Investigating the Effect of Aeration on Residence Time Distribution of a Baffled Horizontal Subsurface Flow Constructed Wetland

Abstract

Constructed wetlands (CWs) are cost-effective and sustainable systems for wastewater treatment, but their hydraulic performance remains a critical challenge. In this study, a lab-scale baffled horizontal subsurface flow constructed wetland was modeled using Computational Fluid Dynamics to investigate the effects of aeration strategies on hydraulic performance, focusing on aeration rates and positions. A gas–liquid two-phase flow system was modeled using the Euler–Euler approach with the Darcy–Forchheimer model in OpenFOAM, simulating 15 cases with varying aeration rates (0.1–0.3 m³/day) and positions (middle of channels vs. bends at the ends of baffles). Results show that the introduction of aeration influenced hydraulic efficiency (HE) and the Morrill Dispersion Index (MDI). Without aeration, the baseline HE was already high (HE = 0.9297) due to the optimized baffle configuration. However, aeration further improved performance, with HE increasing to 0.9594 and MDI decreasing from 1.6087 to 1.4000 when aeration was applied at bends (Position C) at 0.3 m³/day. Aeration at bends was more effective than mid-channel aeration, promoting uniform flow distribution and reducing short-circuiting. These findings highlight the importance of aeration positioning and provide insights for optimizing CW design to balance energy consumption and hydraulic performance.

1. Introduction

Constructed wetlands (CWs) have been globally recognized as cost-effective and sustainable solutions for wastewater treatment. Their ability to integrate natural processes with engineered systems enables the removal of various pollutants, such as organic matter, nitrogen, and phosphorus, while promoting biodiversity and environmental conservation [1,2]. However, the hydraulic performance of CWs is still a critical challenge, as poor hydraulic performance often leads to short-circuiting and uneven pollutant removal, reducing the overall treatment efficiency.

To address this issue, researchers have examined how hydraulic behavior affects treatment outcomes under different design and operational parameters. These include influent modes [3], hydraulic loading rate [4], and system configurations like baffle designs and hydraulic residence time [5]. These structural modifications help reduce dead zones and enhance contact between pollutants and reactive surfaces, thereby improving treatment efficiency [6].

Aeration has been demonstrated as a promising approach to improve the treatment performance of CWs by improving oxygen transfer, promoting microbial activity, and accelerating pollutant degradation [7,8,9]. The installation of baffles in aerated CWs can better facilitate the formation of multistage Anaerobic-Oxic (AO) zones within the system. Such a multistage system, through segmented treatment, can effectively avoid the problem of excessive load on a single stage, thereby achieving more efficient organic matter removal [10]. Additionally, it increases the frequency of nitrification and denitrification reactions, leading to enhanced TN removal efficiency [11]. However, the impact of aeration on hydraulic behavior, especially in baffled CW systems, remains insufficiently understood. Parameters such as aeration rate and aeration position may have an influence on the RTD within the wetland system, potentially affecting both hydraulic and treatment efficiencies.

Recent studies have attempted to address these gaps using different design strategies. For example, Hua, Liu [2] investigated cascading aeration in spiral wetlands and reported improved hydraulic behavior and pollutant removal due to enhanced mixing and reduced short-circuiting. Wu, Fan [12] studied intermittently aerated vertical flow CWs and showed that aeration rate and timing significantly affected treatment efficiency and hydraulic distribution. These findings indicate that aeration strategy design plays a key role in optimizing system performance. However, these works primarily focused on pollutant removal rather than detailed hydraulic behavior.

Furthermore, the configuration of aerators has been shown to impact flow patterns. Alvarado, Vesvikar [13] showed that while additional aerators can improve flow uniformity, the gains plateau beyond a certain number, highlighting the trade-off between hydraulic performance and energy use. Additionally, optimizing aeration strategies in CWs has significant implications for energy conservation. Aeration is often one of the most energy-intensive processes in wastewater treatment systems, accounting for up to 75% of the total energy expenditure [14]. Improper aeration strategies can lead to unnecessary energy consumption without corresponding improvements in treatment performance. Therefore, a deeper understanding of these interactions is crucial for the optimal design and operation of aerated CWs.

