Downhole Risks

Downhole EnvironmentalRisks Associated withDrilling and Well CompletionPractices in theCooper/Eromanga Basins

Damien Mavroudis

March 2001

Report Book 2001/00009

2

© Department of Primary Industries and Resources South AustraliaThis report is subject to copyright. Apart from fair dealing for the purposes of study, research, criticism or review as permitted underthe Copyright Act, no part may be reproduced without written permission of the Chief Executive of Primary Industries andResources South Australia.

3

CONTENTS

Acknowledgments.............................................................................................................................................6

Abstract .............................................................................................................................................................7

Executive summary ..........................................................................................................................................7Drilling fluid impacts ....................................................................................................................................7Effectiveness of cementing for zonal isolation .............................................................................................7

Section One: Drilling fluids .............................................................................................................................8Introduction ...................................................................................................................................................8

Mechanisms of drilling fluid filtration.....................................................................................................8Filtration control materials and techniques..............................................................................................9

Bacterial contamination ..............................................................................................................................10Drilling fluid contaminants .........................................................................................................................12

Mud recap reports ..................................................................................................................................12Clay swelling..........................................................................................................................................15

Conclusion...................................................................................................................................................15Recommendations .......................................................................................................................................16

Section Two: Cements....................................................................................................................................17Introduction .................................................................................................................................................17

Annular gas migration............................................................................................................................17Purpose of cementing..................................................................................................................................18

Primary cementing .................................................................................................................................18Problems arising from poor primary cementing ....................................................................................18Applications of API Cementing .............................................................................................................20Current cementing practices...................................................................................................................21

Cement failure mechanisms ........................................................................................................................21Carbon dioxide effects on cement in well..............................................................................................21Migration of gas through cement pore structure....................................................................................22Casing problem as part of the downhole risk study...............................................................................23High temperature chemistry of Portland cement ...................................................................................25Case history............................................................................................................................................27Microannular formation in cements.......................................................................................................27Mechanism of shrinkage and expansion ................................................................................................27Long-term leaking oil wells ...................................................................................................................28Mud cake removal for cementing job ....................................................................................................28

Remedial cementing....................................................................................................................................29Applications of the squeeze cement job.................................................................................................30Cement plugs..........................................................................................................................................32Reasons for cement plug failure.............................................................................................................32

Testing the quality of cement jobs ..............................................................................................................34Hydraulic testing ....................................................................................................................................34Temperature logging ..............................................................................................................................35Communication tester ............................................................................................................................35Noise logging .........................................................................................................................................35Acoustic logging ....................................................................................................................................35Cement bond log ....................................................................................................................................35Limitations of cement bond log .............................................................................................................39Cement evaluation tool ..........................................................................................................................39Acoustic properties of cement ...............................................................................................................39Channels.................................................................................................................................................41Fast formations.......................................................................................................................................41

4

Preventative techniques...............................................................................................................................41Prevention of gas migration ...................................................................................................................41Use of foamed cement............................................................................................................................41Foamed cement limitations ....................................................................................................................43Displacement properties of foamed cement...........................................................................................43Use of flexible cements..........................................................................................................................43Comparison of foam and flexible cements ............................................................................................43Durable zonal isolation (new cements)..................................................................................................44Downhole corrosion prevention.............................................................................................................45Corrosive agents.....................................................................................................................................45Prevention methods................................................................................................................................46Corrosion testing....................................................................................................................................46

Conclusion...................................................................................................................................................47Monitoring well integrity.......................................................................................................................47Cement time-scale ..................................................................................................................................48Fluid loss ................................................................................................................................................48

Recommendations .......................................................................................................................................48

Appendix 1 ......................................................................................................................................................50

Appendix 2 ......................................................................................................................................................53

Appendix 3 ......................................................................................................................................................55

Appendix 4 ......................................................................................................................................................56

Appendix 5 ......................................................................................................................................................61

Appendix 6 ......................................................................................................................................................64

Appendix 7 ......................................................................................................................................................66

Appendix 8 ......................................................................................................................................................67

References .......................................................................................................................................................68

FIGURES1 The effect of drilling fluid invasion on a permeable formation...............................................................92 Structure of the CMC polymer molecule ...............................................................................................103 Effect of biocide treatment.....................................................................................................................11

4a Downhole drilling fluid loss as a function of depth...............................................................................134b Downhole drilling fluid loss as a function of depth...............................................................................134c Downhole drilling fluid loss as a function of depth...............................................................................144d Downhole drilling fluid loss as a function of depth...............................................................................14

5 Objectives of primary cementing technique ..........................................................................................176 Mechanism for annular gas migration ...................................................................................................187 Common one-stage primary cement job on a surface casing string.......................................................198 Cement bond log response at time of cement placement and some time after ......................................229 Slurry dynamics immediately after placement.......................................................................................24

10 Della 1 cross section ..............................................................................................................................2411 Compressive strength and permeability behaviour of neat Portland cement at 446°F ..........................2612 The accumulation of debris in well 14, SW Pannoniam Basin in Croatia.............................................2613 Defective primary cementing job...........................................................................................................3114 The mechanisms for cement cake build up in borehole.........................................................................31

5

15 Circulation of cement in squeeze cementing .........................................................................................3216 The use of cement plugs for zonal isolation in the abandonment of a well...........................................3317 The use of a cement plug to prevent fluid loss to a thief zone...............................................................3318 Dry test expectations and results............................................................................................................3619 Typical temperature survey showing the probable cement top..............................................................3620 Temperature composite profile log before cement squeeze...................................................................3721 The configuration of a normal CBL tool run in the hole .......................................................................3822 CBL interpretation chart ........................................................................................................................3823 CBL energy transmission as a function of microannulus wavelength...................................................4024 Sonic wave paths....................................................................................................................................4025 Facilities for the generation of foamed cements ....................................................................................4226 Life of the cement sheath for the three primary types of cements used ................................................4427 Typical downhole arrangement of continuous corrosion inhibitor........................................................47A1 Relative static and dynamic filtration in the bore hole ..........................................................................51

6

ACKNOWLEDGMENTSI thank my parents, George and Katarina, for their support and encouragement during the research,preparation and writing of this report. I acknowledge my summer vacation employer the Department ofPrimary Industries and Resources South Australia (PIRSA) and Supervisor, Michael Malavazos in thePetroleum Group of PIRSA for providing me with the opportunity and encouragement to undertake thisproject. His constructive guidance and comments throughout this project have been most appreciated.

Damien Mavroudis2 March 2001

7

REPORT BOOK 2001/0009

DOWNHOLE ENVIRONMENTAL RISKS ASSOCIATEDWITH DRILLING AND WELL COMPLETION PRACTICESIN THE COOPER/EROMANGA BASINSDamien Mavroudis

This project reviewed literature in order to:• evaluate any downhole environmental risks associated with drilling fluids in the Cooper and Eromanga

Basins; and• assess the effectiveness of cementing practices in achieving long term zonal isolation between reservoir

formations penetrated by a well in these basins.

It was found that any potential drilling fluid contamination of Cooper and Eromanga Basin aquifers, inparticular the Great Artesian Basin, is not a major concern.

Cementing and completion practices in the two basins are the main risks to the downhole environment. Manymechanisms are present to cause the cement to deteriorate. As a result, sufficient zonal isolation cannot beguaranteed for an infinite amount of time. The major risk associated with cement failure is cement carbonation.

A system employing the use of logging equipment was devised in order to evaluate whether the cement wasmeeting the criteria it was designed to achieve. Examples of the criteria that can be used to evaluate the integrityof the cement are postulated in this report.

EXECUTIVE SUMMARYThis project reviewed literature in order to:• evaluate any downhole environmental risks

associated with drilling fluids and completionpractices in the Cooper and Eromanga Basins;and

• assess the effectiveness of cementingpractices in achieving long term zonalisolation between reservoir formationspenetrated by a well in these basins.

DRILLING FLUID IMPACTSAlthough an area of concern, drilling fluid impactwas found to constitute only a small portion ofdownhole problems. The use of drilling fluids wasconsidered to be an issue because of the potentialto invade freshwater aquifers, particularly thoseof the Great Artesian Basin (GAB).

In order to determine the significance of fluid lossto the formations, mud recap reports werereviewed. These reports record drilling fluidlosses and provide an estimate of the depth andformation where this occurs. Other potentialimpacts of drilling fluids include:

• microbial contamination in aquifers;• contamination from biocides used to control

microbial activity;• poor mud cake removal.

From the review of the mud recap reports, themost significant areas of fluid loss were in thefirst 4000 ft of the well. The most likely reasonsfor these losses were: unconsolidated Sands; andClay Swelling.

In any event it was determined that due to thepresence of the Bulldog Shale and its lowpermeability, an effective seal would be present toprevent the migration of drilling fluids into moreenvironmentally significant formations. Althoughin deeper wells the fluid losses were difficult toquantify, it was concluded that the majority of thefluid loss was most likely to occur in to theproducing zones of the formation. The mainconsequence of this is reservoir formation damagerather than an irreversible contamination of theaquifer. Once brought ‘on-line’ a well willproduce the majority of drilling fluid lost to theformation.

8

EFFECTIVENESS OF CEMENTINGFOR ZONAL ISOLATIONThe major area of concern was the effectivenessof cements in achieving long-term isolation in thewellbore of individual formations penetrated bythe well. Cements were identified as crucialbecause they are currently the only means forzonal isolation in the wellbore. Zonal isolation isdeemed necessary in the wellbore because itprovides a means for the prevention of cross-flowthrough the wellbore. Such cross-flow isconsidered to be a major risk for aquifercontamination.

The effectiveness of cement zonal isolation in thewellbore was reviewed by investigating thepotential mechanisms of cement failure. Themechanisms identified were:• high temperature;• sour conditions/sweet conditions;• bacterial presence;• cement shrinkage;• formation damage;• poor mud cake removal;• high cement permeability; and• cement carbonation (chemical reactions).

The findings of this investigation concluded:• cement carbonation and deterioration due to

hostile environments was the majormechanism for cement failure;

• it is necessary to establish a cement time-scale for which the cement should continue toprovide zonal isolation to the formationsisolated in the wellbore;

• current cement technology may not be able toprovide long term zonal isolation and newtechnologies need to be considered; and

• wells drilled through a successful drillingprogram will not always be accompanied by acompetent cementing job.

SECTION ONE: DRILLINGFLUIDSINTRODUCTIONSelection of a drilling fluid is a major componentin the drilling of a well. The success of a welloften depends on the performance of a drillingfluid (Darley and Gray, 1991).

As part of examining the downhole environmentalrisks associated with drilling practices in theCooper and Eromanga Basins, consideration wasgiven to the potential impact of drilling fluids.

Apart from the potential to destroy the permeablezones, the real concern regarding drilling fluidswas the possible contamination of undergroundreservoir formations, particularly freshwateraquifers such as the GAB in the Cooper andEromanga Basins. Analysis of drilling fluidimpact involved an investigation into the:• filtration properties that contribute to fluid

loss;• zones in the wellbore where an increased rate

of fluid loss may be present;• rheology of the drilling fluids;• study on the effects of different drilling

fluids;• typical mud composition; and• permeability of the wellbore environment.

It was realised at the outset of this project that anamount of drilling fluid is expected to be lost tothe formation. In fact, it is good practice that a‘controllable’ amount of drilling fluid be lost tothe formation because it enables the formation ofa mud cake to prevent excessive fluid losses to theformation. The principal concern in terms ofdrilling fluids was to determine where the majorfluid losses occur. That is, in the zones wherefreshwater aquifers may be present.

In order to determine where fluid loss occurs, itwas necessary to review mud recap reports.Formations with higher permeability's areexpected to lose a greater amount of drillingfluid.

Mechanisms of drilling fluid filtrationThe potential impact of the drilling fluids loststems from their uncontrolled invasion intoreservoirs penetrated by the well. In particular,from an environmental point of view, the fluidshave the potential to contaminate any aquiferpresent. Drilling fluid invasion can occur throughthree main mechanisms: static, bit and dynamicfiltration (see Appendix 1).

Of these mechanisms, dynamic filtration has beenidentified as the major means of drilling fluidinvasion. Its effect on a permeable formation isillustrated in Figure 1. If well integrity ismaintained, fluid invasion will be confined to asmall region in the formation. In cases where wellintegrity is lost, a significant invasion into theformation may occur.

9

Figure 1 The effect of drilling fluid invasion on a permeable formation (Darley and Gray, 1991)

Filtration into the formation occurs in three steps:• An external mud cake attached to the walls of

the borehole.• An internal mud cake, about 2–3 grain

diameters into the permeable formation.• A zone that has been invaded by smaller

particles during the mud spurt period. Thisinvasion zone can typically extend about aninch into the formation. This zone is typicallydamaged because of pore blocking by thefiner invasive particles. This invasion makesthe recovery of hydrocarbons very difficultdue to the reduction of the permeability nearthe borehole.

The mud cake will only form once a primaryblockage of the pore spaces has begun. This thengives the finer particles a basis to form a mudcake. Particle size distribution has a veryimportant role in forming a low-permeabilityfilter cake. If there are too many large particlesthen the bridge will build too quickly and willform a filter cake that is shallow and thin. Thesubsequent filtration loss will be high. This isknown as filtrate damage. If the particledistribution is too small then the bridge will notbuild quickly enough and the mud solids willpenetrate deep into the formation and causedamage.

Filtration control materials andtechniquesWhere excessive fluid losses are likely to occur,material can be used to minimise fluid loss andmaintain wellbore integrity. These materials assist

in the formation of a mud cake under the dynamicfiltration mechanism. There are several colloidalmaterials available that can be used to controlfluid loss properties of water-based mud. Thefollowing products are the most commonly usedin drilling applications.

Lamellar

Lamellar material controls filtration by aligningthemselves with the normal flow of mud into theformations. This helps to produce highlycompressible filter cakes. Bentonite is an exampleof a lamellar material.

Fibrous

Fibrous material is squeezed into the formationwhere it helps to block the formation and createbottlenecks. This material tends to produce deeppenetration and can be difficult to remove. Filtercakes from such materials are incompressible.Attapulgite is an example of a fibrous material.

Granular

Granular materials enter the formation and blockany pores smaller than three times their size. Thistends to cause shallow solid penetration but,because of their rigid structure, can also allow fordeep filtrate invasion. They tend to form ratherincompressible filter cakes. Barite is an exampleof such a material.

Emulsion

Emulsions tend to control filtration by enteringpores and causing increases in capillary pressuresfor fluids to enter. These types of materials do not

10

form filter cakes. For example, an oil emulsionwill decrease permeability to water withoutaffecting permeability to oil.

As mentioned, mud additives provide protectionagainst water loss through three basicmechanisms: binding of free water, blockingpores and forming a tight filter cake. Thecarboxymethyl cellulose (CMC) polymer as wellas bentonite has the ability to chemically bindwater to the polar sites on the clay platelets or tothe polymer molecules and form a tightimpermeable layer. Mud additives are effectivebecause they bond all of the free water and makeit difficult for the water to escape from thedrilling mud. By binding the water the viscosityof the mud also increases and the mud becomesmore resistant to flow into the porous formation.The benefit of using bentonite and CMC is thatboth of these substances have the ability to buildan impermeable membrane over the porousformation.

BACTERIAL CONTAMINATIONThe communication between surface andsubsurface oilfield environments is, of course,initiated by the drilling process. Drilling requiresthe circulation of fluids from the surface to the bitto help carry cuttings out of the borehole and tocontrol formation pressures in the borehole. Inthis process, chemicals and microbes from thesurface are circulated into the deep subsurfaceenergy-rich oil-bearing strata and hydrocarbon-laden cuttings are brought into the oxygen-richmoderate temperature surface environment.Through this mechanical process microbiologicalactivities can be initiated in surface andsubsurface environments. This does not occurnormally and can lead to the bacterialcontamination of aquifers.

Water-based drilling fluids often contain organicpolymers which act as viscosifiers and fluid losscontrol agents. CMC is one example (Fig. 2).These organic polymers, which tend to be of plantor microbiological origin, can be degraded andused as a food source for the growth of naturallyoccurring oil-field bacteria. This can occurdespite the addition of biocide materials for thehampering of microbial activity. Microbialgrowth in the mud can result in contamination ofthe well and near-wellbore zone (Ezzat et al.,1997). Fouling, corrosion and reservoir souringmay then occur during subsequent operations. Ifbacterial growth is extensive, significantconsumption of the organic polymers can occur

and may result in a loss in the rheologicalproperties of the mud.

Figure 2 Structure of the CMC polymer molecule(Darley and Gray, 1991)

Microbial activity in drilling muds can beinfluenced by several features in the drillingenvironment. Warm temperatures and highnutrient content in the mud tanks can haveundesirable effects in which bacterial growth isenhanced. The process of removing the drillcuttings (shale shakers) can further increase theoxygen content of mud, which is favourable forthe growth of bacteria.

Ezzat et al. (1997) identified bacterialcontamination as causing the following problems:• microbiological corrosion of well tubulars

and screens;• biomass plugging in injection wells and in the

formation; and• hydrogen sulphide production deep in the

formation, leading to reservoir souring.

Apart from this, the potentially hazardous natureof the bacteria if they contaminate a freshwateraquifer needs to be addressed. This will beconsidered in a further investigation for theproject.

Given the additives present in the drilling muds,microbial activity is significant due to thepresence of xanthan gums, starch, CMC,hydroxyethyl-cellulose etc. Biodegradation ofdrilling mud additives results in significantmicrobial growth within the mud. This can raisethe bacteria to a level that may be harmful. Thisincrease is known to affect the wellboreadversely.

Even with a limited amount of fluid loss duringdrilling, the bacteria would accumulate in the nearwellbore zone. Microbial activity would continueduring ‘shut-in’ periods and would be supported

11

by the soluble, carbon-based nutrients from thedrilling mud. Under oxygen-depleted conditionsthe activity of anaerobic sulphate-reducingbacteria would increase, which might have a largeeffect on the wellbore.

