Cement Bond Log (CBL) Evaluation Guidebook QC and Interpretation
Huawen Gai BP EXPLORATION
CBL Evaluation Manual QC and Interpretation
Whether to run a cement evaluation log? What tool to choose? When to run the log? How does the CBL tool work? How do the conditions affect the log? How to carry out QC operation?
• The systematic way to Interpret the Cfl • Historical mIstakes corrected • Squeeze consIderations
DrHuawen Gai Drilling and Completions Branch liP Research Centre Sunbury-on-Thames Middlesex TWJ6 7LN UK Tel. (+44) (0)932 763495 Fax (+44) (0)932 763352
Acknowledgements Many PEs and DEs in BP Explorationhave directly contributed to this manual. I particularlywant to thank David Law, David Munro, Lee Richardson and Daryl Kellingrayof Dyce Aberdeen and Chris Greaves of Westport Lab Houston for their most valuable comments and advice. I want to thank the following people in Drilling and Completions Branch who made significant technical contributions or editorial advice in preparing this manual: Chris Lockyear, Dan Ryan, Ashley Hibbert, George Brown (Production Operation Branch),John Mason, John Bensted and Nigel Brown. The help from Robin Lewis, Ian Palmer and Andy Gardner in associated experiments is most appreciated. Several people from logging service companies assisted in supplying information. I’d like to thankSigveMauritzen, AVince SpinelliandPitakWangvarangkoon ofSchiumberger, and Ruud Henskens of Atlas Wireline for the valuable discussions.
H Gai Sunbury, UK June, 1992
L
Cement Evaluation Manual - QC and Interpretation
1992
CBL Evaluation Manual QC and Interpretation —
Contents: How to Use the Manual and Quick Reference Charts ~1. Things you should know about the tools
i-viii 3
Whether to run a log and Criteria to choose a tool—QCis most important— What the tools can do at their best— Do not interpret logs in isolation
~2. Tool principles andJargon
4
How the measurement is made — How the tool works downhole — Important features ofthe tool structure — What a log looks like — El, £2, E3 etc. — Gates — Transit Time — Stretching — Cycle skipping — Casing arrivals, formation amivals, and mud arrivals— Fastformations— p-annulus— Free pipe
~3. Information Included in the log
8
The log header— The bodyof the log— The log tail — The BP questionnaire
~4. Parameters affecting the log results
9
p-annulus — Eccentricity — Channelling — Casing coating — Fastformations — Mud type and conditions — Temperature — Casing diameter and thickness — Casing damages— Casing standoffand open hole geometiy — Double casing strings — WOCtime — Cementparameters and conditions — Computer keyboard operations
0. Operation QC In three phases
17
Before logging — During logging — After logging
~6. Interpretation
21
Intespretation Chart — QC review— Quick checks— Examine the TTcurves— and the CBL curves and the t’DL log — BPI calculation with example — Special Investigation Chart
~7.
Cementing operation
29
Cementing operation — the CFS
~8. Squeeze considerations
30
Where did the cement go — Whatkind of channel could it be — Where to squeeze
~9. Log examples
33
Log header/tail and scale — ~i-annulus — Eccentricity — Fastformations — Muddensity— Temperature — Green cement — Double casing string
~1O.Data and
Charts for Reference
41
Toolperformance comparison — Tool characteristics — Soundvelocities in muds formations — Casing expansion underpressure — Relationship between eccentricity, amplitude and TTreduction — Interval lengths required for isolation — Amplitude compensation charifor various muds — 3’El readingsfor 100% cementedand 0% cementedpipes
[
Index
47
Cement Evaluation Manual - QC and Interpretation
1992
0 U 0) 0) U
0
Concord.. 99/79901
CBL Evaluation Manual
—
QC and Interpretation
How to use the Manual • Who Isltwrltten for? For PE5 and DEs within BP who maybe involved in planning cement job evaluation,witnessing the logging process, interpreting cement bond logs, or making squeeze job decisions.
• How to use It? Much efforthas been made to ensure that it is easy to use infield applications. Beginner or expert, you can start where you need: Job planning and witnessing Log interpretation Undertaking squeeze job Beginner or requiring basics
(“s”
§1, §5 means section) §4, §6-8 §8 §1-4 and then §5-6
The manual is cross referenced by ~(~*) and supplied with working examples. Wherever you start, you should find the needed information. If you do not get what you want, let us know and we will sort it out for you and improve the manual! • Update your expertise: The manual has some new results from recent research. If you are already an expert in interpretation, you are advised to readthrough at least §6 to updateyour expertise.
Structure
What’s included ~fs’~ S Key knowledge of the tools and their selection (~1) • Principles and jargon used (~2)
5 5
~‘
Operation OC in three phases: (~5) Before the job, during the lob and after the job Log contents and format (~3)
• Operation OC review (~6) InterpretatIon including squeeze consideration (~7) Examples and reference data (~ 9&10)
IS 5
• Cement quality, probability of zonal isolatIon
Conclusions
S Squeeze job recommendation
~
S invalid log Comments/queries to DCB RCS
...~ ‘~— —
— —
S
Comments on the manual o~ general interest
•
Help on log interpretation
The highlighted header at the top of each page tells you where you are in the manua~j
[
Cement Evaluation Manual - QC and Interpretation
1992
The following charts also appear in the ‘Things You Should Know” and “Log Interpretation” sections. They are collected here for easy access or quick reference. Please refer to the appropriate section if any detail of the charts is required.
CBL Evaluation Manual - OCand interpretation
Page ii
This Chart tries to help answer three key questions for decision making.
Whether to run a cement evaluation 1. Aims of the cement
~
Zonal isolation of liners transversing reservoirs should have higher priority
3. Cernenhng service All appraisal wells should be bond logged and so should most production wells
If experienced personnel with good performance records in a field are used, the number of Wr~ai~ be logged can
~?n~te~te
In job planning phase these three factors must be carefully considered to descide if the cement job should be bond logged
What tool to choose? 1. Mud weight etc* For OBM>lOppg & WBM>1 3ppg only CBL type oftool can be used *Please refer to § 10. land § 10.2 for other constraints
2. Achieve aims ofthe logging 3. Logging service company • Capability of the tools (§1.2) • experience of personnel • Importance of isolation • performance of their tool • Possible cement conditions • costs
factors.
4l1I~~~’
Although in most cases both CBL and CET types of tool can be used (~1O.1,the most important is QC), avoidthe CBL when there are: 1) Intervals containing a ~s-annulus (~2.12)which logging under pressure failed to eliminate(~4.2,~9.3) 2) Intervals where isolation is required contain fast formations ~ §9.4)
I I I
When to run the log?~ To avoid green cement (~4.12,§9.8): do not start logging~”)
I within 8hrs after the cement has set I I 2) To avoid l2-annulus especially for CBL: do not reduce thel I pressure in central hole after cementing and before loggingl U~L1,~9.2)
CBLEvaluation Manual - QC and Interpretation
J
Page ii
The Flow Chart on the opposite page offers a systematic way to interpret a CBL log. The actions at each step are briefly explained on this page. See §6 and the given references for detail.
-
check:sJob planning execution • ‘The 5 data sets’ (see p22) S The logging engineer’s comments • The presentation
> Mainly in log header
Take 1 or 2 minutes to see if theIT and the CBL are in the expected range and the VDL log has good contrast.
I E.g.:7”, 29 lb,ft casing The TI should be in I the order of 270~us(see p22) and CBL i should be from 1 to 65mV.
Log completeness
The TI curves are bound to vary. You must know why they did on the log in handj
E.g.:the marked zone is probably due to eccentralisation - No fast formation was confirmed by other logs including VDL.
To confirm TOC,
I
good cementand I ‘free pipe” is to i provide key references for the
BPI
Mainly to substantiate TI and Ampiitude indications
Concentrate on zones of interest. The longer the I interval of hight BPI value, the better chance of I zonal isolation (see p26 for an example). I
CBLEvaluation Manual - QC andInterpretation
SPI
Elf - Elm El - El f c
Page iv
1
CBL Evaluation Manual - OC andInterpretation
Page v
This Flow Chart provides a common sense approach to Special Investigation. A good understanding of what can affect the tool perfomiance and how (in §2 and §4), is very useful.
involves the following actions
Special investigation
which may have to be iterative
1. Finding information~ 2. Analysing abnormal log behavior 1 3. Calculating the probability of zonal isoIati~
Proceedl* ~
____________
Ye~
fthl
accounted for?