Computational Fluid Dynamics (CFD) has emerged as a powerful tool for simulating complex multiphase flow systems in CWs, which enables the analysis of key factors affecting hydraulic performance under various operating conditions. In recent years, many studies have successfully applied CFD modeling to explore the effects of changes in design parameters, including the inlet–outlet configuration [15], vegetation distribution [16], and clogging effects [17,18]. Our previous research [19] also demonstrated the effectiveness of baffle design in improving HE through CFD simulations. While these studies confirm CFD’s utility in analyzing hydraulic performance, they rarely address how aeration configurations interact with flow dynamics in baffled systems.

To address these gaps, this study employs a gas–liquid two-phase CFD model to systematically evaluate how different aeration rates and positions influence RTD in a baffled horizontal subsurface flow constructed wetland (HSFCW). Unlike previous studies that primarily focused on pollutant removal, our research emphasizes hydraulic dynamics under varied aeration conditions. The novelty of this work lies in

(1)

the application of a two-phase CFD model to simulate aeration-driven hydraulic behavior in baffled systems;

(2)

the investigation of aeration configuration (rate and position) as design variables affecting RTD and HE;

(3)

the provision of actionable insights for optimizing aeration strategies that balance treatment performance and energy use.

These findings will contribute to a deeper understanding of the interplay between aeration and hydraulic processes, providing practical guidance for the sustainable design and operation of aerated CWs.

2. Description of the CFD Model

2.1. Modeling Setup

A lab-scale rectangular HSFCW with seven baffles (0.5 m in length) was investigated under different aeration strategies in this study. The reactor, measuring 1.2 m in length, 0.6 m in width, and 0.6 m in depth (Figure 1), was modeled. Based on the physical characteristics of our laboratory-scale reactor [6], the water depth was maintained at 0.4 m, and the reactor was filled with a filter medium with a porosity of 45%, resulting in an effective pore volume of approximately 162 L. Based on a constant flow rate of 2.7 L/h, the nominal residence time $t_n$ of the system was calculated as 48 h. Our earlier study [19] demonstrated that this baffle configuration, featuring seven baffles of 0.5 m in length, significantly enhanced the HE of CWs, increasing it from 59% in a system without baffles to 95%. This configuration was selected for its ability to achieve high HE while also promoting the formation of multistage AO zones in aerated CWs, which is beneficial for pollutant removal. Two aeration parameters were considered, i.e., aeration rates and aeration positions (Figure 2), the combinations of which are shown in

Table 1. Combinations of aeration strategies investigated in simulations.

Aeration Position	Aeration Rate (m3/day)
	0.1	0.15	0.2	0.25	0.3
A	A1	A2	A3	A4	A5
B	B1	B2	B3	B4	B5
C	C1	C2	C3	C4	C5

. The values of aeration rates were determined via the literature [12] and the scale of the reactor. The inlet–outlet configuration, flow rate, and porosity were kept constant for all cases. The inlet was set at 0.3 m from the base of the reactor and outlet was set at 0.4 m. Consequently, 15 cases were simulated to evaluate the effect of different aeration strategies on the hydraulic performance of CWs.

To determine an appropriate mesh size for the simulations, the computational domain was initially discretized using structured cubes measuring 1 × 1 × 1 cm. A grid independence study was then conducted by comparing the results obtained with this 1 cm cell size to those generated using a finer 0.5 cm cell size. The outcome showed that the mean velocity and pressure predicted using the 1 cm cell differ less than 1% from those of the 0.5 cm cells. Therefore, the computational domain was divided into 421,150 structural meshes with a size of 1 × 1 × 1 (cm) for subsequent simulations. The three-dimensional mesh for the computational domain is shown in Figure 1b.

