⚡ electrical engineering

Simulation water movement through the GCL cover using a physical model and HYDRUS-1D software

This study evaluates the performance of HYDRUS-1D software in simulating water movement through various Geosynthetic Clay Liner (GCL) configurations, finding that while the model accurately captures flow trends with high efficiency, it systematically underestimates absolute water volumes due to its failure to account for the hydrophobicity of geotextile layers.

Original authors: Kaveh Ostad-Ali-Askari, Mohahammad-Mehdi Bideh, Mohammad Shayannejad, Mozhde Ramezani, Seyed-Mehdi Hejazi, Peiman Kianmehr, Amirreza Nemati Mansour

Published 2026-08-12

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Kaveh Ostad-Ali-Askari, Mohahammad-Mehdi Bideh, Mohammad Shayannejad, Mozhde Ramezani, Seyed-Mehdi Hejazi, Peiman Kianmehr, Amirreza Nemati Mansour

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the ground beneath our feet as a giant, invisible sponge. For decades, people thought this sponge was so good at cleaning up messy spills that if we buried trash in a landfill, any liquid leaking out (called leachate) would just get filtered clean as it soaked into the earth. But scientists eventually realized that this "sponge" isn't magic; if the trash is too toxic or the leak is too fast, the sponge gets overwhelmed, and dirty water can poison the groundwater we drink. To stop this, engineers invented a special "raincoat" for landfills called a Geosynthetic Clay Liner (GCL). Think of a GCL as a high-tech sandwich: two layers of fabric (like a sturdy mesh) holding a thick layer of super-absorbent clay in the middle. When this clay gets wet, it swells up like a sponge, sealing itself tight to block water. But here's the tricky part: how do we know exactly how much water will get through this sandwich before we build a massive landfill? We can't just wait and see, because by then, the damage might be done. That's where computer models come in. Scientists use software to create a "digital twin" of the landfill, simulating how water moves through these liners to predict if they will hold up.

In this study, a team of researchers decided to put this digital twin to the test. They built a physical experiment with twelve different versions of these clay sandwiches, changing up the ingredients: they used two types of clay (Type A and Type I), two types of fabric (PP and PET), and three different amounts of clay weight (4, 4.5, and 5 kg per square meter). They ran water through these samples in a lab, measuring exactly how much leaked through. Then, they fed the data into a famous computer program called HYDRUS-1D to see if the software could predict the results as well as a human could.

The results were a mix of "great job" and "almost perfect." The computer model was incredibly good at guessing the pattern of how water moved. When the researchers compared the computer's predictions to the real lab measurements, the match was strong, with the model getting about 97% of the infiltration (water soaking in) trends right and 81% of the drainage trends right. However, the computer wasn't perfect at guessing the exact amount. It consistently guessed that slightly less water would get through than actually did in the lab. The researchers found that the model underestimated the volume of water by a small but noticeable margin. They believe this happened because the computer treats the fabric layers like normal, water-loving (hydrophilic) soil, but in reality, the synthetic fabrics in the GCL are water-repelling (hydrophobic). It's like the computer thinks the fabric is a thirsty towel that soaks up water immediately, while the real fabric is more like a waxed jacket that fights the water for a moment before letting it through.

Despite this small "glitch" in the volume numbers, the study confirmed that the computer model is a powerful tool for engineers. It proved that the type of clay matters a lot: the liner made with Type A clay (which has more of a mineral called montmorillonite) was much better at blocking water than the Type I clay. Surprisingly, the type of fabric used (PP vs. PET) didn't make a significant difference in how well the liner blocked water. So, while the computer model needs a tiny tweak to account for the "water-fighting" nature of the fabric, it successfully showed that these clay liners are effective barriers and that we can rely on simulations to help design safer landfills without needing to build a dozen physical prototypes for every single project.

Technical Summary: Simulation of Water Movement Through GCL Covers Using a Physical Model and HYDRUS-1D

Problem Statement
Groundwater contamination resulting from leachate leakage in landfill sites remains a critical environmental challenge. While traditional views once suggested that soil layers could naturally treat leachate, subsequent research has demonstrated that landfills often pose significant risks to groundwater resources. To mitigate this, Geosynthetic Clay Liners (GCLs) have been widely adopted as effective hydraulic barriers. However, the performance of GCLs depends on complex interactions between bentonite types, geotextile layers, and mass distribution. Furthermore, while numerical models like HYDRUS-1D are commonly used to simulate water flow, their ability to accurately predict water movement through the specific hydrophobic and heterogeneous layers of GCLs requires rigorous validation against physical experimental data.