Np~
~.c~yalid io~
* Either go to the next action or resume the main interpretation flow chart on previous page
CBLEvaluation Manual- QC and Interpretation
Page vi
:
This Chart and the one on the next page are concerned with decision of a squeeze job. They ask three crucial questions and offer common sense answers. Do not decide to squeeze before answering these questions!
1. Where did the cement go? Analyse the well and cementing conditions together with the log
Not clearly ~cated?~6.2~
Yes
CheckIf any fluid loss occured during drilling or cementing(~6)
~
Calculate the difference from expected value (Note the hofe gauge, washout: caliper log).
N
1~~(es
Possible cement loss by large quantities. Such cases are usually easily Identifiable.
Possible heavy contamination or bad slurry leading to green cement. Re-run the CBL If possible.
CBLEvaluation Manual- OCand Interpretation
The cement is likely to be still in the annulus but badly bonded tothe casing and maybe to the formation as well.
Possible bad contamination at the cement top or bad mud removal. _____________
Page vii
6 scenarios of communication channels Eccentred casing, mud channel on the narrow side. Heavily contaminated cement which may or may not be solid. Contaminated but solid cement on the wide side with mud channel on the narrow side. Contaminated but solid cement on the narrow side with mud channel on the wide side. Gap between the set cement and the casing Thick mud cake between the set cement and the formation,
kasy: mud cement contaminated cement formation
Well conditions and cementing operation vs potential channels Channel type
Bad casingcentralisation
Deviated wells
‘.4, _________________
Severe iii high deviation
Displacing contamination
cementing operation problems
I,
Washoutsection
~‘o~ei weiis
a, Some delayed communications observed In porous reservoirs are believed to be caused by the disintegration of the mud cake. This type of channel Is hardly detectable with today’s technology
Bond logs provide vital information for squeeze job design particularly in the following areas: 1. Depths/lengths of communicating channels for positioning the perforating gun and bridge plug or packers. 2. Azimuth of communicating channels for perforating shot phase arrangement: a 45 degree channel can be missed! 3. Identify the vent for the channel filling substances.
CBL Evaluation
Manual- OC and interpretation
Page viii
Things You Should Know
This part of the Manual provides the starting point for proper use of the tool and interpretation of the log.
CBLEvaluation Manual - CC and Interpretation
Page
~1. Things you should know 1.1 Whether to run a cement bond log or not can result in substantial expenditure or savings, 1. Aims ofthe cement job Zonal isolation of liners transversing reservoirs should have higher priority
I 2. Existing knowledge ~1 the field I All appraisal wells should be bond logged I and so should most production wells
3. Cementing service company
I
______________________
casing than intermediate
and is a decision which relies largely on experience, logging objectives (e.g. if only TOC is required a temperature log run at the right time would be the best) and government
If experienced with good performance personnel
legislation operations cementing Criteria to (~10. choose 1, §10.2), a particular but areusually tool cangovernedby be dictated by factors the well suchas conditions experience, and
records in a field are used, the number of wells be reduced to be logged can probably
required logging emphasis and cost. Expensive tools do not necessarily give the extra
1.2
factors cementplanning considered must job should tobedescide phase carefully be these bond if thethree logy
information you really need! (This Manual concentrates on the CBL only.) QC is the most important part of cement bondlogging. This is because the effects of most
parameters them as much on the as logpossible are not even quantitatively while logging. knownconditions, (~4).The best 1.3 Every tool has its islimitations under perfect e.g.:thing to do is to eliminate The CBL tool can only give an average measurement of the 360° annulus (~2.3).It is impossible forthe CBLto indicate the position of a channel. TheVariable Density Log (VDL, §2.4) is a qualitative log and does not indicate how much of the annulus is bonded (~6.5).
I
I
1. Mud weight etc* For OBM>1 Oppg &
I [~A~hieve aims ofthe logjI~ijl3. Logging service company I . Capability ofthe tools (§1.2) • experience of personnel
I
I of tool can be used • Possible cement conditions I *PI~serefer to § 10. land II WBM>l3ppg only CBLtYP~j~~ Importance of isolation factors. I § 10.2 for other constraints
1
I
The CET tool and the US! (~io.i)concentrate only on the casing/cement interface in their data measurement. The ‘cement map’ is in fact an interface map. If a channel is beyond
• costs 1 • performance oftheir tool
~Although in most cases both CBL and CET types of tool can be I used (~1O.1,the most important is QC), avoid the CBL when there I are: I 1) Intervals containing a ~.t-annuIus (~2.12)which logging under pressure failed to eliminate(~4.2,~9.3) I 2) Intervals where isolation is required contain fast formations ~ (~4.5,§9.4)
I
1.4 Although cement logging can be quantitative, itis not always accurate because of the this interface it isbond not detectable. many factors that affect the log ~4). Therefore always remember to review the ñill picture including the way the cement job was carried out (~4,§7, §8). Don’t let calculated results override common sense.
1
(~iE~To avoid green cement (~4.12,§9.8): do not start loggin~”~
I within 8hrs after the cement has set I 2) To avoid ~.u-annulus especially for CBL: do not reduce the I pressure in central hole after cementing and before~~n~J §9.2)
CBLEvaluation Manual - CC and Interpretation
Page 2
~CBLEvaluation Manu a - CC and Interpretation
Page 3
Things you should know
~2. Tool Principles andJargon used 2.1 Row the measurement Is made: The CBL tool has a sonde for measurement and an electronics cartridge for signal acquisition and transmission. The sonde works on “piezoelectricity” — a physical property of certain materials such as quartz and piezo-ceramic — ifthe material is deformed, a voltage will be generated on its surfaces and conversely, ifa voltage is applied to the material it will deform accordingly. Mechanical vibrations or “waves” approaching such a material can therefore be converted into voltages and by measuring these voltages the mechanical waves can be analysed. 2.2 How the tool works down hole: The sonde (see Fig.2.1) typically has one transmitter and two receivers which are in a metal mandrel and are 3’ and 5’ fromthe transmitter. When the transmitter is fired, it will send out a cylindrical compressional wave train (usually about 20kHz). This wave train will travel through the mud into the casing/cement/formation structures, where different types of waves such as shear wave wifi be induced by mode conversion phenomenon. Some of the induced waves will travel along the cased weilbore, and on their way they will send their characteristics back to the mud. The receivers in the mud can therefore pick up these waves which carry information about the media. The early part of the received waveform is found to be indicative of the quality of the bond between the casing/cement interface: the better the bond, the lower the amplitudes. The later parts can tell us, for example, how fast the the sound travels in the formation (~2.10, ~4.5). The 3’ receiver is dedicated to measuring the first peak of the received waveform, including itsarrival time and itsmaximumvalue, conventionally called El (~2.5).The arrival time is used to check if the tool is properly centralised for a valid log (~6.3),and the El value is used for bond quality calculations (~6.6).The 5’ receiver records the whole waveform to produce either the VDL or the signature log (~6.5)or both. This provides more information to help detect the bond (~6.5). 2.3 Important features of tool structure: The transmitter and receiversare tube-like andwill respond to mechanical waves without telling their radial directions. This means that the CBL measurement is an average of the circumference and is unable to detect the azimuthal position of an uncemented area in the annulus, known as a channel. Fig. 2.1 CBL sonde CBLEvaluation Manual - QCand interpretation
Page 4
To prevent the housing mandrel from “short-circuiting” the transmitted wave, it is crosssectionally slotted. Consequently the sonde is not rigid and can bend under its own weight. The tool centralisation should take this into consideration (~4. 2). 2.4 What a log looks like: A sample ofa common CBL log is shown in Fig.2.2. The left track is the Transit Time (Ti’) curve (~2.7).Usually in this track there are also a gamma-ray log and a casing collar locator (CCL) log for depth tie-in. The middle track is the CBL amplitude curve which is a continuous reading of El (~2.5).The right track is the VDL log which is produced by applying a simple processing to the waveforms received by the 5’ receiver. The processing is essentially thresholding and stacking: positive peaks are represented by black line segments and negative peaks white ones; these line segments are then stacked along the well depth and the VDL log is created. If the waveforms are stacked without the thresholding treatment, the log created is called signature log. Note these curves may be named with different mnemonics or in different scale, e.g. extra letters may be used to distinguish curves generated by sliding gate (~2.6)from those by fixed gate.
