Study on physical clogging process and practical application of horizontal subsurface flow constructed wetland

Abstract

In view of the urgent concerns pertaining to the proliferation of rural wastewater discharges and the imperative for decentralized treatment, this article examines the phenomenon of physical clogging in a small-scale horizontal submerged flow artificial wetland system. Through a combination of experimental analysis and CFD-EDM numerical simulations, the impact of clogging on the hydraulic efficiency of the system was subjected to rigorous examination. Based on these findings, an innovative design strategy was proposed, namely the addition of vertical baffles. The experimental results demonstrate that this strategy can markedly prolong the duration of complete clogging of the system by up to 15% and enhance the hydraulic efficiency by 21%. Based on this, a small-scale horizontal submerged artificial wetland wastewater treatment system was designed for rural areas and successfully implemented in Fanrong Village, Rui’an City. This resulted in an enhanced wastewater treatment effect and an improved rural landscape. The findings of this study contribute to the development of a more rational and effective sewage treatment solution for rural areas.

Introduction

Decentralized sewage treatment technologies, such as small courtyard-type constructed wetlands, have emerged as the preferred option for rural areas. The integration of constructed wetlands into landscape architecture design is a key area of focus for current research. Constructed wetland technology is an engineered system that mimics the natural construction of wetlands for sewage treatment¹, with subsurface flow wetlands being particularly favored in regions with limited land resources due to their benefits². This technology has been utilized in landscape construction for nearly a century, initially emphasizing landscape design³. As sewage discharge continues to rise, constructed wetlands, serving as a technology that combines sewage treatment and landscape construction, are gaining increasing importance. Yu has highlighted its ecological soundness and water-saving properties⁴. Various studies have identified wetland plants with high adaptability and ecological benefits, analyzing their overall advantages⁵. Courtyard constructed wetlands utilize substrates, plants, and microorganisms to purify sewage treatment, offering cost-effectiveness and efficient purification^6,7. This technology has been widely implemented for rural sewage treatment globally, including systems such as the purification tank in Japan⁸ and the landscape wetland system in Nepal⁹. Regions like Zhuhai and Hainan in China have also adopted this technology for domestic sewage treatment, achieving water quality standards, promoting water recycling, and easing water resource pressures¹⁰.

Constructed wetland sewage technology boasts advantages such as easy maintenance¹¹, low energy consumption, and minimal investment. However, the issue of clogging significantly¹² impacts its decontamination capabilities and service life, leading to disruptions in water flow¹³ and reduced purification efficiency¹⁴. Physical blockages within the matrix pose a particularly challenging problem^15,16. These blockages not only compromise treatment effectiveness and longevity but also escalate maintenance costs^17,18. Therefore, exploring methods to alleviate blockages is imperative. Decentralized sewage treatment technologies take into account local hydrological ecology and enable onsite sewage treatment. This approach offers low construction costs and short cycles, making the integration with constructed wetlands a promising new direction for rural domestic sewage treatment.

Currently, constructed wetlands are categorized into surface flow constructed wetlands, horizontal subsurface flow constructed wetlands, and vertical subsurface flow constructed wetlands based on the direction of sewage flow^19,20,21,22. In horizontal subsurface flow wetlands, sewage enters the substrate through the inlet end and exits after interacting with plants and microorganisms. In the treatment of wastewater, the horizontal submerged flow wetland is limited in its ability to treat high concentrations of pollutants compared to other types of wetlands, and long-term operation is likely to lead to substrate clogging and permeability degradation, which in turn reduces the hydraulic load and purification effect. In addition, it is necessary to balance hydraulic loading and treatment efficiency in operation and maintenance management to ensure stable operation of the system. These challenges require targeted measures during design, operation and maintenance to fully exploit the benefits of horizontal submerged flow constructed wetlands.

Constructed wetland clogging can be categorized into physical, chemical, and biological clogging^23,24,25. The clogging process is intricate and is influenced by various factors such as the quality of sewage water²⁶, wetland matrix composition²⁷, wetland vegetation²⁸, environmental conditions²⁹, wetland design, and operational methods³⁰. The primary cause of physical blockage in constructed wetlands is the accumulation of suspended solids in the matrix pores³¹. This accumulation occurs in two ways based on particle size: either through adsorption on the surface or direct occupation of pores. Larger particles directly occupy the pores, while suspended solids smaller than 100 μm adhere to the matrix surface. These different accumulation modes have varying effects on the hydrodynamic characteristics of the wetland, with direct accumulation significantly impacting hydraulic conductivity. Research indicates that suspended particles larger than 50 μm are prone to causing wetland blockages, leading to reduced matrix porosity and increased interception of suspended solids³².

Research into the role of baffles in reducing clogging in constructed wetlands has received considerable academic attention. However, there are still shortcomings. Wu³³ designed an artificial wetland system with baffles to alleviate clogging, but did not conduct experiments on the alleviating capacity; Xu³⁴ found that the addition of baffles could improve the pollutant removal rate, extend the hydraulic residence time and buffer the hydraulic load fluctuation by actual measurement, but the actual measurement was easily disturbed by environmental conditions; Wang³⁵ used numerical simulation to study the influence of the number of baffles on the hydraulic efficiency of wetlands in artificial wetlands, but did not study the baffle setting position and angle.

This thesis employs the experimental system of horizontal subsurface flow wetlands as its research object, utilising numerical simulation software to simulate the hydrodynamic behaviour of the physical clogging process of the wetland. On this basis, a spatio-temporal distribution model within the substrate was constructed to describe the physical clogging process of the wetland. Based on this, the optimal position and angle of the baffle plate were further explored with the aim of optimising its structure. A small-scale wetland system for rural domestic wastewater treatment was designed and applied in Prosperity Village, Rui’an City. Its practicality and aesthetics were enhanced by combining it with the garden design.

