spe-39290-jpt

Upload: alizareiforoush

Post on 02-Jun-2018

218 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/10/2019 SPE-39290-JPT

    1/4 DECEMBER 1997 1339

    T E C H N O L O G Y T O D A Y S E R I E S

    All of us can readily recognize construction cement for what it is

    because we see it every day and it looks how it is supposed to look.

    Even a poor grade of construction cement is recognizable as

    cement to the naked eye. In our industry, cement is placed in thecasing/openhole annulus for two primary purposes: to isolate pro-

    ducible formation horizons and to support the casing. However, we

    do not have the luxury of looking at the cement with the naked eye

    to determine its quantity and/or quality after placement. We are

    forced to rely on the results of measurements from a variety of elec-tronic downhole tools to define the quality and quantity of cement

    placed around the casing during the primary cement placement.The major problem associated with interpretation of the results of

    these measurements lies within ones definition of good or badcement. All the cement-evaluation tools available today, as well as

    the service companies that design and run them, are caught in the

    vicious cycle of trying to define good or bad cement in the oil-

    well/gaswell annular space. This dilemma arises because we, the

    operators, are asking them to tell us the compressive strength of thecement and whether the cement occupies 100% of the annular

    space. We cannot accept that the cement does not exhibit as high a

    compressive strength in the downhole measurements as it does

    under laboratory conditions. If the cement does not measure up, it

    must be bad cement and we must squeeze to repair it.In truth, when we evaluate a cement in an oilwell/gaswell annu-

    lus, all we really need to know is whether cement exists in the annu-

    lus (regardless of its strength) and whether the cement occupies

    100% of the annulus. We must understand that all we can do isplace cement into a section of the annulus that contains neither

    cement nor particulate matter when we squeeze an annulus to

    repair a primary cement. If 10-psi cement exists in 100% of the

    annulus, no portion of it can be removed for replacement by a

    1,000-psi cement. If the annulus is packed with settled barite (fromthe drilling mud) or formation particles, no portion of the particu-

    late matter can be removed and replaced by cement. Only liquids

    can be removed from the annulus for replacement by squeeze

    cementing. Acceptance of this basic premise can both simplify eval-

    uation of a cement sheath in a casing/openhole annulus significant-

    ly and complicate measurement methods significantly. This leavesus with trying to identify solids or liquids behind the casing, not the

    difference between 250- and 5,000-psi cement.

    CEMENT-SHEATH COMPLICATIONS

    Following placement of a particular cement slurry in a

    casing/openhole annulus, numerous events can, and do, alter its

    physical properties from those seen in laboratory test results.Expecting the annular cement sheath to look like the cement in the

    laboratory tests leads to severely erroneous log interpretations.

    Misinterpretation can be caused by American Petroleum Inst. (API)

    testing procedures, from stress cracking of the cement sheath, or

    from cement contamination.

    API Laboratory Testing Procedures.

    1. Compressive-strength testing under laboratory conditions,

    according to API recommended procedures, calls for heating the

    cement to bottomhole static temperature (BHST) in 4 hours. The

    cement system is then maintained at BHST (and pressure) for therequested testing time (generally 24 hours). The measured com-

    pressive-strength development is reported. This leads to misinter-pretation because percent bonding (or the bond index) is based on

    the relationship of the amplitude (or attenuation rate) of the sonic

    signal at the lowest value of amplitude (the highest value of atten-uation rate) and the remaining cement in the annular cement col-

    umn. No adjustments are made for pressure, temperature, or den-

    sity changes (from the tail slurry to the lead slurry) as the logging

    tool moves up the hole. Cement-column heights can vary signifi-cantly, creating varying temperature and pressure effects on the set-

    ting of the cement. The cement at the top of the lead slurry may not

    even be set by the time the cement sheath is logged because of these

    varying environmental conditions. Field measurements of the time

    required for a well to stabilize to 100% of geothermal static tem-perature have been measured, and the times range from 4 to 40

    hours, depending on drilling times; circulation times and rates;

    inlet-fluid temperatures; casing/hole geometry; fluid type, density,

    and rheology; depth; and physical location.

