This application claims priority to provisional application Ser. No. 61/349,092 filed on May 27, 2010 and provisional application Ser. No. 61/412,671 filed on Nov. 11, 2010. These applications are hereby incorporated by reference in their entirety.
1. Field of the Invention
This invention relates to cementing of casings in wells. More particularly, method and apparatus are provided for preventing entry of fluids from the surrounding rock into the cement before it cures and for attaining higher radial stress in the cured cement.
2. Background of the Invention
The phenomenon of annular fluid flow (called “annular gas flow” when gas comes to the surface) has long been known to occur during cementing of wells. It is caused by fluids from the surrounding rock entering the wellbore before the cement cures. The resulting loss of control of a well has been responsible for loss of life and property for many years. In addition to the well control issue, annular fluid flow of fluids between zones before the cement cures can cause lack of zonal isolation in wells; water flow to surface from shallow pressurized water sands may occur; and casing shoes may not test at expected pressure integrity. All such occurrences can be manifestations of shortcomings in the primary cementing process.
In 1983, Cooke et al described the results of measurements of pressure and temperature in a curing cement column in seven oil and gas wells (“Field Measurements of Annular Pressure and Temperature During Primary Cementing,” J. Pet. Tech., August 1983). In all the wells, pressure in the cement column began to fall as soon as pumping of the cement ended. The paper explains that the pressure falls because cement shrinks in volume during the curing process because of: (1) the hydration reaction and (2) fluid loss from the cement, and at the same time cement develops a gel strength that prevents the cement column moving downward to compensate for the loss in volume. The decrease in volume combined with the gel formation result in a reduction in pressure in the cement column. If this pressure in cement is reduced to a value below the pressure of a fluid in a permeable rock penetrated by the well before the cement has cured sufficiently, the fluid from the rock enters the cement. This is the phenomenon of “annular fluid flow.” Measurements showed that the pressure in the cement column becomes the same as the pore pressure where fluid has entered. Other laboratory observations showed that fluid entering a cement column may rapidly channel up through the cement. This 1983 paper is hereby incorporated by reference herein for all purposes. Some of the field results reported in the paper were analyzed by Zhou et al (IADC/SPE 59137) using a mathematical model.
U.S. Pat. No. 4,407,365 discloses a method for preventing annular fluid flow—by periodically vibrating the casing while the cement is curing, to maintain pressure in the cement above fluid pressure in the pores of surrounding rock. The patent discloses several methods for vibrating the casing. One method is to ignite small explosives at different depths in the casing. The charges may be run on wire line and set off to cause a plurality of pressure pulses at different depths. The limitation of this method is that the amplitude of any vibrations caused outside the casing would be very small and of very limited extent along the axis of the casing. Another method disclosed is to lock a hydraulic jar attached to a drill string into a retaining groove in the casing and to repeatedly activate and re-set the jar during cement curing time. The limitation to this method is that it would be necessary to run a pipe in the casing after cement is pumped, which would be expensive and time-consuming, and it would be difficult to apply a jarring force in more than one location along the casing. Other methods disclosed include using explosive to propel a projectile against the casing wall, using vibrators on electric wire line, driving vibrators by fluid flow down a pipe string inside casing and electrical or hydraulic hammers. There are at least two disadvantages to the use of vibration sources on a wireline or a pipe string: (1) the wireline or string cannot enter a casing until after cement is pumped, and then delivering the vibration sources to a plurality of preferred depths in the casing would be time-consuming and expensive; (2) the power available for a vibrator would be severely limited by the power transmission capabilities of a wireline. Similar limitations exist for use of explosive charges to propel a projectile against the casing wall. This patent is hereby incorporated by reference herein for all purposes.
