Spring 19

Otto HouHemant LGerry MePaul ReidCambridg

Don WilliMontroug

A New Slogan for Drilling Fluids Engineers

For help in pthank Sarah BP ExploratiScotland; Linand Paul WaCambridge, VISPLEX is a

“Zero damage—good; permeability reduction—bad,” long a motto of drilling engineers, is accurate

for most vertical wells. Horizontal wells, however, with more exposure to producing formations,

are different. Laboratory work and reservoir simulation are helping write a more equivocal phrase:

“Zero damage—preferable; permeability reduction—better to avoid, but often allowable.”

wenadvaeten

e, England

amsone, France

1. Renard G and Dupuy JG: “Influence of FormationDamage on the Flow Efficiency of Horizontal Wells,”paper SPE 19414, presented at the 9th SPE FormationDamage Control Symposium, Lafayette, Louisiana,USA, February 22-23, 1990.Browne SV and Smith PS: “Mudcake Cleanup toEnhance Productivity of High Angle Wells,” paper SPE

One of the golden rules for vertical wells isthat formation damage caused by drillingshould, if possible, be eliminated in thereservoir. Generally, this concept has alsobeen extrapolated to horizontal wells andhas led to the adoption of aggressive tech-niques to clean up formation damage. How-ever, in some cases these complicated treat-ments increase rather than decrease risk tothe wellbore. Quantifying the effects of mudsystems and available cleanup techniquesmakes possible an informed choice, specificto the reservoir and well being drilled.

97

reparation of this article, we would like toBrowne, Michael Burnham and Dan Ryan,on Operating Company Limited, Aberdeen,dsay Fraser, Dowell, Houston, Texas, USA;y, Schlumberger Cambridge Research,England. mark of Schlumberger.

27350, presented at the SPE International Symposiumon Formation Damage Control, Lafayette, Louisiana,USA, February 7-10, 1994.Beatty T, Hebner B, Hiscock R and Bennion DB: “CoreTests Help Prevent Formation Damage in HorizontalWells,“ Oil & Gas Journal 91, no. 31 (August 2, 1993):64-70.

The clear objective for any well is that itshould perform to the full potential of the for-mation it penetrates and remain stablethroughout its lifetime. This goal is bestachieved by avoiding formation damage inthe first place, but in most cases this is notpossible. However, if damage is unavoidable,a correlation can be drawn by looking at frac-turing treatments in vertical wells. Duringfracturing jobs, wells may sustain near-well-bore damage similar to drilling-induced dam-age. But this damage may be largely ignoredbecause induced fractures extend thousandsof feet into the formation, exposing more ofthe reservoir to a conductive flow path andsignificantly improving productivity.

Horizontal wells penetrate up to 6000 ft[2000 m]—even more than most inducedfractures—into a reservoir, exposing thewellbore to an area of producing formationat least an order of magnitude greater thanwould be achieved with a vertical well. Thisopens up two opposing factors that drivehorizontal well productivity.

Because of their huge flow area, horizon-tal wells can withstand higher levels ofdamage than vertical wells and still deliverhigher production rates. Conversely, drillingtimes for horizontal sections are generallymuch longer than for vertical wells in thesame formation, giving drilling mud moretime to enter the formation and potentiallycausing more severe formation damage.Also, lower drawdown pressures in horizon-tal wells may reduce cleanup efficiency.1

Therefore, some reduction in permeabilitymay be permissible in horizontal wells, aslong as the wellbore extends far enough intothe formation to ensure sufficient flow area.At the same time, other aspects of thedrilling fluid, like its effect on well drillabil-ity, may be brought to the fore. The trick isknowing which drilling fluids to select tomaximize drilling rate while minimizing riskto the formation.

For too long, decisions about drilling fluidshave been made in isolation. Now, theindustry is developing a strategy that bringstogether the domains of the reservoir,petroleum and drilling engineer with that ofthe fluids engineer. At the heart of this workis the development of a real understandingof how drilling fluid damage affects produc-tivity, with the goal of developing a reservoirengineering tool for drilling fluid design.

3

Skin (damage zone)

Borehole

Static pressurePressure in formation

Dpskin = pressure drop across skin

Flowing pressure

■■Damage skin surrounding a wellbore. The skin factor may berepresented as a dimensionless pressure drop. The magnitude ofthis factor depends on the ratio of the undamaged and damagedpermeabilities in the formation, and on the depth of damage,which is related both to the depth of invasion and fluid loss.

Formation Damage—Invasion of the Production SnatcherFrom a mud standpoint, a well may bedivided into two sections. In the first—fromthe surface to top of the reservoir—the twokey drivers are health, safety and environ-mental (HSE) constraints, and drilling cost.In the second section—the reservoir—HSEconcerns remain of central importance, butthe cost factor is usually overshadowed by aneed to minimize formation damage. Ofcourse, a prerequisite in both sections is thatthe well be drillable with the mud of choice.

Formation damage is considered to beanything that impairs the permeability ofreservoir formations, reducing injectivity orhydrocarbon production. Damage canoccur during all stages of well construction,during remedial treatments and during pro-duction.2 This article concentrates on therelationship between drilling fluids and for-mation damage.

In reality, all reservoirs are damaged tosome extent by drilling fluid. The importantissue is whether this damage significantlyaffects well productivity. One way of quanti-fying formation damage is to use the dam-age skin factor (right). Typically, a poorlyconstructed damaged well will have a posi-tive skin of 20 to 500; a good unstimulatedwell will have a skin of plus five to minusunity; and a well that has been fracture stim-ulated will have a large negative skin.

Today, most vertical wells are completedusing a cemented liner that is then perfo-rated. This is not the case for horizontal wellsthat are most often completed barefoot—thatis open hole—or using prepacked sand con-trol screens, slotted liners or predrilled linerswhere drilling mud may have a greaterimpact on the productivity of a well. Thereare at least two reasons for this effect.

First, oil and gas must be producedthrough the filter cake and mud filtrate-induced formation damage because thereare no perforations that reach beyond thedamaged zone. Second, sand control com-pletions such as prepacked screens may alsobe plugged by the mud (see “How DrillingFluid Reduces Producibility in an OpenholeHorizontal Well,” page 6).

4

Drilling fluid selection for reservoir drillingin horizontal and high-angle wells is a com-plex process. Obviously, the choice of mudshould ensure that the well is drillable, anda wide range of well and formation factorsinfluence this selection. The effect ofdrilling fluid systems on factors such as holecleaning, torque and drag, wellbore stabil-ity, and stuck pipe is central to success orfailure (see “Stickance Tester: Predicting aMud’s Performance,” page 10).

The next and increasingly important factorinfluencing selection is the HSE aspect of adrilling fluid. Some fluids may not be usablein certain situations because of company orregulatory policy. Then come the cost andimpact of the drilling fluid on productivity.This article focuses on the interplay of thesefinal two factors.

Understanding the Real Effectsof a Drilling FluidWhile there appears to be a broad consen-sus on the mechanisms of formation dam-age, there is growing divergence over how itmay be combated or avoided. The need tocost effectively eliminate or at least mini-mize formation damage, so that productivityis maximized, has spawned a specializedarea of fluid design for reservoir drilling andushered in a host of what are called “drill-influids.” Most drilling fluid companies havedeveloped drill-in fluids to allow effectivecleanup following reservoir drilling.

One development has been the introduc-tion of mud systems with a solid phase,which makes up the filter cake, that maysubsequently be removed by washes orbreaker fluids circulated into the well beforecompletion to dissolve or partially break thefilter cake. Theoretically, these treatmentsreduce the pressure required by formationfluid to break through the filter cake once awell is put on production, ensuring an evenflow across the productive part of the hori-zontal interval. In practice, their action isnever uniform across the wellbore and suchtreatments substantially increase drillingcosts and complicate field operations.

Oilfield Review

VISPLEX filter cake (A)

Bentoniteand MMHbridge(D)

Internalfilter cake(E)

Rockgrain

Rockgrains (B)

Pore (C)

100 µ

100 µ

100 µ

■■Minimizing particu-late invasion. VISPLEXfilter cake (A) on theexternal surface of acore (top). Betweenrock grains (B),unblocked pores—onthe order of 30 microns(µ) wide—may be seenimmediately below fil-ter cake (C). At highermagnification, a film-like bridge of bentoniteand mixed-metalhydroxide (D) over thepore throat is high-lighted (center). Afterexposure to KCl-poly-mer mud, internal filtercake (E) is apparentbetween the rockgrains (bottom). Theunusual behavior ofthe VISPLEX fluid mayexplain the low levelof permeability impair-ment seen in labora-tory and field evalua-tions of this system.

2. Krueger RF: “An Overview of Formation Damage andWell Productivity in Oilfield Operations,” Journal ofPetroleum Technology 39, no. 2 (February 1986):131-152.

3. Fraser LJ, Williamson D, Enriquez F Jr and Reid P:“Mechanistic Investigation of the Formation Damag-ing Characteristics of Mixed Metal Hydroxide Drill-InFluids and Comparison With Polymer-Base Fluids,”paper SPE 30501, presented at the 70th SPE AnnualTechnical Conference and Exhibition, Dallas, Texas,USA, October 22-25, 1995.

(continued on page 8)

Saturated salt muds with salt crystals sizedto bridge across the formation and form a sig-nificant part of the filter cake are a typicalexample. After drilling, this cake is washedwith an undersaturated brine that dissolvesthe salt, promoting filter-cake cleanup. Alter-natively, calcium carbonate may be used asthe weighting and bridging agent in bothwater-base and oil-base muds. In this case,the filter cake may then be treated with amild acid to dissolve the carbonate. Also, cel-

Spring 1997

lulosic products that are frequently used forfluid-loss control or as bridging agents maybe dissolved—although often only partially—using dilute acids or oxidizing agents such assodium hypochlorite.

Enzyme breakers have been developed forsome water-base muds (WBM). Theseenzymes are designed to attack polymersand may be used alone or with one of the

above treatments. For example, sized-saltsystems incorporate magnesium peroxidethat when exposed to acid releases hydro-gen peroxide, which degrades polymers. Avariety of solvent and surfactant fluids isavailable to treat oil-base mud (OBM) filtercakes, breaking down the oil-wetting char-acter of the cake and allowing it to dispersein the aqueous, or mixed-phase wash fluid.As with polymer breakers, this treatmentmay also be used in combination withadditives that dissolve the cake.

These treatments are not without problems.Washes may cause significant losses of treat-ment fluid to the formation. These invadingfluids, and the resulting filter-cake residues,may cause significant additional formationdamage—the opposite of what is intended. Ifthe losses are severe, it will be necessary touse expensive and time-consuming lost-cir-culation treatments that may themselvescause damage. Also, severe losses couldeventually lead to well-control incidents.Treatment of some OBM filter cakes pro-duces viscous sludges that cause formationdamage. Polymer sludges may also resultfrom treatment of WBM filter cakes. Acidbreakers may cause corrosion problems.

An alternative is to do away with washesand breakers altogether and back-producethe drilling fluid through the completionhardware (see “Bringing in Wells Without aCleanup,” page 16). Another approach is tominimize particulate invasion of the forma-tion in the first place by creating a filter cakethat may be more easily “lifted” by forma-tion fluid during flowback. An example ofsuch a system is a bentonite/mixed-metal-hydroxide (MMH)/sized-carbonate system.MMH fluids are highly thixotropic, and lab-oratory tests show that they have a lowpotential for formation damage, layingdown a predominantly external filter cakeand thereby avoiding the need for deep-penetrating washes (above left).3

5

How Drilling Fluid Reduces Producibilityin an Openhole Horizontal Well

1. Francis PA, Eigner MRP, Patey ITM and Spark ISC: “Visualisa-tion of Drilling-Induced Formation Damage Mechanisms UsingReservoir Conditions Core Flood Testing,” paper SPE 30088,presented at the SPE European Formation Damage Conference,The Hague,The Netherlands, May 15-16,1995.

There are at least four different mechanisms for

drilling fluid to damage horizontal well producibility

both inside the formation and in the wellbore.

Mud-solids invasion and internal filter cake—

Drilling fluid solids will invade a short distance into

the formation—on the order of a few millimeters—

bridging across or plugging pore throats (next page,top left).1 An internal cake will restrict flow unless it is

removed by treatment or flushed out during produc-

tion. The damage potential of mud solids depends on

the size of particles relative to the size of the pore

throats in the formation being drilled. Shape, flexibil-

ity and degree of dispersion of particles are also

important. Possible exceptions are highly flexible

particles like bentonitic clays that can deform suffi-

ciently, allowing them to penetrate pores smaller than

the diameter of the clay sheets. As a guideline, parti-

cles between one-sixth and one-third of the diameter

of a pore throat may invade a significant distance

into the rock before bridging pore throats; particles

less than one-sixth of the pore-throat diameter gen-

erally do not bridge.

Mud-filtrate invasion—Mud filtrate may interact

chemically and physically with the formation causing

significant damage—for example, mobilizing forma-

tion fines or changing formation wettability due to

adsorption of mud surfactants onto the particles.

Formation-fines mobilization—When water-base

mud invades a rock containing oil, fines mobilization

following filtrate invasion may be triggered by a

salinity change, by a chemical deflocculant in the fil-

trate, or by high fluid-flow velocities in the pore

space. Migrating fines may cause extensive damage

by blocking pore throats. In most formations, mobile

6

particles ranging from 1 to 100 microns (µ) are

believed to be most damaging, since particles

smaller than 1 micron are normally strongly held to

the surfaces of larger mineral grains by Van der

Waals forces and are difficult to dislodge. Particles

above 100 microns are larger than most pore-throat

diameters and so cannot migrate any great distance.

It is usually difficult to carry out successful remedial

treatments to remove damage caused by formation

fines. Sometimes even these treatments can cause

fines to become mobilized—by dissolving intergranu-

lar cements—or can leave reaction products that are

themselves damaging (next page, top center).Changes in wettability—When oil-base mud filtrate

invades a water-wet formation, surfactants or certain

types of polymer in mud filtrate may change the wet-

tability of the rock. Displaced formation water forms

droplets in the pore spaces and thus affects hydrocar-

bon production. In fact, oil-wetting agents are specifi-

cally designed to make weighting agents and drilled

solids particles hydrophobic, so it is inevitable that, if

free surfactant enters the rock in the mud filtrate, the

rock is also likely to become oil-wet. Permeability

damage caused by wettability change is generally

assumed to be permanent. However, because of the

low fluid-loss rates of oil muds, the depth of damage

will often be small. Wettability change generally has a

greater influence on production in tight rocks that con-

tain small-diameter pores (next page, top right).Undisplaced whole drilling fluid—Large-scale flow

loop tests have shown that when screens are uncen-

tralized, mudcake and debris are left on the low side

of the hole even after aggressive cleanup. Whole fluid

left behind in the annulus can pack off on the com-

pletion hardware during production. Also, for wells

with low drawdown, the high gel strength of drilling

mud could prevent or restrict flow from part of the

horizontal section (next page, bottom right).

Drilling fluid damage to completion hardware—As

sand control completion hardware is run into the

well, it fills with the fluid in the well. Mud will flow or

filter through the screen as a result of surge pres-

sures created while running into the well. During this

process, solids in the mud may partially or com-

pletely plug the screen. Susceptibility to mud damage

will vary widely, depending on completion type—

prepacked screens are particularly vulnerable due to

internal plugging (next page, bottom left).Damage profile from a polymer mud—A polymer

mud may damage formation permeability in several

ways. For example, mud solids may invade and create

an internal filter cake; fines can be mobilized and

block pores inside the formation; and certain polymers

carried inside the rock may adsorb onto the rock and

change the wettability, while larger polymers can also

block pore spaces. Each of these processes invades to

a different depth, creating more or less damage. A

damage profile is more useful than a simple average

because it helps explain the consequences and mech-

anisms of invasion (next page, bottom center). In this

case, the damage profile decreases in severity away

from the wellbore. If a formation is not susceptible to

fines damage, then this graph will be different.

Oilfield Review

Spring 1997 7

Undamaged formation

Mud-filtrate invasion

Internal filter cakeExternal filter cake

Completion plugging

Sand-control screen

Changes in wettability

Internal filter cake Mud-filtrate invasion

Mud-filtrate invasion

Mud-solids invasion Formation-fines mobilization

Potential mud plugging Sand-control screen

Residual drilling mud

formation

Damage profile from a polymer mud

Invasion depth, m

Depth of filtrate invasion

Undisplaced whole drilling fluidDrilling-fluid damage to

completion hardware

............

0 0.5 1

100

80

60

40

20

0

Perm

eabi

lity

redu

ctio

n, %

Overalldamage

Polymerdamage

Mud-solidsdamage

Formation-finesdamage

.......................

........................

............

............

............

............

............

............

........................

...........

........................

............

............

............

............

............

............

........................

............

............

............

........................

............

............

............

............

........................

Per

man

ent p

erm

eabi

lity

dam

age,

% Indicates baseline test without breaker

Breaker decreased permeability damage

Breaker increased permeability damage

100

80

60

40

20

0

OBM WBM

OBMA OBMB OBMC WBMA WBMB WBMC

■■The effects of washes on permeability damage. A joint industrystudy showed that, in some cases, washes significantly reducedamage levels; in others, washes increase damage.

Bre

akth

roug

h pr

essu

re, p

si

OBM WBM

25

20

Indicates baseline test without breakerBreaker increased breakthrough pressureBreaker decreased breakthrough pressure15

10

5

0OBMA OBMB OBMC WBMA WBMB WBMC

■■The effects of washes on breakthrough pressure. Data collectedby a joint industry study reveal a wide variation between theeffectiveness of different mud systems when washes are used toremove or destabilize a filter cake. However, contrary to expecta-tions, more breakers were found to increase breakthrough pres-sure than reduce it.

The continuing debate on the pros andcons of washing versus back-production offilter cake in openhole completions has, atleast in part, been driven by the philoso-phies of individual companies. However,new studies into cake properties and dam-age mechanisms are now providing betterinformation for decisions.

For example, a joint industry study of mudcleanup in horizontal wells fully examinedthe role of common washes and breakers.4Small-scale core experiments tested six mudsystems and various breakers. Surprisingly,more of the breakers increased the backflowpressure required to break through the filtercake than reduced it (top). In no case was

8

all filter cake removed. In displacement flowtests, effectiveness of washes on the lowside of horizontal wellbores also proved tobe limited because of the presence of stag-nant whole mud, and large amounts ofresidual mudcake and debris.

Another role of breakers is to removemud-induced damage from the near-well-bore region. In this case, performance var-ied for different mud systems—significantlyreducing near-wellbore damage for somemuds and increasing damage in others(above). For some mud systems there was a

correlation between treatments that inducehigh losses and high levels of damage withthe wash fluids carrying damaging materialsuch as fines or partially degraded polymerdeep into the formation.

The high cost of specialized drill-in fluidsmeans that attempts to drill a well with zeroskin may take up a significant proportion ofwell budgets. However, any savings in themud cost have to be weighed against therisk of reducing productivity rather than pre-serving it. There is much to be gained bydefining the optimum amount of formationdamage that may be tolerated for a givenwell in a given situation.

But what is the optimum skin factor for awell? The answer is not simple. In some set-tings a considerable skin factor may have lit-tle effect on flow. The joint-industry studyreferred to above confirms that some hori-zontal wells can tolerate a significant levelof mud damage before productivity is signif-icantly impaired. In others, only low skinfactors may be tolerated, but these condi-tions cannot be achieved economicallyusing available mud systems. Reservoircharacteristics, well profile, completiondesign and economics all dictate the opti-mum skin factor.

A further determinant is the future role ofthe well. In an exploration well, where theobjective is to find rather than producehydrocarbons, a moderate skin may beacceptable. However, in a marginal devel-opment with a limited number of wells andtight margins, low skin may be ofparamount importance. Many high-anglewells are targeted to intersect multiple sandbodies. For these wells, the main objectiveis ensuring that all potentially productivesections of the well may flow so thatreserves access is maximized. Other wellsare drilled truly horizontal to maintain aconstant standoff with gas or water. Themain driver in these wells is an even draw-down to minimize coning.

Therefore, distribution of the damage isalso important. As part of an extensive study,BP confirmed that the percentage of theinterval flowing, and distribution of theflowing intervals over the length of a hori-zontal well may have a larger impact on

Oilfield Review

0 20 40Near-wellbore permeability reduction, %

Flow

effi

cien

cy

60 80 100

Depth of filtrate damage1 ft2 ft3 ft

50% of formation flowing, but no flow from lower half of well

50% of well not flowing, but from a number of evenly spaced intervals

0 20 40Percentage of intervals contributing to flow

Flow

effi

cien

cy

60 80 100

Number of flowing intervals

32821

60% filtrate damage to 2 ft of invasion

Spring 1997

■■How distribution offlow affects flow effi-ciency. The first wellschematic (top) illus-trates 50% of the for-mation flowing froma single interval inthe heel of the well.The second wellschematic (middle),also shows only 50%of the well flowing.However, this time,flow is divided intosix evenly spacedflow intervals acrossthe length of thewell. The graph (bot-tom), based on datagathered by BP,shows how increas-ing the number offlowing intervalsincreases the flowefficiency of a welleven though thetotal percentage ofthe well contributingto the flow remainsconstant.

■■The impact ofnear-wellbore per-meability reductionon flow efficiency. Asmall reduction inthe near-wellborepermeability—inthis case up toabout 30%—has lit-tle effect on flowefficiency and thedifferences in depthof damage are notsignificant. How-ever, when perme-ability reductionreaches 60% and upto about 80%, theeffect on flow effi-ciency becomes pro-found and the differ-ences in depth ofdamage becomemore marked.

productivity than the reduction in perme-ability around the well (left).5 This work,carried out in Sunbury, England, pro-duced three key findings:• If a given percentage of filter cake is

removed to allow a well to flow, it isbetter for this percentage to be dis-tributed over a large number of smallerintervals, instead of having all the flowconcentrated in a single, large interval.

• The cleanup need not be complete.Rather than remove the filter cake,increasing its permeability to at least0.1 md is sufficient—filter-cake perme-abilities are typically 10-2 to 10-6 md,depending on fluid type, differentialpressure and solids content.6

• Damage by deep invasion of filtrate—on the order of feet—causes only asmall reduction in productivity as longas the reduction in permeability is nottoo great (below left).Studies such as this one by BP illustrate

a central truth. There is no single besttechnique for the cleanup of unce-mented horizontal wells. The comple-tions engineer has a range of options thatmust be assessed for each field and eachwell strategy. The only way of knowingwhich is best is to understand the drillingfluid and its interaction with the forma-tion and completion hardware. Practicaloptions will vary depending on issuessuch as environmental legislation, opera-tional risk or logistics—for example, acomplicated wash strategy may not bepossible if there is insufficient storagecapacity on the rig. There is also, quiteclearly, no guarantee of success.

Thus, although the objective of anydrilling fluid design should be to delivera well with no formation damage,drilling and production practicesinevitably lead to some damage that maynot be removable. But if the well stillproduces to its full potential, this damagecould be termed “affordable.” As yet, thisconcept of affordability is not widelyreflected in industry practices.