Fig. 2.2 Log sample The log interpretation is all about making sense of the curves in the context ofthe cement job and the well.
CBL Evaluation Manual - QC and Interpretation
Page 5
Things you should know Jargon: 2.5 El, E2, E3 etc. (Fig.2.3) El means the amplitude and duration of the first peak,conventionally stated in mV. Similarly E2 is the negative peak following El, E3 is the next positive peak, and so on. 2.6 FIxed gate, Sliding or Floating gate (Fig.2.3) For datareduction, the electronics starts measuring El only when it is about to arrive, and stops measuring when it has passed. This measuring period is called a gate. It is vital that the gate is opened in the right position on the waveform in order to see El. There have been two ways of setting the gate: the fixed gate and the sliding gate. The fixed gate is set by the tool operator to straddle El. Once set a fixed gate will open and close irrespective of the waveform. The sliding gate is triggered open by the waveform when it has first reached a preset detection level. Mostof the timeboth typesofgate will givethe same value ofEl. However, whenthe El position is caused to change by certain conditions (e.g. fast formations, §2.11), a fixed gate could miss it but a sliding gate would pick it up. On the otherhand a sliding gate could be triggered open by E3 instead of El if the latter is lower than the detection level. Therefore both types of gate are now commonly used together. 2.7 Transit Time (Ti’) (Fig.2.3) The Ti’ is the time span between when the transmitter is fired and when the waveform at the receiver has reached a preset detection level. The sliding gate is opened at the TI’. Note: unless El amplitude coincides with the detection level, the Ti’ is not the time when El reaches its maximum value. See also §4.2. 2.8 StretchIng (Fig.2.3) Stretching means the increase in the T~due to the decrease in El caused by, for example, increased bond quality. Because the Ti’ is related to the detection level, a decreased peak wifi reach the detection level later and thus “stretch” the Ti’. 2.9 Cycle skipping (Fig.2.3) When the El value for some reason (e.g. very good bond or severe eccentralisation) becomes lower than the detection level, the first time the waveform reaches the detection level could be part of E3 (or even E5, E7 etc. if the early ones all fall below). The TI’ measurement wifi skip a cycle (or two cycles, three cycles and so on). The TI’ will be increased by roughly an integer number of the wavelength.
CBL Evaluation Manual - QCand interpretation
PageS
2.10 CasIng arrivals, Formation arrivals, Mud arrivals (Fig.2.4) The received waveform is extremely complicated. It is a combination of wave trains which have gone through different media such as casing, formation and mud and consequently carry information about them. Casing arrivals, formation arrivals and mud arrivals are terms to refer to the corresponding portions ofthe waveform. Because the tailof one type ofarrivals will always be eaten by the head of the next, on a single waveform one cannot clearly see the joints of two types of arrivals. However, when the VDL log is generated, the features of these arrivals usually stand out as shown. The features of casing arrivals and mud arrivals on the VDL are straight stripes starting at relatively fixed times. This is because the acoustic properties of the steel casing and that of the mud column are usually homogeneous. The sound velocities in the formations, however, can vary substantially along the well, making the formation arrivals wander in time as shown in wiggly stripes. 2.11 Fast formations
Casing arrivals
Formation arrivals
Mud arrivals
Fig. 2,4
Formations in which sound waves travel typically faster than insteel (57
1.is/ft) are conventionally called fast formations (~4.5,§10.3). On the VDL log, fast formation arrivals will appear before the casing arrivals and override them (~9.4). 2.12 Micro-annulus Micro-annulus refers to a minute gap between the casing and the cement. Such a gap damages acoustic coupling between the casing and the cement although usually it does not permit fluid communication. The development of a micro-annulus and at what size it will invalidate the measurement of bond quality has not been fully understood (~4.1). A micro-annulus can make the CBL log look as if the casing was partially or completely unsupported. On the VDL log there will be strong casing arrivals as well as strong formation arrivals (~9.2).Once a micro-annulus has occurred, it is not possible to quantitatively estimate the bond conditions because the micro-annulus could mask coexistent channels. 2.13 Free pipe Free pipe is a section ofpipe which is not cemented. However, some engineers have been using the term to describe a log which appears as ifthe annulus were not cemented. Inthis case it does not necessarily mean that the annulus is free ofcement or indeed squeezable. See §9.1 on p 34 for a free pipe example.
CBL Evaluation Manual- QCand Interpretation
Page 7
Things you should know
~3. Information included in the log It is not imperative that logs from different service companies have the same format. But a complete log should have four parts: the header, the body, the summary or tail and the questionnaire. See example shown in §9.1. 3.1 The log header should include the following:
GeneralInformatlow log types, well name, log time and date, rig name and type, location,
-
log measurement base, log scale and run number Wellgeometrical data. deviation, depths and bit sizes of hole sections, depths and sizes and weights of casings sections, top and bottom of logged intervals
-
Wellfluids data: type, density
-
Cementing data: type, slurry densities, volumes, additives, retarders, starting and finishing pumping times, lab thickening and setting times, spacer type and density and volume, fluid loss volume
-
Wellpressure and temperature data: pressure applied afterbumping the plug, pressures applied at the time of logging, temperature profile
-
-
Logging equipment data. modules number, calibration status
-
Tool string sketch: centraliser types and positions
-
Logging engineer’s comments: record the aims of the logging, events which may have a bearing on the log and express views on the quality of the log
3.2 The body of the log should include the following where applicable: -
A “freepipe” reading sectiow record about two joints of pipe if available
-
The “main log”: record the main interval(s) of interest
-
The repeat sectiow record about 200m
The title ofeach of these sections should also include the pressure applied even if it was zero. All curve scales and legends should clearly 1 in thebetitle area.and correctly indicated. Less conventional mnemonics should be explained ‘Not In wide practiceyet If you witness logs, you can help speed up this process!
3.3 The log tall Should include summaries of tool operational status, software input parameters, calibration before survey (~9.1).
and tool
3.4 The BP questionnaire (“Log Quality Control Sheet”) Should be completed and signed by the logging engineer, and included as part of the hard copy log.
L
CBL Evaluation Manual- QC and interpretation
PageS
~4.Parameters affecting the Log Results Some key parameters are discussed here. Because most of their effects are not quantitatively known, it is important to understand the mechanism by which they affect the log, and assess which parameter would carry more weight than others in a given situation. 4.1 MIcro-annulus How it occurs: Not fully understood yet but a micro-annulus can be created either by casing contraction after the cement has set, orby casing expansion under high pressures that break the cement bond. Casing contraction can be thermal or mechanical. Casing expansion is usually caused by high pressure such as occurs during a squeeze job. The micro-annulus can therefore be classified into 3 types: thermal contraction, mechanical contraction and expansion. The thermal contraction type is due to the heat released from cement hydration. The casing contracts after the cement hasset and the heat has dispersed. This type ofmicro-annulus depends on the cement sheath thickness and composition, and the thermal conductivity ofthe formation. The mechanical contraction type is caused by reduction in pressure, for example, by changing casing fluid to a lighter one after the cement has set, or holding the pressure inside the casing before the cement has set and release it afterwards. The expansion type is usually caused by squeeze pressure that permanently damages the bond. When the pressure is released, only the casing resumes its previous size but not the cement. How it affects the log: When the cement is bonded to the casing, the acoustic energy is transmitted from the casing to the cement easily and is thus heavily attenuated, When a microannulus has developed, the energy transmission is severely hindered and a large proportion is trapped in the casing. (A gas filled micro-annulus is much worse than a liquid filled one in terms of energy transmission). The casing then rings relatively freely, producing strong casing arrivals on the VDL log. The El amplitude will be high, indicating that little bond exists. Particular problems with micro-annulus are: 1) It is not possible to distinguish a partially bonded annulus with a channel from a cemented annulus which can provide isolation but with a micro-annulus. 2) The effect of a micro-annulus can be so bad that the log may look like that the pipe is completely unsupported. This must have led to a good proportion ofthe failed squeeze jobs.
CBL Evaluation Manual- QC and interpretation
Page 9
Things you should know Becausethe CBLtool measures the bond between the cement and the casing, the micro-annulus that destroys this bond is the most severe factor that affects the log results.
