SOLIDS-FREE GEL, EXPANDING CEMENT REPAIR POOR PRIMARY CEMENT JOBS

Ryan Cox, Keng Seng Chan
Dowell Schlumberger Canada Inc.
Calgary

Ineffectively cemented shallow gas sands can slowly build pressure in the casing annulus, violating environmental regulations.

By use of a unique, solids-free gel to shut off well gas zones, a shallow gas sand was squeezed without fracturing the formation or forming gas channels in the cement.

In the southern and south central plains of Alberta, shallow gas sands occur between 500 and 825 ft.

These sands are interbedded units of shale, sandy shale, sandstone, and minor lignite consisting of three sections.

The upper contact is selected at the top of a thin sandstone bed below a shaly interval.1 It is not uncommon for low-volume, relatively high-pressure gas to be present.

The relatively high gas pressure, low temperature, and shallow depth combine to increase greatly the risk of shallow gas migration after primary cementing. Gas flows from the surface-casing vent can occur when primary cementing jobs are designed and pumped without consideration for the behaviors of these sands.

A number of wells in the Wainwright area of Alberta have inadequate cement tops to cover the shallow gas zones. Wells drilled in the 1950s were not required to be cemented to surface, leaving the shallow gas sands uncemented.

Partial or complete lost circulation and cement fallback after placement during primary cementing can occur in the Wainwright area, leaving upper shallow gas zones without effective zone isolation.

In cases where gas migration is severe because of inadequate cement tops, remedial treatments are required to prevent gas migration to surface and to impede communication between zones. Zonal communication could cause a severe ecological problem if the subsurface gas flows to potable-water formations uphole.

REMEDIAL TREATMENT

The remedial-treatment solution is to obtain an effective hydraulic seal at the gas source. If an effective seal is obtained above the gas source, the gas-zone pore pressure will transmit through the channels or uncemented portion of the annulus and act on the formation around the hydraulic seal.

This will fracture the formation around the seal because of low fracture pressure at the shallow depth. On the surface this can be seen as gas migration outside the outermost casing.

It is, therefore, critical that an effective hydraulic seal be obtained at the gas entry. Many cementing jobs were completed in the past following this concept. The wells were perforated at the known gas source, and a conventional circulation cement squeeze was performed.

However, many of these jobs were not entirely successful. Gas flow from the surface-casing vent valve or outside the casing still occurred sometime after the treatment.

One possible explanation is that gas flowed into the slurry during primary cementing forming gas-cut cement and channels. This failed to form a competent hydraulic seal at the gas source.

Removal of fluids and filter cakes in the annulus is critical because these materials can become pathways for gas migration. The fluids consist mainly of gelled bentonite mud left static for long periods. Evidence of mud dehydration and cement stringers can be seen on bond logs.

Even after extended periods of circulation with a muddispersing chemical wash pumped at turbulent flow rates in near-zero casing standoff conditions, mud-removal efficiency ranged 80-90% in the four wells studied with fluid calipers.' If complete hole cleaning is not achieved, it is imperative to eliminate the gas flow at the source before cementing.

INSUFFICIENT SQUEEZE PRESSURE

Cement-squeeze techniques are typically used to seal gas channels and block gas entry at the formation face. When cement slurry is forced against a permeable formation, the solid particles filter out on the formation face as the cement filtrate enters the formation matrix.

A properly designed squeeze cement job results in cement filter cake filling the openings between the formation and casing. The rate of filter-cake development is a function of differential pressure (net squeeze pressure), slurry fluid loss, time, and formation permeability.3

However, differential pressure is limited by formation fracture pressure and gas pore pressure. The maximum differential pressure without fracturing the formation decreases with the decrease in depth.

For example, given a 0.84 psi/ft fracture gradient and a 0.53 psi/ft gas pore-pressure gradient, if the gas entry is at the depth of 650 ft, a maximum differential pressure of only 200 psi is applied across the cement filter cake at this depth. If the gas entry depth is located at 3,300 ft, the maximum available differential pressure is 1,000 psi.