9

4. Ryan DF, Browne SV and Burnham MP: “MudCleanup in Horizontal Wells: A Major Joint IndustryStudy,” paper SPE 30528, presented at the 70th SPEAnnual Technical Conference and Exhibition, Dallas,Texas, USA, October 22-25, 1995.The work was undertaken as a joint industry projectby Amoco, BP, Chevron, Norsk Hydro, Saga, Shell,Statoil and TBC Brinadd.

5. Early work in this field is reported in:Goode PA and Wilkinson DJ: “Inflow Performance ofPartially Open Horizontal Wells,” paper SPE 19341,presented at the SPE Eastern Regional Meeting, Mor-gantown, West Virginia, USA, October 24-27, 1989.

6. For example, in a typical wellbore, a cake of 0.1-mdpermeability and thickness of 3 mm gives a skin of 5;a cake of 0.01 md gives a skin of 56.

(continued on page 11)

Filtrate

Pressure

Filter cake

Drillingfluid

Torque

Stickance Tester: Predicting a Mud’s Performance

■■Stickance tester. The body of the device is a double-ended, high-temperature, high-pressure (HTHP) mud filtra-tion cell. The top end cap has been modified to allow the entry of a spring-steel wire through an o-ring seal set inthe center of the cap. A new entry port has been drilled to allow the cell to be pressurized. Inside the cell, the steelwire is connected to a 1.5-in. [3.8-cm] polished steel ball that rests on the filter medium at the bottom of the cell.The end of the wire protruding from the cell is attached to an electronic torque gauge.

Stuck pipe during drilling operations is a major non-

productive cost to the industry.1 Stuck-pipe incidents

are generally divided into two main categories:

mechanical and differential sticking. Which of these

problems is more dominant depends on where

drilling is taking place. In the North Sea, mechanical

sticking is the main problem; in the Gulf of Mexico, it

is differential sticking.

Mechanical sticking includes a large number of

mechanisms, including hole collapse and key seat-

ing. Differential sticking is the most common single

mechanism and occurs when part of the drillstring

becomes embedded in the mud filter cake and is

then held there by hydrostatic pressure, which

exceeds the formation pressure. As such, it can occur

only where a filter cake has been established—

across permeable formations. The pipe usually

becomes stuck when it is stationary adjacent to a

permeable zone and there is a significant mud over-

balance. The likelihood of differential sticking

increases with the length of permeable section being

drilled—making extended-reach and horizontal wells

particularly vulnerable.

When it comes to preventing differential sticking, the

nature of the rock cannot be changed. High overbal-

ance pressures may also be needed to maintain well

control or wellbore stability. However, it is possible to

modify mud composition and properties.

Recently, a better understanding of differential

sticking led to the development of a new laboratory

test tool to help design mud systems that avoid dif-

ferential sticking. Work carried out by researchers at

Schlumberger Cambridge Research, Cambridge, Eng-

land has concentrated on the nature of mud filter

cake—in particular thickness, lubricity and strength.2

A true measure of filter-cake properties is not cur-

rently included in the suite of standard American

Petroleum Institute (API) measurements routinely

10

carried out on drilling fluids. Although additional tests

do exist, SCR researchers have developed a new

technique to measure filter-cake properties that can

be related to a fluid’s propensity to encourage differ-

ential sticking. The technique is designed as a low-

cost, simple test that may be carried out at wellsites.

A high-temperature, high-pressure (HTHP) filtra-

tion cell was converted to create a stickance tester

(above). In this test, a filter cake is built up around a

polished steel ball inside the cell. The force needed

to rotate the ball is used to quantify the nature of a

filter cake.

Oilfield Review

300

200

100

0

0 5 10 15 20

Torq

ue, N

. m

Time, min 3/4

Stickance

Spring 1997

■■Typical plot in which thestickance is given by theslope. Good reproducibilityhas been achieved as longas consistent operatingpractices are employed.

1. Bailey L, Jones T, Belaskie J, Orban J, Sheppard M, Houwen O,Jardine S and McCann D: “Stuck Pipe: Causes, Detection andPrevention,” Oilfield Review 3, no. 4 (October 1991): 13-26.

2. Reid PI, Meeten GH, Way PW, Clark P, Chambers BD andGilmour A: “Mechanisms of Differential Sticking and a Simple Well Site Test for Monitoring and Optimizing DrillingMud Properties,” paper IADC/SPE 35100,presented at the 1996 IADC/SPE Drilling Conference,New Orleans, Louisiana, USA, March 12-15,1996.

Quantifying Affordable DamageTo make sense of the notion of affordability,it is necessary to understand the conse-quences of damage. Although new produc-tion logging techniques are being devel-oped, it is still difficult to extract fromhorizontal well tests all the informationneeded to make the required judgements.7Therefore, the productivity effects of forma-tion damage caused by drilling fluid inva-sion—or indeed the magnitude of the dam-age itself—are usually unquantified.

The need to close this knowledge gap hasbeen addressed in work carried out byresearchers at Schlumberger CambridgeResearch (SCR), Cambridge, England. Usingcore-flood experiments, they are determin-ing the formation-damage effects of drillingfluid invasion. Data from these experimentsare then used in accurate reservoir simula-tions that model the effects of this damageon productivity (see “How Core-Flood TestsAre Carried Out,” next page).

New analytical expressions have beendeveloped that relate damage to the produc-tion potential of the formation. From themud, all information on filtration, invasionand cleanup is channelled into the calcula-tions through the skin factor. Formationdamage expresses itself through large posi-tive skin values and hence lower productiv-ity index (PI) values and lower flow efficien-cies—that also take account of wellgeometry, formation thickness, permeabilityanisotropy, reservoir location, length of thewellbore and proximity of other wells.8

To help determine the return, in terms ofPI, from an incremental improvement in theperformance of a mud, numerical simula-tions using data generated by these analyti-cal expressions model the effects of damageon well producibility. These simulationsassess the implication of damage on reser-voir producibility, the implications ofincomplete penetration of the damage if awell is to be perforated (having assessed thedepth of damage from cores), and effects ofincomplete filter cake removal if a well isnot perforated.

7. Bamforth S, Besson C, Stephenson K, Whittaker C,Brown G, Catala G, Rouault G, Théron B, Conort G,Lenn C and Roscoe B: “Revitalizing Production Log-ging,” Oilfield Review 8, no. 4 (Winter 1996): 44-61.

8. Flow efficiency is defined as the flow rate with skindivided by the flow rate without skin, at the samedrawdown pressure.

(continued on page 14)

A test is carried out by placing the filter medium

representing a permeable formation in the cell. The

filter medium is usually filter paper, although cores,

sand packs and simulated fractured formations may

also be used in future versions of the device. The cell

is filled with drilling fluid, the top end cap is installed

and the ball and torque gauge are set in position. The

cell is then placed in a standard HTHP heating

jacket. The mud is heated to the desired temperature

and then pressurized as if a normal HTHP fluid-loss

measurement were being made—typically a differ-

ential pressure of 500 psi [3445 kPa] is used.

As filtration proceeds, a filter cake is built up on

the filter medium and around the steel ball. At pre-

cisely noted intervals—about every 5 minutes—the

torque gauge is rotated and the force needed to free

the ball from the filter cake is measured. This mea-

sures both adhesion of the ball to the cake and the

force needed to break this bond. Torque data are plot-

ted as a function of the three-quarter power of time

(t3⁄4) to account for the buildup of filter cake around a

spherical object. This plot usually gives a straight

line, the slope of which is the differential sticking ten-

dency—stickance (above).Using this apparatus, SCR researchers have

established mud formulation and engineering guide-

lines to reduce the risk of differential sticking. Fur-

ther, treatment options for field muds have been

investigated to help avoid sticking. The stickance

tester is now being prepared for deployment in field

laboratories so that these services may become gen-

erally available.

11

HTHP cell body

Standard HTHPend cap

Valve stem

Reservoir connection, drain tap anddigital pressure gauge

Pressure drain

Reservoir of permeabilitytest fluid

Air or gassupply

Regulator

How Core-Flood Tests Are Carried Out

■■Core-flood equipmentfor initial permeabilitymeasurements. Detail ofthe core holder showshow the rock sample islocked into place.

Over the years, various core-flood experiments have

been performed to assess formation damage by

using equipment that measures permeability changes

in rock cores before and after exposure to drilling

fluid. Researchers at Schlumberger Cambridge

Research (SCR), Cambridge, England have carried

out extensive tests, building up a wide range of data.

The experimental procedure may be divided into

three stages:

Stage One: Sample Preparation and Initial Perme-ability Measurement. The equipment at SCR tests

rock cores that are 25 mm in diameter and up to 32

mm long. Cores are placed under vacuum to remove

entrapped air and then saturated in brine or simu-

lated formation water—this may be unnecessary if

well-preserved reservoir core is used.

Once prepared, the core sample is firmly mounted

in the sample holder so that there is a seal between

the rubber sleeve and the core. The core holder is fit-

ted into a standard high-temperature, high-pressure

(HTHP) fluid-loss cell body that is then filled with the

Core holder fittedto standard HTHP

cell end cap

Rubber sleeve

Extended valve stem

Rock core

PTFE ringSteel locking ring

Detail of core holder

Locking bolt holes

Standard HTHP base

Core holder base

12 Oilfield Review

Motor

Motor/stirrercouplings

Stirrer body with seal and bearings

Cooling water coils

Locking nut for paddle

Paddle stirrer

Core sample in holder

Valve stem

Rubbersleeve

Cell body

Valve stem for pressuring cell

Fluid collection linked to data logger

■■Schematic of equipment for initial permeability mea-surements with the stirrer installed for dynamic filtra-tion tests.

test fluid to be used for the permeability measure-

ment—generally kerosene, crude oil or brine. Finally,

a standard HTHP end cap is secured in place.

The valve stem at the top of the cell is then con-

nected to a 2.5-liter [0.7-gal] reservoir of test fluid

that may be pressurized. The fluid passing through

the rock is collected and its volume logged as a func-

tion of time (previous page).Permeability measurements are made by opening

the valve stem at the top of the cell to pressurize the

fluid inside. The valve stem at the base is opened to

start flow through the sample, and the data logging is

started.

Test fluid is allowed to flow through the sample at

a fixed pressure. The volume of fluid collected versus

time is logged until a constant flow rate is reached,

indicating that the core has reached residual water

saturation. Experience has shown that for most rocks

this constant rate is reached when approximately 100

pore volumes have passed through the core.

At the end of the measurement, flow is stopped by

opening up the regulator and locking off the valve

stem at the base of the cell. After the fluid reservoir

and cell are depressurized, the top end cap is

removed and the cell is emptied of fluid in prepara-

tion for the mud-filtration phase.

Data may now be combined with fluid viscosity

and core size to calculate sample permeability:

Permeability = flow rate x fluid viscosity x sample length .

cross-sectional area x pressure

Spring 1997

Stage Two: Core Exposure to Test Fluid in a Staticor Dynamic Filtration Environment. Filtration—estab-

lishing a filter cake—may be performed under either

static or dynamic mud flow. The filtration phase may

be set for a specified period of time or until a prede-

termined volume of filtrate is collected and may be

performed at temperatures up to 150°C [302°F] and

pressures to 550 psi [3790 kPa].

To perform filtration under static conditions, the

cell is filled with 200 mL mud, the standard end cap

is refitted and a predetermined pressure differential

is applied from a gas source. As with the permeabil-

ity measurement, the volume of fluid collected is

logged as a function of time. Test conditions are var-

ied to mimic reservoir temperature and expected

mud overbalance pressure.

To perform a dynamic filtration test, a paddle stir-

rer is installed in the cell a fixed height above the

core after the mud has been poured into the cell. The

cell is then made up and placed back into the HTHP

heating jacket and the paddle is rotated. Finally, fil-

tration is restarted. Once again, filtrate volume is

recorded as a function of time (right).This stirrer generates a range of flow conditions

from turbulent, where little or no external filter cake

forms, to laminar, which leaves filter cakes similar to

those formed under static conditions. At the end of

the filtration phase, the cell is depressurized before

rotation of the paddle is stopped to ensure that no fil-

tration occurs under different operating conditions.

13

14

160

120

80

40

0

0 50 100 150 200 250 300 350 400

Test

flui

d co

llect

ed, g

Time, min

■■Flow before and after fil-tration.Typically the initialflow (blue) quicklyreaches a constant, whilethe return flow (red) maytake a significant time tostabilize as damagecaused by the drilling fluidmay be cleaned up tosome degree before asteady state is reached.

Valve stem

Resevoir connection, drain tap anddigital pressure gauge

Pressure drain

Reservoir of permeabilitytest fluid

Air or gassupply

Regulator

HTHPcell invertedto enableflowbackthrough core

Extended valve stem

■■Simulating produced fluids flowing through a damaged reservoir.

e

To understand how much damage is toler-able, the undamaged PI of a well must beknown. Herein lies a major snag. Theundamaged PI in horizontal wells is oftenunknown because it is difficult in horizontalwell tests to acquire reliable data for theproductive length of the well and the dam-age skin factor. This difficulty is due to well-bore storage—where fluid compressibilitymasks pressure changes—and the shortduration of early-time radial flow fromwhich skin is calculated (next page, top).

Although a sensible baseline for well pro-ductivity simulations should ideally bedrawn using existing horizontal wells in thesame field, the data uncertainty outlinedabove renders this sort of reference informa-tion unreliable. Therefore, since understand-ing the PI of vertical wells is more straight-forward, SCR researchers use a vertical wellin the same formation as a reference.9

Starting from the influence of formationthickness and anisotropy on the skin factor,researchers derived relationships that com-pare the flow efficiency of a horizontal wellwith that of a vertical well fully penetratingthe same producing formation.10 A novelexpression has been derived that calculatesthe length of horizontal section required tocreate a well with the same skin factor asthe vertical reference well. The expressioncombines all the geometric, reservoir, andformation damage information necessary toassess effects on flow efficiency of the hori-zontal well.

The degree to which an increase in skinaffects productivity of a horizontal welldepends on its drainage area, which intro-duces the concept of “neutral skin.” At neu-tral skin, production from both the horizon-tal well and its vertical reference is equallyimpaired. With a skin value below neutral,production from the vertical well is dispro-portionately reduced compared to its hori-zontal “sister” well. With skin greater thanthe neutral value, the horizontal well suffersa larger proportional production decrease.

Altering the horizontal well—for examplemaking it longer or increasing the drainageradius—may mitigate this effect, and advan-tages of a horizontal well over a verticalequivalent may be enhanced (next page,bottom). This knowledge helps establish theminimum length or drainage area requiredfor a horizontal well. For a given geometry,sensitivity of a horizontal well to skin can beassessed and thus the level of affordabledamage inferred.

9. Renard G and Dupuy JG, reference 1.10. Even if a vertical well has not been drilled, an

approximation of its PI may be estimated using available reservoir information.

Stage Three: Return-Permeability Measurement.Following filtration, another measure of core perme-

ability is made to determine the level of formation

damage caused by the mud. The stirrer is removed

and any remaining mud is poured away. The cell is

then filled with test fluid, and the end cap is fitted

and sealed with a valve stem. The cell is inverted

and replaced in the stand, reversing the direction of

test fluid flow through the core—the equivalent of

producing the formation (top). The same pressure is

used as in the initial permeability measurement,

although there is often a significant time delay befor

a steady flow rate is reached (above). The change in

permeability before and after filtration may then be

calculated.

Oilfield Review

Skin

2.5

2.0

1.5

1.0

0.5

0.0

0 5 10 15 20 25

Drainage radii

1000 ft

2000 ft

4000 ft

Rat

io o

f los

t pro

duct

ion,

(q H/q

V) lo

st

Spring 1997 15

■■Ratio of lost production from horizontal and vertical wells as a result ofdamage skin factors plotted for three similar wells with different drainagearea radii: 1000, 2000 and 4000 ft. For a given well, the neutral skin value isfound at the intersection of the curve describing (qH/qV)lost as a function ofskin with the horizontal line (qH/qV)lost =1. Below this line, the incrementaleffect on flow rate of increasing skin will be greater for the vertical referencewell than for the horizontal well. Above the line the opposite is true andincreasing skin will have a more deleterious effect on the horizontal wellthan on the vertical well. This effect is mitigated by increasing the drainageradius of the well, as can be seen from the graph, where a well withdrainage of 4000 ft remains below neutral for greater skin factors than doequivalent wells with smaller drainage radii. Therefore, placing many hori-zontal wells together in close spacing—thus reducing the horizontal drainageratio—increases the susceptibility of individual wells to formation damage.

• Early-time radial flow is the first radial flow period (in the verti-cal plane), which ends when the effect of the top or bottomboundary is felt. For horizontal wells, this regime is short and diffi-cult to identify because of wellbore storage effects. This is unfortu-nate as it is the only regime in which formation skin damage maybe deduced directly from a well test.

• Intermediate-time linear flow develops if the well is sufficientlylong compared with reservoir thickness—where the spread offlow beyond the ends of the well is negligible compared to itslength. If the well is not long, there will be a long transitionbetween early-time radial flow and the next regime, bypassingthis one.

• Late-time radial flow is the second radial flow period (in the hori-zontal plane) that develops if the reservoir is sufficiently large andwide compared to the length of the well. The well behaves like apoint source in the middle of the formation.

• Late-time linear flow—the second linear flow period—beginswhen the pressure transient has reached all lateral extremities.

■■Consecutive flow regimes observed for horizontal wells.

10099%

70%

36%

12%

Originalemulsifier

formulation

Form

atio

n da

mag

e, %

Emulsifiersystem

replaced

Emulsifierconcentration

reduced

Field sample

80

60

40

20

0

■■Reducing the impact of an oil-base mud. Core-flood testing, carried out by BP withreservoir core under downhole conditions, showed a 99% permeability impairment. Laboratory work revealed an incompatibility between the synthetic-base oil and theemulsifier that caused precipitation in the rock pore throats. Changing the emulsifier and then reducing its concentration cut permeability impairment to 70% and 36%,respectively. In fact, the field sample showed only a 12% reduction in permeability,which had a negligible effect on well productivity.

Bringing in Wells Without a CleanupThree possible options exist when dealingwith the drilling fluid prior to completing ahorizontal well with an uncemented liner orscreen:• displace mud with a low-solids or solids-

free, clear fluid• displace mud with a breaker system• design the mud already in the well to flow

harmlessly through completion equipmentand bring the well into production with-out cleanup.For its part, BP has begun to use comple-

tions programs that, where practical,employ the latter option—back-producingdrilling fluid through prepacked screens.11

Integral to this strategy is a desire to elimi-nate the complications and expenses associ-ated with the other two strategies, whileavoiding any chance that a breaker mightactually decrease rather than increase per-meability. By opting for simplicity, the com-pany reasons it is cutting risk.

16

11. Browne SV, Ryan DF, Chambers BD, Gilchrist JMand Bamforth SA: “Simple Approach to the Cleanupof Horizontal Wells with Prepacked Screen Comple-tions,” paper SPE 30116, presented at the SPE Euro-pean Formation Damage Conference, The Hague,The Netherlands, May 15-16, 1995.

However, BP is reducing risk only if thedrilling fluid system can effectively flowthrough the prepacked screens without leav-ing permanent damage. Also, the well mustflow, lifting sufficient filter cake to enable fullproductivity. Critical to both of these objec-tives is quality control of the drilling fluid inthe field—ensuring that it meets specifica-tions established by laboratory work.

From this, BP has drawn up a series ofguidelines. For example, solids loading mustbe below a critical level to avoid any log-jam effect that could occur as the fluidpasses through the screen; particle size dis-tribution must be carefully controlled as justa few percent of large particles bridging inthe screen may allow the many smaller par-ticles to form an impermeable cake; andparticle cohesiveness must be limited aseven fine particles—such as weightingagents—may agglomerate into much largerparticles. The total volume of mud that willbe flowing per unit area of screen should becalculated and an excess used in the labora-tory tests; the field mud actually used to drillthe horizontal section should also be testedon the screens (above).

At the same time, the drilling fluid must notdamage well productivity by either com-pletely stopping flow from any reservoir sec-tion or significantly increasing near-wellborepressure drop and thus reducing well PI. Forthis reason, the effects of mud filtrate on theformation and the backflow pressurerequired to break through the mud filtrateand establish production must also be tested.

For BP, tests like these now form an inte-gral part of developing the overall well com-pletion plan. In essence there are threecomponents in a design loop: mud systemoptimization to fit the reservoir; cleanupstrategy to ensure selection of the simplesttechnique that leaves no significant mud-related productivity impairment; and sandcontrol screen specification to best accom-modate anticipated downhole needs.

Completing the PictureMany of the steps described above are notnovel. What is new is a much clearer accep-tance that drilling fluid design is one part ofa much bigger process. To understand howa reservoir will perform implies deep knowl-edge of the number and type of wellsneeded; their length, angle and completiontype; and how they will perform—includinganticipated pressure drawdown and waterconing. Drilling fluid design is an integralpart of all of these.

There is a wide range of available fluids. Toselect the right one means that the mud anddrilling engineers must talk to many otherspecialists—reservoir geologists, productionchemists, drillers, completions engineers andlogging engineers—to establish their objec-tives. The task then is to choose a drillingfluid that, in addition to meeting HSE needsand achieving the primary objective of ensur-ing the well can be drilled, helps achievethese shared objectives. In the end, thismeans delivering a well that has sustained nomore than an acceptable level of damage.

The key is knowing what this acceptablelevel is and how a given mud will affect agiven formation in a given drilling situation.This need for understanding has been drivingresearch at BP Sunbury, SCR and elsewhere.The final piece needed to complete the pic-ture is an assessment of actual results overthe lifetime of wells. This process is only justbeginning, but when complete, drilling engi-neers will know that although zero damageis preferable for horizontal wells, permeabil-ity reduction is sometimes allowable.

—CF

Oilfield Review

The Bad Guys and the Good Guysin Petroleum Microbiology

Microorganisms make up 15 of the

24 subdivisions of life on earth—the

animal kingdom occupies just one.

With so much diversity, it is not

surprising that microbes can even

be found in high-pressure, hot,

anaerobic oil wells. Until recently,

the effects—both undesirable and

beneficial—of these organisms on

reservoirs were largely ignored.

This attitude is gradually changing,

however, and bacteria are now being

harnessed to improve recovery.

Catherine BassHilary Lappin-ScottUniversity of ExeterExeter, England

■■Bacteria: diverse and adaptable.

1. For further background reading on this topic: Hurst CJ, Knudsen GR, McInerney MJ, StetzenbachLD and Walter MV (eds): Manual of EnvironmentalMicrobiology. Washington, DC, USA: American Soci-ety of Microbiology, 1996. Costerton JW, Lewandowski Z, Caldwell DE, KorberDR and Lappin-Scott HM: “Microbial Biofilms,”Annual Reviews of Microbiology 49 (1995): 711-745.Bass CJ, Webb JS. Sanders PF and Lappin-Scott HM:1996. “Influence of Surfaces on Sulphidogenic Bacte-ria,” Biofouling 10 (1996): 95-109.

There is more diversity among microorgan-isms such as bacteria than there is in therange of life from artichokes to zebras. Bac-teria have successfully colonized virtuallyevery environment on earth because theycan rapidly adapt to changing conditions,and use a large and varied number of nutri-ents to generate energy. However, until rela-tively recently, the environments of many

Spring 1997

For help in preparation of this article, thanks to Chris Hall,Schlumberger Cambridge Research, Cambridge, England;and Jonathan Getliff, Dowell, Cornwall, England.

