Laboratory testing of pipe splitting operations

Pipeline replacement 'on line' has been practised for many years using pipe bursting technologies, which typically involve multiple fracture and outward displacement of fragments of the existing pipeline using either pneumatic or hydraulic means. One of the most important considerations for the design of a pipe bursting operation is the pattern of ground displacement caused, and consequently the possible damage to existing adjacent services and structures. Due to the requirement for overburst, the magnitude of the typical outward pattern of displacements caused by pipe bursting following construction is much smaller than the temporary displacements that occur while the process is being carried out. The ground movements during construction will often, therefore, provide the worst case for design. However, this pattern of behaviour is complicated by the possibility of residual ground settlements if the works are carried out in loose granular soils or in soft cohesive deposits in which positive pore water pressures are generated. The pattern and magnitude of displacements have been shown by the first two authors to be dependent on several parameters following a comprehensive programme of laboratory modelling and field trials. Parallel work by Advantica Technologies Limited (formerly research, technology and engineering arms of British Gas) has resulted in the publication of tables and charts detailing 'safe working distances' for pipe bursting in relation to cast iron gas mains. Prediction of the ground displacements is clearly vital for the safe operation of these replacement techniques. Pipe splitting is a more recently developed technique that has been proposed for 'on line' replacement of ductile iron and steel pipes. This technique has advantages associated with the continuous and pseudo-static, rather than discontinuous and generally dynamic, form of expansion employed. Prior to its adoption in routine industrial practice, however, it is important that differences in the patterns of displacement to those of pipe bursting are known. To this end, Advantica Technologies Limited is currently sponsoring a programme of full-scale laboratory model testing of pipe splitting operations at the University of Birmingham. This paper describes the testing equipment used to determine the effects on the surrounding ground of the pipe splitting operation, details the programme of tests and reports the results of tests conducted in both clays and sands. From the results it was observed that the McElroy Manufacturing Inc. 'Bullet' pipe splitting device produced a clean split in the ductile iron pipe whereas the U-Mole Ltd. 'Clampburster' produced a much more fragmented host pipe. It has also been found that the orientation of the blade of the pipe splitting device has a considerable impact on the resulting ground displacements. The 'Bullet' device allows the orientation of the blade to remain fixed during the operation, whereas the 'Clampburster' device currently allows rotation of the blade during the operation. The maximum displacements resulting from both splitting devices are very similar. (C) 2002 Elsevier Science Ltd. All rights reserved.