It is important to consider the long-term effectsthat bacteria may have on the wellboreenvironment. If left untreated, it is possible thatthe microbial activity may cause a breakdown ofthe downhole integrity.

Bacteria can invade the formation in the closevicinity of the wellbore. Bacteria left here canresult in a poor cementing job due to microbialactivity. Thus there can be an adverse affect onthe wellbore characteristics and the purpose of theprimary cementing job would have been defeated.Therefore it is important to consider using mudthat does not promote the increased activity ofbacteria. Such drilling fluids employ high salinity,high pH and biocides to curb the microbialactivity.

The major concern with the presence of bacteriain the drilling fluid is with the potentialcontamination of freshwater aquifer supplies thatcould be consumed by humans or livestock. The

possible migration of water-borne bacteria shouldbe considered in detail. A study of the methodsfor controlling bacteria should also be considered,with the intention of analysing the validity ofbacteria control techniques for water treatment.The presence of bacteria can also be hazardous tothe casing of a well through the introduction ofsulphate-reducing bacteria. Hydrogen sulphidecan cause corrosion of the casing (metalimbrittlement) (Ezzat et al., 1997). The effect ofbacteria in freshwater aquifers used foragricultural use or consumption is not yet known.

The use of bactericides for controlling bacteriashould also be investigated. Bactericides aresimply poisons that kill living organisms and killmicro-organisms, including sulphate-reducingbacteria, slime-forming bacteria and algae. Thesemicro-organisms attack polymeric drilling muds,completion and workover fluids, slick waters andfracturing fluids. They can cause a deteriorationof the fluid system and reduce the effectiveness ofthe well treatments. Figure 3 depicts the effect ofbiocides on general aerobic bacteria. After theprescribed biocide concentration has beenreached, further increases in the concentrationlevels do not affect the bacteria.

Figure 3 Effect of biocide treatment (Ezzat et al., 1997)

12

DRILLING FLUID CONTAMINANTSThe following is a list of the chemicals typicallyused in drilling fluids in the Cooper andEromanga Basins (Bowyer, 1994), thesecompounds are considered to be non-toxic(Schlumberger, 2001).

AquagelBentoniteBariteBarium sulphate, a mineral used to increase theweight of drilling fluidBenexBentonite extenderBentoniteClay containing smectite as the essential mineral• presents a very large total surface area• characterised either by the ability to swell in

water or to be slaked and to be activated byacid

• used chiefly to thicken oil-well drilling mudsCMCCarboxymethyl cellulose (sodium)DextridOrganic polymerDurenexResin additive/organic polymerGelBentoniteLignosulphonateMember of the lignin family of organicpolyelectrolytes• modified lignosulphonates may be sodium,

calcium or chromium treated types• as a direct consequence of the ability of

lignosulphates to adsorb on clay surfaces,they are used as an anti-corrosion agent andstabiliser of oil-in-water emulsions.

PACPolyanionic cellulose – a long chain polymer ofhigh molecular weight• can impart viscosity or reduce water-loss

properties to the drilling fluid of eitherfreshwater or saltwater muds

PHPAPartially hydrolysed polycrylamidePolyacrylamideOrganic polymerQ-BroxinModified lignosulphonateSoda ashCommercial term for sodium carbonate (Na2CO3)SperseneModified lignosulphonateVertoilPrimary emulsifier

XC PolymerXanthan gum biopolymer (bacterially producedpolymer)XP-20Chrome lignite

Mud recap reportsMud recap reports were investigated to determinethe areas where the most significant amounts ofdrilling fluids were lost (Appendix 2). Althoughthe number of reports considered was a smallsample of the total number of reports, they were asubstantial aid to determining a perspective ofwhere the major fluid losses occur. Consultationswith field experts and other individuals tended toconfirm the findings from those reports.

As can be seen from the data set (Appendix 2)and the plotted charts, relatively large amounts ofdrilling fluids are lost in the first 4000 ft of aformation (Fig. 4 a–d). Also, large losses occurbetween 8000 and 11 000 ft. The consequences ofthe losses are discussed in the conclusion of thissection.

Mud recap reports were used to determine:• whether significant amounts of drilling fluids

were being lost; and• if major losses were occurring, that they were

not in areas of environmental importance.

The reports describe the surface and downholefluid losses associated with drilling activities.Losses of 500 bbl or more of drilling fluid werenot uncommon in some wells. Losses at this depth(4000 ft) may not be of major environmentalconcern as the Bulldog Shale should provideadequate isolation of the top formation layers.These large losses can be explained by:• unconsolidated sands; and• shales in the layers swelling and adsorbing a

great deal of drilling fluid.

Fluid losses beyond the GAB tend to vary withwellbore depth and the type of formation beingdrilled. In an environmental sense, losses to thissection of the formation are not of majorsignificance. The bulk of the fluid losses to thisregion are most likely to be produced when thewellbore is brought on-line. Therefore, little to nodrilling fluid will remain in the formation.Moomba wells seem to have a fluid loss in the6000–9000 ft interval, whereas losses in otherwells can be expected in the 8000–11000 ftinterval.

13

Figure 4a Downhole drilling fluid loss as a function of depth

Figure 4b Downhole drilling fluid loss as a function of depth

-50

0

50

100

150

200

250

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000

Depth (ft)

Do

wn

ho

le F

luid

Lo

ss (

bb

l)

Moomba 118

Moomba 119

Moomba 125

Moomba 126

-100

0

100

200

300

400

500

600

700

0 2000 4000 6000 8000 10000 12000

Depth (ft)

Do

wn

ho

le F

luid

Lo

ss (

bb

l)

Moomba 128

Moomba 134

Miluna 21

14

Figure 4c Downhole drilling fluid loss as a function of depth

Figure 4d Downhole drilling fluid loss as a function of depth

-50

0

50

100

150

200

250

300

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000

Depth (ft)

Do

wn

ho

le f

luid

loss

(b

bl)

Burke East -1

Moomba 114

Moomba 115

Moomba 116

-20

0

20

40

60

80

100

120

140

160

0 2000 4000 6000 8000 10000 12000

Depth (ft)

Dri

llin

g F

luid

lost

(b

bls

)

Moomba 98

Dullingari-50

Brumby-9

Grenache-1

15

However, fluid losses in these regions are not assignificant as those experienced in the surfaceformations due to the higher degree ofcompaction of this deep formation.

Of interest, however, was the drilling fluidssurvey conducted by APPEA (Appendix 3).APPEA listed the volumes of drilling fluids lostdownhole: it reported a value of 5789 m3 lost toformations drilled in 1996. However, independentresearch and a review of the mud recap reportsindicated a value of 6015 m3 for 1996 fordownhole fluid losses for Santos alone. Thereason for this discrepancy is unclear.

The GAB, on average, tends to start at around the5000 ft mark. Shallow aquifers are found mainlyin the upper formations. As they are not used as apotable water source, fluid losses in this area arenot a major concern, particularly if the BulldogShale is effective in providing upper zonalisolation. Significant interest in possiblecontamination in the GAB exists and steps toavoid contaminating its aquifers are necessarybecause the GAB is important for commercial andlife-supporting activities.

Clay swellingClay swelling has been identified as a reason forfluid loss in the upper areas of a well. For thisreason, it is important describe the method inwhich this water adsorption occurs.

Clay properties are a function of:• structure;• quantities of exchangeable cations; and• chemical composition.

Barshad (1955) discussed the subject of therelationship between adsorption of water (andswelling) and properties of the clay–watersystems. Two types of clay swelling wererealised. The first type is due to the crystal latticeitself (interlamellar or interlayer expansion). Na-montomorillonite exhibits this type of swelling.The second type of swelling is due to theadsorption of water on the surfaces of the clayparticles.

Some clays, for example kaolinite, do not swellupon hydration. Na-montorillonite clays,conversely swell in water to many times theirinitial dry volume. Calcium and magnesiummontmorillonites and illites have intermediateswelling characteristics.

Evidence exists that water molecules adsorbeddirectly on clay mineral surfaces and for somedistance outward have an organised arrangement,much like that of ice (Grim, 1968). The degree ofbonding to the clay mineral surface reflects thedegree of swelling of the clay. That is, the highwater bonding between the clay and the waterreflects a greater degree of clay swelling(Chilingarian, 1981).

Consideration needs to be given to the clayswelling properties of rocks. Rock clays can blockthe wellbore and prevent the flow ofhydrocarbons because of their relative largesurface areas. The degree of hydration of clays inthe vicinity of the wellbore can affect theefficiency of primary production and secondaryrecovery. Although clay swelling in the wellboreis not a consideration in the environmental sense,it does pose a problem for treating zonal isolationissues in any possible remedial action. In thiscase, swelling may be an issue by not permittingre-entry of the hole for log runs, maintenance orremedial work.

CONCLUSIONThe effects of drilling fluid on the downholeenvironment are not of major concern. Thisconclusion was reached from the evidencerelating to the zones where fluid losses occurred.The fact that large losses are encountered mainlyin the top 4000 ft of the formation is not of majorconcern if the sealing qualities of the BulldogShale are adequate. Those properties need to beinvestigated further for a firm conclusion to bereached. In addition, losses into deeper producingformations are not considered environmentallysignificant because the majority of drilling fluidwill be expelled from the formation upon theinitiation of production. Furthermore, thecomponents present in the drilling fluid do notraise any significant concerns regarding toxicity.However, when wellbore integrity iscompromised a potential concern does existbecause excessive amounts of drilling fluid couldbe lost.

Fluid losses in the shallower formations (above4000 ft) are a result of two primary mechanisms:• unconsolidated sand; and• clay swelling.

It is postulated that the primary reason for fluidloss is the unconsolidated sand and lack ofcompaction in the upper formations. Compactingforces are a function of depth: they increase as

16

depth increases. This means that it is easier fordrilling fluids to penetrate surface formationsbecause of the lack of overburden pressure.Therefore, relatively large amounts of fluid lossare expected. Clay swelling is also important,particularly in the thick shale areas encountered inthe Cooper and Eromanga Basins.

However, what should be considered is theproximity of oil wells to consumption waterwells. Although drilling fluid lost to the formationis not likely to affect distant water wells,contamination may occur in water wells that arenearby. This is also a consideration whenimplementing the use of biocides.

Some concern has been expressed (Ezzat, et al.,1997) over the possibility of bacteria growing inwater storage tanks at well sites. It is possible thatbacteria may have sufficient time to accumulate instagnant water deposits. This could causeproblems if introduced to the drilling mud. It maybe necessary to use bactericides in these tanks toensure that the bacteria is killed before the wateris used with the drilling fluid. Acceptable controllimits should be followed when usingbactericides. Regulation may be necessary.Bacteria introduced to the aquifers can poseproblems if the water is produced at some laterstage by a water well intended for humanconsumption. The increased presence of bacteriacould lead to health problems and could besignificant, especially when bacteria levels exceed‘safe’ limits.

Another difficult matter is the introduction offoreign bacteria into the wellbore environment.The effects of introducing surface-bound bacteriato the underground formations through thedrilling process was not investigated in this study.The potential for problems arises since it is notknown whether bacteria introduced to thewellbore may thrive in such conditions therebyincreasing the risk of aquifer contamination.

RECOMMENDATIONSArising from these conclusions, a study should beinitiated into the potential for aquifercontamination through bacterial activity. Also, thetype of bacteria likely to be present in thewellbore should be investigated as those presentmay not pose a significant health risk. Such astudy should also consider the mobility rates ofbacteria which might enter a major aquifer suchas the GAB.

Further study should be conducted into the typesand use of biocides in the treatment of bacteriainfected waters. This information should provide‘safe’ limits for human beings. It should alsoinclude a study into the amounts of biocidecurrently being used in the field withconsideration being given to nearby water wellsthat may be contaminated by a biocide.Examining the hydrodynamics of the GAB maybe useful for estimating potential dispersion rates.

A study into the validity of slimhole technologyshould be considered. Use of slimholes isbeneficial because smaller volumes of drillingfluids are used and thus the fluid losses to thedownhole environment could be reduced.Furthermore, slim-hole rigs are likely to leave asmaller surface indications: this environmentalconsideration is outside the scope of this report.

The use of underbalanced drilling muds indepleted sands should be considered as a meansfor reducing the fluid losses to the formation. Itmay be a consideration that in production, wherethe geology is known, underbalanced conditionscould be used to prevent excessive losses in thefirst few formation layers, especially in the GAB.Production wells in the Cooper and EromangaBasins are drilled into depleted sands. Typicalvirgin pressures in the Moomba area are 4200 psi:depleted sands in this region exhibit depletedpressures of approximately 2100 psi. Sufficientwell control could be achieved with anunderbalanced mud and an experienced drillingcrew.

The geology of the Bulldog Shale should beinvestigated further. Areas where the shale maybe thin or not present are points of concernbecause they may lead to cross-flow in theformation. Such areas should be identified. Theremay be points in the shale where it is thin enoughto cause problems of cross-flow and thuscontaminate the formation itself.

A more stringent means of fluid accountingshould be maintained in the field. While it isrealised that recording fluid loss to the formationis difficult, a more precise means of fluidrecording should be investigated.

A review on the use of bactericides in wellsshould be pursued. Standards need to be set sothat biocides and bactericides are not usedexcessively. An excessive use of biocides wouldresult in elevated concentrations in the drilling

17

fluids. If freshwater did become contaminated,harmful or lethal concentrations of biocides mightbe present in water intended for agricultural useor livestock consumption. It is not known whatconcentration of bactericides are harmful forlivestock or human consumption. The potentialfor biocide-contaminated formation waters totransfer to a water well is also not known. Suchissues warrant further study.

SECTION TWO: CEMENTSINTRODUCTIONThis section considers the cementing practicesand techniques currently employed in the Cooperand Eromanga Basins. Cementing is the mostimportant part of the well completion andabandonment phase of the well. The integrity ofthe cement and the properties that it exhibit arenot only important at the time of cementplacement but also for many years afterabandonment. Ensuring adequate isolation withinthe wellbore of the reservoir sands that are knownnot to be in natural hydraulic communicationwithin the reservoir is a key objective specified inthe ‘Statement of Environmental Objectives forDrilling in the Cooper/ Eromanga Basin’ (PIRSA,2000).

The use of cement is the only current means ofestablishing zonal isolation within wellbores,either through cement behind the casing or plugsbetween the zones. The number of ways identifiedfor cement breakdown to occur means that cementis not an effective means for long-term zonalisolation. In areas where ideal conditions arepresent cements may provide sufficient long-termzonal isolation, although the period of time hasnot yet been identified.

The following causes of cement deteriorationidentified by Marca (1990) are addressed in thissection:• high temperature;• sour conditions/sweet conditions;• bacterial presence;• cement shrinkage;• formation damage;• poor mud cake removal;• high cement permeability; and• cement carbonation.

Since the first use of cements in wells in 1903(Smith., 1990), the use of cement has been veryimportant in isolating different zones within oil,gas and water wells. Zonal isolation is very

significant in well completions as wellproductivity is improved and more control overthe well is achieved.

Figure 5 Objectives of primary cementingtechnique (Smith, 1986)

Annular gas migrationFluid migration may occur during drilling or wellcompletion operations. Inadequate sealing ofvarying formations in the wellbore can lead to themigration of gas. This migration occurs throughthe invasion of formation fluids into the annulusand is caused by a pressure imbalance at theformation face. The fluids can flow to a lowerpressure zone and, in some cases, to the surface(Fig. 6).

Fluid migration from high pressure zones to thoseof lower pressure can lead to the contamination ofthese zones. However, extreme gas accumulationdue to large pressure imbalances can causeblowouts.

The severity of gas migration is not alwaysapparent. Gas migration after primary cementingcan adversely affect the wellbore and evidencemay only be noticed some time later. Remedialcementing procedures are thus required to correctsuch problems.

Gas migration between zones, which does notbuild up at the surface, is difficult to detect. Gasmigration may cause the following problems(Sutton et al., 1989):• impaired gas production;• filling of the above depleted zones; and• the effectiveness of stimulation treatments

may be reduced.

18

Figure 6 Mechanism for annular gas migration (Parcevaux et al., 1990)

It is possible to evaluate the degree to whichdownhole channelling occurs by the use oflogging tools such as noise and acoustic logs.Hydraulic communication testing is notrecommended because the potential exists forcreating communication across effectivelycemented zones and minor defects in the primarycementing job can be aggravated.

When the gas migration problem was firstrecognised it was thought that it resulted frompoor mud removal properties. As has been seen,poor mud removal does not allow for adequatebonding at the casing/cement/formationinterfaces. This can lead to the development ofchannels for fluid migration. Although othercauses of fluid migration have been recognised,the principal cause stems from a mud removalproblem. This is due to the continuous mudchannels in the annulus between two permeablezones favouring annular flow.

PURPOSE OF CEMENTINGThis section of the report presents information on:• the various purposes of cement, namely

primary and remedial cementing;• the failure mechanisms of cement;• the testing quality of cements jobs; and• the preventative techniques.

Primary cementingPrimary cementing is the process of placingcement in the annulus between the casing and theformations exposed in the wellbore. The objectiveof primary cementing is to achieve zonalisolation. By this it is meant that the mixing ofzones such as water and oil or a freshwateraquifer with a saline one is prevented. This isachieved by forming a hydraulic seal between thecasing and the cement and between the formationand the cement (Fig. 5). At the same time it isnecessary to prevent fluid channels in the cementsheath. In many instances the full productionpotential of a well may not be reached if completezonal isolation is not achieved. Sufficient zonalisolation ensures that the environmentalobjectives in drilling the well are met(Appendix 3).