    2. API thickening-time tests are run at a constant bottomhole cir-culating temperature with predetermined slurry heat-up rates. In

    practice, the real-time heat-up rate is much slower than API-rec-

    ommended heat-up rates. At faster heat-up rates, the cement-slur-

    ry-retarder loadings must be increased to yield sufficient thicken-

    ing times for placement. These extra retarder loadings severelydelay the cement-setting process uphole because the temperature

    decreases instead of remaining constant.

    Stress Cracking of the Cement Sheath. After the cement sheath

    has reached a fully developed crystalline state, stress changes causedby casing expansion can create cracks (tensile failure) in the cementsheath. Casing expansion can be caused by excessive casing test

    pressures, excessive temperature changes, or expendable perforat-

    ing guns. The cement sheath, once it has been stress cracked,

    appears as no cement on bond logs or as ultraweak cement to no

    cement on the ultrasonic evaluation logs. Severe stress cracking cangenerally be negated by cementing with low-density, low-compres-

    sive-strength cements with greater ductility. Generally, the greater

    the compressive strength of the cement, the more susceptible it is to

    stress cracking. Ideally, to prevent the stress-cracking problem

    entirely, the cement ductility should approach that of rubber.

    Cement Contamination. Cement mixed with drilling mud, spacer,or formation fluids (water or gas) in the annulus during and after

    OILWELL/GASWELLCEMENT-SHEATH EVALUATIONK.J. Goodwin, SPE, Mobil Technology Co.

    Copyright 1997 Society of Petroleum Engineers

    This paper is SPE 39290. Technology Today Series articles provide useful summary information

    on both classic and emerging concepts in petroleum engineering. Purpose: To provide the gen-eral reader with a basic understanding of a significant concept, technique, or development with-in a specific area of technology.

  • 8/10/2019 SPE-39290-JPT

    2/4

    1340 DECEMBER 1997

    placement yields a sheath of drastically reduced density and com-

    pressive strength. This contamination, especially severe gas influxfrom the formation, yields a cement sheath that is not recognizable

    as cement on a bond log but is readily recognizable as a cementwhen raw acoustic-impedance data from ultrasonic tools or raw

    attenuation-rate data from segmented-type bonding tools are used.

    CEMENT-BOND LOGS

    Cement-bond-log measurements represent a 360 averaging of

    reflected and refracted sonic signals from the casing, the cement

    sheath, and the formation. Depending on the log presentation, the

    results can satisfactorily indicate a total-free-pipe condition; canimply an annular liquid-filled channel (as long as the channel is in

    contact with the casing) or a microannulus at the casing/cement

    interface; and can satisfactorily indicate formation signals, as long as

    the cement is not cut with gas (either from natural-gas influx or pur-

    posefully foamed) and as long as a pressure pass and a nonpres-sure pass are included on the log presentation. Cement-bond logs

    cannot identify channels within the cement sheath or channels at

    the cement/formation interface, very weak or gas-cut cement on

    thick-wall casing, or particulate-matter fill in the annulus. What

    does the bonding index or percent bonding tell us? Essentially noth-ing except that the total-free-pipe condition does not exist at the

    measurement point. The 360 averaging of the returned signal can-

    not satisfactorily differentiate between an annulus that is half full of

    good cement and half full of liquid and an annulus that is totally fullof severely gas-cut cement or an extremely weak cement. Neither

    can 360 signal averaging geometrically locate a channel at the cas-

    ing outer diameter (OD) nor identify the contents of that channel.

    ANNULAR SE GMENTATION FOR ANALYSISCurrently available ultrasonic measuring tools, as well as the seg-

    mented-bond measurement tools, can successfully measure the qual-ity of cement in contact with the casing OD at varying points around

    the casing circumference. These tools, like the cement-bond logs, also

    cannot identify liquid- and/or gas-filled channels within the cement

    sheath or at the cement/formation interface. The major problem withthese types of tools is that logging service companies attempt to cre-

    ate multicolored patterns on their log presentations to beautify, sim-

    plify, make more saleable, and otherwise confound the end user.