How toprevent it: Obviously try not to create the conditions mentioned above under which a micro-annulus may occur. A method practised by some oil companies to prevent it from occurring is pump the cement wiper plug with a light fluid and change back to the weighted mud after the cement have set. Or even circulating the light fluid to cool the casing while the cement is setting. Because a micro-annulus does not usually permit communication and only affect the log, it is more important to eliminate it, if it has occurred, at the time of logging. This can be done by pressurising the casingusing a wireline packoffor sometimes using a heavy mud to increase the hydrostatic pressure (~5.2). Type ofmlcro-annulus Thermal contraction Mechanical contraction Expansion
Be prepared to pressure up to: l000psi Reduced pressure (hydrostatic or wellhead) +l000psi Max squeeze or hydrostatic pressure applied
Limited by~burst pressure of casing; Casing pressure test; Liner top test. * After the recent cementing. See 4.2 for pressure determination during logging.
4.2 Tool eccentridty How it affects the log: When the tool is off the casing centre, the acoustic energy from the transmitter will not reach the casing circumference simultaneously (~2.2).Instead part of the cased weilbore which is closer to the tool forms a shorter path for some of the energy to go through. Consequently this will cause reduction in the CBL amplitude as well as in the Yr as shown in Fig.4.1. The Yr has been used as a log quality indicator. Traditionally 4~.tsthe when the logUis reduction is less than accepted with an error of unknown magnitude. Recent research results tell us that the amplitude has a unique relationship with the amount ofeccentricity but it is a multi-value function of theYrreduction(~1O.5).The importance of these results are two fold: 1) When “minor” eccentricity (e.g. Ti’ reduction 4~ts)occurs the amplitude reduction caused by eccentricity can be compensated for.
r
CBL Evaluation Manual- QC and interpretation
Page 10
2) The fact that the amplitude reduction is not uniquely determined by a given TI reduction reveals the limitation of the U being used as a quality control indicator. How to prevent tool eccentricity: Not only sufficient number of centralisers are needed, but they must be put in the right positions. The following points should be observed: 1) Type ofcentralise?s: The rigid metal type seems to be the best. The rubber fin type is the second and the bow spring type is the worst. Worn centralisers can be weak and ineffective, but they can be checked by visual inspection. 2) Number of centralisers: In vertical wells use minimum of three centralisers and in deviated wells use minimum of five. Always request at least two extra ones for the job in case any faults develop in the mounted ones (~5.l). 3) Where to put them: For vertical wells put centralisers immediately above and below the transmitter-receiver section and on top oftool assembly (CCL or GR). Note the casing collar locator is not an adequate centraliser! Without a centraliser atthe top the CCL and GR section may act as a lever arm to promote eccentering problem. For deviated wells add a centraliser to the centre of each section which does not yet have a centraliser. Preferably always add an extra one at the near receiver which is used for CBL amplitude measurement. 4.3 ChannellIng How it occurs: When the combined conditions of cementing operation and down hole geometries are such that the cement cannot displace all the mud from the intended section of the annulus, pockets of mud may reside in the annulus and formmud channels (~7,~8). Another less recognized type of channel is the mud-cake channel due to filtration often occurring between the cement sheath and the reservoir formation. A channel may not be a problem if it does not communicate. However, you do not know until it does! How it affects the log: Ideally we want to detect any channels and like them to affect the log as much as possible so that we can identify them. Unfortunately only those channels which are immediately next to the casing have a strong bearing on the log. Others are more difficult to observe. This is because of the energy transmission mechanism, as discussed in ~4.1 and ~2.2. When a mudchannel occurs next to the casing,a large portion ofthe acoustic energy in the casing corresponding to the channel will not be transmitted to the formation. As a result more energy is returned to the receiver and the El value becomes higher (~6.6).For channels awayfrom the casing, however, this energy transmission mechanism is not presented in El but later in time, and is usually drowned in the complicated waveform.
CBL Evaluation Manual - QC and interpretation
Page 11
i nings you snouia KnOW A mud-cake channel can sometimes be identified from the VDL log when the cement/casing interface iswell bonded. Because the mud-cake channel provides veryweak coupling, not much energy could go into the formation and the majority would be absorbed by the mud-cake and the cement. As a result the VDL would show weak casing arrivals with little formation arrivals. 4.4 CasIng coating Some casings have a layer of coating such as epoxy. If this layer of epoxy is thick (e.g. >70 mils), it can affect the log in the sameway as a micro-annulus. Butpressurewhen logging will not affect the CBL.
4.5 Fast formations What formations are they: They are usually strongly packed hard formations such as limestone and dolomite (~10.3).The sound velocity in the formations is affected by the forces they are subjected to as well as their microscopic structure. Therefore the velocity may vary slightly in the same type of formation at different locations. How they affect the log: When fast formations are present, the sound wave in the formations is faster than that in the casing. The latter, however, isthe bond quality messenger. The real El is distorted, or even drowned in the formation arrivals. What is measured has therefore nothing to do with cement bond quality. Fast formations make it difficult to evaluate the cement job. How to detect them. Fast formation arrivals are easily seen on the VDL log (~9.4).The Yr will be shorter and the CBL amplitudes may be high. Remember that U reduction can also becaused by tool eccentricity. It is usually easy to tell fast formation from tool eccentricity by examining the VDL log, but it is difficult to see if the log is affected by the combination of the two. The formation arrivals on the VDL log can be confirmed by the open hole sonic log (~9.4). 4.6 Mudtype and conditions How the mudaffects the log: The mud (or other casing fluid) is the medium for the acoustic signal to go to the casing/cement/formation structure and come back. It does not distort the shape ofthe signal but affects the amplitude: any medium will attenuate the acoustic energy by scattering or absorbing. Different mud will have different attenuation rate which affects El amplitude. The sound velocity may also change with different mud conditions, thus affecting the U.
CBL Evaluation Manual- OC andinterpretation
Page 12
Mud parameters that affect the acoustic attenuation and sound velocity are complicated. The mud density and the sizes of particles contained in it are two major ones. The general trend is that the denser the mud, the less attenuative it is (~1o.7).In other words, in dense mud the measured amplitude will be higher than that in lightermud under the same weilbore conditions. Tiny gas bubbles in the mud can affect the log by increasing the TI’ and reducing the El unpredictably. 4.7 Temperature How it affects the log: There are two aspects of the temperature effect: firstly on physical conditions of the wellbore such as casing fluid density, and secondly on the tool. A good tool should be insensitive to temperature changes. The response of such a tool to the changes in weilbore conditions due to the well temperature profile should be stable and repeatable on the log. In most wells this response is small but in high temperature or high temperature gradient wells this can be noticeable. When temperature changes, the sensitivity ofthe transducers will change, and so will that ofthe electronics. The output of the tool will inevitably include some error.This error can be ofsteady state (when the tool is used to the new temperature) or transient (when the tool is not used to the new temperature yet), but usually both. The overall temperature effect cannot be quantified. In HPHT wells it is important to observe the logging time and the repeatability of the log since the log validity can be severely impaired under these circumstances. 4.8 CasIng diameter and Casing thickness How do they affecttbe log: In casings ofdifferent OD, the attenuation change is mainly caused by the different length of the mud path: the larger the casing, the lower the CBL amplitude. The amplitude decay rates also depend on the type of mud in the casing (~l0.7).The mud conditions and the temperature effect are closely linked and should be considered together. A general rule for commonly used casing size is that a millimetre increase in the thickness ofthe casing would cause about lmV increase in the “free pipe” CBL amplitude. The reason is that the thicker the casing the less acoustic energy would be transmitted to the cement, and the energy trapped inthe casing willbe more difficult to damp.Jointsofdifferent weights in one casing string should be seen from the U curves.
CBL Evaluation Manual - QC and interpretation
Page 13
Things you snoula KnOW 4.9 CasIng damage How does itoccur: Casing can beworn by drill-pipe, or splitby excessivepressure. Perforations of course blow holes in it. Corrosion can cause serious pittings in casing. How it affects the log: Casing wear or corrosion can cause tool eccentricity problem and thus reduce the El amplitude. Perforations can damage cement bond, especially for weak cement (compressive strength <2000psi), by cracking the cement. Split casing can cause the tool to become stuck. Severe irregularity in the casingwill be reflected in the log. For example, casing collars damage the continuous transmitting of the acoustic energy and this can be seen from both the U and the VDL log of light or less well-bonded cement.
