American Oil and Gas Reporter - February 2023 - 61

SpecialReport: Resource Science
To see why perforation-specific differences
matter, consider one of the
most useful results from perforation erosion
diagnostics: identifying variations
in the magnitude of cluster-to-cluster
erosion. Ideally, the variations should be
tiny. The main objective of limited entry
completions is to create a uniform fracture
network originating from each cluster
within a stage, as that should increase
well production and reduce fracture-driven
interactions.
Variations in cluster-to-cluster erosion,
from which fracture uniformity can
be inferred, are measured using a metric
called the Erosion Variability Index
(EVI). EVI is calculated by dividing the
standard deviation of eroded area per
cluster across the stage by the mean
eroded area per cluster across that stage.
Low EVIs are desirable because they indicate
uniformity.
Figure 3 shows the EVIs for each stage
in Well 1. The results are mixed, with some
stages, such as Y8, Y3, Y2 and X5, producing
similar EVI values based on both camera
and ultrasound measurements, but
others, such as Y7, Y4, X6 and Y1, showing
notable differences between the sensors.
When EVI values are used to determine
the effectiveness of the fracture treatment
or the stage design, these differences
FIGURE 3
Erosion Variability Indices For Each Stage in Well 1
Ultrasonic EVI Data Camera EVI Data
Y8
Y5
Y6
Y7
Y4
Y3
Y2
Y1
Y0
X9
X8
X7
X6
X5
X4
0.55
1.28
0.81
0.44
0.49
0.64
0.48
0.94
0.59
0.59
1.00
0.88
0.76
0.35
0.68
0.68
0.50
0.57
0.32
1.32
0.80
0.57
0.42
0.70
0.40
EVI
FEBRUARY 2023 61
0.90
1.04
0.87
1.13
1.17
can be of great significance. For example,
stages Y4, Y5 and Y7 have the lowest variability
based on ultrasound measurements
and would be interpreted as the best stages
or stage designs. But in the camera data,
they have the 13th, eighth and 15th highest
variability, which means they would be interpreted
as some of the worst stages.
To understand why the ultrasonic and
optical measurements yield contradictory
results, consider stage Y7. While the average
perforation area values measured
by ultrasound and camera across this
stage were similar at 0.31 square inches
and 0.35 square inches, respectively, the
individual perforation areas shown in
Figure 4 contain significant differences.
Also of significance are four perforations
unmeasured with the ultrasound sensor.
Most of the missing perforations are toward
the heel of the stage in clusters 8
and 5, with the fourth in cluster 3.
When these individual area values are
converted to eroded area and summed at
cluster level, so proppant placement can
be inferred, the significance of the measurement
difference and missing perforations
becomes more apparent. In Figure
5, cluster 8 dominates the erosion comparison
by a substantial margin based on
camera measurements. The contrast with
the ultrasound measurement results from
the ultrasound sensor's failure to capture
the geometries of two highly eroded perforations
because their perforation holes
contained proppant.
However, the video measurements fall
behind the ultrasound measurements for
clusters 5 and 3, as the area of the perforations
missed by the ultrasound sensor
in these clusters is relatively small, such
that the omitted area is outweighed by the
overall variation in measurements for the
remaining perforations. The subsequent
departure between the eroded area measurements,
driven primarily by the missing
datapoints in cluster 8, leads to a significant
difference in the calculated EVI values
for each sensor technology.
These cluster-level eroded area differences
feed through to heel:toe bias
calculations, with a reversal from a
strong (67:33) heel trend calculated
from camera results to a toe trend
(39:61) with ultrasound.
While the effect of inaccurate measurements
is intuitive, the detrimental influence
of a low number of missing data
points should not be underestimated.
With clusters typically containing threesix
perforations, a single unmeasured
perforation can skew results for that cluster
by a significant percentage of the
total. As seen in this example, EVI and
Stage Number
American Oil and Gas Reporter - February 2023

Table of Contents for the Digital Edition of American Oil and Gas Reporter - February 2023