Methods

Experimental study on hydraulic behavior of physical clogging process in horizontal subsurface flow constructed wetland

Construction of experimental

The distribution and collection of water in constructed wetlands can be challenging due to a small aspect ratio, leading to uneven water distribution and non-uniform water flow within the matrix. Research suggests that maintaining an aspect ratio below 3:1 is optimal for effective water management³⁶. In accordance with constructed wetland design specifications, the experimental system for horizontal subsurface flow wetlands has been configured with a length–width ratio of 3:1, with specific dimensions of 0.6 m in length, 0.2 m in width, and 0.25 m in height.

Wetland substrates play a crucial role in precipitating, filtering, and adsorbing pollutants. Material selection should prioritize local availability, cost-effectiveness, and ease of procurement, such as gravel and ceramsite. The design of wetlands should be tailored to the specific wetland type, as different substrates exhibit varying pollutant removal efficiencies, as outlined in Table 1. Therefore, adaptation to local conditions is essential. Utilizing high-quality substrates enhances water flow, plant stability, and microbial survival³⁷. In this experiment, a single matrix of 8–10 mm white gravel was chosen to eliminate extraneous variables and ensure consistency within the experimental system.

Table 1 Filler selection for the removal of different contaminants.

Based on the specifications of the experimental system, the experimental model has a length–width ratio of 3:1 and utilizes gravel with a particle size ranging from 8 to 10 mm as the matrix for the experimental medium system. By referencing the 'artificial wetland design specification’ and considering the specific experimental conditions, the design parameters for the experimental system were established.

The horizontal subsurface flow constructed wetland experimental system device, as illustrated in Fig. 1, was constructed using a colorless transparent acrylic plate with a thickness of 5 mm. The dimensions of the experimental system are 0.6 m in length, 0.2 m in width, and 0.25 m in height. The system is partitioned into a water distribution system and a medium system, separated by a perforated plate³⁸. The medium system is filled with white gravel as the matrix, featuring a particle size of 8–10 mm, a porosity of 0.42, and a matrix height of 0.2 m. The experimental water flows through the peristaltic pump, wetland experimental system (WES), water collection tank, and exits from the top outlet on the right side, maintaining a constant head of 0.19 m from the bottom.

Fig. 1

Experimental system device.

In order to ensure the stable operation of the horizontal subsurface flow constructed wetland experimental system (WES), it is placed in the laboratory and the laboratory water is used as the experimental water source. Due to the complexity and uncertainty of the effects of microorganisms and plant roots, this study focused on the physical clogging caused by suspended solids. The experimental device does not add organic matter, nutrient elements, and does not plant plants. Figure 2 shows the operation process of the experimental system.

Fig. 2

Physical diagram of experimental system.

The functions of the instruments required for the corresponding detection of each index in the experiment are shown in Table 2.

Table 2 Instrument name and function.

Experimental method

Prior to conducting the NaCl tracer method, it is necessary to increase the inlet and outlet water flow of the experimental system to cleanse the experimental system matrix. NaCl was chosen as the tracer due to its suitability for small wetland research and ease of operation. The water conductivity in the experimental system was assessed multiple times, and the average value was considered as the background conductivity. A solution containing 10 g/L NaCl and a suspension solution (utilizing zeolite powder to mimic suspension at a concentration of 800 mg/L) were employed. NaCl was injected in pulses from the wetland inlet, with the inlet conductivity being measured every minute until stabilization. Following the removal of any NaCl residue through device washing, the suspension solution was introduced to simulate physical blockage, and the aforementioned steps were repeated with conductivity values being recorded (refer to Appendix A1). Subsequently, NaCl pulse experiments were conducted to measure conductivity at the onset of Total Suspended Solids (TSS) experiments every three days.

The TSS concentration determination experiment was then executed. To mitigate influencing factors such as biofilm, zeolite powder was utilized to create a suspended solids solution (at a concentration of 800 mg/L), and the experiment spanned 30 days. The water’s TSS concentration was gauged every two days (refer to Table A2). The determination method adhered to the 'Determination of suspended solids in water Gravimetric method GB11901 89’, with each sample being measured thrice for accuracy.

Evaluation indicators

The calculation formula of TSS concentration is as follows:

$$C=\frac{\left(A-B\right)\times {10}^{6}}{v}$$

(1)

In the formula, C (mg / L) is the concentration of suspended solids in water; A (g) is the total weight of suspended solids, filter membrane and weighing bottle after filtration; B (g) is the weight of filter membrane and weighing bottle; v (ml.) is the sample volume. The corresponding TSS removal rate was obtained according to the TSS concentration in the influent and effluent.

According to the fluid reactor theory, the concentration obtained by the tracer experiment is equivalent to the hydraulic retention time distribution density. The measured conductivity was standardized by formula (1)³⁹.

$$N\left(t\right)=\frac{\left(E-{E}_{w}\right){M}_{Nacl}Q}{\left({\lambda }_{Na}+{\lambda }_{Cl}\right)M}$$

(2)

In the equation, N(t) (%/h-1) is the normalized hydraulic retention time distribution density; t (h) is the time of tracer injection; E (S/m) is the conductivity; Ew (S/m) is the background conductivity; Q (m³/h) is the flow rate; MNaCl (g/mol) is the molar mass of NaCl; λNa is the molar conductivity of Na + ions; λCl is the molar conductivity of Cl-; m (g) is the total amount of tracer added.

The short-circuit value (s) refers to the ratio of the hydraulic retention time of the outlet tracer recovery of 16% to the recovery of 50%⁴⁰. Studies have shown that the smaller the s, the slower the peak of the hydraulic retention time distribution density curve. The larger the s, the steeper the peak of the hydraulic retention time distribution density curve. The calculation formula of s is as follows:

$$s={t}_{16}/{t}_{50}$$

(3)

In the formula, t₁₆ (h) is the hydraulic retention time of 16% of the export tracer recovery; t₅₀ (h) is the hydraulic retention time of 50% of the export tracer recovery.

