By WILLIAM S. LYONS
The Geophysical Corner is a regular column in the EXPLORER, edited by R. Randy Ray. This month's column is titled "Seismic Guides Interpretation in the Ferron Coalbed Methane Play."
Seismic Maps Ferron Coalbed Sweetspots
Editor's note: William S. Lyons is a geologist and geophysical interpreter with ChevronTexaco in Denver.
See enlarged image for graphic / photo credit.
Recent work in Utah's Ferron coalbed methane (CBM) trend has demonstrated that seismic can be useful in evaluating CBM plays.
Objective horizons in CBM trends are typically shallow and acoustic impedance contrasts between the coals and interbedded sandstones are high, making many coalbed methane plays attractive candidates for seismic study. Although keeping costs low in these plays is important, 2-D seismic is now relatively inexpensive to acquire and can help greatly to:
Ferron Coal Play
One of the most significant domestic CBM plays has been the Ferron trend along the western flank of the San Rafael Swell, near the town of Price, Utah (figure 1). In this area, intertonguing coals, sands and shales were deposited in a fluvial-deltaic environment on the western side of the early Cretaceaous Interior seaway (figure 2).
Development began on the play's north end, at Drunkards Wash Field in 1992. Since then, the producing part of the trend has been extended both south and north and now spans an area about 65 miles long and six to 10 miles wide.
Despite the lateral extent of known Ferron production, large areas between producing fields remain undeveloped. Limited well control across these open areas has made characterizing the extent, thickness, quality and production characteristics of the coals uncertain, and development has proceeded cautiously.
Coalbed gas contents also decrease southward in the play and project economics become progressively leaner. In these areas there is an additional need to focus development by locating production sweet-spots, or "fairways," quickly.
Because of these factors, Texaco used seismic to help guide its most recent development activity in the play.
A Requisite First Step
In two stages during late 1999, we uniformly reprocessed and interpreted about 140 miles of 1980's vintage 2-D seismic over the Ferron trend. The first stage was conducted over the Drunkards Wash Field, where dense well control was used to determine if seismic could identify the causes of observed production variations. The second stage was intended to expand what was learned at Drunkards Wash into the trend's southern parts, where development is much less complete. Seismic lines from each stage are presented in this article.
The emphasis in reprocessing was on attention to detail, particularly for imaging shallow horizons. Changes observed on seismic had to represent true changes in the Ferron interval, and even subtle features could not be ignored since they might prove to be important in understanding the stratigraphy.
To achieve reliable results, only lines with similar acquisition parameters were chosen. Reprocessing efforts included applying refraction statics, assuring zero-phase data, detailed velocity analyses, correction of CDP geometry errors, careful selection of shallow mutes and using improved deconvolution routines.
All steps proved to be very important in increasing the data's signal to noise ratio, and in optimizing the quality of shallow reflectors.
The Key Link
The Ferron interval in the coal-bearing areas is characterized by an upper sandstone, a middle interval with coal and interbedded siltstones, and a prominent sandstone at the base known as the Lower Ferron Sandstone. Synthetic seismograms from this sequence typically show a distinctive pattern (figure 3). A weak reflection peak is generated from the upper Ferron sandstone due to a low impedance contrast with the overlying siltstones in the Bluegate Shale member. This is followed by a strong trough (or trough doublet) produced from relatively thick sections of high-contrast, negative reflection coefficient coals immediately below the upper sandstone.
At the interval base, the tight sandstone of the lower Ferron once again contrasts sharply with coals, and another high amplitude peak is produced.
Sonic logs were available from a few wells close to seismic lines in the Drunkards Wash area, and synthetic seismograms from these matched the newly-reprocessed seismic quite well. The tool of choice in coal plays is the density log, however, and nearly all the wells in the field have one of these.
Where wells had both sonic and density logs, psuedosonic curves created from the density log produced synthetic seismograms that were nearly identical to those produced from the sonic log.
Drunkards Wash is fairly densely developed at 160-acre well spacing. Since the density logs from these wells could be used reliably to generate synthetics, we were able to tie many nearby wells into the seismic at closely sampled points along each line. Stratigraphic and lithologic changes observed in the wells could then be related directly to changes seen on seismic with high confidence and little interpolation.
Stratigraphic Observations, Interpretations
Figure 4 shows a series of coincident lines that both demonstrate our results and show how lithologic and stratigraphic changes seen on well logs in the Ferron interval translate into changes observed on seismic:
The displays have been hung on a flattened Lower Ferron sandstone and scaled similarly so they can be compared with ease and accuracy.
On the west (or left) side of figure 4, well logs show a thicker Ferron interval. Individual coal seams occur in groups, forming relatively thick upper and lower packages. These are resolved as dual high-amplitude troughs (with maximum negative amplitudes shaded yellow) and are separated by a peak generated by the intervening siltstones and shales. The last high-amplitude peak at the interval base is caused by tight Lower Ferron sandstone.
