Abstract
The utility and evaluation of cutoff values for net pay or net-to-gross
determination have been hotly debated topics since the 1950s. There are
numerous subtleties to cutoffs, but exactly how the values are calculated has
largely been overlooked. Most cutoff users have been content to use a
regression line to calculate the cutoff value.
We show that cutoffs obtained using a regression line are likely to be
inferior to estimates produced by other methods. When four methods were applied
to two field datasets and compared, regression-based porosity cutoffs were
between 1 and 2 pu different than the values that give the smallest number of
errors. Monte Carlo simulations broadly support the results obtained from the
datasets. One method, the “trial-and-error” method, performed well through most
of the tests, reducing errors by 40% from those obtained using the regression
line-based cutoff.
All cutoff estimation methods have errors, caused by the imperfect
relationships we have between variables, such as porosity and permeability.
This study shows we have a choice of methods. Because the better method can be
easily applied in spreadsheet software, this should be a valuable addition to
the petrophysicist’s toolbox.
Introduction
Net pay (NP) may be defined as any interval containing economically
producible hydrocarbon using a specific production method. This represents the
portion of the reservoir that contains sufficient porosity, permeability and
hydrocarbons for economic exploitation. NP can be interpreted as an effective
thickness that is pertinent to identification of flow units and target
intervals for well completions and stimulation programs [Worthington and
Cosentino(1)]. The associated net-to-gross ratio (NGR) corresponds to the
proportion of the total or gross thickness, which is composed of net pay.
Numerous papers have reviewed and proposed methods for NP and NGR
determination. Snyder(2) covers many of the methods in use up to the early
1970s, which used the self-potential (SP) or gamma ray (GR) logs and core
analysis. More recent proposals include using capillary pressure [Vavra et
al.(3)], probe permeameter measurements [Flølo et al.(4)] and percolation
modelling [Li et al.(5)]. Of the large variety of possible methods, one
approach is much more commonly discussed than any other. This method involves
defining threshold values (or cutoffs) for the characteristics of interest and
their surrogates. These limiting values are designed to define those rock
intervals that show potential to contribute significantly to economic
hydrocarbon production.
© 2010. Society of Petroleum Engineers
View full textPDF
(
2,262 KB
)
History
- Original manuscript received:
26 March 2009
- Meeting paper published:
17 June 2009
- Revised manuscript received:
12 January 2010
- Manuscript approved:
23 January 2010
View Cart