The effective volume ratio (e) refers to the ratio of actual hydraulic retention time to theoretical hydraulic retention time⁴¹. When e > 1, it indicates that the water flow in the constructed wetland reaches the outlet directly and quickly, that is, the short circuit phenomenon; when e < 1, it indicates that there is a dead zone in the constructed wetland⁴². The calculation method is as follows:

$${T}_{n}=V/Q$$

(4)

$${T}_{m}={\int }_{0}^{\infty }tN\left(t\right)dt/{\int }_{0}^{\infty }N\left(t\right)dt$$

(5)

$$e={V}_{eff}/{V}_{total}={T}_{m}/{T}_{n}$$

(6)

In the formula, T_n (h) is the theoretical hydraulic retention time, which is defined as the ratio of constructed wetland volume to flow rate ; v (m³) is the volume of constructed wetland; T_m (h) is the actual hydraulic retention time, which is defined as the center of gravity position of the residence time distribution curve ; V_eff (m³) is the effective volume of the tracer, that is, the effective space in which the tracer moves and can eventually migrate to the outlet; V_total (m³) is the total volume, which is the product of the volume of the constructed wetland and the porosity of the medium.

Hydraulic efficiency (λ) is defined as the ratio of apparent hydraulic retention time (HRT) to the time required to reach the maximum tracer concentration at the wetland outlet⁴³. The calculation method is as follows:

$$N={T}_{m}/\left({T}_{m}-{T}_{p}\right)$$

(7)

$$\lambda =e\left(1-1/N\right)$$

(8)

$${\sigma }^{2}={\int }_{0}^{\infty }{\left(t-{T}_{m}\right)}^{2}N\left(t\right)dt/{\int }_{0}^{\infty }N\left(t\right)dt$$

(9)

$${\sigma }_{\theta }^{2}={\sigma }^{2}/{T}_{n}^{2}$$

(10)

In the formula, N is the number of reactor units. When N approaches 1, it means that the water flow is fully mixed flow. When N approaches ∞, it means that the water flow is piston flow. T_p (h) is the peak residence time on the RTD density curve; σ² (h²) is the variance, which is used to characterize the dispersion range of the response curve of the tracer concentration relative to the distribution average. σθ² is the standard deviation, which is used to characterize the hydraulic distribution divergence.

Numerical simulation of hydraulics in physical clogging process of horizontal subsurface flow constructed wetland

Experimental method

As an emerging two-phase flow simulation technology, the CFD-DEM coupling method combines the advantages of CFD in flow field processing and the characteristics of EDEM in particle material simulation, which significantly improves the computational efficiency of numerical solution and the solution accuracy of fluid–solid coupling problem⁴⁴. It shows a broad application prospect.

Procedure of model establishment

The hydraulic model utilized in the CFD-DEM simulation experiment involves simplifying the horizontal subsurface flow constructed wetland system into a z-y profile, as illustrated in Fig. 3. In this representation, the z-axis corresponds to the matrix length in meters, while the y-axis represents the matrix height in meters. The simulation outcomes are presented within this framework. The simulation employs a porous medium model to replicate fluid flow, in conjunction with the particle trajectory model within the Lagrangian discrete model. The micro-particles within the particle trajectory model exhibit effective tracking of the liquid phase flow field. The residence time and its distribution (RTD) of the particle swarm serve as an approximate representation of the RTD of the liquid-phase flow field⁴⁵.

Fig. 3

Simplified model of horizontal subsurface flow constructed wetlands.

After reasonable simplification of the horizontal submerged artificial wetland system, a simulation model was established. In view of the regularity of the model structure, a structured mesh with better convergence and mesh quality was adopted, with a mesh size of 0.02 m. The mesh was imported into Fluent for solving. A particle factory was set up in EDEM software to establish the Fluent-EDEM2020 coupling interface. Regarding the setting of boundary conditions, where the inlet is velocity inlet, the outlet is pressure outlet, the pressure is one standard atmospheric pressure, the rest of the wall boundary conditions are used, and the interior boundary conditions are used between the water distribution area and the matrix system area. The matrix porosity is set to 0.42, the flow of water in the matrix layer is laminar, and the heat transfer is ignored in the calculation process. After the calculation, the discrete phase model in FLUENT was used and particles were injected. The results of the calculations in FLUENT were exported and a .dpm file was generated, which was imported into excel to calculate the residence time of the particles, and ultimately a table of standard deviation as well as mean residence time data could be produced. From this, a graph of the simulated hydraulic residence time variation can be drawn.

Evaluation indicators

In the simulation of porous media model, the mass conservation and momentum conservation equation of fluid mechanics are considered by adding flow resistance⁴⁶.

$$\frac{\partial \rho }{\partial t}+\frac{\partial }{\partial {x}_{i}}\left(\rho {u}_{i}\right)={S}_{m}$$

(11)

In the formula, ρ is the fluid density; u is the fluid velocity vector ; S_m is the mass source term.

$$\frac{\partial }{\partial t}\left(\rho {u}_{i}\right)+\frac{\partial }{\partial {x}_{j}}\left(\rho {u}_{i}{u}_{j}\right)=-\frac{\partial P}{\partial {x}_{i}}+\frac{\partial {\tau }_{ij}}{\partial {x}_{j}}+\rho {\text{g}}_{i}+{S}_{i}$$

(12)

P is static pressure; ρgi is the gravity volume force; τ_ij is the stress tensor; S_i is an additional momentum loss source. S_i is composed of two parts, one is the viscous resistance loss source term, and the other is the internal resistance loss source term, which can be expressed as follows:

$${S}_{i}=-\left({\sum }_{j=1}^{3}{D}_{ij}\mu {u}_{i}+{\sum }_{j=1}^{3}{C}_{ij}\frac{1}{2}\rho \left|{u}_{i}\right|{u}_{i}\right)$$

(13)

In the formula, D is the inertial resistance loss coefficient matrix of viscous resistance;C is the inertial resistance loss coefficient matrix.

