Part
5
How to Run and Cement Liners
L iner tops can be tested with or without
If a leak does exist in a liner and the decision is made to squeeze the shoe, there is a risk of setting the packer below the leak, which could cause one one of three things to happen. Either squeeze cement will be applied to the shoe where it is not needed; or cement could channel around the liner and through the leak back into the pipe putting cement on top of the packer; or the packer (if set below the leak) would effect a good leakoff test, thus hiding the leak. The latter possibility possibility could be avoided avoided by pressuring up on the annulus above the packer to the fracture gradient of the shoe. This will also help avoid collapsing collapsing the casing above the packer with high squeeze pressures.
packers, but there are times one method is preferred over the other .
Glenn R. Bowman , Regional Drilling Superintendent, Ashland Exploration, Houston, and Bill Sherer, Operations Manager, Liner Tools LC and formerly Alexander Oil Tools, Houston TESTING THE TOP TOP of a liner after it has been cemented is necessary to ensure a wall’s integrity. integrity. However However,, whether done with or without packers there are potential problems attendant with either method that can occur if the tests are not properly engineered. A discussion of these problems and ways to avoid them follows.
TESTING LINER TOPS AND SHOES WITHOUT PACKERS If packers will not be used to test a liner, the following considerations should be kept in mind. The positive pressure test of the liner top should be enough to exceed the fracture gradient at the liner shoe by at least one ppg. It should be even even higher with
heavy, solids laden, less heavy, less penetrating type test uids. This high of a test pressure may not seem necessary at rst glance, but many liner tops have been tested at pressures even less than the fracture gradient at the intermediate casing shoe, and the cement job on the overlap was was considered good. This can give a false sense of security. security. If there is communication between the liner top and the intermediate casing shoe due to a bad cement job and the pressure applied to the liner top is below that required to rupture the formation the intermediate casing shoe, false positive test will be achieved. The one ppg or more is added to ensure that the fracture gradient has been exceeded and to overcome possible inaccuracies of the pressure gauges. In addition, barite and other mud solids could plug up very small channels in the liner overlap and mask a leak. Barite and gel can cause small channels to hold positive differential pressures of up to 5,000 psi. The extra test pressure may enable the channel to be ruptu red enough to cause it
to take uid. Pressures required to “crack the rock” are generally
Since the test pressure inside a liner should at least exceed the fracture gradient at the shoe, it would be prudent to apply this same pressure to the liner overlap before it is drilled out and the shoe tested without a packer packer.. Otherwise Otherwise,, a leakoff test on the shoe (if it is a drilling liner) may breakdown the overlap and the operator will again not know for sure where the leak occurred. Another negative aspect of testing liners without packers is that test pressure must not exceed the unknown burst rating of the intermediate casing, especially if mud density is greater inside the intermediate than outside (this is the case when most liner are set). If substantial casing wear is suspected, testing without packer should be considered risky. Wear would be suggested in directional holes when long periods of time are spent drilling inside casing; when intermediate casing is through high dog-legs; and when internal mud weights and circulating temperatures are high while drilling below intermediate casing,1,21,23, etc. In addition, addition, the increase increase in tension due to ballooning during pressure testing should be checked to ensure that tensile rating of the intermediate casing is not exceeded. The following example illustrates the calculations that should be made before pressure testing without a packer.
The slurry stage
higher than those required to extend the fracture, especially if the
fracturing uid is solids laden.
The intermediate stage
Casing
For the same reasons, pressure tests inside a liner should exceed the fracture gradient at the liner shoe by one ppg or more. A leak near the bottom of a liner could go undetected if the breakdown pressure of the formation adjacent to the leak was not achieved.
Consequently, when “bumping the plug” after cementing a liner,
Casing
Formation
the pressure should be increased to the maximum that will be imposed on the liner liner while testing its top. Otherwise, if the plug fails and cement in the shoe joints was contaminated and did not
The set stage Casing
Formation
Formation
set up, it would be possible to pump through the oat equipment and ll the liner with cement if its top has to be squeezed. This is especially true for short liners where fracture gradients at the liner top and shoe may be close. Also, if the liner wiper plug is effective, does not bump up during primary cementing and cement in the shoe joints is green, part of the squeeze job could go down the inside of the liner until the plug bumps. To be sure that the breakdown pressure is not moving the liner wiper plug down, it is good practice to pump in 10 barrels after breaking down the liner top to bump the wiper plug is case it had not bumped during primary cementing.
