SEISMIC DETECTION, INTERPRETATION OF ROSE RUN IN SOUTH-CENTRAL OHIO

James R. Morris
Quaker State Corp.
Titusville, Pa.

With sparse well control along the Rose Run subcrop fairway south-central Ohio, exploration for sand capped Rose Run remnants is seismic dominated (Fig. 1(41442 bytes). The purpose of this article is to describe a method for acquiring high resolution dynamite seismic data and to discuss interpretation techniques for detecting remnants, predicting sand presence, and estimating sand thickness.

STRATIGRAPHY

Rose Run remnants in this part of Ohio consist of a basal carbonate or dolomite of approximately 50 ft, some-times capped with a blocky sand section ranging in thickness from 0-50 ft. When a full section of sand is present, a Beekmantown dolomite section usually caps the sand. Peak porosities in the sand range between 20-30%, making them attractive exploration targets when charged with gas or oil.

RESOLUTION AND DETECTION

Seismic resolution, as defined in the geophysical literature, would require a separate reflection for both the top and bottom of the sand. Using the quarter wavelength criterion for seismic resolution would require our seismic data to have a dominant frequency in our zone of interest equal to the sand velocity divided by four times the sand thickness.

With the sand velocity at about 13,500 ft/sec, a dominant frequency of 67.5 hz would be needed to resolve a 50 ft section of sand. Obviously, less sand would require higher dominant frequencies.

Acquiring seismic data with dominant frequencies this high is a real challenge at the unconformity zone and may not be possible with conventional recording equipment and parameters.

Fortunately for us, resolution in the strict sense does not consider amplitude effects. Detection of the sand via amplitude anomalies and changing reflection character may be enough to make meaningful predictions of sand presence and thickness without measuring sand thickness directly. In either case, we need to acquire in the field and retain in processing as much high frequency, broad bandwidth data as possible.

ACQUISITION TECHNIQUES

Although there are many ways to increase high frequency content in the recording phase, there are two which this author feels are most important in this area of Ohio.

The first involves reducing charge size in the source array. Since charge size is inversely proportional to both dominant frequency and bandwidth, any charge size over that needed to provide good signal return damages our frequency spectrum by narrowing the bandwidth and reducing the dominant frequency. A source array consisting of four holes loaded with one fourth pound each is usually sufficient in this area.

Second, using a lowcut field filter of 27 hz rather than the conventional 18 hz, attenuates the high amplitude low frequency noise while allowing more low amplitude high frequency signal to be recorded. This spectral shaping technique results in a much desired flatter amplitude spectrum boosting the high frequencies.

PROCESSING TECHNIQUES

For maximum resolution and detection capabilities, wavelet processing to zero phase is an important post stack processing step. This process involves estimating the embedded seismic wave let and converting it to its zero phase equivalent. An ideal seismic trace can be thought of as a series of reflection spikes, each of which represents a change in rock properties. A real seismic trace can be thought of as the ideal seismic trace with each one of its spikes replaced with the embedded seismic wavelet. Obviously, the more the embedded seismic wavelet looks like a spike (zero phase or symmetrical), the closer we get to the ideal seismic section.

Post stack spectral whitening is another important step late in the processing flow designed to enhance or boost the high frequencies which were attenuated by the stacking process itself. With most acquisition being 60 fold in this area, this is especially important.

SEISMIC INDICATORS

Data in the study area, a joint venture between Quaker State and KEP Exploration, consisted of 10 seismic lines and six wells acquired and drilled in 1993-94 (Fig. 2(43143 bytes). The acquisition and processing techniques mentioned above were used on all 10 seismic lines resulting in a 2!X to 3 octave bandwidth with strong frequencies up to 105 hz (Figs. 3 (95679 bytes),and 4 (62014 bytes)). Acquisition was done by Hosking Geophysical and Precision Geophysical, while the processing was done by Alvin Hosking and his staff at Associated Geophysical.

The 1 Ross and 1 Mcilvaine were drilled on seismic anomalies in November 1993. The 1 Ross had 24 ft of gross sand and was gas charged, while the 1 Mcilvaine had no sand and was a dry hole. Sonic log interpolation modeling along with a study of their observed seismic signatures resulted in the following list of seismic indicators needed to guarantee sand presence on the remnants:

1. Pull up of the unconformity reflector of more than 4 msec.

2. Positive amplitude anomaly on the unconformity reflector.

3. Rollover on the Lower Black River reflector of more than 4 msec.

4. Rollover of shallow reflectors above the remnant.

PREDICTIONS AND RESULTS

After acquiring the additional seven seismic lines in 1994, only those seismic anomalies with the necessary sand indicators listed above were selected and drilled.

All four selected remnants were predicted to have more sand than the 1 Ross.

Both the 1 Hockman with 43 ft of gross sand and the 3 Ross with 54 ft of gross sand were gas charged, while both the 2 Ross with 51 ft of gross sand and the 2 Mcilvaine with 38 ft of gross sand were wet (Fig. 5(94903 bytes)).

In an attempt to distinguish the gas charged sands from the wet sands, maximum unconformity amplitude values on each of the six remnants were measured. All were of comparable amplitude with the exception of 1 Mcilvaine, which was approximately half as much.

Also, a Precambrian time structure map using the 10 seismic lines was generated but showed no correlation to the wet or gas charged remnants. The most updip remnant was wet, while the most downdip remnant was gas charged. Wet remnants were on structural noses or flanks as well as gas charged remnants.

CONCLUSIONS

With proper care taken to .record high frequencies in seismic acquisition and to retain high frequencies in seismic processing, Rose Run erosional remnants are easily detected in this area of the subcrop fairway. In addition, Rose Run sand presence can be predicted and sand thickness can be estimated. At this time, there does not appear to be a way to distinguish gas charged sands from wet sands.

Copyright 1995 Oil & Gas Journal. All Rights Reserved.