TECHNOLOGY Slim-hole horizontal well improves gas storage field deliverability

Mark E. Gredell
Panhandle Eastern Pipe Line Co.
Houston

Mark A. Benson
Slimdril International Inc.
Houston

Based on a presentation at the American Gas Association Operations Conference, Las Vegas, May 7-10, 1995.

A slim-hole horizontal well in a gas storage field, drilled with a 2,000-ft lateral section under a city, initially produced about four times more than a nearby offset vertical well.

This horizontal well will increase the amount of gas cycled from the storage field each year. The slim-hole well design helped keep costs down considerably.

The ability of the well to cycle gas efficiently from the area under the city will be determined by monitoring the future performance of the well and field. The objectives and conditions of this project were ideal for a slim-hole horizontal well, and the results suggest the potential for more horizontal slim-hole wells in other gas storage applications. The slim-hole well design helped lower total costs.

Approximately one third of the Howell gas storage field underlies the city of Howell, Mich. Efficient cycling of the gas in this portion of the field is limited by the lack of vertical wells.

Among the technical and operational issues addressed on this horizontal well project were prediction of well performance and benefits, environmental and safety concerns of drilling in an urban area, optimizing well design parameters, protecting the integrity of the storage zone, and geosteering in a thin reservoir.

The Howell gas storage field is located in Livingston County, Mich., about 50 miles west of Detroit in the southern portion of the Michigan basin. Detroit is a major market area for Panhandle Eastern Pipe Line Co.s pipeline system. Because of the Howell fields location, it plays a critical role in supplying peak gas needs for the pipeline during the winter heating season.

The need for more gas storage capacity on Panhandles system prompted a reassessment of the possibility of expanding existing fields operated by the pipeline. Because most of the fields have a rather high well density, as is typical of older storage fields, the potential for expansion by infill or replacement drilling is limited.

The Howell storage field, however, was an exception. The portion of the reservoir under the city has few existing well penetrations. Pressure data from field shut-in periods showed the existing wells were inefficient in cycling gas into and out of this area. Horizontal drilling was proposed as a method of accessing the portion of the reservoir underlying the city.

The following were the primary objectives in drilling a horizontal well at Howell:

• Increase the amount of gas that could be cycled each year from the field by more efficiently draining the area under the city and converting a portion of the base gas to working gas

• Extend the high deliverability of the field later into the winter withdrawal season by developing this unused capacity of the field.

Reservoir description

The reservoir at Howell is the Silurian age Brown Niagaran, which is an important productive interval throughout the Michigan basin. The Niagaran is also the reservoir for many other Michigan gas storage fields.

In the Howell field area, the Niagaran (or Brown as it is known locally) is a very hard, dense, crystalline, vuggy dolomite. It is occasionally stylolitic and sometimes fractured, although fracturing is uncommon in the better-quality, more vugular rock. Vugs in the best rock range in size from pinpoint to pencil diameter.

The structure at the top of the Brown Niagaran is an elongated anticlinal feature (Fig. 1)(54749 bytes). A fairway of good quality reservoir rock lies just off the crest of the structure along the northeast side. This fairway parallels the crest. The deterioration of rock quality near the crest accounts for the lack of wells in the area. Permeability and porosity also decrease downdip on the outside edges of the structure.

In contrast to the numerous thick Niagaran pinnacle reefs present in the Michigan basin, the Niagaran in the Howell area is relatively thin and has a generally uniform thickness, ranging 10-16 ft, across the field. Fig. 2 (56286 bytes) is a gross feet isopach of the Brown Niagaran. Note the trend of thicker interval lies off the crest of the structure.

Within this gross thickness, the best quality reservoir rock is about 5-ft thick and is identifiable on porosity logs as intervals with greater than 6% porosity. This high-quality rock occurring in a generally discrete interval at the top of the Niagaran was the preferred target within the formation.

The average porosity is about 10%, and the average permeability is about 50 md, although it can range up to 150 md. Reservoir homogeneity is excellent, with estimates of vertical to horizontal permeability ratios of 0.5 or higher. The average depth of the reservoir is about 3,900 ft. The dip rates in the Niagaran interval and above are low (about 1) but can be locally variable. There is no known faulting at the reservoir level.

