American Oil and Gas Reporter - February 2015 - 56

SpecialReport: Unconventional Resource Science
FIGURE 1
FTA Matrix Porosity Results for 3 (Left)
And 24 Month (Right) Time Intervals
Parameter Influence Identifier
Porosity (%)

Actual

Fuzzy Pattern Recognition (FPR)

Actual

Normalized (Rich Gas-3 Months-Mcf)

Rich Gas-3 Months-Mcf (Actual)

Rich Gas-24 Months-Mcf (Actual)

Normalized (Rich Gas-24 Months-Mcf)

Fuzzy Pattern Recognition (FPR)

Parameter Influence Identifier
Porosity (%)

8
G2-Port (%)

FIGURE 2
FTA Pay Zone Thickness Results for 3 (Left)
And 24 Month (Right) Time Intervals
Parameter Influence Identifier
Pay Thickness (ft)

Parameter Influence Identifier
Pay Thickness (ft)

Actual

Fuzzy Pattern Recognition (FPR)

Actual

Rich Gas-24 Months-Mcf (Actual)

Rich Gas-3 Months-Mcf (Actual)

Normalized (Rich Gas-3 Months-Mcf)

Normalized (Rich Gas-24 Months-Mcf)

Fuzzy Pattern Recognition (FPR)

100

110

120

130
G2-Net Thick (ft)

140

150

160

long-term production.
Once conventional statistical analysis
proved limited, the research then applied
WQA and FTA to this complex dataset.
For WQA, poor, average and good well
qualities were defined according to the best
three and 24 months of production. There
were several regions in which wells with
different qualities overlapped, and because
of this definition, some wells belonged to
more than one class. These class memberships are referred to as fuzzy membership
function, which constitutes one of the technology's unique features.
Integrated fuzzy pattern recognition
was used to deduce understandable trends
from seemingly chaotic behavior. The result of such analysis is continuous curves
rather than a discrete set of steps.
Native Parameters
Two datasets were prepared to perform
the analysis. The first included wells
with 12 months of production history and
56 THE AMERICAN OIL & GAS REPORTER

100

110

120
130
G2-Net Thick (ft)

140

150

the second included wells with production
histories as long as 24 months. Evaluation
considered the impact of native and design
parameters on each well's best three and
24 months of rich gas production.
Native parameters are inherent reservoir
characteristics, and design parameters refer to field measurements that usually are
measured during the operation, including
fluid injection volumes, injection pressure,
and rate. This article discusses the results
of FTA and WQA for the five native parameters of matrix porosity, net thickness, Btu
area, Young's modulus and minimum horizontal stress gradient, as well as the five
design parameters of deviation type, total
number of clusters, injected slurry volume,
injected proppant per stage and inside distance between laterals in the same pad.
When it comes to the parameter of matrix porosity, the Marcellus' relatively
high porosity and low TOC means it is a
system dominated by matrix porosity and
free gas. Therefore, porosity plays an im-

portant role in production. Higher matrix
porosity results in better production both
early and late in well life, although the
trend is sharper early. This is because the
well produces the free gas that exists in the
fracture and inorganic pores in its first
months, which makes the effect of matrix
porosity more significant at production's
outset.
Figure 1 plots the actual data, with gray
dots corresponding to the y-axis on the
right side, in Cartesian scale for reference.
The curved line of dots represents fuzzy
pattern recognition, which illustrates the
changes in dimensionless normalized rich
gas behavior for the first three months of
production for different matrix porosity
values.
Pay zone thickness is a parameter that
varies widely in the Marcellus, ranging
from a few feet to more than 250 feet, with
the thickness generally increasing as one
moves east. The study area's range of net
thickness is 90-170 feet. Differing well trajectories, landing targets and completion
stages make the pay zone fully or partially accessible, which constitutes another important factor in a well's production.
Figure 2 shows FTA identifying a
clear trend of rich gas production expanding according to thickness. The same
trend was observed from the poor to
good wells with thicker net pay in the
WQA results.
As a parameter, Young's modulus is one
of the critical elastic attributes for shale gas
reservoir characterization that indicate
the resistance of rock to elastic deformation under pressure, such that the lower the
Young's modulus, the more flexible the
rock. The Marcellus is among the shale
reservoirs with lower Young's modulus,
therefore, the fracture network's higher closure stress could reduce production significantly. Although increasing closure stress
on network fractures can impact lowermodulus rock severely, the effects may not
be evident during the first couple years of
production.
On the other hand, a large Young's
modulus makes it difficult for the fluid to
produce wide fractures. This will make the
fracture thinner, higher and longer. If we
assume that all created induced fractures

American Oil and Gas Reporter - February 2015

Table of Contents for the Digital Edition of American Oil and Gas Reporter - February 2015