4.10 CasIng standoff and Geometry ofthe open hole How do they affect the log: When the casing has a 0% standoff (i.e. touching the formation), there will be some formation arrivals coming through in the VDL log regardless of the cement conditions in the annulus. The contact between the formation and the casing may also deform the El, giving incorrect bond indications. This can undoubtedly mask channels. The geometry of the open hole would affect the log in two ways (~7).First, if the annulus thickness isless than 19mm (0.75”), the El might be deformed and increase because ofreflection from formation catching the first half of El. (Since this depends also on the cement and the formation, especially the cement/formation interface conditions, this thickness is only a rough guide). Secondly, in washout areas especially in deviated hole, mud removal is much more difficult. The thermal conditions in washout sections are different from the rest of the cemented annulus. This may add to the complication of micro-annulus generation and its detection. Small casing standoff can leave very thin cement in part of the annulus and cause the El deformation problem. Itis reported by Dowel Schlumberger that casing standoff from 100% to 0% could cause 30% increase in the El.
4.11 Double (or concentric) casing strings How do they affect the log: When the annulus created by one casing string inside another (eg. the liner overlap) is cemented, the log usually shows the following features to indicate a good job (~9.8):
CBL Evaluation Manual- QC and interpretation
Page
j
1) The U curves will be increased. In a 7” and 9 5/8” combination for example, the increase 20M~~ is typically 2) The first few cycles of the waveform shown on the VDL log will be narrower, i.e. gaps between the first few stripes are smaller than usual. Also, the apparent casing arrivals can be so strong that formation arrivals are masked. 3) The CBL amplitudes, which are not useful in such an interval, can be high. Arough explanation for these isthat the well-bonded concentric casinghas caused the frequency ofthe first few cycles to increase, while high attenuation has caused El to be skipped, resulting an increase in the U and measurement of E3 as amplitude output (~2.9). 4.12 Wait on cement (WOC) time How does it affect the log: Cement slurries take time to set and bond to the casing. Since El represents the bond quality of the cement to the casing, it willchange from free pipe value, before the slurry thickens, to bonded value, when the cement has set. WOC time should be the minimum time to wait before logging. Logs produced at marginal WOC times can be very tricky because the down hole conditions are not the same as in the lab and itmay well take a bit longer for the cement to set; Beside, the slurry can be contaminated which may stretch necessary WOC time (~7).For foamed cement, this is often complicated by the fact that some foam agents are retarders and are sensitive to contaminations. It is not possible to tell greenish cement from a channel. It is very important not to log the well too early: one hour saved may cost you ten! It is always good practice to monitor the timing of the logging and the slurry properties when interpreting a log (~4.l3). 4.13 Cement parameters/conditions (~7) Whatare they: The slurry density, cement class, additives types (retarders or accelerators) and weighting agents are all things that should be known. How do they affect the log: For neat cement slurry, the heavier it is, the higher the acoustic attenuation it will have when bonded to the casing. Other solids such as bentonite and silica added in the cement may increase the attenuation. The acoustic attenuation of foamed cement is lower than that of the neat cement, resulting in higher El values for 100% bonded casing. However, the CBL response to foamed cement has not been systematically studied and therefore is more difficult to interpret.
CBL Evaluaiion Manual - QC and Interpretation
Page 15
Things you should know 4.14 Computer keyboard operations What are they: Software parameter setting for communicationcontrol and log result formatting (~3,~9.1) How do they affect the log: It isthe engineer’s keyboard operations that make the toolcorrectly function and the data properly recorded. Apparently small errors on the keyboard can spoil all the hard work, and should not be underestimated. In the log presentation, small mistakes such as incomplete information, wrong legend or scale for curves, can cause frustration and time loss in the interpretation, or even big mistakes by inexperienced personnel,
CBLEvaluation Manual - OC and Interpretation
Page 16
0 I.
a C.)
0
~5. Operation QC In three phases Once a decision is made that a log is to be run, the operation QC should be carried out in the three phases as shown in the following chart:
To meet with the logging engineer to set clear objective & to supply sufficient information (~5.1)
To help logging engineer carry out a hi9h quality job by observing the correct procedure (~5.2) ___________________
To accept or reject the log (~5.3) ___________________
5.1 Before logging (Phase 7): Meet with the logging engineer and discuss: 1) The aims of the job (zonal isolation, finding TOC or else?) 2) Well conditions and history including (~3.1): -
-
-
-
Well ambient conditions (deviation, temperature, formation structure, gas or oil reservoir intervals, intervals of interest, any anomalies) Well fluid characteristics (OBM or WBM, weights, gas/solid contents and sizes, fluid change after cement job?) Casing characteristics (sizes, weights, depths) Pressure history (BOP test, casing test, formation test, fluid change). Discuss whether pressure will be needed during logging if pressure was applied to the well after the cement has set or pressure was maintained during the cement setting time.
3) Cement job (~4.13 cement type, density, special additives, estimate setting time, volume, pumping time and any problem in cementing operation). Make sure logging does not start within 8hrs after the lab cement setting time (~4.12,~9.8). 4) Choice of tools (~1),tool string configuration, centraliser types numbers and positions (~4.2), special/back-up equipment needed (such as wireline packoff) The purpose of this meeting is to set clear objectives for the logging engineer, and supply sufficient information for him orher to plan and prepare for the job. The meeting shouldbe held at the earliest possible time. The logging engineer should then produce a job plan which can help you take appropriate actions and plan ahead. The logging engineer’s plan should include: 1) The objective of the specific logging and the approach to achieve it. 2) Estimated time for logging operation, including a detailed plan of wireline rig-up and rigdown. 3) Indication of any further information needed in order to carry out the job successfully and to present the log completely (~3). CBLEvaluation Manual - QC and Interpretation
Page 17
QC Matters 5.2 DurIng logging (Phase 2)’ On the well Site we are to help the logging engineer to carryout a high quality job by observing the correct procedures. The logging engineer should have a detailed plan of when to do what, but he or she may well alter it as things change. A check lists 1) If the well was pressurised or the well fluid was changed not as originally planned, discuss with the logging engineer if pressure should be applied during logging to prevent the possible micro-annulus effect (~2. 12, ~4.1,~9.2).Ifwell head pressureis needed,the following is recommended to determine the pressure required: -
Run a 0 psi repeat section
-
Identify a potential micro-annulus zone (fom-iation signals behind casing signals)
-
Stop the logging tool at the micro-annulus depth
-
Switch the panel to time drive i.e. the screen display as if the tool was moving
-
Tighten the wireline packoff and start pressure-up (pump mud slowly into the casing)
-
Monitor the amplitude until it no longer drops
-
Stop pressure-up
-
Log under this pressure
2) The centralisers are the required type/size and in good condition. Also check that they are correctly mounted and secured in the required position (~4.2). 3) The tool used should be calibrated and the next calibration date has not expired. 4) The tool string is correctly connected and tested before being lowered into the well. Tool calibration before and after the logging run. 5) The parameters/constants Set from the keyboard are correct (j3). 6) Do not exceed the maximum logging speed. 7) The scales and markings are correctly set. 8) Monitorthe curves on the screen. Discussand recordany suspected problemwith the logging engineerand suggest to repeat abnormal sections. Ifthe TI’does not repeat and the difference in amplitude in the repeat section is >10%, it could be a centralization problem. Pull out of hole and change/add centralisers if necessary.
CBLEvaluation Manual - QC and interpretatlor
Page 18
5.3 After logging (Phase 3): To accept the log or otherwise. To present the log correctly and in a complete manner is important, not only for producing valic logs for immediate use, but also for documentation. The logs may be referred to after the wel has produced for several years. Missing information can be irretrievable and render the lo~ meaningless. A check list 1) All the signals have been correctly recorded. Authorise rig down. 2) The scales and legends used are correct (~9.1). 3) The log header, tail and the questionnaire etc. are fully completed with no incorrec’ information. 4) The logging engineer’s comments have included and explained all quality-related incidents the aims of the logging and his or her opinion on how well they were achieved.
5) The log hard copy is delivered on the time agreed. Authorise payment if there are no QC problems, otherwise raise them with the service company. 6) If any event occurred during the logging which may have affected the log, prepare a repor describing that event in detail. Attach a copy of the report to the hard copy of the log wher it has arrived.
CBLEvaluatIon Manual - OCand Interpretation
Page 19
0
a 0,
a. 4,
at
0 -J
~6.CBL Interpretation A systematic way to interpret a CBL log is demonstrated in the interpretation flow charts, Go to the references given if you are uncertain at any stage. You may have to break the main flow chart for special investigation which is shown on page 27, and resume afterwards.