Field development

The Howell field is a depleted volumetric gas reservoir. It was discovered in 1946 and developed on 160-acre spacing with about 15 wells. Only three of these wells were drilled within the city limits as it existed at the time. The formation pressure at discovery was 1,840 psia.

The field was converted to gas storage in 1961. The first year of gas injection was 1962. The total production to the time of conversion was 22.5 bcf. New wells drilled for storage after 1962 resulted in a total of 66 injection/withdrawal wells in the field and three pressure observation wells. None of the new drilling was inside the city limits, although the three old producing wells within the city limits were converted to storage wells. The last drilling in the field was in 1967.

The field facilities include a compressor station with four engines totaling 6,000 hp. The maximum withdrawal rate from the field is 360 MMcfd, and the maximum injection rate is 120 MMcfd. Very little water or condensate is produced in this essentially dry-gas reservoir.

The total capacity of the field is 32 bcf, and the developed working gas capacity was 15.6 bcf prior to the drilling of the horizontal well.

Feasibility studies

Considerable effort was given to prespud planning. The first step in the process of evaluating the potential benefits of a horizontal well was an updated geological description of the reservoir using all available core, log, well testing, and field performance data.

This study suggested the fairway of good reservoir rock extended from the high well density southeast portion of the field into the area underlying the city to the northwest. With this target area identified, several potential drill sites were considered, although the urban nature of the area favored existing well locations. The McPherson 1-35 well location near the southern edge of the city was chosen as the best candidate for further evaluation.

A full-field reservoir simulation was then undertaken to examine some critical points, including the following:

• The amount of additional gas that could be cycled with a horizontal well at the McPherson 1-35 location

• Interference effects with existing vertical wells

• Sensitivities of the results to the rock properties and geological description assumed

• Sensitivities of the results to various well parameters such as lateral length or skin damage.

After a successful history match of actual field and well performance, a horizontal well at the McPherson 1-35 location was added to the simulation. A late-season operating scenario for the horizontal well was assumed, using the well only in the last 30-40 days of a full withdrawal season.

Because the new well would be adjacent to an existing vertical well, the simulated performance of the horizontal well could be compared directly to actual performance of the McPherson 1-35. Incremental recovery of gas with the horizontal well over that of the vertical well showed an additional 1.0 bcf could be recovered under the above assumptions. Simulated flow rates of the horizontal well were 2.5-4 times the rate of the vertical well.

Sensitivity of results showed that changes in porosities and permeabilities in the area under the city had minor effect on the quality of the history match, lending credence to data showing poor drainage of this area with existing wells. These changes also had the greatest effect on incremental recovery from the horizontal well.

Interference effects appeared to be minimal in the simulation assuming a lateral length up to about 2,000 ft. Lengths beyond 2,000 ft showed progressively greater interference effects with existing wells and thus less incremental recovery (Fig. 3)(20664 bytes).

Well location

The drill site selected for the horizontal well was an existing well pad inside an industrial park in the southwest part of the city.

The existing vertical well located there, the McPherson 1-35, was an old completion with 412-in. casing cemented to surface, making a re-entry and kick off from this well bore impractical.

The decision was made to drill the proposed horizontal well, the McPherson 3-35, from the surface on the same pad, 90 ft from the existing wellhead. The drill site was about 2 acres and within a few hundred feet of a bottling plant and other industries. Numerous residences were within a quarter mile of the drill site.

The urban location made noise, safety, and environmental concerns a high priority. Minimizing these concerns absorbed a large part of the planning process prior to the drilling of the well. The rig was equipped with hospital-quiet mufflers.

A closed-loop mud system with all steel pits was used, and all drilling wastes were hauled away to an approved disposal facility. A berm was constructed around half the drill site for runoff and liquids spill control, and silt fencing was erected around the entire drill site. A small wetland identified during the environmental impact surveys was protected.

Although these considerations increased the costs of the well substantially, they were considered very important not only in meeting regulatory approval but in maintaining a good working relationship with the city of Howell.

Because the field is under the jurisdiction of the Federal Energy Regulatory Commission, a special filing was necessary in addition to the usual state permits.

Well plan

To minimize formation damage, the drilling was planned to coincide with maximum reservoir pressures that occur near the end of the summer injection cycle.