The boundary conditions for the CFD simulation in OpenFOAM were set to replicate the flow and aeration processes within the CW system shown in Figure 1a. Both the water and air inlets were defined using the flowRateInletVelocity boundary condition to specify the respective inflow rates for the liquid and gas phases. The top surface, which is open to the atmosphere, was modeled using the pressureInletOutletVelocity condition to allow for free movement of the gas phase and interaction with the environment. The water outlet was assigned a zeroGradient pressure boundary condition to enable free outflow while maintaining pressure balance. No-slip boundary conditions were applied to all solid surfaces, including the baffles, side walls, and bottom surface, to account for frictional effects on the liquid phase. The filter medium was additionally modeled with a porous resistance term using the Darcy–Forchheimer equation to simulate flow resistance through the porous structure. The parameters defining the fluids and pollutant transport used in the model simulations (

Table 2. Parameters of CWs used in the simulations.

Parameter	Value	Unit	Source
water kinematic viscosity	1 × 10−6	m2/s	Empirical value
water density	1000	kg/m3	Empirical value
air kinematic viscosity	1.48 × 10−5	m2/s	Empirical value
air density		kg/m3	Empirical value
surface tension of water/air	7.20 × 10−2	N/m	Empirical value
gravitational acceleration	−9.81	m/s2	Empirical value
porosity	45%	-	tested
average particle size	8	mm	tested
scalar diffusion coefficient	2.00 × 10−6	-	[20]

) are based on empirical values, pertinent literature, and experimental tests. The filter medium is considered to be gravel with a nominal particle size of 8 mm.

2.2. Modeling of Internal Flows

In this study, the HSFCW system was created as a three-dimensional model and simulated using OpenFOAM v2312, which is a free and open-source CFD software developed primarily by OpenCFD Ltd. (Bracknell, UK) since 2004. The twoPhaseEulerFoam was selected as the solver for this study to simulate the effects of aeration process in the baffled CW. This solver is specifically designed for gas–liquid two-phase flow systems using the Euler–Euler approach, which treats both phases as interpenetrating continua. Its capability to model large-scale multiphase interactions makes it particularly suitable for analyzing aerated flow dynamics in a porous medium like CW. The choice of twoPhaseEulerFoam is driven by its ability to efficiently handle multiphase flow in porous environments without explicitly tracking individual gas bubble shapes or interfaces. This is essential for studying the global impacts of aeration, such as gas distribution, interphase momentum exchange, and their influence on hydraulic performance.

The gas–liquid two-phase flow system was modeled using the Euler–Euler approach, which is governed by the mass conservation equation (Equation (1)), and the momentum conservation equation (Equation (2)) for each phase.

$${\displaystyle \frac{\partial \left({\alpha }_k{\rho }_k\right)}{\partial t}}+\nabla ·({{\alpha }_k\rho }_k{u}_{ki})=0$$

(1)

where ${\alpha }_k$ is the volume fraction of phase $k$ ($k$ = liquid, gas), ${\rho }_k$ is the density of the phase $k$, $u_{ki}$ is the velocity of the phase $k$ in the $i$-direction, and the subscript $i=1,2,3$.

In considering the porous structure within the wetland, the Darcy–Forchheimer model was used to add a porosity zone to the fluid domain by adding a momentum loss source term $S_i$ to the momentum equation to simulate the hindering effect of the substrates on the internal fluid [21].

$${\displaystyle \frac{\partial \left({\rho }_k{u}_{ki}\right)}{\partial t}}+\nabla ·\left({{\alpha }_k\rho }_k{u}_{ki}{u}_{kj}\right)=-{\alpha }_k\nabla p+\nabla \left({\tau }_{ij}+{\tau }_{tij}\right)+{\alpha }_k{\rho }_k{g}_i+f_{\sigma i}+S_i$$

(2)

in which $u_{ki}$ and $u_{kj}$ represents the velocity of phase $k$ in $i$- and $j$-directions, respectively; $p$ is the pressure shared by both phases; ${\tau }_{ij}$ and ${\tau }_{tij}$ are the viscous and turbulent stress, since the flow in CWs is generally classified as laminar (Re < 600) [22]; ${\tau }_{tij}$ is equal to 0; $g_i$ is the gravitational acceleration; $f_{\sigma i}$ is the surface tension; $S_i$ is the momentum loss source term described using the Darcy–Forchheimer equation (Equation (3)); and the subscripts i, j = 1, 2, 3:

$$S_i=-\left(C_1{\mu u}_i+C_2{\displaystyle \frac{1}{2}}\rho \left|u_i\right|u_i\right)$$