INTERPRETATION FLOW CHART
CBLEvaluation Manual - QCand Interpretation
Page 20
6.1 QC Review If the log is properly documented, it should contain most ofthe information needed for correct interpretation, The majority of the key parts are included in the log header (~3).Check the following four items. Anything missing or incorrectly recorded will at least cause time loss or an invalid log. Item 1: Theplanning and execution of logging Clear objectives? Effective effort to prevent micro-annulus and eccentricity? Logging actively witnessed — witnessing engineer’s signatures and relevant reports? Item 2: Thefive data sets are complete and correct (~3. 1): 1) 2) 3) 4) 5)
General information Well geometry Well fluids Cementing operation Well pressure and temperature
Item 3:The logging engineer’scomments Clear and relevant in addressingany problems. The answers to the questionnaire can also imply his or her competency. Item 4: The log presentation Header-body-tail-questionnaire, signatures and date complete? 6.2 Quick check of ranges of the various curves 77 Curves: The normal ranges of U curves depend mainly on the casing ID and mud type/ weight (affecting the sound velocity). Temperature and the type of tool (the size of the transducers) also have some effect. A rough guide for calculating this range is TT(~is)—(casingID eg. In mm)/(sound velocity in the mud eg. In mn’z.~us §10.3)
+
170(Ms)
The last item is the distance of T-R spacing (3ft) times the sound speed in steel (57~ts/ft). CBL Curves: The normal ranges ofCBL for unbonded pipes depend mainly on the casing sizes (~1O.8)and mud type/weight (~10.7).The ranges for 100% bondedpipes are less reliable because of the difficulty in controlling the test conditions. For foamed cement there have not been sufficient field data or lab results for fully bounded pipes and the interpretation is therefore more difficult. VDL/Signature log: Check whether the time scale/range are correct (in the above equation replade 170 by 285, which corresponds to Sft), the log has good contrast, and there are any fast formations (~9.4).
CBL Evaluation Manual -00 and interpretation
Page 21
Log Interpretation 6.3 ExamIne the fl Curves The purpose of examining the U curves is to explain the curve variation, ifany, and investigate the log validity. The IT curves variation. The U curves from a properly centralised tool run in uniformly cemented pipes should ideally be straight lines in the expected region (~6.2).Parameters affectingtheU curves include centralisation (~4.2,§9.3), casingID size/weight (ID)changes (e.g. casing collars) (S4.8), casing fluids change over different depths (~4.6,§9.5), fast formations (~4.5, §9.4) and temperature (j4.7, §9.6). The U curves will also vary to indicate stretching, cycle skipping (j2.9) and well bonded double-casing string (~4.11,§9.7). A log validity criterion. A widely quoted log validity criterion in the literature and various manuals is that ifthe U curves vary for more than ±4iis,the log is invalid. Be careful. Variation of the U curves of a real log is rarely within this limit: investigate if the cause was eccentricity! When the tool is not properly centralised, the U can be shorter than normal. Unfortunately the amplitude will also be reduced. The effects of most other factors mentioned above on the amplitude are likely to be relatively small. Traditionally the log is treated as invalid because of the unknownreduction in the amplitude. This nowhas been better understood (~4.2,§10.5) and the amplitude reduction due to eccentricity can be calculated. However, because the uncertainty of the effects of other factors still exists, it is recommended that similar criterion be used with the U reduction limit being 5.us (~10.5). Note the transient temperature effect can make the log invalid (~4.7,§9.6) and so can fast formations and micro-annulus — the amplitudes in the interval of interest are not quantitatively reliable. 6.4 ExamIne the CBL Curves CBL curves: Look for Top of Cement (TOC) if applicable (e.g. a non-liner job), where the CBL curves swing from the low end of the value range to the high (~6.2,§9.1). Is the TOC in the expected region? Check with the annulus size and pumped _____ cement volume. A low TOC measured is often associatedwith slurry loss anda high one incomplete mud removal. Ifthe TOC is not found the slurry has been contaminated at least in the top section (~8). Look for good cement section where the curves are at the low end of the value range. Note ifa lead and tail slurry system was used, a difference should be seen in the good cement sections found, where the tail cement should give a lower reading. These will be useful in qualitatively evaluating the cement job. If no good cement section is found, the problem could be heavy contamination but refer to the Special Investigation Chart on page 27. CBL Evalaation Manual - QCand Interpretation
Lead cement
Tail cement
Page 22
Look for “Free Pipe” (uncemented pipe) section if applicable. Do not be misled by high readings of the CBL curves: check for the micro-annulus (~4.1)!The CBL values for free pipe offer references in cement quality evaluation (~6.6). 6.5 ExamIne the VDliSignature log The VDL log: must have good black and white contrast. It contains much information but now only a small portion is extracted, and this is done by visual examination. The main uses of the VDL log are: --___________ 1) to detect micro-annulus (~9.2) 2) to detect fast formation (~9.4) 3) to confirm free pipe (~9.1) 4) to confirm good bond in double casing strings (~9.7) 5) to confirm good bond to the casing but bad bond to the formation (where the casing arrivalsare extremely low with little or no formation arrivals and CBL amplitude indicates a good bond).
The signature lo& whichis often superimposed on the VDL log, produces the wave amplitude information which is not available on the VDL. This information can be useful in confirming changes in bond quality. However, the signature log is not as easy to use as the VDL in detecting, for example, the formation arrivals and that is why it is often combined with the VDL log. The indications of the U, the amplitude and the VDL logs must be in agreement and their examinations should be in parallel. Ifthey do not agree, there could be a tool problem resulting an invalid log.
L
CBL Evaluation Manual-00 and Interpretation
Page 23
6.6 Bond Percentage Index Calculation BFlDeflnitiow The percentage of the annulus where the cement is well bonded to the casing. The rest of the annulus (100% - BPI) is not well bonded, which may be contaminated cement, localised small gaps between the cement and casing, or a channel. Angle of bonded cement BPI (%)
-
Elf- Elm x 100%
-
The whole arinulus
Elf- Elc
Where the El’s are from the same receiver (e.g., the 3’ receiver), and the subscript m represents the value measured in the zone of interest, f represents free pipe value, and c the value for 100% cemented pipe. The BPI applies to any type ofcement system (neat, foam, etc.). Note when a lead and tail slurry system is used, El c should be selected separately for the two slurries. When Elfand El c are available, the corresponding BPI at a given Elm can be found from the above equation, or as shown in Fig.6.l below.
FIg.6.1 How to findthe BPI graphically
On the Elm axis, mark El~and Elfvalues. Mark the BPI (%) axis by equal intervals from 0 to 100. Draw a straight line from (Elc,lOO) to (EljçO). Given a Elm value, the corresponding BPI can be found as shown. To provide a sealwith high confidence, BPI need to be around 95% orhigher for certain lengths (~l0.6).For gas wells this rule should be applied more stringently.
