A review of sewage sludge dewatering and stabilisation in reed bed system: towards the process-based modelling

Abstract

The sludge treatment reed bed is a sustainable sewage sludge management technology that offers promising sludge dewatering and stabilisation, due to its low energy and labour requirements, and minimal operating and maintenance costs. Despite numerous existing studies and increasing applications of sludge treatment reed bed in both developed and developing countries, the lack of standard system configurations and operating regimes led to several operational problems, including bed clogging and poor dewatering and mineralisation efficiency. Sludge dewatering and stabilisation in STBRs are complex processes governed by system- and operation-related factors such as the substrate media, macrophytes, loading regime, sludge characteristics, and climate conditions. Rather than the complex experimental study, process-based modelling has become a promising approach to investigating the influence of specific factors on the efficiency of sludge dewatering and stabilisation. This paper presents an overview of the critical system and operating parameters in sludge treatment reed beds and a summary of the interactions between the factors. Then, a framework of process-based modelling is proposed, which provides a useful platform to study the interactions of operating parameters and other factors in the complex processes of sludge dewatering and stabilisation, which supports the system design and optimisation of sludge treatment reed beds.

Introduction

In recent years, many countries have included sewage sludge management in their “to-do list” to meet the sustainable development goal of “Clean Water and Sanitation”. Sewage sludge is a semi-solid by-product generated from sewage treatment that exists in a volatile form. In developing countries, sewage sludge is generated from sewage treatment plants and onsite sanitation systems (Tayler 2018). For the sewage sludge from the sewage treatment plant, the most common sludge includes primary and secondary sludge, where the primary sludge is the sludge collected from the primary clarifier, and the secondary sludge is the surplus-activated sludge (Nielsen and Stefanakis 2020). As for the onsite sanitation system, the sludge is known as faecal sludge. Since the septic tank is a standard onsite sanitation system, the sludge produced is also called septic sludge or septage (Chandana and Rao 2022).

Sewage sludge management must comply with the policies, guidelines, and regulatory frameworks (Jain et al. 2022). According to the National Water Services Commission guidelines in Malaysia, the sludge must be dewatered and treated, where the total solid content in liquid form must not exceed 4.0%. Besides, it must attain a dry solid content of at least 20% before being disposed of or used for other purposes. Generally, centralised treatment plants equipped with mechanical dewatering devices are considered the ideal practice for sludge treatment. However, the cost of capital investment and energy consumption of the centralised treatment plant has always been a concern for sewage sludge management in developing countries. In contrast with the high capital and operating expenditures of conventional centralised sludge treatment facilities, decentralised systems utilise simple construction, operation, and maintenance methods, as well as involve lower sludge transportation costs, and have attracted more and more attention as a result (Tan et al. 2020). In addition, the decentralised sludge management system could bring considerable additional income from the sludge treatment process, in which the treated sludge can be applied for agricultural purposes.

The sludge treatment reed bed (STRB), which is an engineered system combining the concepts of conventional drying beds and vertical flow constructed wetlands (VFCW), has been proven to be a promising technology for sludge management. Sewage sludge is loaded onto the top surface of the reed beds intermittently, where it is dewatered through drainage and evapotranspiration, and the sludge residue is simultaneously stabilised. Studies have shown that the sludge dewatered via STRB can reach a dry solid content of 25–40%, which is comparable to mechanical dewatering techniques such as belt presses and centrifuges (Nielsen and Stefanakis 2020). Since the STRB is a chemical-free system, the sludge residue could be recovered as biosolids for agricultural purposes (Collard et al. 2017; Kołecka et al. 2016). In addition to the promising dewatering performance, the STRB can also treat the supernatant (Tan et al. 2020). The liquid percolates through the reed bed and is treated along the way; thus, the final effluent, so-called leachate, is generally desirable in quality. Tayler (2018) indicated that the concentration of Total Solids (TS), Chemical Oxygen Demand (COD), and Biological Oxygen Removal (BOD) in the effluent from STRB were found to be lower than that of the mechanical dewatering methods. In general, the COD concentration in the sludge can be reduced by at least 60%, and most of the ammonium nitrogen is nitrified (Brix 2017). A review article also showed that many studies presented more than 90% removal for TS and COD in the leachate, as well as promising removal efficiencies for nutrients and other pollutants (Jain et al. 2022).

Although the capital cost for a STRB is often higher than a mechanical dewatering device due to its large land requirement and footprint, it offers high energy efficiency and chemical-free dewatering and treatment processes. Further, the overall cost for sludge dewatering by mechanical techniques is 50% to 90% higher than the STRB, mainly due to the high energy input and chemical dosages (Nielsen and Stefanakis 2020; Nielsen and Larsen 2016). These economic savings are crucial for developing countries, where aspects such as operational costs must be weighed against capital expenditure on sludge management facilities. Therefore, the STRB is widely regarded as an eco-friendly solution for sludge treatment due to its natural processes and economic and environmental advantages (Kołecka et al. 2016).

There are no standard design and operational parameters for the STRB, and the application is hindered by several operational problems, including poor macrophyte growth, bed clogging, and slow dewatering and mineralisation rates (Peruzzi et al. 2017). These problems require further studies to improve the competitiveness of STRB against centralised wastewater treatment plants by optimising the technology’s design, operation, and monitoring. Although there have been several review articles on the STRB, standard system configurations and operating strategies have not materialised due to its highly dynamic performance (Jain et al. 2022; Bui et al. 2019; Brix 2017; Tan et al. 2015). Sludge dewatering and stabilisation of the reed bed system is a complex process involving physical and biochemical mechanisms. Its efficiency is affected by various system- and operation factors, including sludge characteristics, reed bed characteristics, loading rate, loading frequency, and climate (Hu et al. 2017). These factors interfere with each other and increase the STRB’s design and operation complexity. For example, although frequent loading and a high loading rate reduces land area requirements, such a loading pattern results in thicker sludge residues and extends the dewatering time needed (Tan et al. 2017; Uggetti et al. 2010).

Rather than the laboratory- and pilot-scale experimental study, process-based modelling has become a promising approach to investigating the performance and aiding the design of the system and operational parameters of STRBs (Bui et al. 2019; Uggetti et al. 2010). Process-based modelling can relate sludge dewatering, including the moisture content and organic content of sludge residue and the quantity and quality of supernatant, to the system-related factors such as reed bed properties and thickness, and operation-related factors such as loading rates and drying period. However, there are only a limited number of numerical models for STRBs. Uggetti et al. (2012) developed a model using Terzaghi’s consolidation theory to simulate sludge dewatering via gravity drainage and evapotranspiration. Tan (2016) adopted the Richards equation and the Activated Sludge Model to predict the leachate flow rate and nitrogen dynamics in a reed bed treating septic sludge. Meanwhile, He et al. (2021) also applied the Activated Sludge Model to analyse the stabilisation of sludge residue on the reed beds.

However, these numerical models still lack the capability to assess the interference between the system and operation parameters in STRBs on the sludge dewatering and treatment performance. Therefore, this study aims to review the performances of the STRB and associated systems and operation parameters in recent articles, and recommend the crucial parameters and their interactions with other factors in the process-based modelling of the sludge dewatering and treatment performance.

Materials and methods

There are numerous publications on sludge dewatering and stabilisation in STRBs in scientific research databases. The publications were collected from the databases of Web of Science and Scopus, where only articles in English were considered. The keywords used to perform the articles search include: “Sludge Treatment Reed Bed”, “Sludge Drying Reed Bed”, and “Sludge Treatment Wetland”. Only the studies published in the recent ten years were considered to ensure the latest developments of STRB are reviewed. In addition, criteria were set to decide whether the publication is taken into consideration, including:

• The studies included were limited to treating sewage sludge, including faecal sludge, septic sludge, primary sewage sludge, and surplus-activated sludge. The treatment of sludge produced from the water treatment or industrial processes, as well as synthetic sludge produced in the laboratory, were excluded;

• The studies should include information on system configurations and operational parameters, including dimensions, reed bed profile, macrophyte type, loading rate, and loading frequency;

• The studies should include information on system performance, including dewatering and stabilisation efficiency or the leachate treatment efficiency;

• In order to distinguish between STRBs and sludge drying beds, studies on unplanted sludge drying beds were also excluded unless the beds were intentionally unplanted; and

• The studies on review, general concepts, economic and environmental assessments, and greenhouse gas emissions were excluded.

Based on the selection criteria, a total of 60 publications from 2013 until 2022 were considered. It is important to highlight that a few publications were found with the same system and operational parameters. These publications were still considered but regarded as the same system; thus, 51 sets of the system and operational parameters were included in the analysis. Figure 1 shows the location of the selected publications, where most studies were from Asia and Europe, with 24 and 20 publications, respectively. In Asia, China had fourteen publications, followed by Malaysia with seven publications. Denmark and Poland contributed five publications each for Europe, while three studies were conducted in Italy. In Africa, three STRB studies were from Burkina Faso, while there were two from Cameroon. Brazil was the main contributor in South America with three publications. There was only one publication from Canada for North America. Among these studies, only twelve systems were operated at full-scale, while the remaining were pilot-, mesocosm-, or laboratory-scale systems. Most of the full-scale STRBs in the selected studies were located in Europe.

Fig. 1

Location of selected publications

Since the number of publications in the last ten years has been relatively low in number, some comparisons were made against research articles published between 2008 and 2012 to examine the roles of the system and operational parameters in STRBs. In this review study, a set of crucial parameters that govern the performance of sludge dewatering and stabilisation were identified. These parameters were further compared with simulation studies to propose a process-based model framework for dewatering and stabilisation processes in STRBs. Due to limited modelling studies for this particular system, simulation studies of similar systems, including treatment wetlands and drying beds, were taken into consideration in the framework development of process-based modelling.