Cement slurry in annulus
Gas fowing through setting cement
Gas or oil in set cement
Figure 22 - How honeycombed honeycombed cement occurs. In the slurry stage (left), the cement column exerts the necessary hydro-
static pressure on on the formation to prevent gas gas or uid ow. ow. In the intermediate or self-supporting stage (center) hydrostatic
These pressure considerations are especially critical for drilling liners whose shoe will be tested to leakoff. If a liner shoe does leak off at some value below expected fracture gradient, the operator will not know whether it is caused by a bad cement job around the shoe, an inherently weak formation at shoe or a leak in the liner string or top. Reprinted from World Oil magazine May 1988 with permission from the authors. ,
pressure is reduced and formation uids begin to move through the cement. Also in this stage, the cement cannot cannot transmit hydrostatic pressure pressure from above. above. In the fully set stage stage (right),
cement is permeated with formation uids or gas, and often cannot be squeezed. (World (World Oil, Oil, January, 1982.)
5.1
w w w . l i n e r t o o l s . c o m
WELL CONDITIONS
leakoff test will not exceed the burst safety factor of the intermediate casing.
A well in South Louisiana had 9 5/8-in., 47 ppf. N-80. LT&C casing set at 10,0000 ft.in 12 ppg mud. An 8 1/2 –in hole was drilled to 13,000 ft. Other pertinent data include the following: x
Pore pressure at 13,000 ft is equivalent to 15.5 ppg mud
x
Increase mud weight to 16.5 ppg before drilling out of 7-in liner
x
Test 7-in. liner shoe to leakoff with 16.5 ppg mud in hole
What pressure will be needed to test liner overlap to same pressure equivalent using 16ppg that will be imposed on liner overlapwhile testing the 7-in. liner shoe to leakoff with 16.5 ppg mud? Surface pressure needed to test overlap to equivalent pressure with 16 ppg = (16) (0.052) (10,000) = 8,320 psi. 8,580 psi (pressure at 10,000 ft with 16.5 ppg mud) – 8,320 psi = 260 psi
Operator policy stipulates that no more than 80% of burst rating will be imposed on intermediate casing. x
Surface pressure required to test overlap with 16 ppg mud to
a sufcient pressure to support a leakoff test on 7-in. liner shoe
Burst rating of 9 5/8-in., N-80 casing is 6,870 psi, tension rating is 905,000 lb x
x
with a 16.5 ppg mud) = 260 + 1,900 = 2,160 psi.
Cement has been drilled to the top of the liner
Will the SF T of 1.7 on 9 5/8-in. casing be exceeded while running the leakoff test without a packer?
The 7-in. liner top is ready to be tested with 16 ppg mud in the well x
x
Casing tension increase at surface due to ballooning while testing 7-in. liner to leakoff will be: 23
The 9 5/8-in. casing is hung off at surface with 400,000lb
Fa = 0.6 (P14 A1 – P04 A 0) Where:
Operator does not want to exceed SFT (Safety factor of tension) of 1.7 on 9 5/8 in. intermediate casing while running the leakoff test on the liner. x
P14 = Pressure inside tube = 1,900 psi P04 = Annulus pressure outside tube = 0 psi. A1 = area of tube ID = 59.19 sq in. A 0 = Area of tube OD = 72.76 sq in. Fa = Axial force, lb (Increase in tension) Fa = 0.6 [1,900 x 59.19) – 72.76 x 0 psi)] Fa = 67.433 lb
PROBLEM.
What positive pressure should be used to test the liner overlap with 16 ppg mud to be sure it will support a leakoff test on the 7-in. liner shoe with 16.5 ppg mud? Should a packer be employed for the leakoff testing? 8
SOLUTION .
The highest pressure imposed on the 9 5/8-in. casing will be the leakoff test pressure used on the 7-in. liner shoe. The anticipated surface pressure to conduct a leakoff test is calculated as follows:
If 9⅝-in. casing is hung off at surface with 400,000 lb, and the increase in tension at the surface due to ballooning is added, then total tension at surface while pressure testing 7-in. liner shoe to leakoff with no packer will be 467,433 lb.
