It is mathematically convenient to describe fractures as simple, vertical, planar features of predictable width and hydraulic continuity.  Published conductivity data are used with the premise that laboratory tests provide a reasonable representation of expected flow capacity of actual fractures.  Presuming homogenous reservoirs and predictable drainage boundaries dramatically simplifies reservoir modeling and development planning.  However, all of these assumptions oversimplify a very complex problem.  This presentation will contrast these assumptions against actual observations and investigate the effect of these simplifications upon fracture treatment design and well performance.  

The speaker reviewed 200 published field studies documenting production changes when fracture designs were altered, and 50 reservoirs in which refracs were attempted.  Frequently, the observed production results cannot be explained with simplistic assumptions of fracture geometry, reservoir homogeneity, and simplistic fluid flow. 

This presentation will summarize evidence from a variety of sources demonstrating that:

§   Fracture geometry is frequently complex, with imperfect lateral continuity

§   Fluid flow regimes are complex – often causing pressure losses 100-times higher than predicted from published conductivity data

§   Reservoirs contain heterogeneities such as boundaries, laminations, lenticular sand bodies, faults, stresses, and anisotropic permeability that influence frac performance.

Hydraulic fractures are key to the development of most low permeability reservoirs, yet they are frequently mischaracterized and poorly optimized.  The primary message of this presentation is that unrecognized opportunities exist to improve well profitability.  Challenging our misconceptions and examining actual field production has yielded techniques to improve frac designs - despite the failure of our simplistic models to recognize those opportunities.