Soil Liquefaction During Earthquakes

How Shaking Turns Sand Into a Liquid

Ordinary dry or lightly loaded sand carries its weight through a network of grain-to-grain contacts. Each sand particle presses against its neighbors, and the friction and interlocking at those contact points give the soil mass its strength, allowing it to support the load of a building above. In loosely packed sand, the grains sit in an open, unstable arrangement with relatively large voids between them, similar to a pile of marbles poured carelessly into a jar rather than one that has been shaken down into a tight, compact stack. When a strong earthquake sends cyclic shear waves through the ground, the loose grains are jostled back and forth many times per second. Each cycle of shaking is a small opportunity for the grains to shift into a denser, more stable packing, the same way a loose pile of marbles settles when you shake the jar. Under normal, slow conditions this rearrangement would happen gradually, and any water in the pores would simply drain away as the grains squeezed closer together. But an earthquake's strong shaking lasts only seconds to a couple of minutes, far too briefly for water to escape through the tiny pore channels, especially in fine-to-medium sand where those channels are narrow. With drainage blocked, the grains cannot actually get closer together, because the water trapped in the pores is essentially incompressible. Instead, each cycle of shaking transfers a bit more of the load that used to rest on the grain contacts onto the trapped pore water itself. This is described using the principle of effective stress: the total stress carried by the soil is shared between the pressure in the pore water and the effective stress carried directly by the grain skeleton. As shaking continues, pore water pressure climbs higher and higher, approaching the total overburden stress pressing down from above. When pore pressure nearly equals total stress, the effective stress between grains drops toward zero, meaning the grains are essentially floating in the pressurized water rather than resting against each other. At that point the soil has lost nearly all of its shear strength and behaves like a dense fluid, even though no water was added and no grains were removed.

Why Loose, Saturated, Fine Sand Is the Danger Zone

Not all soils are vulnerable to liquefaction, and the difference comes down to grain packing, grain size, and the presence of water. Loose sand is dangerous precisely because its grains have far to travel before reaching a denser arrangement, which means many cycles of rearrangement are attempted and a large volume change is being resisted by the pore water. Dense sand, by contrast, already sits in a tightly packed, interlocked arrangement. When shaken, dense sand grains tend to dilate, or expand slightly, rather than contract, which actually lowers pore pressure momentarily rather than raising it, so dense deposits are far more resistant to liquefaction even under identical shaking and identical water content. Saturation matters just as much as packing density. Dry sand above the water table has air rather than water filling its pore spaces, and air compresses easily and escapes readily, so shaking simply rearranges dry grains without ever building up damaging pressure. Only when the pores are completely filled with water, as they are below the water table, can the incompressible-fluid mechanism trap the load and drive pore pressure upward. This is why liquefaction hazard maps focus so heavily on the depth to groundwater; shallow water tables in low-lying areas, floodplains, reclaimed land, and artificial fill are consistently the highest-risk zones. Grain size and soil type add a final filter. Fine-to-medium sand and non-plastic silt are the classic liquefiable materials because their pore channels are narrow enough to trap water during the brief duration of shaking, yet the grains themselves are large enough, and lack the cohesive bonding of clay, to be individually mobile. Clay soils are essentially immune to classic liquefaction because their particles are bound together by cohesive, electrochemical forces between flat, plate-like clay minerals, giving the soil mass an intrinsic strength that does not depend on grain-to-grain friction and is not eliminated simply by a rise in pore pressure. Gravelly soils, meanwhile, tend to drain too quickly for pore pressure to build up in the first place, since their pore channels are wide.

What Happens to Structures Above Liquefied Ground

Once a layer of soil liquefies, it can no longer provide reliable bearing support or lateral confinement, and the consequences at the surface depend heavily on what is sitting on or in that soil. Buildings founded on shallow footings can experience bearing capacity failure, sinking unevenly into the liquefied layer because different parts of the foundation are pressing down on soil that has lost strength unevenly. If one corner of a building sinks more than another, the entire structure tilts, sometimes dramatically, even though the building itself may remain structurally intact, since the failure is happening in the ground rather than in the walls or frame. Buried, lightweight structures such as empty or lightly filled underground fuel tanks, sewage tanks, and pipelines experience the opposite problem. Once the surrounding soil liquefies, it behaves like a dense fluid with a unit weight considerably greater than that of the buried structure, so buoyant forces can push the structure upward, a phenomenon engineers call flotation. Manholes, empty swimming pools, and light utility vaults have been observed popping several feet out of the ground during liquefaction events for exactly this reason. On gently sloping ground or near riverbanks and waterfronts, liquefied soil layers can trigger lateral spreading, in which large blocks of overlying, still-intact soil slide slowly downslope or toward a free face such as a riverbank, carried along on top of the liquefied layer beneath. Lateral spreading can displace the ground by many feet, severing pipelines, cracking foundations, and pulling bridge abutments apart. Slopes that are too gentle to fail under normal static conditions can flow substantial distances once their base layer turns fluid. Once shaking stops and pore pressure gradually dissipates as water slowly drains away, the soil regains its strength, but by then the damage, from tilted buildings to displaced roads, has already occurred and typically cannot be reversed without extensive foundation repair or ground improvement.