Problems arising from poor primarycementingSeveral problems inherent in a poor primarycementing job have been identified by Burdyloand Birch (1990):• the well will never reach its full production

potential;• subsequent efforts to repair the cementing job

may actually end up causing irreparabledamage to the formation;

19

Figure 7 Common one-stage primary cement job on a surface casing string (Burdylo and Birch, 1990)

• lost reserves;• lower production rates;• stimulation treatments may not be able to be

confined to the producing formation;• difficulty in confining secondary or tertiary

fields to the pay zone;• aquifer and reservoir damage from potential

cross-flow; and• potential contamination of aquifers utilised as

a resource.

The latter two issues are relevant to this report.

Zonal isolation

The primary cementing job is intended to provideadequate zonal isolation of the formationspenetrated. Inadequate zonal isolation provides ameans for cross-flow between the communicatingformations. This is not desired because it providesa means for aquifer contamination, continuedcross-flow and, perhaps, a mechanism for naturalpressure depletion.

This report focuses on the downholeenvironmental risks, that is the risk ofcontamination of freshwater aquifers resultingfrom current and continued cross-flow in theformation. In particular, the possiblecontamination of the GAB is a very significantissue not only because of commercial interests butalso because the basin has major environmentalsignificance.

Cement time-scale

The cement time-scale refers to the period of timethat the cement must provide an adequatehydraulic seal between the formations penetratedin the wellbore. What needs to be addressed is asuitable time-scale for the life of the cement.Imposing a time limit is difficult. However, as aminimum, the cement should provide adequatezonal isolation for the life of the producing well.In abandonment an inactive well will need toprovide sufficient zonal isolation to prevent cross-flow. If it is found that isolation is not achieved,then a means for providing continued zonalisolation should be considered. This may require acontinued well maintenance program, especiallyfor wells drilled in highly sensitive areas.

Classes of cements

The choice of the correct cement is crucial inachieving satisfactory isolation. The properties ofthese types of cements are detailed in Tables 1and 2 (Course Notes PTRL 3017, 2000).

20

Table 1 Cement components used in typical class cement

Compounds %API Class

C3S C4AF C3A C2S

Fineness(sq cm/g)

Water/cementratio

A 53 8 8 24 1500–1900 0.46B 47 12 3 31 1500–1900 0.46C 70 13 3 10 2000–2400 0.56D 26 12 2 54 1100–1500 0.38G 52 12 8 32 1400–1600 0.44H 52 12 8 32 1200–1400 0.38J 53.8 – 38.8 – 1240–2480 0.44

Table 2 Function of cement components

Compound Characteristics

Tricalcium aluminate/C3A

(3CaO • Al2O3)• Promotes rapid hydration

• Affects the initial setting and thickening time ofthe cement

• Makes the cement susceptible to sulphate attackTetracalcium aluminoferrite/C4AF

(4CaO • Al2O3 • Fe2O3)• Promotes low-heat hydration

Tricalcium silicate/C3S

(3CaO • SiO2)• Major component produces most of the cement

strength

• Responsible for the strength that the cementdevelops early in its life

Dicalcim silicate/C2S

(2CaO • SiO2)• Hydrates slowly

• Has the properties of slow, small gradual gain instrength over a period of time

Applications of API cementsIn order to improve the chance of effective zonalisolation it is necessary to use the correct cementfor the specific well environment.

API cement A

• typically used a depth of between 0 and6000 ft

• used at temperatures up to 170oF• used when special properties are not required• use of this cement favoured since it is the

most economical of all cements

API cement B

• can be used at depths between of 0 and6000 ft

• intended for use when moderate to highsulphate resistance is required (wellconditions permitting)

• used at temperatures up to 170oF• an economical cement

API cement C

• same depth and temperature range as class Aand B cements

• used when high early strength is required• high in tricalcium silicate

API cement D, E

• class D at a depth range of 6000 to 10 000 ft;E at depths of 10 000 to 14 000 ft

• class D used at temperatures of 170 to 260oF;and class E used at temperatures of 170 to290oF

• used when fairly high temperatures andpressures are encountered

• more expensive than Portland cement• available in types that exhibit high resistance

to sulphate

API cement F

• used in the depth range of 10 000 to 16 000 ft• used at temperatures of 230 to 320oF• used in cases of extremely high temperature

and pressure• types include moderate and high resistance to

sulphate

API cement G, H

• used in depths between 0 and 8000 ft• temperature range up to 200oF without the use

of modifiers• basic cement compatible with accelerators or

retarders• additives can be blended in at bulk station or

at job site

21

API cement J

• used at depth ranges of 12 000 to 16 000 ft• intended for use under conditions of extreme

temperature and pressure 170 to 320oFunmodified

• properties are such that it will not set at atemperature of less than 150oF

• useable with accelerators and retarders

Current cementing practicesCementing techniques in the Cooper andEromanga Basins generally use class A cement inthe surface casing and class G cement in theproduction interval. Two case studies arepresented in this section to illustrate the cementproperties utilised in the Cooper and EromangaBasins. Both wells are located in the Della field:Della 7 reached a true vertical depth (TVD) of6700 ft on February 1980 and Della 20 reached aTVD of 6623 ft on April 2000.

Della 7

Surface cement410 sacks of class A cement were used to cementthe surface casing in place. This cement had thefollowing properties:Slurry weight: 15.6 ppgThickening time: +3 hoursWater requirements: 4.87 gals/sackSlurry volume: 1.1 ft3/sackProduction cement200 sacks of class A cement were used to mixwith:• 4.0% Bentonite by weight of waterThis was followed by 160 sacks of class F cementmixed with:• 15% salt by weight of water

Production casing cement requires that thefollowing properties are maintained:

Class A Class FSlurry weight 11.8 ppg 16.2 ppgSlurry volume 2.4 ft3/sacks 1.0 ft3/sacksWaterrequirements 15 gal/sacks 4.3 gal/sacks

Della 20

Surface cement174 sacks of class A and a tail slurry of 85 sacksof class A cement:

Lead Slurry Tail SlurryBentonite 6% BWOC 0.15 gal/sacksLitefill 25% BWOC 0CaCl2 1% BWOC 0Anti-foam 0.01 gal/sacks 0.01 gal/sacksDispersant 0 0.15 gal/sacksRetarder 0 0.04 gal/sacks

Production cement219 sacks of class G cement followed by 179sacks of class G tail slurry.

Lead Slurry Tail SlurrySilica flour 35% BWOC 35% BWOCUni-FLAC 0.6% BWOC 0.7% BWOCBentonite 16% BWOC 0Anti-foam 0.01 gal/sacks 0.01 gal/sacksDispersant 0 0.1 gal/sacksRetarder 0 0.01 gal/sacksRetarder aid 0.06% BWOCStabiliser 0.55% BWOC

The most significant difference in the cementingtechniques between the two wells was theproduction cement. It is surprising to see that inthe older well, Della 7, class F cement was used.Class F cement is used in areas of hightemperature and pressure. It also offers moderateto high sulphate resistance. Class G cement, asused in Della 20, does not have the hightemperature and sulphate resistant properties ofclass F cement, but it is likely that the cementadditives are adequate in fulfilling the purposesfor which the cement was designed. Class Gcement is used because it offers faster thickeningtime and increasing versatility. Class F cement ismore costly as it is imported and its use is nowlimited to wells that require the specific class Fproperties.

Two questions arise from this finding. Does thecurrent class G cement offer anyadvantages/disadvantages over the old class Fcement? If so, what are they? Answers to thesequestions are beyond the scope of this report.

CEMENT FAILURE MECHANISMSCarbon dioxide effects on cement inwellCarbon dioxide can corrode cement through aseries of chemical reactions. The process isknown as cement carbonation and is often themost likely cause of cement deterioration. Overtime the carbon dioxide can corrode the cement;Thus zonal isolation and loss of casing will occur.

22

According to Bruckdorfer et al. (1986) theprimary mechanism by which cement corrosionoccurs, is:

CO2 + H2O <-> H2CO3 <-> H+ + HCO3-

Ca(OH)2 +H+ + HCO3- -> CaCO3 + 2H2O

C-S-H gel + H+ + HCO3- -> CaCO3 + amorphous

silica gel

CO2 + H2O +CaCO3 <-> Ca(HCO3)2

Ca(HCO3)2 + Ca(OH)2 <-> 2CaCO3 + 2H2O

When excess carbon dioxide is present, calciumcarbonate is converted to calcium bicarbonate,which is water soluble. Calcium carbonate canmigrate out of the cement matrix because of this.The dissolved calcium bicarbonate can then reactwith calcium hydroxide which goes to formcalcium carbonate and fresh water. The resultantwater could go on to dissolve more calciumbicarbonate. Thus the overall result is a leachingof cement materials from the cement matrix. Thisleads to an increase in porosity and permeabilityand a decrease in compressive strength. Corrosionof the cement by carbon dioxide isthermodynamically favoured and it can not beprevented.

Figure 8 the amplitude response data from thecement bond log clearly shows the loss in cementbonding. Although the cement bond indicated bythe log (at time of placement) showed a relativelygood bond, the later log at the same depthindicates poor cement integrity which ultimatelyresults in poor zonal isolation behind the casing.The log can also provide evidence of cementdegradation and deterioration and possiblemigration of crushed fragments from the annuli tothe wellbore, leaving free space behind casing.

The bond of the cement to the casing or theformation interface is affected by:• the bulk volumetric shrinkage of the cement;• the lack of casing and formation roughness;• a mud film or channel at the interface;• a free water channel or layer in deviated

wells;• excessive downhole thermal stresses; and• excessive downhole mechanical stresses.

Figure 8 Cement Bond Log response at time ofcement placement and some time after.Represents the difference in CBL responses attime of cement placement right, 1983 and sometime after re-evaluation left, 1998 (Krilov et al.,2000)

The extent to which cement shrinkage is asignificant problem is difficult to determine. Thetime factor is important because shrinkage is dueto cement hydration which continues with time.As cement begins to set and the hydrationaccelerates, an increase in intergranular stresses isexperienced due to the growth of calcium silicatehydrates. The mechanism for such growth is notbe discussed in this report. Cement hydration isresponsible for an absolute volume reduction ofthe cement matrix and is often known as chemicalcontraction. Normal Portland cement cantypically expect a volumetric shrinkage of 4.6%.This phenomenon, reported in many civilengineering cases, is due to the volume of thehydrated phases being less than that of the initialreactants.

Migration of gas through cement porestructureGuyvoronsky and Farukshin (1963) were the firstto introduce the concept of gas migration throughthe pore structure of a very permeable gelled orset cement. Then Cheung and Beirute (1982)

23

proposed a mechanism indicating that the gas firstinvades the cement pores and then permeates theentire cement matrix, consequently preventing thehydration process from closing all of the porespaces. Parcevaux (1984) was able to prove theexistence of free porosities. These were composedof well-connected pores which began to appearupon the initiation of the setting period. In fact, hestated that gas migration is driven by the unsteadypermeability effect through the cement pores.Following the primary enlargement of the cementpores, a pseudo-steady state is achieved whencommunication has been established throughoutthe cement column and gas channels have reacheda stable size.

Figure 9 shows that immediately after placementthe slurry behaves as a fluid and transmitshydrostatic pressure. A compensation ofvolumetric changes due to hydration and fluidloss is accomplished by a reduction in the heightof the cement column. Continued fluid loss fromthe slurry, as well as hydration, results in thedevelopment of a gel structure that causes thecement to lose its ability to transmit fluidpressure. At this stage it is possible for thepressure to drop and become less than the gaspressure. A potential for gas flow now exists.During this stage the cement becomes self-supporting and further hydration causes a furtherdecrease in pressure. The gel structure restrictspressure as the cement slurry thickens with time.Gas flow can be inhibited by the formation ofstrong bonds between cement particles whichreduces permeability. The critical area isindicated by the shaded region in Figure 9.

Casing problem as part of thedownhole risk studyCasing corrosion in the cooper basin

A consequence of the failure of a cement jobbehind casing is the exposure of the casing tocorrosion mechanisms. To illustrate this the blow-out of Della 1 in 1987 is discussed and thereasons behind this blow-out are investigated. Atthe time it was suspected that the productiontubing, production casing and surface casing allfailed at the 35 m mark below the surface(Martucci, 1989).

In the Della 1 cross-section (Fig. 10), a 12 1/4 in;diameter hole is bored and the 9 5/8 in. surfacecasing, which has a depth of 300 m, wascemented in place. A concentric hole was thendrilled within the surface casing to total depth andthe 7 in. production casing was then set and

cemented in place over the bottom 1000 m.Production casing was suspended within thecasing spool on the ‘Christmas tree’. The finalstep was running the production tubing down theproduction casing and centralising it with apacker. This acted as a barrier between theformation and the surface.

On 16 September 1987 gas was seen ventingthrough a 2 x 2.5 m crater about 30 m from thewellhead. As a result of a wireline run, it waspossible to ascertain that both the productiontubing and the production casing had parted at the35 m mark.

The degree of failure in Della 1 meant that it wasnot possible to remove the surface casing, theproduction casing or the production tubing forexamination. What needed to be addressed wasthe issue of whether or not this was an isolatedevent or if concerns for other wells in the CooperBasin were warranted. The factors leading to theblow-out needed to be determined and criteria forthe assessment of other wells needed to beestablished.

Martucci (1989) reported on the factors whichcould have put Della 1 at considerable risk:• limited allowance for external corrosion due

to the fact that no intermediate casing waspresent;

• the cement sheath around the surface casingwas incomplete at between 15 and 90 m,leaving part of the casing open to corrosionfrom the formation;

• the steel in the N–80 production casing usedover the upper part of the well is moresusceptible to carbon dioxide attack than J–55casing;

• it was an old well cased and suspended afterdrilling in 1970; and

• there was no system for cathodic protectionpresent at the time of drilling the well.

Based on of these factors, the followingassessment criteria were recommended for otherwells that are deemed to be at a potentially higherrisk of critical failure (Martucci, 1989)• no intermediate casing present;• wells with production casing/production

tubing annulus pressure that is less than1.7 Mpa.

• wells more than five years old; and• testing the nature of soils.

24

Figure 9 Slurry dynamics immediately after placement (Bannister et al., 1983)

Figure 10 Della 1 cross-section (Martucci, 1989)

25

In regards to the second point, the use of EUEtubing does not guarantee leak prevention of gasfrom the screwed joints. Over a lengthy period,gas would leak from the screwed joints into theproduction casing/production tubing annulus. Ahigh annulus pressure would indicate thesoundness of the production casing and a lowannulus pressure could signify the possibility of aperforated production casing. In the case of thefourth point, the soils in the Della field werecorrosive. In many instances buried flowlineshave failed 3 to 5 years after operation. Thisindicates how quickly corrosion can take place.

Under the criteria, 42 wells were considered to beat an elevated risk. Following furtherconsideration this figure was revised to 67 wells.Because of this high figure, the following workprogram was employed to check casing integrity;• pressure testing of the wellhead seals;• obtaining samples of annuli gas, liquid and

well scrapings for corrosion activityinvestigations; and

• topping up the casing annuli and pressuretesting the production casing.

Such activity meant that some wells werescheduled for workovers. Workovers wereconducted to stop the substantial communicationbetween the surface casing annulus and theproduction casing annulus. Many wells providedinconclusive information as to the need for aworkover and were therefore placed into aseparate category. Several wells required furtherobservation. One such well, Della 15, was a poorproducer: abandonment of the well was requiredto recover the casing and tubing for examination.

External corrosion of the surface casing bygroundwater was thought to be the most likelycause for the Della 1 failure. Recovered casingfrom Della 15 indicated that corrosion of thesurface casing was from the inside and that theattack on the production casing was much moresevere than that on the surface casing. This helpedto indicate that the source of the problem was dueto the surface casing/production casing annularspace. These wells suffered externally corrodedand parted production casing and some internalcorrosion of the opposing sections of the surfacecasing. The attacks appeared to be more intenseacross the parted region. However, none of thecorrosion logs indicated holes in the surfacecasings. No significant cement bond was presentbehind the surface casing in the top 100 m ofthese wells.

It is likely that the corrosion was due to corrosivechemicals present in the drilling mud being left inthe well at the time of completion. Formationwaters containing corrosive compounds were alsoa possible reason for the failure, particularlylignosulphates in the mud which can decomposeto release CO2 and H2S. It should be recognisedthat unless there are substantial amounts of CO2

and H2S, lignosulphates are not likely to cause thehigh corrosion rates observed.

High temperature chemistry ofPortland cementPortland cement is basically a calcium silicatematerial. When water is added, the tricalciumsilicate (C3S) and dicalcium silicate (C2S), thetwo major components of Portland cement, form agelatinous calcium silicate hydrate (C-S-H gel).This gel is responsible for the strength and solidstability of the set cement at ordinarytemperatures. At well temperatures of less than446oF the C-S-H gel is a very good bindingmaterial: it is the first product to be formed fromthe hydration process. At higher temperatures theC-S-H gel is subjected to metamorphism, whichdecreases the compressive strength of the cementand increases the permeability of the cement onceit sets.

At the high temperatures the C-S-H gel convertsto an alpha dicalcium silicate hydrate (α−C2SH).This is highly crystalline and is much more densethan the C-S-H gel. Because of this, the shrinkagethat occurs adversely affects the integrity of thecement.

The major concern is not whether the strength ofthe cement is sufficient to support the casing butwhether a high permeability will be created as aresult of these curing conditions. In order toprevent interzonal communication, the waterpermeability of well cements should be no morethan 0.1 md. Nelson (1990) found that within onemonth of curing time, the water permeability ofnormal density class G systems was 10 to 100times higher than the recommended limit. Thecompressive strength and permeability changes ofneat Portland cement as a function of time isshown in Figure 11.

