S4 to Full-Scale Test Correlation: New Insights from an ‘Open Source’ RCP Model

Session 10A
1:40 pm

Patrick S. Leevers, Imperial College London

There has been a resurgence of interest in the correlation between results from lab-scale (S4) and full-scale RCP tests. Because the full-scale method tries to replicate installation conditions, it is regarded as definitive. The more widely used S4 test has to sacrifice some realism to achieve its small scale. S4 critical pressure results are therefore lower than those from the full-scale test by a substantial factor. The material-independent ‘correlation factor’ predicted using a gas-dynamics analysis has been accepted as an effective lower bound for data from many PE grades. However, the applicability of this factor to PA12 materials has recently been questioned.

This paper accepts that there is a case to answer and, for evidence, turns to a new fracture mechanics model of RCP in pipe. Unlike the complex, multi-physics computational models of the 1990s implemented using finite-element or finite-volume methods, we have developed a modular, semi-analytical model. Each aspect of the problem is modelled in a simple, transparent manner, to encourage experimentation by other investigators. The objective is to correctly indicate trends rather than to accurately predict results; nevertheless the results agree very well with those from the earlier computational methods. At this stage the model embodies no new assumptions, but it does account quantitatively for residual strain effects.

We conclude that residual strain originating from external cooling increases S4 critical pressure. Full-scale RCP tests are apparently little affected by residual strain but dominated by the depth, density and compaction of the surrounding backfill. The influence of these parameters does depend on the pipe material, because it is mediated by the volume and velocity of backfill material shifted by the flaring pipe wall during fracture. The implications are significant both for the full-scale test method and for pipeline installation practice.

Patrick Leevers, Reader in Polymer Engineering
Christos Argyrakis, Research Student
Department of Mechanical Engineering
Imperial College London