When constructing the physical blockage model of horizontal subsurface flow constructed wetland, it is necessary to ensure that the experimental system is in a relatively stable state. The variation of TSS concentration with time can be expressed by the following formula^47,48:

$$C \left(t\right)={C}_{0}{e}^{-k\left(t-{t}_{0}\right)}$$

(14)

C is the initial concentration of TSS; to is the initial moment; k is the reaction rate constant, which is 0.000096.

The diameter of the bottom of the z-axis of the experimental system is used as the abscissa, and its value is between 0–0.5 m. The y-axis direction of the system is used as the ordinate, and its value is between 0–0.2 m. The spatial distribution expression of TSS concentration is:

$$C \left(z,y,t\right)={C}_{0}{e}^{-k\left({t}_{z,y}-{t}_{0}\right)}$$

(15)

Nash–Sutcliffe efficiency coefficient (NSE) is a normalized statistical data obtained by comparing the residual variance of the simulated data with the variance of the measured data, which characterizes the fitting degree of the measured data sequence and the simulated data sequence with 1:1. The calculation formula is as follows:

$$NSE=1-\left[\frac{{\sum }_{i=1}^{n}{\left({Y}_{i}^{obs}-{Y}_{i}^{sim}\right)}^{2}}{{\sum }_{i=1}^{n}{\left({Y}_{i}^{obs}-{Y}_{i}^{mean}\right)}^{2}}\right]$$

(16)

In the formula, ${Y}_{i}^{obs}$ is the i-th measured data of the evaluated data sequence; ${Y}_{i}^{sim}$ is the i-th simulation data; Y_mean is the mean value of the measured data sequence; n is the number of data sequences. The NSE is between -∞ and 1 (including 1), and the simulation result is the best when the value of NSE is 1. The simulation results are acceptable at 0–1; the simulation results are unacceptable when is less than or equal to 0.

Study on structure optimization of horizontal subsurface flow constructed wetland based on numerical simulation

Experimental method

The fundamental characteristics of the experimental setup and the simplified simulation diagram of the horizontal subsurface flow constructed wetland are elaborated upon in the chapter 2.2.1. To investigate the impact of different baffle configurations on material settlement accumulation and hydraulic performance within the horizontal subsurface flow constructed wetland, three distinct operational scenarios are established (refer to Fig. 4). The first scenario involves a constructed wetland devoid of any baffles. The second scenario features an artificial wetland with vertical baffles, while the third scenario incorporates constructed wetlands with baffles set at 45° and 60°angles. Each baffle measures 0.2 m in width and 0.12 m in height. The inlet velocity is set at 0.0012 m/s, utilizing a velocity-entry condition. The outlet is maintained at a constant pressure, with a pressure-outlet boundary condition of 0 Pa. Given the relationship between viscosity and low speeds, laminar flow is employed instead of turbulent flow. CFD-EDM is utilized to simulate flow velocity and particle accumulation under the three operational conditions, with simulation results from the matrix system on the 10th, 20th, and 30th days serving as illustrative examples.

Fig. 4

Three operating conditions used for the simulation. (a) Condition 1; (b) Condition 2; (c) Condition 3.

The fundamental characteristics of the experimental setup and the simplified simulation schematic of the horizontal subsurface flow constructed wetland are elaborated upon in the chapter 2.2.1. To investigate the impact of the substrate front baffle on physical clogging within the wetland and confirm the enhanced hydraulic efficiency, two operational scenarios were established (refer to Fig. 5): one without a baffle (working condition 1) and another with a vertical baffle (working condition 2). CFD was utilized to simulate particle accumulation under these two conditions, with specific focus on the results obtained after 10, 20, and 30 days.

Fig. 5

A Schematic representation of the two working conditions used for the simulation. (a) Condition 1; (b) Condition 2.

Evaluating indicator

Permeability coefficient

The method of measuring the permeability coefficient is to measure the water depth at different positions, and then use the Darcy formula to calculate the average permeability coefficient at each position. However, this method has certain errors. The calculation formula is as follows:

$$K=\frac{ql}{At\left({h}_{1}-{h}_{2}\right)}$$

(17)

In the formula, K is the saturated permeability coefficient of the filler, m/d; At is the flow area,m2; q is flow, m³/d; l is the vertical distance between two measuring points, m; h1 andh2 are the water depths of the upper and lower reaches, m.

Tracer method

The tracer method is utilized to assess the hydraulic properties of wetlands by injecting a tracer to monitor changes in concentration and analyze the distribution of water flow within the substrate. Typically, the relationship between tracer concentration and time is represented by a peak curve, which provides insights into the flow field distribution within the matrix⁴⁹. By conducting comparative analyses on wetlands with varying degrees of clogging, researchers can establish the connection between hydraulic characteristics and clogging. The Residence Time Distribution (RTD) curve is derived from the conductivity data obtained through the tracer method, enabling the calculation of the actual hydraulic retention time⁵⁰.

When selecting tracers, factors such as cost, recovery rate, and potential harm must be taken into account. Commonly used tracers include dyes, fluorescent dyes, bromides, and salt ions, with a recovery rate exceeding 80% considered a suitable evaluation criterion⁵¹. A minimum of 30–40 samples is typically required to construct a comprehensive tracer curve⁵². Zhang conducted water tracing calculations based on the obtained curve⁵³.

