Water Control

Oilfield Review30

Water Control

Bill BaileyMike CrabtreeJeb TyrieAberdeen, Scotland

Jon ElphickCambridge, England

Fikri KuchukDubai, United Arab Emirates

Christian RomanoCaracas, Venezuela

Leo RoodhartShell International Exploration and ProductionThe Hague, The Netherlands

Today, oil companies produce an average of three barrels of

water for each barrel of oil from their depleting reservoirs.

Every year more than $40 billion is spent dealing with unwanted

water. In many cases, innovative water-control technology can

lead to significant cost reduction and improved oil production.

For help in preparation of this article, thanks to AndrewAcock, Houston, Texas, USA; Kate Bell and AnchalaRamasamy, BP Amoco Exploration, Aberdeen, Scotland;Leo Burdylo, Keng Seng Chang and Peter Hegeman, SugarLand, Texas; Alison Goligher, Montrouge, France; DouglasHupp, Anchorage, Alaska, USA; Lisa Silipigno, OklahomaCity, Oklahoma, USA; and David Wylie, Aberdeen. FloView, FrontSim, GHOST (Gas Holdup Optical SensorTool), MDT (Modular Formation Dynamics Tester), NODAL,PatchFlex, PLT (Production Logging Tool), PosiSet, PSPLATFORM (Production Services Platform), RST (ReservoirSaturation Tool), SqueezeCRETE, TPHL (three-phase fluidholdup log), USI (UltraSonic Imager) and WFL (Water Flow Log) are marks of Schlumberger. Excel is a mark ofMicrosoft Corporation. MaraSEAL is a mark of Marathon OilCorporation. PrecisionTree is a mark of Palisade Corporation.

Oil and water

Water

Free-waterlevel

Oil, gas and water

Gas and water

Free-oillevel

Reservoir containing water, oil and gas.The figure shows the fluid distribution in a typical reservoir before production orinjection begins. Above the free-oil level,water saturation will be at its irreduciblevalue. The transition zone between thefree-oil and free-water levels is character-ized by a gradual increase in water satura-tion to 100%. In this zone, both oil andwater are partially mobile. The thickness ofthe transition zone depends on factorssuch as pore size, capillary pressure andwettability. There is a transition zonebetween the hydrocarbon and water layers where water and oil saturation vary.In general, low-permeability rocks willhave thicker transition zones.

Spring 2000 31

Given the worldwide daily water production ofroughly 210 million barrels [33.4 million m3] ofwater accompanying every 75 million barrels[11.9 million m3] of oil, many oil companies couldalmost be called water companies. Water-handling costs are high—estimates range from 5 to more than 50 cents per barrel of water. In a well producing oil with an 80% water cut, thecost of handling water can be as high as $4 perbarrel of oil produced. In some parts of the NorthSea, water production is increasing as fast asreservoir oil rates are declining.

Water affects every stage of oilfield life fromexploration—the oil-water contact is a crucial fac-tor for determining oil-in-place—through develop-ment, production, and finally to abandonment(below). As oil is produced from a reservoir, waterfrom an underlying aquifer or from injectors even-tually will be mixed and produced along with the

oil. This movement of water flowing through a reservoir, into production tubing and surface-processing facilities, and eventually extracted fordisposal or injected for maintaining reservoir pres-sure, is called the ‘water cycle’ (above).

Oil producers are looking for economic waysto improve production efficiency, and water-con-trol services are proving to be one of the fastestand least costly routes to reduce operating costsand improve hydrocarbon production simultane-ously. The economics of water productionthroughout the water cycle depend on a numberof factors such as total flow rate, productionrates, fluid properties like oil gravity and watersalinity, and finally the ultimate disposal methodfor the water produced. Operational expenses,

Oila

ndw

aterDry o

il

W

ater

ProcessingDemulsifiers/corrosionFacility debottlenecking

TreatingCleaningDischarge

Water shutoffScale and hydrate controlCorrosion inhibitor

Profile modificationWater diversionFluid monitoringGel treatmentsPermeability modifiersDamage removal

> The water cycle. The transport of water through the field startswith flow in the reservoir leading to production, and then surfaceprocessing. Finally, the water isdisposed of at the surface orinjected for disposal or pressuremaintenance.

including lifting, separation, filtering, pumpingand reinjection, add to the overall costs (below).In addition, water-disposal costs can vary enor-mously. Reports vary from 10 cents per barrelwhen the unwanted water is released into theocean offshore to over $1.50 per barrel whenhauled away by trucks on land. Although thepotential savings from water control alone aresignificant, the greatest value comes from thepotential increase in oil production and recovery.

Managing the cycle of water production, sep-aration downhole or at the surface, and disposalinvolves a wide range of oilfield services. Theseinclude data acquisition and diagnostics usingdownhole sensors; production logging and wateranalysis for detecting water problems; reservoirmodeling to characterize flow; and various tech-nologies to eliminate water problems such asdownhole separation and injection, chemical andmechanical shutoff, and surface water separa-tion and production facilities.