For shallow depths, insufficient differential pressure can affect the success of a cement squeeze even when using a higher fluid-loss controlled slurry .4 Increasing squeeze pressure increases the risk of fracturing the gas formation.

It can be concluded that the remedial cement-squeeze treatment efficiency in shallow depths is poor for the following reasons:

• Gas channels are difficult to locate, especially when channels are actually fine capillaries.

• Gas channels may be too small to be penetrated by cement.

• Old mud and mud filtercake removal is incomplete.

• Maximum allowable squeeze pressure at a shallow depth is insufficient to squeeze off the gas entry.

• The squeeze pressure exerted during the job is sometimes large enough to initiate formation fracturing, worsening downhole communication problems.

SETTING FLUID

To overcome these problems, a field technique has been developed to shut off gasflow at the source which doesn't require high squeeze pressures. The procedure involves adequate mud removal before cementing.

The treatment of the gas inflow extends beyond the well bore by use of a solids-free fluid capable of penetrating into the formation matrix microannulus, and capillaries. Special considerations must be given to selection of the subsequent cement system.

The slurry must have the ability to impede gas entry and not allow gas into the slurry. For this reason, a low-temperature, right-angle-set cement was used.

A solids-free Newtonian fluid having an initial viscosity of less than 2 cp was formulated by use of an internal activator. After a given period, depending on the fluid design and the temperature, the viscosity of the solution increases rapidly to form a gel.

For the Wainwright area application, the fluid was designed to set after 80 min at 68 F. with rapid viscosity increase from 2 cp at 80 min to more than 100 cp at 88 min (Fig. 1).

Core-flow experiments for this fluid were conducted with various cores of different permeabilities saturated by brine or diesel fluid. Results of the water-displacement, gel-extrusion resistance are shown in Table 1.

In all cases, the gel was able to withstand a differential pressure greater than 1,200 psi/ft. Long-term shrinkage of the gel was also found to be minimal. Tests conducted at 212 F. with brines and diesel oil found the weight loss of the set gel to be only 0.7% in 2 months.

Although the set gel does not have significant compressive strength, it can gain some compressive strength when the gel is supported by solid particles or sands. The compressive strength increases with the decrease in solid particle size (Fig. 2).

The ability of the gel to plug a small crack or microannulus has been experimentally confirmed by laboratory tests.' The gel is a cohesive system which does not adhere to the face of the channel.

In the tortuous path of a microchannel, extrusion pressure would be very high because the cohesive gel forms colloidal particles to plug the channel pore throats.

RIGHT-ANGLE-SET CEMENT

As cement hydrates, the hydrostatic pressure can decrease to that of water.6 The transition state is defined as the intermediate period during which the cement behaves neither as a liquid nor a solid, and the slurry loses its ability to transmit hydrostatic pressure.7

During this time, a hydrostatic underbalance exists, allowing gas inflow if the adjacent formation pore pressure is greater than the water hydrostatic pressure. The result may be a gas-cut or gaschannelled cement column which will not provide zone isolation.

The transition time can be quantified by a period starting when the gel strength is around 21 lb/100 sq ft and ending when the gas can no longer percolate within the gelled cement. It has been shown that a gel strength range from 250 to 500 lb/100 sq ft is sufficient to prohibit gas percolations

Because gas percolates through the cement slurry only during the transition time, the transition time should be minimized.

Gas percolation can be considered as a particular type of gas migration. Gas in the form of macroscopic bubbles invades the slurry and rises due to buoyancy effects in accordance with Stokes' Law.9 A cement slurry behaves as a non-Newtonian fluid; therefore, this process involves breaking the gel strength of the slurry.

However, gas may also flow at the microscopic level within the pores of the gelled cement structure or directly along the cement/pipe and cement/formation interfaces. Thus, a short transition time to compressive strength is preferred over the transition to a high gel strength.

