Ask a Geotechnical Engineer: Compaction Testing – Just the Top?

Sometimes we are asked if we can “just test the top one foot of pad subgrade” and nothing more.  The reasoning we hear is that “it will bridge over the soft soils, and it should be fine.”  However, achieving an accurate assessment of site conditions necessitates that we test to the full depth of the fill, and this case demonstrates why.

A homeowner in Fresno purchased a finished lot and had a custom home builder construct a house for him.  During construction, when the house framing, roof, and stucco were completed, a large portion of the house started to settle.  The amount of settlement was up to four inches, with major cracking of slabs and footings.

Compaction tests were taken on the subgrade soil in the area of the cracked slab.  The subgrade compaction was above 90 percent and met the CBC code requirements for engineered fill.

Further investigation was performed with auger borings and deeper compaction tests.  The findings revealed that the site had two feet of compacted engineered fill.  Below two feet, a layer of loose, uncompacted, dry soils was present with thickness ranging from four to six feet and relative compaction ranging from 66 to 84 percent.  These loose soils settled significantly by hydrocompaction through surface water infiltration.  To mitigate the problem, the house had to be supported by underpinning or mini-piles.  A repair cost of around a million dollars was estimated.

This problem and subsequent repair cost could have been avoided if the compaction tests were taken to the full depth of the fill instead of just the top two feet.  Furthermore, the conditions of loose fill could have been discovered if a soil investigation was performed prior to building construction.