In-Situ Evaluation of Pipeline Bedding and Padding Spring Constants

Structural response analyses of pipelines using computer models, such as AutoPIPE, incorporate soil “springs” to model the restraint provided by pipeline bedding and padding. These “springs” are referred to as spring constants and are most frequently determined from a limited number of sources in the literature. In some cases representative soil properties are used with theoretical and empirical formulae such as those presented by Nyman (1984). This technique which is referred to herein as the ASCE technique has been used by Alyeska Pipeline Service Company (Alyeska) engineers in their calculations of structural response of the buried pipeline along the Trans Alaska Pipeline System (TAPS) (Hart et al, 1998). This paper describes a field study at a pipeline inspection dig in which spring constants were determined directly from a series of field tests to develop a better understanding of the pipeline restraint the bedding and padding is providing for the pipe. Four other papers which cover other aspects of the work performed at this site are also a part of this conference (Hart et al, 1998, Norton et al, 1998, Stevick et al, 1998, Tonkins et al, 1998). The field work included the drilling of multiple boreholes in which drive samples were taken at about 1 to 2 meter (3 to 5 foot) intervals. In each borehole, pressuremeter tests were also conducted at the same intervals as the drive samples. At two levels in the open pipe trench plate, load tests were conducted. Both nuclear and sand cone density tests were made at several levels in the pipe trench. Results of these tests were correlated to each other. Stress strain relationships were developed from the pressuremeter test and plate load test data which were used independently to develop spring constants. Spring constants were found to vary with the strain level in the soils and were correlated to the drive sample blow count data. Recommended methods for estimating spring constants are presented.