One method used to geosteer horizontal wells is to compare a log trace in the horizontal well, usually a gamma ray log, to the same log in a nearby vertical well (the pilot or template). In this process, the log from the horizontal is squeezed and/or inverted until the character in the well matches the character in the template. Squeezing is necessary because relatively small stratigraphic intervals in the template well will be traversed over much larger distances in a horizontal well. Inversion is necessary because the drill can cover a given stratigraphic interval from bottom to top as well as top to bottom. This technique has been termed " true stratigraphic position" or TSP.
As the name implies, TSP is closely related to true stratigraphic thickness (TST). Indeed, TST can be used in the same manner as TSP. In a geometric sense, TSP uses log correlations to find apparent dips. It is demonstrated here that TST is superior to TSP because it can " look ahead" of the bit, and because it allows the geosteerer to concentrate on features in the drilling well that might not be noticed when the log in the well is squeezed down to the scale of the template.
Display type is important when modeling using TST. Vertical sections can have problems displaying apparent dip and horizons whereas displacement (curtain) sections will properly display both dips and horizons. The source of these display problems with vertical sections is largely from dips whose azimuths are out of the plane of section in places where the borehole has deviated away from the plane of section. Both vector and curtain sections have display problems in areas of high dip or where the borehole is away from horizontal. In those cases, vector sections, which are constructed from dip-normal vectors combined with stratigraphic distances, are much better and displaying the structure.
A common complaint about gamma ray logs from horizontal wells is that they have a " jittery" appearance. This is sometimes an unfair criticism, however, because they sample the stratigraphy at a much higher effective rate than the template wells, giving them a jittery appearance when squeezed down to the template scale. One approach is to stretch, invert, and fault the template log to fit the log in the horizontal log. For example, inversion of the log character when the drill travels up section will cause a " mirroring" or " reflection" effect on logs. Markers that might not even be present in the template log can clearly show a mirror effect on the horizontal log, while the same mirroring is missed when the log is squeezed down to template scale. Taking the idea one step further, it is possible to take a log in a horizontal well, convert it to TST and use it as a template, either to correlate with deeper parts of the same well or to use as a template for other wells using markers that may not even exist in vertical wells.
Figure 1 is a forward model illustrating typical relationships between a pilot log and a log in a horizontal well. The trace in the horizontal log has been stretched and/or inverted relative to the pilot. The inverted part of the log is caused by the bit traveling up section, or from older to younger beds. It creates a typical pattern in horizontal wells known as " reflection" or " mirroring".