In this article, we focus on the detection andcontrol of excess water production. First, wereview the many ways in which water can enterthe wellbore. Then, we describe measurementsand analysis to identify these problem types.Finally, we examine treatments and solutions.Case studies demonstrate applications in individ-ual wells, on a field scale and in surface facilities.

32 Oilfield Review

Lifting

Separation

De-oiling

Filtering

Pumping

Injecting

Capex/OpexUtilitiesCapex/OpexUtilitiesChemicalCapex/OpexChemicalsCapex/OpexUtilitiesCapex/OpexUtilitiesCapex/OpexTotal cost/bblTotal chemicalsTotal utilitiesTotal wellsSurface facilities

$0.044$0.050$0.087$0.002$0.034$0.147$0.040$0.147$0.012$0.207$0.033$0.030$0.842$0.074$0.102$0.074$0.589

5.28%6.38%

10.36%0.30%4.09%

17.56%4.81%

17.47%1.48%

24.66%3.99%3.62%100%

8.90%12.16%

8.89%70.05%

$0.044$0.054$0.046$0.003$0.034$0.073$0.041$0.068$0.010$0.122$0.034$0.030$0.559$0.075$0.010$0.075$0.309

7.95%9.62%8.27%0.45%6.16%

12.99%7.25%

12.18%1.79%

21.89%6.01%5.45%100%

13.41%17.87%13.40%55.33%

$0.044$0.054$0.035$0.003$0.034$0.056$0.041$0.047$0.010$0.091

$0..034$0.030$0.478$0.075$0.100$0.075$0.227

9.29%11.24%

7.24%0.52%7.20%

11.64%8.47%9.85%2.09%

19.06%7.03%6.37%100%

15.67%20.88%15.66%47.80%

$0.044$0.054$0.030$0.003$0.034$0.046$0.041$0.030$0.010$0.079$0.034$0.030$0.434$0.075$0.100$0.075$0.184

10.25%12.40%

6.82%0.58%7.94%

10.58%9.34%6.87%2.31%

18.15%7.75%7.02%100%

17.28%23.03%17.27%42.41%

$0.044$0.054$0.049$0.003$0.034$0.081$0.041$0.073$0.011$0.125$0.034$0.030$0.578$0.075$0.101$0.075$0.328

7.69%9.30%8.55%0.43%5.95%

13.92%7.00%

12.63%1.84%

21.61%5.81%5.27%100%

12.96%17.38%12.95%56.71%

20,000 B/D 50,000 B/D 100,000 B/D 200,000 B/D Average

Surface processing Wells, producers Wells, injectors

SeparationLiftingInjectionCost

1 Well 7000 ftRecompletionTotal 1 wellCost for waterTotal productionTotal waterCost for water lift

0.00251.92

1.2$0.028

kw/bblkw/bblkw/bblPer kw-hr

$1,000,000.00300,000

$1,600,000.00$400,000.00

1,000,0009,000,000

$0.04

Drill and completePer completion3 Completions

bbl @ 90% water cutbbl @ 90% water cut$/bbl

1 Well 7000 ftRecompletionTotal 1 wellTotal injectedCost for water injection

$600,000.00200,000

$1,000,000.0032,850,000

$0.03

Drill and completePer completion3 Completions3 Completions$/bbl

>Water-cycle cost. The table shows typical estimated water-handling costs per barrel—capital and operating expenses (Capex and Opex), utilities andchemicals—lifting, separation, de-oiling, filtering, pumping and injection for fluid production varying from 20,000 to 200,000 B/D [3181 to 31,810 m3/d].

1.0

0

WOR

WOR economic limit

Added recovery

Oil production, bbl

A

B

C

D

>Water control to increase well productivity and potential reserves. As most wellsmature, the water/oil ratio (WOR) increases with production (A) due to increasingamounts of water. Eventually, the cost of handling the water approaches the value ofoil being produced and the WOR “economic limit” (B). Water-control methodologyand technology reduce the well’s water production (C) enabling continued economicoil production. Water control results in increased economic recovery in the well (D).

Spring 2000 33

Water SourcesWater is present in every oil field and is the mostabundant fluid in the field.1 No operator wants toproduce water, but some waters are better thanothers. When it comes to producing oil, a keyissue is the distinction between sweep, good (oracceptable), and bad (or excess) water.

“Sweep” water—Sweep water comes fromeither an injection well or an active aquifer thatis contributing to the sweeping of oil from thereservoir. The management of this water is a vital part of reservoir management and can be a determining factor in well productivity and theultimate reserves.2

“Good” water—This is water that is producedinto the wellbore at a rate below the water/oilratio (WOR) economic limit (previous page, top).3 Itis an inevitable consequence of water flowthrough the reservoir, and it cannot be shut offwithout losing reserves. Good-water productionoccurs when the flow of oil and water is commin-gled through the formation matrix. The fractionalwater flow is dictated by the natural mixing behav-ior that gradually increases the WOR (top right).

