DEVONIAN GAS TECHNOLOGY-3 MICROFRACTURING AND NEW TOOLS IMPROVE FORMATION ANALYSIS

D.E. McMechan, Jim J. Venditto, Tim Heemstra
Halliburton Services
Duncan, Okla.

G. Simpson
Halliburton Logging Services
Houston

L.L. Friend, Ed Rothman
Columbia Natural Resources Inc.
Charleston, W.Va.

Microfracturing with nitrogen, an experimental extensometer, stress profile determination from wire line logs, and temperature logging in air-filled holes are new tools and techniques that add resolution to Devonian shale gas well analysis.

Columbia Natural Resources (CNR) may integrate some of these in its 1992-93 Devonian shale gas well program.

This is the third of five articles detailing the technology evaluated in a joint venture program between Halliburton Co. and CNR in a test well in Roane County, W.Va.1

MICROFRACTURING

Microfracturing creates small fractures by injecting small amounts of fluid at very low rates.

Microfracs are created usually at several different depths to determine stress variation as a function of depth and rock type. To obtain an oriented core containing the fracture, the formation is microfractured during drilling.

These tests are critical in establishing basic open hole parameters for designing the main fracture treatment.

Information obtained from microfracturing includes stress magnitude, stress orientation, calibration of logs, fracture orientation, and fracture deviation.

In the study well, six microfracs were created (Tables I and 2).

The first, above the lower Huron, obtained the stress above the upper pay interval (Fig. 1).
The second, at the top of the first coring point, obtained the stress in the lower Huron and created an hydraulic fracture in the top part of the subsequent core.
The third, in the lower part of the cored interval created by the second microfrac, validated the experimental extensometer (Figs. 2 and 3).
The fourth obtained the stress in the Rhinestreet interval in the lower Alexander siltstone.
The fifth (Fig. 4) was at the coring point in the Rhinestreet.
This also validated extensometer results.
The sixth, at TD, also obtained the stress.

EXTENSOMETER

This study included the first downhole test of the experimental extensometer for measuring borehole deformation during fracturing.

The extensometer is set between compression packers (Fig. 5). The 12 arms open to contact the borehole wall.

When pressure is applied between the packers during a microfrac test, the arms can detect borehole wall deflections as low as 1/1,000 in.

When the formation breaks, the extensometer's displacement pattern determines the fracture direction and directional shear modulus.

An orientation module and quartz-based pressure transducer are in the instrument package. The orientation module includes a tri-axial magnetometer, three accelerometers, and a temperature sensor for compensation.

Gravity and magnetic field measurements are used to orient the extensometer. The resulting three angles (drift azimuth, drift angle, and gravity tool face) describe the instrument's orientation. Magnetic tool face and magnetic tool azimuth, the remaining "driller's angles," are also useful.

The extensometer is run on a tool string that includes a formation tester valve for pressure testing the tubulars.

A wire line unit is needed to transmit power and signals to and from the downhole instrument.

Analog signals are converted to digital signals downhole. These multiplexed digital signals are then transmitted to surface through a single conductor cable.

The field computer displays and stores data from the extensometer and the surface flow and pressure transducers.

The pumping unit is the same as for conventional microfracture tests.

Data for directional modulus calculations are listed in Table 3. Fig. 6 shows the deformation caused when the hydraulic fracture reopens.

From this and similar data a fracturing direction of N60E was determined.

OPEN HOLE LOGGING

Wire line open hole logging in the study well included four major objectives:

Identify naturally fractured zones to increase production.
Estimate the borehole stress profile to improve hydraulic fracture design.
Evaluate fracture height and direction to improve future hydraulic fracture design.
Show the effectiveness of collecting log data in air-filled boreholes. Fig. 7 is a gamma-density log run in the study well, under air-filled conditions.

The circumferential acoustic scanning (CAS) tool and temperature logs identified fractured intervals.

The full waveform sonic (FWS) log, sidewall neutron log, spectral litho-density log (SLD), dual induction log (DIL), and high-resolution induction (HRI) log obtained supporting evidence of fractures.

A six-electrode dipmeter (SED) log is normally run in liquid only, but on the study well it was run successfully in both liquid and in air.

AIR SUITE

The air-filled borehole logging suite, consisting of resistivity, porosity, and dipmeter logs, demonstrated the capability to record accurate log data in an air-filled borehole. The SLD log evaluated the feasibility of recording formation density and photoelectric factor in the air-filled borehole.

The log data from the air-filled well were compared with data from the same well filled with water. The comparison indicated that measurements were valid when there was good pad contact with the formation. Even when pad contact is poor, the information was useful. Elliptical boreholes result from borehole breakout or the collapse of a well bore under stress from the formation.

Hypothetically, the long axis of the well bore is parallel to the least principal horizontal stress. Therefore, poor pad contact could indicate natural fractures in the formation. Thus, SLD log response may validate natural fracture indicators on other logs.

Formation density measurements also indicated that the Devonian shale was not completely homogeneous. Density variations indicated thin mudstone and siltstone beds that may have higher permeability than surrounding beds. These permeable beds may act as conduits to increase production in the same manner as natural fractures.

The sidewall neutron tool contains a pad-mounted detector and therefore can measure neutron porosity in an air-filled borehole. Because the tool maintains contact with the well bore wall, the detrimental effects of the air column on porosity measurement are minimal.

Rugose and elliptical boreholes affect the neutron porosity measurement in a way different from that of the density porosity measurement. When the pad of a neutron tool moves away from the borehole wall, measured neutron porosity is less than actual formation porosity.

