Ercill Hunt Ercill
Hunt & Associates Inc.
Houston
Well logs with deep investigation capabilities are necessary to determine accurately the height of hydraulically produced fractures.
Logs with shallow investigation capability will indicate the height of the fracture near the well bore, but as shown in a test in an East Texas well, fracture height in some formations can be substantially greater away from the well bore.
In the East Texas test, sponsored by the Gas Research Institute (GRI), six wire line surveys were run, including the usual gamma ray surveys. The tools included were:
- Teledyne Geotech's Continuous Microseismic Radiation log (CMR)
- Mobil Oil's Tube-Wave Reflection Log (TWRL)
- Halliburton Services' Tracerscan after the minifracture (Au) and Tracerscan for the main fracture (SCAN)
- Schlumberger Services' Cement Bond Log (CBL)
- Schlumberger Services' Variable Density Log (VDL)
- Schlumberger Services' Cement Evaluation Log (CEL).
The CMR log, a relatively new survey, had the deepest investigation capability. The depth of investigation of the other tools is less than 1 ft.
The fracture heights indicated by the CMR log closely agree with those heights computed using fracture models.
Experimental verification of CMR as an accurate method for determining fracture height was an important research objective for the Gas Research Institute's Staged Field Experiment No. 3 (SFE No. 3).
The six logs were run before and after the stages of completion and fracturing on SFE No. 3 in an attempt to arrive at an independent determination of fracture height.
The fracture heights determined by the above logs are plotted on Fig. 1. The independent estimates of gross fracture height varied considerably.
Four logs, TWRL, VDL, SCAN, and CEL appear to be influenced by the fracture. Results were inconclusive from the Au and the CBL log. Analysis of each of these indicates a different minimum fracture height in this well.
Taking the largest indicated height of these four minimums, yields 192 ft for fracture height. The CMR value for propped fracture height is 250 ft. After an in-depth review of the problems inherent with each of the logs and their analysis, it is apparent that the CMR is indicating the most accurate fracture height.
Plans included two additional logs, Mobil's Shear Wave Anisotropy Log (SWAL) and PetroData's Quadrapole Sonic Log (VAL Systems), both of which were unavailable at logging time. These two logs would have detected shear wave birefringence near the well bore, and offer independent means of finding fracture height.
CMR
The CMR was developed by Teledyne Geotech as part of a research program sponsored by GRI. The CMR determines the height of a created fracture by monitoring the change in continuous microseismic radiation that results from either a change in the elastic properties in the rock around the fracture or the stresses in the rock surrounding the fracture.
The pressure and temperature differences between the fractured region and the host rock are believed to cause the changes in microseismic radiation.
Immediately following the fracture operation, the continuous microseismic radiation is recorded using three component geophones. The geophones are positioned in the well bore, and the microseismic radiation is recorded at several stations below, in, and above the fractured interval.
These signals from each recorded depth are processed to determine the relative amount of the total signal that is vertical or horizontal. The signal will have a stronger horizontal component when the tool is within the fracture zone, and will have a stronger vertical component when the tool is above or below the fractured zone.
In Fig. 2, the arrows indicate the depths where the vertical signal becomes dominant, that is the base and top of the fracture.
In the case of SFE No. 3, the CMR indicates the base of the fracture to be at a depth of 9,375 ft and the top at 9,125 ft. The fracture height is then 250 ft.
The CMR fracture height is compared to lithology and a log-derived stress profile in Fig. 3. It appears that the fracture grew into the shales above and below the perforated intervals and that vertical growth has been contained by stress barriers at 9,125 ft and 9,375 ft.
TWR
When Mobil agreed to run its Tube-Wave Reflection Log (TWRL) on the SFE No. 3 fracture height experiment, it brought new technology to the experiment. The log had successfully indicated fracture height for Mobil in other areas. In addition, the TWRL may eventually be able to infer fracture transmissivity through analysis of the reflected energy. Unfortunately, the TWRL seems only to indicate the absolute minimum fracture height in this well.
Tube waves which are generated by a short, lowfrequency burst of a monopole source, travel down the well bores and are reflected by any change of impedance encountered. A fracture in communication with the well bore should cause a large change in impedance.
Other changes in impedance due to casing collars and fluid density changes are minimized by using a longer wavelength source (17 ft at 300 hz). During the time between the bursts of the source, the acoustic energy is recorded for 100 ms, long enough to record any reflections of interest.
Strong reflections are seen at and between 9,249 and 9,335 ft on the TWRL. These are interpreted to indicate the depths of the lowermost and the uppermost changes in impedance caused by a fracture being in contact with the well bore over that interval. This is an indicated fracture height of 95 ft which is considerably less than indicated by any other method.