Another form of acceptable water productionis caused by converging flow lines into the well-bore (middle right). For example, in one quadrantof a five-spot injection pattern, an injector feedsa producer. Flow from the injector can be charac-terized by an infinite series of flowlines—theshortest is a straight line from injector to pro-ducer and the longest follows the no-flow bound-aries from injector to producer. Waterbreakthrough occurs initially along the shortestflowline, while oil is still produced along slowerflowlines. This water must be considered goodsince it is not possible to shut off selected flow-lines while allowing others to produce.

Since good water, by definition, produces oilwith it, water management should seek to maxi-mize its production. To minimize associatedwater costs, the water should be removed asearly as possible, ideally with a downhole sepa-rator (bottom right). These devices, coupled withelectrical submersible pumps, allow up to 50% ofthe water to be separated and injected downholeto avoid lifting and surface-separation costs.

1. Kuchuk F, Sengul M and Zeybek M: “Oilfield Water: A Vital Resource,” Middle East Well Evaluation Review22 (November 22, 1999): 4-13.

2. Kuchuk F, Patra SK, Narasimham JL, Ramanan S andBanerji S: “Water Watching,” Middle East WellEvaluation Review 22 (November 22, 1999): 14-23; andalso Kuchuk F and Sengul M: “The Challenge of WaterControl,” Middle East Well Evaluation Review 22(November 22, 1999): 24-43.

Injector

Incr

easi

ng ti

me

Water front Producer

Oil and water

Only water

Only Oil

Only Oil

> Good and bad water. Good water needs to be produced with oil. It cannotbe shut off without shutting off oil. Downhole separation may be a solution.Bad water does not help production, and it depletes pressure.

Injector

Producer

Wat

er

Oil

Oil

Water

Simulating water flow in a reservoir.FrontSim streamline reservoir simulationsoftware is ideal for demonstrating whathappens to fluids flowing in a reservoir.The streamlines represent the flow ofwater from injector to producer. The simulator requires geological, structuraland fluid information. The plot shows onequadrant of a uniform five-spot injectionpattern where the water from the mostdirect streamline is the first to breakthrough to the producer. The water fromthese streamlines is considered goodwater because it cannot be shut off without decreasing oil production.

Productionzone

Injectionzone

Oil

Water

Reservoirfluid in

Oil and somewater out

Water out

Downhole separator.Separating water down-hole reduces the costs oflifting the excess water.Typical downhole separa-tors are 50% efficient. Theexcess water is injectedinto another formation.

3. Water/oil ratio (WOR) is the water production ratedivided by oil production rate. It ranges from 0 (100% oil)to infinite (100% water). Also commonly used are theterms ‘water cut’ or ‘fractional water flow’ defined aswater production rate divided by total production rate asa percentage or fraction, respectively. Correspondencebetween these measures can be easily calculated

(for example, a WOR of 1 implies a water cut of 50%). The WOR economic limit is the WOR at which the cost of the water treatment and disposal is equal to the profitfrom the oil. Production beyond this limit gives a negativecash flow. This can be approximated by the net profitfrom producing an incremental unit volume of oil dividedby the cost of an incremental unit volume of water.

“Bad” water—The remainder of this articledeals principally with the problems of excesswater. Bad water can be defined as water that isproduced into the wellbore and produces no oil orinsufficient oil to pay for the cost of handling thewater—water that is produced above the WOReconomic limit. In individual wells, the source ofmost bad-water problems can be classified as oneof ten basic types. The classification of waterproblem types presented here is simplistic—manyvariations and combinations can occur—but it isuseful for providing a common terminology.4

Water ProblemsThe ten basic problem types vary from easy tosolve to the most difficult to solve.

Casing, tubing or packer leaks—Leaks throughcasing, tubing or packers allow water from non-oil-productive zones to enter the production string(below left). Detection of problems and applicationof solutions are highly dependent on the well con-figuration. Basic production logs such as fluid den-sity, temperature and spinner may be sufficient todiagnose these problems. In more complex wells,WFL Water Flow Logs or multiphase fluid loggingsuch as the TPHL three-phase fluid holdup log canbe valuable. Tools with electrical probes, such asthe FlowView tool, can identify small amounts ofwater in the production flow. Solutions typicallyinclude squeezing shutoff fluids and mechanicalshutoff using plugs, cement and packers. Patchescan also be used. This problem type is a primecandidate for low-cost, inside-casing water shut-off technology.

Channel flow behind casing—Failed primarycementing can connect water-bearing zones to thepay zone (below middle). These channels allowwater to flow behind casing in the annulus. A sec-ondary cause is the creation of a ‘void’ behind thecasing as sand is produced. Temperature logs oroxygen-activation-based WFL logs can detect this

water flow. The main solution is the use of shutofffluids, which may be either high-strength squeezecement, resin-based fluids placed in the annulus,or lower strength gel-based fluids placed in the for-mation to stop flow into the annulus. Placement iscritical and typically is achieved with coiled tubing.