Estimated fracture gradient at 13,000 ft (15.5 ppg pore pressure) = 18.3 ppg (After Eaton).
SF T = Tensile rating of 9⅝-in. casing¸ total tension at surface = 905,000 ÷ 467,433 = 1.94. Thus the SF T of 1.7 is not exceeded and it is possible to do all positive pressure testing of the 7-in. liner without a packer.
Fracture gradient + 1 ppg (safety factor to assure initiating rupture) = 19.3 ppg. Bottomhole pressure required to initiate rupture for leakoff at 13,000 ft = (19.3) (13,000) (0.052) = 13.047 psi.
A nal word of caution when working with large OD ush joint liners whose connection strength in tension is a percentage of the tension rating of the pipe body – careful consideration should be given to what pressure limitation should be imposed for bumping the plug. For example, bumping the liner wiper plug with 2,00 psi on an 11¾-in., 60 ppf FJ liner can increase tension load on the liner connection by over 1800,00 lb.
Mud hydrostatic at 13,000 ft with 16.5 ppg mud = 11.154 psi. Maximum surface pressure = bottomhole pressure minus mud hydrostatic = 13.047 – 11.154 = 1.893 psi. Therefore use 1,900 psi as maximum surface pressure required to run leadoff test.
If the liner is not hung off rst, this tension is also added to the tension load on the drill string. This added tension will not be seen on the weight indicator. Safety factors of tension should be checked for both tubulars before deciding what pressure to use when bumping the plug.
Can 1,900 psi with 16.5 mud be safely imposed on the 9⅝ in. casing?
Burst rating of 9⅝-in. = 6,870 psi. 6,870 psi x 80% (Arbitrary safety factor imposed by operator) = 5,496 psi.
As stated earlier, the vast majority of liners are not pressure tested properly. Most are tested below the fracture pressure of the formation with which a leak could be in communication. It would be a low risk procedure to just test the overlap of a production liner to the fracture gradient at the intermediate casing shoe, with the plausible assumption that the newly run liner will have no leaks. A drilling liner should, at the minimum, have the overlap tested to the fracture gradient at the liner shoe to assure that the overlap will hold if a leakoff test is run on the liner shoe without using a packer.
Since 9⅝-in. casing was set in 12 ppg mud, burst pressure due to difference in mud densities alone at 10,000 ft = (0.052) (10,000ft) (16.5 – 12 ppg) = 2,340 psi.
Maximum burst load on the 9⅝-in. at 10,000 ft = 2,340 psi + 1,900 psi = 4,240 psi. Therefore, maximum burst load during the 7-in. liner shoe Reprinted from World Oil magazine May 1988 with permission from the authors. ,
5.2
w w w . l i n e r t o o l s . c o m
TESTING LINERS WITH A PACKER Gas channels in cement mixture
The authors recommend testing all liner tops and shoes with a squeeze packer because of all of the potential problems stated above. Normally, an operator will immediately pull out of the hole after squeezing to pick up a bit and scraper. The retainer, if used for squeezing, and cement has to be drilled before he liner top can be tested. Another trip has to be made to squeeze the top or do a negative test on it.
Formation Micro-annulus between casing and cement (due to pressure inside casing)
Casing Cement
Micro-annulus between formation and cement (shrinkage)
A negative pressure test should be run on liner tops because of the possibility of mud solids plugging up a small channel or the
existence of “honeycombed” cement or a micro-annulus (see Fig. 22). These type environments often cannot be pumped into and give a false sense of security. A negative pressure test should be equal to any negative differential pressure that the well may encounter late in drilling or completion.
Figure 23 - Gas migration routes in a cemented annulus (after Stewart and Schouten51).
