Bowland Shale well placement strategy-Part 2:Fracture simulations using a 3D geomechanical model and implications for stratigraphic and spatial horizontal well locations

Hydraulic fracture modelling is a key component of a shale reservoir well placement strategy as it provides an indication of the typical lengths and heights of stimulated fractures and of the changes to the stress environment in which these are propagating.However,spatial and stratigraphic variations in the stress and geomechanical properties of shales make accurate modelling a challenging task.For the UK Bowland Shale,stacked horizontal wells targeting multiple stratigraphic intervals could be used to avoid large offset faults in a geologically complex area.However,it is not known how these intervals may respond to hydraulic fracturing and predicting the height and length of hydraulic fractures is necessary in order to assess the likelihood of vertical fracture interference across landing zones or propagation towards major faults.In the case of the former,intervals of high effective stress may be key to containing fractures within their desired target.Using a planar hydraulic fracture simulator,and a 3D geomechanical model incorporating dipping stratigraphy,the issue of predicting hydraulic fracture geometry in the Bowland Shale was assessed through a series of modelling exercises using well Preese Hall-1 and horizontal pseudo-wells.When pre-defined landing zones were targeted,narrow and long transverse fractures around 1 km from the well were simulated.When the simulation design mimicked perforation clusters placed at 12 m intervals along horizontal pseudo-wells,the effects of stress shadowing were acute and resulted in irregular fracture geometries.Furthermore,high effective stress intervals performed efficiently as barriers to vertical hydraulic fracture propagation,reinforcing the feasibility of using stacked production for the Bowland Shale.The modelling results were then used to discuss the possible placement of horizontal wells in a mapped,100 km2 region around well Preese Hall-1,where up to 13 sites could be positioned,with a horizontal well length of around 1.5 km.Finally,by drawing on a wellestablished analogue for the Bowland Shale,it was estimated that up to 195 Bcf of gas could be produced from the 13 locations in the area if three stratigraphic intervals are produced from one location.

An analytical solution for spherical fluid flow modeling in porous media with non-uniform initial pressure considering the supercharging effect

The diffusivity equation is a partial differential equation(PDE)which can be used for fluid flow modeling in porous media.Determining reservoir parameters from pressure data(i.e.,pressure transient analysis)is one of the most important steps in the process of field development.This initial evaluation can be used to make decisions about future developments.Wireline Formation Testing(WFT)is one of the most popular techniques for parameter estimation and has received significant attention in recent years.The main problem plaguing WFT is a phenomenon known as the“supercharging effect,”which essentially refers to mud invasion,and this,in turn,alters pressure distribution across the system.In this study,an analytical solution for fluid flow modeling in spherical coordinates with non-uniform initial pressure is presented.This new procedure takes into account the effect of mud invasion,or,in other words,the supercharging effect.The accuracy of this derivation was validated using previous semianalytical solutions(the Laplace method)in addition to field data.New type curves and dimensionless parameters,which can be used for pressure transient analysis,are also proposed.This procedure is applied to the WFT data that was obtained from an oil field in the south of Iran,and an excellent agreement(less than 10%error)was observed.In addition,there is considerable uncertainty regarding the radius of investigation for spherical flow.This is important as this parameter greatly affects the applicability of WFT.The analytical derivation of this study was used to determine a reasonable value for this parameter as well.