Moving oil-water contact—A uniform oil-water contact moving up into a perforated zonein a well during normal water-driven productioncan lead to unwanted water production (belowright). This happens wherever there is very lowvertical permeability. Since the flow area is largeand the rate at which the contact rises is low, itcan even occur at extremely low intrinsic verticalpermeabilities (less than 0.01 mD). In wells withhigher vertical permeability (Kv > 0.01 Kh), coningand other problems discussed below are morelikely. In fact, this problem type could be consid-ered a subset of coning, but the coning tendencyis so low that near-wellbore shutoff is effective.Diagnosis cannot be based solely on known entryof water at the bottom of the well, since otherproblems also cause this behavior. In a verticalwell, this problem can be solved easily by aban-doning the well from the bottom using a mechan-ical system such as a cement plug or bridge plugset on wireline. Retreatment is required if the

OWC moves significantly past the top of the plug.In vertical wells, this problem is the first in ourclassification system that extends beyond thelocal wellbore environment.

In horizontal wells, any wellbore or near-wellbore solution must extend far enough upholeor downhole from the water-producing interval tominimize horizontal flow of water past the treat-ment and delay subsequent water breakthrough.Alternatively, a sidetrack can be considered oncethe WOR becomes economically intolerable.5

Watered-out layer without crossflow—Acommon problem with multilayer productionoccurs when a high-permeability zone with aflow barrier (such as a shale bed) above andbelow is watered out (above). In this case, thewater source may be from an active aquifer or awaterflood injection well. The watered-out layertypically has the highest permeability. In theabsence of reservoir crossflow, this problem iseasily solved by the application of rigid, shutofffluids or mechanical shutoff in either the injectoror producer. Choosing between placement of ashutoff fluid—typically using coiled tubing—or amechanical shutoff system depends on knowingwhich interval is watered out. Effective selectivefluids, discussed later, can be used in this case toavoid the cost of logging and selective place-ment. The absence of crossflow is dependent onthe continuity of the permeability barrier.

Horizontal wells that are completed in justone layer are not subject to this type of problem.Water problems in highly inclined wells com-pleted in multiple layers can be treated in thesame way as vertical wells.

34 Oilfield Review

Injector Producer

> Casing, tubing orpacker leaks.

> Flow behind casing. > Moving oil-water contact.

>Watered-out layer without crossflow.

Spring 2000 35

Fractures or faults between injector andproducer—In naturally fractured formationsunder waterflood, injection water can rapidlybreak through into producing wells (above). Thisis especially common when the fracture systemis extensive or fissured and can be confirmedwith the use of interwell tracers and pressuretransient testing.6 Tracer logs also can be used toquantify the fracture volume, which is used forthe treatment design. The injection of a flowinggel at the injector can reduce water productionwithout adversely affecting oil production fromthe formation. When crosslinked flowing gels areused, they can be bullheaded since they havelimited penetration in the matrix and so selec-tively flow in the fractures. Water shutoff is usu-ally the best solution for this problem.

Wells with severe fractures or faults oftenexhibit extreme loss of drilling fluids. If a conduc-tive fault and associated fractures are expectedduring drilling, pumping flowing gel into the wellmay help solve both the drilling problem and thesubsequent water production and poor sweepproblems—particularly in formations with lowmatrix permeability.

In horizontal wells, the same problem canexist when the well intersects one or more faultsthat are conductive or have associated conduc-tive fractures.

faults or fractures that intersect an aquifer(above right). As discussed above, pumping flow-ing gel may help address this problem.

Coning or cusping—Coning occurs in a verti-cal well when there is an OWC near perforationsin a formation with a relatively high vertical per-meability (below). The maximum rate at which oilcan be produced without producing waterthrough a cone, called the critical coning rate, isoften too low to be economic. One approach,which is sometimes inappropriately proposed, isto place a layer of gel above the equilibriumOWC. However, this will rarely stop coning andrequires a large volume of gel to significantlyreduce the WOR. For example, to double the crit-ical coning rate, an effective gel radius of at least50 feet [15 m] typically is required. However, eco-nomically placing gel this deep into the formationis difficult. Smaller volume treatments usuallyresult in rapid water re-breakthrough unless thegel fortuitously connects with shale streaks.

A good alternative to gel placement is to drillone or more lateral drainholes near the top of theformation to take advantage of the greater dis-tance from the OWC and decreased drawdown,both of which reduce the coning effect.

In horizontal wells, this problem may bereferred to as duning or cusping. In such wells, itmay be possible to at least retard cusping withnear-wellbore shutoff that extends sufficientlyup- and downhole as in the case of a rising OWC.

Fractures or faults from a water layer—Water can be produced from fractures that inter-sect a deeper water zone (above middle). Thesefractures may be treated with a flowing gel; thisis particularly successful where the fractures donot contribute to oil production. Treatment vol-umes must be large enough to shut off the frac-tures far away from the well.

However, the design engineer is faced withthree difficulties. First, the treatment volume isdifficult to determine because the fracture volumeis unknown. Second, the treatment may shut offoil-producing fractures; here, an overflush treat-ment maintains productivity near the wellbore.Third, if a flowing gel is used, it must be carefullytailored to resist flowback after the treatment. Incases of localized fractures, it may be appropriateto shut them off near the wellbore, especially ifthe well is cased and cemented. Similarly, adegradation in production is caused whenhydraulic fractures penetrate a water layer.However, in such cases the problem and environ-ment are usually better understood and solutions,such as shutoff fluids, are easier to apply.