Moving east along the line, the coal sections merge and thin, and the intervening siltstone is lost. Similarly, on the seismic the intervening peak disappears and coal troughs merge to form a doublet. Amplitudes are observed to diminish as the section thins and tuning effects cause destructive interference in the seismic signal.
On the line's eastern end, there is a facies transition to stacked Ferron shoreline sandstones and the coals are absent by non-deposition. The seismic interval thins and amplitudes dim further as the coal-sandstone impedance contrast is lost.
Structural influences are also recognizable. Disruption of the reflectors due to faulting is evident on the east-central part of the line (despite horizon flattening).
Once seismic stratigraphic and lithologic relationships such as these were established in areas of better well control, the seismic could be used to interpret significant changes in the Ferron interval elsewhere. Similar seismic "facies" observed on other lines were used to map significant features of the Ferron interval in undeveloped parts of the play. For example:
Attempts were also made to correlate observed amplitude changes with variations in production. Although no simple association was recognized, interval thickness variation occurring close to the tuning thickness may have caused amplitude fluctuations that obscured meaningful relationships.
Structural Observations, Interpretations
Where structural features such as faults and folds enhance coal productivity, seismic can be particularly helpful in focusing play development -- something clearly demonstrated in our geophysical work at Buzzard Bench field (figure 1).
Initial high water production from CBM wells can be an indicator of permeability and is often a good sign for future productivity in newer coalbed methane plays. However, as Texaco began developing the Buzzard Bench area, many wells there turned out to be low-volume water producers.
Several wells on the field's east side did have dramatically higher water (and gas) production rates (see figure 5). This was encouraging, but on logs, there was no apparent reason for this enhanced production. The wells were suspiciously aligned north to south, however, and production changes in one well seemed to affect the others.
Although this suggested the presence of a fracture or fault system forming a connected pathway of enhanced permeability, well control alone could not confirm it.
The seismic here aided greatly our understanding of the geology behind the production anomalies. Seismic line 4 traverses both the area of poor production to the west (not shown on the line) and the trend of enhanced production to the east.
Where the line traverses the area of poor production, nothing of significance in the Ferron interval was observed. As it crosses the area of better production, however, the line intersects one of the high-volume water and gas-producing wells, the UP & L 14-55, and clearly shows a large fault at the Ferron level. It also shows that the well is in an area of pronounced downward folding (reverse drag-folding) on the fault's downthrown side. Enhanced production in the 14-55 is thus the result of improved permeability from fracturing along the main fault and from small-scale antithetic faulting (below seismic resolution) in this zone of reverse drag folding.
This fault is also found on other lines in the area and can be traced in a north-south direction where other more-prolific producing wells also lie along its downthrown side. With the aid of the seismic, we were able to map a narrow "fairway" of enhanced production parallel to the fault's downthrown side and focus further development along this trend.
Recognition of the fault on seismic also suggested structural causes for other curious complications in this part of the play -- one of which was continued high water production from the high permeability wells beyond the expected volume for normal coal dewatering.
The faulted and folded zone on seismic line 4 at Buzzard Bench resembles a half-graben, a structure typically associated with extensional tectonics. This zone and its master fault can be traced from its termination several miles south of line 4 through the subsurface due north almost to a large fault of the same strike mapped on the surface geologic map.
As it continues northward, the surface fault gets larger and becomes part of the Scofield Graben system in the Wasatch Plateau (figure 1). The connection to that graben system likely explains the half-graben morphology of the Buzzard Bench fault zone on seismic.
With its exposure on the wetter high plateau, the fault system forms a conduit for fresh water recharge (seemingly confirmed by maps of formation water chlorides in this area). This makes dewatering of the coals along the fault down at Buzzard Bench difficult and it may take longer than expected.
Additional wells would help expedite dewatering of coals in the enhanced permeability fairway. Therefore, this seismically-supported geologic interpretation provided another reason to target the area for increased drilling.
Although further structural discussion is beyond the scope of this article, it is important to note that the seismic also helped us understand the formation of other significant structures observed in the Ferron trend, such as the Huntington Anticline, and the timing of their development.
There are additional benefits conceivable in the Ferron area using seismic. Most notably, a definitive link between seismic attributes and production might be possible. Acquisition of new, higher frequency data and perhaps the use of seismic inversion techniques could achieve this objective.
Shear wave attenuation and Vp/Vs ratios from multi-component seismic might also be employed to predict areas of increased cleating, related higher permeabilities and the associated better production.
In the Ferron CBM play, seismic was used in a cost-effective way to:
Seismic is not used often in CBM plays, perhaps because of cost concerns or because of a paradigm that little additional information will be gained from the effort. We learned from experience in the Ferron CBM play, however, that seismic adds valuable geologic information which helps to guide interpretation and focus development.
(Editor's note: )