If a production liner is run in a high pressure gas environment where gas leakage problems are known to occur, it would be prudent to consider running a drillstem test assembly (with BHP bombs) that is capable of closing the bottom of the well to test the liner top. This is recommended because gas leakage around liner tops has been reported to produce as little as 40 Mcf of gas per month from shale gas. Leaks of this size may not be detected by a differential test conducted using a packer but no BHP bombs for a short period of time in which the differential column is opened to the atmosphere. Slight gas seepage through the liner overlap could easily be disregarded as air working out of the test uid of temperature expansion of the test uid. Thorough test ing of the liner top is recommended for production liners when the liner top is left exposed since, during the well’s producing
Annulu s not vente d. A close d sy stem
7-in. protection casing
Tubing
Gas percolating through packer fuid
life, there is sufcient time for pressure to build up from a small gas leak. Often drilling is resumed when leaks like these exist in drilling liners and some have considerably affected mud log evaluations.1 Annu lur gas flo w t hrou gh line r t op @12,888 ft
Another tool the author consider superior for identifying a small gas leak is the gas detector. Liner top leaks have been experienced that kept background gas reading from dissipating during drilling operations even though mud density was greater than any pore pressure in the open hole behind the cemented liner. This was due to a gas leaking through honeycombed cement, which cannot be killed with any mud density.
Honeycombed cement
Production packer
Since honeycombed cement will not accept uids nor transmit hydrostatic pressure because it is solid, an increase in mud density on top of it will not be transmitted to the source of the gas. And increased hydrostatic pressure at the liner top imposed by
7-in. casing set @13,097 ft
heavier mud will not stop the gas ow because the gas, due to its lower density, will lubricate through the mud. As each bubble of gas lubricates up through the mud, pressure in the honeycombed
channel is decreased and more gas is allowed in to ll the void.
Gas migration behind liner through cement High pressure gas zone
Thus there is a continuous gas feed-in that will be detectable by a gas detector. It may have been attributed to backgrou nd gas while circulating, or as unexplained early trip gas readings during bottoms up from trips. A good time to run a gas detector in the liner is before drilling out of it. This provides a test in a closed system, and any gas readings would be suspected of being a leak.
Perforated interval (Cotton Valley Lime) (@14,699 - 14,725 ft)
REPAIRING LINER TOP LEAKS 5-in. liner set @15,140 ft
There are different types of liner top leaks. The most difcult
Figure 24 - This completion had a liner top leak that the operator was unaware of until pressure built up on the intermediate casing. Although the liner passed an interval positive test of 3,000 psi with 16.2 ppg mud in the hole, it still leaked gas. And after traveling through the honeycombed cement, the gas was able to lubricate through the mud regardless of the hydrostatic pressure on the liner top. Because the annulus is a closed system, the gas bubbles were not allowed to expand and pressure built up on the 7-in. casing until it burst. (World Oil, January, 1983.)
and costly to repair are those due to honeycombed cement or a micro-annulus (see Figs. 22 and 23). Many times these type leaks cannot be pumped into for squeezing. 1 There has been much published on why has migrates through cements,40-53 but to explain briey how this occurs, Christian, Chatterji, and Us troot 40 showed that excessive uid losses with cement mixtures can cause dehydration in the annulus and produce a bridge that will interfere with pressure transmission from above to the rock formation. If the drop in hydrostatic pressure below the bridge Reprinted from World Oil magazine May 1988 with permission from the authors. ,
5.3
w w w . l i n e r t o o l s . c o m
function of several slurry and well parameters but is governed primarily by (1) the rare of static gel strength
development of the cement, (2) volume of uid loss from the
cement slurry, and (3) compressibility of the cement slurry.52 Other authors also point out that even with no uid loss, volu metric changes and pressure restrictions caused by the hydration of the cement can also allow gas migration. 50 Stewart and Schouten51 show that gas migration can occur from casing contraction also. Most cement companies now have proprietary anti-migration gas cements that reduce or stop gas migration by casing the cement to develop quick gel strengths (by limiting cement hydration shrinkage) or by causing the cement to lose its permeability on contact with gas. A liner top gas leak can create serious problems. The followin g case history gives a graphic example.
15,900
Case 1: Land-based production liner, East Texas feld, Leon County, Texas. This well was drilled to the Cotton Valley Lime and reached a TD of 15,147 ft. A string of 7-in casing was set at 13,097, Fig 24. Due to the existence of fractures in the Cotton Valley, mud weight at TD could not be raised above 15.9 ppg without a loss of returns. But with 15.9 ppg mud, the background gas would not drop below 125 units and the mud weight was continuously being cut from 15.9 to 15.6 ppg. A reasonable assumption was that formation pressure in the Cotton Valley Lime exceeded 16.3 ppg.