In favor of a shorter height is the fact that the Taylor section of the Cotton Valley formation responds very well to hydraulic stimulation. A contained fracture height with more fracture length over the producing zone is believable.
However, an indicated fracture height of only 95 ft is contrary to all other measurements (Figs. 1 and 3) and also contrary to the hydraulic fracture model results.
We believe that there must be another interpretation of the TWRL. One possibility is that the change in effective impedance is very low except where the fracture is in direct contact with the well bore. The reflections noted at 9,240 and 9,335 ft are the top and base of where the fracture plane intersects the well bore.
If the fracture plane leaves the well bore above and below these depths, the reflections from the fracture may be present but with greatly reduced amplitude.
From borehole televiewer analysis of the open hole stress tests, most of the created fractures are in fact subvertical by 3 or more. If this is the case on SFE No. 3, then the indicated top and base of the fracture by the TWRL could be the uppermost and the lowermost points where the fracture is in direct contact with the well bore.
The upper reflection also coincides with the perforation which is interpreted to be most open; that is, it probably took most of the fracture fluid, by the Tracerscan analysis.
A close look at the reflection log recorded with the source operating at 300 hz does reveal some small reflections above 9,240 ft. Perhaps the impedance change is small when the fracture is not in direct contact with the well bore, but much larger when it is in direct contact. If so, additional processing may reveal a different estimate of the top and base of the fracture.
Since the strong reflections come from zones (9,240 ft) where the other tracerscan survey indicates the fracture is most effective, the TWRL does seem to give a good indication of fracture effectiveness. Perhaps analysis of these reflections will someday provide fracture transmissivity.
RTL
Monitoring changes in radiation levels caused by tagged fluids and proppants which were placed during fracturing, and then relating the change in radiation level to fracture height, has had widespread acceptance in the industry.
A decrease in popularity of this method was caused by lack of control on the distribution of the tagged material, by not knowing whether the tagged material was inside or outside the casing, and by a general lack of quality control on the monitor gamma ray logs.
The introduction of controlled distribution systems such as were used by Pro-Technics on SFE No. 3 solved one problem. The gamma ray spectroscopy logs (such as Halliburton's Tracerscan) are calibrated and can be used to distinguish the relative distance to the radioactive material, that is whether radiation is coming inside or outside the pipe.
When controlled distribution and spectral analysis are combined, more useful information is available.
A major problem still exists, which is the depth of investigation of the tools. That is, how far will a gamma ray travel before it is absorbed by the formation. This is determined by the half-value thickness (HVL) of the formation, which is the thickness of the material necessary to reduce an incident gamma ray by half of its intensity.
The HVL of a typical formation is 3 in. After three or four HVLS, the gamma ray's energy is below the detection level of the tools. It is therefore reasonable to say that the depth of investigation of a gamma ray tool is about 1 ft.
We know that the fracture planes and the well bore axis are not aligned in many cases. When the angle between the well bore axis and the fracture plane is about 4, the inverse tangent is 14.3 ft, which means the fracture is 1 ft away from the well bore for each 14.3 ft vertical from the point where the fracture plane exits the well bore (Fig. 4).
The gamma rays emitted from the radioactive material in the fracture will not have enough energy left to be detected after transversing 1 ft of formation, cement, and pipe. This may explain why the gamma ray logs so often seem to indicate that the fracture only grew about 15 ft or so out of zone.
During a minifracture treatment, the fluid was tagged with gold (Au on Figs. 1 and 3). The gamma ray spectroscopy could not be run until after the TWRL (which has temperature limitations) was run. Then the well was flowed back to reduce wellhead pressure.
When only small amounts of gold are in the well bore, the Tracerscan analysis will distinguish the split of what is inside and what is outside the casing. In this case, the amount of gold in the well bore overwhelms the measurement. The gold in the fracture is not distinguished at all.
During the main fracture treatment, the fluid was tagged with scandium and the proppant was tagged with iridium. In this case, very little of either tag appears to be inside the well bore at logging time. The relative distance curve indicates that the detected radiation is in the formation, at least outside the casing.
The Tracerscan interpretation by Halliburton indicates the base of the fracture to be at 9,333 ft and the top of the fracture to be at 9,183 ft, for a height of 150 ft. The base picked by this author is somewhat lower at about 9,390 ft which gives a fracture height of 207 ft.
CBL AND VDL
The cement bond log is a recording of the amplitude of the first arrival at a receiver spaced 3 ft from a 10-35 khz monopole source. The first arrival is usually from the casing. The amplitude of the first arrival is maximum when the casing is unsupported and minimum when the casing is well cemented.