Campbell A: “Reservoir Biogenics and its Applicationto Improved Oil Recovery,” Emerging Technology Sta-tus Review. Edinburgh, Scotland: Petroleum Scienceand Technology Institute ( PSTI), October 1996.

petroleum reservoirs were considered toohostile for bacterial growth due to low avail-ability of water, and high temperatures,pressures and salinities (above).1

Despite pioneering work during the 1930sand 40s in the USA by Claude Zobell,demonstrating a rich population of bacteriain water produced from shallow hydrocar-bon reservoirs, the possibility of bacteriaexisting in larger, deep reservoirs was largelyignored. The start of North Sea production inthe 1960s led to the realization that bacteriacould produce hydrogen sulfide [H2S] as awaste product and cause reservoir souring.

17

Produced water reinjectedor dumped overboard

High population of activebacteria on seabed

A high number of activebacteria can cause corrosion.

Bacteria can form biofilm oninjection pipe and rock nearthe bottom of injection wells.

Injection water from thesea or produced water

Dormant bacteria inseawater where nutrientsand temperatures are low

In subsurface formations,bacteria are dormant wherefood sources are scarceand active where nutrientsare plentiful.

Bacteria canbe produced withfluids.

Oil-water contact

Bacteria canbe injected withwater.

■■Electron micrograph of a biofilm inside rock. The blocking of pores by bacteria canclearly be seen.

A better understanding of subsurfacemicrobiology from research programs—such as the British Geological Survey andUS Department of Energy Deep Microbiol-ogy Subsurface Program—now shows thatmicroorganisms can grow in temperaturesabove 125°C [257°F], at pH values of 1 to11, in the presence or absence of oxygen,and in up to 30% sodium chloride [NaCl]solutions. As this knowledge develops,microbiologists are turning their attention topetroleum reservoirs as a habitat formicroorganisms (previous page). This articleexamines how bacteria affect oil productionby dividing them into two groups—the “badguys” and the “good guys.”

The bad guys are those groups of bacteriathat use sulfur-based compounds present inseawater—and sometimes in formation oraquifer waters—as part of their energychain, and use the simple carbon com-pounds that are present in formations asfood. Waste from this growth includeshydrogen sulfide, which is poisonous tohumans, and corrosive to tubulars and top-side tanks. Other detrimental microorgan-ism grow profusely around the wellboreregion—in mud filter cake and the forma-tion—blocking rock pores and reducing per-meability; still others break down and ren-der ineffective chemicals that are added tofacilitate production operations andincrease wellbore life.

On the other hand, good microorganismsare those helpful bacteria which, duringtheir growth, produce useful compoundsthat improve oil recovery—for example, sol-vents, acids, gases, surfactants and biopoly-mers. It is these bacteria and their byprod-ucts that can be used constructively inreservoirs. Petroleum microbiologists try tofind ways of suppressing bad bacteria whileencouraging growth of good microbes.

The World of Petroleum MicrobiologyBacteria carry out all their life functionswithin a single cell. Nevertheless, their hugediversity ensures ubiquity. Bacteria arefound virtually everywhere: in the air webreathe, the food we consume and oneverything we touch; and they thrive in the

Spring 1997

■■Possible locations of microorganisms. Activethroughout the oil and gas production cycle, aoffshore, and in shallow zones as well as deep

most extreme habitats—hot springs, ariddeserts, subterranean vents, aquifers, saltylakes and salt deposits. Microorganismshave a remarkable ability to survive adverseconditions and remain in a dormant statewaiting for favorable growth conditions.Many microbiologists believe that this dor-mancy could last for thousands of years.

To date, most petroleum microbiologicalwork has centered on waterflooded reser-voirs that offer a cooled, oxygen-free, salineenvironment, which meets the environmen-tal requirements of many different groups ofbacteria. And bacteria certainly do grow inthese conditions, thriving when nutrients—including reservoir chemicals or seawater—are available, and entering dormant stateswhen food is scarce. Resuscitation from dor-mancy is rapid—perhaps taking two daysafter months of inactivity.

Most bacteria have a natural tendency togrow attached to rock surfaces rather thanfree-floating in the liquid phase. In apetroleum reservoir, bacteria may attach tothe rock, start to grow and then produce

and dormant bacteria are foundt and below the surface, on- and, hot, high-pressure reservoirs.

exopolymers—sugars—that help them attachto each other and rock surfaces. Such growthis termed a biofilm and offers the advantagesof protection from biocides while encourag-ing the bacteria to interact to best use nutri-ents and other resources (above).

Bacteria that are introduced to reservoirsthrough waterflooding will flow over preex-isting biofilms; some bacteria will attachthemselves to these biofilms and grow. Fromtime to time, some bacteria detach from thebiofilm and move with the liquid flow or bytheir own motility and colonize other areasdeeper in the reservoir.

Complete analysis of all potential foodsources supporting bacterial growth is still inprogress. However, analysis of formationwater from many reservoirs has demon-strated the presence of short-chain fattyacids—such as acetate, propionate andbutyrate—that may be utilized by some bac-teria to provide energy.

19

Biofilm-dwelling bacteria coating rock sur-faces and pore spaces not only use externalnutrients from formation fluids, but also uti-lize chemicals from other dead and dyingparts of the biofilm by breaking them downwith enzymes to release essential nutrients,which are then recycled. This process occursto a significant degree in many biofilms,ensuring that energy sources entering, oralready present in, a reservoir are used effi-ciently and economically several times overby many different opportunistic bacteria.

Environmental microbiologists havedemonstrated that bacteria exist in—andmay have originated from—the Earth’s sub-surface.2 In carrying out this work, manyproblems were encountered obtaining sub-surface fluid samples and rock cores thatcontain bacterial cells from the target envi-ronmental niche alone.

Drilling equipment carries chemical, min-eralogical and biological material through-out the borehole, contaminating otherwisepristine environments and making it difficultfor microbiologists to obtain genuine sam-ples at a specific depth in sediments or rockcores. However, technology is now avail-able to obtain cores in presterilized sleeves,sealed for examination in sterile conditionsat surface. Methods for checking theintegrity of these cores have also beendeveloped—for example, fluorescentmarker beads included in drilling muds canindicate contamination if they show up inthe core.

There are several ways of determining themicrobial content of cores. Direct observa-tion of core material using microscopictechniques can show cells associated withrock pore spaces. Cell cultures from subsur-face environments are possible by enrichingportions of rock that have been crushedaseptically and then mixing these with liq-uid nutrient to see what grows. However,this method does not necessarily result ingrowth of representative organisms—merelythose that could grow in the enriched media

20

provided. Today, molecular techniques are

2. Long PE, Onstott TC, Fredrickson JK, Stevens TO, GaoG, Bjorstad BN, Boone, DR, Griffiths R, Hallett RBand Lorenz JC: “Origin of Subsurface Microorganisms:Evidence from a Volcanic Thermal Aureole,” presentedat the International Symposium on Subsurface Micro-

employed to relate the genetic materialfound in recovered bacteria cells with well-documented organisms or with geneticmaterial that is associated with specific,well-characterized functions.

Obtaining uncontaminated fluid samplesat depth also poses problems, and muchresearch and development have gone intodevising equipment that will capture andhold separate sterile fluid samples from dif-ferent depths. Extremely deep samples mayrequire slow decompression to reduce therisk of physically damaging bacterial cells.There has been little evidence, however, inthe scientific literature supporting the exis-tence of genuinely barophilic organisms—those that actively require high pressure—socultures of cells obtained from great depthneed not be maintained at pressure.

One of the biggest issues facing subsurfacesampling is the expense of drilling.Inevitably, petroleum reservoir microbiolo-gists often have to be content with samplesderived from produced fluids. These containa mixture of oil, formation waters, andinjection fluids plus all the added reservoirchemical treatments—many of which areused to mitigate the effects of microbialactivity in the reservoir.

Introducing the Bad GuysThe most notorious villains on the reservoirscene are sulfate-reducing bacteria (SRB),which have relatively simple requirementsfor growth and energy generation—sulfateand carbon. Seawater contains about 2800ppm sulfate and formation waters may con-tain up to 2000 ppm short-chain fattyacids.3 Given suitable conditions, with nooxygen and favorable temperatures, thecocktail recipe is simple: inject seawater,and expect sulfide souring.

Although it has been known for manydecades that SRB are active in shallowwells, the existence of significant bacterialpopulations in deep, hot, high-pressure oilformations was not considered until souringbecame economically significant. Today, theassociation of the onset of reservoir souringwith commencement of seawater injectionin previously “sweet” fields is all too appar-ent and is largely due to SRB activity.

As a group, reservoir SRB have a widerange of temperature tolerances, and theirorigin has been the subject of much debate.It is likely that thermophilic bacteria(tSRB)—those most active between 55 and70°C [129 and 158°F] or higher—havealways been present in the hot, deep subsur-face environment, existing in porous rockmatrices and maintaining viable populationsusing nutrients from deep aquifers anddegraded hydrocarbons.

Work carried out by researchers at theHatherly Laboratories, University of Exeter,England, demonstrated that living tSRB havebeen recovered from open North Sea watersat 10 to 16°C [50 to 60°F]. ThermophilicSRB are most likely brought to surface dur-ing production operations and introducedinto the sea when separated fluids aredumped. These tSRB are extremely hardyand many of them are able to survive pro-longed periods of starvation in seawater atboth surface and reservoir temperatures.Given the resilience of these organisms, it ispossible that they may subsequently be rein-jected into another reservoir where theymay also grow if conditions are favorable.

Starved SRB can exist by just surviving at avery low metabolic level, waiting for theright conditions in order to revitalize. Dur-ing this dormancy, SRB tend to be smallerthan their growing counterparts and maytravel greater distances through the rockmatrix pores during waterfloods. StarvedSRB are less susceptible to standard reser-voir biocide treatments than actively grow-ing populations, making them particularlydifficult to treat effectively. The conse-quences of flushing these bacteria to thesurface during oil recovery and then laterreintroducing them during waterflooding arepotentially grave.

In cooler areas of the oil production pro-cess—such as surface equipment—themajority of SRB are mesophilic bacteria(mSRB) that prefer temperatures of 20 to40°C [68 to 104°F]. Since they are not well

Oilfield Review

biology, Davos, Switzerland, September 15-21, 1996.3. Biofilms in oil reservoirs frequently contain mixtures

of different groups of bacteria, including SRB. Some ofthe organisms are involved in processes that comple-ment the sulfate reduction activity of SRB by slowlyreoxidizing the reduced sulfur products, thus complet-ing the natural cycling of sulfur in the environment.

suited for long-term survival in deep forma-tions and are often associated with corro-sion processes occurring in topside facilities,the origin of this group of SRB is still beingdebated. The most likely source is seawater.With the advent of seawater injection to aidsecondary recovery, it is likely that mSRBenter with the waterflood and attach eitherat injection wellheads or, if formation resi-dence times are short, they are carriedthrough the formation to production well-heads, where they again attach to the metal,creating corrosion cells and copiousamounts of H2S. Once established, thesebacteria are difficult to eradicate since cellsdeep in the protective biofilm survive treat-ments with biocides that cannot penetratethe outer slime layers.

In addition to souring reservoirs, bacterialbiofilms may also plug formations. As thebiofilm increases in depth and maturity, itcovers a greater surface area, bridging porespaces and reducing fluid flow in and

Spring 1997

Project Mink Unit Pilot NIPER1

Year Initiated 1987

Oil Field Delaware-Childers

Formation Bartlesville sandstone

Permeability, md 90

Salinity, % <0.05%

Depth, m 200

Waterflood Yes

Injection Wells 4 out of 21 treated

Producing Wells 15

Oil Viscosity, cp 7

Microorganisms Used Clostridium, Bacillus

Microbial Products Produced Surfactants, acids and g

Nutrients Molasses (cane)

Test Length 2.5 years

Shut-in 12 days

Pre-MEOR2 Oil Production 0.4 B/D per well

Post-MEOR Oil Production 13% improvement in un

WOR3 WOR decreased 35% inwells near treated inject

Comments WOR in off-pattern wellincreased

1. NIPER—National Institute for Petroleum and Energy Research

Microbial Enhanced Oil Recov

around rock pores. Deep in a rock matrixthis can reduce waterflooding efficiency anddivert flow. If this happens near injectioninlets, it may have serious consequences,diverting the flood front away from the tar-get zone.

Incoming seawater carries the sulfate nec-essary for SRB energy generation and lowlevels of carbon sources. Because the injec-tion zone is the place where potential newnutrients enter the reservoir, it is also thearea where most microbial growth is likelyto be found. Injection profiles may quicklybe disrupted in this way.

Chemicals, such as biocides, are periodi-cally injected as slug doses in an intermit-tent treatment program. While these dosesmight inhibit bacterial growth at the injec-tion well, chemical concentration is notmaintained throughout the formation, andtherefore some reservoir zones receivereduced levels of biocide, to which bacteriaare not susceptible.

Phoenix Pilot NIPER SE Vasser Vertz University of Okl

1990 1991

Chelsea-Alluwe SE Vasser Vertz

Bartlesville sandstone Vertz sandstone

16 60 to 181

2.9 11 to 19

122 550

Yes Yes

19 5 treated

47 19

6 2.9

Clostridium, Bacillus Indigenous

ases Surfactants, acids and gases Gases and biom

Molasses (cane) Molasses and amnitrate

1.5 years 262 days

2 days 2 shut-ins (30 da

1 B/D per well No oil productionbefore nutrient in

it 1.2 B/D per well 83 barrels produJanuary 1992 to

ors

s Oil production improved by Permeability mo20% through May 1993 achieved; tertiary

from watered ou

2. MEOR—Microbial Enhanced Oil Recovery 3. WOR—Water/Oil Ra

ery Projects

Reservoir managers also need to be awareof the ability of indigenous bacteria to utilizenot only natural energy sources available ina formation, but also injected energysources. Year after year, many productionchemicals are pumped downhole on a con-tinual basis, giving bacteria ample opportu-nity to adapt and possibly use some part ofthe reservoir treatment for nutrition. Further-more, some chemicals—such as surfac-tants—act on rock surfaces, releasingattached organic material. If this material isoil, mobilization is good news. But if chemi-cals release part of the biofilm that may thentravel deeper into the formation, reattachand grow, this is bad news.

Bacteria can also produce enzymes thatare detrimental. These enzymes maydegrade treating chemicals such as poly-mers on the surface. In this case, if a bacteri-cide is used too late, it may kill only thebacteria, leaving the enzyme to wreakhavoc in oil and gas producing operations.

21

Sand Pilot Phillips Petroleum Co.ahoma

1991

Sand unit North Burbank unit

Burbank sandstone

50

10

884

Yes

1 treated

4 active

3

Indigenous

ass Biopolymer and biomass

monium N2, phosphorus, carbon sequential injection—Phillips patent

In progress

ys and 14 days)

in pilot wells jection

ced June 1992

dification was Permeability modification test oil produced has shown positive results in t well pressure fall-off tests

tio

22 Oilfield Review

Project Institute of Microbiology, Institute of Microbiology, Institute of Microbiology, USSRRussian Academy of Russian Academy of Russian Academy of Sciences and the Tatar Oil Sciences and the Tatar Oil Sciences and the Tatar Oil Research and Design Institute Research and Design Institute Research and Design Institute

Year Initiated 1987 1988 1988 1982

Oil Field Romashkino field, Romashkino field, Romashkino field, Bondyuzhskoe Sarmanovskaya area Zay-Karatayskaya area Aznakaevskaya area

Formation Upper Devonian

Permeability, md 500 500 500 500

Salinity, % 0.25 to 4.0 0.25 to 4.0 0.25 to 4.0 16

Depth, m 1500 to 1700 1500 to 1700 1500 to 1700 1500 to 1700

Waterflood Yes Yes Yes Yes

Injection Wells 3 1 2 1

Production Wells 6 2 5 6

Oil Viscosity, cp

Microorganisms Used Indigenous Indigenous Indigenous

Microbial Products Produced Surfactants, acids and gases Surfactants, acids and gases Surfactants, acids and gases Methane

Nutrients Nitrogen, oxygen and Nitrogen, oxygen and Nitrogen, oxygen and phosphorus phosphorus phosphorus

Test Length 36 months

Shut-in

Pre-MEOR2 Oil Production

Post-MEOR Oil Production 106 B/D increase 33 B/D increase 166 B/D increase

WOR3 85% 40% 95%

Comments 43% increase 10% increase 46% increase Oil >20 to 100% increase, water <20 to 30% increase

Project Institute of Biology of the Institute of Biology of the Institute of Biology of the Institute of Biology of the Romanian Academy Romanian Academy Romanian Academy Romanian Academy

Year Initiated 1987 1990 1989 1990

Oil Field Caldararu Caldararu Bragadiru Bragadiru

Formation

Permeability, md 245 to 248 245 to 248 150 to 300 150 to 300

Salinity, % 0.4 to 0.45 0.4 to 0.45 0.3 to 0.06 0.3 to 0.06

Depth, m 750 to 800 750 to 800 780 780

Waterflood Yes Yes Yes Yes

Injection Wells 2 1 1 treated 1 treated

Production. Wells

Oil Viscosity, cp 26 26 9 9

Microorganisms Used Clostridium, Bacillus, Clostridium, Bacillus, Clostridium, Bacillus, Clostridium, Bacillus, Pseudomonas, Arthrobacter, Pseudomonas, Arthrobacter, Pseudomonas, Arthrobacter, Pseudomonas, Arthrobacter, Mycobacterium, Micrococcus, Mycobacterium, Micrococcus, Mycobacterium, Micrococcus, Mycobacterium, Micrococcus, Enterobacteriaceae Enterobacteriaceae Enterobacteriaceae Enterobacteriaceae

Microbial Products Surfactants, gases, acids, Surfactants, gases, acids, Surfactants, gases, acids, Surfactants, gases, acids, Produced solvents and biopolymers solvents and biopolymers solvents and biopolymers solvents and biopolymers

Nutrients Molasses Molasses Molasses Molasses

Test Length 4 years 4 years 4 years 4 years

Shut-in 2 to 3 weeks 2 to 3 weeks

Pre-MEOR2 Oil Production 2.2 B/D per well 0.7 B/D per well 9.5 B/D per well 1.5 B/D per well

Post-MEOR Oil Production 9.6 B/D per well 2.2 B/D per well 21.5 B/D per well 7.4 B/D per well

WOR3

Comments Microbial flooding recovery Cyclic microbial recovery Microbial flooding recovery Cyclic microbial recovery(wellbore cleanup) (wellbore cleanup)

2. MEOR—Microbial Enhanced Oil Recovery 3. WOR—Water/Oil Ratio

2. MEOR—Microbial Enhanced Oil Recovery 3. WOR—Water/Oil Ratio

Microbial Enhanced Oil Recovery Projects (continued)

There Are Also Good Guys No one should doubt the usefulness of bac-teria. They yield many beneficial products,including pharmaceuticals (such as antibi-otics), food and drink (cheese and beer), andthey degrade human pollution. They arealso responsible for nutrient cycling and soilfertility. In reservoirs, some indigenousgroups of bacteria use nutrients that are pre-sent to produce byproducts that mayenhance oil recovery. During growth, bacte-ria generate acids, surfactants, solvents,biopolymers and gases, all of which arecommonly injected into reservoirs as chemi-cals to improve oil and gas recovery.

The types of bacteria used to improve oilrecovery are frequently isolated from reser-voirs in the first place. When reintroducedinto the reservoir, it is assumed that thesebacteria are best able to survive and out-compete other bacteria present. In somecases, specific nutrients may be pumped intothe reservoir to support the bacteria so thattheir growth and byproducts may improveoil mobility. The introduction of bacteriagenerally follows initial laboratory investiga-tion to optimize microbial growth and indi-cate the most suitable food nutrient package.The bacteria are then transported to the fieldsite and injected into the reservoir.

Spring 1997

Project Alpha Environme

Year Initiated 1987

Oil Field West Geff

Formation Aux Vases sands

Permeability, md 5 to 165

Salinity, %

Depth, m 799

Waterflood Yes

Injection Wells 2

Production Wells 7

Oil Viscosity, cp

Microorganisms Used Mixed Culture

Microbial Products Produced

Nutrients Catalyst, fertilize

Test Length 12 months

Shut-in

Pre-MEOR2 Oil Production 10 to 12 B/D

Post-MEOR Oil Production

WOR3

Comments 40% increase in

2. MEOR—Microbial Enhanced Oil Recovery 3. WOR—Water/Oil

With some applications, the nutrients areinjected together with the bacteria sogrowth can begin quickly. However, withother field applications, it is desirable todrive the bacteria deeper into the formationbefore growth commences so the nutrientsare introduced after injection of the bacte-ria, and microbial treatment may be fol-lowed by a shut-in period to allow the bac-teria to grow before the reservoir isreopened. There are believed to be dozensof different types of bacteria in most reser-voirs, and efforts are centered on identifyingthe desirable bacterium and determining anutrient for it. In some proposed applica-tions, only a nutrient targeted at desirablein-situ bacteria is injected.

Many projects that harness bacteria havebeen carried out around the world (see“Microbial Enhanced Oil Recovery Pro-jects,” page 21, previous page and below).For instance, surfactant-producing bacteriamay be used in a microbial enhanced oilrecovery (MEOR) project to reduce interfa-cial tensions between the oil and rock, andthe aqueous phase, releasing oil from capil-lary pressure. The introduction of surfactant-producing bacteria or stimulation of indige-

ntal Archaeus Technology Group LimitUnited Kingdom

1991

Lidsey field, southern England

tone

0.62 to 4.38

1173

No

1

1

Acetic acid producers

Acids

r Carbohydrates

30 days

7 days

40 B/D per well

50 to 60 B/D per well

45% water cut

oil production Carbonate reservoir, microbial hydraulic acid fracturing

Ratio

nous bacteria that may permeate and growthroughout the reservoir is an attractiveMEOR strategy, offering more extensive cov-erage and better cost-effectiveness than tra-ditional injection of chemical surfactants.The bacterial production of surfactants hasbeen successfully demonstrated in numer-ous field applications from Russia to the UKand the USA.

Gaseous waste from bacterial growthincludes carbon dioxide, methane andhydrogen. The encouragement of bacterialgrowth and subsequent gas productionincreases reservoir pressure and enlivensdead oil. Carbon dioxide also dissolves inoil, aiding its physical displacement. Thebacterial production of gases has been usedin field applications in Bulgaria, Poland andthe Czech Republic.

Acids produced by bacteria—such asacetic and lactic acid—assist oil recovery bydissolving carbonate rocks, increasingporosity and permeability. This has beendemonstrated in simulated reservoir condi-tions and within the reservoir, although, likemany of the microbial techniques, it is fre-quently a combination of methods, forexample both acid and gas production, thatassists recovery.4

23

ed, Research Institute of Exploration and Development, Daqing Petroleum Administration Bureau, Hailung Jang, China

1990

Daqing Oil field

405 to 932

0.7

1151 to 1171

2 treated

2 used for test

58.9

Bacillus licheniformis, Pseudomonasaeruginosa, Xanthomonas campestris, 5GA (similar to Bacteroides)

Acids, organic solvents, gases andalcohols

Molasses, sugar, crude oil and trace minerals

18 months

40 and 64 days

26 and 56 B/D per well

41 and 126 to 185 B/D per well

Single-well stimulation appears to be successful

■■Bacteria samples, before (left) and after (right) producing exopolymer.

