahani, H.A.: The effect of Denver (1991) irrigation by domestic waste water on soil properties. J. Soil Sci. 28. Mah, B.: Slow Sand Filter Agriculture and Agri-Food Canada Environ. Manag. 1, 30–33 (2010) Prairie Farm Rehabilitation Administration Agriculture et 10. Prinz, D., Singh, A.K.: Water resources in arid regions and their Agroalimentaire Canada Administration du rétablissement agri- sustainable management. Ann. Arid Zone 39(3), 251–271 (2000) cole des Prairies. PFRA Water Quality Matters (2001). www. 11. Hoek, W.V.D., Hassan, M.U., Ensink, J.H.J., Feenstra, S., quantu mlynx .com/water Raschid-sally, L., Munir, S., et al.: Urban wastewater: A valuable 2 9. Tchobanoglous, G., Burton, F.L., Stensel, H.D.: Wastewater- resource for agriculture a case study from Haroonabad, Pakistan. Engineering-Treatment-and-Reuse. Metcalf and Eddy Inc., Colombo, Sri Lanka (2002). https: //www.ais.unwate r.org/ais/plugi California (2003) nfile .php/225/mod_label /intro /Resea rch.Report -63.pdf 3 0. Huisman, L., Wood, W.E.: Slow sand filtration. In: Huisman, L., 12. Mara, D.: Domestic Wastewater Treatment in Developing Coun- Wood, W. (eds.) Editio, pp. 15–108. World Health Organization tries, pp. 230–251. Earthscan Publications, London (2004) (WHO), Geneva (1974) 13. Drechsel, P., Keraita, B., Raschid-Sally, L., Adam-Bradford, A.: 3 1. Visscher, J.T.: Facilitating Community Water Supply Treatment: Irrigated Urban Vegetable Production in Ghana Irrigated Urban From transferring filtration technology to multi stakeholder Vegetable Production in Ghana : Characteristics, Benefits and learning, p. 256. Wageningen Universiteit, Wageningen (2006) Risk Mitigation, 2nd edn., p. 247. International Water Manage- 32. Thomas, T.A., Kani, K.M.: Efficiency of slow sand filter in ment Institute (IWMI), Sri Lanka (2014) wastewater treatment. Int. J. Sci. Eng. Res. 7(4), 315–317 14. Qadir, M., Wichelns, D., Raschid-sally, L., Mccornick, P.G., (2016) Drechsel, P., Bahri, A., et al.: The challenges of wastewater 33. Pyper, G.R., Logsdon, G.: Slow Sand Filter Design, pp. 122– irrigation in developing countries. Agric. Water Manag. J. 97, 148. American Society of Civil Engineers, New York (1991) 561–568 (2010) 3 4. Troyan, J.J., Hansen, S.P.: Treatment of Microbial Contaminants 1 5. WHO: Guidlines for the Safe use of Waste Water, Excreta and in Potable Water Supplies : Technologies and Costs. Noyes Data Grey Water. Volume 2 Wastewater Use in Agriculture. WHO, Corporation, California (1989) Geneva (2006) 35. Poynter, S.F., Slade, J.: The removal of viruses by slow sand 1 6. Becerra-castro, C., Rita, A., Vaz-moreira, I., Silva, E.F., filtration. Progress in water technology. Am. J. Int. Assoc. Water Manaia, C.M., Nunes, O.C.: Wastewater reuse in irrigation : Pollut. Res. 9(1), 75–88 (1977) A microbiological perspective on implications in soil fertility 3 6. Van Dijk, J., Oomen, J.H.: Slow sand filtration for community and human and environmental health. Environ. Int. 75, 117–135 water supply in developing countries; a design and construction (2015). https ://doi.org/10.1016/j.envin t.2014.11.001 manual. Voorburg, The Hague, The Netherlands. Report No.: 17. Engwa, G.A., Ferdinand, P.U., Nwalo, F.N., Unachukwu, 11 (1978) M.N.: Mechanism and health effects of heavy