ULTRAFINE CEMENT SEALS SLOW LEAK IN CASING COLLAR

Doug MacEachern
Shell Offshore Inc.
New Orleans

Stephen C. Young
Halliburton Services
Harvey, La.

An ultrafine cement squeeze effectively sealed a difficult casing collar leak in the protective casing in a deep, high-temperature well in Mobile Bay.

The leak was sealed in one operation without perforating the casing, giving greater confidence in casing integrity and allowing the well to be drilled to total depth (TD). Restoring pressure integrity of the casing with this procedure saved approximately $250,000.

The well is in the Fairway field, located at the mouth of Mobile Bay within Alabama state waters approximately 2.5 miles south of Dauphin Island. The main geological objective in Mobile Bay is the Jurassic Norphlet formation, a single eolian sand package that ranges in depth from 20,000 to 22,000 ft.1 Natural gas produced from the Norphlet formation in the Fairway field is sour with concentrations of H2S ranging from 100 ppm to 1.6% (16,000 ppm).

WELL DATA

Fig. 1 is a schematic showing the drilling plan for a typical well drilled through the Norphlet formation at 21,800 ft true vertical depth (TVD).

The thermal gradient for wells in the Fairway field is about 1.5 F./100 ft. Pore pressures are low, and the wells are only slightly geopressured. Water-based mud weighing between about 10.0 ppg and 11.5 ppg is used for drilling the interval covered by the protective casing. Water-based or oil-based muds with densities between 11.5 ppg and 13.0 ppg are used to drill out of the protective casing and on to TD.

In accordance with the well plan, a string of 10-3/4 in. x 9-5/8 in. protective casing was run and cemented in place at 16,502 ft (Fig. 2). A casing integrity test was performed before the float collar was drilled out.

The plans included testing the casing to a 16.2 ppg equivalent mud weight. This procedure required a surface pressure of 5,300 psi on top of the hydrostatic pressure from the 10.1 ppg mud.

A leak in the casing was detected as the surface pressure reached 3,000 psi. Pumping operations were stopped and the pressure dropped to 1,900 psi after 15 min. A retrievable packer used to isolate sections of the protective casing helped find the leak between 12,978 ft and 12,983 ft. According to the casing tally, the leak appeared to be at or near a collar in the casing string.

Because the volume loss associated with the leak was low, it was assumed that the connection was leaking and that no damage to the body of the casing had occurred.

OPTIONS

Casing integrity had to be fully restored. Therefore, several options were considered to correct the problem, including mechanical seals such as casing patches or lead seals and squeezing the leak with cement or other sealants.

The low leak off rate suggested that conventional Portland cement would probably not penetrate the opening to effect a seal. Perforating the casing would, in all likelihood, be required to successfully place conventional cement behind casing to establish a seal. In place of cement, epoxy resins could be used to penetrate the existing opening without perforating; however, epoxy resins are more difficult to handle in field operations than cement.

After consideration of the various options, squeeze cementing with ultrafine cement was selected as the first choice to attempt to seal the leak. Ultrafine cement was selected because:

It is composed of American Petroleum Institute (API) cement which can be easily handled in field operations.
Ultrafine cements have a documented, high success rate when used to squeeze casing leaks.2
No perforations would be required--a conventional squeeze job would require perforations in the casing. Perforations add the risk of creating a larger leak and requiring multiple squeeze attempts with only a marginal chance for success.

If necessary, perforations could be added later if ultrafine cement was not effective. The cost of squeezing with ultrafine cement was much less than the costs estimated for the other options.

ULTRAFINE CEMENT

Ultrafine cement is defined as a very finely ground cement with an average particle size of 6 mu and a maximum particle size of 15 mu. In comparison, standard API cement is 5-7 times larger than ultrafine cement (Figs. 3-4). The ultrafine cement's small particle size makes it well-suited for squeeze jobs, especially casing and collar leaks in which cement must penetrate very narrow or tight areas. 2

The composition of the ultrafine cement used on this job was 20-30% finely ground cement and 70-80% hydraulic slag. The hydraulic slag material is cementitious and has a high silica content, which aids high temperature stability at temperatures up to about 275 F. No silica flour was added to the cement to prevent strength retrogression because of the chemical nature of the hydraulic slag.