4. Moses V, Brown BJ, Burton CC, Gralla DS and Cor-nelius C: “Microbial Hydraulic Acid Fracturing,” inProceedings of the 1992 US Department of EnergyInternational Conference on Microbial Enhancementof Oil Recovery, Upton, New York, USA, September7-11, 1992: 207-229.

5. Wagner M, Lungerhausen D, Murtada H and Rosen-thal G: “Development and Application of a NewBiotechnology of the Molasses In-Situ Method:

In the laboratory, the bacterium Clostrid-ium tyrobutyricum, isolated from a reservoirand growing anaerobically in a mediumcontaining 6% molasses, produces high lev-els of organic acids—11,400 mg/L butyricacid with 3200 mg/L ethanol and butanol—plus 350 mg/L of carbon dioxide (80%) andhydrogen (20%). These products can sup-press SRB growth and enhance methane-generating bacteria—methanogens.5

In a field study in the Russian Romashkinoreservoir of porous carbonate rock, 107cells per ml of Clostridium tyrobutyricumwere inoculated into the reservoir, whichhad been primed with 2% by weight ofmolasses fed through the injector wells,gradually increasing the molasses concen-

24

tration. Metabolic products suppressed thegrowth of SRB, and high levels of methanewere produced. The well was shut in forseven months and on recommencing injec-tion, the fluids were switched graduallyfrom waterflood to molasses solution. Gen-eration of methane, carbon dioxide andhydrogen increased the gas/oil ratio fromabout 4 m3 gas/m3 oil at the start to 8.5 m3

gas/m3 oil. Over two years, the rate of oilproduction was virtually doubled.

Additionally, some bacteria produce copi-ous quantities of exopolymer during growth(above). This combination of an increase inthe number of bacterial cells and produc-tion of these polymers can provide a densephysical blocking agent or plug that caneffectively impede flow within reservoirs.

This effect may be employed positively.When exopolymer-producing bacteria areintroduced into the reservoir during water-flooding, they follow the path of least resis-tance and are carried to the highest perme-ability zones where little additional oilremains. Some of the bacteria injected intothe reservoir will attach to the high-perme-ability rock. Then, when a suitable nutrientis injected, the bacteria produces exopoly-mers that block this zone. Thus, bacteriamay be used as a cost-effective, environ-mentally friendly, blocking agent, whichsubsequently diverts the waterflood into lesspermeable zones that still contain oil, butno blocking bacteria (next page). In a similarmanner, bacteria that produce, or behaveas, surfactants can cause emulsions inwatered-out zones, improving sweep andrecovery efficiency.

When microbial flow diversion is consid-ered for a reservoir, the candidate bacteriamay either be extracted from produced flu-ids or be specific exopolymer producers thatare known to survive at the particular reser-voir temperatures and conditions. Work atthe University of Exeter has resulted in sev-eral bacteria cultures that may be used todivert waterflooding. For this microbiologi-cal method—Selectively Induced Flow Tech-nique (SIFT)—to work optimally, however,reservoir data currently not readily availableare required.

Ideally, a sterile sample of fluid is neededfrom each reservoir zone to perform the fullrange of analyses for anions, cations andhydrocarbon derivatives to discover whatthe potential nutrients for the bacteria mightbe. The products of bacterial activity such aspolymers, acids and gases should also bedetectable. Full details of key parameters—for example pH, redox, dissolved gases,temperature, salinity, fluid flow rates, pres-sure, and geology—are needed for eachzone to help determine how bacteria willbehave downhole.

Oilfield Review

Detailed Evaluation for Selected Wells In theRomashkino Carbonate Reservoir,” in Proceedings ofthe 5th US Department of Energy International Con-ference on Microbial Enhanced Oil Recovery andRelated Technology for Solving Environmental Prob-lems, Dallas, Texas, USA, September 11-14, 1995:153-173.

Waterflood injectionflows preferentiallythrough high-permeability zonesor streaks in theformation.

Dormant bacteriaintroduced duringwaterflooding followthe path of leastresistance–morepermeable zonesfrom which oil hasbeen produced.

Nutrients are injectedto stimulate growthof the bacteria, whichbecome active andproduce exopolymers.These polymers andthe increased numberof bacterial cells form adense blocking agent.

As the more perm-eable zones areplugged, waterfloodinjection is divertedinto less permeableareas that stillcontain oil.

��������������������������������������������������������������������������������������������������������������������������������

Flow

����������������������������������������������������������������������������������

Flow

����������������������������������������������������������������

����������������Flow

������������������

Flow

Flow

■■Microbial flow diversion scheme.

Future Reservoir Bacteria “Farming”There are other potential good guys—bacte-ria that help control SRB. The SRB in reser-voirs do not live alone, but exist in commu-nities with other groups of bacteria. Many ofthese groups interact by providing growthrequirements for their neighbors, with somegroups even living off the waste of others.Looking to the future, it is possible to envis-age a situation whereby beneficial bacteriaare farmed to control detrimental bacteriathrough their growth and activities.

�������

Spring 1997

6. For additional information on these topics: Parkes RJ, Cragg BA, Bale SJ, Getliff JM, Goodman K,Rochele PA, Fry JC, Weightman AJ and Harvey SM:“Deep Bacterial Biosphere in Pacific Ocean Sedi-ments,” Nature 371 (September 29, 1994): 410-413.Rueter P, Rabus R, Wilkes H, Aeckersberg F, RaineyFA, Jannasch HW and Widdel F: “Anaerobic Oxida-tion of Hydrocarbons in Crude-Oil by New Types ofSulfate-Reducing Bacteria,” Nature 372 (December1, 1994): 455-458.L’ Haridon S, Reysenbach AL, Glénat P, Prieur Dand Jeanthon C: “Hot Subterranean Biosphere in aContinental Oil Reservoir,” Nature 377 (September21, 1995): 223-224.

Research by Mike McInerney and col-leagues at the University of Oklahoma, Nor-man, USA, has demonstrated that a bac-terium called Thiobacillus can mitigate thegrowth effects of some SRB. Their studiesshowed that as the SRB grew, they producedhydrogen sulfide, but the Thiobacillus keptsulfide levels low by converting it to sulfate.For this scheme to work, nitrate must beadded to the reservoir. In this manner it ispossible to control reservoir souring bymanipulating the ecology of the bacteriaresiding there. This promising technologyhas been studied only in the laboratory attemperatures no greater than 30°C [86°F]. Itwould be necessary to ascertain whetherthis technique would be successful at thehigher temperatures and more extreme con-ditions found in hydrocarbon reservoirs.

�������

Given time, all microbial communitieswill adapt to the prevailing conditions, mak-ing efficient use of all available energysources.6 An open system, such as an oilreservoir, is in a state of constant flux withdifferent populations of bacteria dominatingthe zones that most suit their energyrequirements, and temperature and chemi-cal preferences. It is impossible to achieve astate of sterility within a reservoir. However,given good data collection and manage-ment practices, it will be possible to create abalance that enhances production and eco-nomical reservoir management. It is thispromise that is expanding the role of micro-biologists in reservoir management.

—CF

25

Specialty Chemicals in the Oil Field

Achieving greater oilfield efficiency

and productivity depends on wellsite

operations that cost-effectively maximize

recovery of oil and gas reserves, while

minimizing the impact on the

environment. Pivotal to these operations

are specialty chemicals

that impart unique capabilities and

functionality for well drilling, completion

and intervention services. The

last decade’s progress in upgrading

chemical quality, deliverability and

environmental compliance is paying

off for operators in terms of field

performance and longevity.

26

Brian AinleyDavid ClouseDon HillTulsa, Oklahoma, USA

Mike CatrettNalco/Exxon Energy Chemicals, L.P.Sugar Land, Texas, USA

Greg KubalaSugar Land, Texas, USA

For help in preparation of this article, thanks to BrianDarling, Jon Elphick, Terry Greene, Tom Griffin and JoeMiller, Dowell, Sugar Land, Texas, USA; Jim Thompsonand Terry Whittle, Dowell, Tulsa, Oklahoma, USA; andBill Bailey, Brent Diez, Melody Lindley, Jeff Schiller andSteve Sears, Nalco/Exxon Energy Chemicals, L.P., SugarLand, Texas, USA.

The old expression ”garbage in, garbageout” has particular meaning when appliedto oil and gas reservoirs. Hydrocarbon-bearing formations are highly susceptibleto damage and plugging from a variety ofsources—both natural and induced. Thepermeability and porosity of virgin reser-voirs may be altered dramatically unlessdrilling, completion and interventionpractices are conducted with the utmostdiligence and attention to detail. If not,well productivity and ultimate reserverecovery suffer, while field maintenance,workover and environmental protectioncosts skyrocket.1

Many services performed in the oil fieldrely on specialty fluids and additives thatfulfill specific functions within the wellboreor formation. This article describes the typesof specialty chemicals that are employeddaily to drill and treat oil and gas wells andhow a pervasive focus on their quality, relia-bility and deliverability is helping operatorsget the most from field developments.

Oilfield Review

1. Kruger RF: “An Overview of Formation Damage and

The Role of Specialty Oilfield Chemicals Well drilling, completion, treating andworkover fluids perform to their utmostbecause of the specialty chemicals that areadded to impart unique properties and func-tionality. These chemicals fall into a broadvariety of categories, with a staggering overallnumber of different compounds and blends(below). If designed and manufactured toproper physical and performance standards—defined and confirmed through extensivefield application—specialty chemicalsbecome invaluable solutions to overcomeproblems that plague oil and gas wellsthroughout their lifetimes.2

If these chemicals are prepared, stored,mixed or pumped incorrectly, however, theycan become a well’s worst nightmare—leading to significant problems, such asplugging or precipitation, because of thepresence of, and interactions caused by,inferior materials. Whether it’s a drillingfluid that causes excessive formation dam-age (see “A New Slogan for Drilling FluidsEngineers,” page 2) or a fracturing fluid thatleaves flow-restricting polymer residue inthe proppant pack, increased costs, reducedefficiency and lower profits can be the endresult of faulty selection or application.3

Spring 1997

����

���

���

�������Drilling and

completion fluidadditives

DeflocculantsDefoamers/foamersFluid-loss reducersLubricantspH controllersPipe-freeing agentsPolymeric viscosifiersShale inhibitorsSpecialty saltsSurfactants

��■■Categories of specialty oilfield chemicals using the lifetime of a typical oil or gas well. Sping, completion, stimulation and production otives vary from well to well, those listed here

During the boom years of the late 1970sand early 1980s, attention to chemical qual-ity control and performance consistencywas often lax. Operator and service com-pany personnel and facilities were stretchedto the limit just getting wells drilled andtreated on schedule without costly mistakes.Often, there was simply insufficient time tofine-tune field formulations to achieve opti-mal results. The same was true for chemicalsuppliers, working all-out to satisfy demandfor their products during a period of peakactivity. There was little chance to concen-trate on improving in-plant production anddistribution procedures.

The situation was complicated further bygrowing demand for more sophisticated flu-ids. Over the years, simple fluids had givenway to more complex ones. By the time thetotal depth of a well was reached, for exam-ple, a drilling mud might contain 20 ormore distinct chemical types, many ofwhich had been added to offset the effectsof other components present during earlierphases of drilling. A large number of addi-tives means that a complicated set of chemi-cal and physical interactions have to bethoroughly analyzed before the impact ofthe total fluid system on the formation canbe understood.

Stimulationadditives

BactericideBreakersClay stabilizCorrosion inCrosslinkerDiverting agFluid-loss reMutual solvpH controllePolymeric v

gellingSurfa

Cementingadditives

AcceleratorsDispersantsExtendersFluid-loss reducersGas-migration

controllers/latexesRetardersSpacers and

chemical washes

����

���

ed at the wellsite. Hundreds of different chemicecialty additives provide the necessary fluid prperations at bottomhole temperatures and pre

represent a typical set.

Rising to the ChallengeFollowing the mid-1980s oil crisis, a newquality drive emerged throughout the oilfield—reinforcing industry efficiency andproductivity initiatives already in place.These initiatives first led operators and ser-vice companies to restructure and stream-line their operations in an effort to improveprofitability. Industry-wide consolidationand a refocusing on core competenciesaccompanied a host of cost-reduction steps.

When attention then turned to providinggreater quality and value in each phase ofthe business, operators—concerned aboutthe need to concurrently lower costs andimprove well performance—began request-ing more detailed information about thechemical additives present in fluids beingpumped by service companies. Servicecompanies, in turn, demanded more infor-mation from their chemical suppliers.

At the same time, a rising tide of publicand governmental concern about health,safety and environmental (HSE) issues—from personnel exposure to potentiallyharmful materials in chemical plants and atthe rigsite, to protection of marine life andaquifer quality—prompted a concertedreevaluation of oilfield chemicals and their

27

fluid

s

ershibitors

sentsducers

entsrsiscosifiers/ agentsctants

Productionchemicals

BactericidesDemulsifiersGelsInhibitors:

corrosion/paraffin/scale/asphaltene

Oxygen scavengersSurfactants

����

���

���

���

al compounds are pumped downhole dur-operties required for basic drilling, cement-ssures. Although the most important addi-

Well Productivity in Oilfield Operations,” Journal ofPetroleum Technology, (February 1986): 131-152.

2. Drilling, Completion and Workover Fluids; Cement-ing; Fracturing; and Acidizing supplements to WorldOil (1996).

3. Hawkins GW: “Laboratory Study of Proppant-PackPermeability Reduction Caused by Fracturing FluidsConcentrated During Closure,” paper SPE 18261, pre-sented at the 63rd SPE Annual Technical Conference& Exhibition, Houston, Texas, USA, October 2-5,1988.

Initial Driver

Mid-80s oil crisis Industry-wide focus on efficiency/productivity• Restructuring• Consolidation• Core

competancies• Streamlining• Cost reduction

Pressure to improve product• Quality/value• Performance

First Response Subsequent Drivers Second Response

Expansion of HSE regulations

Initiatives targeted at product• Quality• Reliability• Deliverability• HSE compliance

Market drivers and industry response. The oil crisis of the mid-1980s sparked major effi-iency and productivity actions within the industry—highlighted by restructuring,treamlining and cost-reduction steps. Next came a concerted push for improved quality,alue, performance and HSE compliance in products and services that has been realizedhrough broad initiatives by specialty chemical manufacturers and service companies.

4. Over 90% of the papers cited were presented at the:SPE/UKOOA European Environmental Conference,Aberdeen, Scotland, April 15-16, 1997.The Third International Conference on Health, Safetyand Environment, New Orleans, Louisiana, USA, June9-12, 1996.The Second International Conference on Health,Safety and Environment, Jakarta, Indonesia, January25-27, 1994 and published in: “Environmental Considerations,” SPEreprint series, no. 37, 1992.

effects on both the surface and subsurface.A host of regulations that had impacted oil-field operations since the 1970s and newlegislation, enacted principally within theUSA and the North Sea, combined to dra-matically affect chemical approval, usage,handling and disposal (below).

Governmental decrees, coupled with theindustry’s commitment to doing business ina more open manner, focused increasedattention on fluids pumped into a well ordischarged in the vicinity of the wellsite.Operators wanted details of any practiceswith potential negative impact so that theycould fulfill obligations to regulatory agen-cies and answer questions from environ-mental groups.

Specialty chemical suppliers were facedwith a wide range of challenges andqueries. To their credit, they reacted swiftlywith a well-directed, comprehensiveapproach. As a result, there have beentremendous strides over the past decade inproduct quality control, reliability, deliver-ability and HSE compliance (above)

During the 1990s, the drive for continuousimprovement and higher standards has ledthe oil field beyond regulatory compliance.The industry now expects more from itselfand has begun to evaluate resource con-sumption and environmental burdens asso-ciated with oilfield activities. The concept ofsustainable development—a belief thatoperators and service companies can meet

■■csvt

1963

Clean Air Act

1964

Wilderness Act

196

WatQua

SolidDispAct

■■Environmental actions impacting theapplication of oilfield chemicals. The USAhas been the leader in enacting legislationprotecting the environment, with many ofthe major milestones shown on the timeline. Several regulations have directly orindirectly influenced the use of specialty oil-field chemicals. As the United Nations (UN)and other organizations representing coun-tries around the world became involved,the impact broadened, as shown by mile-stones designated with an asterisk(*).

the world’s energy needs without compro-mising the environment for the future—isbeing employed at all levels to integratequality and HSE goals into everyday busi-ness strategies and action plans. This evolu-tion has been documented in over 350papers published since 1992.4

The Modern Specialty Chemical PlantChemical manufacturing and blendingplants are now operated to much stricterstandards, with broader checks and balanceson product quality. Advanced process con-trol and optimization of reaction conditionshave improved product reproducibility andincreased product cost-effectiveness. In-plant safety and environmental aware-ness, packaging and inventorying, and distri-bution practices have been scrutinized andupgraded. At the same time, research con-ducted by service companies and specialty

1968

Wild andScenicRivers Act

National TrailSystem Act

1970

National EnvironmentaPolicy Act

OSHA and Ecreated

5

erlity Act

Waste osal

From 1880 to 1960, the US Congresspassed a total of eight acts related to theenvironment. Six more followed from 1960to 1969. From 1970 to 1990, however, thetrend accelerated dramatically, and 50acts were passed—ranging from the cre-ation of the Environmental ProtectionAgency (EPA) to establishment andamendment of the Superfund Act. In addi-tion, international actions by OSPARCOM—the Oslo and Paris Commissions for the Pro-tection of the Marine Environment of the

chemical manufacturers has led to a newcrop of innovative, value-added materialsand application methods that have extendedthe capabilities of well operations to deeper,higher temperature and higher pressureenvironments (next page).

l

PA

1972

Oslo Commission*

UN Plan for Protection of the World Environment*

Clean Water Act

Coastal Zone Management Act

Ocean Dumping Act

1972

Paris Commission*

Safe DrinkingWater Act

Northeast Atlantic—set the foundation forlaws governing protection of oceans andcoastlines from hydrocarbons, and regula-tions for disposal of offshore platforms.

US laws intended to preserve unex-ploited natural wilderness and wetlandsareas have reduced or prevented seismicactivity in certain areas and prompted thedevelopment of sophisticated seismic andwireline tools and software to limit environ-mental impact. Air and water emissionand waste restrictions have led to

1976

Resource Conservation and Recovery Act—RCRA

1977

Toxic Substance Control Act—TSCA

1986

Superfund Amendment Reauthorization Act—SARA

1991

Wetlands Executive Order

1996

ISO 14001/14004Standards for EnvironmentalManagement*

1980

Comprehensive EnvironmentalResponse, Compensation and Liability Act—CERCLA/Superfund

and 1980, respectively, it took severalyears for the agencies involved to promul-gate enforceable regulations. Oil and gasE&P activities were exempt from these reg-ulations during the industry’s restructuringperiod in the mid- to late-1980s. In return,9.7 cents of each barrel of produced orimported oil went to Superfund.

In the North Sea, actions by OSPAR-COM have impacted exploration, drilling,cementing and stimulation practices.Drilling fluids have moved from oil-base

HSE awareness

Product reformulation/New product R&D

Modern specialty chemical plant

Distribution efficiency

Inventory management

Reaction condition optimization

Process control

Packaging and labeling flexibility

Quality control/ISO performance

■■The modern oilfield chemical plant. Today’s facilities bear only passing resemblance to those of 15 years ago. In-plant logistics have been improved. Adoption of ISO quality standards, computer control of reaction and blendingprocesses, and advances in packaging, warehousing and tracking have combined with heightened HSE awarenessand product optimization studies to increase plant throughput and product quality.

improved techniques to decrease or elimi-nate unwanted off-gas and water produc-tion. Disposal considerations havechanged the nature of oilfield chemicalpackaging from small, disposable contain-ers to large, reusable containers. Radiationlaws have spurred development of surfaceand downhole tools that rely on nonra-dioactive instead of radioactive sources.

In the US, the dominant laws affectingthe oil field have been RCRA and CER-CLA/Superfund. Although passed in 1976

to synthetic or water-base systems. Stimu-lation treatments now include corrosioninhibitors, crosslinkers and other additiveswith lower toxicity.

On a global basis, developing nationsare facing similar environmental issues.Chile, for example, is expected to issueenforceable environmental regulationslater this year. Many countries are consid-ering requiring ISO 14000 or a similar stan-dard as a management system to fosterproactive, beyond-compliance environ-mental activities.

30 Oilfield Review

Warehouse

Administrationbuilding

Laboratory

Warehouse

Shippingdock

Tote-fillingbuilding

Tanktruckunloadingstation

Tankfarm

Waste containmentbasin

Reactors

1. Receive raw materials2. Store raw materials3. Blend/react components4. Package finished products5. Store finished products6. Ship finished products

Drummingbuilding

Boiler house

Pilot plant

Control building

14

5

6

3

2

Fire water reservoir tank

Pumpbuilding

Run-off retention area

Dry blenders

Warehouse

Administrationbuilding

Laboratory

Warehouse

Shippingdock

Tote-fillingbuilding

Tanktruckunloadingstation Boiler house

Reactors

Run-off retention area

Pilot plant

Control building

14

5

3

2

Fire water reservoir tank

Pumpbuilding

Tankfarm

Drummingbuilding

Dry blenders

1. Receive raw materials2. Store raw materials3. Blend components and package products4. Store finished products5. Ship finished products

■■Material flow for liquid (top) and dry (bottom) products. The logistical complexity of specialty chemical manufac-turing facilities has prompted studies targeted at optimizing material flow and sequencing within each sector of theplant to improve product scheduling and deliverability.

■■The Dowell Chemical Manufacturing Plant in Tulsa, Oklahoma, USA. This facility, which achieved ISO 9002 certification in 1992and ISO 9001 certification in 1996, has one of the most consistent product on-time delivery records in the specialty chemical manu-facturing business.

5. The International Standards Organization, based inGeneva, Switzerland, is the main body that has estab-lished quality procedures and controls adopted by theoil and gas industry. Its ISO family of programs havebecome the recognized standard for a quality system.

Given the number of raw materials, reac-tion intermediates and finished products—along with packaging, labeling and storageoptions—specialty chemical manufacturingplants are among the most logistically com-plex facilities to be found anywhere in theworld. Today’s plants have adopted qualityand productivity programs that have beenproven to be effective in other industries.Some have been introduced out of neces-sity, due to the complex nature of the opera-tion. Others are a direct result of applicationof general quality standards, while still oth-ers reflect guidelines established throughInternational Standard Organization (ISO)certification or mandated by environmentalregulations.5

Spring 1997

Compared to a decade ago, materials flowhas been streamlined to simplify in-plantlogistics and support new product deliveryconcepts, such as just-in-time manufactur-ing. For both liquid and dry products, theoptimization process has affected theamount of space allocated to various func-tions—such as raw material receipt and stor-age, reaction and blending, packaging, fin-ished product storage and shipment—aswell as their proximity and interactions (pre-vious page). The result: increased plantthroughput, greater productivity of plantpersonnel, improved product delivery andshortened order lead times.