Results and discussion

Sludge characteristics

As mentioned previously, the primary source of sewage sludge includes onsite sanitation systems and sewage treatment plants. Figure 2 shows the classification of sludge types in this review study. There are 37% (19 out of 51) STRB systems treating sludge from various onsite sanitation systems, where the majority were treatment of sludge removed from septic tanks, so-called septic sludge. Meanwhile, four studies investigated sludge treatment from both septic tanks and pit latrines, which were regarded as faecal sludge in this review. A total of 57% (29 out of 51) of STRB systems were categorised as sludge treatment systems for sewage treatment plants. These systems were further classified into the treatment of surplus-activated sludge from secondary treatment only, and a mixture of sludge from primary, secondary, and other treatment processes in the sewage treatment plants. In addition, three systems acted as secondary treatment units following the primary STRBs or other sludge management facilities, which were considered “Other” in this review.

Fig. 2

Sludge types in the selected studies (n = 51)

In STRBs, the priority of the design goal is to achieve desirable sludge dewatering and stabilisation performance. Therefore, the solids content is the foremost parameter in terms of sludge characteristics. The characteristics of septic sludge, faecal sludge, surplus-activated sludge, and the mixture of sludge from primary, secondary, and other treatment processes reported in the selected studies are summarised in Table S1 in the online resource. In terms of solids content, total solids (TS) or dry matter (DM) is the common parameter indicating the total amount of solids in the raw sludge, which provides crucial information for considering operational parameters of STRBs for solid–liquid separation (Chandana and Rao 2022).

The average TS concentrations for septic sludge in most references are above 10,000 mg/L, where the highest TS concentration was 57,600 mg/L from Miri, Malaysia (Tan et al. 2017). In contrast, the TS concentrations of septic sludge in Belo Horizonte were generally below 2400 mg/L, and the average concentration was as low as 706 mg/L (Calderón-Vallejo et al. 2015). The wide range of TS concentrations in the selected studies reflects the high variability in septic sludge quality. Several factors, such as tank dimension, desludging frequency, user habits, and climate, have significantly influenced septic sludge quality (Vincent et al. 2011). Total suspended solid (TSS) is another quality parameter indicative of solids content, and the TSS concentration of septic sludge was typically 70–90% of TS (Kinsley et al. 2018; Sonko et al. 2014; Jong and Tang 2014b).

On the other hand, the average TS concentration of faecal sludge was higher than the septic sludge, as Osei et al. (2019) reported an average concentration of 52,857.6 mg/L. It is consistent with the findings of Jain et al. (2022) and Pandey and Jenssen (2015), where the septic sludge typically has a more extended stabilisation duration than the faecal sludge removed from the pit latrines, resulting in a better solid decomposition. The faecal sludge quality also varied significantly, depending on the removal mechanism and frequency. In contrast, the TS concentrations of surplus-activated sludge were significantly lower than the septic sludge and faecal sludge, where the concentrations were typically below 8000 mg/L. Troesch et al. (2009) indicated that septic sludge is produced from the anaerobic decomposition of organic matter and sediments of inert solids composed of many fine and coarse particles that result in a high TS concentration.

Meanwhile, Vincent et al. (2011) indicated that the extracellular polymeric substances produced in the activated sludge process stimulate the flocculation of microorganisms and organic matter, leading to a more consistent particle size distribution. The same study also highlighted that the high specific surface areas of the fine particles in the septic sludge characterise high water retention capacity, resulting in poorer dewaterability. However, the TS concentrations of sewage sludge from an activated sludge treatment plant reported by El-Gendy and Ahmed (2020) were as high as septic sludge, where it was attributed to the mixing of surplus-activated sludge with the primary sludge, as well as the preliminary thickening of sludge in conventional drying beds.

Volatile solid (VS) is the quality parameter that measures the amount of organic solids in the raw sludge, where the efficiency of sludge stabilisation in STRBs is usually evaluated based on the decrease in VS/TS ratio of the sludge residue over the loading-resting cycle. Based on Table S1, the VS/TS ratios of septic sludge ranged from 44 to 60%, and only one study showed that the ratio reached 68% (Meng et al. 2020). The mixture of sludge from septic tanks and pit latrines had VS/TS ratios between 55 and 66%. Compared to the faecal sludge in pit latrines, the relatively low VS/TS ratio of septic sludge is due to the prolonged anaerobic decomposition of organic matter in the septic tanks. The ratios of surplus-activated sludge varied widely in the selected studies, ranging from 40 to 70%. Vincent et al. (2011) demonstrated that the VS/TS ratio of surplus-activated sludge was generally lower than septic sludge, where the concentrations depended on the scale of sewage treatment plants, sewage characteristics, and sludge age.

Based on the summary of sludge characteristics, the concentration levels of a wide range of hazardous pollutants in the sewage sludge, including chemical oxygen demand (COD), biochemical oxygen demand (BOD), nutrients such as total nitrogen (TN), ammonium (N-NH₄) and nitrate (N-NO₃), total phosphorus (TP), and phosphate (P-PO₄), and pathogenic compounds such as total coliforms and Escherichia coli were much higher compared to the high-strength wastewater as stated in Henze and Comeau (2008). These quality parameters, in particular COD, were crucial in designing STRBs in terms of the leachate treatment capability. The highest average COD concentration of septic sludge reached 35,526 mg/L (Tan et al. 2017), but an average COD of below 500 mg/L was also reported (Calderón-Vallejo et al. 2015). The COD levels in the septic and faecal sludge were generally proportional to the TS concentrations. However, although the TS concentrations of surplus-activated sludge were relatively low, the COD concentrations were usually high. For instance, Wang et al. (2022b) indicated that the TS concentrations of the surplus-activated sludge were between 7500 mg/L and 15,000 mg/L, but the COD concentrations ranged from 27,400 mg/L to 35,600 mg/L. Zhong et al. (2021) also observed the same trend, where the average COD and TS concentrations were 46,380 mg/L and 9424 mg/L, respectively. The concentrations of nutrients in sewage sludge, particularly TN and TP, were also relatively high compared to high-strength wastewater, but the concentrations varied from batch to batch. The highly variable concentrations of COD and nutrients make it challenging to ensure that leachate treatment in STRBs meets effluent standards.

Recent studies have highlighted that the treatment of heavy metal content in dewatered sludge is critical if the sludge was to be utilised for agricultural applications as compost, as high concentrations can lead to toxicity and be hazardous to the environment (Boruszko 2018). However, the selected studies did not report heavy metal contents in the raw septic and faecal sludge. Chandana and Rao (2022) indicated that heavy metal concentrations in septic and faecal sludge were usually below 1 ppm, and thus, the contamination caused by the heavy metals could be negligible in the management. However, a wide range of heavy metals was detected in the surplus-activated sludge in Table S2 of the online resource. The results among these studies are consistent, where most studies found a very high concentration of iron (> 40 mg/L), while manganese and zinc concentrations were lower (close to 1 mg/L). Other heavy metals remained at a low concentration in the raw surplus-activated sludge. In addition, the detection of emerging pollutants, including antibiotics (Ma et al. 2020, 2021) and polycyclic aromatic hydrocarbon (Cui et al. 2015), was also reported.

System and operating parameters

A conventional STRB consists of several key components: a reed bed composed of multi-layer granular porous media, macrophytes and a sludge distribution and drainage system. The STRB does not require pre-treatment other than a screen that removes large objects that can damage the sludge distribution devices. Some designs divide STRB into two treatment stages: the first stage dewaters the raw sludge, while the second stage treats the supernatant from the first stage treatment (Jong and Tang 2014a; Chen and Hu 2019; Kim et al. 2018; Andrade et al. 2017). The dewatering of sludge in STRB is through draining and evapotranspiration (Kołecka et al. 2016). In a conventional STRB, a distribution system is used to discharge the sludge to the top of the reed bed. During the loading period, the particulate matter in the influent sludge is physically retained on the top surface of the reed bed, producing a layer of sludge residue, also known as a sludge deposit or sludge cake (Tan et al. 2017). Then, the liquid component, which is also known as leachate, percolates through the reed bed and is released via the drainage system for disposal or further treatment (Nielsen and Larsen 2016). This liquid–solid separation through draining is the primary dewatering mechanism in STRB. Some systems recirculate the leachate to the reed bed to ensure the effluent quality is up to standard (Magri et al. 2016).

Typically, the STRB is fed intermittently in batches to enhance the sludge dewatering and stabilisation efficiency, where the loading rate and frequency are the most critical operational factors. The STRB operation by controlling the hydraulic loading rate (HLR) is the most straightforward, as the amount of sludge to be fed is directly measured by volume. However, such an operation regime might lead to a large variability of the resulting solids load in each loading cycle due to the varying sludge characteristics (Calderón-Vallejo et al. 2015). On the other hand, the operation of the STRB by solids loading rate (SLR) is more conventional, where the loading volume is determined based on the total solids concentration of the raw sludge. This approach has been proven to be a better operating regime for controlling the buildup of sludge residue and improving long-term performance (Tan et al. 2017). The climate is a crucial factor in the selection of SLR, where Gholipour et al. (2022) concluded that the average SLR in tropical climates could be up to 101 kg TS/m²yr, while it is only 70 kg TS/m²yr in the temperate region. However, measuring the solids content of the raw sludge to control the SLR is time-consuming, so the HLR operating regime is still a viable method when the sludge characteristics are relatively constant (Tan et al. 2020).