26

Figure 11 Compressive strength and permeability behaviour of neat Portland cement at 446oF. Plots 1 and 2represent normal density class G cement. Plot 3 represents class H cement and plot 4 indicates cement oflower density. Furthermore, the permeability of the high density class H system was barely acceptable(Nelson, 1990)

Figure 12 The accumulation of debris in well 14, SW Pannonian Basin in Croatia (Krilov and Loncaric, 2000)

27

Case historyThe most severe case of permeability problems isseen in well 14 at the gas condensate field in theSW Pannonian Basin in Croatia (Fig. 12).Analysis of the well cuttings showed that itconsisted of high bottom-hole temperatures(356oF at 3350 m) combined with sour gas andhigh salinity brine as formation fluids had theability to cause significant problems in the well.

In 1988 a break in production was encountered.At this point after a steady decrease in theproduction period the hydrocarbon content in aone-month period dropped drastically (gas from250 000 m3/day to 100 000 m3/day; condensatefrom 20 m3/day to 10 m3/day). A sharp increase inwater production was noted too. Approaching theend of 1993 the wellhead pressure dropped from180 bar to 70 bar. The well was subsequently shutin. Log analysis data indicated that a formation ofdebris was present below the packer. Only 5 m ofthe 25 m of the perforated zone was leftuncovered as a result of this debris.

The debris could have resulted from:• long-term exposure in a harsh downhole

environment causing cement deterioration andopening the path for drastic waterbreakthrough after a loss in zonal isolation;and

• more intensive sour brine breakthroughcausing a more aggressive cement corrosionprocess resulting in the deterioration.

Debris recovered from the well were shown to bedeteriorated cement, mixed with corrosionproducts and downhole scale fragments. It wasconcluded that these constituents were the resultof a 15-year exposure to aggressive downholeenvironments (sour gas brine).

Microannular formation in cementsIt is possible that a microannulus can be formedbetween either the casing and the cement or thecement and the formation. Such occurrences canbe determined through the use of a cement bondlog response or through the observation of gasmigration problems.

One example of microannulus formation is givenin terms of the radial displacement of the casingresulting from wellbore temperature and/orpressure changes. This occurs predominantlywhen the wellbore pressure is decreased, ie. achange in mud weight when the cement has set.

This type of microannulus is known as an innermicroannulus. An outer microannulus is formedwhen there is cement bulk shrinkage. This is aworst-case scenario but a realistic one. A clearunderstanding of these mechanisms is essential inorder to identify extreme cementing problems insome cases.

The use of expanding cements can help to preventthe formation of microannulus. Theoretically,expanding cement will fill any gap and willensure that good bonding is achieved betweeneither the casing and the cement or the cementand the formation. Expanding cement is known tomove only in the direction of the formation andnot in the direction of the casing.

Mechanism of shrinkage andexpansionThe primary mechanism behind the phenomena ofshrinkage and expansion has to do with theformation of hydration products. These productshave different volumes compared with thehydrating components. Changes in externalcement sample dimensions are referred to as bulkshrinkage and bulk expansion.

Cement chemical shrinkage

Cement chemical shrinkage is the basicmechanism operating during the hydration ofPortland cement. As the volumes of the water andcement are larger than the volume of thehydration products, a volume contraction occurs.Measurement of total chemical shrinkage isconducted by placing cement slurry in a reservoirunder free water access. The amount of water thatthe cement adsorbs during hydration correspondsto the total chemical shrinkage.

Experimental results have shown that themagnitude of the bulk shrinkage depends on theenvironment in which the cement exists. Freeaccess to additional water might make up for thebulk shrinkage and a visible change in the volumewill not be observed. But the absence of freewater and pressure application makes excessiveshrinkage possible. Chemical shrinkage is a linearfunction of the percentage of the four majorclinker minerals. This shrinkage is thereforedependent on cement class. The water-cementratio influences the magnitude and the rate of thechemical shrinkage. The shrinkage rate rises alsowith an increase in the curing temperature.

28

Cement expansion

Cement expansion is an increase in bulk volumeof the initial cement volume. This can be achievedby the addition of cement expanding agents to thePortland cement. Cement expansion of cement isrelated to the chemical and mineralogical changeswhich result from the hydration andrecrystalisation of the expanding agents calciumsulphate, calcium sulphate hemihydrate or sodiumsulphate. The magnitude of the expansion isdependent on the amount of expanding agentadded, the cement powder, the slurry design andthe curing conditions. Although expansion occurs,an overall total chemical shrinkage is stillmaintained.

The general basis for using an expanding cementhas to do with microannulus prevention. Aproblem arises if the properties of the cement arenot adequately controlled. A cement may possessexpansion properties that could damage theformation through excessive expansion.Unconsolidated formations limit the use ofexpanding cements in terms of inner annulusformation.

Baumgarte et al. (1999) reported that anexpanding cement in a soft formation may be atrisk of creating an inner annulus. This wouldoccur as the expansion moves radially outwards inthe direction of the path of least resistance.Because of this observation, expanding cementsystems tend to work best in hard formationswhich can accept the expanding force of thecement. The hard formations can pre-stress thecement and thus build a good hydraulic seal. Thecement only begins to expand after the cementhas set. It is therefore important to realise that thebond between the formation and the cement andthe casing and the cement will only improve withtime.

Long-term leaking of oil wellsOil and gas wells can develop leaks well after theborehole has been abandoned. Severalmechanisms are believed to be the cause of suchleaks:• channelling;• poor mud cake removal;• shrinkage; and• high cement permeability.

These leaks are instigated by cement shrinkagewhich in turn gives rise to the propagation offractures. These fractures propagate from the

action of the slow-moving gas at high pressurebehind the casing wall. Many different aspectsmust be considered when dealing with this asmany factors can contribute to the long-termbreakdown of the cement design.

These further considerations consist of:• workability;• density;• set retardation;• mud cake removal;• entrainment of formation gas;• shale sloughing;• pumping rate; and• mix consistency.

The consequences of cement shrinkage are onlylikely to be realised over time. In North Americawhere tens of thousands of wells are eitherinactive or abandoned, some wells are leaking gasto the surface. This is a result of cementshrinkage and microannular formation.

Some of the migrating gas can enter shallowaquifers where traces of sulphurous compoundscan sour the water making it non-potable.Furthermore, it can cause the presence ofcorrosive waters that can deteriorate the cementand the casing. Escaping gas can cause furthertrouble by entering household systems andflowing when the taps are turned on or byentering agricultural wells and locking the well.This gas influx is likely to increase theconcentration of gas over time. The extent towhich this is a problem needs to be determined.

It is the view of Dusseault (2000), that if theseevents are occurring then the standards for oil-well cementing and abandonment are either notwell founded or are based on a flawed system.When the problem to be rectified has beenestablished, practices can be based on correctphysical mechanisms which might give a betterchance of success.

Mud cake removal for cementing jobPoor mud cake removal is another area of cementfailure. In order to cement successfully, acomplete displacement of the drilling mud duringthe placement of the cement must be achieved.Many methods of mud removal exist, including:• pre-flushes;• centralisers;• casing movement; and• conditioning of the drilling mud.

29

These procedures displace the mobile mud butreally need to be applied vigorously in order toremove some of the gelled mud. There is somedifficulty in removing the mudcake completely asshown in laboratory testing (Haberman et al.,1991). Even when it is partially removed by thetangential flow of fluids the filtration rate is notchanged. An increase in filtration rate can beachieved by using mechanical scratchers or usingturbulent spacers at high displacement rates.

An interesting point to investigate is the strengthof the mud cake during the cementing process. Islow permeability enough to control the fluid lossfrom cement slurries or do the filtration propertiesof the cement slurry dictate the overall fluid loss?An experiment by Haberman et al. (1991)determined the function of pressure using aregular HTHP mud fluid-loss cell at constanttemperature (175oF). Haberman et al. found thatthe filtration rate was independent for theoverbalance pressure as expected for a dispersed,compressible mud cake. During each test the fluidloss was assumed to be approximately constantover the short time intervals used.

Haberman et al. were unable to supply anyevidence as to the mud cake being removedmechanically during cementing. This supportedthe laboratory evidence which suggested thestrong mechanical nature of the filter cake. Thefilter cake was held tightly onto the wellbore bythe overbalance pressure. This test was conductedunder conditions present in the Mississippi RiverDelta. Therefore the conclusions from thisexperiment only apply to wells drilled in thisregion. It is difficult to predict whether similarresults could be expected in fields such as those inthe Cooper and Eromanga Basins. Theexperimental conditions in the Mississippi regionwere:• extensive sand intervals, commonly with

permeability's greater than 100 md;• typical (TVD) was in the vicinity of 10 000 ft

and the casing was cemented up to a depth ofapproximately 5000 ft;

• the bottom hole circulating temperature wasmeasured to be about 160oF;

• no scratchers were used;• centralisers were used across production

zones; and• casing was reciprocated.

The following observations were made as a resultof the experiment:• cement placement did not alter the fluid loss

substantially, indicating that the mud cakewas not removed or even altered by thecementing process;

• in the absence of gelation effects, fluid-losscontrol additives in the cement slurries had noaffect on the fluid loss after cementing – fluidloss was controlled by the filtration propertiesof the mud cake;

• the significant reduction in fluid loss caused alarge reduction in the overbalance pressure;

• downhole fluid loss can be determinedaccurately by measurements made at thesurface; and

• the magnitudes of downhole fluid losses wereequivalent to a low-temperature, low-pressureAPI drilling mud fluid loss of approximately1 cc/30 min. when normalised to the samesurface area as the API test.

The experiment provided an insight to the longdebate about whether fluid losses were dominatedby mud filter cake or cement filtration properties.It was found that in an overbalance situation(most common) the integrity of the mud cake isquite high. Thus it indicated that mud cake wasnot removed by the cementing process. Thisexperiment did not involve the use of scratchers:it is recommended that the effectiveness ofscratchers to remove mud cake build up beinvestigated. The use of scratchers for increasingthe effectiveness of the primary cementing jobshould be investigated also. It is likely thatscratchers will improve cementing practicesbecause they provide a mechanical means for mudcake removal. Perhaps the poor mud removalproperties in some wells could be the reason forpoor primary cementing. Brumby 9 is one wellwith poor cementing properties: it is discussedlater in this report. Pre-flushes are not used beforecementing in the Cooper and Eromanga Basins.Instead, two well volumes are circulated in orderto remove the mud filter cake. The effectivenessof this technique is not known.

REMEDIAL CEMENTINGIn many cases the primary cementing job fails andthere is a need for a remedial cementing job to beperformed in order to achieve the designrequirements of the primary cement job (Fig. 13).Many remedial jobs involve the technique ofsqueeze cementing or cement plugs. Marca (1990)noted the purpose of the remedial cementing as:

30

• repairing a primary cement job that failed dueto the cement by-passing the mud or aninsufficient cement height in the annulus;

• eliminating water intrusion;• repairing casing leaks caused by corroded or

split pipe;• sealing off lost-circulation zones;• attempting to stop fluid migration into a

producing zone;• abandoning a non-productive or depleted

zone;• decreasing the producing gas–oil ratio by

isolating the gas zones from adjacent oilintervals; and

• plugging all or part of a zone in a multizoneinjection well so as to direct the injection intothe desired intervals.

The cement slurry in a remedial job is subject to adifferential pressure against a filter of permeablerock. The physical considerations that must beacknowledged relate to filtration, deposition offilter cake and potential fracturing of formations.The differential pressure exerted on the slurrycauses the cement to lose part of its water to theformation. As a result the slurry begins todehydrate and a cake of partially dehydratedcement is formed. The rate of filter cakeformation (Fig. 14) is a result of four factors:• formation permeability;• magnitude of differential pressure applied;• capacity of the slurry to lose fluid at

downhole conditions; and• time.

It is important to tailor the cement slurry to suitparticular formation characteristics. For example,a cement slurry with high water-loss capabilitiesis likely to choke the wellbore with filter cake.The design of the cement slurry should dependupon formation characteristics. Ideally, thesqueeze cement slurry needs to be constructed tocontrol filter cake growth and allow uniform filtercake to build up over all permeable surfaces.

Applications of the squeeze cementjobOne of the major reasons for a failed primarycementing job is poor mud displacement, whichcauses the cement slurry to channel through thedrilling mud. Consequently, voids and pockets orchannels are left behind the casing, resulting ininsufficient hydraulic isolation between thevarious permeable zones. If this condition is leftuncorrected more problems are likely to arise:

• cross-flow between formations which are atdifferent pressures; and

• potential contamination of freshwateraquifers.

These problems can impact on the environment.Other impacts such as those affecting theproduction potential of the well are outside thescope of this report.

One of the most difficult tasks in performing aremedial procedure is determining where toperform the perforations. Once this has beenestablished, correct circulation of the cement willhelp to provide a better remedial job. Figure 15depicts the downhole circulation of cement afterthe location of perforations.

Typically, two situations exist behind the casing:• the mud channel to be repaired is against a

permeable formation and during the squeezejob the cement filter cake builds so that overtime it fills the void; and

• circulation is established between two sets ofperforations and a ‘circulation’ or ‘channel’squeeze is performed to replace the mud inthe channel with cement.

In order for these operations to be succeed it isnecessary to maintain downhole pressure belowthe formation fracturing pressure. If a fracturewas to occur then the ensuing path of leastresistance would lead to large amounts of cementslurry being lost. This could not only damage theformation but could also contaminatehydrocarbon zones and/or fresh aquifers.

The issue of damaging zones of commercialinterest is not an environmental one. What is ofconcern is the possible reservoir damage thatcould lead to a loss in natural reservoir drive notonly in the local field but also neighbouringfields. Damage that could deplete the reservoirpressure of neighbouring in fields is a cause forconcern.

In many cases it is not possible to perform aremedial operation: the integrity of the wellboremay not be strong enough to support such a job.Fracturing which could damage the formationmay then occur. It is not always true that in orderto achieve a good squeeze job a high pressuregradient is needed. A high pressure is likely toexceed the fracture gradient of the formation andlead to lost control of cement placement and

31

Figure 13 Defective primary cementing job (Marca, 1990)

Figure 14 The mechanisms for cement cake build up in borehole (Baret et al., 1990)

32

Figure 15 Circulation of cement in squeeze cementing (Marca, 1990)

formation invasion. A fracture has the ability toextend across various zones and open unwantedchannels of communication between previouslyisolated zones. Careful monitoring of thehydrostatic pressure can ensure that such damagedoes not arise.

One misconception that exists concerning thecement slurry is that it penetrates the pores of therock. If this was to occur the permeability of therock would need to be in excess of 100 darcies(Marca, 1990)! The only way that the cementslurry can actually enter the formation is throughfractures or vughs. Only the mix water and thedissolved substances in the cement slurry canenter the formation. The solids in the cement forma filter cake on the face of the formation (Fig. 14).

Cement plugsAnother procedure used for controlling lostcirculation areas and attempting to repair adamaged cementing job involves the plugging ofthe formation with a cement plug. Plugs are usedto sidetrack a fish, initiate directional drilling,plug back a zone or a well, solve lost circulationor provide an anchor for openhole tests. They arealso the main means for providing zonal isolationto wells that are to be abandoned (Fig. 16).

The use of lost circulation material is commonalso in the setting of plugs. It prevents the loss ofexcessive amounts of fluid to the formation(Fig. 17).

Reasons for cement plug failureThe basis for determining the success of aplugging is the depth of the top of the cementplug and the hardness of the cement. These aretested by cement bond logs and cement evaluationtools (CET) and also by radioactive tracers. Mudcontamination is a major cause of cement plugfailure as it affects the compressive strength ofthe cement significantly.

Successful plugging involves the placement of thecement: the chances of success are greater whenthe plug is set in wellbores of near gauge.Furthermore, plugs have a better chance ofmeeting their design properties if they are set inhard rock formations. Plugs set in soft formationsare not likely to bond strongly with the formationbecause insufficient resistance may prevent agood cement filter cake from forming.

33

Figure 16 The use of cement plugs for zonal isolation in the abandonment of a well (Marca, 1990)

Figure 17 The use of a cement plug to prevent fluid loss to a thief zone (Marca, 1990)

34

Three principal means of plug failure have beenidentified as of for concern:

Mud contamination

This is a major cause of cement plug failure, aswell as of cement failure in general. Mudcontamination affects compressive strength. Itgenerally results from a poorly centralised pipe. Iftubing or drillpipe is not central in the hole it willrest against the side of the well and slurry comingout of the bottom will follow a path of leastresistance. This causes cement channels in themud and cement mixes with the mud when thepipe is pulled out of the hole.

Insufficient cement volume

If a plug is set in a poorly constructed part of ahole that may contain a washout it is likely thatinadequate amounts of cement would be in place.This means that the top of the cement columnwould not have reached the required height. Inlarge sections mud displacement is difficult andtherefore the chances of contaminating the cementslurry are increased. The plug should be placed inthe most well-calibrated part of the hole. Toincrease the chances of success Bradford andDees (1982) and Spradlin (1982) recommended aminimum of 500 ft (152 m) for plug height. Smithet al. (1983) recommended that a plug should bebetween 300 and 900 ft. Bradford and Deesargued that the extra cement is economical,particularly when examining the cost of repeatingthe job, waiting on cement and retesting the plug.

Water loss

Water loss considerations are essential whendetermining the placement characteristics ofcement. If filtrate invades a formation thenformation damage is likely to occur. Water losscan prevent the complete hydration of the slurrywhich typically leads to weak cement.Furthermore it leads to poor cement-formationbonds. For a drilling operator it is extremelyimportant to be able to calculate and control thewater loss.

TESTING THE QUALITY OFCEMENT JOBSA clear understanding of the design purpose ofthe cement is needed if accurate means of cementjob evaluations are to be employed. Otherwise itis impossible to determine whether the cementingjob has fulfilled its requirements. A good cementjob needs to:

• provide adequate zonal isolation;• protect sensitive areas such a freshwater

aquifers;• provide good cement bonds between the

casing and cement and between the formationand cement;

• support loading of casing sufficiently; and• protect the casing against corrosion by

preventing cross-flow of fluids behind thecasing.