    These colorations could provide fully satisfactory definitions of the

    annular-fill material if the cement never became contaminated, wasnever stress cracked, reached full strength over the entire column at

    the same instant (hopefully before logging), and was mixed in thefield as exactly as it was mixed in the laboratory. Unfortunately, these

    things never happen in reality. Most frequently, combinations ofthese events cause the cement acoustic impedance to be less than

    that of the drilling mud and sometimes equal to or less than the

    acoustic impedance of water spacers. How do you differentially colorthese types of cements vs. water or drilling mud? You cant. However,

    if we are to identify the annular components for a squeeze/no-squeeze decision realistically, we must differentiate between these

    types of materials. It is not necessary to define whether the cement

    has 5,000- or 50-psi compressive strength or to determine whether

    the cement is good or bad. It is only necessary to differentiatebetween a crystalline material (cement or particulate annular fill)

    and a noncrystalline material (drilling mud, water, or gas).

    To surmount the interpretation difficulties created by the col-

    oration problems, it has become necessary to ask for (or demand) the

    raw-data presentations. For the ultrasonic logging tools, the raw-datapresentation consists of acoustic-impedance measurements (on a

    scale of 0 to 5 MRayle) from each transducer. For the ultrasonic tools

    with the rotating transducer, the raw-data presentation consists of

    acoustic-impedance measurements every 5 or 10 up to 40 (on a

    scale of 0 to 5 MRayle). For the segmented-type bond tools, the raw-data presentation consists of unfiltered attenuation-rate measure-

    ments from each tool segment (on a scale of 0 to 9 db/ft). Acoustic

    impedance is the product of material density and the composite

    velocity of sound through the material. Consequently, a liquid, such

    as water, drilling mud, or a gas, must by definition exhibit a constantacoustic impedance (or attenuation rate). Depending on the quality

    of the tools transducers, a constant acoustic-impedance value

    should, and does, yield a straight-line (or constant) value on the log

    presentation. The following examples illustrate the technique.These examples are intended to demonstrate the types of acoustic-

    impedance (or attenuation-rate) signatures resulting from variousmaterials occupying the annular space, not tout one service compa-

    ny tool or format over another. Please note that, for each instance

    (i.e., water, cement, and gas-cut cement), definitive and identifiablesignatures are exhibited that, in reality, represent the material and its

    condition occupying the annular space immediately outside the cas-

    ing. All the ultrasonic logging tools and the segmented bond tools are

    capable of producing this type of information; information that is

    required to assist in making an intelligent decision about the require-ment for remedial repair of a primary cement sheath.

    CEMENT-EVALUATION RE COMMENDATIONS

    The cement-evaluation guidelines offered are intended to make the

    job of the completions, production, or operations engineer easier.The recommended cement-evaluation-log presentation format

    Cement-evaluation tools are not recommended if the primary cement job was conducted according to recommended primary cementingpractices and no problems were experienced with mixing or placement of the primary cement. These practices include the following.Casing is 100% centralized.Hole is circulated until a minimum of 85% of the annular volume is circulating.Drilling mud is thinned as much as possible before cementing.Casing is reciprocated or rotated during circulation and cementing.

    A minimum of 500 linear ft of water (or spacer if required for pressure control) in the annulus is pumped ahead of the cement.Cement slurry is mixed within 0.2 lbm/gal of planned density.Cement is circulated to the surface.

    Flow or loss circulation problems are solved before cementing.Bond logs are not recommended in 95/8-in. or larger casing.If bond-log tools or Western Atlas SBT (segmented bond tool) are used for cement evaluation, a pressure pass and a nonpressure pass

    must be made. The casing pressure used should be equal to any changes in hydrostatic or test pressure after the cement hasset plus 500 psi.

    A bit and scraper should be run to clean the casing wall before running any cement-evaluation log.Cement should have a minimum compressive strength of 250 psi at the top of the cement column before a cement-evaluation log is run.If a cement-evaluation log is deemed necessary, bond logs are not recommended. Any of the evaluation logs that can segment the

    annulus and present readable raw data is recommended. These include Schlumbergers CET (cement evaluation tool) or USIT(ultrasonic imaging tool), Halliburtons PET (pulse echo tool) or CAST-V tool, and Western Atlas SBT.