As mentioned, the STRB is loaded alternatively, and the dry period between two consecutive loads is known as the resting period. During the resting period, evaporation and transpiration release the residual liquid in the sludge residue on the top surface of the bed, which is another important dewatering mechanism in STRB (Brix 2017). Nielsen (2011) and Jong and Tang (2016) emphasised the importance of alternating loading and resting periods to reduce the risk of overloading and waterlogging in the bed. At the same time, the sludge residue also undergoes aerobic mineralisation that stabilises the volatile substances to improve the quality of the final sludge residue (Nielsen and Larsen 2016). The intermittent loading regime creates a saturated–unsaturated cycle that promotes an aerobic condition in the reed beds, which improves the removal efficiency of the organic matter in the leachate. Therefore, the duration of this dry period is a key factor in the rate of sludge residue accumulation, sludge dewatering by evapotranspiration, and the stabilisation of volatile substances in the sludge residue (Kołecka et al. 2016).

With regards to the system configuration, the reed bed, also known as substrate media, is a multi-layered granular media that acts as a filter to retain the particulate pollutants physically. This granular media is also an attached growth reactor that removes dissolved pollutants from the leachate through biological and chemical reactions. The substrate materials generally include gravel and sand (Bui et al. 2019), but materials such as biochar (Greenway et al. 2022; de Rozari et al. 2018), zeolite (Wang et al. 2022a), and palm kernel shell (Jong and Tang 2015) were also adopted to enhance the pollutant removal capacity of the system. The particle size of the granular material is crucial to the permeability and filtration efficiency of the STRB system, but substrate thickness does not significantly affect the dewatering and stabilisation performance (Uggetti et al. 2010). As mentioned, the filtration of particles produces a sludge residue layer at the top surface of the substrate, which reduces the permeability of the bed. The excessive accumulation may eventually lead to waterlogging (Tan et al. 2020). Accordingly, the sludge residue layer is usually considered part of the substrate as it governs the filtration efficiency and hydraulic retention time, and it is even referred to as an indicator of the bed’s acclimatisation status (Wang et al. 2022a). In general, sludge residue is removed when the depth causes bed clogging, where prolonged surface ponding is observed.

Numerous studies have highlighted the role of reeds in promoting transpiration to improve dewatering efficiency and alleviate clogging problems with the extensive root system in the STRB (Hu et al. 2021; Uggetti et al. 2010). A rhizosphere layer formed by the extensive growth of reed roots also stimulates air exchange in the substrate and increases the surface area for biofilm development (Nielsen and Larsen 2016). Although there have been many other macrophytes used in the study of STRB, the common reeds (Phragmites species) are the common option mainly due to their excellent tolerance to high-strength influent, micropollutants, and fluctuations in influent flow, as well as their capability of removing the pollutants with high efficiencies (Hu et al. 2021; Hu et al. 2020a, b). Furthermore, the installation of ventilation pipes through the bed vertically has become more common as it enhances the aerobic mineralisation of organic matter by creating efficient air exchange, ensuring a better quality of sludge residue and leachate in the STRB (Nielsen and Larsen 2016). Recently, several studies modified the conventional STRB by introducing earthworms (Hu et al. 2020a, b), active aeration (Plestenjak et al. 2021), electro-oxidation (Wang et al. 2021), and biological power generation (Zhong et al. 2021) to improve the system performance. Gholipour et al. (2022) referred to these studies as the state-of-the-art system configuration in STRBs. In this review article, these advanced features are not considered as there is insufficient research to determine their roles in sludge dewatering and stabilisation. Thus, it is difficult to include them as input parameters in the proposed modelling framework. A summary of the system configurations and operational parameters of the STRB systems among the selected studies is presented in Table S3 of the online resource.

Substrate media

A full-scale STRB system consists of several reed beds arranged in parallel to implement an alternating loading-resting cycle in each bed (Brix 2017). It was reported that the number of beds is usually at least eight, with some even up to 24 (Nielsen and Larsen 2016). The number of beds is subject to the duration required for the desired degree of sludge dewatering and stabilisation, where the properties of substrate media play vital roles in controlling the influent infiltration and retention time, filtering the particles, and acting as a reactor for leachate purification through physical and biochemical reactions (Jain et al. 2022; de Rozari et al. 2015). The selection of substrate media, including the materials, grain sizes, and depth, is a critical technical issue in STRB and is mainly based on the principle of attached growth systems for wastewater treatment. In general, the substrate media is a multi-layered granular media, and the upper layer of the substrate media is known as the filter or growth layer, where the macrophytes are found. The grain size of the medium in this layer is often small to enhance the filtration capacity and provide water retention to support the macrophyte growth (Nielsen and Larsen 2016; Troesch et al. 2009). Underneath the top layer, one to three layers with different sizes or materials (also known as the main substrate layer) are usually filled for the purpose of influent percolation and treatment. Then, a drainage layer is created at the bottom of the reed bed to maximise the outflow of the leachate from the bed (Kołecka et al. 2016).

Figure 3 shows the typical substrate materials and thicknesses in STRBs among the selected studies. Natural materials such as gravel and sand are the most common substrate materials in the selected studies since they are abundant, relatively low in cost, and promising in retaining pollutants. A total of 35% of the studies (18 out of 51) involved STRBS consisting only of gravel in the substrate and involved at least two substrate layers with different grain sizes, where the size distribution usually increases with depth to achieve a high degree of water permeability (Brix 2017). The only exception is Wang et al. (2022b), which employed a single-size substrate media. The gravel used in the selected STRB systems varied in grain size, ranging from 2 to 100 mm. Kołecka et al. (2016) explained that the dewatering rate of STRBs is dependent on the grain size of the substrate media, and the coarse-grained materials used resulted in large pore spaces and higher hydraulic conductivity, which is vital to ensure that sufficient dewatering takes place (Brix 2017; Stefanakis 2020). However, the pollutant removal could be limited by the low filtration capacity and low surface area of the coarse-grain substrate media in biofilm development and sorption, and the low water retention capability could also lead to plant wilting. In addition, according to Nielsen and Stefanakis (2020), heavy metals are likely to be bound to the gravel media, thus filtering them out of the drainage system with less than 16% of the final heavy metal mass leaving the bed.

Fig. 3

Materials and depth of substrate media in the selected studies (n = 51)

Sand has a smaller grain size than gravel, typically less than 2 mm in diameter, with characteristically low permeability (Wang et al. 2019). A total of 33% of the selected systems (17 out of 51) adopted a top sand layer on the media, which acts as the primary filter layer. There are insufficient data to conclude the effect of the sand layer on the sludge dewatering efficiency, but the advantage of the sand layer in improving pollutant removal due to better filtration capacity has been highlighted (Wang et al. 2019; Panuvatvanich et al. 2009). In addition, the sand layer acts as an interphase that separates the granular medium and the sludge residue layer and prevents the solids from entering the deeper substrate layer of the STRBs. It enables the restoration of the reed bed from clogging by simply replacing the sand layer (Uggetti et al. 2010).

A total of 14% of the selected systems (7 out of 51) employed industrial by-products or artificial materials such as slag, zeolite, ceramsite, graphite, palm kernel shell, biochar, and activated carbon in the substrate media of STRBs to improve the removal of organic matter, nitrogen, phosphorus, and heavy metals in the leachate (Greenway et al. 2022; Wang et al. 2022a; Zhong et al. 2021; Jong and Tang 2014a). The significant advantages of these materials include better sorption capacities due to the greater surface area, chemical reactivity, and cation exchange capacity (Jain et al. 2022). Although favourable results were obtained, the cost and local availability of these materials are important concerns (Wang et al. 2019), so these materials were mostly found in state-of-the-art systems with electro-oxidation (Wang et al. 2022a) and biological power generation (Zhong et al. 2021). In addition, the long-term performance of sludge dewatering and leachate purification, including the permeability and sorption capacity of the substrate media over the treatment durations, requires further studies.

Among the selected studies, most systems had substrate media depths greater than 0.5 m (23 out of 51), where the vast majority were between 0.60 and 0.69 m (10 out of 51). Although the substrate media with a depth between 0.30 and 0.49 m have been found in thirteen studies, the number of systems with a substrate media depth below 0.30 m has been relatively low. The depth of the substrate media is variable, and it has been reported to have a limited effect on the performance of STRBs (Uggetti et al. 2010). For example, a STRB system in Brazil has a 0.75 m substrate media (0.45 m large gravel layer, 0.20 m small gravel layer, and 0.10 m coarse sand layer from bottom to top) (Magri et al. 2016), and its removal efficiency of TS and COD were similar to the system in Palestine with a 0.40 m substrate media (0.15 m gravel layer (Ø 25–40 mm), 0.15 m gravel layer (Ø 10–25 mm), and 0.10 m fine sand layer from bottom to top) (Afifi et al. 2015) in treating the septic sludge with similar characteristics. Panuvatvanich et al. (2009) also indicated that the reed bed with a 0.20 m sand layer has a better nitrogen removal efficiency than a system with a 0.10 m sand layer, but there is no further improvement when the sand layer was increased to 0.40 m. Accordingly, a substrate media depth between 0.40 and 0.50 m should be sufficient to deliver promising sludge dewatering and leachate purification, while the selection of substrate materials and grain sizes is more critical to the system’s performance.

Macrophytes

The roles of macrophytes in STRBs are summarised as follows:

• The presence of the macrophytes provides a large surface area of the leaves that ensure a high transpiration rate that helps in sludge dewatering (Hu., Chen and Chen 2021; Rozkošný et al. 2020).

• The root penetration and stem movement due to wind create cracks in the sludge residue and substrate media, enlarging the openings for influent infiltration and improving the sludge dewatering via draining (Brix 2017).

• Macrophytes release oxygen to the sludge residue through the extensive root system, which creates an oxygen-rich layer that enhances aerobic mineralisation (Nielsen and Larsen 2016; Pandey and Jensen 2015; Uggetti et al. 2010).

• The rhizosphere of the macrophytes acts as a site for biofilm attachment for the leachate treatment (Bui et al. 2019).

• The macrophytes uptake the pollutants, such as nutrients and heavy metals, subsequently improving the leachate purification efficiency (El-Gendy and Ahmed 2020; Peruzzi et al. 2009).