In many carbonate reservoirs, the fracturesare generally steep and tend to occur in clustersthat are spaced at large distances from eachother—especially in tight dolomitic zones. Thus,the probability of these fractures intersecting avertical wellbore is low. However, these fracturesare often observed in horizontal wells wherewater production is often through conductive

4. Elphick J and Seright R: “A Classification of WaterProblem Types,” presented at the Petroleum NetworkEducation Conference’s 3rd Annual InternationalConference on Reservoir Conformance ProfileModification, Water and Gas Shutoff, Houston, Texas,USA, August 6-8, 1997.

5. Hill D, Neme E, Ehlig-Economides C and Mollinedo M:“Reentry Drilling Gives New Life to Aging Fields,” Oilfield Review 8, no. 3 (Autumn 1996): 4-17.

6. A fissure is an extensive crack, break or fracture in a rock.

Injector

Producer

Fault

Fault

> Fractures or faults between an injector anda producer.

> Fractures or faults from a water layer (vertical well).

> Fractures or faults from a water layer(horizontal well).

> Coning or cusping.

Poor areal sweep—Edge water from anaquifer or injection during waterflooding through apay zone often leads to poor areal sweep (right).Areal permeability anisotropy typically causes thisproblem, which is particularly severe in sand chan-nel deposits. The solution is to divert injectedwater away from the pore space, which hasalready been swept by water. This requires a largetreatment volume or continuous viscous flood,both of which are generally uneconomic. Infilldrilling is often successful in improving recovery inthis situation, although lateral drainholes may beused to access unswept oil more economically.

Horizontal wells may extend through differentpermeability and pressure zones within the samelayer, causing poor areal sweep. Alternatively,water may break through to one part of the wellsimply because of horizontal proximity to thewater source. In either case, it may be possibleto control water by near-wellbore shutoff suffi-ciently up- and downhole from the water.

Gravity-segregated layer—In a thick reservoirlayer with good vertical permeability, gravity seg-regation—sometimes called water under-run—can result in unwanted water entry into aproducing well (below). The water, either from anaquifer or waterflood, slumps downward in thepermeable formation and sweeps only the lowerpart of the reservoir. An unfavorable oil-watermobility ratio can make the problem worse. Theproblem is further exacerbated in formations withsedimentary textures that become finer upward,since viscous effects along with gravity segrega-tion encourage flow at the bottom of the formation.Any treatment in the injector aimed at shutting offthe lower perforations has only a marginal effect in

sweeping more oil before gravity segregationagain dominates. At the producer there is localconing and, just as for the coning case describedearlier, gel treatments are unlikely to provide last-ing results. Lateral drainholes may be effective inaccessing the unswept oil. Foamed viscous-floodfluids may also improve the vertical sweep.

In horizontal wells, gravity segregation canoccur when the wellbore is placed near the bot-tom of the pay zone, or when the local criticalconing rate is exceeded.

Watered-out layer with crossflow—Watercrossflow can occur in high-permeability layersthat are not isolated by impermeable barriers(below right). Water production through a highlypermeable layer with crossflow is similar to theproblem of a watered-out layer without crossflow,but differs in that there is no barrier to stop cross-flow in the reservoir. In these cases, attempts tomodify either the production or injection profilenear the wellbore are doomed to be short-lived

36 Oilfield Review

because of crossflow away from the wellbore. It isvital to determine if there is crossflow in the reser-voir since this alone distinguishes between thetwo problems. When the problem occurs withoutcrossflow, it can be easily treated. With crossflow,successful treatment is less likely. However, inrare cases, it may be possible to place deep-pene-trating gel economically in the permeable thieflayer if the thief layer is thin and has high perme-ability compared with the oil zone. Even underthese optimal conditions, careful engineering isrequired before committing to a treatment. Inmany cases, a solution is to drill one or more lat-eral drainholes to access the undrained layers.

Horizontal wells completed in just one layerare not subject to this type of problem. If a highlyinclined well is completed in multiple layers,then this problem occurs in the same way as in avertical well.

Knowing the specific water-control problem isessential to treating it. The first four problemsare relatively easily controlled in or near thewellbore. The next two problems—fracturesbetween injectors and producers, or fracturesfrom a water layer—require placement of deeperpenetrating gels into the fractures or faults. Thelast four problems do not lend themselves to sim-ple and inexpensive near-wellbore solutions, andrequire completion or production changes as partof the reservoir management strategy. Any oper-ator wishing to achieve effective, low-risk, rapidpayout water shutoff should initially concentrateon applying proven technology to the first sixproblem types.

Injector Producer Injector Producer

> Poor areal sweep.

> Gravity-segregated layer. >Watered-out layer with crossflow.