    TABLE 1GENERAL RECOMMENDATIONS

  • 8/10/2019 SPE-39290-JPT

    3/4

    1342 DECEMBER 1997

    should make determining the presence and the competency of the

    cement sheath in place much easier. However, in cement-sheathanalysis, the one caveat that one must always be aware of is that agood cement sheath at the cement/casing interface does not neces-

    sarily guarantee that a channel does not exist within the cement

    sheath or at the cement/borehole wall. Most often, those types of

    channels can be detected only when some fluid (oil or water) or gas

    is flowing at a sufficient rate for the flow to be detected with either

    a noise or temperature log. Large amounts of money are wastedannually because of poor or meaningless bond logs or because of

    misinterpreted cement-sheath quality. Recent developments in

    cement-evaluation tools provide much more meaningful interpreta-

    tions than common bond logs. The cost of these newer tools ismuch greater; however, if an unnecessary squeeze is circumvented,

    the overall cost is much less. If it is absolutely necessary to run a toolto evaluate a cement job (as opposed to running one out of curios-

    ity or habit), then the newer tools are always recommended over the

    common bond logs. Table 1 gives general recommendations, andTable 2 lists recommended log presentations and settings. Figs. 1through 4 show typical log signatures for various tools.

    INTERPRETATION GUIDELINES

    Cement is cement is cement. The terms good cement and bad

    cement imply the same thing: that there is cement in the annulus.

    Regardless of the cements compressive strength, it cannot beremoved from the annulus so that the bad cement can be replaced

    by the good cement. In fact, if the annulus were filled with sand (or

    barite or any other type of small insoluble particles), it also could

    not be removed from the annulus for replacement by good cement.Liquid is liquid is liquid. In log interpretation practices, a liquid

    is any noncrystalline material (drill mud, water, gas, or other suchmaterial) that can be removed from the annulus and replaced by

    TABLE 2RECOMMENDED LOG PRESENTATIONS AND SETTINGS

    Recommended Presentations

    CET or PETFour acoustic impedance tracks, each containing raw-data measurements (acoustic impedance) for each two transducers.

    A relative-bearing curve.Acoustic-impedance tracks on a scale of 0 to 5 MRayle (Z).A normalized W2-W3 crossplot (CET only).

    USITThirty-six acoustic impedance tracks, each on a scale of 0 to 5 MRayle.

    A relative-bearing curve.CAST-V

    Use the altcast.cls presentation only.A minimum and maximum value of acoustic impedance for each of nine annular sections (deleting average and maximum

    acoustic-impedance curves is recommended for clarity).Acoustic-impedance curves on a scale of 0 to 5 MRayle.Presentation is oriented so that the low side of the hole is in Sec. E and the high side of the hole is in Secs. A and I.

    SBTSix attenuation rate tracks (one from each pad).

    A relative-bearing curve.Attenuation-rate curves on a scale of 0 to 9 db/ft.Nonfiltered data.

    A cement map scaled only above and below the free-pipe attenuation.

    Running Rules

    CETStandard tool may be run in water-based-mud (WBM) weights up to 12.5 lbm/gal only.Modified tool (EPS Mod-1) may be used in WBMs gal.Not recommended in inverse-emulsion muds. Modified tool can be run in oil-based mud (OBM) weight up to 12 lbm/gal.Maximum casing size is 95/8 in. A special tool is available for 13

    3/8-in. casing only.Run with and without pressure if the microannulus is less than 0.006 in.Maximum temperature is 350F.Recommended logging speed is 4,000 ft/hr.Should be accompanied by a gamma ray log.

    USITCan be run in WBM weights up to 16.0 lbm/gal.Can be run in OBM weights up to 11.6 lbm/gal.Can be run in inverse-emulsion muds.Should be run with the General Purpose Inclinometer Tool package for log/hole orientation purposes.Maximum temperature is 350F.Run with and without pressure if the microannulus is larger than 0.006 in.

    PETCan be run in WBM and OBM weights to 14 lbm/gal.Should not be run in inverse-emulsion muds.Can be run in casing sizes up through 133/8 in. if extended transducer cans are used.Recommended logging speed is 3,600 ft/hr.Maximum temperature is 350F.Run with and without pressure if the microannulus is larger than 0.006 in.