On the other hand, when the pad of the density tool moves away from the borehole wall, the measured density porosity is greater than the actual formation porosity.

A similar effect is observed for both tools in gas-bearing formations when there is good pad contact. Because of the neutron tool's greater depth of investigation, lost pad contact did not affect the neutron tool as much as the SLD tool.

The neutron log, in combination with the density log, indicated possible mudstones and siltstones that might be permeable conduits of formation gas to the borehole.

A temperature log detected cooling anomalies and indicated gas entry into the borehole. The temperature log is one of the most effective methods of detecting gas entry into an air-filled well.

As gas enters a well bore from a formation, considerable pressure drop across the formation wall causes the gas to expand. This expansion produces the cooling effect detected by the temperature log.

Some of the cooling anomalies in the study well were located at the microfracs. This indicated that gas was entering the borehole over these intervals.

A comparison of the two induction logs run, the dual induction log (DIL) and the HRI log, showed that the HRI tool identified thinner features than the DIL. The HRI has a superior 2-ft vertical bed resolution.

The DIL was run in an air-filled borehole, but the HRI log was run with 2% KCI water in most of the hole. The upper portion of the HRI log was recorded in air.

The coil configuration of the HRI tool allows deep, medium, and shallow resistivities to measure beds 2-ft thick from the same point. For the HRI, the depth of investigation for deep resistivity measurement is 91 in. This is approximately 40% deeper than the corresponding DIL measurement.

A comparison of the HRI log with the SLD and borehole stress logs showed that SLD porosity increased primarily when the HRI deep resistivity measurement (HRd) was greater than 35 ohm-m. In some instances, the SLD porosity increased with no apparent correlation to the HRI log.

This porosity increase was interpreted as indicating naturally fractured zones. The comparison of the HRI log and the borehole stress log showed that the calculated fracture pressure closure decreased when HRd exceeded 20 ohm-m. The highest stress zones occurred when HRd was less than 20 ohm-m.

In more highly resistive zones, the HRI log read higher resistivities than the DIL log. The HRI log showed that the Devonian shale was not homogeneous and has many zones that consist of layers of 2-3 ft beds.

Because the DIL tool has a coarser vertical resolution, its measurements are averaged over these bedding features. These thin beds were most likely associated with higher permeability fingers that could act as conduits to feed gas to an hydraulically induced fracture.

The DIL tool was the standard resistivity device run in the Devonian shale formation. Similar to the HRI tool the DIL tool can measure resistivities in air-filled boreholes; however, its deep resistivity measurement has a 65-in. depth of investigation and 60-in. vertical resolution.

LIQUID SUITE

The CAS tool measures high-frequency acoustic waves reflected from the borehole wall. It is used primarily to identify natural fractures and bedding planes. Additionally, the CAS tool can measure the direction of hydraulically induced fractures created by microfrac tests conducted at various intervals in the open hole portion of a well.

The CAS tool log indicated that the direction of the hydraulically induced fractures created by the microfrac test was in the ENE direction. The borehole, from the cross sections, was eroded around the induced fractures created by the microfrac.

The direction of the fracture from these cross sections is ENE. Each cylinder in the figure represents a different view around the borehole. The hydraulically induced fracture can be seen between the breakouts.

FWS data recorded were used to calculate a stress profile and thereby identify possible barriers to fracture extension. The FWS log records sonic waveforms for calculating compressional and shear travel times. These were used to compute the stress profile.

This profile was plotted on the borehole stress log as the fracture closure pressure. The log showed a relatively flat profile, but the subtle changes in stress located the perforation intervals. The stress profile is the key input to the hydraulic fracture design. If longer fracture extension is achieved, the probability of an early screenout can be reduced.

Three microfracs were performed over the interval logged by the FWS tool. The deepest such microfrac was at 5,010 ft, and the closure pressure recorded from the microfrac was 1,830 psi. The value estimated from the log was essentially the same.

The middle microfrac was at 4,780 ft. The closure pressure recorded from this microfrac was about 1,500 psi. The estimated value was 1,750 psi, a 17% difference.

The third test was conducted at 4,560 ft. The average closure pressure recorded from the microfrac was about 1,740 psi. The estimated value was essentially the same as the microfrac measurement.

The microfrac and log comparison demonstrated that the log-based stress estimates were valid. The favorable estimates also suggested that regional stress, plasticity, and low porosity had little effect on the accuracy of the poro-elastic model used to generate the stress curve.

DIPMETER

A six-electrode dipmeter (SED) verified the feasibility of deploying the tool both in air-filled and water-filled wells.

Because the dipmeter is a pad-contact tool, the air column should not interfere with the measurement. In addition to conventional pads, knife-blade contacts similar to those found on pads run in oil-based muds improved the contact with the borehole wall.

The calculated dips were more scattered on the air-filled log because of occasional loss of pad contact.

Test results show that the SED tool performed better in the water-filled well, but the comparison of the computed results proved satisfactory in either case.

The SED six-arm caliper measurements improved the description of the well bore shape. In the depth track of the computed dipmeter log, a borehole diagram was constructed from the six caliper measurements.

These periodical plots indicated the shape of the borehole. The ellipticity of the borehole can be correlated with natural fractures and can be used to determine the orientation of borehole breakout and hydraulic fractures.

REFERENCE

Sweeney, J., "Oil and Gas Report and Maps of Wirt, Roane, and Calhoun Counties, West Virginia," West Virginia Geological & Economic Survey, Bulletin B-40, 1986.