For example, 89% of all North Americanfield orders are now shipped within twodays from the Dowell specialty chemicalplant in Tulsa, Oklahoma, USA, a 33%improvement from the three-day average afew years ago. Shipments to overseas loca-

tions typically take two weeks today, insteadof the previous three (above).

Better packaging techniques for liquidproducts have greatly improved accuracy.Instead of filling containers according tovolume, which is subject to variations of +/-1% based on the temperature of thematerial at the time of loading and otherfactors, weight has become the standard.State-of-the-art mass flowmeters provide anaccuracy within 0.15%.

31

32

Distributed Organization

Operations & Control

Operations & Control

Planning Coordinator

Liquid Production Material Control Dry ProductionShipping & Receiving

Blending & Reacting PlanningPackaging SamplesCustomer Service

Production Self-Directed Work Teams

Customer Service

Blending & Reacting Packaging Shipping & Receiving

Integrated Organization

■■Change in plant organizational structure. The traditional distributed organization (top) is giving way to an integrated, concurrentstructure (bottom) that will be composed of self-directed work teams dedicated to particular process streams within the plant. This struc-ture will empower employees to become intimately involved in all aspects of quality control and quality assurance programs.

■■Just-in-time manufacturing flow. Just-in-time production is generally defined as a sys-tem of managing operations with little or no delay time or idle inventories between oneprocess and the next. Modern specialty chemical plants have improved raw materialflow and chemical production by adopting just-in-time processes similar to those used insmall-parts manufacturing plants. Just-in-time production is most evident in a continuousprocess in which material arrivals are timed to coincide with the production run. Batchprocesses vary from this methodology along the following lines:

• For high-volume products purchased by many customers, some finished productsare typically held in inventory, but with a minimum trigger level that causes addi-tional batches to be scheduled to maintain that level.

• For medium-volume products, a plant typically carries inventories of the raw materi-als. Products are not made until an order is received.

• For low-volume products, particularly those sold to a single customer, the only rawmaterial inventories are those used in other, high-volume products. When a cus-tomer orders the product, the plant, in turn, orders the raw materials specific to thatproduct.

• High-volume raw materials have a minimum trigger level for reordering, but low-volume raw materials are ordered only as needed.

Batch is placed on dock and shippedto customer

Customer places order on supplier

Order triggersbatch ticket to startmanufacture

Inventory of stockraw materialselectronically pulled

Orders placed on other raw materialsuppliers

Batch is manufactured and packaged

The combination of packaging improve-ments, along with less off-specificationmaterial produced and tighter quality con-trol, has meant fewer product returns fromthe field and more satisfied customers. Anadded benefit is minimization of waste atthe wellsite and at the manufacturing plant.

Equipment and logistics improvements areonly part of the story, however. There hasalso been a revolution in information sys-tems and organizational work practices.Technological advances in linked computersystems and software, increased interventionby regulatory agencies and an emphasis onreduced inventory levels have been instru-

Oilfield Review

■■The Nalco/Exxon Energy Chemicals Plant in Sugar Land, Texas, USA. Nalco/Exxon wasone of the pioneers of just-in-time specialty chemical manufacturing for the oilfield mar-ket and an innovator in applying new informational systems. The Sugar Land plantreceived ISO 9002 certification in 1992.

6. Duncan E, Gervais I, Le Moign Y, Pangarkar S, StibbsB, McMorran P, Nordquist E, Pittman T, Schindler Hand Scott P: “Quality in Drilling Operations,” OilfieldReview 8, no. 1 (Spring 1996): 20-35.

7. Schonberger RJ: Building a Chain of Customers: Link-ing Business Functions to Create a World-Class Com-pany. New York, New York, USA: The Free Press,1990.

mental in driving improvements in plantefficiency. These changes have occurred inboth supplier plants and customer facilities.

Organizationally, major gains are beingachieved by encouraging people on theplant floor to directly influence productquality and deliverability. Rather than sepa-rating functions as in the past, there is amove toward self-directed work teams thatoversee all aspects of the planning, prepara-tion, packaging and shipping of particularproduct streams. This concurrent organiza-tion instills a sense of pride and ownership,not unlike the strides that have beenachieved in automotive assembly. Potentialproblems are caught sooner. Employees areencouraged to submit suggestions for furtherimprovements, with a promise of rapidmanagement review and response (previouspage, top).6

In total, there have been a multitude ofchanges that are having pronounced bene-fits both for plant and field operations. Theremainder of this article focuses in greaterdepth on four:Within the plant—

• Improved deliverability using just-in-time principles

• Quality control through organizationaland informational changes

• Chemical product reformulation At the wellsite—

• Minimizing waste discharge

Improved Deliverability Using Just-in-Time PrinciplesMany service companies and other specialtychemical customers now place smaller, morefrequent orders with shorter lead times,thereby reducing their inventories and carry-ing costs. For many plants, the volume ofchemicals shipped has not changed appre-ciably, but the number of orders hasincreased significantly. Manpower and costsassociated with order processing are linkedmore closely to the number of orders, ratherthan order size. Thus, chemical suppliershave adopted more sophisticated means ofprocessing orders to keep from increasingstaffing levels. This has led to adoption ofjust-in-time manufacturing principles at facili-ties like the Nalco/Exxon Energy ChemicalsPlant in Sugar Land, Texas, USA (above).7

In the current marketplace, service compa-nies strive to minimize inventories and applymore sophisticated scheduling and inventorymanagement methods. This is contrary to tra-ditional practices in which large inventorieswere maintained to avoid running out ofmaterials. Today, without an inventory cush-ion, on-time shipments become critical, andthe communication link between supplierand customer must be flawless.

Spring 1997

On the manufacturing side—with a singleplant producing as many as 500 productsstarting from as many raw materials—inven-tory costs are significant. Methods to mini-mize raw material inventory can be key tokeeping production costs low. The 20/80rule-of-thumb applies—about 20% of theraw materials are used in about 80% of theproducts. The balance may be used onlyoccasionally—with most of the remainderappearing in only one to three products.Large inventories increase the probability ofoverstocking, with a corresponding negativeimpact on overall costs.

Production scheduling formerly was “eye-balled” by an experienced individual basedon historical norms. For just-in-time produc-tion, scheduling requires integrateddatabases that track customer orders, pro-duction status, raw materials in plant inven-tory and in transit, equipment constraints,and environmental regulations that limitwhich products may be made in whichpieces of processing equipment (previouspage, bottom).

33

Traditional

Ideal

Vesseltime

Vesseltime

Batch hold time

Packagingtime

Packagingtime

Warehouse time

Docktime

Invoicetime

Customertime

Customertime

Paymentreceived

Paymentreceived

■■Cycle-time history. By using just-in-time principles and changingfrom traditional business practicesbased on large-batch production,specialty chemical plants have cutthe total cycle time to make anddeliver a product. These methodssignificantly reduce the timebetween receipt of an order andreceipt of payment after productdelivery.

Traditional product costing tends to bebatch-based, driving production to largerbatches and increased inventories. Newmethods had to be developed to reflect thetrue cost structure more accurately in achanging market. With the trend to smallerbatch sizes and quicker equipmentchangeover to meet order demand, the tra-ditional chemical production line hasevolved into something more akin to asmall-parts assembly line.

Large general-purpose reaction and blend-ing vessels with long batch times are usedless frequently. Instead, smaller vessels withrapid turnover—segregated to similar pro-cess families to reduce waste and wash-ing—are now the mainstay of plants, alongwith automated in-line blending equipmentfor selected product/chemistry lines. Com-puter systems throughout the productionfacility now bring up-to-the-minute informa-tion, such as batch status and inventory con-sumption, directly into the schedulingoffice, moving plants like Nalco/Exxontoward a make-to-order facility.

Packaging and labeling are additionalareas that have seen radical change. Prod-ucts are shipped in a variety of packagetypes and sizes, including traditional 55-gal[208-L] drums, returnable tote tanks anddisposable containers, such as small pails.Returnable tote tanks represent an addi-tional capital resource and require somealterations to the normal manufacturing andpackaging cycle. As we will see later, how-ever, substantial overall cost savings andenvironmental benefits result from their use.

Many customers demand customized con-tainer labeling to fit their facility and inven-tory management needs. Bar coding oncontainers and portable radio-frequencyreaders make storage and access easierwhen dealing with numerous, random loca-tions. Tracking becomes critical in optimiz-

34

ing warehouse space and managing mini-mum inventories, which in many cases maybe only a few containers. Bar coding alsoreduces random errors in shipping, some-thing to be avoided at all cost in a reduced-inventory, just-in-time delivery market.

Quality Control Through Organizationaland Informational ChangesAs noted earlier, fundamental changes inorganizational work practices haveimproved production efficiency for specialtychemicals. The quality and ISO processesprovide structure in what used to be a rela-tively unstructured business and form thebasis for a set of recognized guidelines anda common operating language for rawmaterial vendors, chemical manufacturersand clients.

This common approach has led to a majoradvance in product quality assurancethrough application of standardized testmethods, as well as procedures for equip-ment calibration and maintenance. Byreducing the testing required to statisticallyvalidate processes, cycle times have beendecreased (above). The ISO structure, by itsvery nature, provides a means for betterassimilating the increasingly complex infor-mation that is being generated, defining dis-ciplined standards and tools for data organi-zation that allow productivity improvementsin a diverse, dynamic business environment.

In the past, it was difficult or impossible toset product specifications and define rawmaterial evaluation methods that were mutu-ally agreeable to raw material suppliers, spe-cialty manufacturers and clients alike. Today,as a result of adoption of quality processes,specifications are routinely established anddiligently adhered to. Raw material vendorsunderstand that their performance will becontinually monitored using a comprehen-sive set of criteria—where price is only oneparameter. Other factors include on-timedelivery, correct labeling, correct loading inthe delivery truck, completeness of paper-work, and container condition. In combina-tion, these data provide input to a rating sys-tem in which each supplier’s performance isdetermined, compared to a site standard andthen ranked against the performance of othervendors. A critical part of the process is pro-viding comprehensive feedback to vendorson areas for improvement. If noncompliancewith site standards occurs, formal writtendocumentation outlines the deficiencies andrequests explanation of causes and provisionsfor short-term fixes and longer-term solutions.The rating system is useful for other purposesalso, such as identifying suppliers who con-sistently exhibit superior performance andmight be considered as candidates to partici-pate in alliances (next page, top).

Multifunctional teams—including repre-sentatives from research, marketing, engi-neering, purchasing, quality assurance andHSE—evaluate and select vendors. For criti-cal high-volume products, clients may alsobe involved in the vendor selection process.The multifunctional team, in conjunctionwith a similar team from the vendor, deter-mines the product specifications, qualityassurance testing procedures and communi-cations protocol.

These specifications and quality assurancedata are compiled into a common informa-tional database that also includes operatingprocedures, health and safety information,maintenance records and the most recentenvironmental regulations. With networkedcomputers, employees throughout the planthave access to the same, up-to-date infor-mation on raw materials and finished prod-ucts. Common databases allow plant

Oilfield Review

Vendor Certification Process

Initial analysis

Vendor’s productsFinancial healthLogistics capabilitiesReputationWorldwide supply position

QA exercise

Chemical plant team membersPurchasingQuality assuranceResearchEngineering

Vendor team membersPurchasingQuality assuranceResearchEngineering

Teams meet to determineProduct specificationsCommunications protocol

Pricing

Contract development

■■Vendor certification process. The vendorselection process includes an analysis ofthe raw materials, the supplier’s stability,the supplier’s shipping and timing capa-bilities, and finally price. Once a multi-functional team selects a raw materialvendor to supply a product, groups fromthe vendor, chemical plant and often theend-user meet to determine raw materialand product specifications, quality assur-ance testing and logistics for shippingmaterials worldwide.

Productivity improvement using computers and databases

Purchasing Environmental regulations

Productspecifications

QA and testingNetwork servers

Marketing Database

employees to respond accurately andquickly to client inquiries related to productmanufacture, testing, and environmentalcompliance.

Changes in environmental regulations sig-nificantly affect specialty chemical plantsbecause of the wide variety of raw materialsemployed and the types of products that aremanufactured. Just a decade ago, a plantwould have only a few environmental engi-neers, but today nearly half of the engineerswork on environmental issues and regulatorycompliance. Information on production andplant practices must be retained longer andin more detailed formats to comply with reg-ulatory guidelines. Networked computer sys-tems allow instant access to particular prod-uct information in the event of unannouncedenvironmental audits. Without linked infor-mation systems, data retrieval would be diffi-cult, slow and costly (below). The qualityprocess has brought disciplined measuresand tools to help manage the collection anduse of massive amounts of data in this highlycomplex business.

Chemical Product ReformulationSeveral forces drive the reformulation—orreengineering—of specialty chemicals. Themost notable are the need to:

• improve performance• reduce cost• minimize safety or environmental

hazards.Chemical manufacturers and service

companies have addressed these reformu-lation problems singly and in combination,with a beneficial impact on wellsite execu-tion efficiency, well performance, environ-mental protection and overall chemicalusage and cost.

Research studies targeted at improving theunderstanding of mechanisms controllingchemical interactions have led to revisedmaterial design specifications and exten-sions in functionality that allow field tem-perature and concentration ranges to bebroadened. Process optimization within

Spring 1997 35

EngineeringManufacturing

■■Productivity improvement using computers and databases. Part of the revolution in thespecialty chemical manufacturing industry has centered on information systems andorganizational work practices, not just new equipment and chemical processes. Net-worked computers allow information about products to be retained with greater detailand retrieved more easily for compliance with environmental regulations. Through com-mon databases, information on production runs and product testing is immediatelyavailable to personnel in all areas of the plant.

Improved toxicityrequired

Review product toxicity and regulations

Use/recommendproduct

Reformulationfeasible

New chemistryrequired

Best available technology

Resourcesavailable

Develop newchemistry

Final toxicityevaluation

Find alternate components;reformulate

Evaluate performance

Acceptabletoxicity and

biodegradation?Commercialize

Productperformanceacceptable

Toxicity screenevaluation

Engineering controls

No

No

No

No

No

No

Yes

Yes

Yes

Yes

Yes

Yes

■■Decision tree used in the product refor-mulation process. Reformulating a productto reduce cost or improve performance is atime-consuming task. When environmen-tal concerns become part of the equation,the process takes on additional complexityand requires a comprehensive, stagedapproach that focuses on product perfor-mance, toxicity and biodegradability.

specialty chemical plants—centered onparametric studies of reaction and blendingconditions such as temperature, pressure,exposure time and raw material additionsequencing—has increased product yieldand reduced byproducts. Product consis-tency, reproducibility and reliability haveimproved, dovetailing directly with comple-mentary programs geared toward qualitycontrol and quality assurance.

New delivery methods that permitincreased additive concentrations and pin-point placement of materials, such asencapsulation techniques for fracturing fluidbreakers, have provided breakthroughs inpumping and treating procedures.8 Multi-functional additives have helped limit the

36

number of fluid components needed for agiven functionality, thereby reducing com-plexity, simplifying chemical and physicalinteractions within the fluid and the forma-tion, reducing inventory requirements, andincreasing mixing and blending efficiency atthe wellsite.

Reformulation often is not elective—itbecomes mandatory if one or more rawmaterials is restricted or no longer manufac-tured. To lower risk, especially for criticalchemicals that maintain well productivity, itpays to anticipate—particularly if availabil-ity is limited to a single vendor and continu-ity of supply cannot be assured. In theseinstances, reformulation entails extensivetesting and evaluation of a suite of alterna-tive materials from multiple vendors todetermine the most cost-effective replace-ment candidate. Care must be taken toensure that product performance does notsuffer and that there are no unexpected

handling or mixing constraints. Industry-wide, a concerted effort on product refor-mulation has helped guarantee a continu-ous supply of laboratory-optimizedproducts and spurred competitive pricingamong raw material vendors who offer sim-ilar product lines.

The oil crisis of the mid-1980s and enact-ment of environmental regulations haveundoubtedly had the greatest impact onspecialty chemical reformulation efforts.Chemical-cost reduction became a poten-tially quick and easy way to reduce wellsitecosts and improve profitability in the face ofdepressed oil prices. As with replacementmaterials that broaden production anddelivery capabilities, the main goal was tointroduce alternative, lower-cost materialsthat do not demonstrate adverse effects,while still adhering to the product’s originalperformance specifications.

For example, a high-volume corrosioninhibitor used in matrix acidizing stimula-tion treatments could no longer be pro-duced because one of its key ingredientswas no longer available. Thorough labora-tory testing led to a suitable material thatsatisfied four objectives simultaneously,exceeding expectations established at theoutset. The price of the final product couldbe reduced by 18% due to lower raw mate-rial costs and the new inhibitor demon-strated better dispersability in acid—a keycriterion for product performance. Corro-sion rates on samples of well tubing showedprotection equal to the existing material,and the reformulated product posedreduced handling and mixing hazards.

Product reformulation, however, takes onits most important aspect when safety orenvironmental considerations come intoplay. The basic chemistry of many specialtyoilfield products makes them potentiallyharmful if discharged into the environment.Reducing their impact may require elimina-tion of materials banned by regulatory man-date or incorporation of components thatoffer reduced levels of toxicity. In bothcases, the choice of replacements must besuch that acceptable product performanceis maintained. When environmentally sen-sitive materials form part of the reformula-tion equation, the evaluation and testingprocess becomes more complicated than inthe case of replacing materials for cost or

Oilfield Review

* For Skeletonema costatum

Chemical product

Product toxicity* (EC 50) Surfactant

Existing ReplacementExisting Replacement

Surfactant componentbiodegradability

Stimulationsurfactant

0.4-1.0 mg/L

0.27 mg/L

3.94 mg/L 80%82%62%

78%25%—

42%27%

0% 94.6%

85%—

1.89 mg/L

123

12

2.57 mg/L

0.32 mg/L

Acid-corrosioninhibitor

Acid-gellingagent

■■Results of product reformulation efforts for the North Sea. Three products were reformu-lated to improve environmental acceptability. Each new material showed lower toxicityvalues—as measured by EC50 standards on Skeletonema costatum (algae)—andimproved surfactant biodegradability. [* EC50 is a European Community-approved pro-cedure for measuring the toxicity of various chemicals to marine life. The EC50 value isthe concentration at which 50% of the species exposed to the chemical survive.]

8. Gulbis J, King MT, Hawkins GW and Brannon HD:“Encapsulated Breaker for Aqueous Polymeric Fluids,”paper SPE 19433, presented at the 9th SPE FormationDamage Symposium, Lafayette, Louisiana, USA,February 22-23, 1990.

9. O’Neill JE and Hill DG: “Reduction of Risk to theMarine Environment from Oilfield Chemicals—Bal-ancing Environmental and Technical Needs,” paper

performance reasons. Interplay amongcomposition, performance, toxicity andultimate environmental impact must beconsidered, requiring a phased approach todesign, testing, cross-checking and applica-tion9 (previous page).

In 1995, a replacement program for specificproducts used in the North Sea was initiatedat the Dowell product center in Tulsa. The ini-tial objective was to replace eight chemicals,selected based on criteria established byclients and usage volumes. The goals were toeliminate any nonylphenol surfactants—which had been banned in selected areas ofthe North Sea—decrease product toxicityand improve biodegradability.

Four products in particular—a stimulationsurfactant, acid-corrosion inhibitor, acid-gelling agent and brine viscosifier—requiredextensive reformulation work. Each pre-sented the team of chemists working on theprogram with different challenges—dictatedby chemical composition, complexity andintended use. For the stimulation surfactant,a suitable reformulated product was devel-

Spring 1997

SPE 35946, presented at the Third International Con-ference on Health, Safety and Environment in Oil andGas Exploration and Production, New Orleans,

oped that exhibited lower toxicity without asignificant loss in product performance. Forthe acid-corrosion inhibitor, an alternativeproduct showed improved biodegradabilityfor all but one component, and dramaticallyreduced toxicity. Reformulation of the acid-gelling agent resulted in a product withequivalent performance, and eliminated thenonylphenol surfactant previously included.Studies of the brine viscosifier showed thatnovel chemistry would need to be devel-oped before a long-term solution could beachieved (above).

While this highly successful program isonly one example of numerous studiesbeing conducted within the industry, it wasalso time-consuming and expensive—requiring 18 months and costing about$75,000 US per product. Extending suchstudies to reformulate an entire slate of hun-dreds of specialty chemical products is amajor undertaking. It is critical to balancethe objective of improved environmentalcompatibility with the cost and technicalrequirements involved so that maximumbenefit can be achieved in the most expedi-tious and cost-effective manner.

Minimizing Waste DischargeDaily, millions of barrels of drilling andtreating fluids are pumped into wells aroundthe world. The industry’s goal is to pump theminimum volumes necessary to achievedesign objectives—in other words, use thehighest-performing, most time- and cost-effective fluids possible. In most cases, thismeans minimizing the amount of fluid lostto the formation. In wells that contact signif-icant intervals of highly permeable forma-tions, this is no small challenge. Equallyimportant—since the introduction of stricterenvironmental rulings—is the reduction orelimination of waste streams at the surfacethat require treatment or disposal, particu-larly since this treatment or disposal may bevery expensive (next page).

Downhole, progress in fluid design andspecialty additive formulation has pushedoperating limits forward. Today, we seedrilling fluids with lower static anddynamic filtration rates that still giveacceptable penetration rates; fracturing flu-ids with greater fluid efficiency, that allowthe creation of deeper, wider fractures atthe same stimulation treatment volume; andloss-controlled matrix stimulation fluids thateliminate wormholing and rapid acidspending, react with more formation sur-face area and provide an etching patternthat leads to greater stimulation.10

At the surface, from the spudding of a wellthrough completion, first oil or gas delivery,and ongoing production, the industry hasdramatically reduced the volumes of fluidrequired for disposal on and around thewellsite. Partially closed or completely self-contained, non-effluent drilling mud treatingsystems have eliminated the need foronshore reserve pits and offshore dischargesinto the marine environment. The advent

37

Louisiana, USA, June 9-12, 1996.

■■Minimizing fluid lost to the formation and wellsite waste discharges. Fluid lost duringwell drilling and treatment can be reduced through application of innovative chemicaltechnology that simultaneously decreases formation damage. At the wellsite, use ofclosed-loop mud treatment equipment, continuous-mix systems and container (tote tankand drum) recycling programs can dramatically reduce cost and environmental impact.

Drilling StimulationReduced static and dynamic fluid loss

Increased fracturing fluid efficiency

Closed drilling mud treatment system Continuous versus batch mixing

Tote tank and drum recycling

Minimize fluid lost to the formation

Minimize surface waste discharge

Mitigation of acid wormholing

Water

Additive 1

Pump Additive 2

and routine use of continuous-mix systems,replacing batch-mix systems, have signifi-cantly decreased or eliminated tank bottomsby allowing on-the-fly preparation anddelivery of fluids with adjustable

38

properties.11 Continuous-mix processes cangreatly improve technical and environmen-tal performance and represent the applica-tion of hazard control through engineeringadvances and risk management.