Aqui

fer

Cumulative oil, bbl

WOR economic limit

Log

WOR

> Recovery plot. The recovery plot shows the increasing trend in water/oilratio with production. If the extrapolated WOR reaches the economic limitwhen the cumulative oil produced reaches the expected recoverablereserves, then the water being produced is considered good water.

> Production history plot. A time, days plot of the water and oil flowrates against time can be helpful in identifying water problems. Anysudden simultaneous change indicating increased water with areduction in oil is a signal that remediation might be needed.

Spring 2000 37

Well Diagnostics for Water Control In the past, water control was thought of as sim-ply a plug and cement operation, or a gel treat-ment in a well. The main reason for the industry’sfailure to consistently control water has been alack of understanding of the different problemsand the consequent application of inappropriatesolutions. This is demonstrated by the number oftechnical papers discussing the treatments andresults with little or no reference to the geology,reservoir or water-control problem. The key towater control is diagnostics—to identify the spe-cific water problem at hand. Well diagnostics areused in three ways:• to screen wells that are suitable candidates for

water control• to determine the water problem so that a suit-

able water-control method can be selected• to locate the water entry point in the well so

that a treatment can be correctly placed.When a reliable production history is avail-

able, it often contains a wealth of informationthat can help diagnose water problems. Severaldifferent analytical techniques using information,such as water/oil ratios, production data and log-ging measurements, have been developed to dis-tinguish between the different sources ofunacceptable water.

Recovery plot—The recovery plot is asemilog plot of WOR against cumulative oil pro-duction (above). The production trend can be ex-trapolated to the WOR economic limit todetermine the oil production that will beachieved if no water-control action is taken. If theextrapolated production is approximately equalto the expected reserves for a well, then the wellis producing acceptable water, and no water con-trol is needed. If this value is much less than theexpected recoverable reserves, the well is pro-ducing unacceptable water and remedial actionshould be considered if there are sufficientreserves to pay for intervention.

Production history plot—This plot is a log-logplot of oil and water rates against time (belowleft). Good candidates for water control usuallyshow an increase in water production and adecrease in oil production starting at about thesame time.

Decline-curve analysis—This is a semilogplot of oil production rate versus cumulative oil(below). A straight-line curve can be expected fornormal depletion. An increased decline may indi-cate a problem other than water, such as severepressure depletion or damage buildup.

1000

100

10

Oil a

nd w

ater

pro

duct

ion

rate

, B/D

1

0.1120,00080,000 100,00060,0000 20,000 40,000

Cumulative oil, bbl

Water

Oil

> Decline curve. Any sudden change in the slope of the usualstraight-line decline in oil production rate is a warning that excesswater, as well as other problems, may be affecting normal production.

10,000

1000

0

100

10

1

0.1

Time, days

Barre

ls p

er d

ay

10 100 10,0001000

Water flow rate

Oil flow rate

Diagnostic plots—A diagnostic log-log plotof WOR versus time can be used to help deter-mine the specific problem type by making com-parisons with known behavior patterns (left).Three basic signatures distinguish between dif-ferent water breakthrough mechanisms: openflow through faults, fractures, or channel flowbehind casing; edgewater flow or a moving OWC;and coning problems.7 Edgewater flow interpre-tations have been constructed from numericalsimulation and field experience.8 The time-derivative of the WOR also can be used, but theuncertainty or noisy nature of field measure-ments generally limits its application. The inter-pretation engineer can learn to recognize themany variations in these profiles and minimizethe problem of nonuniqueness, when combinedwith other data.

The usefulness of WOR diagnostic plots indetermining multilayer water encroachment isillustrated by an example in a field operated by amajor North Sea operating company. A medium-size reservoir with a moderate-to-high energyshoreface structure had been heavily bioturbated,giving rise to substantial permeability variations(next page, top). No significant shale barriers werepresent, and the 360-ft [110-m] thick reservoirfrom X590 to X950 ft [X180 to X290 m] gentlydipped into an aquifer. The edges of the reservoirwere bounded by sealing faults and truncated byan unconformity. A vertical well was perforatedacross 165 ft [50 m] in the middle of this unit. NoOWC or gas-oil contacts (GOC) were present inthe reservoir.

The WOR-diagnostic plot generated frommonthly well-test data shows the effect of the per-meability variation in the reservoir strata (nextpage, bottom). The plot illustrates watering-out ofhigh-permeability layers, which contribute tocrossflow in the reservoir. The ratio of break-through times (1800:2400:2800) gives an indication

38 Oilfield Review

100

10

1.0

0.1

WOR

WOR

100

10

1

0.0001

0.1

0.01

0.001

WOR

100

10

1

0.000110,0001000100

Time, days

101

0.1

0.01

0.001

WOR

WOR

WOR'

WOR'

WOR

> Diagnostic-plot profiles characterizing water breakthrough mecha-nisms. An open flow path (top) shows a very rapid increase. This profileindicates flow through a fault, fracture or a channel behind casing,which can occur at any time during the well history. Edgewater flow(middle) generally shows a rapid increase at breakthrough followed bya straight-line curve. For multiple layers, the line may have a stair-stepshape depending on layer permeability contrasts. A gradual increase(bottom) in the WOR indicates the buildup of a water cone early in thewell’s life. It normally levels off between a WOR of 1 and 10. The slopeof WOR decreases. After the water cone stabilizes, the WOR curvebegins to look more like that for edge flow. The magnitude of the slope,WOR’, is shown in red in the two lower profiles.