Figure 4 shows the species of macrophytes used in the STRBs in the selected studies. The number of systems with common reeds (Phragmites, Phragmites australis, and Phragmites karka) outnumbered those with other macrophytes (28 out of 51). Common reeds have been widely used in the STRB systems regardless of climate types because of their excellent productivity, extensive root development, and wide availability in the local area. According to Bakhshoodeh et al. (2020), the root of the common reeds can easily penetrate up to 0.6–1.0 m deep and withstand a wide pH range (pH 4.8–8.2). Its adaptability to different climates is also an important feature as the macrophytes in STRBs. In addition, common reeds demonstrated excellent tolerance to high-strength influents, micropollutants, and fluctuations in influent flow (Hu et al. 2021). A previous study also highlighted the excellent transpiration rate of Phragmites australis compared to other macrophytes (Panuvatvanich et al. 2009). In addition, Jong and Tang (2014b) further commented that the continuous growth of roots and rhizomes in the sludge residue layer and the filter media, as well as the movement of the reed stems, prevent the bed from clogging and keep the bed under an aerobic condition. However, a study indicated that macrophytes played a minor role in maintaining the drainage rate of the reed beds during the loading period, as no significant difference was observed between the planted and unplanted systems (Kinsley et al. 2018). On the other hand, Hu et al. (2020b) revealed that the sludge dewatering and NH₄ removal efficiency in the STRB planted with Phragmites australis was approximately 90% higher than the systems with T. angustifolia, and such macrophytes also contributed to the effective removal of heavy metals (Peruzzi et al. 2009) and polyaromatic hydrocarbon (Cui et al. 2015). However, the cultivation of Phragmites australis in STRBs could be restricted in the areas where it was not registered as a native species (Mariñelarena et al. 2021).

Fig. 4

Species of macrophytes in the selected studies

Several species of macrophytes, including Canna indica, Echinochloa pyramidalis, and Cynodon dactylon Pers, are capable of withstanding in the STRBs and providing promising sludge dewatering and leachate purification. However, studies of the feasibility of these macrophytes in STRB are limited to a particular climate, as Gholipour et al. (2022) reported. For example, the systems with Canna indica were all located in the region under a humid continental monsoon climate (Wang et al. 2022a, 2022b; Zhong et al. 2021; Hu et al. 2017), while Echinochloa pyramidalis and Cynodon doctylon Pers were found in the regions under Equatorial Guinean climate (Nzouebet et al. 2022; Kengne et al. 2014; Sonko et al. 2014) and humid subtropical climate (Silva et al. 2019; Andrade et al. 2017; Calderón-Vallejo et al. 2015). The climate is one of the crucial factors in selecting macrophytes in the STRBs, where the growth of a particular type of macrophytes could be unfavourable under low temperatures and high humidity and may eventually wilt during the winter. Furthermore, macrophytes prefer higher humidity conditions, as the high water exchange rate within the plant is crucial for biomass growth (Tran et al. 2019). Therefore, further studies are required to investigate the performance of the macrophytes other than common reeds in the STRBs under different climate conditions.

Loading regime

In STRBs, the sewage sludge is fed intermittently in batches, where the loading rate is one of the most critical design parameters. The loading rate is crucial to the buildup of sludge residue and associated stabilisation; hence, it is an essential factor in determining the bed dimensions and resting period to ensure the desired performance of STRBs (Kołecka et al. 2018; Uggetti et al. 2010). Furthermore, Brix (2017) emphasised that the loading rate of STRB should be as high as possible without compromising sludge dewatering and stabilisation efficiency. This review article classified the loading regime as solid loading rate (SLR), hydraulic loading rate (HLR), or uncontrolled. Figure 5 shows the distribution of loading regimes in the selected studies. The analysis revealed that 39% of systems (20 out of 51) adopted HLR in operating the STRBs, whereas 28% (14 out of 51) used SLR as the loading regime. In addition, there is an increasing trend that the amount of sewage sludge fed on the STRBs was dependent on the sludge production at the source (Mariñelarena et al. 2021; Silva et al. 2019; Andrade et al. 2017).

Fig. 5

Loading strategy of selected studies (n = 51)

SLR is a conventional way to describe the treatment capacity of an STRB, which is given as kg of total solids (TS) or dry matter (dm) per square metre per year (kg TS or DM/m²/yr) (Brix 2017). Climatic conditions and sludge quality are the governing factors for selecting SLR (Gholipour et al. 2022; Bui et al. 2019). A hot and dry climate and raw sludge typically guarantee a faster dewatering rate and produce a relatively dry biosolids in the stabilisation process (Khomenko et al. 2019). Accordingly, the SLRs for the STRB systems in the regions with tropical, subtropical and arid climates were usually high, ranging from 100 to 350 kg TS/m²/yr (Nzoubet et al. 2022; Tan et al. 2017; Magri et al. 2016; Pandey and Jensen 2015; Afifi et al. 2015). However, excessive sludge loading jeopardises the sludge dewatering efficiency and leads to an operational problem of bed clogging (Tan et al. 2020). A study revealed that the high SLR significantly reduced the dewatering efficiency of STRB, as the average percentage of supernatant reduced from 66.56% to 22.82% when the SLR was increased from 100 to 350 kg TS/m²/yr (Tan et al. 2017). Several studies also observed a lower dewatering efficiency in the systems with SLR of 150 kg—200 kg TS/m²/yr (Magri et al. 2016; Pandey and Jensen 2015; Sonko et al. 2014). Accordingly, an extended resting period should be provided to ensure appropriate liquid lost through draining and evapotranspiration in the system with a high SLR to meet the targeted dewatering efficiency (Tan et al. 2020). In addition, Sonko et al. (2014) found that clogging is more frequent in beds with a high solid load in the sludge application, even though a longer resting period was provided. This finding further emphasises the importance of SLR in the dewatering efficiency of STRBs.

The SLR of the STRB systems located in the temperate regions with continental, Mediterranean and hemiboreal climates, ranged from 25 to 75 kg TS/m²/yr (Kowal et al. 2021; Meng et al. 2020; Masciandaro et al. 2017, 2015; Caicedo et al. 2015; Peruzzi et al. 2013). Many studies recommended that the SLR for temperate climate conditions should be at a maximum of 60 kg TS/m²/yr (Tayler 2018; Brix 2017; Nielsen and Larsen 2016), as evapotranspiration, photosynthesis, and microbial activity rates are significantly reduced under low temperatures, especially during the winter in temperate regions (Varma et al. 2021). However, it was also reported that a low loading rate could hinder the growth of macrophytes (Hu et al. 2017). Stefanakis (2020) determined that a SLR below 60 kg TS/m²/yr was too dry for a system under the Mediterranean climate, causing the common reeds in the system to wilt and decrease the evapotranspiration rate and overall dewatering efficiency.

In addition to sludge dewatering, SLR is also a key factor in the removal efficiency of pollutants in the leachate. An operational concern is that the optimum temperatures for the primary biological processes in the STRB are within the range of 15–30 °C. However, it is difficult to maintain the efficiency of leachate purification in most temperate countries due to the change in temperature over the year, so the SLRs should remain below 60 kg TS/m²/yr to ensure the desired treatment efficiency (Stefanakis 2020). On the other hand, it was also observed that the increasing SLR was capable of achieving excellent removal efficiency for TS, COD, and TN (Tan et al. 2017; Magri et al. 2016; Afifi et al. 2015). Further, a high solid load increases the accumulation rate of sludge residue on the reed bed, subsequently improving the system’s filtration capacity and treatment efficiency (Molle 2014). Based on the literature review, SLR ranges of 50–60 kg TS/m²/yr and 100–150 kg TS/m²/yr were widely referred to in evaluating the operation of STRBs in temperate and tropical regions, respectively.

Recent studies suggested that the HLR is a more appropriate loading regime than SLR in the design of STRBs due to its simplicity in operation (Tan et al. 2020) and significant impact on the drainage rate (Kinsley et al. 2018). The HLR is presented as the volume of sludge per square metre of reed bed per day (m³/m²/day) or per loading over the loading-resting period (m³/m²/loading), which is usually simplified as the depth of sewage sludge fed per day (cm/day or mm/day). Kinsley et al. (2018) observed that the variation in the drainage rate of STRBs depended on the HLR, while SLR has limited influence on such a dewatering mechanism. However, Tan et al. (2020) observed that the drainage rate did not significantly vary with the HLR, where the dewatering efficiency was affected by other factors, such as the accumulation rate of sludge residue and rest periods between the loadings. In fact, some studies assessed the adequacy of HLR in dewatering performance by estimating the corresponding solids load from the TS concentrations of the raw sewage sludge (Osei et al. 2022; Hu et al. 2020a, b). Accordingly, the effect of HLR and its interaction with other factors on the overall dewatering efficiency of STRB still requires further studies.

Many studies have discovered that a high HLR deteriorates the removal efficiency of organic matter and nutrients in the leachate by creating a faster percolation rate over the substrate media, which shortens the contact time of the pollutants with the attached biofilm (Karolinczak and Dabrowski 2017). However, a low HLR also influences the leachate purification performance as it causes water shortage in the reed bed and hinders the growth of macrophytes, which in turn affects the development of attached biofilm and pollutant uptake (Hu and Chen 2018). For the STRB systems located in tropical and semi-arid regions, the HLRs were mainly above 10 mm/day, with the highest reaching 57 mm/day (Osei et al. 2022, 2019; Tan et al. 2020; Bui et al. 2018; Kouawa et al. 2015). On the other hand, HLRs below 5 mm/day were reported for STRB systems in temperate regions (Hu et al. 2020a, b; Rozkošný et al. 2020; Chen and Hu 2019; Mennerich et al. 2017); Chen et al. 2016) to over 10 mm/day (Zhong et al. 2021; Plestenjak et al. 2021; Kim et al. 2018; de Rozari et al. 2016; Cui et al. 2015; Calderón-Vallejo et al. 2015), to even be as high as 100 mm/day (Karolinczak and Dabrowski 2017). In summary, the HLRs adopted in the selected studies varied widely, and only a limited number of studies proposed the design of HLRs for the STRB under particular climates (Kinsley et al. 2018).

