We performed scaled model tests to study the relation between completion and treatment parameters and the near wellbore geometry for highly deviated, cased perforated wellbores.

We have instrumented our casings to observe the moment fluid flow around the casing causes a significant increase in the amount of pressure acting on the wall of the borehole. When fluid flow around the liner occurs extensively before fracture initiation, this leads effectively to an open hole situation at the moment of initiation. Otherwise the perforations will be the only possible initiation sites.

In our experiments with a perforation phasing of 90° the perforations on top of the borehole acted in most cases as initiation sites. We performed stress calculations with a simple analytical model to understand the influence of the perforation angle on initiation pressure and fracture propagation.


Reduced communication between well and fracture may hamper propped fracturing from a deviated well. Several workers1–6 have performed laboratory experiments to investigate the fracture geometry near deviated wellbores. They found a fracture starting from one perforation might fail to link-up with adjacent perforations because it turns away from the wellbore directly and avoids the other perforations. During the initiation stage multiple fractures may initiate from one perforation. When the fracture turns from the initiation plane towards the preferred fracture plane the fracture may break up into multiple fractures. These multiple fractures partly overlap and compete with each other. This reduces the width of each separate fracture strongly. Failure to link-up neighboring perforations and the occurrence of multiple fractures reduce the communication between the fracture and the well.

To avoid the risk of fracture initiation from the wall of the wellbore, one may shoot the perforations in all directions. At least some of the perforations will be positioned in, or close to the preferred fracture plane. Fractures initiate from these perforations and the proppant can pass through them from the wellbore to the fracture. The spacing between the perforations in one direction increases when perforations are placed in all directions. This reduces the probability of link-up of starter fractures.

We performed a number of hydraulic fracturing experiments in which we varied the perforation phasing. We placed the perforations in a row, or with a phasing angle of 90°. In our experiments we varied also the product of flow rate and viscosity. To investigate the sensitivity of our results to changing field conditions, we varied the horizontal stress contrast.

In this paper we will first give a brief description of the stress distribution near the wall of the wellbore to obtain some understanding as to which perforations act as initiation sites. Then we describe the experimental set-up and the configuration of our models. Next we show the results of the experiments and discuss the near-wellbore fracture geometries of these experiments.

Experimental set-up

The experiments presented in this paper differ from most of the published experiments1–6 in two ways. In the first place, we cemented the casing while the block was loaded. This difference in cementing chronology has a large impact on the contact stress between the casing and the block. Weijers et al.7 showed a large impact of the cementing chronology on the near wellbore fracture geometry.

We scaled the experiments according to de Pater et al.8 We injected a fracturing fluid with a very high viscosity. As model material we chose a material with an extremely low permeability and a low toughness.

This content is only available via PDF.
You can access this article if you purchase or spend a download.