Spring 2000 39

of the permeability ratios in these layers. Thecumulative oil produced and the relative perme-ability-height products of the layers might be usedto estimate the remaining reserves in the lowerpermeability parts of the formation from X590 toX670 ft [X204 m].

The observed WOR response shows that lay-ers with higher permeabilities have watered out.Although there is no direct evidence of verticalconnection between these layers, an understand-ing of the depositional environment and theimpact of bioturbation can help resolve thisissue. Some communication between the high-permeability layers is likely, as well as possiblevertical communication within the remaining low-permeability zone. Any near-wellbore attempt tocontrol water from the high-permeability layerswill depend on vertical isolation over a largeareal extent between the remaining reservesabove X670 ft and the watered-out layers below.This can be confirmed with MDT ModularFormation Dynamics Tester measurements oflayer pressures, vertical interference testing,shale correlations and production logs.

Shut-in and choke-back analysis—The pro-duction history of most wells includes periods ofchoke-back or shut-in. Analysis of the fluctuatingWOR can provide valuable clues to the problemtype. Water-entry problems, such as coning or asingle fracture intersecting a deeper water layerwill lead to a lower WOR during choke-back orafter shut-in. Conversely, fractures or a faultintersecting an overlying water layer has theopposite effect. Such systems are not stableover geologic time but certainly can be inducedduring production.

7. Chan KS: “Water Control Diagnostic Plots,” paper SPE 30775, presented at the SPE Annual TechnicalConference and Exhibition, Dallas, Texas, USA, October 22-25, 1995.

8. Yortsos YC, Youngmin C, Zhengming Y and Shah PC:“Analysis and Interpretation of Water/Oil Ratio in Water-floods,” SPE Journal 4, no. 4 (December 1999): 413-424.

X590

X680

X770

X860

X950300025002000150010005000

Mea

sure

d de

pth,

ft

Horizontal permeability, mD

Wellbore

Perforations

> Horizontal permeability variations in a North Sea reservoir. Significant permeability variation results in effective layer isolation, thereby encouraging preferential flow along high-permeability layers. The well is perforated in the middle sectionof the reservoir.

10

1.0

0.1

0.01

0.0011000 2000 3000 4000 5000

Production time, days

WOR

1

2 34

> Diagnostic plot from monthly well-test data. The plot shows howaquifer water breaks through at about 1800 days (point 1) with asharp increase in WOR corresponding to a sudden water satura-tion change at the flood front. This breakthrough is most likely tobe from the highest permeability layer. The WOR gradually risesuntil 2100 days as normal for edgewater flow. The water inflowstabilizes from point 2 indicating that the layer is virtually wateredout, leading to a constant WOR. This value suggests that the firstlayer to break through contributes approximately 14% of the totalpermeability-height product—the key formation factor determiningthe flow rate. At 2400 days (point 3), the breakthrough of water isseen through the interbedded high-permeability layers. The curveappears to be less steep at this breakthrough because the WORis starting at a higher value. At the end of this period, the WOR isapproximately 0.24, suggesting that 10% of the permeability-height product comes from the second layer, which has wateredout. The last distinctive increase (point 4) represents final break-through of the remaining high-permeability layers.

3000

2000

1000

0

Flow

ing

pres

sure

, psi

Flow rate, B/D1000 2000 3000 4000

Oil Water Total flow rate

Water

Oil

100 mD, 4 ft

20 mD, 20 ft

> Multilayer NODAL analysis. The modeled well (insert) used for the NODAL analysis has two layers, each with a differentthickness and permeability. The multilayer analysis shows theindividual and total flow rates of the oil and water layers as theyare produced together at different pressures.

One well from the Middle East showed a pro-duction rate of 7000 bbl [1112 m3] of water perday and 400 bbl [64 m3] of oil per day after eachshut-in (above). These rates reversed after a fewdays of production. Production data suggest thatthe apparent cause was a conductive fault con-necting the oil reservoir to a shallower watered-out reservoir. In wells with the water source at ahigher pressure than the oil, choking back thewell causes the WOR to increase. The choke-backtest offers a useful diagnostic method to distin-guish between these two problems.

When production history data are of low qual-ity, a short-term production choke-back test canbe performed with several different choke sizes.The pressure should be monitored along withWOR from a separator or, preferably, a three-phase flowmeter, to accurately determinechanges in the WOR with drawdown pressure.This can be performed only if the well has suffi-cient wellhead pressure to flow at several ratesand so should be done early in the life of the well.