There is also an increasing trend that the loading regime of STRBs was not controlled by either SLR or HLR, but rather depended on the amount of sewage sludge received from the source. Such a loading regime was usually observed in full-scale systems (Mariñelarena et al. 2021; Kołecka et al. 2018; Boruszko 2018; Masciandaro et al. 2015; Peruzzi et al. 2015; Nielsen 2014). In septic sludge management, different trucks were used to remove sludge from various origins, where highly variable volumes and characteristics in its composition were expected, restricting the control of hydraulic and solid loads in the operation of STRBs (Calderón-Vallejo et al. 2015). For example, the loading of a STRB system treating septic sludge depended on the sludge collected by the clean-pit truck, which resulted in a significant difference in HLR, ranging from 6.3 to 21.4 m³/m²/yr, corresponding to SLRs of 10.5–914.5 kg TS/ m²/yr (Andrade et al. 2017). The highly variable HLR and SLR were also observed in the STRB treating surplus-activated sludge (Mariñelarena et al. 2021; Masciandaro et al. 2015; Nielsen 2014). Although the average SLR of these systems were comparable to the recommended loading rate, uncontrolled loading poses a higher risk of overloading the reed bed, which can lead to insufficient sludge dewatering and stabilisation and poor leachate purification.

Resting period

The duration of the drying period between loading periods, the so-called resting time or resting period, is a key design parameter to ensure sufficient sludge dewatering and stabilisation in STRBs (Brix 2017). In general, sludge dewatering through drainage is observed at the beginning of the resting period, followed by further moisture content reduction through evapotranspiration (Iannelli et al. 2013). The solid loading rate is a prerequisite in determining the resting period of STRBs, and so climate conditions are critical factors for the design of loading frequency (Jain et al. 2022; Kim et al. 2018). During the resting period, the residual water content in the reed beds continuously drains out or is released via evapotranspiration, so the resting period is typically longer than the loading period (Uggetti et al. 2010). The formation of surface cracks in the sludge residue layer due to the continuous loss of water content is crucial in maintaining the dewatering capacity and enhancing the stabilisation of sludge at a deeper level (Tan et al. 2017; Nielsen and Larsen 2016; Uggetti et al. 2010). If the beds are overloaded and are without a sufficient resting period, waterlogging could lead to anaerobic conditions in the reed beds, subsequently reducing the efficiency of stabilising volatile organic matter in the sludge residue (Nielsen and Larsen 2016). Pandey and Jenssen (2015) further highlighted that a longer resting period reduced the volume of sludge residue in the reed beds, eventually extending the system’s lifetime and reducing the maintenance cost.

There are no standard design criteria for the resting period, but the duration of the resting period is usually longer than the loading period. Among the selected studies, only one system has a more extended loading period than a resting period. The loading regime of the STRB is based on the ease of the operation, where the bed was loaded for five days (weekdays) and rested for two days (weekend) to align with the working hours of the operators (Karolinczak and Dabrowski 2017). However, overloading and poor leachate quality were observed during the loading period due to insufficient rest time. To simplify the operation, approximately 30% of STRB systems in the selected studies (16 out of 51) adopted a weekly basis loading regime with a one-day loading period followed by a six-day resting period (Wang et al. 2022a & 2022b; Osei et al. 2019; Bui et al. 2018; Pandey and Jensen 2015; Jong and Tang 2014b; Sonko et al. 2014). However, the feasibility of this loading-resting regime has not yet been determined since the hydraulic performance of the STRBs still relied more on the loading rate, as low water recovery through drainage was still observed when high SLR was applied (Tan et al. 2017). In temperate regions, a longer resting period was usually implemented to enhance the sludge dewatering and stabilisation (Velychko and Dupliak 2021; Kim et al. 2018; Masciandaro et al. 2017; Larsen et al. 2017; Collard et al. 2017; Peruzzi et al. 2015;), where the longest loading-resting period can be up to one-day loading period followed by a 28-day resting period (Plestenjak et al. 2021). It is also a common practice to have a final resting period ranging from several months to year before the emptying of sludge residue to ensure adequate dewatering as well as stabilisation of volatile matter and heavy metals (Peruzzi et al. 2017; Magri et al. 2016; Pandey and Jensen 2015).

The climate is crucial to the dewatering of sludge residue during the resting period, as Kim et al. (2018) reported that the dry matter content may reach 30% with a 3-day loading period and 15-day loading-resting period during the summer, but the dry matter content only achieved 20% during the winter and rain events. In a full-scale STRB system, a 7–10 days resting period was implemented during the summer season, while the resting period was extended to 14–20 days during the autumn, winter, and spring (Peruzzi et al. 2017). The maturity of the reed bed also plays a vital role in selecting a resting period. Due to better evapotranspiration after the growing period of the macrophytes, a shorter resting period can be practised to enhance the treatment capacity (Kim et al. 2018; Kinsley et al. 2017). Hu et al. (2021) and Sonko et al. (2014) observed that a shorter resting period was beneficial to the growth of macrophytes in the reed beds, which in turn improved the dewatering efficiency due to the better evapotranspiration rate. Therefore, it is still challenging to determine the optimum resting period and loading frequency due to the complex processes in the STRBs, which are also controlled by factors such as the loading rate, climatic conditions, and growth of plants and biofilms in the STRBs.

Sludge residue layer

The application of sewage sludge leads to the accumulation of sludge residue on the reed bed as a result of solids sedimentation and filtration. Andrade et al. (2017) found that a SLR of approximately 80 kg TS/m².yr resulted in a rate of sludge residue accumulation of about 7 cm/yr in a system located in subtropical regions, while Calderón-Vallejo et al. (2015) showed an accumulation of 12 cm/yr with an average SLR of 75 kg TS/m².yr. On the other hand, Brix (2017) reported that a 10 cm/yr residue accumulation was obtained with SLRs of approximately 50–60 kg TS/m².yr under a temperate climate, which agreed with Nielsen et al. (2014). Kinsley et al. (2018) estimated an average rate of 5.7 cm of sludge accumulation per m of septic sludge applied, corresponding to 0.21 cm per kg TS/m². The difference in the sludge accumulation rate is attributed to the high evapotranspiration rate under the hot climate, which leads to significant water loss that reduces the depth of the sludge residue layer due to the shrinkage effect (Tan et al. 2020).

The sludge residue layer is crucial to the dewatering efficiency of STRB, as the continuous loading of sludge compacts the layer and reduces the effective void space, subsequently increasing the filtration capacity and deteriorating the permeability, so-called specific cake resistance (Christensen and Keiding 2012). The hydraulic properties of a sludge residue layer are highly dependent on the organic content, where the sludge residue with high VS content characterises low permeability and subsequently limits the dewatering efficiency (Khomenko et al. 2019). In other words, raw sewage sludge with high concentrations of fats and oils directly reduces the dewaterability of the sludge residue. In addition, this sludge residue is rich in nutrient content, especially nitrogen and phosphorus, which could lead to excessive biofilm growth and further reduce permeability (Pucher and Langergraber 2019). Therefore, the excessive sludge residue on the top surface of the reed beds due to high SLR is the root cause of bed clogging, which can be observed through prolonged surface ponding and a low drainage rate (Molle 2014). However, the sludge residue could be beneficial to leachate purification, as it not only increases the filtration capacity of the reed bed but also improves the removal of dissolved organic and nitrogen compounds. The slow leachate percolation rate due to the low permeability of the thick sludge residue layer extends the hydraulic retention time of the leachate in the bed, which could enhance interaction between the pollutants and the attached biofilm in the substrate medium to deliver a more complete treatment process (Varma et al. 2021; Bakhshoodeh et al. 2020).

The permeability of sludge residue could vary over the resting period due to the formation of shrinkage cracks and extensive root system of macrophytes, which creates preferential flow pathways for the influent and alter the drainage efficiency of the reed beds (Tan et al. 2017). The high volatile solid (VS) content in the influent results in more water-bound due to the capillary action (Bui et al. 2019). Accordingly, the continuous loss of capillary-bound water in the sludge residue through transpiration via macrophytes is crucial to dewatering, which eventually leads to deformation and shrinkage of the sludge residue when a sufficient resting period is provided (Brix 2017). The shrinkage of sludge residue usually causes cracks, allowing the influent to “bypass” the sludge residue layer and accelerate the drainage flow following feeding. Such a deformation in the sludge residue layer creates a “preferential flow pathway” (PFP) which has a positive impact on the drainage rate of the STRB system. However, these PFPs are undesirable for leachate purification, as the hydraulic conductivity increases drastically with the presence of PFPs, reducing the treating duration and the efficiency of sludge dewatering.

Based on the selected studies, the resting period is the primary factor that affects the stabilisation of sludge, which is usually measured as the reduction of VS content in the sludge residue. Therefore, the depth of the sludge residue was found as a factor of sludge stabilisation as the bottom layer experiences a longer resting period (Caicedo et al. 2015). Kołecka et al. (2017) observed a higher organic content in the top layer (71.9 ± 6.0%) compared to the bottom layer (52.4 ± 4.9%). In addition, the stabilisation efficiency was also found to be better when the sludge residue was deeper in the reed beds (Silva et al. 2019). Due to the varying degrees of stabilisation across the depth and bed maturity, many studies suggested a final resting period ranging from a few months to a year for complete stabilisation of persistent organic compounds and faecal bacteria indicators in the sludge residue (Peruzzi et al. 2017).