Monitor curve
Gamma Ray
The well was safely logged and the decision made to run a 5-in production liner. The liner was run to 15,140 ft with its top located at 12,888 ft. It was cemented using 90 bbl of 17 ppg,
Cement bond curve
Casing collars Correct depth
16,000
low uid loss cement displaced by 15.9 ppg mud. Full returns were achieved while cementing. Top of cement was found at 12,753 ft or 135 ft above the liner top. While drilling cement on top of the liner, the well kicked momentarily and then died. Mud weight was increased to 16.2 ppg and the operator tested the liner top with 3,000 psi on the mud with no leakoff. This was an equivalent mud weight test of 20.6ppg. Thus, the liner overlap cement job was believed to have been successful. No negative test was performed. The well was completed with a production packer inside the top of the 5-in. liner, leaving the overlap in communication with the 7-in. intermediate casing.
Casing collars recorded 15.5 ft deep
Later during production, pressure built up on the 7-in. casing from gas percolating through the honeycombed cement in the overlap, rupturing the 7-in. and causing an underground
GR FR
ow. A total of 156 rig days and $2,565,000 (1977 dollars) were required to repair the well and place it back on production. The authors have had very little experience with special ce-
Figure 25 - Portion of a cement bond log showing excellent bonding of a special cement used to control annular gas migration.
ments to control gas migration. When attempts were rst made to control gas migration with a “compressible” type cement, the job was successful. A description of it follows.
reduces the pressure below formation pressure and if the formation contains gas, then it can feed in and percolate up through the cement causing microcapillaries that will allow passage of gas
Case 2: Land-based production liner, Hemphill County, Texas.
but not uids. If the top of the cement is not fully set before the
The Anadarko basin is noted for problems of gas leakage through liner tops5,6 when high-pressure gas zones such as the Morrow are encountered. The well was drilled to a TD of 16,100 ft using 14.2 ppg oil base mud. A string of 7-in. casing was set to 13.922 ft and a 6⅛-in hole drilled. There were mud seepage losses at about 10 bbl per hour while drilling the 6⅛-in hole. The well was logged and a decision made to run a 5-in. completion liner. Due to excellent well bore conditions, it was decided that it would be feasible to reciprocate the liner .
migrating gas reaches it, the result is an entire column of honeycombed cement. This occurs as the cement slurry undergoes transition form a liquid to a solid. This can create a liner top leak (if cement was circulated above the liner top) that often cannot be squeezed. Kulakofsky goes on to show that the likelihood for this occurrence is a Reprinted from World Oil magazine May 1988 with permission from the authors. ,
5.4
w w w . l i n e r t o o l s . c o m
51
Stewart, R. B., and Schoute n, F. C., “Gas invasion and migration in cemented annulus: Causes and cures” IADC/SPE 14779. Dallas,
A 5-in. LT&C liner was run with one centralizer allowed to oat on every joint and four wireloop cable wipers on every joint 40 ft above, through, and below the pay interval. The liner was cemented with 93 bbl of 16.8 ppg, low water loss compressible cement followed by a 25-bbl, 14.5 ppg mud spacer. The liner was reciprocated 32 ft throughout the entire job until the plug was bumped. The liner was set from 13,596 to 16,098 ft. Top of cement was found above the liner at 13,140 ft.
Texas. February 1986. 52
ACKNOWLEDGEMENT
After drilling cement to the top of the liner, the liner top was tested successfully and the well completed in the Morrow formation from 15,777 to 16,038 ft. The well produced 1.5 MMcfd
The authors thank their respective managements for permission and encouragement to publish this article and for their progressive management philosophy that encourages maximized engineering
with no annular gas ow through the liner top. No squeezes were
efforts on all eld operations. The authors also thank drilling
necessary prior to perforating for production. A portion of a bond log was run is shown in Fig. 25.
foreman Leon Pate and Ray Guidry, and Tim Alexander for sharing their expertise and Judy BenSreti for typing the manuscript. The authors would also like to state that they have read so much literature and talked to so many people concerning the subject matter that they realize that the manuscript does not completely constitute original thinking. Any credit not given to previous authors where credit is due is regretted and unintentional.
LITERATURE CITED 1
Lindsey, H.E. and Bateman, S.J. “Improve cementing of drilling liners in deep wells.” World Oil. October 1973. 5
Suman, G.O., and Ellis, R.C. Cementing Handbook. World Oil. 1977.