Stimulation vessels operating in the NorthSea were among the first to employ continu-ous-mix technology. A typical stimulationtreatment may require up to 200,000 gal[757 m3] of hydrochloric acid containing avariety of chemical additives. Previously,materials were batch mixed up to a day inadvance. The potential waste to be disposedof from tank bottoms could be as high as 15to 20% of the total treatment volume, or upto 80,000 gal [302 m3]. Today, with continu-ous-mixing methods, a similar size treat-ment generates no waste.12

Of the various practices that have aidedthis process, one that has had a majorimpact is reduction in the number of dispos-able chemical containers—through recyclingand alternative technology. Until the early1990s, 55-gal steel and plastic drums werethe preferred method for delivering liquids tothe wellsite—comprising nearly 100% of thepackaging produced by many specialtychemical manufacturing plants. Drums wereconvenient, widely accepted, readily avail-able and relatively cheap to transport.

The problem came with the end user inthe field. Inventories of used drums at well-sites and at service company field officesgrew, creating a major disposal problem.Direct disposal costs, often in the range of$7 US per drum, and the potential for envi-ronmental liability were a growing con-cern. As business activity increased, so didthe number of drums requiring disposal,increasing costs and cutting into margins.Rules governing drum disposal in the USAvary, depending on size and constructionmaterial. The combination of logistics andaccompanying documentation, compli-cated by numerous stocking points anddrum usage exceeding tens of thousandsannually, presented a sizable challenge forservice companies. The challenge extendedto Canada, where drum disposal costswere even higher and it was necessary incertain locations to stockpile containersuntil a credible disposal firm could beidentified. With time, the problem beganaffecting many oil-producing countriesaround the world.

To meet the challenge in the USA, somecompanies opted for interim solutions—implementation of 55-gal drum recyclingprograms or providing materials in smallervolume containers that were easier to dis-pose of or could be recycled. The formerentailed empty container transport back tothe origination point, washing, disposal ofwash materials and refilling. In many cases,

Oilfield Review

■■Reusable chemical tote tank. This standard Nalco/Exxon 4- × 4- × 4-ft [1.2- × 1.2- × 1.2-m] reusable chemical tank holds 375 gal [1419 L].

10. Fluid efficiency is defined as the volume of fluidremaining in the fracture divided by the volume offluid pumped into the fracture. Higher efficiencymeans less fluid lost to the formation.

11. Geehan T, Helland B, Thorbjornsen K, Maddin C,McIntire B, Shepherd B and Page W: “Reducing theOilfield’s Environmental Footprint,” Oilfield Review2, no. 4 (October 1990): 53-63.

12. O’Neill and Hill, reference 9.

the costs incurred exceeded those for drumdisposal and, with time, several companiesdropped or cut back on such programs. Inthe latter case, some companies are nowfocusing on recycling smaller, 5-gal [18.9-L]containers, and the containers are returned,shredded, remanufactured and then reused.

A successful long-term approach has beenthe use of returnable stainless steel andcomposite tote tanks, available in a varietyof sizes with 100-, 150- and 330-gal [378-,567- and 1249-L] most common. (above).Built for durability, these containers have alife expectancy of five years or more. Withinmany companies, the tote-tank program wasbegun on a trial basis as a supplement tocontinuing delivery of the bulk of their prod-ucts in drums. As the benefits of thisapproach were demonstrated, the initiativegrew and currently many service companiesand specialty chemical manufacturingplants use tote tanks as the primary vesselsfor liquid products. The Dowell Tulsa plant,for example, ships 98% of its liquid prod-ucts in tote tanks and the remaining 2% incomposite, instead of steel, 55-gal drums.

Spring 1997

Tote tanks reduce overall material deliverycosts to the wellsite. While requiring an ini-tial capital investment and ongoing freightand handling costs for transportation backto the chemical plant, these costs are morethan offset by savings in drum costs, dis-posal fees and reduction of environmentalexposure. Long-term savings outweigh ini-tial investment and maintenance costs. Asthe number of tote tanks has increased,suppliers have learned how to optimizedelivery and return logistics to lower trans-portation costs.

What the Future HoldsProducing oil and gas as cheaply and effi-ciently as possible requires cost-effectivespecialty chemical additives tailored to pro-vide optimal well drilling, completion andintervention services. The past decade hasseen a concerted effort by chemical manu-facturers and service companies to improvequality and reliability, and extend the opera-tional capabilities of these materials. Byusing the latest tools and techniques, muchhas been accomplished, with major benefitsboth in the plant and at the wellsite.

For the future, there will be a continuingdrive for efficiency and productivity in everyaspect of oilfield operations. This willinclude expansion of synergistic efforts inthe specialty chemical sector—just-in-timemanufacturing and inventory practices,quality control and quality assurance pro-grams that utilize the latest informationtechnology, product cost and performanceoptimization through reformulation, andeven greater emphasis on environmentalcompatibility. The cornerstone of today’s oil-field business is delivering solutions, ratherthan simply supplying products and ser-vices. This is the key to successfully leadingthe industry forward, and is especially truefor specialty oilfield chemicals.

—DEO/KPR

39

New Dimensions in Modeling Resistivity

Over sixty years ago, resistivity modeling

emerged as a means to design electrical

logging tools and interpret their

responses in simple, layered formations.

Computational advances now make it

possible to rapidly predict unexpected

responses from modern tools in complex

formations, and to more accurately

extract true resistivities and structural

geometries from resistivity logs.

Barbara AndersonVladimir DruskinTarek HabashyPing LeeMartin LülingRidgefield, Connecticut, USA

Tom BarberGreg GroveJohn LovellRichard RosthalJacques TabanouSugar Land, Texas, USA

David KennedyMobil E&P Technical CenterDallas, Texas

Liang ShenUniversity of HoustonHouston, Texas

■■Resistor network. The resistor networks were analog computers for two-dimensional (2D)resistivity models. These devices led to predicting tool responses relative to true formationresistivity. The last resistor networks used at Schlumberger were two racks about 6 ft [1.8 m] tall and 12 ft [3.6 m] long with nearly half a million resistors.

In 1927, Conrad Schlumberger changed thecourse of petroleum exploration when hepurposefully tipped his surface electricalprospecting array vertically and sent it log-ging down a wellbore.1 Existing resistivitymodeling methods were applied to the newgeometry to predict the vertical arrayresponse.2 Thus modeling is not new; but itis finding new uses.

In the logging vernacular, the term model-ing refers to computing a logging instrumentresponse in the presence of the environmentsurrounding the logging instrument. Themodel attempts to capture all the detailrequired to account for and duplicate anobserved instrument response, and it is usedas an aid to log interpretation or tool design.

A resistivity model consists of a resistivitydistribution for the environment and adescription of the instrument sensors. Theresistivity distribution includes the resistivityof various borehole and formation regionsand locations of their boundaries. Depend-ing on the logging situation, the boundariesmight include the locations of geologicalbedding planes, borehole wall, invasionfronts, and the relative angle between all ofthese and the instrument axis. The result of

Spring 1997

1. Schlumberger AG: The Schlumberger Adventure. NewYork, New York, USA: ARCO Publishing, Inc., 1982.

2. Allaud LA and Martin MH: Schlumberger The Historyof a Technique. New York, New York, USA: John

For help in preparation of this article, thanks to DavidAllen, Robert Freedman and Koji Ito, Schlumberger Wire-line & Testing, Houston, Texas, USA; and T. S. Ramakrish-nan, Schlumberger-Doll Research, Ridgefield, Connecti-cut, USA.AIT (Array Induction Imager Tool), CDR (CompensatedDual Resistivity), DIL (Dual Induction Resistivity Log),DLL (Dual Laterolog Resistivity), FMI (Fullbore FormationMicroImager), INFORM (Integrated Forward Modeling),MicroSFL, Phasor, PLATFORM EXPRESS, PrePlus, RAB (Resis-tivity-at-the-Bit) and SFL (Spherically Focused Resistivity)are marks of Schlumberger. RtBAN is a mark of Z&SConsultants, Inc.

modeling is a synthetic tool response, orlog, which can be compared to an observedlog for interpretation or to the resistivity val-ues in the model for tool design (see “Whatis a Model?” page 43).

In the early days of resistivity modeling,scale models and analog computers—forexample resistor networks—helped to char-acterize tool responses in cases too compli-cated for mathematical models, such aswhen radial and vertical resistivity variationswere to be considered simultaneously(above). The resistor network became one ofthe earliest applications of resistivity modelsin the creation of log correction chartsdescribing the effects of one-dimensional(1D) and two-dimensional (2D) environ-mental heterogeneities (such as changes inthe vertical direction like shoulder beds andin the radial direction as with invasion) onlogs. Typical cases were computed for thecharts, but intermediate or extreme casesencountered in practice required interpola-tion within a chart or among several charts.The drawback to this kind of modeling wasthe difficulty in varying the parameters ofthe problem; each new case required a newscale model to be built or hundreds to thou-sands of resistors to be exchanged.

With the development of digital comput-ers, direct numerical solutions of the elec-tromagnetic equations can be used to com-pute resistivity response in models having

complex geometries. To achieve this, theseequations are approximated by a large sys-tem of simultaneous equations; the next taskis to find numbers that simultaneously sat-isfy all the resulting equations. The geomet-rical restrictions are much less severe thananalytical and scale models. Realistic prob-lems can be formulated, and the solutionsappear as simple and unthreatening num-bers rather than as arcane and difficult-to-compute mathematical functions. Unfortu-nately, the computational burden isenormous; solutions for realistic three-dimensional (3D) problems by this methodrequire extensive machine computation (see“The Vocabulary of Resistivity Modeling,”page 46). On the other hand, machinecapability has steadily improved for morethan 50 years, and it is now possible toobtain useful numerical solutions to for-merly intractable problems.

Most log analysts appreciate that tool-response modeling plays an important rolein instrument design. Less well recognizedis that both the apparent resistivity recordedby the logging tool itself, and the variouscorrections thereto, are derived from mathe-matical models. The models help relate toolresponses to true formation resistivity, Rt.

41

Wiley & Sons, 1977.

3. Anderson B, Minerbo G, Oristaglio M, Barber T,Freedman B and Shray F: “Modeling ElectromagneticTool Response,” Oilfield Review 4, no. 3 (July 1992):22-32.Gianzero S: “The Mathematics of Resistivity andInduction Logging,” originally published in The Tech-nical Review 29, no. 1 (March 1981): 4-32;reprinted in Resistivity Logging. SPWLA Reprint Vol-ume Series 33, Houston, Texas, USA: University ofHouston (June 1992): 1-19.

4. The logging corrections are for parasitic effectsincluding adjacent beds, simple invasion of drillingmud, borehole, and skin effect (which causes a nonlinear decrease in signal strength in conductiveformations).

5. Koelman JMVA, van der Horst M, Lomas AT, KoelemijAT and Bonnie JHM: “Interpretation of ResistivityLogs in Horizontal Wells,” Transactions of the SPWLA37th Annual Logging Symposium, New Orleans,Louisiana, USA, June 16-19, 1996, paper G.

6. Anderson B, Barber T, Singer J and Broussard T:“ELMOD—Putting Electromagnetic Modeling toWork to Improve Resistivity Log Interpretation,”Transactions of the SPWLA 30th Annual Logging

In the last decade, modeling provided thebasis for automatic environmental correc-tion algorithms such as Phasor processing,and is the basis for the log products of theAIT Array Induction Imager Tool. For both ofthese tools, shoulder effect and thin-bedresponse are corrected to the resolution lim-its of the tool at low dip angles. Furthermodel-based processing allows responsecorrections for moderate dip angles (< 50°)for these tools.

Although modern resistivity tools performmost environmental corrections as the logis being recorded, oil company archives arebulging with logs run with older DIL DualInduction Resistivity Log and SFL Spheri-cally Focused Resistivity tools and equiva-lents from many service companies. Forthese tools, modeling is the only way tounderstand the complex effects that causedtheir logs to read completely wrong inmany situations.

Tool-response modeling, as practiced inmodern log interpretation, extends the inter-pretation of instrument responses beyondthe simple resistivity distributions envisagedin the apparent resistivity function and itsfirst-order corrections—as provided in chart-books—to more realistic geometricalarrangement of the borehole and formationenvironment. The idea is to discover theprobable distribution of resistivity thatwould give rise to an observed toolresponse. In many cases, only tool-responsemodeling can decode what the instrumentresponses reveal about the formation beinglogged—in many cases, if not most, intu-itions gained from study of, and experiencewith using, charts are misleading, becausethe charts are not applicable.

Resistivity modeling was reviewed in Oil-field Review five years ago.3 The two majoradvances chronicled at the time weregeosteering with logging-while-drilling(LWD) resistivity tools and the combinationof analytical and numerical modeling meth-ods to speed calculations. At that time, twoneeds were identified. First, modelersrequired enhanced computational capabilityto develop models in the 3D domain formore realistic analysis in complex forma-tions. Such improvements—already on thehorizon—would, in part, come from devel-oping new advanced algorithms that speedup lengthy calculations. Second, furthertechnological improvements were neededto help log analysts model logs quicklyenough to be practical. Some of this capa-bility was already under development insome oil company research groups.

42

Symposium, Denver, Colorado, USA, June 11-14,1989, paper M.

This article discusses the progress made in the past five years. We first show how 3D modeling is used for strongly dippingformations and horizontal well evaluation.Then, we discuss the impact of recent com-putational breakthroughs in modeling capa-bilities and practical interpretation usingresistivity modeling.

Why Model Resistivity? Reserves estimates are based on log-derivedmeasurements of formation resistivity. Allother factors being constant, as the forma-tion hydrocarbon content increases, so doesthe formation resistivity. In fact, using Rt,Archie’s law allows computation of the frac-tional volume of the formation containinghydrocarbons. However, a few operatorsuse raw resistivity logs as a hydrocarbonindicator—where resistivity is high, theyperforate the well. This is not the best use ofresistivity data.

The log analyst looking at resistivity logsfrom older tools—uncorrected for environ-mental effects—often has little to serve as aguide for resistivity interpretation. Manytimes the uncorrected resistivity, for exam-ple from the deep induction tool, is consid-ered an accurate enough estimator of Rt.This may sometimes be true, but all resistiv-ity instrument responses, whether frominduction or laterolog tools, are influencedby the resistivity distribution in a large vol-ume surrounding the logging instrument.Correct interpretation would require manycorrections for bed boundaries, borehole,invasion and other environmental or geo-metrical effects.

This means that what looks like the best,or highest, resistivity reading on a rawinduction log may simply be an artifact ofsome nearby boundary layer of contrastingresistivity. Many high-resistivity anomaliesdo not correspond to resistive beds at all,but rather to the interface between twobeds, each with lower Rt than the apparentresistivity, Ra, indicated on the log. Undersimple enough conditions, such as in a ver-tical well through thick beds with simpleinvasion, chartbook corrections can suffice.4For these applications, many software sup-pliers offer chartbook-based corrections forborehole and invasion for the DIL tool.

Wells are not always ideal for logging.Low-cost drilling efforts frequently lead toboreholes that are not uniform. Deep inva-sion also affects tool responses. In the lastdecade, thousands of deviated and horizon-tal wells have been drilled to optimize pro-ductivity. These wells have logging tool-to-formation orientations not contemplated inthe design of the apparent resistivity

response and the corresponding chartbookcorrections. Resistivity responses in suchformations appear counterintuitive, evenweird, and are not correctable by chart-book algorithms.

However, the physical principles—embod-ied in Maxwell’s equations—of such instru-ment responses are well understood, andwith modeling capable of honoring enoughdetail of the tool and environment, suchresponses are predictable. Artifacts such aspolarization horns, which appear as largetransient overshoots in the tool resistivityresponse at bed boundaries, can be under-stood on the basis of such models.

Another Dimension in Resistivity ModelingThe growing interest in complex formationgeometries, such as dipping beds, invasionand anisotropy, has led to progress in devel-opment of sophisticated models andenhanced computational efficiency. Asinterpretation problems have become moredetailed, model requirements have grownfrom 1D, to 2D, and finally to 3D codes, tohandle the geometry of the problems athand. Significant recent developments havebeen the use of new numerical techniques,which make solutions in 3D models muchfaster. Now, they are beginning to be usedfor difficult log interpretation problems.5

Compared to current methods, early mod-eling codes were slow in execution andtherefore cumbersome to use for forwardmodeling. Models had to be modified ateach iteration by editing detailed text files,and comparison of observed and syntheticresponses required hard-copy plotting of thesynthetic log, and visually inspecting for dif-ferences at each iteration.6 Nevertheless, thevalue added was significant.

Oilfield Review

What is a Model?

Predicting the response of a logging tool is a compli-

cated, detailed process. The boundaries of regions of

differing resistivity (or other material electromagnetic

properties) and the properties within each region

must by specified by the model. The tool must also

be introduced into the model with its transmitters

and receivers located at specified points, with the

transmitters exciting the model medium in some

specified manner. The interaction of the disturbance

produced by the transmitters proceeds through the

medium and is detected by the tool’s receiver coils or

electrodes.

The response of the each receiver is thus pre-

dicted. Resistivity modeling belongs to the class of

so-called boundary-value problems. In relatively sim-

ple cases, these problems may yield analytic formu-

las, but, in general, a system of equations must be

solved explicitly (though approximately) for each new

case considered—such as when the tool is moved or

the model changed—and the result is a set of num-

bers rather than a formula. This type of mathematical

modeling is called numerical modeling. Two exam-

ples of numerical modeling in use today are the

finite-element method (FEM) and finite-difference

method (FDM).

Given a system of equations governing a tool

response, it is straightforward to compute the tool

response. One simply plugs in model parameters and

after computation, the desired tool response quanti-

ties are written down. The result is an array of tool

responses that would be observed if a physical exper-

iment were performed. This is called the forwardproblem. For cases where the governing equations

are linear in the resistivities, this process is partially

reversible—given the tool response, the model

1. Kennedy WD: “Induction Tool Forward Modeling: A Rigorousand Systematic Approach to Model Construction,” Transac-tions of the SPWLA 36th Annual Logging Symposium, Paris,France, June 26-29,1995, paper G.

2. Anderson BI and Barber TD: Induction Logging. Sugar Land,Texas, USA: Schlumberger Wireline & Testing,1997.

Spring 1997

parameters can be estimated by multiplying the vec-

tor of observed tool responses by the generalized

inverse of the same matrix used in the corresponding

forward problem. This is called the inverse problem.

There are variations upon this theme. For some sys-

tems, the output of the system—tool response— can

be represented as the mathematical operation of con-volution of the tool input with the impulse response of

the tool. Given the output (a log), the input (a forma-

tion) can be correctly determined if the transfer func-

tion (the Fourier transform of the impulse response)

does not pass through zero for any value of its argu-

ment. Under this restriction, the input is determined

from the output by the process known as deconvolu-tion.

Unfortunately, this is not the case for many tools.

For example, the 6FF40 is known to have a blind fre-

quency corresponding to about 5-ft intervals.1 As a

result, there is a nonunique inversion—for practical

purposes, the thin-bed response at blind frequencies

turns out to be indistinguishable from the whole-

space response.

Moreover, the response of resistivity tools is usu-

ally nonlinear with respect to formation resistivity

because of skin effects. Thus, in the strict sense men-

tioned above, neither deconvolution nor inversion is

possible—at least not in a single, deterministic step.

Furthermore, the forward problem itself is relatively

difficult to solve. It typically is represented by a large

system of equations. The solution of the equations

may be relatively straightforward, but since the sys-

tem is large, extensive computational resources and

time are required.

The inversion or deconvolution of a nonlinear sys-

tem requires that the forward problem be “lin-

earized” by some means. Phasor processing of an

induction log response provides an example of lin-

earizing an instrument response. The induction tool’s

receiver in-phase or resistive voltage, the R-signal, is

deconvolved using a filter designed to correct the

zero resistivity response. The tool’s quadrature or

reactive voltage, the X-signal, is processed with fil-

ters and fitting functions to approximate the differ-

ence between the actual R-signal and a linear esti-

mate of the R-signal.2 The two are added, and the

result is a linear output with the shoulder effect cor-

rected. In enhanced resolution Phasor processing,

the log resolution is corrected for beds as thin as 3 ft

[1 m].

A more versatile technique finds a solution by itera-

tively solving the forward problem. If the iterations are

performed manually, the process is called forwardmodeling. If the iterations are carried out by an algo-

rithm seeking to minimize the differences between

the observed log and a synthetic log calculated in a

model by adjusting the model to reduce the differ-

ences at each step, the process is called inversemodeling. Note that for either technique, an important

part of the process is an efficient means of solving

the forward problem so that the differences can be

obtained rapidly and the iteration carried forward as

fast as possible.

Equally important, especially for inverse modeling,

is the construction of the initial model. If the initial

model is good enough, the differences between

observed and synthetic logs vanish on the first com-

parison. Failing this, the closer the initial model is to

the actual formation, the fewer the iterations

required, and the final solution is reached rapidly.

Thus, the best possible initial model is essential to

inversion. An initial model would be based upon the

highest resolution data available plus any available

43

Y1

X1 X2 X3 XNX

Z1

Z2

Z3

ZNZ

Y2

Y3

YNY

σ(x, y, z)

Borehole center line

■■Typical 2D finite-element grid for resistivity problems.The grids used for defining the nodes used in finite-ele-ment models usually conform to the natural geometryof the environment, in which edges and intersectionsbetween different geometrical regions with differentproperties (resistivities) lie on the grid.

estimates—such as using FMI Fullbore Formation

MicroImager readings to get bed boundaries—of

radial resistivity variation within the bed boundaries.

The type of inverse modeling described above is

known as model-based inversion. This means that

the form that the solution is most likely to take is

decided in advance, and only solutions of that form

will be found by the technique.

Differences Between FEM and FDMIn the finite-element method, Maxwell’s equations

are written as integral equations of the resistivity and

electromagnetic fields. The next step is a process of

discretization throughout the medium surrounding the

tool and formation—the step of converting continu-

ous equations to a finite dimensional system of

equations that can be solved with a digital computer.

For FEM, discretization is done based on a variational

principle. For example, this could mean the total

energy of the system is minimized at points on a

gridded network throughout the media environment

(above right). This is a reasonable way to solve the

problem, since the static electromagnetic potentials

always adjust themselves to seek a minimum energy.

The topology of the discretization is usually selected

to conform with the problem geometry. The dis-

cretization step leads to a large set of simultaneous

equations for the electromagnetic field in terms of

the media resistivities, which are represented by a

set of large matrices.

The flexibility of FEM translates to complexity in

numerical implementation and the codes can be

more difficult to optimize. The form of the matrices is

unstructured, having its nonzero elements spread

throughout the matrix, because these matrices reflect

the topology of the discretization.3 The final step is to

find a solution to the matrix equations representing

the electromagnetic fields. For a problem of a given

size, the FEM method is capable of the highest con-

formity to geometrical complexity of any numerical

■■Grid and coordinate system used for 3D finite-differ-ence method calculations. The staircase shows how adipping boundary is implemented in the Cartesian gridsystem. The material properties (resistivity) of the cellsintersected by the staircase boundary are averaged in

3. Lovell, reference 12, main text.4. Druskin V, Knizhnerman L and Lee P: “Solutions of Maxwell’s

Equations Using Krylov Subspaces from Inverse Powers ofthe Stiffness Matrix,” presented at the Thirteenth AnnualReview of Progress in Applied Computational Electromagnet-ics, Monterey, California, USA, March 17-21,1997.