(3)

where $\mu$ is the fluid viscosity, and the coefficients C₁ and C₂ must be specified in OpenFOAM, which can be calculated by solving the following equations [23]:

$$C_1={\displaystyle \frac{150}{d_p^2}}{\displaystyle \frac{{\left(1-\epsilon \right)}^2}{{\epsilon }^3}}$$

(4)

$$C_2={\displaystyle \frac{3.5}{d_p}}{\displaystyle \frac{\left(1-\epsilon \right)}{{\epsilon }^3}}$$

(5)

in which $d_p$ is the mean particle size of the filter and $\epsilon$ is the porosity of the substrates. Furthermore, the fluid domain is considered as equivalent to being spatially averaged in the process of solving the porous media model [17]. Hence, the porosity value used in the simulation is the spatial average value determined by the experiment.

2.3. Simulating Tracer Transport

To simulate tracer transport in the CW system, a scalar field was introduced to represent the concentration of the tracer. This was achieved by adding a scalarTransport function to the controlDict file in OpenFOAM. The scalarTransport function enables the solution of an additional scalar transport equation alongside the existing flow fields, capturing the behavior of the tracer as it is transported and diffuses through the domain. The convection–diffusion equation is as follows [24]:

$${\displaystyle \frac{\partial T}{\partial t}}+\nabla \cdot \left(\mathit{U}T\right)-{\nabla }^2\left(D_{T}T\right)=0$$

(6)

where $T$ is the transported scalar, $\mathit{U}$ is the fluid velocity, and $D_T$ is the diffusion coefficient divided by the fluid density; both are supposed to be constant. Then RTD functions can be obtained by monitoring the value of scalar at the outlet.

2.4. The RTD Model

The time it takes for individual fluid particles to pass through a tank (or reactor) varies based on the specific hydraulic pathways they follow [25]. This duration is commonly referred to as the exit age or residence time. The RTD curve, typically represented as E(t), describes the statistical distribution of the exit ages for all fluid particles passing through the reactor. The RTD curve is expressed in units of the inverse of time and is characterized by the following equation:

$$E(t)={\displaystyle \frac{Q\left(t\right)C\left(t\right)}{{\int }_0^{\mathrm{\infty }}Q\left(t\right)C\left(t\right)dt}}$$

(7)

where $Q\left(t\right)$ is the flow rate at time t, $C\left(t\right)$ represents both the outlet tracer concentration at time t and the value of the outlet scalar at time t, and $dt$ is the sampling time interval.

From the RTD analysis, the following key parameters were derived to evaluate the hydraulic performance:

$$t_m={\int }_0^{\mathrm{\infty }}t\cdot E\left(t\right)dt$$

(8)

The parameter HE $\lambda$ enables for a comparison of various systems. This parameter is a measure of the CW’s ability to evenly distribute flow throughout the occupied water volume, as well as to provide proper mixing or recirculation [26]. HE was calculated with Equation (9) [27]:

$$\lambda ={\displaystyle \frac{t_m}{t_n}}$$

(9)

where $t_n$ is the nominal residence time.

The Morrill Dispersion Index (MDI) is a widely used parameter to evaluate the HE and flow uniformity of CW. It is defined as the ratio of the time required for 90% of the tracer to exit the system ($t_{90}$) to the time required for 10% of the tracer to exit ($t_{10}$) [27,28]:

$$\mathrm{M}\mathrm{D}\mathrm{I}={\displaystyle \frac{t_{90}}{t_{10}}}$$

(10)

This index provides insight into the degree of mixing, short-circuiting, and dead zones within the system. An MDI value close to one indicates near-ideal plug flow, where fluid particles move uniformly through the system, maximizing HE. Conversely, a higher MDI value suggests significant dispersion and poor flow uniformity, often associated with short-circuiting or stagnant zones.

