A limited amount of experimental data exists concerning the axial resistance of pipelines exposed or buried in rockfill. This paper presents some theoretical considerations in the interpretation of buried axial tests together with additional results from some recent testing. The different sources of experimental data are synthesized in order to better quantify the axial resistance afforded by rockfill. Arching theory and finite element analysis (FEA) results from a companion paper (Milewski, 2021) are utilised to help collate the data from disparate sources.


Geotechnical quantification of upheaval and axial restraint are both key aspects of pipeline buckling design. Various methods of pipeline restraint are employed in offshore oil and gas developments including trenching & burial and the use of concrete mattresses. In some regions such as the North Sea, rockfill cover (either on top of a trenched or exposed pipeline) is a popular approach as it affords greater upheaval restraint than natural fill and is readily available from Norwegian quarries.

Limited published data exists for estimating axial resistance in rockfill, and there is a need to collate this data in a coherent way. The first step towards this – accounting for reduced normal force due to a finite berm instead of uniform cover - has been presented in the companion paper Milewski (2021). The present paper seeks to combine all the available experimental data in order to derive trendlines to assist the determination of axial resistance in routine design.

Interpretation of Normal Force

The axial resistance of a pipeline buried in rockfill is the product of a friction coefficient and the normal force surrounding the pipe. To interpret the experimental data it is necessary to consider the definition of this normal force. One key assumption necessary regarding the stress distribution around the pipe is that the presence of the pipe does not affect the stress distribution. In reality the stress distribution will be more complex (e.g. Sheil et al., 2018) but use of a 'nominal' normal force represents a practical and consistent approach for design estimations.

This content is only available via PDF.
You can access this article if you purchase or spend a download.