One of the most severe dangers that underground petroleum of gas pipelines face, is permanent ground deformation (PGD), such as the one applied by the activation of seismic faults that cross it, the lateral spreading of liquefied soils or slope failures (cyclic or plane ground failures). The main reason behind this is that pipelines are (a) linear structures and as such, it is impossible for its axis to not cross one or more fault zones or areas with possible ground failures (landslides, lateral spreading) and (b) because this kind of displacements are much  larger than those applied by seismic motions and are permanent.

 

From the various causes of PGD, this thesis focuses to the design of underground pipelines versus displacements caused by the activation of seismic faults that cross the pipeline axis, due to fact that this kind of displacements cannot be avoided by dealing with the cause of the failure itself (such is the case with ground failures where special design measures can be taken for the local improvement of the ground in the area, the application of drainage network, etc.), but the pipe design must be able to resist the applied deformations from the fault. Figure 1 presents the deformed form of a pipeline that crosses a strike-slip fault with a random angle.

 

The common “traditional” methods for the dealing with the applied PGD due to active fault offset can be categorized according to the main mechanism that they utilize in order to prevent the developed strain in the pipeline:

(a) reduction of the side friction between the pipe and the ground (e.g. use of coatings – geogrids, fill with pumice stone)

(b) reinforcement of the pipeline ability to cope with the applied deformations (e.g. increase of pipe wall thickness, change of steel type), and

(c) reduction of the ground resistance to the lateral displacements of the pipeline near the fault area (e.g. increase of excavation width, construction of external underground culvert from reinforced concrete).

 

Figure 1:  Deformed form of a pipeline that crosses a strike-slip fault

 

All the above mentioned methods can be applied for small to medium fault offsets, up to 2.5-3.0 pipe diameters. Exception is the construction of an external culvert from reinforced concrete, method that can protect the pipeline against very large fault displacements, but is expensive and its cost increases geometrically with the increase of the fault offset. Thus, in the Geotechnical Department of NTUA, an alternative design method with the use of flexible joints is being researched for pipelines at areas where large PGD are encountered. In this research effort, a new analytical methodology has been developed for the estimation of the beneficial effect of flexible joints to the developing pipeline strains.

 

This thesis, as part of this research effort, focused at the parametric investigation and evaluation of the validity of this analytical methodology for pipelines with flexible joints that cross strike-slip faults. In addition, an economic-technical evaluation of the proposed design method has been made, against the “traditional” design methods, in order to define practically its application limits.

 

From the extensive comparison with the numerical analyses that have been made, it is clear that the proposed analytical methodology can accurately estimate the phenomenon and can give accurate results for almost all checked cases. Moreover, from the comparison of the results between the analytical and the numerical methodologies and the relative error of each design parameter, the following application limits for the analytical methodology have been defined:

·         Vertical fault offset larger than one and a half pipe diameter (Df>1.5D)

·         Crossing angle of the pipeline with the fault trace equal or more than 60^o.

 

It must be clarified that the afore mentioned limits are not a practical limitation of the use of the analytical methodology, since the use of flexible joints is proposed only for large ground deformations (>2.5-3.0D). In addition, the use of flexible joints is not recommended when the pipeline crosses the fault with small angle, since in this case the developing strains are mainly due to axial deformations, and thus the use of such flexible joints is pointless.

 

From the economic-technical comparison between the proposed design method with flexible joints and the “traditional” methods, it has been found that their use is beneficial for the pipe and can reduce the developing strains under the limit strain rate of 0.5%, for fault offsets up to 3.3D when the crossing angle is 60^o and up to 10D for vertical crossing angle. Thus, the proposed design method has the largest beneficial effect between all the “traditional” methods, except from the construction of an external culvert.  The use of flexible joints either every 6m or 8m, may be a more expensive method for large PGDs from the use of pumice stone or the increase of pipe wall thickness, but those methods have an application limit up to 3.5D. Moreover, the use of flexible joints is less expensive from the construction of external culvert in most cases, and easier to be applied, since it does not need additional construction provisions (the flexible joints are just being welded at the selected positions).

 

More specifically, for vertical crossing angle, the increase of the pipe wall thickness is recommended for applied displacements d_f  ≤ 3.5D, where D is the pipe diameter. The use of flexible joints is the most effective and economic solution for larger applied displacements 3.5D < d_f ≤ 10D, while, for even larger displacements d_f > 10D the use of external culvert from reinforced concrete is recommended.

 

For the case of smaller crossing angles, the increase of the pipe wall thickness is recommended for applied displacements d_f  ≤ 3.0D, while for  larger displacements the use of pumice stone as fill material has been found to be the best solution, both technically and economically.