Two technical articles that help to elucidate the requirements for a process to maintain pressure in a cement column by vibrating the casing are: (1) “Primary Cementing Improvement by Casing Vibration During Cement Curing Time,” SPE Production Engineering, August 1988, and (2) “The Rheological Properties of Cement Slurries: Effects of Vibration, Hydration Conditions, and Additives” SPE Production Engineering, November 1988. The first article reports that axially vibrating a casing in a 200-ft well with a large electromagnetic vibrator attached to the top of the casing maintained pressure in the cement as it cured and also increased radial stress in the cement, resulting in a very good cement bond log. The increase in radial stress in the cement will increase the resistance to flow between the cement and the wellbore. During the vibration process the surface of the cement in the annulus dropped during each vibration period. The second article reported that breaking the gel structure of cement in a rheometer required only a small amplitude vibration, which was not sensitive to frequency, but that the structure began forming again in a very short time period after it was broken—in the range of 1 minute. Chemical additives in the cement affected gel strength during curing. These two articles are hereby incorporated by reference herein for all purposes.
FIGS. 1A and 1B illustrate why it is critically important in cementing some wells to minimize loss of pressure in the cement after it is pumped and before it cures. FIG. 1A illustrates well 10, which penetrates zones Z1 and Z2. Wellbore 11 has been formed, casing 12 has been placed in the wellbore and cement 13 has been pumped into the annulus outside the casing. The two characteristics of the strata penetrated by the well that are important for cementing are fracture gradient (the pressure gradient that will create a fracture in the earth) and pore pressure. Pressure that can exist in the cement slurry as it is pumped is limited by the fracture gradient in the earth, represented by line 14, on the right. The fracture gradient is represented as slightly less than normal in Zone 1, so this zone will limit pressure in the cement slurry. Cement slurry density and viscosity are selected such that the Equivalent Circulating Density (ECD—line 16) of the cement is less than fracture gradient throughout the cement column and static head is higher than pore pressure in any zone. Pore pressure is represented by line 18, on the left. Pore pressure is slightly higher than normal in Zone 2. In some wellbore conditions, the difference in pressure between highest allowable cement pressure and the highest pore pressure in a zone is small. Therefore, the allowable pressure drop in the cement column before cement pressure drops to pore pressure in a permeable zone may be, for example, only 200-300 psi. Consideration of the fact that cement pressure drops rapidly after pumping in some wells (August 1983 J. Pet. Tech. paper, referenced above) leads to the conclusion that a method to limit pressure drop in the cement after pumping that will keep pore fluids from entering the cement column should be available for application soon after cement-pumping ends. As the cement cures, gel strength increases, which means that breaking gel strength in the cement column, such that the cement will flow, will become more difficult as time-after-pumping increases. Maintaining pressure in the cement column will not only prevent fluid entry into the cement while it is curing, it will also cause flow of cement in a radial direction outward, leading to higher radial stress when the cement has cured.
Later references disclose other methods for vibrating casing during cement curing time. U.S. Pat. No. 5,361,837 discloses a method for preventing annular fluid flow using tube waves in the casing. The tube waves are induced in casing by pressure variations at the surface caused by opening and closing of valves to pump in and out a liquid. The patent discloses that studies showed that casing vibration having a longitudinal displacement of at least 0.25 inches along the wellbore axis is normally more than sufficient to break the gel strength of cement slurry around the region of vibration and that the tube waves can cause longitudinal displacement of about 1.0 to 1.5 inches at the bottom of a casing string. The disclosure posits that extensional waves near the bottom of the casing, in the region of the hydrocarbon zone, are sufficient to prevent annular fluid flow. No evidence is presented, however, that vibration only near the bottom of a casing string will allow the pressure in cement to increase near the bottom of the casing.
U.S. Pat. No. 5,152,342 discloses apparatus and method for vibrating a casing string during cementing, with the vibrating device located near the bottom of the casing string. Cement slurry being pumped down a casing flows through a device, powering the device and causing vibrations in the casing.
U.S. Pat. No. 6,725,923 discloses apparatus that includes hammers that oscillate in a radial direction and hit the wall of tubes when the flexible suspension to which the hammers are attached is pulled. It is stated that the resulting vibrations in the casing can improve cementing.
