The primary design condition in conventional onshore oil and gas pipeline standards is based on limiting the hoop stress under the maximum operating pressure to a fraction of the material specified minimum yield strength. This fraction is referred to as the utilisation factor and is varied as a function of location, temperature and joint type to achieve a level of conservatism consistent with the failure susceptibility and failure impact of the pipeline.

While this methodology has served the industry well for many decades, recent industry trends have created a need to consider other design methods. For example, the current utilisation factors are not necessarily applicable to the high steel grades (X100 and above) being considered for large diameter pipelines. For these grades, the utilisation factor results in thinner pipe walls, and since susceptibility to major failure threats, such as corrosion and equipment impact, is more influenced by wall thickness than hoop stress, lower overall reliability may result.

This indicates that the utilisation factor is not always the best safety control parameter, and that it cannot be used with confidence outside the conventional range of design parameters for which it has been empirically proven.

Another example is the design of pipelines subject to high bending loads caused by soil deformations due to frost heave and thaw settlement. This type of loading is not adequately addressed in existing design standards, nor is it amenable to the elastic stress-based design approach. To withstand bending loads, line pipe capacity is best represented by its tolerance to longitudinal deformations or strains. Safe strain limits have shown to exceed the material yield point by a significant margin.

Article continues below…
Advertisement

The RBDA approach

The common theme in these examples is the large degree of uncertainty resulting from limited precedent. This uncertainty can be addressed using a class of design approaches called reliability based design and assessment (RBDA), in which safety is measured by the reliability, defined as the probability that the pipeline will not fail within a specified period of time. The pipeline is then designed and operated to meet a specified reliability target considering all applicable threats.

In addition to allowing systematic consideration of uncertainty, the RBDA approach addresses the key limitations of the traditional hoop stress design approach, offering significant advantages. By directly addressing the expected loads and failure threats, resources are targeted where they are most effective in preventing failures.

As reliability is defined as a lifetime requirement, design and maintenance decisions can be considered in combination, leading to more optimal overall solutions. Failure threats are also analysed from basic principles, providing the flexibility to deal with unique loads, new technologies and unconventional design conditions.

Recognising these benefits, the Pipeline Research Council International sponsored the development of a comprehensive RBDA methodology for natural gas pipelines, which is now incorporated as a non-mandatory Annex in the Canadian Standard CSA Z662.

Early applications of the methodology have demonstrated the challenges associated with full acceptance and wide application. To start with, the reliability targets are defined in terms of a calculated failure probability and although these targets are intended to help approach the goal of eliminating all incidents, some see the mere acknowledgment of a failure probability as contradictory to this goal. The probabilistic calculations needed to apply the approach are complex and unfamiliar to many pipeline engineers, resulting in the methodology being thought of as a ‘black box’ and viewed with skepticism.

Most of these challenges can be addressed by developing a simplified version of the RBDA approach, often referred to as the limit states approach. In this approach, deterministic design and assessment rules are developed that meet the reliability targets underlying the RBDA methodology. This creates a simple and transparent methodology that capitalises on the main benefits of RBDA while addressing most of its limitations. The simplified methodology is adequate for many situations, but full RBDA is still required for specialised and high impact applications.

Reliability based and limit states methods have been successfully used in many industries and have become the exclusive basis for codified design of such systems as buildings, bridges and nuclear structures. The limit states approach has also been used as a basis for API and DNV codes for offshore pipelines. Given these precedents and the potential for RBDA to help the onshore pipeline industry address key design and assessment issues, the author believes that the challenges associated with adopting the methodology will be overcome and that it will gain wider acceptance over the coming years.