With oil spill prevention and pipeline safety top of mind in Canada and other northern regions, C-CORE’s geotechnical team undertook research to evaluate the soil response around a high-temperature pipeline being designed for northern Alberta.

Pipeline trenches are typically backfilled with the excavated native soil, which is loosely placed around the pipe, tamped down with low ground pressure equipment and crowned above the trench to allow for some settlement under its own weight. The material can be expected to become denser as it settles. However, pipeline trench backfill is a little-studied material, particularly in the context of placement during winter conditions; a better understanding of its strength and stiffness is needed to improve risk analysis related to pipeline restraint and movement under loading from the thermal cycle.

Therefore, C-CORE conducted a field program in northern Alberta along a proposed pipeline route. Test sites were chosen to represent variable soil conditions and pipe bend features, as well as for proximity to an existing trenched pipeline (some eight years old), allowing for a comparison of freshly excavated backfill, aged backfill and native undisturbed soils. The tests were performed in the summer and then repeated at the same location in the winter to evaluate the effects of cold weather construction.


The plate load testing apparatus used in the field was developed specifically for this program. Field trials were carried out to test and evaluate the apparatus and testing procedure before its actual use on this project. The testing setup and procedures were loosely based on the standardized test to measure the bearing capacity of foundation soils. The equipment setup was modified for these tests by rotating the equipment to test in the horizontal direction, adding a second bearing plate, and modifying the setup to measure the relative displacement of each plate independently and relative to the surrounding soils. This simple design proved to be very robust, easy to use and effective.

The testing procedure was also successful in evaluating gap formation after cyclic loading. The field team made a small excavation (0.5 m x 0.5 m) by hand, to between 0.7 m and 1.0 m deep, and placed two aluminum plates (305 mm x 305 mm) at the base parallel to each other on opposite sides with a hydraulic ram positioned between them. A load was then applied to push the plates laterally into the soil; the displacement of each plate relative to the surrounding soil and the corresponding applied load on the plates was measured throughout the test.


Overall, the test results show a significant reduction in the strength and stiffness of backfill compared to the native undisturbed soils and that backfill material becomes only slightly more dense as it ages. The significance of this is that pipe–soil interaction within backfill materials will have less soil restraint and greater deformations compared to assumptions based on the properties for undisturbed native soils.

During winter construction conditions, exposed soil stockpiles were observed to freeze quickly when ambient temperatures were between -20oC and -35oC.  Despite best efforts, some frozen material was placed within the trench during backfilling at the elevation of the pipe.  The results from the winter field testing show that partially frozen backfill material is stronger and stiffer than “fresh” backfill, as measured during the summer field testing program.

It is assumed that the strength and stiffness of the partially frozen backfill material will reduce as the material thaws. However, whether it will return to the values measured during the summer field testing program for “fresh” backfill or whether it might better represent the “aged” backfill material testing in the summer program is unknown. The existing pipeline where the “aged” materials were evaluated during the summer testing program was also constructed in winter conditions and then allowed to “age’ for 8 years.