Depending upon the kind of cement, shrinkage results in the formation of a microannulus between the casing and cement or between the cement and the formation.10 Recently it has been found that the bulk shrinkage (external volumetric reduction) occurs after initial set is generally only a few tenths of 1%.11

The microannulus is generally too small for particle migration, but gas flow is possible. Therefore, an expansive cement should be considered for long-term gas migration control.

A unique cement system satisfies the expanding and right-angle-set criteria. The cement exhibits a 5-min transition period from 40 Bc to 100 Bc (Bearden Consistency Index) in consistometer measurements (Fig. 3). This behavior is termed "right-angle set."

The system develops compressive strength rapidly after reaching 100 Bc. Within 10 min, 290 psi compressive strength is attained. The start of the transition period can be adjusted from 30 min to 3 hr depending upon cementing conditions in the well.

Table 2 shows three different slurry formulations which yield thickening times from 1 to 3 hr. These slurries show good rheology control, very short transition time (less than 5 min), and rapid compressive-strength development.

Fig. 4 shows the unrestrained expansion behavior of right-angle-set cement at 68 F. Expansion increases almost linearly with time for the first 80 days. This early period is very critical because the cement-hydration process slows down after about 80 days. Accordingly, the rate of expansion slows and eventually stops.

FIELD TEST

During primary cementing of a well in the Wainwright area, the well had severe lost circulation. The cement job was designed to have the cement top at the surface, yet there was no cement return after the displacement,

A cement evaluation log showed the top of cement at a depth of 1,296 ft. In addition, the temperature and noise logs showed that gas was flowing up the uncemented open hole into the annulus between the surface casing and production casing.

The location of the gas source was estimated to be between 646 and 692 ft.

Although no volumetric gas-flow measurement was obtained, the 12-hr annulus pressure buildup was 160 psi.

This problem is regarded as serious vent flow according to the Alberta Energy Resources Conservation Board's drafted guidelines.

For this well, 7-in. surface casing was set at 590 ft with 41,12-in. production casing set at 2,632 ft. Open-hole caliper logs indicated an average open-hole size of 6 1,12-8 1/3 in. to the bottom of the surface casing.

The gas source was only 56 ft below the bottom of the surface casing. The correct treatment should try to shut off the gas entry at the source below the surface casing.

If the effective zone isolation is implemented above the bottom of the surface casing, the gas may still flow up to the bottom of the surface casing and then migrate outside of the surface casing.

The major design consideration was the difference between pore and fracture pressure. With a bottom hole gas-zone pore pressure of 420 psi and a bottom hole fracture pressure of 610 psi, it was difficult to control gas inflow into the slurry during placement while simultaneously preventing hydrostatic overbalance from fracturing the formation,

For the remedial treatment, the casing was perforated at 692 ft, and fluid circulation to surface was established. A cement retainer was then set at 567 ft. With an estimated porosity and gas source height, 6 bbl of solid-free solution were required to achieve a radial penetration of 11/2 ft. The fluid was injected into the permeable gas source at a pressure less than the bottom hole fracture pressure.

Before the injection of the solid-free fluid, gas was seen in the returns during the pretreatment hole cleaning and fluid circulation. After the placing of the solids-free fluid, no gas was found to return to surface in subsequent circulation before the cementing. Right-angle-set cement was then pumped into the annulus.

SQUEEZE PROCEDURE

The following is the detailed job procedure with a few observations:

1. Held 150 psi back pressure to prevent gas inflow. Circulated water down the tubing and up the casing/open hole annulus at 6 bbl/min for 15 min. Gas bubbled to surface during circulation.

2. Circulated 30 bbl of mud-dispersing chemical wash solution at 6 bbl/min. Gas bubbling to surface was still observed.

3. Ran a fluid caliper. Pumped 3 bbl of water tagged with 41/2 lb of ironoxide colorant and displaced with water at 6 bbl/min. The fluid calliper volume yielded a 161/2-bbl path of circulation.