CBL EvaluatIon Manual-00 end Interpretation
Page 24
Log Interpretation BPI Calculation Example Below is a section around the shoe ofa 7’, 29 lb/ft linerjob run atthe Gulf ofMexico. Assuming that the log has passed the QC and quick examinations, let us see how the BPI is calculated. Suppose the cementing operation and other well conditions allow us to believe that at least a section of
perfectcement job can be achieved, we can then select the lowest reading, 5mV, as 100% bonded value Elc. The “free pipe” value Elf is nonexistent(hopefully!) in a linerjob, we therefore look itup from §10.8 anduse 62mV (or we can use an available value from a log with close conditions as the job in hand). Between points Aand B the average reading of7mV gives the BPI-(62-7)/(62-5)-96.5%; BetweenB and C the average reading is about9mV. The effect of the slight U reduction (3—4~.ts)caused by eccentricity can be compensated for (~10.5),in this case by increase the amplitude by about 10% to lOmV. The corresponding BPI here is therefore (62-10)/(62-5)91%. If El~cannot be clearly def’med on the log, we can also use a reasonable value elsewhere, e.g. 2.4mV from §10.8. The BPI for the two intervals will be (62-7)/(62-2.4)92% and (62-l0)/(62-2.4)—87% respectively. As there exist about90% bonded intervals for30ft, the probability of zonal isolation is high. Notemore than 50%reductionin Elc(from5mVto 2.4mV) has changed the BPI valuefor <5%. Variations in Elf does not affect the BPI much either. This means that for conventional jobs the selection of the reference valuesis important but not critical for a reliable interpretation. Preferablyboth 0and Elfare decided from the log whenever possible. This method applies to all types of slurry design. CBL EvaluatIon Manuai-QC and Interpretation
Page 25
SPECIAL INVESTIGATION FLOW CHART~
* Either go to the next action or resume the main interpretation flow chart on previous page
C~L Evaluation Manual - QC and interpretation
Page 26
6.7 Finding Information If information in the log is incomplete, or other information is required such as open hole logs or some detail of cementing operation, contact the right person to obtain it. Keep an updated list of contacts for necessary help: their name, specialities, company, base, telephone and fax numbers etc.. 6.8 AnalysIng abnormal log behaviour 1) Review all factors which are likely to affect the log (~4). 2) Eliminate one by one around the key problem in hand until only such factors whose effects may be significant. 3) Judge whether these factors have invalidated the log — a good understanding of all the factors (study §4) may be very useful. 6.9 CalculatIng the probability ofzonal Isolation Cementing operation can offer valuable common sense evaluation which should not be discounted (~l.4).If any steps in the operation were altered from planned, it is suggested that the post-job CPS be run (~7).A starting point is to judge the likelihood of at least some sections which are channel free. Ifno such sections can be assured and the log shows a very bad cement job, a remedy job may have to be considered (~8). For the BPI calculation, the El value for 100% bond (Sl0.8) may have to be found elsewhere if that value for the given conditions is not available in the literature. The log of a well of close location with similar conditions can prove useful to provide this reference. The value can be chosen by finding the lowest reading at intervals where isolation was required and achieved. Make sure there were no fast formations or micro-annulus (~4. 5,~4.1). 6.10 Keep a record of Interpretation It is recommended that the interpretation of a log is recorded, eg, on the back of a log, for ease of later reference. The rubber stamp issued together with this manual is for this purpose. The table from the stamp is designed to summarise the log with minimal amount of information. Stamp the back of the log and fill in the form as you complete the interpretation.
CBL Evaluation Manual - QC and interpretation
Page 27
Log interpretation
~7. Cementing Operation 1. What to check~ Do not come to conclusions on the quality of the cement job without reviewing the cementing operation. Every step of the operation is carefully designedto ensure a good job. Problems may occur if any recommended step was not followed correctly. Common sense can often tell you the nature of the problem. Cement placement itself is a large and complex subject. But simplistically speaking, there are somerules ofthumb. For example, the more centralized the casing string, the better the cement job; Equally important are the displacement rates of mud, spacer and cement slurry. Usually the higher the flow rates, the better the displacement. The flow rates are limited by the formation strength and the displacement facilities. The rheology and density of the fluids are similarly important. The idea is to remove the viscous mud by spacer as completely as possible, then to displace the spacerby the cement slurry. The spacer here is designedto make easythe removal of mud and the placement of cement. Usually a heavier fluid following a lighter one can help the displacement. Remember, non-vertical wells are not only difficult to log, but also difficult to cement because ofthe casing centralisation problem and the complicated flow regimes in an inclined annulus. Without being deeply involved in cementing technology, you can always get some idea about the quality of the cement job by checking the following: -
Was there any lost circulation?
-
Was the casing string centralised?
-
Were the pumping rates and displacement timing as planned?
-
Were the well-site-measuredfluids properties (density, viscosity, additives quantities etc.) as designed?
Answer “No” to any of these questions could mean a less than perfect cement job. Ifthere were losses to the formation or other operational problem encountered during cement placement, it could be a bad job and log interpretation must take this into account. 2. The Cement Placement Simulator (CPS) This simulator, developed by the Fluid Mechanics Team at RCS, can tell you how to carry out the cement job and what the cement job should look like. The CPS is located at Dyce, Houston and Sunbury and wifi also be incorporated in DEAP. If a bad cement job is suspected it is suggested that the CPS be run. The result is another reference for interpreting the log.
CBL Evaluation Manual - QCand Interpretation
Page 28
:
Log Interpretation
~8 Squeeze considerations A bad cement job indicated by a bond log (whichever type) does not mean you can go ahead and squeeze. The following questions must be answered as best as you can before any action: Where did the cement go? What kind of channel could it be, and where and how to squeeze? The following Charts may help you answer these questions.
1. Where did thecement go? Analyse the well and cementing conditions together with the log Not clearly~~d?~
62~
Yes
Check if any fluid loss occured duririu d~IIing or cementing(~6) _____________________
__________________________
Possible cement loss by large quantities. Such cases are usually easily identifiable.
Possible heavy contamination or bad slurry leading to green cement. Re-run the CBL if possible.
CBL Evaluation Manual - QC and interpretation
Calculate the difference from expected value (Note the hole gauge, washout: caliper log).
~
The cement is likely to be still In the annulus but badly bonded tothe casing and maybe to the formation as well.
Yes
Possible bad contamination at the cement top or bad mud removal.
Page 29
6 scenarios of communication channels Eccentred casing, mud channel on the narrow side. Heavily contaminated cement which may or may not be solid. Contaminated but solid cement on the wide side with mud channel on the narrow side. Contaminated but soiid cement on the narrow side with mud channei on the wide side. Gap between the set cement and the casing Thick mud cake between the set cement and the formation.
key: mud cement contaminated cement formation
Well conditions and cementing operation vs potential_channels Channel type
Bad casingcentralisation
Deviated wells
Displacing contamination
Cementing operation problems
I, I,
________________
Severe in high deviation
Washout section
Often in wail. horizontal ~~1
Some “delayed” communications observed in porous reservoirs are believed to be caused by the disintegration of the mud cake. This type of channel is hardly detectable with today’s technology
Bond logs provide vItal information for squeeze job design particularly in the following areas: 1. Depths,lengths of communicating channels for positioning the perforating gun and bridge plug or packers. 2. Azimuth of communicating channels for perforating shot phase arrangement: a 45 degree channel can be missedi 3. Identify the vent for the channel filling substances.
L~~Evaiuatlon Manual~QC and interpretation
Page 30
at at U C
at at at
C at at
C.
E ‘C
9.1 A Log Example Header [on this page], body and tail [on next page], and questionaire (~3)
~ion~cC~
Log Examples
CBL Eva’uation Manual - OC andInterpretation
Page 32
CBL Evaluation Manual - QC and Interpretation
Page 33
Log Examples RUN1
RUN2
~ AMPLITUDE
aEFORE PRESSURING
AFTER PRESSURIZIFICJ TO3000FSI& RELEASING
AMPLITUDE *
9.2MIcro.annukis F~.ashs1om~rho1ding5Op~onthecasing~r3daysRun1wasmnwithOpsiandRun2wsswith7OOpsi.Adequate pxessuxe cen eliminate some miao-annulus. Fig.bwlogsninbefoieandasngwsspiessuzeiisedto the casing and thecement TT
(US
)~
aUO.00
3OOO~l.Excessivepressuxecanbreakthebondbetween
___L_~M~___~_.
5~._OOO
100.00 1200.0Q
VDL
(US
1200.0
9.3 ToolEccentricity Eccentrldty causes reductionintheTrand intheCBLamplitude (~1O.5).Addan exn~centmliser aithe 3’ receivercan alleviate the problem(~6.2).Never run the CBL knowinglyoff-centred (eg with 711 rubber centraliser for liner, running in95/8” casing above the liner). CBL Evaluation Manual - OC and Interpretation
Page 34
TRAVEL TIME
290 •
T~A/
~
190 —— -
API 0
100 CC
9.4 Fast Formations
ATTENUATION OS/IT 20 0 — — PEAK AMPLITUDE — — MV 0 120200 AMI’LIIIED
‘LIGNATUITE
VARIABI E DENSITY 1200 200
‘200
R
The TT fell below the expected value (~5.2).The VDL and Signature logs also showed that formation arrivals overtook the casing arrivals. The formation arrivals can be confirmed by the open hole sonic log: the delta-T curve should closely mirror the formation arrivals on the VDL, as shown below.