AUTHORS
6
Glenn R. Bowman is the regional drilling superintendent for Ashland Exploration’s Houston Region. He graduated from Marietta College with a BS degree in petroleum engineering and has held various drilling engineering positions before joining Ashland in 1984. He was most recently drilling manager for Wainoco Oil and Gas in Houston. Mr. Bowman is a member of SPE and has authored several other papers for World Oil on liners and bottomhole drilling semblies.
Lindsey, H.E. “How deep Anadarko wells are designed and equipped.” World Oil. February 1, 1979. 21
Kulakofskv, David S., “Cement leakage diminished,” Gulf
Coast Oil Reporter , July 1982.
Durham, Kenneth S. “How to prevent deep well liner failure,”
Parts 1 & 2. World Oil. October and November 1987. 23
Goins, W.C., “Better understanding prevents tubular buckling problems.” World Oil. February 1980. 40
Christian, W. W., Chatterji, J., and Ostroot, G. W., “Gas leak age in primary cementing- A eld study and laboratory investigation.” SPE Paper 8257, 1979.
Bill Sherer is the operations manager for Liner Tools LC in Houston, and worked for Alexander Oil Tools from 1984-2001 concentrating on the B&W liner hanger line. Mr. Sherer worked for B&W from 1965 to 1979 and later as a consultant for running liners from 1979 until 1984. Mr. Sherer specializes in optimization techniques for cementing liners and has personally supervised the running of over 300 liners.
41
Tinsley, J. M., Miller, E., Sabine, F. L., and Sutton, D. L., “Study of factors causing annular gas ow following primary cementing,” J. Pet. Tech., August 1980. 42
Sabins, Fred l., Tinsley, John M., and Sutton, David L., “Transi tion time of cement slurries between the uid and the set state,”
ADDITIONAL INFORMATION
SPE 9285, Dallas, Texas, 1980. 43
For more information regarding high rpm liner rotation, centralization, and primary cementation please visit our web -sit e at the bottom of this page.
Cheung, P. R., and Beirute, Robert M., “Gas ow in cements,”
SPE 11207, Dallas, Texas, September 1982. 44
Levine, D. C., Thomas, E. W., Bezner, H. P., and Tolle, G. C.,
“How to prevent annular gas ow cementing operations,” World Oil, October 1980. 45
Cook, Clyde and Carter, L. G., “Gas communication in direc tional wells.” Drilling-DCW, Feb. 1976. 46
Cook, Clyde and Carted, L. G., “Gas leakage associated with static cement,” Drilling-DCW, March, 1976. 47
Sykes, R. L., and Logan, J. L., “New technology in gas migration control,” SPE 16653, September 1987. 48
Sutton, David L., Sabins, Fred, and Faul, Ronald, “Preventing annular gas ow—two parts,” Oil and Gas Journal, Dec. 1984. 49
Liner Tools LC
Stehle, Don, Sabins, Fred, Gibson, Jim, Theis, Karl, and Ven-
Specializing in Liner Primary Cementing
ditto, J.J., “Conoco stops annular gas ow with special cement” April 1985. 50
Bannister, C. E., Shuster, G. E., Wooldridge, L. A., Jones, M.
J., and Birch, A. G., “Critical design parameters to prevent gas invasion during cementing operations,” SPE 11982, San Francisco, Calif., October 1983. Reprinted from World Oil magazine May 1988 with permission from the authors. ,
5.5
w w w . l i n e r t o o l s . c o m
Showcase: The Mechanical Rotating Liner Hanger Optimal for medium to long length liners with severe down-hole conditions requiring high burst and collapse.
Applications: Used to run, cement, and rotate a liner at high RPM. Can be drilled into the hole. Optimum for all wells including deviated and S curved wells.
Features: Recessed, tongue and groove slips are pro-
tected. Unique design allows rotation and reciprocation while cementing. High burst and collapse provided by a casing barrel. Resists hostile down-hole environments with optimum material selection. Controlled and evenly timed slips load th e casing uniformly, eliminating casing failures due to point loading. Optimum slip angle maximizes the hanging capacity of the liner hanger. Simple to operate, requiring multiple right hand rotations to set the hanger.
Reprinted from World Oil magazine May 1988 with permission from the authors. ,
5.6
w w w . l i n e r t o o l s . c o m
