One deteriorating slope, four lifelines: applying the Accumulated Geotechnical Deterioration (AGD) framework to climate-driven infrastructure damage, monitoring and operational recovery in the Toledo–Samoré corridor, Colombian Andes

In the steep, rain-soaked mountains of the Colombian Andes, the ground is not a static foundation but a living, breathing entity that changes with the weather. For engineers and managers who build roads and lay pipelines across these slopes, the central challenge is not just knowing where a landslide might happen, but understanding how the soil itself is changing over time. When heavy rains soak into the earth, they weaken the bonds between soil particles, making the ground softer and more likely to slide. However, during dry periods, the soil can regain some of its strength. This cycle of weakening and partial recovery is what researchers call accumulated geotechnical deterioration. It is a concept that shifts the question from "Is this slope stable?" to "How is the condition of this slope changing, and what does that mean for the structures resting on it?" This distinction is vital because a slope that looks dangerous today might behave differently tomorrow if the soil has had time to recover, or conversely, a seemingly safe slope might fail if it has been weakened by repeated wetting and drying.

In the Toledo–Samoré corridor, a narrow valley where four critical pieces of infrastructure run side by side, researchers have documented a dramatic example of this process. Here, a national road, an oil pipeline, a gas pipeline, and high-voltage power lines all traverse the same unstable hillsides. In August 2021, a massive reactivation of four landslides struck this area, affecting all four lifelines at once. Yet, each structure failed in a completely different way. The road lost its base support as the river washed away the soil beneath it; the oil pipeline was exposed and pushed by the moving earth; the gas pipeline was sheared apart by the sudden rupture of the ground; and the power lines suffered from shifting foundations. This paper brings together the geological history of the slope with the specific engineering responses to each structure, showing how a single event in the ground can produce four distinct outcomes depending on what is built upon it.

The story begins with the ground itself. Between 2017 and 2021, researchers measured the soil on one of the largest landslides in the area, known as Landslide D. They found that the soil's ability to hold together, its cohesion, dropped significantly, and the soil layer became thicker and looser. By August 2021, after weeks of heavy rain, the slope failed catastrophically. The ground moved roughly 70 meters, a massive displacement that tore through the infrastructure. However, the story did not end with the failure. In the years that followed, the researchers continued to monitor the site. They found that the soil began to recover. The cohesion increased, the soil became stiffer, and the likelihood of another massive slide decreased. This recovery was not just a theory; it was measured in the lab and confirmed by geophysical tests that showed the ground regaining its strength.

The most striking evidence of this state-dependent behavior came from comparing two wet seasons. In 2021, a period of intense rainfall totaling over 1,150 millimeters in just sixteen days triggered the catastrophic failure. Remarkably, in 2024, the region received almost the exact same amount of rain over a similar period—1,151.3 millimeters. If the slope's response depended only on the amount of rain, it should have failed again. It did not. The ground moved only a few centimeters, and the infrastructure remained intact. This difference proves that the condition of the soil matters more than the rainfall alone. In 2021, the soil was already weakened and ready to slide; in 2024, it had recovered enough to withstand the same deluge.

This understanding of the ground's changing state directly influenced how engineers repaired the damage, particularly for the oil pipeline. The pipeline had been pushed out of the ground and was initially running on a temporary, flexible bypass. To restore it permanently, engineers first tried a high-tech solution: drilling a tunnel horizontally deep underground to bypass the landslide entirely. This method, known as horizontal directional drilling, failed. The geology of the area was too complex; the rock was folded, fractured, and filled with water. When the drillers tried to push through, the rock collapsed, and the equipment got stuck. The attempt to tunnel through the earth proved impossible because the ground was not a solid block but a shattered, water-logged mess.

Faced with this failure, the engineering team turned to a different strategy based on the monitoring data. They realized that while the landslide was still moving, it was doing so very slowly, creeping at a rate of less than 12 centimeters per year. Instead of trying to stop the landslide or tunnel through it, they decided to build a protective shell around the pipeline. They dug a trench and installed a corrugated steel casing, placing the rigid oil pipe inside it. This design allows the steel casing to move with the slow creep of the ground while protecting the pipe from being crushed or bent. The system includes inspection windows so engineers can watch the movement and adjust the casing if the ground starts to move faster. This solution, built in early 2024, successfully returned the oil pipeline to normal operation.

The other lifelines tell a different story of recovery. The national road was rebuilt on a new embankment, but it remains vulnerable to the river washing away its base, a problem that requires constant monitoring of the water flow. The gas pipeline, which was sheared apart in 2021, is still operating on a temporary flexible bypass. Unlike the oil pipeline, a permanent fix has not been implemented because the gas line crosses multiple landslide zones, and the cost and complexity of stabilizing the entire hillside are too high. The power lines were restored after the emergency, but the exact mechanism of their failure remains a hypothesis based on the ground movement, as they were not instrumented with sensors.

The value of this research lies in connecting the dots between the soil and the structures. By tracking the soil's strength and stiffness over time, the researchers provided a clear picture of the ground's condition. This information allowed the engineers to make a calculated decision to reconnect the oil pipeline at the same site, using a design that accommodates the slow, monitored movement of the earth. It also explained why the gas pipeline could not be fixed in the same way. The study demonstrates that infrastructure management in a changing climate requires more than just reacting to disasters; it requires a continuous understanding of how the ground is deteriorating and recovering. The ability to distinguish between a slope that is about to fail and one that is merely wet and recovering is the key to keeping these vital lifelines running.

The path forward involves using this knowledge to build better warning systems. The researchers propose combining their ground measurements with satellite data and new sensors to create a digital model of the entire corridor. This would allow operators to see the state of the ground in real time and make decisions before a failure occurs. While these advanced tools are not yet in place, the core lesson has already been proven: the ground is not a fixed backdrop for human engineering. It is a dynamic system that changes, and by listening to its signals, we can build solutions that survive with it.