fractured reservoir CASE HISTORY
Most of the porosity for this reservoir is within the heavy mineral zone, at the top of the log interval. Mineralogy here is quite complex, containing barite, fluorite, quartz, calcite, anhydrite, and clay in a relatively un-cemented form. A drillstem test recovered these minerals along with oil. No drilling samples of this material came to surface.
In addition to production from the heavy mineral zone, there appears to be some potential reservoir in the very low porosity carbonate below. This limestone is near the front of the Bent Horn overthrust fault, which strikes roughly east-west and dips (now) to the south.
Displays of all helpful logs (computed as well as original data) for well A-02 are shown above. The initial indication of fractures was from the caliper log (4-arm) reading oval hole in certain zones, and from difficulty in keeping the sonic log from skipping. The dipmeter correlation curves also show discrete events on one or two curves that can be interpreted as high angle (relative to the hole) fractures. Hole breakout on the caliper, noisy density correction (not shown) and noisy density log all point to fractures.
The best dipmeter example is at 9198 (fracture dips north at 40 degrees) and others at 9240, 9329 and 9355-9362. Some fracture planes are so steep that dips were not computed by the dipmeter program due to too low a search angle. The results using a higher search are shown. In some cases, new (higher) dips are found and replace lower, less coherent dips. In other zones no new data were found and some were lost.
Other zones of stress release fractures are indicated by the oval hole indications on the borehole geometry log. The long axis of the hole is perpendicular to the strike of the thrust sheet front, and therefore parallel to the tension cracks caused by spreading of the front. This orthogonal system of fractures would present nearly vertical fractures at the borehole, and only those which the dipmeter pads touched and which also depart locally from the vertical will show up on the computed dipmeter.
The sonic amplitude variable density log was run in open hole below casing shoe at 10100, and in cased hole above this. Due to poor cement bond, the VDL is not very helpful in locating fractured zones through casing, but the portion in open hole (not shown) indicates that the majority of the rock is fractured.
The same logs for well O-51 are presented above. The amplitude VDL log is in open hole in this case. It shows fracture zones at 7470-7506, 7636-7664 and 7920-8060. These zones are confirmed by the dipmeter and the borehole geometry log. Other smaller fractured zones are present on the log. The distinctive "chevron" patterns on the VDL indicate semi horizontal fractures, as opposed to the very regular "straight" sections which indicate competent undisturbed rock.
The limestone in this well dips to the east and the long axis of the hole enlargement is roughly east-west, which is perpendicular to the dip strike and parallel to the direction of hole deviation. This is different from well A-02 where the hole enlargement is perpendicular to strike, but also perpendicular to the hole deviation direction. The "all quality" dipmeter with the high search angle shows why some zones lose their coherent dips when the higher search angle is used. The scatter of possible dip answers is very large.
The low porosity nature of the zone creates other interpretation problems. The neutron porosity from the CNL (original log) is very low and often a little behind zero. The sonic log (knowing we are in fairly clean limestone) shows some porosity in the range of 0 - 4% on both wells. This encouraged us to look at computation methods that would utilize the more optimistic porosity instead of the CNL (original) porosity.
Two different approaches were taken. First, the Coriband program was forced to use sonic log porosity (corrected for shale content) by shifting density and neutron data above the highest sonic porosity. Clay parameters were shifted the same amount, so that shale content and matrix density are still computed correctly. Using the sonic limit option then forces the program to use sonic porosity instead of crossplot over the entire interval. These were dubbed "Sonic" Coribands; results are shown in Figures 28.40 and 28.41.
The second approach was to improve neutron porosity. This is done by using the actual count rates recorded on tape and calculating a porosity from each of the near and far count rates and from their ratio. It is apparent from this data that the negative porosity on the original log is generated by incorrect hole size corrections, and completely distorts the true porosity picture. The log could be shifted by about 1% to attain a more reasonable result, but the logarithmic conversion from ratio to porosity is not handled properly this way.
The computed log which agreed most closely to the sonic porosity was chosen, which in this case is the CNL porosity from count rate ratios. Differences between sonic and neutron porosity are due to secondary porosity, different response to vertical fractures, and shale correction inaccuracies. No vuggy porosity was seen in cores. Coriband was then run with the results shown. Saturation calculations can be improved by lowering the formation factor exponent (M) to some value between 1.0 and 2.0, determined by the ratio of fracture porosity to matrix porosity. This would have to be done on a zone by zone basis, and has not been attempted in these wells.
The A-02 well produced oil at rates up to 3,300 barrels per day through perforations from each of several fractured low porosity zones similar to the one at 9100-9130. It is possible that there was communication behind casing to the main (heavy mineral) reservoir at 9034-9054. There is also some reservoir volume in the fracture porosity and in the low matrix porosity.
O-51 well tested a small amount of gas with water from comparable
rocks. It is structurally higher but obviously in a different
reservoir. The porous heavy mineral zone at the top is also missing.
Page Views ---- Since 01 Jan 2015
Copyright 1978 - 2017 E. R. Crain, P.Eng. All Rights Reserved