   However, the calculated total tubing and annulus volume was 191/2 bbl. This implied that less than 3 bbl of mud or other unknown material was left in the annulus.

4. Pumped 6 bbl of water containing 1/4 gal/bbl of surfactant into the gas source at gpm.

5. Pumped 6 bbl of solids-free fluid into the gas source at 11 gpm with an average surface pressure of 225 psi.

6. Maintained pressure above pore pressure for 1 hr until solids-free fluid was set and gained sufficient

   strength.

7. Circulated 12 bbl of water to clean any residual solids-free gel from the annulus to prepare for cementing. Gas was no longer bubbling to surface.

8. Mixed and pumped 21 bbl of right-angle-set cement with a 1-hr thickening time.

9. Observed 3 bbl of cement returns. Pulled out of the retainer and backwashed residual cement from the tubing.

10. Applied 100 psi through the surface-casing vent valve. Waited 20 min and repressured to 125 psi. Waited 10 min and repressured to 150 psi. There was no bleed off at the final stage.

To date, 10 months after the treatment, no gas vent flow or gas pressure buildup inside the casing annulus has been detected. No gas migration outside the surface casing has been observed.

CONCLUSIONS

• Shallow gas formations in the Wainwright area exhibit high gas pore pressure and low fracture pressure.

• the effectiveness of a conventional cement squeeze treatment is limited by low effective squeeze pressure at the shallow depths. Increasing squeeze pressure increases the potential of fracturing the gas formation.

• for effective remedial treatments of gas migration control at shallow depths, the job design should eliminate gas entry into the well bore before cementing.

• A solids-free setting fluid with an expansive right-angle-set cement is a good squeeze solution for wells with the cement top below the bottom of the surface casing.

ACKNOWLEDGMENT

The authors wish to thank Dowell Schlumberger Canada Inc. for permission to publish this article and Jean de Rozieres for his assistance with the laboratory experiments and with the preparation of the article.

REFERENCES

1. Glass, D.J., "Lexicon of Canadian Stratigraphy," Canadian Society of Petroleum Geologists, Vol. 4, 1990.

2. Chan, K.S., et al., "Surface-Set Cements and Their Successful Applications for Shallow Gas Migration Control in Southeastern Alberta," CIM/SPE Paper 90-114, 1990.

3. Suman, G.O., Jr., and Ellis, R.E., "World Oil's Cementing Oil and Gas Wells Including Casing and Handling Procedures," Ann Arbor, Mich., 1977.

4. Hook, F.E., and Ernst, E.A., "The Effects of Low-Water-Loss Additives, Squeeze Pressure, and Formation Permeability on the Dehydration Rate of a Squeeze Cement Slurry," SPE Paper 2455, 1969.

5. De Rozieres, J., laboratory results.

6. Levine, D.C., Thomas, E.W., Bezner, H.P., and Talle, G.C., "Annular Gas Flow After Cementing: A Look at Practical Solutions," SPE Paper 8255, 1979.

7. Tinsley, J.M., and Sutton, D.C., "Study of Factors Causing Annular Gas Flow Following Primary Cementing," SPE Paper 8257, 1979.

8. Sykes, R.K, and Logan, J.L., "New Technology in Gas Migration Control," SPE Paper 16653, 1987.

9. Nelson, E.B., Well Cementing, Schlumberger Educational Services, Chapters 8 and 13, 1990.

10. Gotsis, C., Roy, D.M., Licastro, P.H., and Kaushal, S., "Thermal and Thermomechanical Analysis of a Cylinder Cementatious Plug Hydrating in a Borehole," American Concrete Institute Publication SP95-4, 1984.

11. Dreq, P., and Parcevaux P.A., "A Single Technique Solves Gas Migration Problems Across a Wide Range of Conditions," SPE Paper 17629, 1988.

Copyright 1991 Oil & Gas Journal. All Rights Reserved.