The determination of cumulative clogging and matrix porosity serves to characterize the extent of clogging in constructed wetlands. The process involves cleaning the blockages, drying and weighing them at 105 °C, determining the accumulated blockages by further drying in a muffle furnace at 550 °C until a constant weight is achieved, and calculating the percentage of volatile organic compounds based on the sample loss mass. The substrate’s porosity is calculated by comparing the wetland drainage volume to the wetland volume, reflecting its water retention capacity. These measurements are crucial for comprehending the functionality and enhancing the efficiency of constructed wetlands.

Cumulative clogging and matrix porosity

The determination of cumulative clogging and matrix porosity serves to characterize the extent of clogging in constructed wetlands. The process involves cleaning the blockages, drying and weighing them at 105 °C, determining the accumulated blockages by further drying in a muffle furnace at 550 °C until a constant weight is achieved, and calculating the percentage of volatile organic compounds based on the sample loss mass. The substrate’s porosity is calculated by comparing the wetland drainage volume to the wetland volume, reflecting its water retention capacity. These measurements are crucial for comprehending the functionality and enhancing the efficiency of constructed wetlands.

Experimental results and analysis

Experimental study on hydraulic behavior of physical clogging process in horizontal subsurface flow constructed wetland

Comparison of hydraulic efficiency before and after clogging experiment

The standardized conductivity data was utilized to generate the hydraulic retention time distribution curve (RTD) of the horizontal subsurface flow constructed wetland experimental system both before and after the physical clogging experiment. This curve, depicted in Fig. 6, serves as a valuable tool for assessing hydraulic efficiency in a clear and straightforward manner. The curve exhibits a high and steep peak value, with a significant short circuit value (s). The timing of the peak (T_p) decreases as the peak becomes earlier, while the trailing duration (T_m) decreases as the trailing becomes shorter⁵⁴. Notably, the introduction of suspended solids solution impacts the timing of the peak. In the absence of this solution, the peak occurs rapidly and with a narrow width, indicating efficient system operation and fast water flow rates.

Fig. 6

RTD density plots obtained from NaCl tracer experiments.

Table A1 and formula (4) through (10) from the experimental data of the subsurface flow wetland indicate, as shown in Table 3, that the introduction of suspended solids results in an elevation of the short circuit value (s) and a decline in hydraulic efficiency. The effective volume ratio (e) for both sets of data was found to be below 1, suggesting the presence of a substantial dead zone. Typically, a higher λ value signifies superior hydraulic efficiency. With the addition of suspended solids, there is a notable increase in the standard deviation (σθ2), indicating that the water flow tends towards a fully mixed flow regime. The residence time is minimal, approaching that of piston flow. A system designed for piston flow requires a smaller reaction vessel, resulting in shorter reaction times when reactants are reduced to equivalent concentrations, thereby enhancing efficiency⁵⁵. The inclusion of suspended solids exacerbates the dead zone and diminishes hydraulic efficiency.

Table 3 Relevant hydraulical parameters before and after the blockage experiment.

The changes of TSS concentration and porosity of the system during the clogging experiment

The evolution of porosity in the horizontal subsurface flow constructed wetland experimental system over the course of the experiment is depicted in Fig. 7. The porosity of the system exhibited a decreasing trend as the experiment progressed. Variability was observed in the TSS value of the initial effluent and the TSS removal rate. Following a distinct inflection point on the 8th day, the system demonstrated a tendency towards stability. The TSS concentration remained within the range of 200–450 mg/L, while the TSS removal rate fluctuated between 45 and 70%. Analysis indicates that the system achieved a relatively stable state by the 5th day.

Fig. 7

Changes in the three indicators during the experiment. (a) Porosity versus time of the experimental system; (b) TSS concentration versus time in the experimental system; (c) TSS removal rate of experimental system changes with time.

The change of hydraulic retention time of the system during the clogging experiment

The porosity of the horizontal subsurface flow constructed wetland experimental system experienced a reduction from 40.43% to 35.69%, leading to diminished fluid volume and a shortened hydraulic retention time. The impact of this porosity decrease on other system performance aspects remains uncertain. By determining the porosity value, the theoretical hydraulic retention time (Tn) of the experimental system decreased from 0.622 h to 0.549 h. By evaluating the conductivity value, the actual hydraulic retention time (Tm) of the experimental system decreased from 0.595 h to 0.314 h, confirming that the blockage in the constructed wetland resulted in a shortened hydraulic retention time.

Upon comparing the variations in Tm and Tn in the horizontal subsurface flow constructed wetland experimental system (refer to Fig. 8), the findings reveal that Tm consistently falls below Tn at each operational stage, indicating the presence of a short-circuit phenomenon during system operation. Both Tm and Tn exhibited negative correlations with the system’s operational duration, displaying a linear relationship. Notably, the rate of decrease in Tm was observed to be faster than that of Tn, suggesting that clogging was not solely due to the reduction in porosity of the experimental system but also linked to the disruption of hydraulic behavior within the system caused by clogging.

Fig. 8

Changes in theoretical versus actual residence times during the experiment.

The change of hydraulic efficiency of the system during the clogging experiment

Based on the evaluation parameters calculated for the hydraulic behavior of the experimental system with the physical blockage process of the added suspension solution (refer to Fig. 9), the standard deviation (σθ²) observed during the experiment falls within the range of 0.021 to 0.181, indicating a proximity to fully mixed flow conditions. The effective volume ratio (e) ranges from 0.671 to 0.965, suggesting a substantial dead zone extent, with no discernible temporal trend in the e value. The hydraulic efficiency (T_p) values range from 0.168 to 0.204, allowing for classification into three categories based on T_p magnitude⁵⁶: superior hydraulic efficiency (T_p > 0.75), moderate hydraulic efficiency (0.75 ≥ T_p > 0.5), and poor hydraulic efficiency (T_p ≤ 0.5).

Fig. 9

Porosity and hydraulic residence time during the blocking experiment.