44

method, but is computationally slow.

The finite-difference method uses discretization

based on a direct difference approximation of the dif-

ferential operator form of Maxwell’s equations. This

leads to a grid requirement restricted to a Cartesian

topology (right). Although the grid topology does not

conform to the geometry of the resistivity discontinu-

ities found in the formation environment, it can be

made to approximate this geometry. The real benefit

of this approach lies in the next step, in which the

matrix equations for the discretization of the differen-

tial equations are usually structured because of the

Cartesian topology, and always sparse because the

derivatives are local operators. With these advan-

tages, matrices lend themselves to fast, specialized

computational methods that allow rapid solutions to

resistivity models with extremely complex geometry.4

the calculation.

Oilfield Review

7. Richard Hardman, personal communication, 1997.8. Anderson B and Gianzero S: “Induction Sonde

Response in Stratified Media,” The Log Analyst 24no. 1 (January-February 1983): 25-31. Anderson B, Safinya KA and Habashy T: “Effects ofDipping Beds on the Response of Induction Tools,”paper SPE 15488, presented at the 61st SPE AnnualTechnical Conference and Exhibition, New Orleans,Louisiana, USA, October 5-8, 1986.

9. Druskin V and Knizhnerman L: “A Spectral Differen-tial-Difference Method for the Numeric Solution of3D Nonstationary Problems of Electrical Prospect-ing,” Izvestia Academy of Sciences, U.S.S.R. Physicsof Solid Earth 24, no. 8 (1988): 641-648. (Englishedition published by American Geophysical Union.)

10. Druskin V and Knizhnerman L: “Krylov SubspaceApproximation of Eigenpairs and Matrix Functions inExact and Computer Arithmetic,” Numerical LinearAlgebra with Applications 2, no. 3 (1995): 205-217.See also Druskin V and Knizhnerman L: “SpectralApproach to Solving Three-Dimensional Maxwell’sDiffusion Equations in the Time and FrequencyDomains,” Radio Science 29, no. 4 (July-August,1994): 937-953.

11. Hardman RH and Shen LC: “Charts for CorrectingEffects of Formation Dip and Hole Deviation onInduction Logs,” The Log Analyst 28, no. 4 (July-August,1987): 349-356.

12. Lovell JR: Finite Element Methods in Resistivity Log-ging. Ridgefield, Connecticut, USA. SchlumbergerTechnology Corporation, 1993.

13. Hardman RH and Shen LC: “Theory of InductionSonde in Dipping Beds,” Geophysics 51 (March1986): 800-809.Anderson et al, 1986. reference 8.

14. Howard AQ and Chew WC: “Electromagnetic Bore-hole Fields in a Layered, Dipping-Bed Environmentwith Invasion,” Geophysics 57 (March 1992): 451-465.

(continued on page 49)

Consumers of well log data concludedthat to gain maximum benefit from resistiv-ity logs, the numerical modeling of toolresponses would be the standard for the1990s. Several oil companies, among themShell, BP, Chevron and ARCO, began todevelop or acquire in-house modeling capa-bilities. And one major oil company, Mobil,formed a task force of researchers, practic-ing petrophysicists, computer scientists andacademic consultants to make logging toolresponse modeling convenient, practicaland routinely applicable. This commitment,combined with several technical break-throughs, brought rapid success to this effort(see “An Efficient User Interface for Resistiv-ity Modeling,” page 49).

As early as the late 1980s, computationalefficiency was benefiting from the newlydeveloped fast Hankel transform (FHT)applied to induction log modeling, whenRichard Hardman invented a method ofoptimizing the use of the FHT.7 The numberof computations was further reduced byexploiting various symmetries of the induc-tion antenna array and reciprocity. Availableby 1991, the resulting 1D computer pro-grams were from 100 to 600 times faster,depending on dip, than the standard codesused by University of Houston Well LoggingLaboratory.8 A recent revision to Hardman’s1D induction modeling will be available inGeolog6 software from MINCOM, Pty. Ituses shallow measurements, such as SFLlogs, as input to assign bed boundaries.Model viewing and editing are performedrapidly through a graphical interface.

During the last decade, at the CentralGeophysical Expedition in Moscow,Vladimir Druskin and Leonid Knizhner-man—using 3D modeling techniques origi-nally developed for surface electromagneticprospecting problems—developed a break-through in a new, efficient 3D finite-differ-ence method (FDM) for resistivitymodeling.9 Their approach, along with otherrecent developments in their mathematicaltechniques, has led to remarkable improve-ments in the speed and accuracy of full 3Dresistivity modeling capability for bothinduction and 2-MHz propagation tools.10

Spring 1997

15. Anderson B, Barber T, Druskin V, Lee P, Dussan V,Knizhnerman L and Davydycheva S: “The Responseof Multiarray Induction Tools in Highly Dipping For-mations with Invasion and in Arbitrary 3D Geome-tries,” Transactions of the SPWLA 37th Annual Log-ging Symposium, New Orleans, Louisiana, USA,June 16-19, 1996, paper A.

Horizontal and Highly Deviated Wells—3D Resistivity ModelingInterpretation of resistivity tools in horizon-tal or highly deviated wells is far more com-plicated than in vertical wells, particularlywhen drilling-induced resistivity changesare important to the tool response. Resistiv-ity tools have been designed for focusing invertical wells, where horizontal layers andaxial symmetry are assumed. When theseconditions hold, interpretation proceduresare straightforward—the recommended pro-cedure is to apply borehole corrections,process to improve or match vertical resolu-tion, and correct for invasion, in sequence.Although dip-induced shoulder effects canbe corrected for angles up to 60°, in prac-tice invasion complicates the inter-pretation.11 At higher angles, shouldereffects can be different on all sides of thesonde, and invasion may be highly nonsym-metrical. As a result, apparent resistivityresponse interpretation using the standardmethods is not possible in horizontal andhighly deviated wells.

For laterolog-type tools, such as the DLLDual Laterolog Resistivity tool and the RABResistivity-at-the-Bit tool, effects of the bore-hole and possible invasion cannot beignored. In the presence of dip, such toolsmust be modeled with 3D codes. Fortu-nately, their low-frequency operation meansthat they can be modeled with Poisson’s orLaplace’s equation—a significant simplifica-tion of Maxwell’s equations—and finite-ele-ment method (FEM) codes for these equa-tions are relatively fast to compute. Forexample, a full 3D FEM solution for the DLLresponse in the presence of dip, pinchouts,anisotropy, invasion and tool eccentricitytakes 10 to 15 seconds per tool position.12

For induction-type tools, such as the AITresponse and CDR Compensated DualResistivity response, the effects of resistivemud drilling-induced alterations and bore-hole signals can be negligible, but the fre-quency effects on the tools are not. Thus, thefirst attempts to model their response inhighly deviated wells were based on solvingfor the electromagnetic fields in layered 1Dgeometries, which omitted both boreholeand invasion. These early attempts have metwith reasonable success.13 This class ofmodels is widely applicable, fast to com-pute, and commercially available. Subse-quent efforts based on 2D models haveincluded a borehole and axisymmetric inva-sion; these are also commercially availableand fast enough to be practical for con-

structing model-based interpretations.Unfortunately, nonaxisymmetric invasionalso exists, even at early stages of invasionencountered by LWD tools.14 Even wheninvasion is axisymmetric, in the presence ofdip this geometry requires a 3D solution.

Fast FDM modeling techniques are nowavailable for solving the full 3D electromag-netic problem necessary for interpretinginduction logs in deviated wells with inva-sion, and for understanding tool responsesin other difficult environments includingcomplex fracture systems, faults and pin-chouts, and effects of resistivity anisotropy.Finally, realistic logging problems, withcombined effects of invasion, dip and shoul-der beds, can be analyzed within reason-able time using the new 3D FDM inductionforward-modeling programs.15

45

The Vocabulary of Resistivity Modeling

Rm

r2

Rxo Rann

Rt

z

φρ

r1

rbh

■■1D invasion geometry. A common geometricalarrangement in well logging is a cylindrical wellbore sur-rounded by coaxial cylindrical shells bounding cylindricalregions of differing resistivity.

Geometrical Dimensions in ModelingThe physics of resistivity modeling is described by

Maxwell’s equations. Depending on the assumptions

made about the spatial distribution of resistivity,

these solutions can take the form of simple formulas,

giving the electromagnetic field for any desired point

in space. In complicated cases, the field cannot be

expressed in terms of simple formulas, but must be

laboriously computed at many points in space simul-

taneously. The degree of complexity is customarily

summarized by referring to geometries of differing

dimensionality—zero-dimensional (0D), one-dimen-

sional (1D), two-dimensional (2D) and three-dimen-

sional (3D).

One way to decode this jargon is to consider how

resistivity varies in a coordinate system.1 Consider,

for example, the cylindrical coordinate system speci-

fied by the coordinates ρ, φand z. If the resistivity, R,

is constant everywhere, then it is referred to as 0D. If

the resistivity, R, in a geometry varies with only a sin-

gle coordinate direction, then that geometry is

referred to as 1D. The familiar layer-cake geometry of

sedimentary geological structure is an example of

how the resistivity varies only in the z-coordinate

direction or, R = R(z) (below left). In essence, each

layer has a constant resistivity, but the resistivity

varies between the layers.

46

Rt0

Rt4

Rt3

Rt2

Rt1

y

x

z

φ

ρ

■■1D layered formation model geometry. The familiar layer-cake geometry of a sedi-mentary geological structure is a 1D resistivity geometry if the resistivity varies onlyin the z-coordinate direction.

■■2D layered formation vertical and radial direcas two-dimensional. Thieling programs today.

Another commonly used geometrical arrangement

in well logging features a cylindrical wellbore sur-

rounded by coaxial cylindrical shells bounding cylin-

drical regions of differing resistivity (left). In this

case, R = R(ρ). Thus,1D geometries are either verti-

cally or radially layered geometries.2

Similarly, if resistivity varies in the vertical and

radial directions and is axisymmetric, then the geom-

etry is two-dimensional (below right). In this case, R= R(ρ, z) in the notation introduced above.

When R = R(ρ, φ, z), the geometry is said to be

three-dimensional (next page, top). This system of

nomenclature can be extended without modification

to cases in which resistivity at a point varies with

direction—resistivity is a tensor.

To illustrate this, Maxwell’s equations are formu-

lated for the electric field of a magnetic dipole in an

infinite and homogeneous conductive medium, and

the result is a simple formula. The medium is zero-

dimensional, and the formula is the basis for calibrat-

ing the induction tool’s voltage measurement to

apparent resistivity. If the equations are formulated

for the electric field of a vertical magnetic dipole on

the axis of a radially layered medium or within a hor-

izontally layered medium, the results are still formu-

las for a specified ρ and z. The field can be computed

Oilfield Review

Rt0

∞

∞

Rt1

Rt2

Rxo1

Rxo2

Rm

Rm

with invasion model geometry. If resistivity varies in bothtions (and is axisymmetric), then the geometry is referred tos 2D geometry is used in many commercial forward mod-

Rsh1

Rt

Rsh2

Rm

Rxo

Rm

■■3D dipping formation with asymmetric invasion model geometry. 3D geometries such as this and those for frac-tures, nonuniform invasion fronts and dipping beds with laminated anisotropy are used for most realistic formationmodels.

1.This geophysics convention is adopted in this article.2.The third possibility for a 1D geometry, R = R(φ), does not cor-

respond to a case of much utility for resistivity modeling ofinduction tools, which have no sensitivity to φvariations, and isnot used.

Side view

Transmittermagnetic dipolein center ofborehole

Center lineof borehole-formationgeometry

Electric fieldin formation

Eφφ

Eφ

Top view

Rm Rt

mm

■■Electric field around apoint magnetic dipoleon the borehole axis.The magnetic fieldcomponents can beobtained from the sin-gle component of elec-tric field using Fara-day’s law of induction.The electric field isconcentric and sym-metrical with the axisof the borehole and istherefore completelyindependent of theazimuthal angle φ.

Transmittermagnetic dipolein eccenter ofborehole

Center lineof borehole-formationgeometry

Electric fieldin formation

Eφ

Eφ

Eρ

RmRt

E

mm

■■Electric field around apoint magnetic dipoleoff axis from the bore-hole, representing aneccentered tool in aborehole. In this eccen-tered tool geometry, thecylindrical symmetry ofthe induced electricfield is destroyed. Now,two components of theelectric field arerequired (a radial Eρ,and an azimuthal Eφcomponent) to com-pletely describe thenonconcentric electricfield in the formation.

without reference to the field at other points, but the

formulas are complicated and contain semi-infinite

integrals to be performed, after substituting parame-

ter values, to get a numerical answer. The geome-

tries are one-dimensional; these computations form

the basis for invasion corrections (radially layered 1D

geometry) and bed-thickness corrections (vertically

layered 1D geometry), and most commercially avail-

able 1D codes.

Models with geometries of higher dimensionality

usually cannot be solved analytically—the solutions

cannot be expressed as simple formulas. In these

higher dimensional cases, numerical methods must be

used to obtain the field values. Although there are

many different numerical methods, they all share the

same property: to compute the field at a single point,

the field must be known at many adjacent points. For

some methods, the field must be computed at every

point, and these computations are time-consuming.

Consequently, computations in 2D are about three

times slower than 1D computations, and 3D geome-

tries require two orders of magnitude more time than

1D computations.

A possible source of confusion arises when more

realistic modeling is attempted. To illustrate, consider

a vertically oriented, point magnetic dipole on the

axis of an axisymmetric, two-dimensional medium

(top right). At each point in space, the electromag-

netic field has six components: three components of

electric field and three components of magnetic field.

However, in the plane of the source, as long as the

Spring 1997

source is on the axis of symmetry, two of the electric

field components will be identically zero.

The magnetic field components can be obtained

from the single component electric field using Fara-

day’s law of induction. Thus, it is sufficient to formu-

late this problem in terms of the azimuthal compo-

Side view

nent of electric field alone. However, if the dipole is

moved off the axis of symmetry to represent an

eccentered tool in a borehole, the cylindrical symme-

try of the electric field is broken; now two compo-

nents of the electric field are required—radial and

azimuthal—to completely describe the electric field

(bottom).Numerical modelers sometimes refer to the

dimensionality of their models according to the num-

ber of non-zero field components necessary to com-

pletely specify the field. However, in this example, the

medium remains 2D irrespective of the source loca-

tion or orientation, even though the mathematical for-

mulation may have to account for a greater or fewer

number of field components according to the source

orientation and position.

47

Top view

20 ft

1 ohm-m

100 ohm-m

10 ohm-m

1 ohm-m

15 in.

■■Three-dimensional model for an invaded, dipping reservoir. The overlying and underly-ing formations are shales, while the reservoir, containing 90% oil saturation, is invaded to15 in. [38 cm] by saline water.

Resistivity, ohm-m1

-16

-14

-12

-10

-8

-6

10 in.

20 in.

30 in.

60 in.

90 in.

3Dmodel

Polarizationhorns

-4

-2

0

2

4

6

8

10

12

14

16

10 100

60° Dip

1000

True

ver

tical

dep

th, f

t

Rxo

ri =15 in.

Rt

■■AIT Array Induction Tool logs in dippingbed with invasion. The results of the 3Dmodeled induction responses (blue curves)show the 10-in. response in agreementwith the invaded zone resistivity, Rxo,(dashed green line). However, in the cen-ter of the thick hydrocarbon zone, thedeeper tool responses are reading onlyabout 40% of the true formation resistivity,Rt, (solid green line). This corresponds toan oil saturation of only 60%, a significanterror. The deepest reading 90-in. responsealso shows strong polarization horns at thebed boundaries.

Resistivity, ohm-m1

-16

-14

-12

-10

-8

-6

-4

-2

0

2

4

6

8

10

12

14

16

10 100 1000

True

ver

tical

dep

th, f

t

10 in.

20 in.

30 in.

60 in.

90 in.

10 in.

20 in.

30 in.

60 in.

90 in.

No dip

Rxori =15 in.

Rt

3Dmodel

2Dmodel

■■The effect of invasion without dip. In theabsence of dip, the case of a boreholethrough an invaded layer can be com-pletely modeled in 2D (red curves). Treat-ing the problem in 3D yields the sameresults (blue curves). Note that the deep90-in. tool response is greater than the Rt inthe center of the bed, while the shallower60-in. tool response is nearly equal to Rt.Also note the absence of horns at the bedboundaries. These effects are differentfrom those seen in dipping formations.

Resistivity, ohm-m1

-16

-14

-12

-10

-8

-6

-4

-2

0

2

4

6

8

10

12

14

16

10 100 1000

True

ver

tical

dep

th, f

t

10 in.

20 in.

30 in.

60 in.

90 in.

10 in.

20 in.

30 in.

60 in.

90 in.

60° DipNo invasion

Rt

3Dmodel

1Dmodel

■■The effect of dip without invasion. In theabsence of invasion, the case of a boreholethrough a dipping layer can be com-pletely modeled in 1D (red curves). Treat-ing the problem in 3D should, and does,yield the same results (blue curves). In thiscomparison, the bed boundary hornsappear on the shallow 10-in. reading aswell as on the deep 90-in. tool response.The deep 90- and 60-in. tool responses arereading much less than Rt, while the shal-low 10- and 20-in. responses are close to Rtin the center of the thick bed.

48 Oilfield Review

An Efficient User Interface for Resistivity Modeling

Modeling the Effects of Dip and Invasion.An example comes from a case in which theAIT tool logged through a 20-ft [6.1-m]resistive bed at an angle of 60° from vertical(previous page, top).16 The shallow 10-in.log reads near flushed zone resistivities, Rxo,as expected. However, the deep 90-in.curve falls way below Rt, and the 20-, 30-and 60-in. curves cross each other in a dis-orderly fashion, indicating a mixture ofshoulder bed and invasion effects (previouspage, bottom left).

To disentangle the different effects andcompare the relative effects of invasion andshoulder beds, the limiting cases of invasionwith no dip (previous page, bottom center),and dip with no invasion were modeled(previous page, bottom right).

First, in the case of invasion with no dip,the expected curve separation occurs, withthe 10-in. curve reading close to Rxo and thedeeper reading curves increasing insequence towards Rt, with the 90-in. curvereading above Rt. Next, dip with no inva-sion results show that there is a significantlylarge shoulder-bed effect on the deep curvesin this 20-ft bed at 60° dip. At center-bed,the readings decrease in sequence from theshallow to deep curves, with the 90-in. cen-ter-bed curve reading only 25 ohm-m in the100-ohm-m bed.

Comparing the combined dip and invasionresults with the results of the nondippingcalculations shows the 10-in. curves readthe same in the nondipping bed. The 20-and 30-in. dipping curves read a bit lowerthan those in the nondipping bed, indicatinga slight shoulder effect. Comparing the com-bined dip and invasion results with those ofthe noninvaded dipping bed model showsthat the 60- and 90-in. curves behave thesame, in both cases reading below Rt. Thisindicates that the deep-reading logs are notinfluenced by the shallow invasion, but thatthere is a considerable shoulder-bed effect,which in this case of conductive invasion(Rxo less than Rt) could be mistaken fordeeper invasion. Modeling of dip and inva-sion together and separately helps to inter-pret this difficult case. It is important toarchive all AIT raw tool logging responses inorder to take full advantage of the new, faster3D modeling capability.

Spring 1997

16. As a test, to verify that the curve crossover was notsimply an artifact of 3D code inaccuracy, the FDMcode was compared with an analytical solution inboth cases, since each limiting case can be solved ina 2D analytical model. Agreement within 3% wasachieved in all cases.

Before computational codes were improved, the time

the log analyst spent editing the model geometry file

between computations was comparable to computa-

tional time. By 1991,1D induction responses could be

computed at the rates of hundreds of feet per second.

As a result, editing time subsequently came to be

viewed as a major bottleneck. The Mobil Resistivity

Modeling Task Force was formed to specify and

develop a graphical user interface capable of display-

ing observed log curves in a user-selectable format,

and also capable of rapidly creating, deleting and

otherwise modifying resistivity models in the same

windows as the data using the point, click and drag

features available in the then new X-windows sys-

tem.

The interface is not itself a numerical modeling

code; rather, it is a graphical editor useful for quickly

constructing and modifying models. It incorporates

facilities to rapidly and automatically build initial

models using user-specified log response curves,

and is useful for the specification of 2D axisymmetric

models and 1D models with constant relative dip.

Commercially available modeling codes currently

adapted for use include forward and inverse 1D for

6FF40 and ILD deep induction logging, 2D for con-

ventional electric survey, 2D for laterolog and spheri-

cally focused logs, and 2D for induction. Other suit-

able codes are under development by various

vendors. The Oklahoma benchmark formation can be

computed in well under one second; log responses in

2D geometries take significantly longer (though not

prohibitive) times to compute over similar intervals.

However, the incentive (and reward) for 2D modeling

is the same as for 1D modeling—hydrocarbon pore

volume increases on the order of 5 to 15% from a

more accurate Rt analysis.

The resulting program was being routinely used by

Mobil log petrophysicists by 1992. The marriage of

fast,1D induction forward modeling to the graphical

user interface made forward modeling a practical

tool for routine log interpretation. Mobil licensed its

graphical user interface to Z&S Consultants, Inc.

where it has been available under the trade name

RtBAN since about 1993; to date it has been licensed

to a number of major and independent oil companies.

Petrophysicists experienced with the program and

familiar with the 6FF40 response regularly model

3000 to 5000 ft [900 to 1500 m] per day.

49

50 Oilfield Review

Resistivity, ohm-m

h varied

24 in.

Bed boundary

1-16

-12

-8

-4

0

4

8

12

16

20

24

28

32

10 100

Wel

lbor

e-to

-bou

ndar

y d

ista

nce,

h ft

10 in.

20 in.

30 in.

60 in.

90 in.

10 in.

20 in.

30 in.

60 in.

90 in.

90° Dip

3Dmodelwith

invasion

1Dmodel,

noinvasion

RtRxo2D

model

■■Shallow (24-in. diameter) invasion in ahorizontal well approaching an overlyingshale. Here the deep-reading 3D modeled60- and 90-in. curves (blue) both track theno-invasion 1D results (red) and approachRt as the tool moves farther away from theboundary, while the shallower 10-, 20-,and 30-in. curves are all shifted towardsRxo. In all cases, far from the bed bound-ary, the FDM results approach those of theno-bed-boundary, invasion-only 2D model(black points).

Resistivity, ohm-m

h varied

48 in.

Bed boundary

1-16

-12

-8

-4

0

4

8

12

16

20

24

28

32

10 100

Wel

lbor

e-to

-bou

ndar

y d

ista

nce,

h ft

90° Dip

RtRxo

10 in.

20 in.

30 in.

60 in.

90 in.

10 in.

20 in.

30 in.

60 in.