A slim-hole well design was chosen to allow minimal rig requirements and conserve space. The trajectory for the horizontal well was planned to be downdip directly to the east throughout its length. The geological data, available from the old vertical well on the same pad, were used to plan the horizontal well. Although formation dips were low, some uncertainty in formation tops was anticipated because of subtle changes in formation thickness and localized variations in dips.

The well path planned was a medium-radius design to ensure landing the hole properly within the storage zone. The formations encountered in the curve are a carbonate-evaporite sequence.

At critical points in the well path were two thick salt sections, the Salina B and A-2 salts. Below the A-2 salt is the A-1 dolomite, a ratty and tight section. The A-1 anhydrite, the seal for the storage zone, lies directly above the Brown Niagaran. The A-1 is a very dense anhydrite and is generally about 15-20 ft thick. A geologic column with lithologies encountered in the radius section is shown in Fig. 4 (57508 bytes).

The well plan called for 858-in. casing to be set at about 2,000 ft. A 778-in. hole was planned from that point into the storage zone, with a 600-ft radius in the curve to ensure that casing could be run to the bottom of the hole. The kick-off point was planned to be at 3,304 ft in the Salina B salt because of the enhanced ability to kick off in a softer salt and to avoid steering problems in the carbonate section.

The importance of protecting the integrity of the storage zone dictated plans to set 512-in. casing through the curve and at least 50 ft into the Niagaran.

Based on sensitivity results from the simulation, a 2,000-ft lateral with a 434-in. hole size was planned. Because of the hard, competent nature of the Niagaran, the decision was made to complete the lateral open hole (Fig. 5)(50782 bytes).

Radius

The bottom hole assembly used in drilling the radius included a 778-in. F47H rock bit, a 434-in., medium-speed positive displacement motor (PDM), an orienting/circulating sub, and two nonmagnetic drill collars (Fig. 6)(46377 bytes). A gamma ray logging-while-drilling (LWD) tool was used along with a rotatable wet connect.

The rotatable wet connect allowed real-time gamma ray logging while in the rotary mode.

A saturated-brine drilling fluid was used in the radius to minimize washouts in the massive salt sections. After a successful kick off directly to the east of the surface location, the bottom hole assembly performed as predicted in the radius, building 11/100 ft in the salt sections and 9/100 ft in the carbonates.

Drilling continued to 4,104 ft measured depth (MD), the first indication of the top of the Niagaran. After circulating samples, drilling continued to 4,179 ft MD, where the gamma LWD and samples confirmed the top of the Niagaran had been encountered higher than expected. Because of the angle build rate at the time and the thinness of the target, correcting the angle enough to land the hole properly raised concerns about running the casing.

The decision was made to turn the angle down, drill all the way through the base of the Brown Niagaran, log the interval, and plug back for a sidetrack. This unplanned log run was very useful in allowing landing the hole properly in the second attempt.

The sidetracking operations were uneventful, and the hole was landed 67 ft into the storage zone at close to 90 inclination, in good position for drilling out of the casing shoe. The hole was reamed prior to running the 512-in. casing, and the string was run and cemented with no problems.

Lateral

The operations in the lateral had two primary concerns:

• The lateral section of the hole was drilled with a brine system both to attempt to minimize formation damage by using a clear drilling fluid and to maintain the weight necessary for control at the high reservoir pressures existing at the time of drilling.

• The decision to use the gamma LWD tool was based on the importance of keeping the well bore within the target zone, and the gamma in combination with the mud logger data provided the information needed to accomplish this goal.

After the landing collar, cement, and float shoe were milled, a new bottom hole assembly was picked up to drill the lateral. This assembly consisted of a 434-in. MP-48 polycrystalline diamond compact (PDC) bit, 338-in. Slimdril International steerable motor, 338-in. orienting/circulating sub assembly, and two Monel drill collars ( Figs. 7 (62466 bytes)and 8 (66226 bytes)).

After the lateral was begun, the first 21 ft of the formation were drilled at very high rates of penetration, some exceeding 1 fpm, in very good quality reservoir rock.

During a connection, a kick was taken, unloading most of the annular hole volume. Because no problems were encountered while the 778-in. hole was drilled under the same conditions, one of the factors in this kick was believed to be the much higher ratio of drilled gas to small annular hole volume in the 434-in. slim hole compared to the 778-in. hole. After the well was killed and the mud weight stabilized, drilling resumed.