Methodology
The study employed a dual approach combining physical experimentation and numerical simulation:

Physical Modeling:
- Specimens: Twelve distinct GCL configurations were tested, comprising two types of sodium bentonite (Type A and Type I), two types of geotextiles (Polypropylene [PP] and Polyester [PET]), and three bentonite mass per unit areas (4, 4.5, and 5 kg/m²).
- Characterization: Bentonite samples were analyzed for chemical composition (XRD, XRF, Flame Photometry, Titration) and free swelling index.
- Permeability Testing: A custom-designed permeability testing device was utilized to measure hydraulic conductivity under a constant pressure head of 30 kPa (simulating landfill bottom conditions) for 6 hours. Tests were conducted in triplicate for each configuration.
- Data Collection: Cumulative infiltration and drainage water volumes were recorded at various time intervals.
Numerical Simulation (HYDRUS-1D):
- Model Setup: A 1D soil profile was constructed to match the exact thickness of the GCL samples.
- Calibration: The Levenberg-Marquardt optimization algorithm was used to calibrate the van Genuchten-Mualem hydraulic parameters ( $\alpha, n, K_s, \theta_r, \theta_s$ ) using data from the first experimental replicate. The pore-connectivity parameter ( $l$ ) was fixed at 0.5.
- Validation: The calibrated parameters were applied to a direct (forward) model to simulate flow for the second and third replicates without further adjustment.
- Performance Metrics: Model accuracy was evaluated using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Coefficient of Determination ( $R^2$ ), Modeling Efficiency (EF), and Coefficient of Residual Mass (CRM).

Key Results

Hydraulic Conductivity: Statistical analysis (ANOVA) revealed that bentonite type and mass per unit area significantly influenced saturated hydraulic conductivity ( $p < 0.01$ $p < 0.01$ ), whereas the geotextile type (PP vs. PET) had no significant effect.
- Type A vs. Type I: Contrary to initial expectations based on mineral content, samples with Type A bentonite (higher montmorillonite content and free swell index) exhibited lower hydraulic conductivity (better performance) than Type I samples. For instance, the PA5 sample (Type A, 5 kg/m²) showed the lowest conductivity (0.05 mm/min), while TI4 (Type I, 4 kg/m²) showed the highest (0.18 mm/min).
Model Performance:
- Correlation: There was a strong correlation between simulated and measured trends. $R^2$ values for infiltration ranged from 0.98 to 0.99, and for drainage, from 0.9973 to 0.9998.
- Efficiency: Modeling efficiencies (EF) were 0.81 for drainage and 0.97 for infiltration.
- Error Analysis: The model demonstrated a systematic underestimation of absolute water volumes. The RMSE was 1.45 for drainage and 4.71 for infiltration. The Coefficient of Residual Mass (CRM) was negative (-0.06 for drainage, -0.26 for infiltration), confirming this tendency to underestimate.

Significance and Claims
The paper asserts that while HYDRUS-1D provides a robust estimation of overall flow dynamics and is highly effective for simulating water movement in GCLs, it has specific limitations regarding the physical properties of synthetic materials.

Hydrophobicity Limitation: The authors attribute the systematic underestimation of water volumes to the model's reliance on conventional soil hydraulic functions (Mualem-van Genuchten), which assume hydrophilic porous media. The model does not fully account for the hydrophobic nature of the PP and PET geotextile layers, which resist initial wetting until a breakthrough pressure is reached.
Practical Implications:
- Material Selection: The study concludes that the mineralogical properties of the bentonite (specifically Type A) are the dominant factor in barrier effectiveness, outweighing minor variations in mass per unit area.
- Geotextile Independence: Since geotextile type showed no significant impact on hydraulic conductivity, engineers are free to select geotextiles based on mechanical survivability, chemical resistance, or cost without compromising hydraulic integrity.
- Screening Tool: Despite the minor discrepancies caused by hydrophobicity, the high statistical efficiency validates HYDRUS-1D as a powerful preliminary screening tool. It can significantly reduce the time and financial costs associated with exhaustive physical permeameter testing in large-scale geotechnical projects.

Conclusion
The research successfully validated the use of HYDRUS-1D for GCL simulation while highlighting the need to interpret results with an understanding of the material's hydrophobic characteristics. It establishes Type A bentonite as the superior choice for landfill liners among the tested configurations and confirms that geotextile selection can be decoupled from hydraulic performance concerns.

More like this