Dangerous soil liquefaction can occur away from earthquake epicentres in drained conditions – Physics World

Dangerous soil liquefaction can occur away from earthquake epicentres in drained conditions – Physics World

Time Stamp: November 7, 2023 1:52 PM

Liquefaction in New Zealand
Floating on land: this storm drain was pushed up through the road in Christchurch by liquefaction caused by the 2010 Canterbury earthquake. (Courtesy: Martin Luff/CC BY-SA 2.0)

Contrary to conventional wisdom, soil liquefaction during earthquakes can occur away from epicentres, in drained conditions, and at relatively low seismic energy density levels. The finding by an international team of researchers could allow us to better assess and prepare for earthquake hazards.

One of the most catastrophic and unsettling of earthquake-related hazards is soil liquefaction. This occurs when seismic shaking temporarily increases the space between individual soil grains, causing a loss of solidity. The soil starts to behave like a viscous liquid, into which vehicles, buildings and other structures can sink. At the same time, buried infrastructure like pipelines can “float” to the surface (see figure). Liquefaction can also cause the ground to spread and crack, and even trigger landslides.

While soil liquefaction can be a devastating effect of an earthquake, it can have useful applications. Civil engineers deliberately induce liquefaction to improve soil quality before construction and minimize the risk of seismic liquefaction. This can be done by blasting, dynamic compaction and vibroflotation, which involves a large vibrating probe.

Undrained conditions

Traditionally, seismic liquefaction has been associated with undrained conditions (soil that does not naturally drain of water) near the epicentres of earthquakes. However, geoscientists have also observed liquefaction occurring away from the epicentre with lower levels of seismic energy.

“This is quite a common scenario,” explains Shahar Ben-Zeev, a seismologist at the Hebrew University of Jerusalem. For example, he notes, “many of the liquefaction events that occurred during the famous Canterbury 2010–2011 earthquake sequence that caused an enormous amount of damage in Christchurch, New Zealand, occurred in the far-field, under very low seismic energy density input.”

To understand how this is possible, Ben-Zeev and colleagues did both grain-scale simulations and physical experiments on the response of layers of water-saturated, cohesionless grains to horizontal shaking. The physical experiments were undertaken in a transparent box, within which an array of pressure transducers allowed measurements of both grain motion and pore pressure.

Interstitial fluid flow

The researchers found that, even in drained conditions, seismic shaking can trigger interstitial fluid flow within soils, leading to the build-up of excess pore pressure gradients and, as a result, the loss of soil strength. Drained liquefaction was seen to unfold rapidly — guided by the movement through the soil of a compaction front at a speed that is constrained by the rate of seismic energy injection.

“The classical undrained mechanism is perceived as a cumulative process, i.e., the pore pressure rises gradually over time,” explains Ben-Zeev. However, he adds: “In the drained scenario, the pressurization is rapid and more instantaneous. Accordingly, we found that the control parameter for drained liquefaction is the seismic power (the rate of the seismic energy density input into the soil).”

The findings, the team noted, also have implications for how we interpret liquefaction-related geological features associated with past earthquakes that have not been measured using seismic instruments.

“Decision and policy-making procedures regarding earthquake preparedness rely on earthquake catalogues, mainly the reoccurrence time interval of a certain earthquake magnitude in a region, Ben-Zeev explains. One way to construct a catalogue that goes back before  instrument records, he notes, is to examine soft-sediment deformation in the geological record.

“If evidence of soil liquefaction events is found, it is possible to calculate ground motion parameters that triggered liquefaction, and then to constrain epicentral distance and magnitude,” he says. “Our study, which showed that liquefaction can be initiated under relatively low intensity shaking, calls for the re-examination of possibly overestimated paleo ground motion.”

Not fully explained

Oliver Taylor, a geotechnical engineer with ECS Limited who was not involved in the study believes that the work is significant: “[Ben-Zeev and colleagues] provide a thorough insight into soils that liquefy outside the classical undrained regime. This is something that has been observed in-situ, yet not fully explained by our current understanding.”

However, Taylor notes that the team only tested the loosest possible soil condition on an uncompacted uniform sand. “The issue with this,” he adds, “is that it only creates the ‘worst-case’ scenario from which the results are ‘validated’ – and may not be representative of the in-situ conditions where low energy-density liquefaction was observed”.

Calling the study “very interesting”, Chi-Yuen Wang − an applied geophysicist at the University of California, Berkeley – points out that it is “unclear why [the] simulation did not consider the compressibility of the porous soil, given that the latter is the major component of storage of soil at shallow depth, which controls the evolution of pore pressure.”

With their initial study complete, Ben-Zeev and his colleagues have been using the same theoretical framework to explore the mystery of how soil liquefaction can occur many times at the same location. This is not expected to occurs because the initial episode should densify the soil and prevent re-liquefaction in the future.

The study is described in Nature Communications.

Time Stamp:

More from Physics World