3. Results and Discussion

3.1. RTD Analysis

Based on the results of the CFD models, the normalized value of the passive transport scalar was collected at the outlet to generate RTD curves for all cases. The RTD curves for the aerated CW system are shown in Figure 3a–c for three aeration positions (A, B, and C) under varying aeration rates (A1–A5, B1–B5, C1–C5), and with no aeration as a control. The findings indicate a consistent trend across positions: as the aeration was integrated in the system, the shape of the RTD curve became slightly taller and narrower compared to the control group (no aeration). However, there was no significant change in the RTD shape with increasing aeration rate. This pattern suggests that higher aeration rates have very limited impact on the RTD.

Figure 3d presents the comparison of RTD curves for three different aeration positions under a fixed aeration rate (0.2 m³/day). It can be observed that the location of the aeration point has a more significant impact on the RTD curves than the aeration rate, particularly when the aeration point is positioned at the bend at the end of the baffles (Position C). This indicates that aeration at the bend promotes a more uniform flow distribution and a longer mean residence time, resulting in flow patterns closer to ideal plug flow. In contrast, the RTD curves for the two scenarios where the aeration points were placed in the middle of the channels (Positions A and B) show some improvement compared to the non-aerated control group, but the differences between these two positions are not significant.

The reasons for this phenomenon can be better understood through flow field diagrams (Figure 4). Positioning aeration points at the center of each baffle section enhances local flow uniformity and mixing, as seen in case (Figure 4a), where high-velocity regions are concentrated around the aeration point, promoting localized mixing. However, the effect of aeration is limited to the surrounding aera of the aeration point, and this localized improvement fails to address issues such as short-circuiting and dead zones in the corners. Low flow velocity problems persist in these areas. In contrast, relocating the aeration point to the bend at the end of the baffles (as shown in Figure 4b) creates a wider, high-velocity zone that extends over a larger area of the wetland. This configuration targets the transitional flow zones, significantly reducing flow stagnation and resulting in a more uniform velocity distribution across the system. In Case C3 (Figure 4b), the transitions between baffle sections become smoother, allowing the flow to more closely approximate ideal plug flow conditions, ensuring a more consistent RTD. Consequently, this aeration layout effectively optimizes the HE, making it a more robust strategy for improving the overall performance of the wetland system.

3.2. Hydraulic Performance Analysis

Table 3. Values of parameters defining the hydraulic performance of HSFCWs.

Case	Aeration Rate (m3/day)	${\mathit{t}}_{10}$ (h)	$\mathbf{M}\mathbf{D}\mathbf{I}$	${\mathit{t}}_{\mathit{m}}$	λ
0	0	34.5	1.6087	44.63	0.9297
A1	0.1	35	1.5500	44.89	0.9352
A2	0.15	35	1.5357	44.95	0.9365
A3	0.2	35	1.5214	45.03	0.9381
A4	0.25	35.25	1.4965	45.15	0.9406
A5	0.3	35.25	1.4823	45.23	0.9423
B1	0.1	35.75	1.5455	44.92	0.9358
B2	0.15	35.75	1.5315	44.99	0.9373
B3	0.2	35.75	1.5105	45.10	0.9396
B4	0.25	36	1.4861	45.26	0.9429
B5	0.3	36	1.4792	45.33	0.9444
C1	0.1	36	1.4514	45.49	0.9477
C2	0.15	36	1.4375	45.65	0.9510
C3	0.2	36	1.4306	45.79	0.9540
C4	0.25	36	1.4236	45.95	0.9573
C5	0.3	36.25	1.4000	46.05	0.9594

presents four evaluation indicators used to assess the hydraulic performance of CWs, including a short-circuit indicator $t_{10}$, a mixing indicator MDI, the mean residence time $t_m$, and HE λ. MDI and HE are the most commonly used and reliable indicators for evaluating hydraulic performance of CWs, but they focus on different aspects [29]. MDI assesses the flow uniformity and degree of mixing within a system, which is particularly useful for identifying short-circuiting and dead zones. In contrast, HE evaluates the overall system efficiency by comparing the mean residence time $t_m$ to the nominal residence time $t_n$, making it more suitable for determining how effectively the system utilizes its design capacity. In general, the larger the value of $t_{10}$ and the smaller the value of MDI (≥1), the closer the wetland flow is to the plug flow. Additionally, the closer the value of $t_m$ is to $t_n$, and the larger the value of λ (≤1), the closer the flow characteristics are to ideal plug flow.

