HORIZONTAL-WELL PILOT WATERFLOOD TESTS SHALLOW, ABANDONED FIELD

J.L. McAlpine
White Buffalo Petroleum Co.
Tulsa

Sada D. Joshi
Joshi Technologies International Inc.
Tulsa

The suitability of using horizontal wells in a waterflood of shallow, partially depleted sands will be tested in the Jennings field in Oklahoma.

The vertical wells drilled in the Jennings field intersect several well-known formations such as Red Fork, Misner, and Bartlesville sand. Most of these formations have been produced over a number of years, and presently no wells are producing in the field. In the 1940s, 1950s, and 1960s, wells were drilled on 10-acre spacing, and the last well was plugged in 1961.

The field was produced only on primary production and produced approximately 1 million bbl of oil. Because the field was not waterflooded, a large potential exists to produce from the field using secondary methods. To improve the economics for the secondary process, a combination of horizontal and vertical wells was considered.

Criteria for selecting the reservoir was as follows:

To make the process economic, the reservoir should not have been waterflooded.
The reservoir should have structural traps. The traps will help contain the fluid movement to only the reservoir structure.
The reservoir should be fairly thick (above 50 ft), with a dome-type structure. This structure will facilitate waterflood from below by injecting water in the lower reservoir portion, near the flanks, and producing from horizontal wells drilled near the structure top.
The reservoir should have fairly good vertical communication.
If possible, the reservoir should be shallow so that horizontal drilling costs are minimal and commensurate with expected reserves from a horizontal well in a partially depleted field.
Additionally, because vertical wells will be used as injectors, a shallow reservoir would minimize injection costs.

RESERVOIR SELECTION AND TESTING

The Bartlesville sand in the Jennings field in Creek County, Okla., was selected as a horizontal well candidate based upon the preceding criteria. Fig. 1 shows that the field contains the good structural traps that are required for pressure maintenance using water injection.

The old well records and electrical log records were obtained from the Oklahoma well log library in Tulsa. On the average, vertical wells were drilled at 10-acre spacings. This provided sufficient data to generate structure top and base maps (Fig. 1).

A detailed geological analysis was conducted to obtain the total pore volume, estimated to be about 40 million reservoir bbl.

It was difficult to estimate reservoir properties such as permeability, and present pressure and saturations from old well logs. Therefore, a decision was made to drill a vertical well as a test well for information gathering. Initially, this would be a production well that later if necessary could be converted to an injection well.

The vertical well, Oller No. 1, was drilled in April 1991. The well was cored through the Red Fork, Misner, Bartlesville sand, and Wilcox. The cores were used to estimate porosity, permeability and the location of tight streaks in the vertical plane.

Figs. 2 and 3 show core data plots or permeability against porosity. This information was further confirmed by running gamma ray, resistivity, spontaneous potential, and neutron-porosity logs.

Based upon the log and core information, the Bartlesville sand was divided into three zones, namely:

The top, 9 ft thick with 13% porosity, 6 md horizontal permeability, and 0.6 md vertical permeability
The middle, 31 ft thick with 16% porosity, 18 md horizontal permeability, and 2.4 md vertical permeability
The bottom, 30 ft thick with 17% porosity, 66 md horizontal permeability, and 33 md vertical permeability.

In addition to core analysis, drill stem tests (DST) were also conducted in the Red Fork and Bartlesville sand. The DST analysis showed a present reservoir pressure of about 780-800 psi.

Using structural maps, log information, core information, and DST information, a three dimensional reservoir simulator was developed for the field. The present oil-in-place is estimated to be about 16 million st-tk bbl of oil in the structure. Several reservoir simulation runs were made by varying the:

Number of water injection wells
Rate of water injection
Number of horizontal wells
Production rate from a horizontal well.

It was decided that more than three horizontal wells will be required to obtain expected oil recovery in an economic time frame. It is interesting to note that even after 20 years, the production rates above 30 bo/d are estimated. These rates indicate a significant oil production potential.

To verify the horizontal well model, the production data from the vertical well, Oller No. 1 is being monitored.

The production data of Oller No. 1 is matched using a vertical well model. In addition, a horizontal well has recently been drilled and is under production testing.

HORIZONTAL WELL

As noted, the first horizontal well was drilled as a test well. The performance of this well will dictate the future course of action.

The geological structure is fairly well-known due to a large number of vertical wells drilled in the 1950s. The detailed geological analyses made us comfortable about the structure. It was decided to drill the first horizontal well in the top structure region in the middle zone. This would provide over 60 ft of vertical stand-off from the vertical injectors in the structure flank areas.

Additionally, the reservoir pressure at 2,700 ft was only 800 psi while a fresh water gradient of 0.43 psi/ft would give a bottom hole pressure of 1, 1 61 psi. Thus, the drilling fluid would have to be reduced in weight (density) by adding air or the drilling has to be conducted with significant lost circulation.

The lost circulation may pose a well control problem. To circumvent these problems, the vertical and curved portion of the well was drilled and cemented to the end of the curved.

First, 9 5/8-in surface casing was set at 657 ft. Then 7-in. casing was set at 90 and cemented to the vertical.

A horizontal portion of 1,480 ft length was drilled using a specially designed mud. As expected, there was lost circulation while drilling.

Surface rotary drilling was used to drill the horizontal.

Because measurement-while-drilling (MWD) was not run, multiple short surveys determined the directional well location.

The well was drilled with an approximate 600 ft turning radius. About 5 days were required to drill 1,480 ft horizontally.

The well was left open hole and washed lightly with acid. In early July, a pump was run into the hole. As time progresses, the well is expected to give increasing oil cut, especially after a large portion of the lost drilling fluid is recovered.

Additionally, water injection has been started in flank wells which are vertical.

Production will be monitored to fine-tune the models. This will help optimize the number of wells required to obtain the best possible rate of return.