Analysis

In the constructed wetland system, an increase in suspended solids can result in clogging, causing a notable decrease in the hydraulic efficiency of the system. Following a stabilization period of 5 days, it was observed that while the effluent concentration and removal rate of TSS tended to stabilize, the porosity of the system decreased significantly from 40.43 to 35.69%. Moreover, both the theoretical hydraulic retention time and the actual hydraulic retention time were reduced, suggesting that factors beyond physical blockage contribute to the system’s hydraulic behavior. Post-clogging, the system exhibits short-circuiting and mixed flow phenomena, along with an expansion of dead zones, all of which contribute to an overall decrease in hydraulic efficiency.

Hydraulic numerical simulation analysis of physical clogging process in horizontal subsurface flow constructed wetland

Hydraulic retention time simulation results

The simulation results are exported as .dpm files to Excel software to calculate the particle residence time, indicating the hydraulic retention time of the simulation experiment. Comparing the measured values with the simulated values (Fig. 10), NSE, which is used to evaluate the efficiency of the hydrological model, is used to evaluate the simulation results⁵⁴, and the fitting degree is evaluated by comparing the residual variance with the measured data variance.

Fig. 10

Comparison between measured data and simulated data.

As far as this study is concerned, the NSE value calculated by substituting the measured and simulated values of hydraulic retention time into the above formula is 0.81, indicating that the simulation results are feasible and the effect is excellent.

CFD-DEM simulation results of flow field and hydraulic retention time

The results of the simulation demonstrate the flow field and velocity distribution within the horizontal subsurface flow constructed wetland experimental system, as depicted in Fig. 11. The dark blue region represents a stagnant area with zero flow velocity, indicating a substantial dead zone within the system and a limited effective watershed area. These findings align with the measured results, suggesting that system blockages result in a large unused area and reduced hydraulic efficiency.

Fig. 11

Simulation of the flow field distribution of the CFD-DEM experimental system.

Furthermore, the simulation reveals that the effective watershed area of the system remains relatively constant over time, consistent with the measured effective volume ratio (e). However, there is an observed increase in effective flow velocity with experimental duration, particularly at the system outlet (Fig. 12).

Fig. 12

Change diagram of outlet flow rate of added suspended solution.

Utilizing CFD-DEM simulation, the study analyzed the spatial and temporal distribution of water flow residence time within the experimental system matrix on days 6, 12, 18, and 24 (Fig. 13). The results indicate that the distribution curves for these time intervals progressively advance, implying a reduction in the time taken for water flow to reach specific locations within the system. This trend suggests issues such as short-circuiting within the experimental system, leading to a continual decrease in hydraulic retention time.

Fig. 13

Simulation of the spatiotemporal distribution of hydraulic dwell time in experimental systems.

In summary, with the increase of operation time, the effective watershed of horizontal subsurface flow constructed wetland remains unchanged, but the flow velocity is accelerated, and the time of water flow through the substrate is shortened, which leads to the decrease of hydraulic retention time and affects the contact time between water flow and substrate.

Analysis of spatial and temporal distribution model of physical clogging in horizontal subsurface flow constructed wetland

Through the chapter 2.2.1, the measured results show that the horizontal subsurface flow constructed wetland system needs 5 days to reach a stable state. Therefore, this paper takes the simulation of TSS concentration (Fig. 14), particle accumulation (Fig. 15) and porosity distribution (Fig. 16) on the 6th, 12th, 18th and 24th days as an example for simulation calculation. The colour intensity observed in the simulation results is directly proportional to the total suspended solids (TSS) concentration, accumulated material, and temporal and spatial distribution of porosity. In this context, redder colours indicate higher concentrations, while bluer colours indicate lower concentrations and accumulated material.The results show that the TSS concentration increases with time, and the vertical space decreases with the depth of the matrix. The treatment efficiency is stable and consistent with the measured results. The particle accumulation increases with time, and decreases with the depth of the matrix in the vertical space. However, the simulation does not consider the influence of the bottom boundary and the perforated plate, or causes deviation. The porosity decreases with time and matrix depth, indicating that the bottom of the system is easy to be blocked.

Fig. 14

Temporal and spatial distribution of TSS concentration in the experimental system.

Fig. 15

Temporal and spatial distribution of accumulated material in the experimental system.

Fig. 16

Temporal and spatial distribution of porosity in the experimental system.

The developed model enables the real-time monitoring of TSS concentration and matrix porosity changes over time at any location within the system, facilitating the identification of clogging processes. Model predictions indicate that the lower regions of the system, particularly the outlet on the right side at the bottom, are more prone to clogging, resulting in inadequate drainage and premature backwater formation on the system’s surface.

Analysis

Utilizing numerical simulation, the NSE coefficient of 0.81 was employed to validate the accuracy of the flow field and distribution of hydraulic retention time within the simulated system. The observed decrease in hydraulic retention time with an increase in system flow rate aligns with empirical measurements. Furthermore, the temporal and spatial clogging distribution model highlights that the lower wetland regions and outlet zones are predisposed to initial clogging, consequently impeding drainage. These findings hold significant implications for the effective maintenance and management of the wetland system.

Analysis of research results of horizontal subsurface flow constructed wetland structure optimization based on numerical simulation

Experimental results

The flow velocity within the matrix area under baffle conditions experienced fluctuations during operation. The impact of baffle placement on the flow field is depicted in Fig. 17. The dark blue region represents a stagnant area with zero flow velocity, indicating a substantial dead zone within the system and a limited effective watershed area. Installation of the baffle resulted in a deceleration of flow rate compared to the absence of a baffle, leading to a decrease in flow rate. The baffle serves to impede eddy currents, diminish flow velocity, and influence particle sedimentation. However, an inclined baffle may excessively obstruct water flow, potentially causing flooding. Hence, the design of constructed wetlands should consider solid particle concentration to ensure adequate flow rates.