U.S. Pat. No. 5,377,753 discloses a method for breaking the gel strength of cement in an annulus by applying pressure pulses in a fluid above the annulus.
U.S. Patent Application Publication 2009/0159282 discloses inducing pressure pulses in the cement in the annulus before the cement has cured “for bonding a wellbore to a casing.”
All prior art methods disclosed for inducing vibration into a casing to prevent pressure drop in the cement column have been limited by applying vibration only at the top end or the bottom end of the casing or, if vibration is induced at intermediate points along the casing, by placing apparatus in the casing after pumping of cement has ended (the top plug has been “bumped”). No method or apparatus is known for inducing vibrations into a casing string by sources mechanically coupled to the casing at locations spaced apart along the casing and inducing these vibrations “near simultaneously” (defined herein as within a time period before gel strength of the cement re-builds to its original value after it is broken by vibration), beginning soon after cement pumping ends. What is needed is method and apparatus for inducing an impulse or vibrations at a selected location or at selected locations along a casing string beginning soon after pumping of cement ends and continuing for a selected time during the cement curing period. (“Soon” depends on the time required for the cement to build gel strength to a selected value. For most cements, this time is preferably less than 30 minutes.)
Apparatus and method are provided for pumping down and mechanically coupling to the casing a source or sources of impulses or vibrations that are activated by pressure changes in the casing and then retrieved or drilled or milled from the casing or moved to a segment of the casing that is not to be used in further well operations. Power for the source of the impulses may be supplied by fluid pressure changes in the casing resulting from alternately pumping in and releasing fluid from the casing. Sources for the impulse or vibration may be pumped to the locations along the casing string by launching them into the displacing fluid while cement is being displaced from the casing or dropping them after the plug has been bumped. The devices for applying impulses to the casing may be locked in place (mechanically coupled) in sections of the casing adapted for receiving the devices or may be locked by a locking mechanism in the device. In one embodiment, the source of an impulse may be a mechanical or hydraulic jar, such as that known in the industry. The jars may be activated by an increase in hydrostatic pressure in the casing. Potential energy stored in the jar may be derived from a pressure increase in the casing. Other sources of energy, such as chemical reactions may be used to induce the impulses or vibrations in the casing. Alternately, the devices may be vibrators driven by flow of fluid under pressure that is created by increase and decrease in pressure in the casing or from other sources. The devices in the casing are operated as the cement cures to maintain pressure in the cement above pore pressure in zones in contact with the cement for a selected time and to increase the radial stress in cured cement. After cement curing, the devices may be recovered to the surface, where they may be re-used, or they may be expendable devices that are removed by drilling or milling from the casing (casing above production casing) or they may be moved to a segment of the casing where they are not interfering with further operations in the well, such as a rathole (production casing).
Referring to
The damping of amplitude of an impact or vibration from a source may be predicted using finite element analysis and rheology data providing wellbore viscoelastic properties of the cement at wellbore conditions as a function of time after pumping and time after breaking the gel. (Such properties for a cement at room temperature are provided in the SPE Production Engineering article of November 1988, referenced above.) Shrinkage data for the wellbore cement vs time may be obtained as discussed in the March 2009 article referenced above. Impact data for a jar or other source may be available from the manufacturer or may be measured for the conditions of use. Strain gage measurements of the amplitude of casing displacement and measurements of pressure in the cement may be used to calibrate predictions of amplitude and determine how the amplitude of casing displacement affects pressure in the cement. One model of the cementing process (without vibrations or impulses) after pumping, such as may be used to predict pressure after pumping has ended was published by Zhou et al (“New Model of Pressure Reduction to Annulus During Primary Cementing,” IADC/SPE 59137, February 2000, referenced above). Segments of the cement column where gel strength is not broken by impacts or vibration may be caused to move because gel strength is broken in other segments of the cement column, resulting in higher pressure gradient along the cement column where gel strength is not broken. Pressure gradient along the cement column may also be increased by application of a pressure at the surface of the annulus during cement curing, a practice that has long been known in industry, in combination with the methods taught herein. Such surface pressure is limited by fracture gradients in the earth, which usually make this approach ineffective when used alone. A pressure gauge in the fluid above the cement in an annulus may be monitored to detect movement of the top of the cement column when the casing is vibrated. Preferably the gage and connections are liquid-filled. The compressibility of the gauge system (pressure change per volume change) may be used to indicate the volume of shrinkage of cement compensated for by vibration of the casing.
