American Oil and Gas Reporter - September 2021 - 58

sample every 0.25 milliseconds.
The 40-level vertical array in the Boxwood
55-1-12 Unit 5PH was deployed
with 50-foot receiver spacing totaling
2,000 feet of array extension. Two horizontal
arrays in the Boxwood 55-1-12
Unit 3H were deployed as a single splitwhip
array with 100-foot receiver spacing
to the toe. The horizontal arrays had
1,300 and 1,400 feet of length, respectively,
and were separated by a 1,500foot
interconnect cable.
For Thresher 55-1-12 Unit A 16H, a
single split-whip array with 100-foot
spacing was deployed to straddle the
well's heel, with one array in the horizontal
section and the other in the
vertical section. The vertical and horizontal
arrays had 1,300 feet and 1,400
feet of length, respectively, separated
by 1,500 feet of interconnected cable.
The standard Geospace DS150 geophones
had magnetic coupling in the
vertical sections and gravity coupling
with no magnetic components for the
horizontal arrays.
Figure 1 illustrates the position of
the treatment and monitor wells. On the
first (toe) half of acquisition, events
were detected primarily on the Boxwood
55-1-12 3H's horizontal arrays. As the
completion progressed, events easily
were detected by the Boxwood 55-1-12
5PH array. Events in the heel-half were
detected by Thresher 55-1-12 Unit A
16H arrays as the completions moved
toward the heels.
All data was natively synchronized
FIGURE 2
Perforation Shot Relocations
Plan View
Depth View
on a single recording system, including
all channels on all arrays. In other words,
every single channel for all the receivers
in all the monitor wells was connected to
and recorded by a single acquisition system,
as if they were a single array, and
merged into a single trace gather in the
processing phase. This specific setup is
critical because the native sample is not
recoverable if each array is recorded independently
and merged. The approach
eliminates the possibility of synchronization
uncertainty.
Processing
Clearly identifiable P- and S-wave arrivals
must be evident from a given array
to be incorporated into processing and
locating that event. For this reason, data
from the array most distant from an event
usually was excluded from multiarray
event location. This meant that events
associated with the toe-half stages excluded
data from the Thresher 55-1-12 Unit A
16H array. Similarly, horizontal arrays
in the Boxwood 55-1-12 3H were used
to process heel-half stages only if the
sources generated clearly identifiable Pand
S-wave arrivals.
Triggering is the first step in the proprietary
processing workflow. It detected
microseismic events in the continuously
recorded data. By applying image processing
techniques, the algorithm automatically
classified the continuous data
and categorized it based on noise level.
Data segments with medium to low noise
are likely to contain microseismic events
with visible P- and S-waves.
The highest-quality events' direct Pand
S-wave arrival times were used for
velocity model building. The key differentiator
of our processing is that we built
the velocity model based on both perforation
shots and microseismic events.
The advantage of incorporating microseismic
events to constrain the velocity
model is that it significantly increases
reservoir sampling compared with the
traditional method of relying solely on
perforation shots.
Events carry velocity model information
along their ray paths. By adding
them to the velocity model building, sampling
of velocity variations in different
azimuths and geologic formations is improved,
eliminating extrapolation of the
observed reservoir character along many
different azimuths that could introduce
significant event location error.
This allows the construction of a layered
azimuthally anisotropic velocity
model that fits the travel time data of
both perforation shots and events. During
post-processing, additional high-quality
events are incorporated to further constrain
the velocity model, and all the events are
reinverted using this final velocity model.
To eliminate unknown origin times, only
sources with both P- and S-waves picked
in at least one vertical and one horizontal
array were used.
Second Path
The second way events are located is
using joint inversion or tomography and
250 ft.
250 ft.
58 THE AMERICAN OIL & GAS REPORTER
American Oil and Gas Reporter - September 2021

Table of Contents for the Digital Edition of American Oil and Gas Reporter - September 2021