With the design purpose stated, it is possible toperform tests to determine how well the cement isperforming. The evaluation of the procedureneeds to be made after the job has been completedso that the level of attaining the objectives can bedetermined. Several techniques can be used toassess the integrity of cement behind the casing.A guide for assessing the quality of cement jobs(incorporating the results of some techniquesdiscussed in this report) is in Appendix 5.

Hydraulic testingHydraulic testing of the degree of zonal isolationprovided by the cement is a commonly done byeither dry or pressure testing. Zonal isolation isintended to prevent cross-flow behind the casingand helps to prevent the corrosion of the casingby preventing formations penetrated in thewellbore from communicating.

Pressure testing is typically performed after everysurface or intermediate casing cement job afterdrilling the casing shoe. The pressure inside thecasing is increased until the pressure exerted atthe casing shoe becomes greater than the pressureexpected to be applied. A casing shoe that doesnot pressure up indicates a poor cement job, and aremedial job is required to correct this.

Dry testing (also known as a drill stem test[DST]) is a good tool for determining theeffectiveness of a squeeze job or cement seal atthe top of a liner (Jutten & Morriss, 1990). Thistest aims to show that upon a reduction of casingpressure the formation fluids do not invade thewellbore. A successful cementing job will showno downhole pressure change during the openingof the downhole valve or during the followingshut-in period. Figure 18 shows the resultsexpected from a test performed before cementsqueezing and the result from a good DST run.

35

Temperature loggingTemperature surveys are useful in detectingcement in the annulus several hours after cementplacement (Jutten & Morriss, 1990). The reasonfor this is the exothermic nature of cementhydration. The heat generated during the curing ofthe cement raises the temperature of the wellborewhich in turn induces a deviation from the normaltemperature gradient (Fig. 19). The kinetics ofcement hydration will be affected by thecirculation of fluids prior to and duringcementing. Thus the longer the circulation, thelower the temperature. This results in longerthickening times and smaller heat increases in thewell. Temperature logs may not be all thatsuitable for evaluating the cement job in deepwells because of the large temperature differencesbetween the top and the bottom of the well.Temperature increases are typically larger in abigger annulus because the amount of heatgenerated is proportional to the volume ofcement.

Communication testerIf channelling behind the casing is expected, thetemperature logs can be effective tools inidentifying the problem. Figure 20 shows atypical case where flow behind the casing isoccurring. This is the first temperature surveyprior to the injection of 80 bbl of diesel oil. Thesecond temperature survey, run only a short timeafter injection, shows a large temperaturedecrease above the perforations and temperaturevariations down to the oil-water contact. Also,these variations indicate communication behindthe casing. In such cases, remedial cementingmust be performed to seal the annulus and reducethe water content.

Noise loggingNoise logging is a means of detecting whetherflow is occurring behind the casing (Jutten andMorriss, 1990). This tool works on the premisethat flowing gas, water or oil produces noise. Itcan give an indication not only to what is flowingbehind the casing but also of the magnitude of theproblem. The tool relies on a succession of staticnoise measurements and so it is difficult toperform when the tool is moving. Therefore it haslittle use in the oil and gas industry.

Acoustic loggingAcoustic logging is a very popular means ofcement evaluation in the oil and gas industry(Jutten and Morriss, 1990). The response from the

acoustic log is related to the acoustic properties ofthe surrounding environment. As a result it ispossible to determine the quality of the acousticcoupling between the casing, cement andformation. Good acoustic coupling indicates agood bond. But it does not necessarily mean thatadequate zonal isolation has been achieved. Thelack of a reliable relationship between theacoustic coupling and hydraulic isolation is amajor limitation on this technique. However, itcan give a general idea of wellbore conditionswhen the acoustic properties of the cement andformation are known.

According to Jutten and Morriss (1990), fairlyreliable data is obtained from the acoustic logwhen there is:• good quality control procedure of the field

log;• knowledge of the well and casing data;• a good estimate of the relevant cement

properties; and• a clear understanding of the previous well

history cases.

Cement bond logFor the purposes of controlling fluid migration itis important that an effective bond between thecement and the formation and between the cementand the casing be achieved. One way to monitorthis bond is through a cement bond log (CBL).When used originally the amplitude of theacoustic signal in a firmly cemented pipe wasonly a fraction of that of a free pipe. From theirfirst use it was immediately established that aCBL is the primary technology in cementintegrity monitoring (Fig. 21). A CBL has theability not only to determine the bond between thecement and the casing but also between thecement and the formation. The accuracy of a CBLis sufficient for determining the compressivestrength of cement under favourable conditions.

Many factors can affect the amplitudemeasurement of the CBL (calibration of tool,pressure, temperature etc.). So it is necessary tospeak of attenuation rates. The basic premise isthat the attenuation rate helps to quantify theresults as a function of cement.

Attenuation rates are linearly related to thepercentage of the circumference of the casingbonded by the cement. The concept of bond index(BI) was derived by Pardue et al. (1963). Thevalidity of a BI was confirmed by Jutten and

36

Figure 18 Dry test expectations and results (Marca , 1990)

Figure 19 Typical temperature survey showing the probable cement top (Jutten and Morriss, 1990)

37

Figure 20 Temperature composite profile log before cement squeeze (Jutten and Morriss, 1990)

38

Figure 21 The configuration of a normal CBL tool run in the hole (Jutten and Morriss, 1990)

Figure 22 CBL interpretation chart (Jutten and Morriss, 1990)

39

Parcevaux (1987) for the percentage of cementedarea, regardless of the shape and fluid of the non-cemented area.

A CBL interpretation chart in, was constructedfrom the attenuation rate variations as a functionof the cement, casing size and thickness (Fig. 22).It presents a relationship between the CBL andthe cement compressive strength. The chart wasmodified for lightweight cements by Bruckdorferet al. (1983).

Limitations of cement bond logsTraditional cement bond logs are limited by thefact that in cemented sections a high amplitudecan be the result of either channelling ormicroannulus formation (Sheives et al., 1986).Sometimes it may be difficult to determine theexact case.

Detection of microannulus with cementbond logs

Measuring of the CBL amplitude is a function ofthe shear coupling of the casing to the mediumbehind the casing (Fig. 23). If cement is presentbehind the casing and there is sufficient bond toprovide a good shear coupling, a reduction in theamplitude will be recorded. An amplitude of 40–50 dB for a 3–foot spacing can be expected,depending on the properties of the cement. Thesensitivity of the CBL is such that even very smallmicroannuli will indicate large amplitudes whichis evidence of poor bonds. This would thenindicate the potential presence of poor hydraulicseals even though zonal isolation may still bepresent. An advantage with the (CET) is thatsmall gaps (<0.004 in.) have a relativelyinsignificant effect on the calculation of the bond.With such small bonds it is likely that the cementis still providing good hydraulic seals. Eventhough the microannulus may not appear to be aproblem at the time of testing, its significance forfuture problems is not known.

Determination of a microannulus using a CBL canbe done by comparing a pressured log with anunpressured log. At a stage in the depth of thewell the hydrostatic pressure will be sufficient toclose the microannulus. This is noted by adecrease in the amplitude of the CBL log.Furthermore, CET logs have the ability to besensitive to gas-filled microannuli because theacoustic impedance contrast between steel and airis very great. Due to this, the bond value dropsbelow the free pipe value and most of the gasexists behind the casing.

Cement evaluation toolA CET has been utilised to overcome the CBLlimitations (Ataya et al., 1987). This evaluationtool is intended to improve the ability todistinguish between good and bad cement bonds,as well as to identify the channels behind thecasing accurately.

This section comprises the predictions of the CETand the communication results. The CET is usefulin determining whether primary or subsequentcementing operations are the reasons forcommunication. Remedial operations such ascement squeezing can be used to correct theproblem. Like the CBL, the CET is limited todetection of fast formation arrivals. The pulseecho tool (see below) is useful in areas of fastformations.

Good cementing jobs have competent shear andhydraulic bonds between the formation and thecasing. The majority of cement jobs support thecasing in the hole. It is the hydraulic bond thatblocks the flow of fluid across a cementedinterval and reduces the danger of casingcorrosion. Effective hydraulic bonding means thatcement seals the formation as well as the casing.The CET determines the degree of this bond.

Typically, the CET is run in combination with theCBL to help provide a greater insight into thecement condition. The CET is a high frequencyultrasonic device with eight focused transducersarranged in a helical pattern on the sonde. A ninthtransducer is aligned with the tool axis and is usedto monitor the velocity of sound in the fluid. Thetransducers act as both transmitters and receivers,emitting an ultra-sonic pulse perpendicular to thecasing wall and detecting the reflected wave.Figure 24 depicts the sonic wave paths throughdifferent medium in the wellbore.

Acoustic properties of cementLog responses in a cased hole primarily dependon the acoustic properties of the hardened cement.While acoustic properties of rocks are known,there are difficulties in cement acoustic propertiesbecause the physical properties of cement changewith time. Thus the log responses also changewith time. Also, cement is physically not the samealong the entire casing string. This can produce alarge difference in the log response on longstrings where a large temperature difference existsbetween the bottom and top of the cement(Sheives et al., 1986).

40

Figure 23 CBL energy transmission as a function of microannulus wavelength (Jutten and Morriss, 1990)

Figure 24 Sonic wave paths (Jutten and Morriss, 1990)

41

ChannelsThe pulse echo tool (PET) is useful for thedetection of channels in cement. PETs have theadvantage over the CBL in being able to isolatethe channels and to give a better circumferentialresolution for the bond value. Instead the PETgives a low ‘average’ bond amplitude over a3–foot interval (Sheives et al., 1986)

Fast formationsCBLs are not favoured in fast formations becausethey do not give accurate data relating to bondformation. In fast formations the sound velocityof the formation is more than the casing soundvelocity and therefore the formation arrivalsignals can interfere with the normal CBL casingarrivals. The PET only measures the cement–casing bond. This results in less interference fromthe formation signals on the measurement.However, consideration must be given when thecement sheath is small and the acousticimpedance of the material behind the casing isdifferent to that of the cement. Reflections fromthis boundary can interfere with the resonancewindow, thereby causing inaccurate bondmeasurements.

PREVENTATIVE TECHNIQUESPrevention of gas migrationBy reducing the critical zone it is possible toreduce the possibility of gas migration (Bannisteret al., 1983). This can be achieved by:• preventing large amounts of water loss from

the slurry to a permeable formation;• decreasing initial hydrostatic pressure exerted

by the fluid column;• limiting/preventing the annular pressure drop

until sufficient interstitial cement bonding hasoccurred to prevent gas flow; and

• mechanical or chemical modifications, nearthe gas zone, helps the cement form animpermeable barrier against gas migration:

• filter cake inhibition: formation of animpermeable cement filter cake againstthe gas-bearing zone; and

• gas-induced inhibition: chemicalmodification of the cement slurry bythe incoming gas to form animpermeable barrier to additional gasflow.

Use of foamed cementIn some situations foamed cement can be asolution for cement problems that are related to

formations that may contain fractures, have a lowfracture gradient, be highly permeable, or arevuggy or cavernous (Davies & Hartogl, 1981).The cement column in the wellbore exerts aweight on the formation that is dependent on theweight of the cement being used and the height ofthe cement column. In many cases the formationmay be too weak and thus the anticipated pressureexerted by the mud column may exceed thefracture gradient of the formation. As a result,some of the weaker formations can fracture andfail. Low-density cements have been developedfor use in these type of environments. Suchcements are typically created by mixingconventional cement slurries with microspheresor even gas. The lowest probable conventionalcement slurry density is 11.0 lb/gal. Anythinglower than this value will have a permeability thatis too high and a compressive strength that is toolow for adequate zonal isolation.

Although hollow ceramic or glass spheres areused in ultra-low density cements, they requirespecial handling techniques. Also, rheologicalcement properties must be maintained accuratelyto prevent the spheres from floating. In addition,the collapse pressure of these microspheres isabout 7250 psi. This implies that only limiteddepths can be used for such slurries, as thecollapse pressure may be exceeded.

For these reasons foamed cements are oftenpreferred. They are less expensive and mucheasier to design that microsphere systems.Furthermore, it is possible to mix foamed cementat lower densities and yet maintain betterproperties. Davies and Hartog (1981) noted thatfoamed cements have the following advantages:• they can develop a relatively high density in a

short period of time;• they can cause less damage to water-sensitive

formations;• they can reduce the chance of annular gas

flow; and• they allow cementing of thief zones.

The addition of gas does not affect the cementplacement properties. Also, the density can bevaried easily by changes in the gas concentration.Because the reduced chance of cement loss topotential producing zones, the prospect ofincreased well productivity and adequate zonalisolation is a benefit.

Nitrogen is the most commonly used gas in afoamed cement. Along with a surfactant to act as

42

a foaming agent, chemicals are used to improvefoam stability. Foamed cements are characterisedby their ‘quality’ which is the ratio of the gasvolume to the total foam volume. Figure 25 showsthe process for mixing and creating of foamedcements. Two extreme structural situations canarise depending on the quality of the cementbeing used. Foamed cements are compressiblefluids and thus the quality of the cement willchange through the process of circulation. Thisoccurs through severe pressure variations.Expected values would see a 1000 psi pressuredecreasing when flowing down the casing wherepressures may exceed 10 000 psi. The quality willagain decrease as it flows up the other side of theannulus. Thickening time does not depend on thequality.

It is possible to estimate the quality of the foamedcement by taking into account the compressibilitylaws for nitrogen and its solubility in the baseslurry. Foamed cements made in field conditionswhich involve high pressure and high shear rateshave been found to be more stable than thosecreated in laboratory conditions. It has also beenshown that higher pressures promote the creationof smaller bubbles (Kopp et al., 2000).

Pressure is not the only parameter that isimportant when considering foamed cements,

which are three phase systems with many changestaking place at the interfaces. Foamed cement isin constant evolution due to the reorganisation ofthe bubbles that may shrink, grow or coalesce. Asa result it is virtually impossible to produce twosamples with the same initial bubble-sizedistribution.

Cured cement has a cement matrix with a networkof pore structures. With a sufficient quality, eachgas bubble is adjacent to several other gasbubbles. Outside forces such as dehydration willcause some interfaces to rupture; as a result theadjacent bubbles will form a channel of two ormore bubbles. Foam stability is of majorimportance not only to the bonding properties ofthe cement but also to the ability of the cement toperform the task it was designed for. Unstablestructures result in a pore structure that isimperfect and interconnected. Typically thisoccurs during the setting of the cement and iscaused by the rupture of unstable nitrogenbubbles upon contact with other bubbles. Thisresults in coalescence and larger gas pockets. Asponge-like structure is encountered that haslower compressive strength, higher permeabilityand poor bonding properties.

Figure 25 Facilities for the generation of foamed cements (Rozieres, et al., 1990)

43

Foamed cement limitationsRozieres et al., (1990) imposed the followingboundary conditions on the design process offoamed cement:• fracture and pore pressure profile;• permeability of the formation;• density of the lead slurry;• safety factors; and• length of the foam column.

The foamed slurry should be designed to have apermeability less than one-tenth than that of thecritical formations. In order to support the casing,the compressive strength of the cement needs tobe in excess of 100 psi (and even above 500 psi ifit is required by regulations). The slurry must alsobe able to contain the pore pressure of theformation. Because of these limits, a lower limitis placed upon the foamed cement density.However, it is the fracture gradient of theformations that determines the upper limit of thefoamed cement density.

The main advantage of foamed cement is that ithas ductile properties that allow it to deform asthe casing is pressurised. Unlike conventionalcement it will not crack. Furthermore, foamedcement can have very favourable tensile strengthsand displacement properties. Thus it is veryeffective in zonal isolation as demonstrated inwells drilled in Wyoming (Kopp et al., 2000).

The cost of foamed cement is often higher thanthat of conventional cement. However, improvedzonal isolation practices means that considerablecost savings could be made over the life of thewell.

Displacement properties of foamedcementDuring pumping foamed cement has the ability todevelop higher dynamic-flow shear stress thannormal cements. This increases the mud-displacement properties of the cement. Slurrydensity is determined by the gas content or qualityand depends on the pump rate of the base slurry,foamer, stabiliser rates and nitrogen rate. Thevolume of gas used to foam the cement decreases,allowing slurry pressure to remain almostconstant during the system’s transition period.This helps to control gas migration and formation-fluid influx, which limits migration channels inthe set cement sheath.

Use of flexible cementsThe lack of ductility from conventional cementshas been identified as a major reason for thefailure of primary cement jobs (Faul et al., 2000).Flexible cements have been used to preventcement sheath cracking. Using of cements withvulcanised rubber is very costly: foamed cementsare an alternative. Foamed cements exhibitimproved ductility over conventional cements andhave the ability to stay at least one magnitudemore ductile than normal cements. This ductilityenables the cement sheath to flex as the casingexpands, thereby helping to protect it fromcracking. Limits apply to the cement quality:above 35% the quality is too porous for zonalisolation and below 20% it becomes too brittle.

As the pressure in the reservoir decreases, theeffective stress action on the reservoir sandincreases: the increased stress can cause largecompaction strain. This strain can deform thecasing. Even if the casing is not breached it canprevent workovers and recompletions. Foamedcement is also chosen for primary cementingpurposes because it exhibits excellentdisplacement properties and is especially usefulwhen there are concerns over reservoircompaction or salt-formation flow.

Comparison of foam and flexiblecementsThe tensile strength of foamed cement make theselow-density cements excellent for zonal isolationin many operations. The low compressive strengthof the foamed cement does not increase the risk offracture initiation and propagation duringhydraulic-fracturing treatments. Increasedwellbore pressure during casing pressure tests orfracture stimulation treatments are tensile innature. The sheath’s ability to withstand thesestresses is evaluated by the cement’s mechanicalproperties and tensile strength. The cements,compressive strength is of minimal importance(Deeg et al., 1999).