metal toxicity in 37. Logsdon, G.S., Kohne, R., Abel, S., LaBonde, S.: Slow sand humans. In: Poisoning in the Modern World, pp. 2–24. IntechO- filtration for small water systems. J. Environ. Eng. Sci. 1(5), pen, London (2019) 339–348 (2002). https ://doi.org/10.1139/s02-025 18. Lente, I., Ofosu-Anim, J., Brimah, A.K., Atiemo, S.: Heavy 3 8. Ellis, K.: Slow sand filtration. CRC Crit. Rev. Environ. Control metal pollution of vegetable crops irrigated with wastewater in 15(4), 315–354 (1985) Accra, Ghana. West Afr. J. Appl. Ecol. 22(1), 41–58 (2014) 39. Guchi, E.: Review on slow sand filtration in removing microbial 19. Tom, M., Fletcher, T.D., McCarthy, D.T.: Heavy metal contami- contamination and particles from drinking water. Am. J. Food nation of vegetables irrigated by Urban stormwater: A matter of Nutr. 3(2), 47–55 (2015) time? PLoS ONE 9(11), 1–21 (2014) 4 0. Das, B.M.: Principles of Geotechnical Engineering, p. 42. Cen- 20. Gottinger, A.M., Mcmartin, D.W., Price, D., Hanson, B.: The gage Learning, Boston (2010) effectiveness of slow sand filters to treat Canadian rural prairie 4 1. Van Der Hoek, J.P., Orlandini, E., Graveland, A., Smeenk, water. Can. J. Civ. Eng. 38(August), 455–463 (2011) J.G.M.M.: Slow sand filtration: effect of grain size and filtration 2 1. Weber-shirk, M.L.: Enhancing slow sand filter performance with rate on operation and performance. In: Graham, N.J.D., Collins, an acid-soluble seston extract. Water Res. 36, 4753–4756 (2002) R. (eds.) Advances in Slow Sand and Alternative Biological 2 2. Rooklidge, S.J., Burns, E.R., Bolte, J.P.: Modeling antimicro- Filtration, pp. 201–210. John Wiley and Sons, Chichester (1996) bial contaminant removal in slow sand filtration. Water Res. 39, 4 2. Rolland, L., Molle, P., Lie, A.: Influence of the physical and 331–339 (2005) mechanical characteristics of sands on the hydraulic and bio- 23. Bagundol, T.B., Awa, A.L., Enguito, M.R.C.: Efficiency of slow logical behaviors of sand filters. Desalination 248, 998–1007 sand filter in purifying well water. J Multidiscip. Stud. 2(1), (2009) 86–102 (2013) 43. Langenbach, K.: Slow Sand Filtration of Secondary Effluent for 24. Muhammad, N., Ellis, K., Parr, J., Smith, M.D.: Optimization of Wastewater Reuse: Evaluation of Performance and Modeling slow sand filtration. In: Reaching the Unreached: Challenges for of Bacteria Removal Herbert Hieronymus Druck amd Verlag the 21st Century. New Delhi, India, pp. 283–285 (1996). https [München]. Technical University München, München (2009) ://wedc.lboro. ac.uk/resou rces/confer ence /22/Muham me.pdf 4 4. Di Bernardo, L., Escoba, R.A.: Influence of sand uniformity 2 5. Sadiq, R., Al-Zahrani, M.A., Sheikh, A.K., Husain, T., Farooq, coefficient on slow sand filtration performance. In: Graham, N., S.: Performance evaluation of slow sand filters using fuzzy Douglas, J., Collins, R.M. (eds.) Advances in Slow Sand and 1 3 190 International Journal of Energy and Environmental Engineering (2021) 12:175–190 Alternative Biological Filtration, pp. 179–188. John Wiley and 55. Khatri, N., Tyagi, S., Rawtani, D.: Recent strategies for Sons, Chichester (1996) the removal of iron from water: a review. J. Water Process. 