Were there any natural justice in the
world, this article would be the perfect Victorian morality tale: If you
do your homework, spend your money wisely and are not afraid to take prudent
risks, you should be rewarded, right?
Unfortunately, sometimes Mother Nature wants
to play tricks on us -- and our belief in the technology we use becomes
shaken.
In this case we found that advanced
geophysical modeling allowed us to recognize a phenomenon we had not seen
before. Unfortunately -- at the time -- we did not have the additional
technology we needed to distinguish this phenomenon from commercial hydrocarbons!
But before we get to that, let's discuss
something that has been widely discussed in recent years in the geological
literature: the issue of paleo-oil-water-contacts (paleo-OWCs).
In outcrop, bitumen-stained sandstones
can be easily recognized. Geologists from Norway to Abu Dhabi routinely
identify residual oil or tar mats below present-day OWCs in cases where
there has either been a seal failure, change in hydrodynamic regime or
subsequent regional tilting.
What is considerably less common is
the identification of paleo-OWCs by geophysical means.
However, as this article shows, this
phenomenon may be more widespread than previously thought and could, in
fact, be responsible for some previously unexplained exploration failures.
Figure 1
shows four nearby structures (Cockroach, Louse, Tick and Flea), with seismic
lines crossing each. Geophysical modeling shows that the porous Pliocene
sandstone reservoirs in this basin tend to be lower acoustic impedance
than the surrounding shales, so they tend to show up as amplitude anomalies
on compressional ("P-wave") seismic lines, especially when filled with
hydrocarbons.
On Figure
2 we can see "bright spots" in the Cockroach structure at two-way
times of 1.9 seconds (gas) and 2.3 seconds (oil). Actually, all four structures
have amplitude anomalies between one and three seconds.
An amplitude extraction over a 45 milliseconds
window around the main reservoir at Cockroach shows structurally conformable
high amplitudes (black), with an outer secondary "bath tub ring" of high
amplitudes (Figure 3). Several circular "no
data" zones are due to mud volcanoes, which pierce the structure.
Barely visible on this display is the
shallow gas cloud that obscures a significant percentage of the reservoir
(40 percent overall) by absorbing and scattering the seismic energy. The
inner amplitude anomaly corresponds to the present-day OWC, and has been
penetrated by wells. The outer ring is asymmetric.
Note the position of the two wells
on the southern flank. The logs for these two wells (Figure
4) show that the upper well penetrated the present day OWC. The lower
well penetrated a flushed zone. The outer "bathtub ring" corresponds to
the paleo-OWC.
Both regional tilting and a seal failure
have occurred, as shown schematically in Figure
5. A thick, hydrocarbon-bearing shale underlies this structural trend.
Pliocene compression, plus sedimentary loading of the shale on one flank
of the structure, initiated argillokinesis and uplift, which created a
linear sill.
This, in turn, caused further asymmetric
loading and steepening of the northern flank.
Finally, the diapir rose to a level
at which explosive ex-solution occurred within the shale, and dissolved
gases penetrated through the overlying formations.
Hydrocarbons then leaked to the surface,
either directly through the mud volcano diatremes or via faults that became
active at this time, until the migration conduit was largely sealed by
fine clastics (satellite radar data in this area show numerous seeps).
An amplitude anomaly remains at the position of the paleo-OWC.
Figure 6
shows a seismic line that crosses both Cockroach and Louse, which is the
southernmost of the series of structures parallel to Cockroach (Figure
1). Bright spots can be seen between 2.5 and 3 seconds at Louse. These
amplitude anomalies are favorably located at the top of the structure.
However, exploration wells found that
the main reservoir was wet -- and there was non-commercial gas, with some
questionable liquids, at a deeper level.
In detail, Louse is seen to be highly
faulted. Structural modeling has shown that little tectonic movement occurred
throughout most of the Pliocene. The faulting occurred in Latest Pliocene
time, which corresponds to the onset of mud diapirism at Cockroach. The
same structural loading that steepened the reverse-faulted northern flanks
at Cockroach also caused extensive normal faulting at Louse, allowing
hydrocarbons to leak out at this time.
Some people have explained the disappointing
results by arguing that hydrocarbons never filled the main reservoir at
Louse, and suggest that the faults we see on the seismic were never wide-ranging
enough to act as a migration pathway. This would make Louse somewhat unique
in a basin with so many surface seeps.
Further along the trend, the Tick structure
(Figures 1 and 7)
also shows structurally conformable amplitude anomalies at deeper levels
but not at the shallower main reservoir level.
Figure 8
is one such amplitude map.
From the drilling results, we know
that these are hydrocarbon-related, whereas the (non-structurally conformable)
high amplitudes seen at 1.2 seconds and 2.1 seconds correspond to porous
sands.
Note the lack of younger faults at
Tick: the structure was relatively unaffected by the Latest Pliocene tectonics
that affected both Cockroach and Louse, so entrapped hydrocarbons could
not leak to the surface.
The final structure to examine is Flea,
at the trend's northern end. Figure 9 shows
amplitude anomalies at 1.5 and 2.5 seconds. Both are structurally conformable
and highly faulted, with the shallower anomaly also showing a stratigraphic
component on 16-bit seismic data (not illustrated) due to channeling.
Figure 10
shows the amplitude map at the lower level. On drilling, both levels showed
only residual gas: the hydrocarbons had leaked out.
In retrospect, we now realize that
a typical structure in this basin could have a multiplicity of false positive
and false negative hydrocarbon indicators on conventional compressional
("P-wave") data.
- First, scattering associated with a shallow gas
cloud could obscure the amplitude anomaly associated with either the
top of the reservoir or OWC (or GWC).
- Second, velocity effects due to the shallow gas
could obscure the fact that an amplitude anomaly is structurally conformable,
even on 3-D data.
- Third, 8-bit data do not show stratigraphic components
of trapping.
- Fourth, porosity -- rather than fluid fill --
is the primary driver of amplitude response in this area, so that porous
beds could be mistaken for pay.
- Fifth, leakage of hydrocarbons due to mud volcanism
or recent faulting could leave behind a paleo-OWC that may not be distinguishable
on the seismic from the real OWC, if one still exists.
Although the study area may be unique
in having all these effects simultaneously, there is reason to suppose
that many basins around the world could exhibit one or more of these problems,
which could mean the difference between an exploration success or failure.
At the time of this exploration campaign,
the technology needed to distinguish between residual and commercial hydrocarbons
and to image through gas clouds was not widely available. Now, with multicomponent
data -- which also utilize shear waves ("S-wave") -- we have the ability
to derive density from seismic data and to make accurate structural maps
even in the presence of shallow gas.
Intelligent explorationists will undoubtedly
take advantage of these new techniques to distinguish residual hydrocarbons
from commercial accumulations.
And no doubt Mother Nature, in her
turn, will find some new way to baffle us!