Besides, sludge dewatering has been found to directly affect the stabilisation performance. The current study has proven that the decomposition of these organic matters is at optimum under aerobic conditions, where the residual water content in the sludge residue decreases the mineralisation efficiency as it prevents atmospheric aeration (Plestenjak et al. 2021). Brix (2017) emphasised the importance of cracks in the sludge residue layer as their presence promotes oxygen diffusion and aerobic decomposition. In addition, the installation of vertical ventilation pipes is suggested to enhance the mineralisation of the sludge residue in the deeper layers, which has been observed with higher VS and organic matter removals (Meng et al. 2020).

The removal of total organic materials is generally up to 25% via sludge mineralisation in the STRBs (Nielsen and Larsen 2016). The STRB is also believed to be efficient in eliminating pharmaceutical compounds, hormones, and wastewater toxicity, where it can remove most of the heavy metals found in the influent sludge (Moreira and Dias 2020). Therefore, the final sludge residue of an STRB with low hazardous organic compounds and heavy metals allows it to be used in agricultural activities (Nielsen and Larsen 2016).

Discussion

Interactions of system and operating parameters

The design objective of STRBs is to deliver sufficient dewatering that transforms the sludge from liquid to sludge residue with minimum moisture content, as well as stabilises the volatile content in the sludge residue so it can be safely applied as organic fertiliser for agricultural purposes or disposed of without creating environmental problems (Uggetti et al. 2010). Sludge dewatering and stabilisation in STBRs are complex processes governed by many factors, including the properties of substrate media, loading regime (sludge loading rate and resting period), macrophytes, as well as raw sludge characteristics (solids content and influent quality), and climate conditions (Plestenjak et al. 2021). In order to formulate a mathematical model for STRBs, it is essential to assess the interactions between these factors in the dewatering and stabilisation processes.

Velychko and Dupliak (2021) divided the dewatering process into three stages: the formation of surface ponding on the reed beds, the leachate discharge with the solids retention in the reed beds, and the evapotranspiration in the surface pond and sludge residue layer. A water balance equation has been widely used in the literature in investigating sludge dewatering efficiency in STRBs:

$$\begin{aligned} {\text{Sewage}}\;{\text{sludge}}\;{\text{applied}}\;{\text{to}}\;{\text{reed}}\;{\text{beds}} + {\text{Precipitation}} = & \;{\text{Drainage}} + {\text{Evapotranspiration}} \\ + & {\text{Bound water}} + {\text{Ponding}} \\ \end{aligned}$$

(1)

Here, the sewage sludge applied to reed beds is measured as hydraulic loading rate (mm), precipitation is obtained from the nearest weather station (mm), drainage is measured from the volume outflow of leachate (mm), bound water is the residual water content in sludge residue (mm), and ponding is the depth of surface ponding (mm) (Kinsley et al. 2018). This equation shows that the dewatering efficiency, which is measured as the water lost through drainage and evapotranspiration, can be reduced by the surface ponding and water content remaining in the sludge layer as bound water. However, the direct monitoring of the evapotranspiration rate is always tricky due to the influence of macrophytes and weather, so the typical indicators of sludge dewatering include the volume of outflow leachate and the TS contents in the sludge residue layer (Tan et al. 2020; Hu et al. 2017).

Gravity drainage is the primary dewatering mechanism in which the solids remain on the bed surface while the liquid is lost as leachate by percolating through the sludge residue layer and substrate media. Then, evapotranspiration releases the residual liquid content in the sludge residue and the upper layer of the substrate medium. In general, the permeability of the substrate medium governs the leachate outflow, but it has no obvious effect on the evapotranspiration rate. However, it has been found that the sludge residue accumulation on the bed surface, which results in a much lower hydraulic conductivity, imposes a specific resistance that limits the water infiltration and drainage efficiency (Christensen and Keiding 2012). Besides, the physical filtration of solids in the substrate medium gradually fills the void spaces and decreases the porosity, subsequently reducing the permeability of the reed beds (Vymazal 2018). Therefore, a high solid load due to the high loading rate or high solid contents in the raw sewage sludge could decrease the permeability of the reed beds, resulting in low dewatering efficiency through drainage (Tan et al. 2017). In addition to the solids load, studies also indicated that the high VS content in raw sewage sludge increased the bound water in the sludge residue, which could also limit the sludge dewatering (Khomenko et al. 2019; Bui et al. 2019).

The hydraulic loading rate significantly affects the drainage rate of the reed beds (Kinsley et al. 2018). Since the STRB is usually loaded intermittently in batches, a higher load generally produces deeper surface ponding on the bed during the early loading phase, where faster leachate outflow is observed due to the larger hydraulic head difference. As the sludge continuously infiltrates, the head difference decreases, gradually slowing down the outflow rate (Molle et al. 2006). On the other hand, if the resting period is not long enough to dry out the prolonged surface ponding due to the high hydraulic load over the resting period, waterlogging in the reed beds could lead to a severe overflow problem (Brix 2017; Pandey and Jenssen 2015). As a result, the duration of the resting period is closely related to the hydraulic loading rate, where the resting period must be sufficient for the complete infiltration of sewage sludge loaded, so evapotranspiration can take place to release the residual water content in the sludge residue (Uggetti et al. 2012). The varying permeability of the sludge residue layer over the resting period due to the formation of shrinkage cracks could also lead to an arbitrary leachate outflow rate from the STRBs (Tan et al. 2020). A review article has emphasised the importance of macrophytes types and density in the evapotranspiration rate (Jain et al. 2022), and the presence of macrophytes also favours the drainage efficiency of the reed bed by creating preferential flow paths in the sludge residue layer through the stem movements and the extensive root system (Uggetti et al. 2010).

Sludge stabilisation is a combined process of mineralisation and humification of volatile organic matter in the sludge residue as a result of the biochemical reactions stimulated by microorganisms and macrophytes. Accordingly, the sludge characteristics, particularly the influent VS concentration, govern the sludge stabilisation efficiency in STRB (Uggetti et al. 2010). In addition, the solids loading rate governs the accumulation of sludge residue in the bed and determines the degree of stabilisation required (Kinsley et al. 2018). Peruzzi et al. (2015) identified the specific pathways of sludge stabilisation, where the mineralisation of fresh organic matter and stable organic matter was observed during the initial commissioning and operating phases of the STRBs, respectively. However, the humification process that stabilises the persistent organic compounds and faecal bacteria was only prevalent during the final resting period. Therefore, the rest period is also one of the crucial parameters for sludge stabilisation in STRBs. Raw sewage sludge is continuously loaded on the sludge residue formed by previous operations, so the degree of stabilisation also varies across the depth of the sludge residue, where the mineralisation of stable organic matter is usually observed in the bottom layer (Nielsen et al. 2014). An extended resting period ranging from a few months to a year is generally needed to ensure that the sludge residue is safe to be applied as biosolids or disposed of (Peruzzi et al. 2017). It is important to highlight that the duration of the resting period is closely related to the solids loading rate, as a longer period is needed for the sludge stabilisation when more solids are retained in the bed (Tan et al. 2017).

On the other hand, substrate media and hydraulic loading rate have no direct effect on sludge stabilisation. Nevertheless, the high water content of sludge residue may slow down the mineralisation of organic matter because it hinders oxygen diffusion and the occurrence of aerobic mineralisation. Accordingly, the permeability of substrate media and hydraulic load, which are the critical parameters for the drainage rate of the STRBs, indirectly affects the performance of sludge stabilisation. Moreover, the formation of cracks in the sludge residue layer due to the water loss during the resting period is vital to ensure the desirable mineralisation of organic matter by promoting aerobic mineralisation in the deeper layer (Brix 2017). The extensive root system of macrophytes across the sludge residue also creates an oxygen-rich condition that enhances sludge stabilisation (Nielsen and Larsen 2016).

The simultaneous purification of the leachate produced from the dewatering process is another important feature of STRB. Solids, COD, BOD, nitrogen compounds (including organic nitrogen, ammonium, and nitrate), phosphorus, and faecal bacteria are the main pollutants in the raw sewage sludge. As mentioned previously, the high solids load increases the accumulation rate of sludge residue, which enhances the reed beds’ filtration capacity and improves the solid removal rate in the leachate. Since the biological process is the primary removal pathway for dissolved pollutants such as COD and ammonium, a dependency between influent concentration and removal efficiency was also observed, where a higher pollutant load promotes biofilm growth in the substrate media, subsequently increasing the removal rate (Karolinczak and Dąbrowski 2017). The grain size of the granular medium determines the specific surface areas of the substrate, which is crucial to the biofilm development and sorption capacity, eventually governing the degree of leachate treatment.

The drainage rate controls the hydraulic retention time of the leachate in the bed and, ultimately, the treatment efficiency of the leachate. The extended hydraulic retention time has been found to promote more extensive interaction between the contaminants and the attached biofilm in the substrate medium, resulting in more complete pollutant removals (Varma et al. 2021). Karolinczak and Dąbrowski (2017) indicated that a higher HLR reduced the removal efficiency of TSS, BOD, and COD. In contrast, the high hydraulic load improved total nitrogen removal efficiency, as the high moisture content in the substrate media limited oxygen diffusion and favoured denitrification. The restoration of aerobic conditions in the reed beds mostly takes place during the resting period, and the formation of shrinkage cracks in the sludge residue due to the continuous water loss has been found to enhance oxygen diffusion. The macrophytes provide additional surface area for biofilm attachment and transfer oxygen into the substrate through the extensive root system, thereby facilitating the aerobic decomposition of pollutants (Bui et al. 2019). In addition, the macrophytes improve the leachate quality by absorbing the pollutants such as nutrients and heavy metals. A summary of the interactions between the system and operational parameters and sludge dewatering and stabilisation is presented in Table 1.