Fig. 17

Flow field distribution diagram under different working conditions.

Particle movement undergoes alterations during operation, as illustrated in Fig. 18, depicting the effect of baffles on particle dynamics. Red indicates the highest accumulation of material, whereas blue represents the lowest concentration.Conditions 2 and 3 exhibit enhanced settlement efficiency due to baffle obstruction. In Condition 3, the 60° baffle prevents particle overflow, resulting in higher settlement efficiency and increased particle accumulation. Nevertheless, the vortex generated by the 60° baffle may cause particles to float, compromising long-term stability compared to a vertical baffle.

Fig. 18

Distribution diagram of accumulated material under different working conditions.

The impact of a vertical baffle in the water distribution area on solid particle settlement in horizontal subsurface flow constructed wetlands was assessed through CFD simulation, as shown in Fig. 19. Red indicates the highest accumulation of material, whereas blue represents the lowest concentration.Condition 2 (with a vertical baffle) notably enhanced particle sedimentation and delayed physical clogging of the wetland. Particles obstructed by the baffle predominantly settle in the water distribution area, facilitating cleaning and subsequent maintenance. Therefore, it is advisable to incorporate a vertical baffle in the water distribution area or include a sedimentation tank baffle when optimizing the structure of horizontal subsurface flow constructed wetlands.

Fig. 19

Distribution map of particulate particle concentration under different working conditions.

Following the completion of CFD simulation, the outcomes are exported as a .dpm file, which is imported into excel for subsequent analysis of particle residence time and computation of hydraulic performance metrics for each operational scenario (refer to Table 4). The findings reveal that the flow divergence (σ_θ²) value for operational condition 2 (featuring a vertical baffle) is recorded at 0.056, significantly lower than the 0.217 observed for condition 1, suggesting enhanced hydraulic performance approaching that of piston flow. Moreover, the effective volume ratio (e) for condition 2 is notably higher, leading to improved spatial utilization efficiency and an approximately 15% extension in the duration before complete blockage occurs. The hydraulic efficiency (λ) attains 0.475 in condition 2, representing a 21% increase compared to condition 1. The investigation validates that the proposed optimization strategy effectively enhances the hydraulic performance of horizontal subsurface flow constructed wetlands and mitigates physical clogging issues.

Table 4 Experimental hydraulic performance calculation results.

Analysis

Through CFD analysis, it has been determined that the presence of vertical baffles can enhance particle settling in wetland systems and decrease particle accumulation downstream of the substrate. However, it was observed that the substrate located upstream of the baffle is susceptible to blockages. Experimental results indicate that the wetland system incorporating vertical baffles (condition 2) exhibits superior hydraulic characteristics and reduced risk of physical blockages compared to the system lacking baffles (condition 1). Specifically, condition 2 prolongs the time until complete clogging of the wetland by 15% and enhances hydraulic efficiency by 21%. Therefore, it is advisable to incorporate vertical baffles in the water distribution zone or introduce sedimentation tanks with vertical baffles at the inlet of the wetland when optimizing the design of horizontal subsurface flow constructed wetlands. This approach can enhance the overall performance of the wetland system and improve its effectiveness in addressing clogging issues.

Practical application

Design of horizontal subsurface flow constructed wetland sewage treatment system based on structural optimization scheme

Design basis

This chapter addresses the issue of rural sewage discharge by proposing a decentralized, small horizontal subsurface flow constructed wetland system that has been structurally optimized for the treatment of rural domestic sewage. The system was implemented in a village in Rui’an City to enhance rural prosperity. Fanrong village has 898 households and 3,577 residents, with an average of four individuals per household. The average daily sewage output per household is estimated at 0.6 m³/d. The primary component of the wetland system is a horizontal subsurface flow constructed wetland bed with a length-to-width ratio of 3:1, occupying an area not exceeding 10m². Locally abundant gravel and pebbles from Fanrong village were chosen as the substrate material, considering the economic conditions prevalent in rural areas. The fundamental design parameters are detailed in Table 5.

Table 5 Geometric dimensions of modular constructed wetlands.

Design construction

Based on established design criteria and parameters, in conjunction with industry standards such as the 'Code for Design of Artificial Wetlands’ and 'Guidelines for Operation and Maintenance of Rural Domestic Sewage Artificial Wetland Treatment Facilities’, a compact horizontal subsurface flow constructed wetland sewage treatment system tailored for rural settings was developed. The constructed wetland bed is segmented into three distinct sections. The initial section comprises a 1-m-long sedimentation tank linked to the sewage tank, featuring two vertically positioned baffles within. Each baffle measures 1.2 m in width and 0.6 m in height, with the first baffle positioned 0.3 m from the edge, and a 0.3-m separation between the two baffles. The subsequent middle section spans 2.2 m and serves as the primary treatment zone housing the internal filling treatment matrix. Finally, the lower section consists of a 0.4-m-long collecting canal that connects to the reservoir (refer to Fig. 20).

Fig. 20

Structure diagram of the horizontal subsurface flow constructed wetlands system.

Process flow

The preliminary stage involves precipitating domestic sewage to eliminate suspended solids and bacteria. The treated effluent is transferred to a constructed wetland to remove suspended particles, nitrogen, and phosphorus pollutants. This system treats farmers’ sewage, facilitating reuse or discharge. The horizontal subsurface flow constructed wetland is used for this, facilitating recycling efforts. The treated water is suitable for irrigation and other uses, facilitating both sewage treatment and recycling.