    SBTCan be run in any density drilling mud.Can be run in WBMs, OBMs, and inverse-emulsion muds.Can be run in any casing size up to and including 16 in. (maximum expanded tool diameter = 15 1/2 in.).Should be run with and without pressure in all cases.Maximum temperature is 350F.Maximum recommended logging speed is 2,400 ft/hr.

  • 8/10/2019 SPE-39290-JPT

    4/4

    DECEMBER 1997 1343

    good cement. If solids cannot be removed from the annulus for

    replacement by good cement and liquids can be removed, then

    those facts limit the information we need from a cement-evaluation

    log. All we really have to determine from a cement-evaluation log

    is the presence and position of solids fill or liquid fill in the annu-lar space. We do not need to know the compressive strength, den-

    sity, or bonding percentage of the cement. In addition, we can do

    nothing about it at this point.

    All we have to do now is learn to recognize what a liquid looks

    like on a cement-evaluation log. The percent-bonding curve (theamplitude curve) cannot differentiate between partial liquid fill or

    partial solids fill of the annulus because it averages the reflected sig-

    nal from 360 of the casing. The presence of a liquid-filled channel

    can be determined from a bond log (on the microseismogram) ifthe log is run with and without pressure. Otherwise, a liquid-filled

    channel cannot be differentiated from a microannulus.

    Because the CET, SBT, CAST-V, USIT, and the PET (defined in Table

    1) segment the annulus for analysis, the existence and position of a

    liquid-filled channel is much easier to identify. Note that a liquid-filledchannel must be in contact with the casing before it can be identified.

    No cement-evaluation tools currently available can identify a channel

    within or outside of the cement sheath. It is possible to identify a per-

    fect primary cement job on an evaluation log and still have interzon-

    al communication outside of the cement sheath. These outside chan-nels can be identified only with a noise or temperature log.

    Recognizing a liquid-filled segment of an annulus or an entire

    annulus filled with liquid is very easy. Remember that the velocity

    of sound through the casing wall is primarily a function of the

    shear strength and density of the material in intimate contact(acoustically coupled) with the casing. Therefore, all the physical

    measurements in cement-evaluation logging (i.e., amplitude, atten-

    uation rate, and acoustic impedance) are functions of the shear

    strength and density of the annular material. Variations in thosemeasurements indicate variations in the density and/or shear

    strength of the annular material. Conversely, liquids do not exhib-

    it shear strength and exhibit practically no change in density over

    meaningful lengths of the annulus. Because the physical propertiesof the liquids do not change, the physical measurements of ampli-

    tude, attenuation rate, or acoustic impedance would not normallybe expected to change either. In fact, they dont. Therefore, a liquid

    in any segment of the annulus being evaluated should, and does,

    exhibit an unchanging value of amplitude, attenuation rate, or

    acoustic impedance (a straight line).

    SI METRIC CONVERSION FACTORS

    ft3.048* E01=m

    F (F32)/1.8 =Cgal3.785 412 E03=m3

    in.2.54* E+00=cm

    lbm4.535 924 E01=kg

    psi6.894 757 E+00=kPa

    *Conversion factor is exact.

    K.J. Goodwinis Associate Drilling Engineering

    Adviser at the Mobil E&P Producing TechnicalCenter in Dallas. He teaches Mobils cement-

    ing courses and conducts remedial-cementing

    and cement-sheath-evaluation seminars for

    Mobil and for state and federal regulatory

    groups. He also provides technical assistance

    for company drilling and production operations worldwide in

    these areas. During his career, he has worked extensively in

    stimulation, cementing, and lost circulation. Before joining

    Mobil in 1985, he worked for Dowell and the Western Co. of

    North America. Goodwin holds a BS degree in chemistry from

    Northwestern State U. in Oklahoma. Currently an SPE Short

    Course instructor, he was a 199495 Speakers Bureau speak-

    er and a 199091 member of the Forum Series in NorthAmerica Steering Committee.

    Fig. 1Known liquid signature from the CET. Fig. 2CET acoustic-impedance gas-cut cement signature.

    Fig. 3PET acoustic-impedance uncontaminated cement signature.

    Fig. 4Segmented-bond-tool presentation.