Table 1 Interactions between the system and operational parameters and sludge dewatering and stabilisation

A Framework for process-based modelling of STRB system

The main challenge of the process-based modelling of the STRB system is the highly nonlinear sludge dewatering and stabilisation process, where the sludge residue layer is usually regarded as a “black box” in the analysis (He et al. 2021). The existing sludge dewatering (Uggetti et al. 2012) and sludge mineralisation models (He et al. 2021) were still incapable of integrating the key factors in investigating the adequate loading rate and resting period. Process-based numerical models could be a better tool for analysing the dewatering and pollutant degradation mechanism in STRBs. This study proposed a modelling framework to numerically describe the processes of sludge dewatering, stabilisation, and leachate purification by considering the interactions between the crucial system and operating parameters shown in Table 1. Generally, the model should consist of several sub-models, including a porous media flow sub-model, an evapotranspiration sub-model, a reactive transport sub-model, a sludge cake sub-model, and a bio-kinetics model. Figure 6 shows the modelling framework developed based on the sub-models proposed.

Fig. 6

framework for process-based modelling of STRBs

Porous media flow sub-model

The porous medium flow sub-model shall characterise the water infiltration and percolation in the reed bed, which is crucial in predicting the dewatering efficiency through drainage. Typically, the sewage sludge is loaded in the STRB intermittently while the leachate is freely drained, and thus, the drainage flow rate is described by the passage time of percolates through the substrate media. During the loading period, temporary surface ponding occurs due to the low permeability of the sludge residue layer, and the leachate outflow reaches the maximum rate immediately after the discharge is observed due to the largest hydraulic head difference (Tan et al. 2020). Then, the continuous infiltration of the sludge reduces the head difference and slows down the leachate outflow rate, where the flow pattern is very similar to the VFCW treating sewage (Giraldi et al. 2010). Accordingly, Richard’s equation is a viable model for simulating the transient variably saturated flow in the STRB with intermittent loading, which has been proven to be effective in describing the hydrodynamics in the VFCW (Yuan et al. 2020). It is a partial differential equation that relates the water percolation in the unsaturated porous medium to the hydraulic head difference between two points, which includes the effect of gravity and water pressure, and the associated hydraulic conductivity. A general expression of a one-dimensional Richard’s equation is as follows (Richards 1931):

$$\frac{\partial \theta }{{\partial t}} = \frac{\partial }{\partial z}\left[ {K\left( h \right)\frac{\partial h}{{\partial z}}} \right] + \frac{\partial K\left( h \right)}{{\partial z}} - S$$

(2)

where θ is the volumetric water content, h is the hydraulic head, K(h) is the unsaturated hydraulic conductivity at a given hydraulic head, z is the vertical coordinate (cm), t denotes the time (s), and S is the sink term. Due to the nonlinear behaviour, a numerical method such as the finite difference or finite elements method is required to solve Richard’s equation. Several studies have found that the finite difference method is promising in solving Richard’s equation in simulating the hydraulic behaviour of VFCW up to two dimensions (Langergraber and Sˇimunek 2005; Giraldi et al. 2010; Pálfy et al. 2016). In addition, these existing models are capable of switching the upper boundary between the prescribed head-controlled and flux-controlled conditions to simulate the formation of surface ponding due to the intermittent loading mode.

The van Genuchten–Mualem (VGM) model (van Genuchten 1980) is a common approach in predicting the unsaturated hydraulic conductivity as a function of volumetric water content and properties of porous media, and it has been widely applied in the simulation of VFCW (Langergraber and Sˇimunek 2005; Giraldi et al. 2010; Pálfy et al. 2016). This model consists of two equations, in which the first equation estimates the volumetric water content based on the pressure head and soil properties, and the second equation determines the unsaturated hydraulic conductivity from the volumetric water content obtained in the first equation:

$$\theta \left( h \right) = \theta_{r} + \frac{{\theta_{s} - \theta_{r} }}{{\left[ {1 + \left| {\alpha h} \right|^{n} } \right]^{m} }}$$

(3)

$$K\left( h \right) = K_{s} \left[ {\frac{{\theta \left( h \right) - \theta_{r} }}{{\theta_{s} - \theta_{r} }}} \right]^{l} \left\{ {1 - \left[ {1 - \left( {\frac{{\theta \left( h \right) - \theta_{r} }}{{\theta_{s} - \theta_{r} }}} \right)^{\frac{1}{m}} } \right]^{m} } \right\}^{2}$$

(4)

where θ_r and θ_s are the residual and saturated water contents, respectively, α is the inverse of the air-entry pressure, n is a measure of the pore-size distribution and m = 1–1/n, K_s is the saturated hydraulic conductivity, and 1 is an empirical pore-connectivity parameter. In summary, θ_r, θ_s, α, n and m describe the water retention capacity of the porous media at a given hydraulic head. In addition, the permeability of the substrate media can also be reduced due to the retention of solids. The Kozeny–Carman equation was used in several models to estimate the actual hydraulic conductivity due to the reduction of pore spaces in the substrate of VFCW (Pucher and Langergraber 2019):

$$K = \frac{{K_{0} }}{{\left[ {\left( {1 + \rho \frac{{D_{{{\text{vtot}}}} }}{{\varepsilon_{0} }}} \right)^{x} \left( {1 - \frac{{D_{{{\text{vtot}}}} }}{{\varepsilon_{0} }}} \right)^{y} } \right]}}$$

(5)

where D_vtot is the total volumetric specific deposit in the pore spaces, K₀ and ε₀ are the hydraulic conductivity and porosity of the clean porous media, respectively. ρ, x and y are empirical parameters.

In STRBs, a significant fraction of stagnant water is observed in the sludge residue layer caused by the irregular pore size and organic content, and Richard’s equation could overestimate the water percolation rate as it assumes an equilibrium flow. A dual-porosity model that divides the porous media into mobile and immobile regions was used in the flow model to improve the accuracy of the simulation. However, the additional parameters for the porous medium properties in the dual-porosity model increase the complexity of the model calibration and validation.

4.4 Sludge residue sub-model

The accumulation of sludge residue can be up to 0.21 cm per kg TS/m² (Kinsley et al. 2018). The sludge residue layer acted as a filter cake to trap the influent solids, which significantly reduced the permeability of the reed bed and decelerated the gravity drainage flow (Bui et al. 2018; Tan et al. 2020). Accordingly, a reliable simulation of the buildup of sludge residue is important to investigate the dewatering dynamics in STRB systems. However, no existing model integrates the sludge residue accumulation in the simulation of hydraulic behaviour in the reed bed. There are two challenges in simulating the buildup of sludge residue: estimating sludge residue thickness and its associated permeability when the system is fed. In the one-dimensional model developed by Tan (2016), an equation proposed by Abboud (1993), which assumed that the inner structure of newly formed and existing sludge reside are homogenous and incompressible, was adopted to predict the thickness of sludge residue formed during the feeding period as a function of hydraulic loading rate, solids concentration, and filtration capacity:

$$\rho_{s} \left( {1 - \overline{n}} \right)\frac{{{\text{d}}L}}{{{\text{d}}t}} = \sigma \rho_{s} \frac{{q_{p} }}{{n_{s} }} - K_{y} \rho_{s} \left( {1 - \overline{n}} \right)L$$

(6)

where L is the thickness of the newly formed sludge residue layer, ρ_s is the density of solids particles, $\overline{n }$ is the average porosity of the sludge residue layer, σ is the retention parameter, q_p is the input flux rate of raw sludge, K_y is the erosion rate, and n_s is the porosity of influent sludge. The n_s is estimated from the density of water (ρ_w) and solid particles (ρ_s) and the associated mass fraction of particulates in the raw sewage sludge s:

$$n_{s} = \frac{1}{{1 + \left( {\frac{s}{1 - s} \frac{{\rho_{w} }}{{\rho_{s} }}} \right)}}$$

(7)

The existing model described the system as a fixed domain with a uniform mesh. The current approach considered the increment of sludge residue thickness obtained from the equation to reform the mesh of the top boundary in the simulation to impose a flow resistance. This method lacks robustness when the loading rate and solids concentration are high, as studies have indicated that the thickness of the newly formed sludge deposit was comparable to the existing sludge residue layer, as shown in Fig. 7. A moving mesh method can be a feasible solution to the problem. This method has been found to be adequate for a range of one-dimensional moving boundary problems, including the finite difference Richards Equation model (Lee et al. 2015). However, further studies are required to integrate the sludge residue accumulation into the flow sub-models by taking into account the variations in the overall depth of the reed bed.

Fig. 7

Variation in top boundary due to accumulation of sludge residue

Moreover, the assumption of filtration capacity and hydraulic properties of the sludge residue layer remaining constant over the loading period was found to be inadequate when the solids load is high, as the particles might fill up the pore spaces in the sludge residue and increase the specific resistance, subsequently deteriorating the dewatering capacity of STRBs (Dominiak et al. 2011). Several studies experimentally proved that the permeability of the sludge residue layer was significantly reduced due to the compression, eventually becoming impermeable at high concentrations (Jadhav et al. 2019). Therefore, a compressible cake filtration model (Christensen et al. 2010) is a viable alternative to predict the accumulation of sludge residue in STRBs. However, this model consists of several equations that assume the sludge residue as a compressible layer that traps incoming solids above the filter medium, in which the variation of filtration capacity under the feeding condition can be promisingly predicted to improve its accuracy. This model is fundamentally based on gravity drainage theory (Severin and Grethlein 1996), which considers the compressibility of organic compounds, the settling rate of particles in influent, and the changes in specific resistance due to the buildup of sludge residue. Nevertheless, further studies are still required to include the variations in permeability due to the formation of shrinkage cracks in the simulation of hydraulic behaviour, where the plastic limit of the sludge residue could be a reference to predict the occurrence of cracks (Vincent et al. 2012).

