Good forecast for pipelines
David Hodges of Kiwa Quality Services looks at how ongoing research and data collation has been fundamental to identifying pipeline defects and predicting pipes' working lifetimes
A knowledge of the properties of pipe materials combined with improved data processing, enables relatively accurate forecasting of how and when things will go wrong. There have been several examples of failures in pipelines which were believed to have complied with all the standards and guidelines available at the time of manufacturing. Static test methods, often given in standards, do not necessarily reflect the nature of in-service failures, while research using dynamic testing procedures has reproduced failures more closely resembling those seen under operational conditions. Improved data quality and a better understanding of failure processes has helped to reduce premature failures. This has led to design improvements in products where weaknesses are known.
Inspections to assess the rate of physical change are part of the process for quantifying how much longer a pipeline can be expected to operate safely and productively. Determining a reliable value for the deterioration rate of the pipeline and using this to predict the deterioration process in the future helps determine the right course of remedial action.
Early research began by studying strengths and weaknesses of different pipeline materials, applications and ages. This is particularly difficult as pipelines are made in a very wide range of materials such as ductile iron, steel, GRP, cement, PVC and polyethylene. Information was then obtained on the design characteristics, manufacturing processes and test methods employed. In addition, information had to be obtained on service history, including operating conditions, maintenance procedures and records and any pipeline surveys.
Failure can be caused by a number of factors:
- external corrosion,
- internal corrosion,
- impact damage,
- erosion damage,
- incorrect operation conditions,
- loss of ground support,
Research increasingly focuses on the reliability of newer materials such as PVC and polyethylene. This includes crack propagation, the long-term effects of surface damage and the effects on the pipe and quality of the water it contains when the pipes are laid in contaminated soil - all of which can affect the lifetime of a pipeline.
Recent improvements in non-destructive testing technology, such as scanning using magnetic and ultrasound sensors, can result in the precise identification of defect size and location in a wide range of materials in varying operating conditions. The application of such sensors is now the foundation for almost all lifetime predictions.
The probability of failure increases as the pipeline ages. Curves of failure probability versus time can be used as the basis for scheduling further inspections and for planning rehabilitation or replacement by predicting the time of failure if no remedial action is taken. The basic assessment methodology usually uses techniques to determine the defect size that would be present in the pipeline when it went into service and the defect sizes that can be tolerated at the operating pressures. The fatigue life can then be calculated by determining the number of pressure cycles required to enlarge the defect to a size which would fail at the operating pressure. Using this approach it is possible to determine the nature and distribution of defects that could be present in the pipeline and the probability of failure. A study allows the operator to determine which pipeline is at greatest risk from fatigue and schedule the inspection of the pipelines accordingly.
Another approach is to study the existing defect population in the pipeline, the likely future deterioration rates and the potential for reducing these rates through, for example, reduction of pressure, the dosing of corrosion protection additives and the use of lining techniques such as epoxy resin, spray lining or polyethylene liners.
Assessment analysis techniques can also be used to quantify the benefits of repairs, which has led to substantial improvements in the prediction of lifetimes.
There is still a long way to go in pipeline research before we can produce
a computer model showing the effects of service on the lifetime of pipelines.
Much progress has been made but we still do not fully understand what causes
a number of failures. There is a need to identify fully the mechanisms involved
in the ageing process particularly in plastics. We are unlikely to stop the
ageing process altogether but slowing it will delay failure and therefore lifetimes
will be extended. We do not have all the answers yet but we have already gone
a long way in identifying many of the factors involved, and we are now able
to model many of their effects. We may not yet have a pipeline able to last
several hundred years without maintenance but through continued research we
are getting closer to it.