From this point, the rate of penetration was limited to control the amount of gas influx into the annulus. The maximum safe penetration rate was determined by monitoring the amount and trend of background gas and connection gas at the mud logging unit. Drilling continued with the penetration rate limited to 20-30 ft/hr through the remainder of the lateral section.

In keeping with these planned drilling parameters, the mud logger was critical in monitoring the connection and background gas to adjust the penetration rate as the lateral operation continued.

Careful monitoring of cuttings was also important in steering control, particularly because the penetration rate could not be reliably used as a qualitative indication of rock quality or formation changes.

The real-time gamma log proved to be very useful in this respect also and needed to be integrated with the mud log data continuously. The gamma and mud log information complemented each other very well.

Special attention was given to maintaining low solids in the mud system to minimize formation damage, torque, and drag and to sustain proper mud weight. High-viscosity polymer sweeps were run periodically and appeared effective in keeping the hole clean, in addition to providing better cuttings quality.

The drilling of the lateral was completed with no other problems. The hole was successfully kept within the target interval of the formation throughout the 2,000-ft length. Close attention to operations and the quality and continuity of rock encountered contributed to keeping the well bore within the narrow target.

Several factors were important in accomplishing the drilling goals of this project:

• Considerable planning went into the well before the rig was moved on location.

• There was flexibility to adjust drilling parameters as conditions warranted.

• It was possible to steer the lateral very well, partially because of good monitoring of the cuttings and gamma ray counts and partially because of favorable geologic conditions (low dips and good quality rock).

• The directional drillers, drilling contractors, mud loggers, and supervisory personnel were capable and competent.

• Good communication, including daily morning meetings, existed among all parties present at the well site.

• The bottom hole assembly and bits performed well and required no lost time in looking for the correct combination of drilling assemblies under the specific conditions of this well.

Completion

The well was completed open hole. To ensure cleanup of the well bore, a coiled tubing unit unloaded the well with nitrogen and performed a small 6,000-gal acid wash with 28% HCl.

The coiled tubing showed the hole to be relatively clean because there were few tight spots encountered as it was run in the hole.

The well cleaned up and stabilized quickly on flow back. Surface restrictions prevented flowing the well at more than 17 MMcfd during this initial cleanup. After the well was tied into the field gathering system, the well flowed at stabilized rates up to 100 MMcfd.

The well was tested under the constraints imposed by the daily operational needs of the pipeline. Modified multipoint isochronal tests were conducted using surface pressure gauges. Downhole gauges were not used because of the high rates expected and risk of running downhole tools.

Results indicated the well was performing at rates about four times the rate of the offset vertical well at equivalent pressure drawdowns.

Future performance

The principle goal of the project, to increase the available amount of working gas from the field, will be evaluated by monitoring field and well performance through at least two full withdrawal cycles.

The operating plan for the well at this point is to use it only in the last portion of the withdrawal season. The data will be entered into the reservoir model to compare actual recoveries and performance with those predicted in the simulation. The spring field shut-in tests will show how the horizontal well has influenced patterns of pressure distribution in this portion of the field.

On the basis of these results, the manner and timing of the use of the well can be adjusted to maximize the efficiency of gas recovery in the portion of the reservoir drained by the well.

To minimize the risk of formation damage, the well will not be used for gas injection operations for at least the first 2 years. Evaluation of pressure distribution patterns will help determine if it is necessary to use the well for injection in the future.

Results

The benefits of a horizontal well in the Howell storage field were quantified by use of detailed reservoir description and reservoir simulation, and the simulation results were used to justify project costs.

Extensive predrill planning, with the input and advice of all the contractors on the well, was considered crucial to accomplishing the drilling objectives.

The integrity of the storage zone was protected by successfully running and cementing casing into the formation.

Careful mud logging and LWD monitoring enabled the lateral extension to reach the planned 2,000 ft and stay entirely within the target.

Through commitments of considerable time and expense, environmental and safety concerns were addressed with minimal disruption to area businesses and residents.

Acknowledgment

The authors wish to thank Karen Beaver, Larry Jenkins, and George Hodges of Panhandle Eastern Pipe Line Co. for their critique of this article and Panhandle Eastern and Slimdril International Inc. for permission to publish the article. The authors also thank the Gas Research Institute which cofunded this project. n

The Authors