American Oil and Gas Reporter - January 2016 - 76

JANUARY 2016
is still a large total spread and a lot of experimentation with optimum length in
the study area, even though operators
generally converged in mid-2010 on a
10,000-foot perforated interval, as shown
in Figure 2C.
So, the data imply that in the more
mature unconventional basin plays, operator decisions on some drilling and
completion variables are converging to
best-practice median values.

Improving Type Curves
With play-optimized drilling and completion practices, average type curves (by
an assigned acreage grade) generally have
increased since development in the study
area started, as shown in Figure 1.
As the Bakken and Eagle Ford have
matured, year-over-year improvements
have slowed to where even with high
grading of acreage in 2015, there has not
been much difference in their type curves
compared with 2014. However, the
Delaware and Midland basins experienced
significant improvements in 2015 type
curves compared with 2014 curves,
demonstrating an approximately 10 percent
increase in daily production during wells'
peak months.
Some convergence in utilizing best
practices in perforating/completing intervals is taking place, as is evidenced by
a plateauing of completed interval size
for the Eagle Ford, Permian, and Bakken.
However, plots of completed interval
median values for the Delaware and Midland basins show that there is still experimentation going on in the Midland Basin
with regard to optimizing lateral wellbore
lengths, as seen in Figure 2A. This contrasts with what appears to be a convergence on a shorter median lateral wellbore
in the Delaware Basin shown in Figure
2B.
Interestingly, within the Bakken, there

Forecasting Reserves
Decline rates for ultralow-permeability,
low-porosity shale reservoirs-whether
producing liquid or gas-are much higher
than for conventional reservoirs. High
initial production rates usually are followed
by steep declines, which then level asymptotically, approaching the economic
rate cutoff.
Important drivers for the decline behavior exhibited by unconventional reservoirs are rock properties, wellbore spacing,
treatment design, and choke management.
Peak rates are driven highly by initial
reservoir pressure and operational considerations. Somewhat surprisingly, public
data indicate that peak monthly rates are
better predictors of total production in
lower-quality reservoir grades than in
high-quality reservoirs.
This relationship is still being studied,
but may be caused by a lack of perme-

FIGURE 1
Horizontal Well Type Curves
650

550
500

600

450

500

400

450

350

400
350

BOE/d

300

250

200

150

100

50
0

500

450

450
400

400

Bakken

Midland Basin

350

300

BOE/d

250

200

150

100
50

Delaware Basin

550

Eagle Ford Oil Window

76 THE AMERICAN OIL & GAS REPORTER

ability enhancements, such as natural
fracturing that don't exist in lower-grade
acreage, but may exist in better-grade
acreage, and are difficult to normalize.
Operators with lower-grade acreage may
have less variability in predicting estimated
ultimate recovery, and therefore, obtain
earlier visibility into their actual reserve
bases, whereas operators in better acreage
likely need to realize that their reserves
forecasting will be more prone to uncertainty.
Public production data show that peak
rates normalized by perforated length increased over the past few years while
EURs stayed fairly flat, as seen in Figure
3 (Eagle Ford). A steady increase in perforated intervals can be seen.
Estimating ultimate recovery can be
challenging in ultratight, naturally fractured
reservoirs. Arps' method for exponential
and hyperbolic decline with b-factors between zero and one is used widely as an
industry standard, regardless of the fact
that these reservoirs almost always show
b-factors greater than one (superhyperbolic), especially in early years of production.
Since it is valid only for boundary
dominated flow (BDF), the Arps' method
incorrectly predicts for the linear transient
flow regime seen in shale reservoirs. For
BDF reservoirs, the majority of the EUR
is produced after the flow regime leaves
transient conditions. The exact opposite
is seen in shales, where production data
indicate transient flow is in effect for
many years, during which most of the
EUR is produced. BDF conditions are in
effect only after about 70 percent of EUR
is recovered.
Basin Variability
EURs can vary greatly, based on the
type of decline curve analysis (DCA)
model used. Industry research shows that
overestimates can be up to 150 percent
when an inappropriate model and/or insufficient production data are used. With
more representative production data, more
appropriate models have been developed
over the past five years. Unfortunately,
because of the variability in reservoir
properties, natural fractures and operational
choices, there is no one-size-fits-all DCA
model.
Type curves for entire basins can be
misleading for individual operators with

American Oil and Gas Reporter - January 2016

Table of Contents for the Digital Edition of American Oil and Gas Reporter - January 2016