Laboratory tests show ultrafine cement to have repeatable slurry pump times, good Theological properties, and adequate compressive strength. 23

The benefits of ultrafine cement's small particle size have been proven in other jobs which require sealing a casing collar leak or other very narrow areas. In these cases, traditional API cements form a bridge on the affected area, but ultrafine cement penetrates to provide a much more complete seal.

Other applications of ultrafine cement include penetrating gravel packs to isolate fluids or zones, sealing highly permeable zones, and squeezing microchannels.

PROCEDURE

The seal procedure incorporated proper plug setting techniques and was performed as follows:

Open-ended drill pipe was tripped in the hole to 13,218 ft. The drillstring consisted of 12,888 ft of 5-in. drill pipe and 330 ft of 2-3/8 in. tubing.
The pipe rams were closed, and the casing was pressured to 2,000 psi to determine empirically the fluid compression.
A 15-bbl, 10.2-ppg, high-viscosity gelled pill was spotted from 13,218 to 13,028 ft.
Two stands were pulled to 13,028 ft. The well was circulated while ultrafine cement was batch-mixed in a 50-bbl blender.
A balanced plug was spotted from 13,028 to 12,775 ft. The plug consisted of 20 bbl of freshwater ahead plus 0.1 gal/bbl surfactant, 105 sacks (20 bbl) of ultrafine cement, and 5 bbl freshwater behind (Fig. 5). Table 1 lists the slurry properties for this job.
Seven stands were pulled, the pipe rams were closed, and the well was pressured up according to the schedule in Table 2. The pressure slowly increased from 3,000 psi to 3,050 psi over 15 hr as a result of thermal expansion.
Pressure was bled from 3,050 psi to 3,000 psi and held for an additional 7 hr.
The pressure was then bled from 3,000 psi to 0 psi. The volume of cement squeezed into the casing collar was estimated as less than 1 bbl.
The drillstring was tripped out of the hole, and a drilling assembly was picked up. The total time waiting on cement was 36 hr.
After the plug was drilled out, the entire 10-3/4 in. x 9-5/8 in. casing string was tested with 3,000 psi surface pressure on top of 10.1 ppg mud (bottom hole pressure of 9,800 psi at 12,980 ft) for 30 min. No leak was detected.
Prior to penetration of the Smackover formation and 21 days following the first successful integrity test, the 10-3/4 in. x 9-5/8 in. casing string was tested with 3,000 psi surface pressure on top of 10.1 ppg mud (bottom hole pressure of 9,800 psi at 12,980 ft) for 30 min. No leak was detected.
Another integrity test was performed after the 7 in. x 6-5/8 in. production liner was run, but prior to the running of the 7-5/8 in. x 7 in. production tie back string. The 10-3/4 in. x 9-5/8 in. casing was tested with 1,300 psi surface pressure on top of 12.6 ppg mud (bottom hole pressure of 9,800 psi at 12,980 ft) for 30 min. No leak was detected.

This squeeze operation was successful and significant: The squeeze job with ultrafine cement preserved casing integrity, the casing collar leak was successfully squeezed on the first attempt without perforating the well, the volume of cement squeezed into the casing collar was less than 1 bbl, and the squeeze held while the well was drilled to 21,800 ft TD.

ACKNOWLEDGMENT

The authors thank Shell Offshore Inc. and Halliburton Services for permission to publish this article and K.M. Cowan of Shell Development Co. and A.F. Strange and L.F. Funkhouser of Halliburton Services for their contributions.

REFERENCES

Stancliffe, R.J., "Reservoir Characteristics of Fairway Field, Offshore Alabama," paper presented at the 1989 Shell Oil Co. Geological/Petrophysical Conference, Houston, Oct. 3, 1989.
Meek, J.W., and Harris, K.L., "Repairing Casing Leaks Using Small-Particle-Size Cement," IADC/SPE paper 21972, presented at the 1991 IADC/SPE Drilling Conference, Amsterdam, Mar. 11-14, 1991.
Ewert, D.P., Almond, S.W., and Bierhaus, W.M., "Small Particle Size Cement," SPE paper 20038 presented at the 1990 California Regional Meeting, Ventura, Calif., Apr. 4-6, 1990.