The overall trends in Table 3 indicate that both increasing aeration rates and changing aeration positions can improve the hydraulic performance of CW. As the aeration rate increases from 0 to 0.3 m³/day, there is a consistent improvement in HE and a reduction in MDI across all positions. This is consistent with previous findings. For example, Hua, Liu [2] demonstrated that a cascading aeration design significantly enhanced the effective volume ratio and ammonia removal by improving hydraulic behavior and reducing short-circuiting in spiral flow wetlands. A similar trend was observed in previous studies. Butterworth, Dotro [30] pointed out that aeration enable easy water transport through settled solids in the CW system and ameliorated the HE value of CW.

Moreover, the results of Hreiz, Potier [31] confirm that the spatial layout of aerators critically affects RTD and turbulence in channel reactors. Their study showed that sidewall and distributed aeration generated more uniform mixing and flow patterns than central configurations. Notably, the impact of aeration becomes more pronounced when the aeration point is placed at the bend, with Position C consistently achieving the highest λ values and the lowest MDI values at all aeration rates. At an aeration rate of 0.3 m³/day, HE increases from 0.9297 without aeration to 0.9423 for Position A, 0.9444 for Position B, and 0.9594 for Position C, highlighting the superior performance of aeration at bends compared to mid-channel positions. Similarly, the mixing indicator MDI, which reflects flow uniformity, decreases from 1.6087 in the control to 1.4823, 1.4792, and 1.4000 for Positions A, B, and C, respectively, under the same aeration rate. These results emphasize that aeration at bends not only reduces short-circuiting more effectively but also enhances the RTD closer to ideal plug flow. Moreover, while increasing the aeration rate from 0.1 to 0.3 m³/day improves λ across all positions, the improvement is most pronounced for Position C, demonstrating its effectiveness in optimizing hydraulic performance. This finding underlines the critical role of aeration positioning in achieving enhanced HE in wetland systems. This spatial effect aligns with findings by Hreiz, Potier [31], who emphasized the role of aeration-induced turbulence in shaping plug-flow-like behavior.

On the other hand, with increasing aeration rates, the hydraulic performance improved across all cases, albeit to a limited extent. Specifically, HE increased and MDI decreased. This outcome is consistent with the conclusions drawn by Munoz, Drizo [32], who reported that aeration increases HE by improving water mixing and dispersion, as evidenced by changes in tracer response curves and higher numbers of continuously stirred tank reactors (CSTRs). This improvement helps reduce preferential flow paths and short-circuiting. However, as also demonstrated by Alvarado, Vesvikar [13], over-aeration does not necessarily lead to proportional hydraulic benefits. Their CFD-based scenario analysis in a full-scale aerated pond found that running more aerators increased energy use without significantly improving RTD performance. This supports our observation that continuously increasing the aeration rate does not benefit treatment efficiency. Wu, Fan [12] explored the optimization of organics and nitrogen removal in intermittently aerated vertical flow CWs, focusing on aeration times ranging from 1 to 10 h per day and aeration rates from 0.1 to 2.0 L/min. The results identified 4 h per day at an aeration rate of 1.0 L/min as the optimal condition, achieving removal efficiencies of 97.2% for COD, 98.4% for NH₄⁺-N, and 90.6% for TN. However, increasing the aeration rate to 2.0 L/min led to a decline in TN removal efficiency due to excessive oxygen reducing the anoxic zones necessary for effective denitrification, underscoring the need to balance oxygen supply for simultaneous nitrification and denitrification. Additionally, Alvarado, Vesvikar [13] emphasized that energy-efficient operation is achievable with fewer aerators, reinforcing the importance of identifying the threshold at which aeration becomes energetically inefficient or even hydraulically disruptive. Moreover, further increasing the aeration rate could potentially induce turbulence, disrupting the original plug flow pattern and transforming the CW into a completely mixed flow system, resembling a CSTR. Therefore, it can be inferred that there is a threshold for improving hydraulic efficiency by increasing the aeration rate. This threshold requires further investigation under different design conditions, particularly through tracer studies and multiphase CFD modeling, to define optimal energy–performance tradeoffs. Additionally, since aeration is an energy-intensive process, a comprehensive cost–benefit analysis is recommended to evaluate the specific benefits of increased aeration.