The practical application of horizontal subsurface flow constructed wetland sewage treatment system in rural landscape

Plant selection

When establishing a small-scale constructed wetland in Rui’an City, Wenzhou, it is crucial to select plant species suited to the local climate. A diverse selection of trees, plants, and aquatic vegetation should be included to enrich the landscape and promote ecological diversity. It is crucial to select cost-effective, durable, pollution-resistant, and aesthetically pleasing plant varieties that respect the preferences of local residents. By respecting the choices of the community members, a harmonious and natural sewage treatment system can be integrated seamlessly into the landscape. The recommended plant configuration method involves placing taller trees such as cinnabar and pomegranate at the upper level, followed by middle plants like Acer palmatum and tortoise shell winter green, lower plants including Fragrant burdock and February orchid, and aquatic plants such as calamus and water lily.

Landscape construction design

Horizontal subsurface flow constructed wetlands are a crucial element of sewage treatment systems, with a primary focus on ecological considerations. It is essential to prioritize ecological integrity and landscape preservation while ensuring effective sewage purification. Proper maintenance of plants is necessary to sustain the wetland’s purification capacity and prevent overloading. Additionally, integrating the wetland with other purification equipment is vital to achieve system harmony and unity. By blending seamlessly with the local natural environment and human characteristics, engineering footprints can be minimized, resulting in dual benefits of enhanced landscape aesthetics and ecological functionality.

Following an evaluation of current conditions, a renovation plan has been developed for the small courtyards in Fanrong village, which collectively span less than 20 square meters. These underutilized and dispersed courtyards have the potential to be transformed in a way that enhances their utility and aesthetic appeal through a process of thoughtful planning and landscape design. The plan, shown in Fig. 21, aims to optimize the courtyard space with open areas, scenic spots, dry features, and wetland zones. The plan enhances connectivity and functionality in the courtyards, catering to residents’ needs and dividing the space to balance daily activities and visual appeal.

Fig. 21

(a) The location of the project site; (b) Landscape plan; (c) Current status of the site.

Operation monitoring based on physical blockage simulation of horizontal subsurface flow constructed wetland wastewater treatment system

Prediction of clogging time of horizontal subsurface flow constructed wetland

Based on the theoretical framework outlined in Chapter 5.1.2. regarding physical clogging, this chapter presents a proportional model for optimizing the structure of horizontal subsurface flow constructed wetlands. The model predicts clogging time using computational fluid dynamics coupled with discrete element method simulations. The findings indicate that the system achieves stability within 4–5 months, but complete blockage occurs after 14 years and 3 months, surpassing the anticipated service life of 10–15 years specified in the "Constructed Wetland Design Specification". Consequently, the optimization approach proposed in this research not only aligns with the stipulated criteria but also enhances operational efficiency and hydraulic performance significantly.

Water quality monitoring method of horizontal subsurface flow constructed wetland

A compact horizontal subsurface flow wetland system was implemented for the purpose of wastewater treatment originating from communal establishments and rural residences within Fanrong village. The sewage received initial processing through screening and sedimentation procedures, subsequently undergoing enhanced purification within an oxidation pond and a horizontal subsurface flow constructed wetland. Following a period of 5 months of operational activity, an analysis was conducted to ascertain the principal contaminants remaining in the treated effluent, with specific monitoring methods outlined in Table 6.

Table 6 Contaminant and detection methods.

Prevention and recovery of physical blockage in horizontal subsurface flow constructed wetland sewage treatment system

In the management of the horizontal subsurface flow constructed wetland sewage treatment system, it is essential to focus on preventing physical blockages and implementing comprehensive preventive measures in design and operational maintenance to delay matrix blockages. These actions are crucial for extending the wetland’s operational lifespan.

Prevention of blockage

In addressing the issue of physical clogging in horizontal subsurface flow constructed wetlands, a proactive approach is essential. This involves selecting a matrix with a larger particle size and high porosity during design and maintenance to improve permeability. Additionally, pretreating sewage to decrease suspended solids, promptly removing withered plant residues, and implementing protective measures are recommended. Restoring water intake and operation can also help. Rigorous daily maintenance and management practices are important for preventing pollution and forming blockages in wetlands.

Recovery of blockage

In instances where a horizontal subsurface flow constructed wetland experiences physical obstruction, it is advisable to initiate maintenance procedures by clearing and replacing heavily blocked substrate, particularly in the water inlet region, to ensure the continued efficient functioning of the wetland. Furthermore, the introduction of small saprophytic organisms like earthworms can aid in the decomposition of organic material obstructing the matrix and in clearing the pores, thereby mitigating blockages effectively. Additionally, implementing a sedimentation tank equipped with a vertical baffle upstream of the water inlet area represents a viable optimization strategy for minimizing clogging issues.

Discussion

This article examines the hydrodynamic behavior of a horizontal subsurface flow constructed wetland in the presence of physical clogging. It presents a spatio-temporal model of clogging, proposes a structural optimization scheme, and offers a verification of the proposed scheme. In light of these findings, the study provides guidance for wastewater treatment and wetland landscape creation in Prosperity Village. The principal conclusions are as follows:

• (1) The experimental system was stabilized within a period of 4–5 days. During this time, the suspended material caused a notable decrease in hydraulic efficiency, a reduction in porosity, a contraction in the hydraulic residence time, and the emergence of undesirable flow regimes.

• (2) A CFD-DEM simulation was conducted to verify the experimental results. It was observed that an increase in the water flow rate resulted in a decrease in the actual hydraulic residence time. The blockage model reveals changes in material accumulation, porosity, and TSS concentration.

• (3) The addition of a baffle improves deposition efficiency. The optimized structure requires vertical baffles, extends clogging time by 15%, improves hydraulic efficiency by 21%, and the system is more stable.

This article presents an initial investigation into the physical clogging mechanism of a horizontal submerged artificial wetland wastewater treatment system. Further research should concentrate on extending the experimental period to allow for complete clogging, incorporating plant growth factors to enhance the model, and improving the validation of practical engineering experiments. These avenues of enquiry will serve to enhance the accuracy of the model and facilitate the integration of academic and engineering applications.