Summary
Experience shows that high-performance fractures (HPFs) may retain
near-unit-flow efficiency (equivalent to zero skin in a vertical well) and
rarely fail, even in highly deviated wells. This may be partly because overly
simplistic models of the well flow behavior lead operators to maintain wells at
lower production rates than could have been achieved for the same amount of
injected proppant with a vertical-well-completion design. Rigorous models that
account for widely accepted rock-mechanics fundamentals indicate that the
fracture-to-well connection is compromised in deviated wells and lead to
questions concerning whether the bulk of the flow to the well actually passes
through the fracture.
Distributed volumetric sources are used in this model to rigorously model a
wide variety of possible fracture geometries such as an expanded wellbore
because of halo effect, flow strictly through the fracture, and combined flow
through a single fracture and to remaining flowing perforations not connected
to the fracture. The model also includes turbulent-flow effects that may occur
for radial-flow conditions in the fracture plane or in the reservoir opposite
wellbore sections not connected to the fracture, along with high-velocity flow
through the perforation tunnels. It also computes the effective flow area at
the resulting face between the reservoir and the completion to check whether
flow velocity exceeds conditions that would risk production of reservoir fines,
and it estimates the screen-flow velocity on the basis of the number of flowing
perforations.
This comprehensive view of the HPF completion enables a thorough analysis of
the risks of flowing the well at high rate. Field examples show that the new
model depicts real field conditions in calculating the total skin, flow
fractions, and the flux for HPF completions in high-rate gas and oil wells.
Complete inflow-performance behavior for likely flow patterns for HPF wells
in oil and gas reservoirs is provided.
© 2010. Society of Petroleum Engineers
View full textPDF
(
1,189 KB
)
History
- Original manuscript received:
1 August 2009
- Meeting paper published:
5 October 2009
- Revised manuscript received:
23 April 2010
- Manuscript approved:
29 June 2010
- Published online:
14 October 2010
- Version of record:
17 November 2010