A reflection seismic experiment was conducted over the Minneola complex of oil and gas fields in Clark county, southwest Kansas in an attempt to acquire high-resolution data that might distinguish a Morrowan-Atokan clastic section containing thin productive sandstones near the top of a channel fill to estuarine/marine section, from one that does not contain these sandstones. Three lines were acquired. One was acquired along the length of a channel while the other two crossed several channels and intersected the first line. Interpretation of the data was aided by well control and seismic modeling of well logs. Two wells along the lines contained sonic and density logs and were used to generate synthetics that tied well with the seismic data. The newly acquired seismic data also tied fairly well with previous seismic data. The data were acquired using:
1. 2# Pentalite or 2.5# Seisgel minihole-dynamite source @ 20 ft.,
2. a 21-bit 120 channel Bison recording system
3. 30 Hz geophones (12/group, buried), 1 10 ft group interval, 220 ft shot point interval, 1 ms sample interval, 3 sec recording time, and no notch filters.
Previous well log seismic modeling suggested that frequencies between 120 and 180 Hz were needed on the high end and between 20 and 30 Hz on the low end were needed for the experiment to work. What was obtained from the final stack were high frequencies at the roll off point no higher than about 50-60 Hz at the depth of interest. Therefore, the final results of this experiment combined with well log modeling suggest that the frequencies obtained were too low for the presence of the thin sands to create a seismic signature that was significantly different from a seismic signature created from just shales. This conclusion was supported by data, which did not show a significantly different seismic signature between a sand and no sand channel of the same thickness. However, the results are somewhat inconclusive because the combination of extremely wet conditions and electrical noise generated from power lines and other sources resulted in a very poor signal to noise ratio, particularly for frequencies above 60 Hz. If conditions were dryer, the electrical noise may not have been as bad, and the results may have been different. Another possibility for the lack of high frequency data is the source. Although small charges typically produce high frequency data, the source may have been too weak to get high frequencies down to and back from the depth of investigation.
Alternatively, the low Q near surface ground conditions (loess) may have caused absorption of much of the higher frequency signal at both source and receiver. One result of the experiment though was the ability to roughly determine channel thickness based on seismic amplitude. Although this is not a totally new result, the additional data helped to better define the channels mapped from previous seismic data, and supported the use of seismic amplitudes for channel thickness determination. The data also suggest that the channels may be more complex than the 2-D data illustrates. If a detailed knowledge of the channel geometry would be useful for further development of the field, than a 3-D survey of the area would be helpful. However, unless another high resolution seismic acquisition method is found which could record higher frequencies back from the depths of investigation, identifying the thin Morrow sandstones from a pure shale section would be difficult with reflection seismology alone.