Flexible cement is one of the most durablecements that can be used (Fig. 26). Flexiblecement has a much longer life than foamedcement. But the limiting factor in its use is themajor cost involved and so it is unlikely to beadopted for regular use. Strong reasons areneeded to justify using this cement. Vulcanisedrubber cement can offer a longer life insour/sweet well applications.

44

Foamed cement is a good system as demonstratedin Wyoming where six wells were analysed(Kopp et al., 2000). Two wells were cementedwith conventional high-strength, non-nitrifiedcement across the formations. Tracer tagsindicated poor zonal isolation and stimulationtreatments caused communication between thehigh and low pressure zones Foamed cement wasused in the other four wells and this helped toobtain good zonal isolation. Furthermore, thestimulation treatment remained in the zone, littlefracture growth occurred outside the targetformation and communication did not occurbetween the high and low pressure zones. In theconditions present in Wyoming, foamed cementproved to be beneficial for zonal isolation. Nowork has been done to establish the validity ofusing foamed cement in wells in the Cooper andEromanga Basins.

Durable zonal isolation (new cements)Downhole changes may cause sufficient stress toalter the cement sheath and remove the zonalisolation that was created upon primarycementing. Apart from chemical changes,mechanical failure is the single biggest reason forfailure of a cement job. Mechanical failure leadsto the formation of cracks while debonding leadsto the creation of a microannulus. Regardless ofwhether the failure is mechanical or chemical

both mechanisms are highly likely to create a pathof potential fluid conductivity.

Many mathematical models have been proposedto account for cement properties that will preventthe loss of integrity (Le Roy-Delage et al., 2000).In order to avoid mechanical damage, cementswith a ratio of a high tensile strength to Young’smodulus and a low Young’s modulus comparedwith that of the rock are the best cements in termsof mechanical durability. However, theserequirements are functions of the downholespecific well environment (well geometry, casingproperties, rock mechanical properties andexpected loading history). It is possible thatmechanical damage can be caused by:• large increases in wellbore pressure (pressure

testing, mud weight increase, perforating andgas production);

• increase in wellbore temperature (steaminjection and geothermal production); and

• formation loading (compaction and faulting).

Weaker formations generally result in poorcements because the environment is not likely tosupport cement deformation. In cases of atemperature increase, the thermal properties ofsteel, cement and rock need to be considered.Furthermore, mechanical damage can also be dueto cement shrinkage. Therefore non-shrinkingcements should be used.

Figure 26 Life of the cement sheath for the three primary types of cements used (Kopp et al., 2000)

45

Microannuli formed between the casing and thecement are referred to as inner microannulus andthose between the cement and the formation asouter microannulus. Detection of microannuliinvolves the use of CBL. Prevention ofmicroannular formation is limited simply to theuse of non-shrink cements. Cements that expandupon hydration are often preferable in conditionsthat can accept such a physical change.

Research has shown that expanding cementsactually expand towards the formation and not inthe direction of the casing (Baumgarte et al.,1999). This annulus experiment tried to determinethe behaviour of an expanding cement sheath andwhether or not it provided an acceptable approachto microannular prevention. An intention also wasto determine the optimal conditions for using suchcements. The experiment indicated that casedwells cemented with expanding cements in softformations (for example, unconsolidatedsandstones) were at significant risk ofexperiencing debonding between the cement andthe casing string as the cement moved radiallyoutwards in the direction of least resistance. Theoptimal condition when using expanding cementswas in hard formations that could provideresistance to the expanding cement. The successof the cement bond in part has to do with theelastic properties of the cement sheath.

Downhole corrosion preventionPart of successful well completion planningrelates to the casing protection. If the failure ofthe cement is known or anticipated, techniquescan be utilised to enhance the longevity of thecasing. Metals in the wellbore corrode when theycome into contact with corrosive gases in thepresence of water. The primary means of controlis to remove the contact that exists between thegases and the metal. Further controls rely onmetallurgy, the wellbore environment andcathodic protection. In general these techniquesare used in combination or even to obtain the bestpossible type of protection.

As demonstrated, problems can occur if effectivecorrosion prevention methods are not in place.Uncontrolled corrosion can lead to thereplacement of equipment, lost production time,contamination of subsurface formations, blowoutsand well abandonment due to casing leaks. Interms of the downhole environment the majorconcern lies with the possible contamination ofsubsurface formations, particularly freshwateraquifers.

In some cases corrosion cannot be detected untilthe damage is done. This is disturbing because thecasing is the last line of defence in the wellsystem. The type of corrosion under considerationis the oxidation of the iron metal. This does notmean that oxygen is present. However, it doesrefer to the loss of electrons by the metal to gain apositive charge.

Fe – 2e <-> Fe2+

When this reaction occurs the metal loses an atomto the electrolyte and a void in the metal canoccur. This decreases the inherent strength of themetal and thus presents serious problems to theintegrity of the casing design.

Corrosive agentsCorrosive components in a reservoir environmentinclude hydrogen sulphide, carbon dioxide andorganic or inorganic sulphide. Oxygen acceleratescorrosion. It rarely exists in the wellboreenvironment but is introduced through injectingwater, gas or fire floods and circulating drillingfluid. The use of acids for well stimulation canhave an effect on corrosion.

Carbon dioxide

CO2 not only has an affect on the well casing, butis also can cause leaching of the cement in thewellbore. CO2 forms a corrosive environmentwhen it reduces the pH of water to below 7.0.CO2 testing is difficult primarily because it mustbe measured at system conditions: CO2 mayescape when the system temperature is raised orthe system pressure is reduced.

Hydrogen sulphide

Difficulties arise in predicting corrosion throughhydrogen sulphide because the iron sulphideproduced by corrosion is insoluble at the normalpH level. This can form a film which protects themetal. However, in the presence of CO2 the pHlevel is lowered and the increased solubility ofiron sulphide prevents the formation of theprotective film. The presence of oxygen canaccelerate the corrosion by either an aeration cellbeing formed (the oxygen-free environmentresults in a pitting of corrosion) or carbon dioxidemay lower the pH sufficiently to make the ferrichydroxide somewhat soluble and fresh metalavailable for corrosion. Hydrogen sulphide canreact with dissolved oxygen to form free sulphurwhich is also corrosive.

46

Prevention methodsOnce the corrosion problem is identified, it ispossible to implement a prevention or treatmentplan. Corrosion prevention plans should be inplace before a well is drilled. Well programsimplemented without regard to corrosion controlcan make it a very costly matter should corrosionset in. The existing means of corrosion preventionor treatment are outlined below.

Fluid mechanics

In wells handling highly corrosive gas at highvelocity, care must be taken to ensure thatturbulent effects are not created by suddenreductions in pipe diameter or flow direction.Weeter (1982) showed that larger tubing stringswith low flow velocities may have less corrosionthan a smaller sized tubing handling the sameamount of fluid.

Fluid additives to prevent or inhibitcorrosion

Inhibitors are often used to prevent the oxidationof bare metal. Typically they tend to be oil-soluble organic chemicals that prevent water andcorrosive compounds from reaching the baremetal surface.

Removal of corrosive gases

Corrosive gases such as carbon dioxide, hydrogensulphide and oxygen present problems. But it isextremely expensive to remove these gases fromthe wellbore.

Correct selection of materials for tubulargoods and fittings

Corrosion resistant materials should be used in allpotentially corrosive environments. However,their use is governed by cost. It is often tooexpensive to justify their use. Furthermore,corrosion resistant materials may not provide thecorrect mechanical properties required for the job.For example, many stainless steel applications,which are in fact corrosion resistant, often lackthe physical strength as well as the malleability arequired.

Coatings

Protective coatings may be applied in instanceswhere it is not practical to use non-corrosivematerial. The coating can be an application ofcement or various kinds of plastics to the inside ofthe tubing. It is difficult to apply the coatingevenly to eliminate small gaps in the coatingsurface. If areas of the metal are left uncoated

then the metal will corrode at those points therebydefeating the purpose of the protective coating.

Cathodic protection

Corrosion can be avoided by electricallypreventing the oxidation of iron. This is done byapplying an electrical charge that actually forceselectrons on to the pipe. The current is suppliedby a voltage source such as a transformerrectifier. The well casing acts as the cathode onthe negative side and the positive side is a pieceof metal that helps to complete the circuit aftergrounding through earth. Sufficient amounts ofelectrons supplied to the casing prevent corrosion.Platinum and graphite electrodes arerecommended because they can last between 10and 20 years if correct operational procedures arefollowed. Figure 27 shows a typical downholecathodic protection arrangement.

Corrosion testingThere are a several means of determining whethercorrosion is taking place in the wellbore.

Metal coupons

A metal coupon is essentially a piece of metal thatis as close to the metal present in the wellbore asphysically possible. This method has limitations:• it is not convenient to place these coupons

downhole; and• coupons need to be placed in areas where the

water can be trapped so that readings areaccurate.

Surface measurements are not indicative of theextent of corrosion at the bottom of the hole or ofcorrosion at any intermediate depth in the hole.

Corrator probes

Corrator probes are used to monitor a continuouselectrolyte. If connected to a recorder the probescan provide a measure of the continuouscorrosion rate for both general and pitting types ofcorrosion. However, they are limited in the sameway as the metal coupons: they can only monitorcorrosion at the installation point.

47

Figure 27 Typical downhole arrangement of continuous corrosion inhibitor (Weeter, 1982)

CONCLUSIONThe major concern for cement deterioration isleaching which is a thermodynamically favouredreaction. Concern increases because additives arenot present to prevent or eliminate cementcarbonation in hostile environments. The rate atwhich the cement is worn away is not known.This is an area for further study. Also, theformulation of a cement carbonation model indiffering downhole environments needs furtherinvestigation.

The good bonds detected at the time of casing inthe case history had deteriorated substantiallysome 15 years later and had caused a plunge inproductivity. Cross-flow issues were also aconcern. There was a substantial loss of cementintegrity behind the 7 in. casing along theperforated interval. Log responses indicated poorbonding or free pipe. This indication followed thegood bonds in the first stage of cementing. Thelog indicates cement degradation anddeterioration and possible migration of crushedfragments from annuli to the wellbore, leavingfree space behind the casing. However, the casehistory shows the need for continued cementmonitoring and highlights the fact that cementsare not a permanent solution.

The case history was one of the more extremecondition wells. Conditions in this well werefavourable for cement deterioration. Cement inwells exposed for long periods of time toformation brine saturated with sour gas (rich inCO2) under high temperatures (>180oC/ >356oF)

is at major risk. Furthermore, high temperatureshelp to intensify carbonic acid leaching whichaffects cements in these conditions and increasesthe rate of degradation. This will lead to cementdeterioration with a loss of compressive strength.The conditions in the Cooper and EromangaBasins may not be as severe as that expressed inthe case history. However, with time it is likelythat cements exposed to even moderate conditionswill deteriorate. It is important to understand thatthe risk of cement failure increases over time:once initiated the risk is most likely to growexponentially.

Monitoring well integrityAn array of tools exists for the evaluation of acement job. At this stage the CBL seems to be themost accepted tool for the evaluation of zonalisolation in the wellbore. Pulse echo logs havealso been considered. There seems to be somesuccess in their use with enhanced channelresolution, suitability in fast formations andsensitivity to a gas-filled microannulus. Whilesuch equipment may not be 100% accurate, it issatisfactory for providing evidence of the integrityof the cement job.

The most conclusive test for evaluating howeffective a cementing program has been is time.The effective evaluation of the cements usedshould involve the study of already cementedwells.

48

Cement time-scaleThe well in the case history started to fail 15years after the cement had been put in place.Deterioration most probably occurred long beforethis time. So what timeframe for the isolation of aformation is required? This report does notaddress the timeframe for the life of cement butone possible outcome concerning cement life maybe that the technology does not exist to be able toprovide isolation timeframes. Absolute favourableconditions would need to be present to ensure thatcement integrity is maintained for an infinite timeafter well abandonment. This is never likely tooccur since wellbore cement is exposed todynamic conditions and streams of potentiallycorrosive compounds.

Della 1 indicated that the majority of older wellslacked cement in areas near the surface and inother sections of the well. This allowed formationfluid direct communication with the casing string.The risk of external corrosion would be increasedwithout sufficient means of corrosion prevention.Modern-day cementing involves running cementwell into the surface casing which allows forbetter protection of the casing from contact with aformation.

Fluid lossThe loss of fluid from the cement has never beena real concern. Only mix water and dissolvedsubstances have the ability to enter the formation.Very large permeability's must exist in order forthe cement to invade the formation. Furthermore,the cement slurry does not contain anycompounds of known toxicity and fluid losses areso small as to not constitute a need for concern.

RECOMMENDATIONSAny disputable formation penetrated in thewellbore which is likely to be subject tocontamination should be isolated until sufficientevidence can be obtained to confirm or negate thisneed.

Present cementing technology does not to provideadequate zonal isolation for long periods. As aresult, it is recommended that further work beconducted into the adequacy of cementingpractices. The research should include a modelfor predicting the relative time-scale for cementintegrity in certain environments (for example,corrosive and high temperature environments).Consideration should be given to wellabandonment practices. It may be viable to

consider filling the abandoned hole with shalesand clay similar to those originally encountered inthe wellbore. Such a technique will needinvestigation.

Zones exist in some older wells that haveformations exposed directly to casing. This posesproblems, especially in cases where the cementwas the primary means for corrosion prevention.The risk of corrosion increases if sufficient casingprotection is not offered. This was a case in Della1 where the casing was not cemented between 15and 90 m. Many older wells have received sometreatment (cathodic protection) for corrosion.However, the adequacy of this practice needsinvestigation. It is possible that corrosion mayhave jeopardised the integrity of the casing topoint where failure is imminent.

Further study needs to be conducted into the ratesof cement leaching. It is necessary to construct amodel that indicates the rate of cement leachingthat can be expected in given environments. Oneexpected outcome from the study would be toprovide a means for evaluating the degree ofcement deterioration. This study should alsoinvestigate the level of cement deterioration andprovide a point at which the cement is likely to beprevented from fulfilling the zonal isolationrequirements. Such a point will most likelyrepresent a percentage weight reduction orthickness loss at which the properties of thecement are affected.

A study should be conducted into the validity ofthe mud removal procedures. No pre-flushes areused in the Cooper and Eromanga Basins. Insteadthe two hole volumes are circulated. The use ofscratchers for mud cake removal is minimal. Isthis procedure adequate in removing the mud cakefrom the wellbore? Wells that encounteredproblems with the cementing job should beinvestigated to determine the effectiveness of themud removal program. Practices such as pre-flushes and scratchers need further investigation.The practices may require changing or therheology of the mud properties may need to bealtered for more favourable removal.

Examining the validity and cost effectiveness offoamed and vulcanised rubber cements should beconsidered. A means for increasing cement lifecould reduce the costs involved in remediationwork. A study should be made of this cementtechnology to conditions specific to the Cooperand Eromanga Basins. The use of foamed cements

49

is gaining interest, particularly in areas where theformation is known to be weak.

Continued cross-flow from poorly isolatedwellbores is likely to provide a mechanism for thedepletion of natural reservoir energy and ispotentially a source of contamination of freshwater aquifers. Is there a stage where the pressurein the sands reaches an equilibrium and cross-flow ceases? Will the natural reservoir energycontinuously be supplemented by gas cap or waterdrive mechanisms and thus enable cross-flow foran indefinite period of time? Answers to thesequestions will be useful in helping to address theproblems associated with cross-flow.

Current cementing technology is not sufficient inproviding an indefinite zonal isolation. Newmethods need to be considered, particularly whenconsidering the abandonment of the well. Theseconclusions were some of the outcomes arisingfrom this detailed investigation into the downholeenvironmental risks associated with drilling andcompletion practices in the Cooper and EromangaBasins. By pursuing these lines of enquires,benefits may become apparent to the oil and gasindustry.

50

APPENDIXESAppendix 1 Types of filtrationSTATIC FILTRATIONThe static filtration equation represents Darcy’s law of fluid flow. This is the same equation that is usedwhere there is a pressure-induced flow of fluids through. In this case the mud cake is visualised as being athin cylinder.

wheredVf /dt = filtration rate (cc/s)k = mud cake permeability (Darcies)A = flow area (cm2)µ = filtrate viscosity (cp)∆P = pressure drop across the mud cake (atm)hmc = mud cake thickness (cm)

This is not a perfect model because it does not account for permeability, mud cake thickness and thefiltration rate (which is not constant). The greatest variation in permeability occurs early in the filtrationprocess where the only resistance to flow is caused by the filter paper. This means that there is a large initialpermeability that decreases quickly as the mud cake develops and reaches a steady value when the mud cakehas developed to the point where its growth is balanced by its erosion. This initial large permeability resultsin what is known as spurt loss.

To improve the model it is necessary to express the effect of changing filtration rates as the filtrationproceeds. Considering a relationship between the volume filtered and the thickness of the filter cake doesthis. The volume of solids filtered from the mud is equal to the volume of solids in the mud cake:fsmVm= fschmcA

where fsm = fraction of solids in the mudVm = volume of mud that has been filteredfsc = fraction of solids in the mud cakes

The volume of the mud that has been filtered is equal to the volume of the mud cake added to the volume ofthe filtrate:Vm = hmcA + Vf

where Vf = volume of filtrate

Then rearrange the equation in terms of hmc.

