Page 215 - Well Logging and Formation Evaluation

Well Deviation, Surveying, and Geosteering 205

known to be small. Note that the small-scale variation in the FWL depth
is probably not real but due simply to inaccuracy/scatter in the calculated
S w. Also the fact that the FWL appears to go much deeper as the wellbore
rises is almost certainly spurious and a result of inaccuracy in the satura-
tion/height model. However, such a plot could at least be used to confirm
a general dipping of the FWL along the wellbore trajectory and to estab-
lish that the well was some 50ft above the contact.

13.5 ESTIMATING THE PRODUCTIVITY
INDEX FOR LONG HORIZONTAL WELLS

A technique that has often been found to give good results in estimat-
ing PI is as follows. Using the established poroperm relationship, deter-
mine k all along the horizontal wellbore. Integrating this function from
the top of the objective to TD will yield the gross product k*h for the well.
Make a graph of k*h vs. the PI (in b/d/psi drawdown) for a number of
wells already producing in the area. Often a good correlation is found.
This enables one to predict the PI while the well is being drilled, taking
into account sections of the well that will be nonproducing due to poor
permeability. This information may be important because the total length
a well needs to be drilled may be shortened (thereby saving money) if a
PI threshold has been reached above which surface facility limitations will
negatively affect production. It allows comparisons to be made between
wells drilled under different conditions and helps identify formation
damage in wells that produce far below the established trend.