The presence of a very small gap between the casing and the cement, called a microannulus, is enough to cause the received signal to be greatly reduced. This reduced amplitude occurs even when the gap is too small for fluid to pass through.
When the presence of a microannulus is suspected, well bore pressure is applied to expand the diameter of the casing enough to close a microannulus. Then the amplitude of the CBL curve will represent the degree of bonding between the pipe and the cement.
A microannulus was interpreted to be present in SFE No. 3. Prior to perforating the well, a CBL was first recorded without any added well bore pressure. The CBL curve indicated that a significant portion of the interval from 8,800 to 9,600 ft had only poor to moderate pipe-to-cement bonding.
After 2,000 psi of pressure was added at the surface, and the CBL rerun with the added pressure in the well bore, most of the interval appeared to be well bonded. The interpretation is that the well bore pressure closed the microannulus.
After perforating and breaking down the perforations/formation, the extra internal 2,000 psi pressure can no longer be maintained. As a result, the microannulus is always present, but in varying degrees.
Changes in CBL amplitude caused by the fracture operation corrupting the integrity of the bond would be lost in the much larger amplitude changes brought about by only minor changes in the effective thickness of the microannulus. Therefore, the CBL amplitude cannot be used to infer fracture height on the SFE No. 3 well. For this reason the CBL is also left blank on Figs. 1 and 3.
The variable density log is recorded simultaneously with the CBL. The VDL is the amplitude of the full acoustic waveform at a 5-ft spacing from the same acoustic source used for the CBL. Whereas the first arrivals (CBL) come from the casing, the later arrivals (VDL) of the waveform come from the formation.
Generally, the amplitudes of these later arrivals are large when the annulus between the pipe and formation is well cemented, and are weak or nonexistent when no cement is present in the annulus.
Some changes in the late arrival amplitudes were noted on VDL logs recorded after the various completion and fracturing stages. These VDL amplitude changes are reflecting changes in acoustic impedance of the annulus-to-formation region.
The changes in acoustic impedance are interpreted to be caused by the fracture corrupting the acoustic integrity of the region. Note that a fracture could come in contact with the cement sheath and not change the impedance enough to be detected. Therefore, the changes in the VDL amplitudes are also probably only a minimum height indicator. The height from the VDL is interpreted to be 1 64 ft.
CEL
The cement evaluation tool is an ultrasonic pulsed echo device that investigates the integrity of the cement sheath. The pulsed echo recording system permits the use of time gates such that only data coming from the cement sheath are analyzed.
The microannulus and the formation do not influence the gated echo. The tool is very sensitive to changes in acoustic impedance within this narrow region. One would expect to observe changes in impedance with the CEL tool which could not be distinguished with lower frequency tools such as the VDL.
The CEL tool displays the processed ultrasonic data in such a way that channels in the cement sheath can be observed. The quality of the cement is given in eight different segments. It was hoped that the fracturing operation might cause changes in the integrity of the cement sheath that appeared as a channel in the cement sheath to the CEL tool.
Changes in CEL response after fracturing were noted from 9,160 to 9,352 ft indicating a height of 192 ft. Again, these changes are changes in acoustic impedance of a narrow band between the outside of the microannulus and the formation. When the fracture leaves the plane of the well bore, one would then expect little or no change to be seen in the CEL tool response. As such, this also is a minimum height response.
HEIGHT MEASURED
After review of the measurements used to check for height on the SFE No. 3, the minimum height is believed to be 192 ft. This is arrived at by considering each of the four reliable measurements, TWRL, SCAN, VDL, and CEL as being minimum indicators. Then by reasoning that the maximum of the four is still a minimum height, the minimum height of the fracture is 192 ft (Fig. 1).
The CMR indicates the fracture height to be 250 ft. The CMR has a depth of investigation of a few tens of ft. If the fracture is subvertical by a few degrees and since the CMR investigates farther away from the well bore than any of the other tools, one would expect the CMR indicated height to be taller than the other devices.
While this is not a proof statement, it does appear that the shallower investigating devices indicate a shorter fracture, and the deeper investigating device sees a taller fracture, which appears logical if one believes the fracture is subvertical.
The CMR should see the top of the fracture unless it is subvertical by more than a few degrees or the fracture is very tall. In the latter case, even when subvertical by only a few degrees, the fracture top and base are far from the hole for a tall fracture. In such a case, even the CMR height is a minimum height.
Any future tests should include the four tools used and one or more measurements of shear-wave anisotropy. The fracture should cause shear-wave birefringence that can be measured with the shear-wave anisotropy tools. This measurement of birefringence would come from a region slightly deeper than measured by the minimums (TWRL, SCAN, VDL, and CEL) but shallower than the CMR.
Copyright 1991 Oil & Gas Journal. All Rights Reserved.