)1( −=

sm

sc

fmc

f

fA

Vh

Then by using hmc in Darcy’s law and combining the equations the result will give:

A

tA

f

fPkV

sm

scf )1(2 −∆=

This equation has the advantage because at a constant fraction of solids in the mud cake the volume of mudfiltered through it can be determined. This means it is a useful way of determining the tendency of a drillingmud or cement mixture to lose water into a permeable formation. It is possible to determine the solidfraction in the mud cake by either drying the mud cake and measuring the density of the filtrate or assuming

mc

f

h

PkA

dt

dV

µ∆=

51

that the filtrate has a zero solid fraction. It is then possible to rewrite the filtrate volume/time relationship in

terms of t and constant c:

tcV f =

By using this equation it is possible to make a quick estimate of how much fluid will be lost when drillingthough permeable formations. This equation is most useful, however, for determining the water lossproperties of a drilling mud as well as the determining the spurt loss. The disadvantage of the equation isthat it does not take into account pore plugging effects or formation permeability. Spurt loss is the volumeintercept on a graph of filtration volume plotted against root time. Also, API water loss is defined as thevolume filtered through an API filter press under 100 psig in 30 min. The square root relationship (as above)means that the API water loss and spurt loss is also equal to twice the volume flowing through an API filterpress and a spurt loss of 7.5 min:

V30 –Vsp = 2(V7.5-Vsp).

Bit filtrationBit filtration, the first filtration type is due to the action of the drill bit. Very little filter cake forms on thebottom of the hole because the action of mud jets is highly erosive. In addition, the action of the bit is suchthat every time a bit tooth strikes, a fresh surface of rock is exposed. Beneath the bit filtration is restricteddue to an internal filter cake that forms in the pores of the rock just ahead of the drill bit.

Dynamic filtrationUnder dynamic conditions the growth of the filter cake is limited by the erosive behaviour of the drillingfluid. When the rock is first exposed the rate of filtration is very high and the mud cake grows quickly. Therate of growth decreases with time until eventually the erosion rate equals the formation rate and thethickness of the mud cake remains constant. In dynamic conditions the rate of filtration depends on thethickness and permeability of the cake: these are governed by Darcy’s law. However, under static conditionsthe cake thickness increases to infinity (Fig. A1).

Figure A1 Relative static and dynamic filtration in the bore hole (Outmans, 1963)

52

The filtration rate for dynamic filtration is given by the following equation.

where k1= cake permeabilityτ = shear rate exerted by the mud streamf = coefficient of internal friction of cakes surface layerδ = thickness of the cake subject to erosion(-v+1) = is a function of the cake’s compressibilityQ = dynamic filtration

Permeability varies the most early in the filtration when the only resistance to fluid flow is from theformation. The permeability is greatest initially and it decreases gradually as the mud cake develops on thefilter paper. The permeability of the filter paper reaches a constant value when the mud cake has reached arelevant level. This effect is known as mud spurt.

Mud spurt occurs when the drilling mud comes into contact with the filter paper. The particles in the mudrange from large to small. The smaller particles pass through the filter paper easily yet the larger particlesbecome trapped in the pore spaces of the filter paper and in a sense block it up. Gradually the finer particlesin the mud stick to the larger particles in the pore space and start to clog the pores. It should be noted thatparticles of a critical size are needed to initiate the clogging of the pores. If the particles are too big thenthey will not enter the pores. Particles that are too small will simply pass through the pores and not getstuck. Once the pore space has been completely clogged by the particles a bridge is formed and no moresolids can pass through the pore, no matter how small they are. At this stage only filtrate can invade theformation. The mud spurt time is very much dependent on the size and amount of bridging particles present.A region of three zones is identified in the permeable formation (Fig. 1).

)1(

)/( 11

+−=

+−

νµδτ νfk

Q

53

Appendix 2 Mud recap report data – Drilling fluid lost to the formationDepth (ft) Volume (bbl)

Moomba 98 994 67

2525 35

3530 33

6432 5

6615 3

6950 4

7550 37

7964 23

8469 68

8825 98

8825 64

8825 5

8825 5

8825 6

8825 6

8825 6

8825 6

8825 6

8825 6

8825 6

8825 6

8825 6

8825 6

8825 6

Total Fluid Loss (bbl) 513

Dullingari 50 591 60

2484 150

3016 78

3016 56

3026 0

3681 62

4280 66

5141 65

5734 30

Total Fluid Loss (bbl) 567

Brumby 9 1592 20

Total Fluid Loss (bbl) 20

Grenache 1 2188 48

3522 150

9726 29

10014 60

10082 17

10329 20

10738 40

10738 80

10738 70

Total Fluid Loss (bbl) 514

Depth (ft) Volume (bbl)

Burke East 1 5406 34

5970 89

6386 149

7375 183

7946 187

8331 91

8627 78

8698 17

8698 150

8698 30

8698 34

8698 2

Total Fluid Loss (bbl) 1044

Moomba 114 2470 261

3020 27

4270 10

5500 10

6487 34

7136 111

7348 108

7847 79

9047 129

9047 18

9047 26

9047 15

Total Fluid Loss (bbl) 828

Moomba 115 577 19

3020 105

3020 114

4699 51

5500 10

6330 57

7092 52

7300 108

7456 55

7991 20

8640 93

8640 15

Total Fluid Loss (bbl) 699

Moomba 116 0 20

1733 191

1733 40

1733 0

3310 202

5485 136

6212 103

6718 190

6946 106

7153 21

7513 75

7835 52

7835 115

7835 13

Total Fluid Loss (bbl) 1264

54

Depth (ft) Volume (bbl)

Moomba 118 811 36

1710 192

1710 11

1710 8

3465 145

5023 236

5051 29

5426 143

5192 61

6397 161

6566 72

6771 76

6858 125

7070 57

7201 50

7481 86

7824 63

7824 70

7824 30

Total Fluid Loss (bbl) 1651

Moomba 119 5849 92

6416 95

6850 129

6946 67

7462 130

7588 16

7588 24

7588 27

Total Fluid Loss (bbl) 580

Moomba 125 5357 186

6530 140

7220 79

7400 100

7675 116

8541 70

9092 48

Total Fluid Loss (bbl) 739

Moomba 126 6282 1

6919 2

7092 0

7236 0

7373 0

7442 0

8125 1

8275 4

8275 70

8634 36

8634 20

8634 0

Total Fluid Loss (bbl) 134

Depth (ft) Volume (bbl)

Moomba 128 6465 46

7145 86

7460 0

7988 113

8670 17

9069 185

9069 0

9069 0

9069 0

Total Fluid Loss (bbl) 447

Moomba 134 6171 5

6815 2667153 757394 867766 848589 1778671 1138835 959109 619110 1349110 399110 409292 1059400 319580 309713 789836 1279942 106

10112 13710202 3510434 3910507 7610507 4010507 810507 4810507 0

Total Fluid Loss (bbl) 2036

Miluna 21 1820 594

1820 33

Total Fluid Loss (bbl) 627

55

Appendix 3 Environmental objectives and assessment criteriaMinimise lossof reservoirand aquiferpressures andcontaminationof freshwateraquifers.

This objective seeks to protect thewater quality and water pressure ofaquifers that may potentially be usefulas water supplies, and to maintainpressure in sands that may hostpetroleum accumulations elsewhere.

To address this objective, the risks ofcrossflow between formations known tobe permeable and in natural hydraulicisolation from each other, or wherethere is insufficient information todetermine that they are permeable or inhydraulic communication, must beassessed on a case by case basis andprocedures implemented to isolatethese formations.

The following geological formations inthe Cooper-Eromanga Basins maycontain permeable sands (aquifers)which may be in natural hydraulicisolation from each other (fromshallowest to deepest):

• Eyre formation;

• Winton formation;

• Mackunda formation;

• Coorikiana sandstone;

• Cadna-owie formation;

• Namur sandstone;

• Adori sandstone;

• Hutton sandstone;

• Poolowanna formation;

• Cuddapan formation;

• Nappamerri Group formations,Walkandi and Peera Peeraformations (multiple sands);

• Toolachee formation (multiplesands);

• Daralingie formation (multiplesands);

• Epsilon formation (multiple sands);

• Patchawarra, Mt Toodna or Purniformations (multiple sands);

• Tirrawarra sandstone or SturatRange formation;

• Merrimelia Boorthanna and CrownPoint formations (multiple sands);

• Basement reservoirs.

Drilling and Completion Activities

• Casing design (including setting depths) have beencarried out in accordance with company definedprocedures which satisfy worst case expected loadsand environmental conditions determined for theparticular well.

• Casing set in accord with design parameters andcompany approved procedures.

• Sufficient isolation between any of the formationslisted in the adjacent column – where present – issubstantiated (eg through well logs, pressuremeasurements or casing integrity measurements).

• For cases where isolation of these formations is notestablished, sufficient evidence is available todemonstrate that they are in natural hydrauliccommunication.

Producing Wells

• Monitoring programs, carried out in accord withcompany approved procedure(s), demonstrate nocrossflow or fluid migration occurring behind casing.

• Casing integrity and corrosion monitoring programs,carried out in accordance with company approvedprocedure(s), show adequate casing condition tosatisfy the objective.

Inactive Wells

• In the case where a well is suspended for aprolonged period of time:

• Monitoring methods for detecting fluid migration,carried out in accord with company approvedprocedures for this purpose, are in place and showno fluid migration.

Well Abandonment Activities

• Plugs set to isolate aquifers through the well bore,designed and set in accord with defined proceduresto satisfy worst case expected loads and downholeenvironmental conditions.

• Plugs have been set to isolate all aquifers which arepresent which are not in natural hydrauliccommunication nor have been isolated by cementbehind casing.

56

Appendix 4 APPEA 1996 drilling fluids survey

57

58

59

60

61

Appendix 5 Goal attainment scaling for oil-well cementsGoal attainment scaling for oil-well cements is proposed as method in which the downhole integrity of thewellbore can be investigated and evaluated. It is similar to the assessment concept used to assess therestoration of abandoned wellsites in the Cooper Basin.

A scale for assessing the cement can be used to grade the cement on the basis of how well it is fulfilling thetask it was intended for. The grading of the cement will be based:

Score Outcome

-2 Much less than expected-1 Less than expected0 Expected

Objectives of the cement jobIn order to grade the level at which the cement is fulfilling its task, it is necessary to identify the purpose ofa cementing job. The purpose of a cement job in the abandonment of the well is to provide:• zonal isolation for formations penetrated in the wellbore (zonal isolation is essential for formations that

are not naturally in hydraulic communication; timeframe for such isolation needs to be provided on acase-by-case basis;

• some structural support to the wellbore and help maintain the integrity of the formation; and• protection of the casing through formation fluids not coming into direct contact with the casing.

When the cementing job does not meet the lowest level that it was expected to reach, concerns arise as to itseffectiveness in isolating the formations. In an environmental sense it is very important that the zones beisolated to prevent cross-flow and the potential contamination of freshwater aquifers.

The aim of this assessment is to grade the level at which the objectives of the well abandonment areachieved. For objectives that are not reached an outcome of at least 0 means appropriate corrective measurescan be taken.

What constitutes a good cement job?The most important part of a cement job is the adequate isolation of zones penetrated in the wellbore. Agood cement job isolates these areas by means of a valid hydraulic seal and provides protection from cross-flow. Furthermore, a good cement job fulfils the objectives of well abandonment. In order to achieve aquality cement job it is necessary for the cement to create a sufficient bond with the wall of the formation.The cement must also have a permeability low enough to inhibit migration and have sufficient properties toquench the formation of a microannulus. A good cement job should also have reached the required height inthe wellbore and achieved a hardness that has the required shear strength to support the casing in thewellbore.

Evaluation of the cement jobCBL and CET are the most common means of evaluating the quality of the cement job. These tools are usedto determine the degree of contact between the formation and the cement. In part this determines the degreeof zonal isolation present within the wellbore. The CBL provides a reasonable determination of the cement-to-formation bond. The CET provides a comprehensive determination of the cement-to-casing bond. Thesetools, when used in conjunction, are extremely good at determining the effectiveness of the cement bond.Another method for testing the effectiveness of the job is to use temperature sensitive logs. These are mostcommonly used to detect the cross-flow behind the casing.

In order to grade the quality of the cementing job it is recommended that an independent body carry out thetests. While PIRSA should be required to survey the data and impose a grade, cost incurred by this processshould be borne by industry: it should form part of a competent well completion program.

62

What tests should be performed?In order to grade the cement system, a series of tests based on the same grading system should be concluded.In the event that a grade indicates that the job has been completed poorly, a remedial program should besubmitted by the relevant operator. This program should address the problems identified by the independentbody and how they are to be corrected. If a remedial program cannot be conducted due to a poor formationor other factors, another party should be contracted to assess the validity of the situation.

Tests should be conducted on those wells that have problems known to be associated with the cementingprogram or wells drilled into hostile environments. The tests could be implemented at the time of wellabandonment and then at 5 and 10 year intervals. Follow-up tests could be conducted if a concern regardingthe cement properties existed. However, cost considerations could make such an exercise prohibitive: testingover a long period remains the only method for determining the integrity of the cement but the cost wouldnot make this a viable option.

The following tests should be performed:

Zonal isolation

This test is conducted to evaluate the degree to which the cement is providing sufficient zonal isolation.Formation temperature testers should be used to detect the potential cross-flow of formation fluids behindthe casing. The following scale provides a grading system:

Score Outcome

-2 No zonal isolation due to a complete failure of the cement design. Little to no cement exists inareas that were marked for zonal isolation. Extremely poor cement job. Cement did not reachrequired height.

-1 Zonal isolation is minimal. Failure of cement system with small amounts of cement existing inareas identified for zonal isolation. Poor cement program.

0 Zones identified for zonal isolation have been isolated. Cement program was of a satisfactorylevel. Cement reached expected height.

DST may be a another tool that can be implemented to test the quality of the cement seal. A successfulcementing job will show no downhole pressure change during the opening of the downhole valve or duringthe following shut-in period. The expected dry test results are in Figure 18.

Cement bond

The purpose of this test is to evaluate the degree of bonding present between the cement and the formationand between the cement and the casing. This is investigated by CETs and CBLs.

Score Outcome

-2 No bonding between the cement/casing and cement/formation. Severe channels allowing themigration of fluids between formations.

-1 Some bonding, but formations are still able to communicate hydraulically. Formation ofmicroannulus has occurred.

0 Cement bond is sufficient to prevent cross flow. The bond achieved by the cement issatisfactory.

Cement durability

Cement durability only has validity as a test for some time later in the well time-scale. It is acceptable toneglect the first test at the time of well abandonment as cement durability is really a measure of how wellthe cement is able to combat the downhole environment and ensure continued zonal isolation. It is expectedthat the score of the cement durability in a wellbore will decline as time proceeds. Action should be taken tocorrect the level of durability before it becomes unacceptable. The grading system could be applied here butthe guidelines are difficult to impose.

63

Costs of programsThe remedial programs and wellbore assessments are not only costly but also are time consuming. Testingeach well is certainly not proposed. However, wells that have had problems at the time of completion, or arein an area known to accelerate the degradation of cement, need to be checked. It is recommended that thisprocess be included in the operator’s abandonment program. The operator has a responsibility to prove thatcementing procedures are meeting the requirements they were designed for. Goal attainment scaling for oil-well cements should be used in aging wells as a means of demonstrating the effectiveness of the wellprocedure in place.

64

Appendix 6 Mud programThe function of the mud is to:• cool the bit and drill string to prolong bearing life and reduce pipe damage from the heat;• bring cuttings produced by the drill bit to the surface (a good mud will keep cuttings from sticking above

the bits of the collars: it is important to maintain a clean hole);• suspend the cuttings when the pump is shut off so that the cuttings are prevented from falling down

around the bit and collars;• build a mud cake in the bore so that the borehole is prevented from caving in and losing drilling fluid;• reduce formation invasion and improve electric logging conditions for formations; and• control downhole pressures through the weight of the mud column.

In the field barite is used to provide weight to the mud: this helps to control any pressure kicks. Gels areused to help build viscosity in the mud. They help form mud cakes in the borehole as well as suspendingcuttings in the mud system. A good mud cake must prevent the mud flowing into the formation. Adding lostcirculation material (mica, cellophane flakes and plastic) can do this. It is important that in zones of lostcirculation the drilling mud is monitored continually in order to prevent any loss of mud that could lead to ablowout.

There are five main types of mud: freshwater, saltwater, oil-based, surfactant and emulsion. The best mudfor drilling the formation needs to be chosen. This is done by evaluating the formation in which the drillingis going to take place. The mud is required to have turbulent flow in the wellbore as opposed to laminarflow. Turbulent flow enables the cuttings to come to the surface flat rather that tumbling up the annulus. Italso helps to increase the annular velocity and aids in cleaning the wellbore.

In petroleum engineering the characteristics of the drilling fluid are essential because they not only help tocontrol penetration rates and extend the life of the bit but they also help to prevent any pressure kicks causedby any unexpected drilling regions. The following properties must be known in order to form a reliabledrilling mud.

Viscosity of the drilling fluidThe resistance to flow of a drilling fluid depends on friction between:• solids;• solid and liquid phases;• liquid phases; and• the particles of the drilling fluid.

The first component that affects resistance to flow is known as viscosity. This is caused by the solid clayparticles in the mud rubbing together (mechanical friction). There are two types of viscosity: effective andplastic. The second component that affects resistance to flow is known as the yield strength. This is causedby the intermolecular attraction that exists between the clay molecules which tend to bind together to formstructures. This tendency depends upon:• the surface properties of the mud;• the concentration of the solids in the mud; and• the electrochemical environment of the solids in solution.

Another important property of the mud is the gel strength. This is the minimum energy required to start themud to flow (setting properties).

Components used in drilling fluidsMany clays and additives can be used to create a drilling mud. The major components used are bentoniteand barite.

65

Bentonite

Bentonite is used not only because it is cheap and readily available, but it also:• increases the hole cleaning ability;• it reduces the water seepage or the filtration into formations that are permeable;• it forms thin mud cakes that have low permeability;• it help to keep the hole stable when the cementing is poor; and• it helps to avoid, and in some instances to overcome, the loss of circulation.

