Development Characteristics of Expansive Soils along the Handan Section (Hebei Province, China) of the Middle Route of the South-to-North Water Diversion Project and Their Influence on Slope Stability

Imagine a stretch of earth that behaves like a living thing, breathing in water and swelling, then shrinking back as it dries. This is the nature of expansive soil, a material found in many parts of the world that contains clay minerals capable of absorbing significant amounts of water. When these soils get wet, they expand and can push against foundations, roads, and canal walls with enough force to cause cracks and structural damage. When they dry out, they shrink, leaving behind gaps that can later fill with water and trigger further movement. Engineers and geologists have long known that these soils are tricky, but the real danger often lies not just in the soil itself, but in how water reaches it. In many cases, a layer of soil on the surface might look solid, but hidden cracks or fissures deep underground can act as highways, allowing rain to bypass the surface and soak into the vulnerable layers below, where the soil is most sensitive to moisture.

This dynamic is particularly critical for the massive water-transfer projects that move water across vast distances, such as the South-to-North Water Diversion Project in China. Along the canal in the Handan section of Hebei Province, engineers faced a specific challenge: the canal cuts through a landscape rich in these moisture-sensitive soils. The question was not just whether the soil was present, but how it was distributed, how it would react to the heavy summer rains typical of the region, and whether water could travel deep enough to weaken the slopes holding the canal in place. A team of researchers set out to map the hidden geology, test the soil's strength in the lab, and simulate how rain would move through the ground to answer these questions. Their work reveals a complex picture where the location of deep cracks determines which parts of a slope become vulnerable to water.

The researchers began by looking at the ground itself, tracing the history of the soil along the 80-kilometer stretch of the canal. They found that the expansive soils here are not just a single layer of dirt but a mix of ancient lake deposits and weathered rock. Some of this material sits right on the surface, while a much larger portion is buried deep underground, hidden beneath layers of loess or gravel. The team mapped out an area of about 220 square kilometers where these soils exist, noting that the buried types are far more extensive than the ones visible above ground. They identified three main types of material: claystone from the Neogene period, clay from the Lower Pleistocene epoch, and clay that formed as the older claystone weathered over time. Most of the soil they tested fell into the category of moderately or weakly expansive, meaning it has the potential to swell but not as violently as the strongest varieties. However, they discovered a specific pattern in where these soils are found: the samples they collected were heavily concentrated at elevations between 70 and 90 meters, a height range that coincides with where the canal cuts into the earth.

To understand how this soil behaves when it gets wet, the team took hundreds of samples back to the laboratory. They subjected the soil to a series of tests, measuring how much it swelled when water was added and how its strength changed. The results were clear: as the soil absorbed water, it lost its ability to hold together. In one set of experiments, they tested the soil at different moisture levels and found a critical tipping point. When the water content in the soil rose from 20 percent to 25 percent, the soil's ability to stick together, known as cohesion, dropped dramatically by nearly 75 percent. This means that a small increase in moisture can turn a firm slope into a much weaker one. The tests also showed that the soil's strength depends on its history; soil that had been dried out and then re-wet behaved differently than soil that was simply saturated, highlighting that the path the water takes matters just as much as the amount of water present.

The final piece of the puzzle was to figure out how rain actually reaches these deep, vulnerable layers. In the field, the surface of the canal slope is often covered with a layer of artificial fill, which might seem like a protective shield. However, the researchers suspected that deep cracks, or fissures, could act as shortcuts for rainwater. To test this, they built a digital model of a typical canal slope and simulated rainfall under different conditions. They compared a slope with no cracks to slopes with deep cracks reaching down 7 meters. In the model without cracks, the rainwater only penetrated about 2 meters into the ground, even after days of heavy rain. The water simply soaked into the top layer and stopped. But when they introduced the deep cracks, the story changed completely. The cracks acted as channels, allowing the rain to rush down to the bottom of the slope, reaching the potential slip zone where a landslide could start.

The simulations showed that the speed at which this deep layer became saturated depended on both the intensity of the rain and the location of the cracks. Under heavy rainfall of 3 millimeters per hour, a point near the bottom of a slope with a crack at the front reached saturation in about 100 hours. Similarly, when a crack was located at the rear of the slope, the lower points near that crack also responded rapidly, reaching saturation in approximately 100 hours. If the rain was lighter, it took much longer, or the water never reached that depth at all. Crucially, the location of the crack mattered. A crack at the front of the slope wetted the bottom layer faster than a crack at the back, showing that the specific geometry of the ground controls where the water goes. The researchers concluded that the danger is not a uniform soaking of the entire slope, but a localized weakening where the water finds a path through the cracks. This means that a slope might look stable on the surface while the soil deep down is losing its strength, waiting for the right conditions to fail.

The study does not predict exactly when a landslide will happen, nor does it claim that every slope in this area is doomed. Instead, it provides a clear framework for understanding the risk. The danger arises from the combination of three factors: the presence of moisture-sensitive soil, the existence of deep cracks that allow water to bypass the surface, and the intensity of rainfall that can fill those cracks. The researchers emphasize that while the soil is inherently weak when wet, the real trigger is the water's ability to reach it. This insight shifts the focus from simply looking at the soil type to investigating the hidden fractures in the ground. For the engineers managing the canal, the takeaway is that monitoring and maintaining the slopes requires looking beyond the surface, paying close attention to where cracks might form and ensuring that water is diverted away from these pathways before it can reach the vulnerable layers below. By connecting the geology, the lab tests, and the water flow, the study offers a practical way to assess and manage the stability of these critical waterways.