For the past four years, NOVA Gas Transmission Ltd. (NGTL) has included the backfill shear strength in buoyancy control design for natural gas pipelines built and operated in muskeg soil laden terrain. Incorporation of muskeg soil shear strength has resulted in major cost reductions. In typical muskeg soil laden terrain, the weighting requirement for NPS 4 and smaller pipelines has been eliminated, while the need for weighting of larger pipelines has been significantly reduced. Traditionally, muskeg soil shear strength has not been considered due to an absence of adequate theoretical and performance data. NGTL has constructed a series of laboratory and field studies in an attempt to quantify the contribution of “muskeg soil shear strength.” The results of this research program, when combined with a rudimentary field investigation, are part of NGTL’s current buoyancy control design methodology for pipelines traversing muskeg soil laden terrain. This paper presents a practical methodology used by NGTL for buoyancy control. Design examples are provided to calculate two design components; the maximum tolerable muskeg depth and optimum weight spacing. The paper concludes with a discussion of construction and operational considerations which influence buoyancy control design.

Directional drilling technology as related to the installation of small and large diameter pipelines at water crossings is continually improving. The current technology provides pipeline owner/operators with a valid installation option for both environmentally and geotechnically (i.e. unstable slopes) sensitive crossings. Route selection and design teams need some general information to assess whether directional drilling is feasible at a particular crossing. To aid this assessment, a study was commissioned by NOVA on the general aspects of directional drilling such as profiles, work area size, and current technology limits. Highlights of the study are presented. In addition, six case histories where directional drilling at water crossings was either considered or used are discussed. Finally, the authors' views of the future of this technology within Alberta are offered.

A common challenge for pipeline designers is the placement and safe operation of pipelines within unstable slopes. Consequently, special design and operation procedures must be created to maintain the integrity of the pipeline through its operating life. Nova Gas Transmission Limited (NGTL) has developed a methodology to monitor pipeline integrity in slow moving (creeping) unstable slopes. This methodology uses Pipeline-Soil Interaction models to produce parameters that are in turn placed in Pipeline Integrity Assessment Techniques such as finite element analysis. For slope movements, pipeline integrity is based on pipeline strain criteria that are established from regulated codes or NGTL’s risk-based criteria. The result is that pipe strain can be estimated over time given a particular soil type and predicted ground movement. The ability to predict when a pipeline is reaching a critical strain allows NGTL to effectively quantify the risk and associated cost for various remedial measures based on a given operating life (life-cycle cost). These remedial measures can take the form of strain relieving outages, re-routing of the pipeline, directional drilling, or conventional geotechnical remedial practices (dewatering, slope grading, buttressing, etc.). Two case studies are presented detailing this present state-of-practice methodology at NGTL.

Perhaps the greatest challenge to geotechnical engineers is to maintain the integrity of pipelines at river crossings where landslide terrain dominates the approach slopes. The current design process at NOVA Gas Transmission Ltd. (NGTL) has developed to the point where this impact can be reasonably estimated using in-house models of pipeline-soil interaction. To date, there has been no method to estimate ground movements within unexplored slopes at the outset of the design process. To address this problem, rainfall and slope instrumentation data have been processed to derive rainfall-ground movement relationships. Early results indicate that the ground movements exhibit two components: a steady, small rate of movement independent of the rainfall, and; increased rates over short periods of time following heavy amounts of rainfall. Evidence exists of a definite threshold value of rainfall which has to be exceeded before any incremental movement is induced. Additional evidence indicates a one-month lag between rainfall and ground movement. While these models are in the preliminary stage, results indicate a potential to estimate ground movements for both initial design and planned maintenance actions.