Barite

The main purpose of barite in a drilling fluid is to increase the density of the mud. Barite has a high specificgravity and because it is not soluble in water it does not react with other components in the mud.

66

Appendix 7 Effect of temperature on the rheology of drilling fluidsThe rheological properties of drilling mud under downhole conditions may be very different from thosemeasurements made at ambient pressures and temperatures. Temperatures in downhole conditions dependon the geothermal gradient. Elevated temperatures can influence the rheological properties of drilling fluidsin any of the following ways:• physically – an increase in temperature decreases the viscosity of the liquid phase;• chemically – all hydroxides react with clay minerals at temperatures above 200oF (and temperature does

not disturb low alkalinity mud but high alkalinity mud can react at temperatures above 200oF); and• electrochemically – an increase in temperature increases the ionic activity of any electrolyte and the

solubility of any partially soluble salt that may be present in the mud.

FlocculationA mud consists of many colloid particles that have the ability to remain indefinitely in suspension becauseof their extremely small size. In pure water they cannot agglomerate (form a mass) because of theinterference between the highly diffuse double layers. But if an electrolyte is added, in this case salt, thenthe particles can approach each other so closely that the attractive forces predominate and the particles canthus agglomerate. This phenomenon is known as flocculation. If the concentration of clay in suspension ishigh enough, flocculation can cause the formation of a continuous gel structure. The gels commonlyobserved in aqueous drilling fluids are the result of flocculation by soluble salts.

Temperature effect on water lossAn increase in temperature could also work to decrease the plastic viscosity by decreasing the amount offree water bound to the clay and polymer particles. This effect can only occur at temperatures that are highenough to break the bond between the particles.

67

Appendix 8 Effect of cement additivesCement slurry has poor fluid retention properties and can lose a significant amount of water to a formationrapidly. Cement additives provide protection against water loss and help to control other properties of themud. They also:• vary cement density;• increase or decrease cement strength;• accelerate or retard setting time;• control filtration rate;• reduce or increase slurry viscosity;• provide bridge for lost circulation control; and• improve the economics of production.

Cement densityIn some cases long cement columns may be constructed without risking formation breakdown. Typicallythere are three methods that can be used in order to decrease the density of the cement slurry. The firstinvolves adding lightweight materials such as bentonite and attapulgite, which permit an increased mix withwater and at the same time prevent the separation of water. Secondly, it may be possible to add hollowceramic spheres that have a low specific gravity which can provide reduced density slurries. Thirdly, foamcements are sometimes used: for example, the addition of nitrogen plus a surfactant can create a mud with adensity of 7–8 lbs/gal.

Slurry viscositySometimes it is necessary to use friction-reducing additives to assist the cement slurry in removing annularmud. Also, viscosity of the cement mixture may be required in order to ensure that it does not invade theformation. Interparticle friction reducers are basically dispersing agents that can reduce the apparentviscosity of the slurry. Salt is an example of a friction reducer.

Control of filtration rateIn order to control the filtration rate it is helpful to form a very thin, tight filter cake. This can be achievedby adding materials such as bentonite and CMC to reduce the fluid loss from the cement slurry. Bentoniteand CMC both have the ability to bind water chemically to their polar sites on the clay platelets or polymermolecules. Such additives help to provide protection against water loss by binding the free water in theslurry. It is common to run tests on the additives to ensure that they will meet the requirements of thespecific situation.

Rheology of cement slurriesIt is important for a drilling operator to know the properties of the cement slurry. In most cases accuratepredictions of the friction pressure are known. These help to avoid fracturing the lower stressed formations.Which can lead to a loss in circulation. It is often difficult to determine realistic rheological properties ofcement slurries due to the variability of fluid properties.

68

REFERENCESAldrich, C.H. & Mitchell, B.J., 1975. ‘Strength, Permeabilities and Porosity of Oilwell Foam Cement’, ASME paper

75–Pet–10.

Andersen, L.B., Rokke, G. & Veire, G., 1998. ‘A New Method for Well Specific Risk Analysis in Drilling Operations’,SPE paper 46613.

API Task Group on Cements for Geothermal Wells, 1985. ‘API Work Group Reports Field Tests of GeothermalCements’, Oil & Gas J.

APPEA, 1996. ‘1996 Drilling Fluids Survey Report’, APPEA Ltd, Canberra.

Armstrong, K., Bruney, J., Campbell, J.A., Scrupholme, P.L. & Visser, J.P., 1998. ‘Principles for Social andEnvironmental Assessment’, SPE paper 46857.

Ataya, A.A. & Youssef, F.Z., 1993. ‘Cement Evaluation Tool and its Application to Communication Testing’, SPEpaper 15757.

Bannister, C.E., Shuster, G.E., Wooldridge, L.A., Jones, M.J. & Birch A.G., 1983. ‘Critical Design Parameters toPrevent Gas Invasion During Cementing Operations’, SPE paper 11982.

Bannister, C.E. & Lawson, V.M., 1985. ‘Role of Cement Fluid Loss in Wellbore Completion’, SPE paper 14433.

Baret, J., Michaux, M. & Nelson., E.B., 1990. ‘Cement Additives and Mechanisms of Action’. In: Nelson, E.B., (ed.),Well Cementing, Elsevier, New York.

Baret, J., Daccord, G. & Yearwood, J., 1990 ‘Cement/Formation Interactions’. In: Nelson, E.B., (ed.), Well Cementing,Elsevier, New York.

Barshad, I., 1955. ‘Adsorptive and Swelling Properties of Clay-water System’, Clay Technol. Bull., 169: pp. 70–77.

Baumgarte, C., Thiercelin, M. & Klaus, D., 1999. ‘Case Studies of Expanding Cement to Prevent MicroannularFormation’, SPE paper 56535.

Berg, E., Romijn, H. & Wolting, B., 1998. ‘Environmental Evaluation Systems for Well Drilling and CompletionChemicals – The Present Situation and Future Views’, SPE paper 46596.

Berra, M., Fabbri, F., Faceotti, M., Pezzouli, M., Ricciardulli, R., Romano, G. & Tarquini, B., 1988. ‘Behaviour of aCementing Hydraulic Binder Under Severe Geothermal Conditions’, Geothermics, 17: pp. 785–813.

Bessems, E., 1982. ‘The Prevention of H2S Produced by Sulphate-reducing Bacteria During Oil Production’, SPE paper10855.

Bettge, G.W., Bland, R.G. & Chang, C.Y., 1987. ‘Toxicity of Drilling Fluids Photobacteria and Mysid ShrimpBioassays’, SPE paper 16689.

Bleler, R., 1990. ‘Selecting a Drilling Fluid’, JPT, pp. 832–34.

Bradford, B.B., 1982. ‘Setting Cement Plugs Requires Careful Planning and Execution for Successful Cementing Job’,Oil and Gas J., 13 Dec. 1982, pp. 102–04.

Brady, M.E., Craster, B., Getliff, J.M. & Reid, P.I., 1998. ‘Highly Inhibitive, Low-Salinity Glycol Water-Base DrillingFluid for Shale Drilling in Environmentally Sensitive Locations’, SPE paper 46618.

Borghini, S. & Bartolomeo, M., 1998. ‘The Role of Environmental Costs and Liabilities in the Decision Making Processof Oil & Gas Exploration and Production Companies’, SPE paper 46654.

Bruckdorfer, R.A., Jacobs, W.R. & Masson, J.P., 1983. ‘CBL Evaluation of Foam-Cemented and Synthetic-CementedCasings’, SPE paper 11980.

Bruckdorfer, R.A., 1986. ‘Carbon Dioxide Corrosion in Oilwell Cements’, SPE paper 15176.

69

Burdylo, L. & Birch, G., 1990 ‘Primary Cementing Techniques’. In: Nelson, E.B., (ed.), Well Cementing, Elsevier, NewYork.

Callet, F., Guillot, D., Hendriks, H. & Vidick, B., 1990 ‘Mud Removal’. In: Nelson, E.B., (ed.), Well Cementing,Elsevier, New York.

Carter, G. & Smith, D.K, 1958. ‘Properties of Cementing Compositions at Elevated Temperatures and Pressures’,Trans. AIME, 213: pp. 20–27.

Cheung, P.R. & Beirute, R.M., 1982. ‘Gas Flow in Cements’, SPE paper 11207.

Chilingarian, G.V. & Vorabutr, P., 1981. Drilling and Drilling Fluids, Elsevier, New York.

Ch. Noik & Rivereau, A., 1999. ‘Oilwell Cement Durability’, SPE paper 56538.

Chuckwu, G.A., Hatzignatiou, D.G. & Talabani, S., 1993. ‘Comprehensive Description and Evaluation of Polymers asDrilling Fluids’, SPE paper 26071.

Darley, H.C.H. & Gray G.R, 1991. Composition and Properties of Drilling and Completion Fluids, Gulf PublishingCompany, Texas.

Davies, D.R. & Hartog, J.J., 1981. ‘Foamed Cement – A Cement With Many Applications’, SPE paper 9598.

Deeg, W.F., Griffith, J.E., Crook, R.J. & Benge, G., 1999. ‘How Foamed Cement Advantages Extend to HydraulicFracturing Operations’, World Oil, November 1999, pp. 51– 53.

Dees, J.M. & Spradlin, W.N., 1982. ‘Successful Deep Open-Hole Cement Plugs for the Anadarko Basin’, SPE paper10957.

Drecq, P. & Nelson, B.E., 1990 ‘Special Cement Systems’. In: Nelson, E.B., (ed.), Well Cementing, Elsevier, NewYork.

Drecq, P., Parcevaux, P. & Rae, P., 1990 ‘Prevention of Annular Gas Migration’. In: Nelson, E.B., (ed.), WellCementing, Elsevier, New York.

Duell, A., East, L.E., McIntyre, K.W. & Tucker, J., 2000. ‘Improved Zonal Isolation Through the Use of SealantsBefore Primary Cementing Operations’, SPE paper 59131.

Dusseault, M.B., Gray, M.N. & Nawrocki, P.A., 2000. ‘Why Oilwells Leak: Cement Behavior and Long-TermConsequences’, SPE paper 64733.

Economides, M.J., 1990. ‘Implications of Cementing on Well Performance’. In: Nelson, E.B., (ed.), Well Cementing,Elsevier, New York.

Ekrol, N., Johnsen, S., Reed, M. & Rye, H., 1998. ‘Accumulated Concentration Fields in the North Sea for differentToxic Compounds in Produced Water’, SPE paper 46862.

Ezzat, A.M., Rosser, H.R. & Al-Humam, A.A., 1997. ‘Control of Microbiological Activity in Biopolymer-BasedDrilling Muds’, SPE paper 39285.

Faul, R., Reddy, B.R., Griffith, J.E., Fitzgerald, R. & Waugh, B., 2000. ‘Next-Generation Cementing Systems to ControlShallow Water Flow’, OTC paper 11977 presented at the Offshore Technology Conference, Houston, 1–4 May2000.

Frisch, G.J., Graham, W.L. & Griffith, J., 1999. ‘Assessment of Foamed-Cement Slurries Using ConventionalEvaluation Logs and Improved Interpretation Methods’, SPE paper 55649.

Froelich, B., Dumont, A., Pittman, D. & Seeman, B., 1982. ‘Cement Evaluation Tool: A New Approach to CementEvaluation’, JPT, pp. 1835–41.

Gray, G.R., 1958. ‘Chemicals in Drilling Mud’, Oil & Gas J., pp. 91–98.

70

Griffin, T.J. & Rozieres, J., 1990 ‘Foamed Cement’. In: Nelson, E.B., (ed.), Well Cementing, Elsevier, New York.

Grim, R.E., 1968. Clay Mineralogy, McGraw-Hill, New York, NY, pp. 596.

Guyvoronsky, A.A. & Farukshin, L.K., 1963. ‘Hydrostatic Pressure of Cement Slurry’, Neftyanik, no. 10: pp. 30–32(translated from Russian).

Haberman, J.P & Delestatius. M., 1991. ‘Downhole Fluid-Loss Measurements from Drilling Fluid and Cement Slurries’,SPE paper 22552.

Harms, W.M & Febus, J.S., 1985. ‘Cementing of Fragile-Formation Wells with Foamed Cement Slurries’, SPE paper12755.

Iliyan, I.S., Cotton, W.J. & Brown, G.A., 1983. ‘Test Results of a Corrosion Logging Technique Using ElectromagneticThickness and Pipe Analysis Logging Tools’, JPT, pp. 801–08.

Jutten, J. & Morriss, S.L., 1990 ‘Cement Job Evaluation’. In: Nelson, E.B., (ed.), Well Cementing, Elsevier, New York.

Jutten, J. & Parcevaux, P., 1987. ‘Relationship Between Cement Bond Log Output and Borehole GeometricalParameters’, SPE/IADC paper 16139.

Kopp, K., Foreman, J., Carty, B. & Griffith, J., 2000. ‘Foamed Cement vs. Conventional Cement for Zonal Isolation –Case Histories’, SPE paper 62895.

Krilov, Z. & Loncaric, B., 2000. ‘Investigation of a Long-Term Cement Deterioration Under a High-Temperature, SourGas Downhole Environment’, SPE paper 58771.

Krueger, R.F., 1962. ‘Evaluation of Drilling-Fluid Filter-Loss Additives Under Dynamic Conditions’, SPE paper 451.

Le Roy-Delage, S., Baumgarte, C., Theircelin, M. & Vidick, B., 2000. ‘New Cement Systems for Durable ZonalIsolation’, SPE paper 59132.

Marca, C., 1990 ‘Remedial Cementing’. In: Nelson, E.B., (ed.), Well Cementing, Elsevier, New York.

Martinez, S.G. & Ruiz de Galarreta, V.A., 1998. ‘Evaluation of Hydrological Contamination in an Oil and GasProduction Area Including Different Anthopic Activities. Yacimiento Centenario, Neuqen. Argentina’, SPE paper46585.

Martucci, S., 1989 ‘Casing corrosion in the Moomba area, Cooper Basin, South Australia’. In: O’Neil, B.J., (ed.), TheCooper and Eromanga Basins, Australia.

Mumallah, N.A., 1987. ‘Chromium (III) Propionate: A Crosslinking Agent for Water Soluble Polymers in Real OilfieldWaters’, SPE paper 15906.

Nelson, E.B., 1990 ‘Thermal Cements’. In: Nelson, E.B., (ed.), Well Cementing, Elsevier, New York.

Outmans, H.D., 1958. ‘Mechanics of Differential-Pressure Sticking of Drill Collars’, Trans. AIME, 213: pp. 265–74.

Parcevaux, P.A., 1984. ‘Mechanisms of Gas Channelling During Primary Cementation – Methods for Prevention andRepair’, Chemische Produkte in der Erdolgewinnung, Clausthal-Zellerfeld, 6 Sept. 1984.

Parcevaux, P.A. & Sault, P.H., 1984. ‘Cement Shrinkage and Elasticity: A New Approach for a Good Zonal Isolation’,SPE paper 13176.

Pardue, G.H., 1963. ‘Cement Bond Log – A study of Cement and Casing Variables’, JPT, pp. 545–54.

PIRSA, 2000. ‘Statement of Environmental Objectives for Drilling and Well Operations in the Cooper/Eromanga Basin– South Australia’. Department of Primary Industries and Resources South Australia, Adelaide.

Robinson, W.W., 1940. ‘Chemicals in Drilling and Production Operations’, API Drilling Production Practices, pp.236–245.

71

Sasaki, S., Kobayashi, W. & Okabayashi, S., 1984. ‘Strength Development of 2CaO.SiO2 – Silica Cement Under HighTemperature and High Pressure Conditions’, SPE paper 13406.

Schlumberger, 2001. Personal correspondence with Schlumberger’s mud engineer.

Shen, J.C., 1989. ‘Effects of CO2 Attack on Cement in High Temperature Applications’, SPE paper 18618.

Sheives, T.C., Tello, L.N., Makir, V.E., Standley, T.E & Blankinship, T.J., 1986. ‘A Comparison of New UltrasonicCement and Casing Evaluation Logs with Standard Cement Bond Logs’, SPE paper 15436.

Smith, R.C., 1986. ‘Successful Primary Cementing: Fact or Fiction’, SPE paper 14135.

Smith, R.C., Beirute, R.M. & Holman, G.B., 1983. ‘Improved Methods of Setting Successful Whipstock Cement Plugs’,SPE paper 11415.

Sutton, D.L. & Ravl, K.M., 1989. ‘New Method for Determining Downhole Properties that Affect Gas Migration andAnnular Sealing’, SPE paper 19520.

Taylor, H.F.W., (ed.), 1964. The Chemistry of Cements, Academic Press Ltd, London.

Tinsley, J.M., Miller, E.C., Sabins, F.L. & Sutton, D.L., 1979. ‘Study of Factors Causing Annular Gas Flow FollowingPrimary Cementing’, SPE paper 8257.

Warren, B.K. & Smith T.R., 1993. ‘Static and Dynamic Fluid-Loss Characteristics of Drilling Fluids in a Full-ScaleWellbore’, SPE paper 26069.

Watters, L.T. & Sabins, F.L., 1980. ‘Field Evaluation of Method to Control Gas Flow Following Cementing’, SPE paper9287.

Webster M.A., Brew, J. & Grimison, A.G, 2000. ‘Hydrodynamics in the Queensland Sector of the Copper/EromangaBasins: Identifying Non-Conventional Exploration Plays Using Water Pressure and Chemistry Data’. SPE paper64281.

Webster, W.W. & Eikerts, J.V., 1979. ‘Flow After Cementing – Field and Laboratory Study’, SPE paper 8259.

Weeter, R.F., 1982. ‘Downhole Corrosion – Prevention and Treatment’, SPE paper 9984.

Wright T.R., 1978. ‘Guide to Drilling, Workover and Completion Fluids’, World Oil, June 1978, pp. 53–98.