Fracture control studies for new gas transmission pipelines usually produce a specified minimum Charpy energy, often including “correction factors”, which will ensure that a crack will arrest in the body of the pipe. The basic pipeline parameters such as pressure, pipe grade, diameter and wall thickness will be fixed early in design, and the reservoir and process engineering design will set limits on the extremes of the gas composition.

The inverse case, where the gas composition in an existing pipeline is to be changed from the original design basis, is more challenging. Changes in composition can arise from ageing of the reservoir supplying a pipeline, or opportunities for the operator to generate additional revenue from 3rd party access. Sales gas specification limits for general purpose natural gas transmission often have broad limits, which can be met by a wide range of compositions.

As a wide range of gas compositions can give the same crack driving force, determining the composition limits is a “many to one” problem without a unique solution. This paper describes the derivation of an envelope of richer gas compositions which gave an acceptable probability of crack arrest in an existing pipeline which had originally been designed for a very lean gas mixture. Hence it was necessary to limit the amount of rich third party gas to ensure that the crack driving force did not increase sufficiently to propagate a long running fracture.

Manufacturing test data for the linepipe were used with the EPRG probabilistic approach to derive a characteristic Charpy energy which would achieve a 95% probability of crack arrest in 5 joints or fewer. After “uncorrecting” the high Charpy energy, the value was used with the Battelle Two Curve model to analyse a range of gas compositions and derive an envelope of acceptable compositions. Sensitivity studies were carried out to assess the effects of increasing the temperature and of expanding the limits for nitrogen and carbon dioxide beyond the initial assumptions.

It is concluded that for a specific case it will be possible to solve the inverse problem and produce composition limits which will allow increased flexibility of operation whilst maintaining safety.

This content is only available via PDF.
You do not currently have access to this content.