Experimental and Numerical Investigation of Stability of Horizontal Boreholes during Horizontal Directional Drilling
This thesis focuses on the stability of horizontal boreholes in sand and saturated clay during Horizontal Directional Drilling. For saturated clay, new criteria for distinguishing tensile and shear failure were developed and case studies were used to support the criteria. New ‘risky’ and ‘dangerous’ zones along the drilling path were defined. Lateral earth pressure coefficient at rest influences the failure mechanisms and an overview of this coefficient was presented including lessons learned from the tests for obtaining this coefficient. Hydro-sand and sandy gravel were used for the tests presented in this thesis and the influence of compaction effects has been investigated. Contrary to conventional understanding, no obvious compaction-induced earth stresses were observed. For sand, a series of medium-scale mud loss experiments have been successfully conducted. The relationships between maximum allowable mud pressure (P_max) and burial depth, pump rate and multi-layer stratigraphy (dense sand overlain by loose sand or dense sandy gravel) were examined. Values of P_max obtained using finite element analyses were close to the P_max results for experiments at different burial depths. Based on a parametric study undertaken using numerical modelling, a new design equation for calculating P_max in uniform sand has been developed. Case studies were also used to support the design equation and a new ‘dangerous’ zone was identified along the drilling path. In addition, stress paths (relationships between radial and hoop stresses at the crown of the borehole) calculated using finite element analyses that deviate from the elastic closed form solution have been investigated. Based on the experiments, P_max is independent of pump rates and dependent on parameters associated with the multi-layer stratigraphy. In practice, it is conservative to use the design equation when dense sand is overlain by dense sandy gravel and a reduction factor was developed for the design equation when a loose layer rests on top of the dense layer. Furthermore, mud travelling paths from the primary infiltration zone around the borehole along shear planes to the ground surface have been examined. Ground surface movement was monitored using Geo-PIV combined with a new post-processing method for diminishing ‘pseudo displacements’ in three-dimensional applications, and it is found that the maximum ground movement occurred just before mud flowed out of the soil rather than the moment when mud pressure reaches a peak value.