Famous Historical Cases of Liquefaction Damage

The 1964 Niigata earthquake in Japan produced what remain some of the most widely reproduced images in geotechnical engineering: entire reinforced-concrete apartment buildings in the Kawagishi-cho district tipped over onto their sides, some resting at angles beyond thirty degrees, while their structural frames stayed almost entirely undamaged. The buildings were relatively short and rigid, so once the loose, saturated, sandy fill beneath them liquefied, they rotated and settled as essentially intact rigid blocks rather than crumbling. Niigata, together with the 1964 Great Alaska earthquake in the same year, is widely credited with drawing serious engineering attention to liquefaction as a distinct and predictable hazard, spurring decades of research into how to identify and mitigate it. The 1989 Loma Prieta earthquake in California caused severe liquefaction in the Marina District of San Francisco, an area built largely on loose, saturated sandy fill placed after the 1906 earthquake and for the 1915 Panama-Pacific Exposition. Ground failure there damaged and destroyed numerous buildings and ruptured underground utilities, despite the Marina being located a considerable distance from the earthquake's epicenter near Loma Prieta peak, illustrating how artificial fill and shallow groundwater can create serious hazard even away from the strongest shaking. More recently, the 2011 Christchurch earthquake sequence in New Zealand produced widespread and severe liquefaction across the city, ejecting enormous volumes of sand and silt to the surface through cracks in the ground, a process called sand boiling, and damaging thousands of homes built on the loose, saturated alluvial soils of the Canterbury Plains. The 2011 Tohoku earthquake in Japan similarly caused extensive liquefaction damage in reclaimed land around Tokyo Bay, far from the main rupture zone offshore. Each of these events reinforced the same lesson: liquefaction hazard is controlled less by earthquake magnitude alone and more by local soil conditions, groundwater depth, and how the ground beneath a city was originally formed or filled.

Assessing and Mitigating Liquefaction Risk

Geotechnical engineers evaluate liquefaction potential using a framework generally called the simplified stress-based procedure, which compares the cyclic shear stress that a design earthquake is expected to impose on a soil layer, called the cyclic stress ratio, against the soil's inherent resistance to liquefaction, called the cyclic resistance ratio. This resistance is typically estimated from field tests such as the standard penetration test or cone penetration test, which measure how much a soil resists being probed or driven into, giving an indirect but reliable measure of density and, therefore, liquefaction susceptibility. When the imposed stress ratio exceeds the soil's resistance ratio, engineers classify the site as susceptible to liquefaction under that design earthquake scenario. Where liquefaction risk is identified beneath existing or planned structures, several ground improvement techniques can reduce the hazard. Vibro-compaction and vibro-replacement use vibrating probes to densify loose sand in place, directly attacking the loose packing that makes liquefaction possible. Stone columns and prefabricated vertical drains shorten the drainage path for pore water, allowing pressure to dissipate quickly enough during shaking that it never approaches the critical threshold. Deep soil mixing blends cement or other binders into the soil to create stronger, less liquefiable columns or grids within a loose deposit. For existing buildings where large-scale ground treatment is impractical, engineers sometimes turn to deep foundations, such as piles driven through the liquefiable layer into stable soil or rock below, so that the building's load bypasses the weak layer entirely rather than relying on it for support. Ultimately, liquefaction mitigation is as much about mapping as it is about construction technique. Detailed hazard maps that combine groundwater depth, soil type, and historical shaking intensity let planners steer critical infrastructure, hospitals, and dense housing away from the highest-risk zones in the first place, which remains the most cost-effective way to reduce future liquefaction losses.

Frequently asked questions

Try it live

Everything above runs in your browser — open Soil Liquefaction During Earthquakes and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.