Impulse source 54 may employ a mechanical or hydraulic jar, which is well known in industry. The jar is energized by pressure increases and decreases in fluid pressure 82 outside the source, as illustrated in
Vibration source 66 may be an oscillating, vibrating or rotating vibrator driven by flow of fluid. Such vibrators are well known in industry. For example, a tool described in SPE 90737, “Downhole Impulses vs Downhole Impacts Improve Recovery of Stuck Retrievable Packers,” 2004, may be used. This paper is incorporated by reference herein for all purposes. Another source of vibration (vibrator) may be a water hammer, such as used in impact drilling. Such hammers are available, for example, from Wassara AB of Stockholm Sweden. A Model W 80 Wassara hammer will produce vibrations at a frequency of 65 Hz and 210 Joule/blow with a flow rate of about 32 gal/min through the hammer, according to the manufacturer. Thus, five seconds of vibration may be produced by flow of about 2.7 gallons of water through the hammer and into an accumulator. Such hammers may be designed to produce different frequencies of vibration at selected flow rates through the hammer. Optimum frequency ranges may be selected by comparing results of vibration at various frequencies. (The impact surface of jars or hammers may have lower modulus materials to cause more low-frequency output of energy.) Pressure port 65 transmits fluid from the casing through vibrator 66 to compress gas in chamber 68 by moving piston 67 (a hydraulic piston accumulator). Alternatively, a spring accumulator may be used to receive water driven through a hammer. Accumulators are readily available from many sources in industry. A detent mechanism in the accumulator may be used to prevent flow into the accumulator until a selected over-pressure has been applied. Release of pressure in the casing may cause flow through vibrator 67 in the reverse direction if the vibrator allows two-way flow. If it does not, fluid pressure in chamber 68 may be relieved through a by-pass channel around vibrator 66 having a one-way check valve (not shown), allowing piston 67 to return to stop 67A. Fishing neck 61 operates a mechanism to retract locking dogs 63 from a receiving groove and cups 62A after use of the vibration source has been completed, using methods well known in the wireline retrievable tool industry. Materials used to form apparatus 60 may be selected to allow apparatus or parts of apparatus 60 to be drilled from the casing or released such that it can be moved to a location that will not interfere with further operations in the well.
Alternatively, the plugs and impact or vibration sources may be launched from a radial launcher, such that the total length of the launcher is less than that of
Alternatively, it may be preferable to place the impact or vibration sources in the casing during or after cement pumping without rubber cups such as 52A, 62A or 72A. The apparatus may fall by force of gravity until it is locked to the casing. Fluid flow area around the sources and weight of the sources may be selected to attain a suitable fall velocity with or without flow of fluid downward in the casing.
With sources of impact in place, pressure inside casing 34 is increased to a value selected to activate all the impact sources, which are set to activate at about the same surface pressure in the casing. Preferably, the rate of increase of pressure inside the casing is rapid, to assist in applying the impacts near simultaneously. All impact sources preferably activate within a time span of about 5 minutes, more preferably within a time span of 3 minutes and most preferably within a time span of 1 minute. All time spans less than the time for the cement to re-form a gel structure to its original value after the structure is broken are defined herein as “near simultaneous.” The time spans for the cement to re-form a gel structure after the structure is broken may be measured for the cement of interest and at well conditions using the methods described in the November 1988 paper referenced above and reported in
Other methods of triggering impacts and multiple locations in the casing may be used. For example a coded series of pressure pulses may be used to activate a chemical reaction, which creates a pressure in a device and releases an impact source. Devices such as described above may be energized by changes in casing pressure and activated by sound pulses sent down the casing of fluid in the casing. Preferably the forces of impact will be designed to activate near simultaneously at all locations along the casing. Preferably the first impacts in the casing will be applied shortly after cement pumping ends, i.e., shortly after bumping the plug.