Quoted from “Cement Evaluation” Bigelow E L, Western Atlas International, Houston, 1990 CBL Evaluation Manual - CC and Interpretation
Page 35
Log Examples
9.5 CasIng Fluids Effect on the Log The mud weight changed from 1O.5ppg above 7028ft to 16.6 ppg below. The effect on the rr is obvious butnot so on the CBL amplitude. The well temperature profile canmake a difference to the mud (eg in density) if the mud has been static in the hole for somtime, This can cause the iT to vary (~6.2). Quoted from DS Cement Bond Logging Field Reference Manuai. CBL EvaluatIon Manual - CC andInterpretation
Page 36
9.6 Temperature
Effect on the Log
The temperature in the section shown was about 320°Fand 300m below was about 345°F.The tool was run to TD in 1.5hrs, experiencing high temperature gradiant, and the logging was finished in 2hrs. The tool was properly centralised but the CBL amplitude and the VDL were affected by transient temperature effect.
9.7 Double Casing String Below D is 7” (291b/ft) inside 9 5/8” (561b/ft) and the annulus was fully cemented. The TT has increased by 20its and the VDL showed strong “pipe ring” with first few cycles being narrower. The CBL amplitude here may be E3 (~2.5,~6.11). CBLEvaluation Manual - QC and Interpretation
Page 37
Log Examples
TIME 4 HRS.
AFTER CEMENTATION
a
HRS.
ze
HRS.
9.8 Green Cement Remember that the down hole conditions are not as ideal as in laboratory. Cement slurry can easily get contaminated by mud, spacer orformation fluids. The thickeningtime and setting time can be much longer than quoted (S4.1, §6.12,13).
CBL Evaluation Manual - QC and Interpretation
Page 38
Parameters Original design purpose Calibration conditions Casing sizes Mud density Micro-annulus Heavily contaminated cement Cement/formation bond Eccentralisation Vertical resolution Azimuthal resolution Shape of channel (on casing/cement interface) Cement thickness Compressive strength Double casing string Casing stand-off Casing thickness Casing wall conditions Mechanical complexity Operational robustness Costs
CBL
CET
USI
cement bond down hole
bond/corrosion down hole
4.5” - 13 3/8” high(eg>l8ppg) unquanfifted effect
4.5” -95/8” OBM
4.5”- 133/8” OBM
severe effect
less 9effect (liquid filled) could indicate as channel” not measured less sensifive 3ins 45 degrees detectable
less effect (liquid filled) could indicate as channel” avoided less sensifive 1.Sins 10 degrees good measurement
less effect derived less effect less effect measurable strong effect
less effect idea abandoned less effect less effect measured detectable
complex easy to get wrong
easy to damage
typically 2xCBL
typically 4xCBL
openhole uphole
amplitudeattenuated can be detected sensitive 3ft non-directonal no effect values debatable derived (un~usfiflably) severe effect strong effect strong effect some effect simplest OK relatively low
good?
*This is because the acoushc impedance of heavily contaminated cementis in a similar range as that of mud.
10.1 Performance Comparison of CBL, CET and US! CBL Evaluation Manual - QC and Interpretation
Page 39
Data and Charts for Reference
1 _______ I 10.2 CharacterIstics of Commonly Used CBL Tools (quoted from “CBL Field Referenc Manual” DS) CBL Evaluation Manual - OCand Interpretation
Page 40
SOUND VELOCITY MATERIAL TYPE
POROSLfl
(%)
Casing Dolomite Limestone ~ Calsite
-
~ Anhydrite
-
Z Gypsum
-
~
Quarts —
Halite
—
Dolomite Limestone ~ Sandstone ~ Shale —
~
-
5 to 20 5 to 20 5 to 20
OBM WBM Water Water + 10% NaC1 Water + 20% NaC1 Sea Water Kerosene Air (15 psi, 0°C) Air (3000 psi, 100°C)
-
AT (Ms/ft)
(ft/a)
(rn/a)
57.0 43.5 47.6 49.7 50.0 52.6 52.9 66.6
17,500 23,000 21,000 20,100 20,000 19,000 18,900 15,000
5,334 7,010 6,400 6,126 6,096 5,791 5,760 4,572
50.0 to 66.6 20000 to 15000 6096 54.0 to 76.9 18500 to 13000 5639 62.5 to 86.9 16000 to 11500 4877 58.8 to 143 17000 to 7000 5181 209 to 222 198 to 205 208 192 182 199 230 920 780
4785 to 4505 5050 to 4878 4800 5200 5500 5020 4340 1088 1280
ACOUSTEC IMPEDANCE (MRayl) 41.6 20.2 17.3 16.6 18.2 13.6 15.2
9.3
to 4752 to 3962 to 3505
to 2133
1458 to 1373 1539 to 1487 1463 1585 1676 1531 1324 331 390
17,0 to 11.5 14.8 to 9.4 12.6 to 8.2 12.0 to 4.3 1.7 to 2.7 1.8 to 3.0 1.5 1.7 1.8 1.6 1.1 0,0004 0.1
10.3 Speed of Sound In Various Types of Formation and Mud (edited from “Well Cementing”, Dowel Schlumberger, 1990)
10.4 CasIng Expansion Under Pressure (quoted from “Cement Evaluations” Atlas, 1990) CBL Evaluation Manual - QC and interpretation
_______________ Page 41
Data and Charts for Reference
Notes:
Eccentricity vs Relative El
Eccentricity vs IT reduction
IT reduction affected by cement
TI reduction vs reIativ~ El
1) Cement does not affect the relationship in (a). 2) Mud affects relationships in (a) (b) and (c) but less in (d). 3) Cement affects amount of Yr reduction as shown in (c), consequently Yr-El relationship is not unique. 4) Free pipe U-El relationship in (d) can be used for El conpensation.
10.5 Relationship among Tool Eccentricity, CBL Amplitude and Tr Reduction CBL Evaluation Manual - OCand interpretation
Page 42
Data and Charts for Reference For BPI95%
10.6 Interval Lengths Required for Isolation
10.7 Amplitude Compensation Charts for Various Muds, Casing ID and Thicknesses (quoted from “The Fluid-Compensated Cement Bond Log” Nayfeh et al, SPE Formation Evaluation, pp335-341, 1986) CBL Evaluation Manual - OCaid Interpretation
Page 43
Casing size (in)
weight (Ib)
4 1/2
9.5 11.6 13.6
5
15.0 18.0 21.0
51/2
15.5 17.0 20.0 23.0
7
23.0 26.0 29.0 32.0 35.0 38.0 40.0
75/8
26.4 29.7 33.7 39.0
95/8
40.0 43.5 47.0 53.5
10 3/4
40.5 45.5 48.0 51.0 54.0 55.5
Free pipe Elf(mV)
100% bonded (class H) E1 0(mV)
The data above were derived from a chart by Pardue et al [Cement Bond Logging - A Study of Cement andCasing Variables,JPT May 1963, pp545-555]. Note they can onlybe used asa general guide.
10.8 CBL 3’ ReceIver El Readings for 100% Bonded and 0% Bonded Pipes CBLEvaluation Manual- QC and Interpretation
Page 44
Index Amplitude After logging Before logging Body, log BPI, Bond Percentage Index Casing, arrivals Casing, coating CCL Cementing Centralisers CET Channel Check list CPS Criteria, tool selection Cycle skipping Density Diameter, casing Double casing string During logging El, E2, E3, etc Eccentridty,centrallsation Examples, logs Fast formations Features, tool Five data sets Foam cement Formation arrivals Free pipe Gate, fixed and sliding Geometry, open hole Green Cement Header, log Interpretation Isolation Jargon, explained Keyboard operation Manual structure Micro-annulus Mud arrivals OBM Peak
4,5,18,22 19 17 8,32 24 7 12 5,11 3,17,11,28 8,18,11 3,39 3,5,7,9,11,24,30 18,19 28 2,3 6,22 8,13,15,17,28,36 13 14,37 18 6 5,10,34 31-38 7,23,35 4 21 15,21,24 7 7 6 14 15,38 8,31 20 2,9,20,43 5-7 16
Pressure Principles Questionnaire Review, QC Signature log Sonde Squeeze job Standoff Stretching Tail, log Temperature Test: BOP, casing, formation Thickness TOC TT, Transit Time USI VDL WBM WOC
2,8,9, 10, 17,18,21 4,5 8,19,21,33 21 4,5,23 4 29 14 6,22 8,31 8,13,17,21,22,37 10,17 10,13,14 17,22,28 6,21,22 39 3,5,7 17,41 15,38
2,7,9,18,22,23 7 17,41 6,7
CBL Evaluation Manual - QC and interpretation
Page 45