In this study, a configuration with seven baffles was used to create multi-stage AO conditions that better reflect practical applications. However, the numerical improvements in HE achieved (92.97–95.94% improvement across all scenarios simulated) through aeration appeared relatively limited. This may because the hydraulic performance without aeration was already highly optimized due to the baffles, with HE and MDI values approaching those of ideal plug flow conditions. Therefore, to determine whether the aeration strategy has the same impact on CWs with different HE, and also to identify the threshold for aeration rate, future studies should simulate a broader range of scenarios. These should include configurations with varying baffle configurations and aeration rates, and these simulations should be validated with tracer experiments.

3.3. Future Study

While this study provides insights into the hydraulic effects of aeration rate and positioning in baffled CWs using CFD modeling, there are several limitations. First, the simulations are based on an idealized, lab-scale system with fixed boundary conditions and do not account for factors such as influent variability, temperature fluctuations, or long-term clogging. Second, the model does not incorporate biofilm growth, oxygen gradients, or microbial kinetics, which are critical to understanding pollutant degradation. Third, no experimental validation such as tracer studies or in situ measurements of microbial activity was performed, which limits the generalizability of the findings.

Therefore, further experimental work could include an investigation of biofilm communities, as the positioning and intensity of aeration may influence biofilm development and composition. Key areas to assess include the rate of biofilm formation, its stability under shear stress, and its microbial diversity. Understanding these dynamics could help determine whether changes in biofilm characteristics correspond to variations in treatment efficiency or process type. Sampling within different reactor stages could provide insights into the type of treatment occurring in specific zones, offering a deeper understanding of how biofilm communities contribute to pollutant degradation.

Moreover, future studies should explore how aeration positioning affects microbial activity and substrate degradation. While HRT reflects the potential for contact between pollutants and bacteria, it does not fully capture spatial differences in air–bacteria interactions. Uneven air distribution may result in localized oxygen gradients, potentially influencing microbial metabolism and digestion rates in different zones of the wetland. Experimental work incorporating microbial sampling and spatial oxygen profiling would provide valuable data to assess how aeration position impacts actual treatment mechanisms.

Another area to be explored is the economic feasibility of these configurations. A cost–benefit analysis could evaluate whether the HE and treatment efficiency improvements achieved by aeration justify the associated energy consumption. This analysis should extend across a broader range of scenarios to identify conditions where baffles alone are cost-effective, where they become cost-prohibitive, and where aeration—or a combination of both—may be necessary to meet more stringent legislative standards or achieve zero-discharge objectives.

4. Conclusions

This study evaluated the effects of aeration rates and positions on the hydraulic performance of baffled HSFCWs using CFD simulations. The results demonstrate that aeration has positive effects on hydraulic parameters, with improvements observed in HE and a decrease observed in MDI. Without aeration, the baseline hydraulic performance was already high (λ = 0.9297, MDI = 1.6087) due to the presence of seven optimally configured baffles. Increasing aeration rates consistently improved hydraulic performance, but the magnitude of improvement was not significant. Aeration further enhanced performance, with λ increasing to 0.9594 (2.97% improvement on non-aerated baffled CW) and MDI decreasing to 1.4000 when applied at the end of a baffle where the flow “bends” around a corner into a new flow channel (Position C) with an aeration rate of 0.3 m³/day. “Bend” aeration proved to be more effective than mid-channel aeration (Positions A and B), as it facilitated better mixing, reduced short-circuiting, and promoted uniform flow patterns throughout the wetland. In contrast, mid-channel aeration resulted in smaller improvements at the same aeration rate because the aeration in the middle of the channel does not affect the low-velocity areas at the corners. These findings indicate that the combination of bend aeration and high aeration rates is crucial for achieving both efficient pollutant removal and improved hydraulic performance. However, considering energy consumption and the design requirements for pollutant removal, designers should adjust the aeration rate based on practical conditions.