Richard A. Fagin
Teleco Oilfield Services Inc.
Houston
Jane E. Trusty, Lisa R. Emmet, Mark K. Mayo
American Exploration Co.
Houston
Dual propagation resistivity (DPR) measurements and modeling helped steer a horizontal well through a thin oil zone to avoid a nearby oil/water contact.
The horizontal well came on-line flowing 380 bo/d with no water production. Offset vertical wells produce an average of 33 bo/d with an 83% saltwater cut.
Until recently, horizontal drilling efforts have been focused on thick reservoirs in which the stratigraphic position of the well bore within the target formation was not critical. Navigation through much thinner reservoirs is now possible with a technique that uses an electromagnetic propagation resistivity tool run with a measurement while-drilling (MWD) collar.1
A complex model of expected tool response is constructed from detailed information regarding the resistivities of the target formation and the zones above and below the formation. The model then becomes the principal navigation guide during horizontal drilling. This technique has proven applicable for steering within a thin oil column above an oil/water contact.
RESISTIVITY MODELING
Electromagnetic propagation resistivity logging while drilling has been available since 1986. A recently created program generates a model of electromagnetic propagation resistivity response in a horizontal well from existing vertical well log data.1
The program is based on a 2-mhz electromagnetic propagation resistivity tool developed by Teleco Oilfield Services Inc. The MWD tool measures two separate formation resistivities. The phase difference resistivity measurement (Rpd) provides high vertical resolution, but the amplitude ratio measurement (Rar) has a greater depth of investigation.2 The program computes expected log values for both types of resistivity in any series of parallel layered media of various thicknesses and true resistivities.
A line normal to the bedding plane and the tool center line form the orientation angle used in these calculations. The model generates logs plotted in true vertical depth (TVD), but the actual field logs are much longer because they are recorded in measured depth.
Fig. 1 shows a simple model of DPR response as the tool approaches and enters a bed, travels horizontally within the bed, and then exits the bed. A significant feature of the modeled response is the distinct spike or "horn" that occurs as the tool enters and exits the bed at a high inclination angle. These elevated values do not represent actual resistivities; rather, these are products of the DPR design caused by resistivities that vary widely across the boundary between two beds.'
These spikes serve as useful indicators of an approaching bed boundary. The spike indications of bed boundaries can help determine when to start correction runs with a steerable motor or to stop drilling if the stratigraphy does not permit further lateral displacement.
A comparison of field logs with the modeled response can also determine tool position within the reservoir.
FIELD APPLICATION
American Exploration Co. used this navigation technique in the Midway field unit (MFU) in Lafayette County, Ark. Production in the field comes from the Smackover formation, a partially dolomitized oolitic limestone. The reservoir has a partial water drive and has been under a waterflood since 1956.
The reservoir has an oil column approximately 30 ft thick and is underlain by 150 ft of water. The oil/water contact tilts, following the structural attitude of the top of the Smackover formation. A shale/anhydrite formation caps the reservoir.
The main objective of horizontal drilling within the Smackover was to increase well bore footage in the oil column, thereby increasing the oil production rate and the drainage area. An additional objective was to remain within the oil column and not penetrate the water zone, thereby decreasing or eliminating water production.
Fig. 2. displays the horizontal well path of MFU 2-15 well in relation to the structure on the Midway Smackover map. The dashed line from the surface location (SL) to the "X" represents the well path up to the Smackover, and the solid line from the X to the bottom hole location (BHL) represents the horizontal displacement within the Smackover.
Fig. 3 shows a cross section of the MFU 2-15 well path from the kickoff point in the Cotton Valley, through the Buckner anhydrite, to the Smackover reservoir. The MFU 2-15 well penetrated a total of 670 ft of oil-productive reservoir rock.
The gamma ray and resistivity logs from a nearby vertical well indicate a large resistivity contrast at the oil/water contact (Fig. 4). The gamma ray contrast between the Smackover limestone and the caprock above (Buckner anhydrite) makes it possible to use MWD gamma ray data to determine whether the well path has penetrated the caprock.
However, the shallow depth of investigation of the gamma ray device prevents its use as a predictive indicator in this example. The mud log samples provided the best indication that the well path was nearing the caprock; a tight dolomitic zone directly above the bed boundary supplied distinct lithologic evidence of this boundary.
Under these conditions of resistivity, DPR modeling can very effectively predict the tool's approach to the oil/water contact. Fig. 1 shows the modeled DPR response at an inclination of 88 based on the resistivity data from the offset vertical well in Fig. 4.
Note that a very obvious spike occurs in both resistivity measurements at the oil /water contact. In addition, the deeper reading Rar measurement decreases to approximately 50 ohm-m at 5 vertical ft above the oil/water contact, making it possible to predict the contact before it is penetrated.
Fig. 5 is a portion of the actual field log from the MFU 2-15 horizontal well. The top of the Smackover limestone occurs at 6,820 ft measured depth (MD) on the DPR log.
Note that the Rar reading spikes off scale at 6,813 ft, a common response in formations with resistivities exceeding 100 ohm-m.3 oil saturations in this well were found to be at or near original oil saturations, a condition which had not been anticipated. Tool parameters could have been reset to display the amplitude ratio curve had these oil saturation values been anticipated. The Rpd curve remains on scale throughout the section, indicating high formation resistivities.
At the final log data point (7,450 ft MD), the Rar curve had not come back on scale. These high Rar values were well above the 50 ohm-m cutoff point that would have indicated penetration of the oil/water contact. The well path at this point had an inclination of 86, and the formation dipped at 7.
American Exploration decided to drop angle and continue drilling because comparison of the model and actual log data indicated that the well path was safely above the oil/water contact. However, the drilling assembly stuck differentially during an attempt to slide the motor, and American Exploration chose to stop drilling at 7,491 ft MD. The well was subsequently completed with 670 ft of horizontal section in the reservoir.
The actual field logs in this well differed from the modeled response because of the higher than anticipated resistivities encountered. In general, the modeled response will differ from actual field data if the estimated bed thicknesses, formation resistivities, borehole conditions, or invasion profiles vary significantly from the parameters used for modeling. These factors should be taken into consideration when using DPR data for navigation of horizontal wells.
This successful horizontal completion met all three of the main objectives: increased oil production, increased drainage area, and decreased water production. American Exploration plans additional horizontal wells for the Midway field and will also use MWD data as a navigation guide in these wells.
REFERENCES
- Wu, J.Q., Wisler, M.M., and Barnett, W.C., "Bed Boundary Detection Using Resistivity Sensor in Drilling Horizontal Wells," paper presented at the Spwla 32nd Annual Logging Symposium, Midland, Tex., June 1991.
- Wisler, M.M., "Real-Time Electromagnetic Propagation Resistivity and Memory for MWD," paper presented at the 11th European Formation Evaluation Symposium, Oslo, Norway, September 1988.
- Fredericks, P.D., Hearn, F.P., and Wisler, M.M., "Formation Evaluation While Drilling with a Dual Propagation Resistivity Tool," paper presented at the SPE 64th Annual Technical Conference and Exhibition, San Antonio, October 1989.
Copyright 1991 Oil & Gas Journal. All Rights Reserved.