90 in.2D

model

3Dmodelwith

invasion

1Dmodel,

noinvasion

■■Deep (48-in. diameter) invasion in a hori-zontal well approaching an overlyingshale. In this deep-invasion geometry, onlythe deepest 3D modeled 90-in. curve (blue)tracks the no-invasion 1D result (red). The10-in. curve reads Rxo, and the 20-, 30-,and 60-in. intermediate curves are shiftedtoward Rxo. The 3D FDM calculation is car-ried up to, but not through, the bed bound-ary because of uncertainty about theshape of the invasion just below theboundary. In all cases, far from the bedboundary, the FDM results approach thoseof the no-bed-boundary, invasion-only 2Dmodel (black points).

Horizontal invasion, in.

Vert

ical

inva

sion

, in.

20

10

0

-10

-20

0° Dip

20

10

0

-10

-20

35° Dip

20

10

0

-10

-20

60° Dip

20

10

0

-10

-20

70° Dip

20

10

0

-10

-20

80° Dip

20

10

0

-10

-20-30 -20 -10 10 20 300

90° Dip

■■Anisotropic invasion profile. The effect ofinvasion at 1-week intervals is shown forsix borehole deviations from vertical tohorizontal in a formation with a perme-ability anisotropy khorizontal/kvertical of tento one. The invasion profile is character-ized by its aspect (long axis to short axis)ratio, which depends on permeabilityanisotropy, time and dip.

17. A cap shale is frequently called a shale-seal or sim-ply a seal, because it represents an impermeable

Horizontal well10 in.

20 in.

30 in.

60 in.

90 in.

10

1

1 2 3

ri at 1=16 in. ratio

Aspect (long/short axis) ratio

Elliptical Resistive Invasion - Rxo > Rt

Res

istiv

ity, o

hm-m

Rxo

Rt

Horizontal well10 in.

20 in.

30 in.

60 in.

90 in.

10

100

11 2 3

ri at 1=16 in. ratio

Aspect (long/short axis) ratio

Elliptical Conductive Invasion - Rxo < Rt

Res

istiv

ity, o

hm-m

Rxo

Rt

■■AIT log sensi-tivity to theinvasion profileaspect ratio forresistive andconductiveanisotropic inva-sion. With resis-tive invasion,the four deepest(20-, 30-, 60- and90-in.) curvesare fairly flat,while the shal-low 10-in. curvealone decreasestowards Rt asthe invasion axisratio increases.For conductiveinvasion, theshallow 10-in.curves read sys-tematically highand increasetoward Rt as theinvasion shapebecomes moreeccentric. Themedium-range20- and 30-in.curves alsoincreasetowards Rt in asimilar manner.

Invaded Horizontal Well Near a Cap Shale.Horizontal well logs in any environment aredifficult to analyze without 3D modeling. Atypical and interesting problem is the caseof invasion in a permeable sand below acap shale interface (previous page, top leftand center).17 In these examples, the log-ging sonde remains parallel to the bedboundary, while the distance between thedrainhole and the boundary is varied.

Spring 1997

The 3D FDM model is computed for shal-low (24-in. diameter) and deep (48-in.diameter) invasion depths and comparedwith the results of a 1D analytical modelwithout invasion. The 1D model shows thetypical curve order reversal and polariza-tion-induced horns on the inductionresponses at the bed boundary. Also, forcomparison, the no-bed-boundary, invasion-only limiting responses were computed witha numerical 2D model.

These results show that for shallow tomoderate invasion, the deepest curve canbe used to infer Rt and proximity to theshale cap, while the shallowest curve indi-cates Rxo. The relative separation betweenthe intermediate curves can be used withcaution to estimate depth of invasion. It isclearly still possible to get good results fromthe AIT induction in horizontal wells, eventhough the tool was designed for verticalwells with axially symmetric geometries.

Asymmetric Invasion. As a refinement onthe above model geometry, which assumesa cylindrical invasion front, it is interestingto examine the causes of nonuniform inva-sion, and then forward model the toolresponse in such environments. Nonuniformeffects can be caused by gravity segregationor permeability anisotropy. As a well devi-ates from vertical to horizontal, gravity cancause invading fluid to behave differently inthe top and bottom halves of the borehole.Permeability anisotropy causes an ellipticalinvasion shape in deviated wells. The inva-sion shape becomes more elongated relativeto the borehole as permeability anisotropy,well deviation and quantity of filtrate inva-sion increase (previous page, right).

The 3D FDM computes the effects of theshape of the invasion profile, after threeweeks of elliptical invasion, on AITresponses (left). The results show that volu-metric contribution of Rt is increasing as theshape of the equal area invasion frontbecomes more elongated. The sensitivity toRt becomes greater because a larger portionof the induction current circulates outsidethe invaded zone as the front elongates.

The modeled AIT responses directly reflectthe relatively shallow invasion depth in thevertical and horizontal portions of the well.The filtrate invades preferentially in the hori-zontal direction along the higher khorizontal,and less filtrate invades in the well’s verticaldirection. In the horizontal borehole, theformation fluid remains closer to the loggingtool above and below the drainhole, whichincreases the influence of Rt on the shallowresistivity curves.

51

barrier on top of a reservoir.

18. The invasion depths are strongly influenced by theformation fluid mobility (permeability/viscosity): highmobility results in shallow top and deep bottom inva-sion depths. See Anderson et al, 1996, reference 15.

19. Terry R, Barber T, Jacobson S and Henry K: “The Useof Modern Logging Measurements and New Process-ing Algorithms to Provide Improved Evaluation inDeeply Invaded Gas Sands,” Transactions of theSPWLA 35th Annual Logging Symposium, Tulsa,Oklahoma, USA, June 19-22, 1994, paper FFF.

10 in.

20 in.

30 in.

60 in.

90 in.

10

100

10 10 20 30 40 50 60 70 80

Initial ri = 6 in. Equal area ri = 12 in.

1D model

Lower depth of invasion, h in.

3D model

AIT Sensitivity to Gravity Segregation - Rxo > Rt

Res

istiv

ity, o

hm-m

Rxo

Rt

■■AIT log response to gravity segregation in a horizontal well. Becausethe invasion is pulled downward away from the drainhole, the sensi-tivity to Rxo is decreased compared to cases in which buoyancy is nota factor. This means that in isotropic formations with high mobility,gravity prevents the filtrate from penetrating the formation over a sub-stantial region around the horizontal drainhole (top and sides), thusresulting in a greater influence of Rt on all the AIT curves.

Rxo

h

5.3 in.

10.6 in.

Rt Rt

θ

Rxo

■■Gravity segrega-tion of invasion in ahorizontal well.Gravity sweeps thedenser water down-ward (left), reduc-ing the invasion onthe top side whileproducing a hydro-dynamically unsta-ble mixing of theinvading filtratewith movable oil onthe bottom side ofthe drainhole.Geometry for grav-ity segregation in ahorizontal well ismodeled as a box(right).

Buoyancy, caused by gravity, is anotherphenomenon that can greatly distort theshape of the invasion front (top). The waterfiltrate invading an oil zone in a horizontalwell behaves much differently on the bot-tom side than on the top side of the drain-hole. As a result, the resistivities of theinvaded zones at the top and bottom of thedrainhole are quite different.

On the top, Rxo is confined to a rathershallow invaded zone, but below the resis-tivity is a gradual transition—from Rxo toRt—over a relatively thick region.18 Model-ing the gravity-induced asymmetric invasionprofile in a horizontal well shows that theeffect on AIT responses is noticeable onlyon the most shallow curve—the 10-in.response (above).

52

Field Logs and 3D Modeling. Computermodeling predicts many 3D effects, butwhat do real logs look like? A field log fromthe Middle East in a horizontal well withmoderately salty mud invasion illustrates theneed for 3D modeling in interpreting induc-tion logs in deviated wells (next page, topright). Below 150 ft, the logs reveal signs ofthe horizontal wellbore approaching a shalebed from below—the deep 90-in. curvewith signs of horns and the 20- and 30-in.curves in reverse order read lower than theshallow 10-in. AIT and the MicroSFL logs.However, the low-reading 20-in. curvecould not be modeled by a single interfacewith the known shale resistivity of 4 ohm-m.

The low-reading 20-in. curve along withthe reversed curve order suggests an annulusinvasion profile.19 A short zone from 100 to250 ft was chosen for modeling. With poros-ity log data, a known formation water resis-tivity, Rw, and the tool response, a formationmodel with an annulus profile was gener-ated. From the earlier discussion on capshales and invasion, it is reasonable toassume that the shallow AIT logs (10-, 20-and 30-in. curves) respond to the invasionprofile, while the deeper logs (60- and 90-in.curves) respond to nearby bed boundaries.

In the formation model, invasion occurs inthe sands with only slight variations in thedepth of invasion to fit with variations inporosity and deep induction response (nextpage, left). The effect of the shale was to cutoff the upper part of the invasion. Thus asthe invasion and annulus profileapproached the cap shale, they are trun-cated at the bed boundary (next page, bot-tom right). As modeling progressed, itbecame clear that the behavior of the 90-in.log is influenced as much by invasion as it isby the bed boundaries. Only when the inva-sion radius was adjusted appropriately werethe excursions to 2000 ohm-m reproduced.

The qualitative agreement between thefield logs and modeled logs suggests that theannulus profile is real. The behavior of allthe logs is a complicated mixture of inva-sion effects and high-angle, bed-boundaryeffects. This model ignored any gravity orpermeability anisotropy effects, becausethese are negligible compared to the first-order effects of bed boundaries and inva-sion, and because the sensitivity of the AITlogs to these effects is not large.

All the unknowns surrounding invasion inhorizontal wells lead to the non-uniquenessof any solution that matches the field logs.Additional information, such as Rxo or shaleresistivity, Rshale, becomes essential to helplimit the uncertainty. This example showsthat, just as in vertical wells, it is inadequateto assume a simple invasion model in hori-zontal wells. Only modern multiarray toolsand full 3D modeling can successfully pro-vide quantitative interpretation of resistivitylogs in horizontal wells.

Oilfield Review

Spring 1997 53

1000

100

GR

10

150

0100 150 200

Depth, ft250 300 350

Res

istiv

ity, o

hm-m

GR

AP

I

MicroSFL

10 in.20 in.

30 in.

60 in.

90 in.

■■Field logs from a horizontal well underlying a cap shale. The GRresponse increases from 300 ft to 150 ft as the borehole approachesthe cap shale.

200 ohm-m

Shale4 ohm-m

Borehole Invasion20 ohm-m

Annulus2.2 ohm-m

Borehole onbed boundary

Borehole 6 in.belowbed boundary

Borehole 12 in.belowbed boundary

200 ohm-m

Shale4 ohm-m

200 ohm-m

Shale4 ohm-m

■■3D formation model with annulus profile geometry in ahorizontal well. Rt was 200 ohm-m. The borehole was mod-eled at the shale-formation interface (top) and at 6 in. (mid-dle) and 12 in. (bottom) below the interface. No invasionwas permitted into the impermeable shale.

1000

Res

istiv

ity, o

hm-m

TVD

wel

l pos

ition

, ft

100

10

1

0 30-2

0

2

4

6

60 90 120 150 ft

10 in.20 in.30 in.60 in.90 in.Rxo

Rann

Rt

Boreholer1

r2

Shale 4 ohm-m

Reservoir200 ohm-m

■■3D simulation results of AIT logs (top) in a horizontal well with capshale. The borehole trajectory and invasion profile relative to the capshale are shown (bottom). The model simulates the observed fieldlogs (top), and suggests that the borehole entered a steeply dippingcross-bedded formation below the cap shale, known to exist in thereservoir.

Rxo

Rxo

Rt

Rt

ri

Form

atio

n re

sist

ivity

pro

file

Distance from wellbore

Step profile

Invasionmidpoint

Rxo

Distance from wellbore

Distance from wellbore

Ramp profile

Rt

Slope profile

Annulus profile

Rt

Rann

Rxo

r1

r2

r1

r2

ri

■■Characteristics of typical modeled inva-sion profile geometries. The simple stepinvasion profile and the ramp profilerequire three parameters; the slope profileis a four-parameter model; the annulusprofile is a five-parameter model.

Resistivity, ohm-m

Bed boundary

Rh=Rv=10 ohm-m

Rv=20 ohm-mRh= 4 ohm-m

Borehole axis

-16

-12

-8

-4

0

4

8

12

16

10

True

ver

tical

dep

th, f

t

10 in.

20 in.

30 in.

60 in.

90 in.

10 in.

20 in.

30 in.

60 in.

90 in.

3Dmodel

1Dmodel

■■Comparison of AIT logs using 3D and 1Dmodels in an anisotropic dipping bed. Inthe lower, isotropic layer, the inductioncurves read Rt, but in the anisotropic,upper zone the tool reads neither Rh nor Rv,but a weighted average of the two.

Anisotropy. In highly deviated wells, induc-tion and propagation tools can detect resis-tivity anisotropy—largely invisible to thesetools in vertical wells—because they havebeen designed to measure currents or fieldsin planes normal to the tool axis. The originsof electrical anisotropy are linked with thesame phenomena that cause permeabilityanisotropy—formation bedding geometry orgrain size in homogeneous sand beds.20

Often thin, anisotropic, conductive shalebeds or laminations are mixed with high-resistivity pay zones.21

Although some new properties can beobtained from the use of both horizontaland vertical resistivity, it is horizontal resis-tivity, Rh, that is most desired in log interpre-tation.22 In vertical wells, inductive toolsread the geometric mean of the horizontalbedding resistivities. In deviated or horizon-tal wells, it is difficult to define a simplemixing law to represent the way the induc-tion tools read an average of the horizontaland vertical resistivity, Rv, which will bevery different from the average horizontalresistivity the tool reads in a vertical well.

54

A version of the 3D FDM model has beendeveloped that can account for any direc-tional variation in resistivity, which willaccount for any differences in vertical andhorizontal resistivities. A simple dipping bedexample, with no invasion, shows an induc-tion tool’s responses, logging at a dip of 45°as it moves from an isotropic bed into aanisotropic bed (left). The induction curves all are in good agreement with Rt deep in theisotropic bed, but as the tool moves into theanisotropic bed, all the curves tend towards aweighted average of Rh and Rv. In a verticalwell, the log curves would all read Rh.

To investigate how anisotropy further com-plicates the already complex interpretationof invasion in a horizontal well, anisotropywas added to the previous example of ahorizontal well near a cap shale with inva-sion. The invaded zone was assumedisotropic; the overall responses are similar tothe previous case—polarization hornsappear at the bed boundary, and in theinvaded zone shallow tool response curvesread close to Rxo and the induction logs sep-arate—but there is a difference in the deeptool responses. In the anisotropic bed, thedeep tool responses tend towards an aver-age of Rv and Rh, instead of reading close toRh. Without the model, it would be impossi-ble to determine the presence of anisotropyor accurate formation resitivities based onthe behavior of the resistivity curves alone.

Knowledge of vertical and horizontal resis-tivities is important for analyzing thinly lam-inated dipping formations, where both resis-tivity values are crucial for estimating thesand lamina resistivity and the net-to-grossratio (the sand fraction) simultaneously.23

Armed with the 3D modeling capability andknowledge of the relative dip, the formationresistivity anisotropy can be determined. Ofcourse, the inverse is also true, if the forma-tion resistivity anisotropy is known, then therelative dip or deviation angle can bederived from the induction tool responses.

Oilfield Review

20. For a detailed introduction to anisotropy: AndersonB, Bryant I, Lüling M, Spies B and Helbig K: “OilfieldAnisotropy: Its Origins and Electrical Characteris-tics,” Oilfield Review 6, no. 4 (October 1994): 48-56.

21. Klein JD: “Induction Log Anisotropy Correction,”Transactions of the SPWLA 32nd Annual LoggingSymposium, Midland, Texas, USA, June 16-19,1991, paper T.

22. Hagiwara T: “A New Method to Determine Horizon-tal-Resistivity in Anisotropic Formations Without PriorKnowledge of Relative Dip,” Transactions of theSPWLA 37th Annual Logging Symposium, NewOrleans, Louisiana, USA, June 16-19, 1996, paper Q.

23. Hagiwara T: “Macroscopic Anisotropy Approach toE-Log Evaluation in Laminated Sand-ShaleSequences,” presented at the 3rd SPWLA ReservoirCharacterization Archie Conference, Galveston,Texas, USA, November 8-11, 1992.

24. Lüling MG, Rosthal RA and Shray F: “Processing andModeling 2-MHz resistivity Tools in Dipping, Lami-

1 : 600

950

1000

1050

1100

1150

1200

Virgin Formation

Invaded Zone

Borehole

AIT 30 in. Quality Curve

RTohm.m

0.2 2000 -5 5

0.2 2000Resolution Matching

Formation Evaluation

Resistivity Analysis

Forward Modeling

Invasion Analysis

■■First look. Deconvolution produces beds and resistivities for the initial formation model.The display shows the borehole and radii of invasion on the left track. AIT and MicroSFLfield logs along with simulated logs are shown in the middle track. Rxo, and Rt are alsoshown in this track. The right track contains log quality curves.

The response of propagation tools toanisotropy is even more pronounced thanthe response of induction tools. At high dip,the vertical component of resistivity for bothtools is multiplied by a term proportional tofrequency. This term is an order of magni-tude greater for 2-MHz propagation toolsthan for induction tools. The anisotropicresponse also varies greatly with the trans-mitter-receiver spacing. The highly nonlineartransforms from phase shift to resistivity andattenuation to resistivity behave differently,amplifying the anisotropic effects when thetwo logs are compared.24

Resistivity Modeling as a Log Analyst’s ToolModeling results in the dipping bed andhorizontal well examples indicate that resis-tivity logs in complex formations containgeometrical information, but extracting it isa challenge. At the very least, the use of for-ward modeling has now become a key toolfor log analysts in understanding the forma-tion properties that combine to produce thelogging tool responses.

A new workstation program is currentlybeing developed for GeoQuest, calledINVASION, to provide log analysts with thetools for invasion-based, resistivity-modelingformation analysis. The programs are basedon forward modeling for layered formations,and the interpretation allows for evaluatingdynamic reservoir properties—early-timepermeability, water cut and fractional fluidflow. The system helps take the drudgery outof the most time-consuming activities associ-ated with modeling. This is done with agraphical interface and interactive parameterselection, which promotes a more accurateRt evaluation from multiple resistivity mea-surements (above right).

Spring 1997

nated, Anisotropic Formations,” Transactions of theSPWLA 35th Annual Logging Symposium, Tulsa,Oklahoma, USA, June 19-22, 1994, paper QQ.

25. Barber TA and Rosthal R: “Using Multiarray Induc-tion Tool to Achieve High-Resolution Logs with Min-imum Environmental Effects,” paper SPE 22725, pre-sented at the 66th SPE Annual Technical Conferenceand Exhibition, Dallas, Texas, USA, October 6-9,1991.

26. Generalized (finite conductivity) 1D tool geometricalresponse functions are derived using a solution simi-lar to the single scattering Born approximation for-malism traditionally used in quantum mechanics.See Gianzero S and Anderson B: “A New Look at

The AIT induction logs are corrected forenvironmental effects either at the wellsiteor by a preprocessing program called PrePlus,which corrects for apparent dip and effectsof shoulders. The resulting logs are verticallymatched in resolution.25 The logs are thenprocessed by a resistivity iterative inversionprogram, using 1D radial tool response func-tions as a forward model.26 The combinationof separate vertical processing and radialprocessing is called 1D+1D processing. Thisstep, along with log squaring, gives the ana-lyst a first approximation of the formationbeds and resistivities. At this point, using agraphical log display, the log analyst canreview the initial formation model andmake refinements to the model with aninteractive interface.

This interactive task permits the analyst tomanually define invasion resistivity profilesand the formation bedding and resistivitydescription through a tabular interface (pre-vious page, right). With this interactive task,the analyst can explore the sensitivity of logdata to changes in the formation model.

55

Skin Effect,” The Log Analyst 23, no. 1 (January-February, 1982): 20-34.Also see Thadani SG and Hall HE Jr: “PropagatedGeometrical Factors in Induction Logging,” Transac-tions of the SPWLA 22nd Annual Logging Symposium,Mexico City, Mexico, June 23-26, 1981, paper WW.

950

1000

Resolution Matching

Formation Evaluation

Resistivity Analysis

Forward Modeling

Invasion Analysis

Quality WaterCut

SOR

SW AIT 10 in. AIT 20 in.ohm.m

10 10 0

0.04

0.04

400

400

0.04

0.04

400

400 0

Fractional Flow

2001

10 ohm.m

■■Interpreting invasion. Model-based fractional flow inversion logs enable reservoirengineers to predict water cut throughout the reservoir. With porosity and the fractionalflow computed from the invasion profile, a water-cut log is computed that helps deter-mine which part of the reservoir to perforate. In many cases, it may be necessary tominimize water cut, and in other cases a higher water cut may be acceptable for opti-mum oil production.

Sensitivity analysis is especially importantwhen Rxo and Rt are dissimilar. With thesame task, the user can change Rt, and inva-sion profiles (invasion and annulus radii andresistivity Rxo, and Rann) and quickly recom-pute, using forward modeling, the syntheticlog for verification with original logs. For lat-erolog responses, the forward model usesthe 3D FEM for dipping beds and a 2D FEMotherwise. For induction tools, the forwardmodel is based on 2D hybrid FEM, and a1D analytical code is used for dipping beds.Formation models can include invasiongeometry as either a step function, ramped,annulus profile or no invasion.

During the next phase of interpretation,after the analyst is satisfied that the final for-mation model accurately represents the log-ging environment, there are two new pre-liminary functions to compute invaded zonefluid properties.27 First, with invasionparameters, LWD and wireline resistivitiesand porosity logs, the volume of mud filtratethat invaded the formation around the bore-hole can be computed. This allows a time-lapse permeability analysis. Also, if the wellis vertical in a clean or shaly sand and hasbeds thicker than 6 ft [1.8 m], the analystcan compute a formation water-cut log—useful for reservoir engineering, and a frac-tional-flow log, both based on the fluiddynamics reflected in the formation invasionprofile (above).

This graphically interactive, resistivitymodeling-based formation evaluation program facilitates interpretations in com-plex formations.

56

—RCH

27. Ramakrishnan TS, Al-Khalifa J and Al-Waheed HH:“Producibility Estimation from Array-Induction Logsand Comparison with Measurements—A CaseStudy,” to be presented at the SPWLA 38th AnnualLogging Symposium, Houston, Texas, USA, June 15-18, 1997.

Outlook for Resistivity ModelingThe recent developments in code efficiencywill lead to full use of 3D models for inter-pretation applications. Induction, propaga-tion, laterolog and eventually other loggingtools, such as nuclear and acoustic, will bemodeled in more realistic formations. Petro-physical relationships will be incorporatedin the models, which means the formationwill be described in log analysis terms:lithology, porosity and saturation.

Trends started with AIT and PLATFORM

EXPRESS equipment will continue—sophisti-cated tool environmental corrections willbe built into the logs using results from

detailed physical and numerical modelingduring the tool design and developmentphases. The power of analysts’ tools basedon resistivity modeling, such as in theRtBAN, INFORM and INVASION programs,will be brought to bear on petrophysicalinterpretation in complicated environments.Many consumers of well log data believethat to gain maximum value from resistivitylogs in general, and induction logs in par-ticular, the numerical modeling of toolresponses will be an indispensable facet ofinterpretation for the 21st century.

Oilfield Review