19.3: Liquefaction and the Role of Soil Type

Liquefaction

A major hazard in earthquakes is a phenomenon termed liquefaction. Liquefaction involves the intersection of three factors: loose soil, wet soil (near water table), and seismic shaking. When unconsolidated saturated soil is subjected to earthquake movement, an unfortunate behavior is that the water creates pressure between loose soil particles, which then “float” in such a way that the ground loses its ability to hold heavy objects, which then proceed to sink into the newly-mobile ground.

Basically, earthquakes take seemingly firm soil and temporarily turn it into a mud slurpee.

If people built only on bedrock, then liquefaction would not be a concern. However, many inhabited areas are constructed on flat land near water, which is the perfect recipe for liquefaction. The California Geological Survey maintains a digital database of some areas of California susceptible to liquefaction.

Even if the ground is not saturated with water, and thus not likely to experience liquefaction, earthquake shaking can still produce a number of undesired effects. All else being equal, bedrock–actual consolidated rock–behaves the best in earthquakes. The accelerations of the quake transmit through bedrock to buildings, but at least the bedrock jolts and then returns to normal. Bedrock does not keep moving and settling long after the quake has passed.

If the soil is natural but unconsolidated–say, soil deposited in a valley or by an alluvial fan–then the soil will boost the apparent earthquake shaking by continuing to move after the quake has passed. Having your house on consolidated soil will make the earthquake appear bigger and last longer than the quake actually is.

Another possibility is soil that is artificial, and this is by far the worst. Natural soil at least has been around for a while and experienced many quakes to help it settle, but artificial fill may be only a few decades old. When earthquake waves hit this very poor soil, substantial settling may occur, and the shaking of the quake appears augmented compared to what it would be on bedrock.

In the 1989 Loma Prieta earthquake, one of the most heavily damaged areas was the region near Oakland and Alameda, including the Bay Bridge, which partially collapsed, and the I-880 freeway, which completely collapsed, killing 42 people. Despite the lives lost, there was one valuable lesson: the different soils there were recorded by seismographs and provide a textbook example of the role of soil in earthquake damage.

Figure \(\PageIndex{1}\): Shaking in the Oakland-Alameda area in the 1989 Loma Prieta quake, by soil type. "1989 Loma Prieta" by USGS is in the public domain. Access a detailed description.

In Figure \(\PageIndex{1}\):, we find three distinct soil types: bedrock, sand and gravel, and soft mud. The sand and gravel is largely natural, while much of the soft mud is, in fact, artificial fill. The associate seismograms show startling differences between regions only a few kilometers from each other.

If one were on the Oakland hills bedrock, the 1989 Loma Prieta quake would have felt much less than other parts of the Bay Area. In fact, many people doing activities such as driving were unaware a quake had even happened.

If one were on the sand and gravel areas, the shaking was much more intense, and seemed to last longer than the 15 seconds of ground movement the bedrock areas experienced (note how the seismogram from bedrock quickly returns to normal compared to sand and gravel).

Worse still, in the soft mud (artificial) areas, the seismographic record shows even greater intensity and duration. The differences in the apparent shaking are so stark it is as if three completely different earthquakes happened.