Oil is produced from the Manderson Field, in Wyoming's
Big Horn County, from a fracture system with possible significant lateral
connectivity -- so characterizing the fractures in this field is of great
importance.
Amplitude Versus Offset (AVO) analysis of azimuthally
restricted seismic data is used to determine the orientation and density
of fractures at selected locations corresponding to well sites in the
Manderson Field.
The result of this analysis is compared to fracture
strike analysis from oriented cores for three wells.
Introduction
The Manderson Field, discovered in 1951, is located
on a sharp, asymmetric, northwest-plunging anticline.
It produces oil and gas from Pennsylvanian, Permian
and Cretaceous horizons, although the Permian Phosphoria is the most productive
zone. It is a complex interval consisting of a thick unit of medium-to-thick
bedded, fractured carbonate.
Purchased by KCS Mountain Resources in 1995, the
field showed several wells with cumulative production in excess of original
oil-in-place estimates.
Low matrix porosity, production history and strong
pressure support suggest that oil is produced from a fracture system with
significant lateral connectivity.
Four oriented cores and one Formation Micro-Imager
(FMI) log were taken by KCS between 1996 and 1998. Analysis of fractures
in the cores and the FMI show varied fracture orientations at different
wells within the field.
A 3-D seismic survey was undertaken in 1996 to improve
structural definition of the reservoir. The survey was reprocessed in
1998 to further improve structural characteristics -- and to detect those
fractures that strongly influence production.
Method
The seismic data from the Manderson 3-D was acquired
over a full 360 degree azimuthal range -- a suitable candidate for the
test of measurements of azimuthal anisotropy from pre-stack seismic data
using seismic AVO techniques. This method examines differences in the
AVO response with respect to azimuth to predict the primary fracture strike
and the relative density of cracks.
The Manderson 3-D covers the locations of three
wells for which oriented cores have been analyzed for fracture azimuth.
This experiment is to determine whether fracture
strike and fracture density can be determined from 3-D seismic data in
the Manderson Field. The requirements for this analysis are full azimuthal
coverage in the 3-D seismic (excluding the edges) and sufficient source-to-receiver
offsets to measure significant differences in azimuthal AVO effects in
the zone of interest in the Phosphoria carbonate at 1,200-1,400 meters
(approximately 6,700 feet).
The five locations chosen for these tests are centered
on the wells 43-33P, 34-28P, 34-18P, 42-24P and 12-18P.
Well 43-33P is the most productive well in the field,
and wells 34-28P, 34-18P and 42-24P have fracture strike analysis. Well
42-24P was abandoned shortly after drilling, without achieving commercial
production.
A modification of the AVO method of Lynn et al (1996)
is used to estimate the fracture strike and density at these locations.
The theory behind this method is the acoustic velocity of shear waves
in a fractured medium is faster parallel to the fractures than perpendicular
to them. This is known as shear wave birefringence (Figure
1).
Figure 1 also
shows how the AVO gradient is related to the shear wave impedance. So
AVO in a fractured zone should show the largest difference between values
parallel and perpendicular to the dominant fracturing direction.
The primary direction of the anticlinal folding axis
is 140 degrees (Figure 2). The most
dominant fracture direction may be expected to be parallel and perpendicular
to the folding axis at 140 degrees or 50 degrees respectively.
The AVO azimuthal analysis uses eight azimuth ranges centered on 5, 27.5,
50, 72.5, 95, 117.5, 140 and 162.5 degrees. Each range is a 22.5 degree
azimuthal cone, so the error associated with any estimated fracture strike
is ±11.25 degrees.
These orientations, starting at 5 degrees, are based
on the estimate of the primary direction of the anticline. They give complete
coverage of all possible fracture strike directions.
Examples
Azimuthal differences in AVO response are shown
in the change in amplitudes with shot-receiver offset as shown in the
boxes in Figure 3. These amplitudes
are larger at long offsets in the gather on the left than the one on the
right.
The fracture azimuth predicted by the seismic data
falls within the range of values estimated from the oriented core (Table
1). This azimuth appears to be the average value for the open fractures.
For example, the strike orientation rose diagram for the open fractures
of the 34-28P well (Figure 4) shows
three significant strike directions, at 0, 60 and 105 degrees. If all
these fracture strikes are averaged then 84 degrees is the expected response.
The average fracture strike is a useful value, because
the largest volume of open fractures will be encountered by drilling horizontally
perpendicular to this average fracture direction.
Well 43-33P is the most prolific well in the Manderson
Field, and it has the largest value of crack density. Wells 34-28P, 34-18P
and 42-24P indicate that there may be correlation between crack density
and open fracture aperture.
Combine this with the high value observed for well
43-33P and crack density appears to have some correlation with open fracture
aperture.
There appear to be two predominant fracture strikes
indicated by the azimuthal AVO analysis: one at 90-110 degrees
follows the East-West faulting direction seen in Figure
2, the other at 130-170 degrees follows the folding axis of the
anticline.
All measurements are performed on the unmigrated
seismic data. The values derived from these data may be map-migrated to
correctly position the detected fracture strikes and crack densities.
A window from the top Phosphoria to 10 milliseconds below is used. The
average value for this zone is shown on the map.
Conclusions
Fracture strike and crack density are estimated
from the seismic AVO response. The seismic AVO results show consistent
fracture strikes and crack densities at the test locations implying that
these values are robust.
The most predominant estimated fracture strikes
at these locations coincide with the major geologic features in the area,
which makes sense from a geologic standpoint. The fracture strikes derived
using AVO agree with the average fracture strikes determined from oriented
cores.
Crack density may correlate with average fracture
aperture, and it generates significantly higher values at the location
of the most prolific wells in the Manderson Field.
It is important to note that this AVO method probably
finds the average fracture direction. If the fractures have different
aperture or spacing in different directions, then the seismic may find
a weighted average.
For example, if fractures at 0 degrees are 0.1mm
thick and fractures at 60 degrees are 0.2mm thick, then the weighted average
is 40 degrees.