After the impulse sources have been activated for the desired time, a wireline can be lowered into the well and latched on to the fishing neck of each device and the devices can successively be withdrawn from the well for re-use. Electric wireline, slick line, swab line or coiled tubing may be used. A lubricator may be used for the line. As explained above, expendable apparatus may also be used.
The use of vibration or impulses to break the gel strength in cement has been primarily discussed herein, but it should be understood that the methods and apparatus disclosed may be used with other methods for increasing pressure in curing cement. It has long been recognized that applying pressure at the surface of a cement column in an annulus can sometimes be helpful in preventing annular fluid flow. The limitation of this method is that fracture gradient in the wellbore often prevents application of enough pressure to be effective in moving a cement column. However, at times application of pressure at the surface can increase pressure in cement, as observed, for example, in the August 1983 paper referenced above (see
When all the factors that control pressure in a cement column are considered, the pressure in the cement column as a function of time can be predicted based on gel strength of cement in the range of each source of impact or vibration, gel strength outside the range of a source, pressure applied at the surface of the cement column and its density, fluid loss rate from the cement column and shrinkage in volume of the cement slurry as a function of time after pumping. Whether fluid enters the wellbore will depend on whether pressure in the cement column is maintained above the pore pressure in any permeable zone intersecting the cement column long enough for the cement to build sufficient strength to exclude fluid at the pore pressure.
Although the present invention has been described with respect to specific details, it is not intended that such details should be regarded as limitations on the scope of the invention, except to the extent that they are included in the accompanying claims.
Number | Name | Date | Kind |
---|---|---|---|
3335801 | Wilsey | Aug 1967 | A |
4407365 | Cooke, Jr. | Oct 1983 | A |
5152342 | Rankin et al. | Oct 1992 | A |
5361837 | Winbow | Nov 1994 | A |
5377753 | Haberman et al. | Jan 1995 | A |
5439290 | Wicks et al. | Aug 1995 | A |
6725923 | Ivannikov et al. | Apr 2004 | B1 |
7311143 | Engels et al. | Dec 2007 | B2 |
20090159282 | Webb et al. | Jun 2009 | A1 |
Number | Date | Country |
---|---|---|
2004073929 | Sep 2004 | WO |
Entry |
---|
“Field Measurements of Annular Pressure and Temperature During Primary Cementing,” J. Pet. Tech., Aug. 1983. |
“Primary Cementing Improvement by Casing Vibration During Cement Curing Time,” SPE Production Engineering, Aug. 1988. |
“The Rheological Properties of Cement Slurries: Effects of Vibration, Hydration Conditions, and Additives” SPE Production Engineering, Nov. 1988. |
“Cement Shrinkage Measurement in Oilwell Cementing—A Comparative Study of Laboratory Methods and Procedures,” SPE Drilling and Completion, Mar. 2009. |
“New Model of Pressure Reduction to Annulus During Primary Cementing,” IADC/SPE 59137, Feb. 2000. |
“Downhole Impulses vs. Downhole Impacts Improve Recovery of Stuck Retrievable Packers,” 2004. |
United States Patent and Trademark Office, Office Action Summary for U.S. Appl. No. 13/293,770, mailing date of Jul. 18, 2013; p. 1-41. |
Number | Date | Country | |
---|---|---|---|
20110290485 A1 | Dec 2011 | US |
Number | Date | Country | |
---|---|---|---|
61349092 | May 2010 | US | |
61412671 | Nov 2010 | US |