4 5. Błażejewski, R., Murat-Błażejewska, S.: Water retention time Eng. 19(13), 291–304 (2017). https ://doi.org/10.1016/j. in intermittently dosed sand filters. Pol. J. Environ. Stud. 18(2), jwpe.2017.08.015 289–292 (2009) 56. Zhang, B., Gao, L., Fazal, S., Mahmood, Q., Laghari, M., Sayal, 46. Kaczkowski, J., Guerra-Mondragon, N., Lammers, R., Cox, K., A.: Biosand filter containing melia biomass treating heavy metals Mckenna, R., Wink, S., et al.: Drinking Water Issues in Rural and pathogens. Pol. J. Environ. Stud. 25(2), 859–864 (2016) Colombia. Colombia (2012) 57. Harmon, S.: Iron Removal: A World Without Rules–A complete 4 7. Campos, L.C., Su, M.F.J., Graham, N.J.D., Smith, S.R.: Biomass guide to iron removal methods, equipment & their limitations. development in slow sand filters. Water Res. 2002(36), 4543–4551 Water Technology Magazine, Berlin (2003) (2001) 58. Ayres, D.M., Davis, A.P., Gietka, P.M.: Removing heavy metals 48. Eliasson, J.: Hydraulic loading rates. Washington (2001) from wastewater. Maryland (1994) 4 9. Missimer, T.M., Jones, B., Maliva, R.G.: Intakes and Outfalls for 5 9. Muhammad, N.: Removal of Heavy Metals by Slow Sand Filtra- Seawater Reverse-Osmosis Desalination Facilities, pp. 198–200. tion. Loughborough University, Loughborough (1998) Springer International Publishing, Switzerland (2015) 6 0. Shabarova, T., Villiger, J., Morenkov, O., Niggemann, J., Ditt- 5 0. Abudi, Z.N.: The effect of sand filter characteristics on removal mar, T., Pernthaler, J.: Bacterial community structure and dis- efficiency of organic matter from grey water. Al-qadisiya. J. Eng. solved organic matter in repeatedly flooded subsurface karst Sci. 4(2), 143–155 (2011) water pools. FEMS Microbiol. Ecol. (2014). https ://doi. 51. Anggraini, K.A.S.: Optimization of Slow Sand Filtration Design org/10.1111/1574-6941.12339 by Understanding the Influence of Operating Variables on the 6 1. Great Lakes Upper Mississippi River Board of State Public Health Suspended Solids Removal. Karlsruhe Institute of Technology, and Environmental Managers. 1987. Recommended Standards for Karlsruhe (2018) Water Works. Albany, NY: Health Research Inc., Health Educa- 52. WHO: Health Guidelines for the Use of Wastewater in Agriculture tion Services Division. and Aquaculture. World Health Organization-Technical Report 62. Visscher, J. T., R. Paramasivam, A. Raman, and H. A. Heijnen. Series, Geneva (1989) 1987. Slow Sand Filtration for Community Water Supply, Plan- 53. Mwabi, J.K., Mamba, B.B., Momba, M.N.B.: Removal of Escheri- ning, Design, Construction, Operation, and Maintenance. Techni- chia coli and faecal coliforms from surface water and groundwa- cal Paper No. 24, The Hague, Netherlands: International Refer- ter by household water treatment devices/systems: a sustainable ence Center for Community Water Supply and Sanitation. solution for improving water quality in rural communities of the Southern African Development Community Region. Int. J. Envi- Publisher’s Note Springer Nature remains neutral with regard to ron. Res. Public Health 10(9), 139–170 (2012) jurisdictional claims in published maps and institutional affiliations. 5 4. Mbir, B.I., Tetteh-Narh, R.: Using slow sand filtration system with activated charcoal layer to treat salon waste water in a selected community in Cape Coast, Ghana. J. Adv. Chem. Eng. 5(4), 8 (2015) 1 3