Evapotranspiration sub-model

There are several well-established models for the evapotranspiration rate. The Penman–Monteith equation has been widely used to estimate the evapotranspiration rate in STRBs during the resting period (Velychko and Dupliak 2021; Uggetti et al. 2012). This equation combines the effect of both radiation and aerodynamics to simulate the volume of water loss to the atmosphere within a specific time, which provides a simple option to estimate the evapotranspiration rate based on fundamental weather data such as wind speed, temperature, and vapour pressure. A Penman–Monteith equation is presented as follows (Velychko and Dupliak 2021):

$${\text{ET}}_{c} = K_{c} {\text{ET}}_{0} = K_{C} \frac{{0.408\Delta \left( {R_{n} - G} \right) + \gamma \frac{{C_{n} }}{T + 273}u_{2} \left( {e_{s} - e_{a} } \right)}}{{\Delta + \gamma \left( {1 + C_{d} u_{2} } \right)}}$$

(8)

where ET_c is the crop evapotranspiration, K_c is the coefficient, ET₀ is the reference evapotranspiration, 0.408 is the unit conversion coefficient for the heat flux, ∆ is the slope of the vapour pressure curve, R_n is the soil heat flux, G is the soil heat flux, γ is the psychrometric constant, T is the mean daily temperature at the height of 2 m, u₂ is the average wind speed at the height of 2 m, e_s is the saturation vapour pressure at the given temperature, and e_a is the actual vapour pressure. C_n and C_d are the coefficients for the macrophytes, where the C_n and C_d are 900 and 0.34 for short macrophytes between 0.12–0.5 m height, while the values are 1600 and 0.38 for the macrophytes above 0.50 m height. The calculated evapotranspiration rate can be integrated into the flow model as the sink term to investigate the reduction of volumetric water content in the sludge residue over the loading-resting cycle.

However, the effect of evapotranspiration is not only observed at the bed surface. A fraction of the water content in a certain depth of the sludge residue can be extracted by the root of macrophytes and lost to the atmosphere, resulting in a varying dryness across the sludge residue profile. In a simulation study of VFCW, Giraldi et al. (2010) adopted a formula to predict the transpiration rate in a deeper layer based on the distribution of root density and hydraulic head, as well as the hydraulic conductivity and atmospheric pressure (Varado et al. 2006). Although this transpiration model provides a more comprehensive simulation in the study of dewatering across the reed beds, the calibration and validation are challenging because direct measurements of transpiration rates and water content are often expensive and time-consuming.

Reactive transport sub-model

The removal mechanisms of dissolved and particulate pollutants in STRBs depend on the interaction between the contaminants and substrate media. Therefore, modelling pollutant transport in the substrate media is crucial to assess the leachate quality, which has been comprehensively studied in many models developed for VFCWs. The advection–dispersion equation is the most common transport model used in describing pollutant transport (Bresler 1973):

$$\theta \frac{\partial c}{{\partial t}} = \frac{\partial }{\partial z}\left[ {D_{T} \frac{\partial c}{{\partial z}} - qc} \right]$$

(9)

where $\theta$ is the volumetric water content in the pore spaces of substrate media, c is the solute concentration in aqueous, t is the time, z is the vertical coordinate, q is the unit flux rate, and D_T is the hydraulic dispersion coefficient. It is important to highlight that the volumetric water content and flux rate are obtained from the porous medium flow sub-model. The D_T is expressed as follows:

$${\mathrm{D}}_{\mathrm{T}}= {\mathrm{D}}_{\mathrm{L}}\left|\mathrm{q}\right|+\uptheta {\mathrm{D}}_{\mathrm{D}}$$

(10)

where D_L is the empirical longitudinal dispersivity based on porous media properties, and D_D is the molecular diffusion coefficient that depends on the properties of pollutants.

Generally, the transport of dissolved pollutants in the liquid phase is the main focus of the transport sub-model. However, there are also existing models that consider the transport of particulate pollutants, as well as the exchange between solid and gaseous phases, to describe the reactions between the pollutants and substrate media, including adsorption–desorption, filtration and oxygen diffusion (Petitjean et al. 2012; Giraldi et al. 2010). The simulation of pollutant adsorption in the porous media can be found in several well-established models, where the associated reduction of pollutant concentration is described as a sink term in the transport model (Hua et al. 2018). On the other hand, Giraldi et al. (2010) proposed a transport model for particulate pollutants based on the advection–dispersion equation by considering the filtration and excluding the dispersion effect as follows:

$$\theta \frac{\partial s}{{\partial t}} = - q\frac{\partial s}{{\partial z}} - qfs$$

(11)

where s is the concentration of particulate pollutants and f is the filter coefficient. The transport of particulate pollutants is crucial in the simulation of STRBs in order to assess the deterioration of drainage dewatering due to the reed bed clogging. Petitjean et al. (2012) proposed a flow model that considered both liquid and gaseous phases to simulate the oxygen transport in the substrate media. However, the calibration and validation of the model were complex, and it is only required when oxygen is a limiting factor in the STRBs that limit the sludge stabilisation and leachate purification.

Bio-kinetics sub-model

The primary degradation mechanism of organic and nitrogen compounds in STRBs is through microbial processes. In the numerical model, the fate of the pollutants is determined by a set of mass balance equations, which include multiple reaction processes that either consume or produce the pollutant under specific conditions (Tan et al. 2021). The bio-kinetics sub-model is usually integrated with the transport sub-model to simulate the transport and fate of pollutants in the substrate media at the same time, and thus, the leachate quality can be assessed based on the discharge standards. Most bio-kinetic sub-models for VFCWs were developed based on the Activated Sludge Model (ASM) (Henze et al. 2006), which simultaneously predicts the fate of different pollutants with the presence of microbes to predict the degradation through several biochemical processes, including hydrolysis, aerobic decomposition of organic compounds, and nitrification. CW2D (Langergraber and Sˇimunek 2005) was regarded as one of the most advanced models considering the microbial degradation of organic, nitrogen, and phosphorus compounds in VFCWs under aerobic conditions. In addition to the leachate purification, ASM is also implemented to model the sludge stabilisation process in STRB. He et al. (2021) modified the ASM1 and ASM3 to simulate the degradation of volatile organic matter based on two cell decay theories, namely death-regeneration and endogenous respiration. The model delivered promising results for the stabilisation process, and it can also be used as a tool to study the roles of several parameters in sludge stabilisation, including the loading rate, moisture content, and influent concentrations. However, further studies are required for the dependency of sludge stabilisation on the dewatering efficiency, especially the formation of shrinkage cracks which could promote oxygen diffusion and enhance the stabilisation of volatile organic matter.

Conclusion

STRBs have great potential as a decentralised system for managing sludge removed from onsite sanitation systems and wastewater treatment plants, which can deliver promising dewatering and stabilisation efficiency compared to the mechanical systems while positively impacting environmental and economic aspects. According to recent publications over the past decade, the main conclusions are as follows: (1) A multiple-layer substrate constructed by sand and gravel with a thickness between 0.5 to 0.8 m is widely used in STRBs, and the grain size of the substrate materials should be selected to balance hydraulic conductivity, filtration efficiency, and hydraulic retention time. (2) Common reeds (Phragmites species) is the only macrophyte enhancing dewatering and contaminant removals and demonstrating tolerance to high-strength pollutants and fluctuations in the influent flow for STRBs under different climate conditions. (3) The loading regime based on hydraulic load rate (HLR) is gaining popularity over conventional solid load rate (SLR) in STRBs because of its ease of operation and significant impact on drainage rate and macrophytes growth. However, the HLRs adopted in the selected studies varied widely, and the current review only observes that a high HLR deteriorates the leachate quality due to the shorter hydraulic retention time in the substrate media, while the influence on drainage rate still requires further study. (4) SLR ranges of 50–60 kg TS/m²/yr and 100–150 kg TS/m²/yr were widely used to operate STRBs in temperate and tropical regions, respectively, as the temperature is a factor in evapotranspiration and microbial activity rates. (5) The loading rate is a prerequisite in determining the resting period of STRBs, where the resting period must be sufficient for the complete infiltration of sludge loaded so evapotranspiration can take place to release the residual water content in the sludge residue. A weekly loading regime with a one-day loading period followed by a six-day resting period was a common loading regime for STRBs to simplify the operation, but a longer resting period was usually needed in temperate regions to ensure a complete sludge dewatering and stabilisation. (6) The accumulation of the sludge residue layer imposes a specific cake resistance on the top surface of the reed bed that increases the filtration capacity while deteriorating the permeability. (7) The characteristics of the sludge residue layer vary over the resting period due to the formation of shrinkage cracks, which enhances the sludge dewatering through gravity drainage by creating preferential flow pathways and promotes stabilisation of volatile solids due to the extensive oxygen diffusion.

This study proposes a modelling framework that comprise five sub-models to develop a process-based model, which can be a useful platform to study the interactions of operating parameters and other factors in the complex processes of sludge dewatering and stabilisation in STRBs and support the system design and optimisation. For the sludge dewatering, the porous medium flow sub-model is aimed to simulate the gravity drainage flow from the reed beds. A sludge cake model is also needed to estimate the buildup rate of sludge residue due to the solid loads that affect the hydraulic performance, while integrating with the evapotranspiration model to describe the dewatering of sludge residue through the evapotranspiration process. In addition, the combination of the transport and bio-kinetics model simulates the transport and fate of pollutants in the substrate media, and thus, the leachate quality can be predicted. The bio-kinetics model can also be implemented to describe the stabilisation of organic matter in the sludge residue. These sub-models can adopt common models such as the Richard equation, the Penman–Monteith equation, the compressible cake filtration model, the advection–diffusion equation, and the activated sludge model (ASM).