Abstract
The formation of channels in propped fractures, in the presence of fibers, has been demonstrated experimentally under a wide variety of conditions. The presence of channels can have a significant impact on the effective fracture conductivity and consequently the productivity of a given well. While single channels are observed in experiments, a question is how many channels may be present in real proppant packs and how long do they grow? The length is especially important when proppant flowback control additives are used as "tail-ins" with the last part of the proppant pack. To complement the experimental study, the understanding of channel propagation and the interaction between multiple growing channels was modeled using finite difference codes.
Modeling was performed using a 2-dimensional explicit finite difference program. The formation of channels was predicted by using two parameters – a critical flow rate and a curvature limit. Results demonstrated that the curvature limit is a critical parameter in determining the stability of channels. Calibration of the model with laboratory experiments with proppant and fibers provided clear explanation as to why channel formation provides proppant pack stability at high flow rates.
For a given flow rate, channels grew until a stable length was attained; increasing the flow rate would cause the channel to grow further until a subsequent stable length was attained. Multiple channels were generated but a dominant channel grew once the length of the channel was greater than the spacing between channels. The presence of one, or several widely spaced, long channels will have a much larger impact on productivity than many short channels. Bilinear flow conditions resulted in channel growth to a stable equilibrium length. This is expected to occur in real fractures. A consequence of this behavior is that fractures should be flowed back at high rates to grow channels that will then be stable at subsequent production rates.