NODAL analysis—The design of a productionsystem depends on the combined performance ofthe reservoir and the downhole tubing or reservoir“plumbing” system (above right).9 The amount ofoil, gas and water flowing into a well from thereservoir depends on the pressure drop in the pip-ing system, and the pressure drop in the pipingsystem depends on the amount of each fluid flow-ing through it. The deliverability of a well oftencan be severely diminished by inadequate perfor-mance or design of just one component in the sys-tem. An analysis of a flowing wellbore and theassociated piping, known as NODAL analysis, isfrequently used to evaluate the effect of eachcomponent in a flowing production system fromthe bottom of a well to the separator.

NODAL analysis is also used to determine thelocation of excessive flow resistance, whichresults in severe pressure losses in tubing sys-tems. The effect of changing any component inthe system on production rates can be deter-mined.10 For example, a commonly held belief isthat choking back a well that produces water willreduce the water cut. This is certainly the case forconventional coning. In other cases, it depends onthe problem type as well as the reservoir pres-sures. For example, if a well is shut in for anextended period of time, the WOR (measuredwhen the well is put on line again) will depend onthe water problem and pressures involved.

A 35° inclined North Sea black-oil producer is perforated and producing from five differentlayers. Each layer is known to be isolated fromthe others by impermeable shale barriers with nocrossflow between them. A nearby injector andan aquifer provide pressure support. The wellproduced 29,000 B/D [4608 m3/d] with a watercut of 90%. A recent production log in this wellshows significant shut-in crossflow from lowerlayers into the upper—possibly a thief—layer.NODAL analysis was performed to match the PLTProduction Logging Tool analysis for both shut-inand flowing conditions, thereby providing confi-dence in any prediction of anticipated additionaloil production obtained from various water shut-off treatments (next page, top).

Although NODAL analysis is a standardmethodology for modeling wellbore response,there are two important considerations in its usein this application. First was the need to calibratethe computed flow responses in the face ofaggressive shut-in crossflow, and second, a rela-tively high number of separate layers wereinvolved. The analysis included six steps.

40 Oilfield Review

9. Elphick J: “NODAL Analysis Shows Increased OilProduction Following Water Shutoff,” presented at thePetroleum Network Education Conference’s 2nd AnnualInternational Conference on Reservoir ConformanceProfile Modification, Water and Gas Shutoff, Houston,Texas, USA, August 19-21, 1996.

10. Beggs HD: Production Optimization Using NODALAnalysis. Tulsa, Oklahoma, USA: OGCI Publications, Oil & Gas Consultants International, Inc., 1991.

11. A switch angle determines when primarily verticalmultiphase correlations should be replaced by primarilyhorizontal ones. Is important to note that there are nomultiphase-flow pressure-drop correlations in the publicdomain suitable for all inclination angles.

14,000 1.8

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

12,000

10,000

8000

6000

4000

2000

0 200 400 600 800 1000

Wat

er/o

il ra

tio

Tota

l liq

uid

rate

, B/D

Time, days

WOR

Liquid rate

> Production rates during choke-back. Production dataduring the choke-back period in a Middle Eastern wellshow that choking back the production rate 50% resultsin a dramatic increase in the WOR.

Laye

rs

Zonal flow rates, STB/D-6000 -4000 -2000 0 2000 4000 6000 8000

Option 1 oilOption 1 water

L1

L2

L3

L4

Option 1 (shut off just Layer 5)

Laye

rs

Zonal flow rates, STB/D0 1000 2000 3000 4000 5000 6000 7000 8000

Option 2 oilOption 2 water

L3

L4

Option 2 (shut off Layers 1, 2 and 5)

NODAL analysis to predict benefits of watercontrol. The two options proposed for this wellwere to either simply shut off Layer 5 with a plugand produce from the upper layers, or shut offLayers 1, 2 and 5, leaving Layers 3 and 4 to produce. The first option (top) would produce an expected net increase in production of 1328BOPD [211 m3/d], whereas the second choice(bottom) predicts a net increase in production of1647 BOPD [262 m3/d]. The second option is moreexpensive and probably requires setting a plugto isolate Layer 5 and cementing Layers 1 and 2.The operator chose option 1.

Spring 2000 41

• Model construction—Basic model constructionrequired a detailed deviation survey, pressure-volume-temperature (PVT) properties, charac-teristics of the reservoir in the near-wellboreregion for each layer and perforation locations.

• Geology—Geological information about thedepositional environment around the well wasnecessary to estimate the degree and lateralextent of impermeable barriers. The wellexhibited good lateral extent of such barriers.Elsewhere in the field, variation in depositionalenvironment caused uncertainty in the continu-ity of permeability barriers, degrading confi-dence in the sustainability of the localizedshutoff treatments.

• Layer pressures—Individual layer pressureswere obtained from shut-in data. Formationskin damage factors were initially assumed tobe zero.

• Correlation selection—A multiphase flow cor-relation comparison was conducted on thebasic system to determine the degree of varia-tion exhibited by the models and the impact ofcorrelation parameters, such as switch angles.11

This step involves matching well-test data.

• Shut-in crossflow—First, the shut-in crossflowexhibited by the PLT tool measurements wasmodeled, enabling skin damage for each layerto be evaluated. The process required a trialand error approach, in which rough estimates(from earlier tests) of each layer’s productionindex were sequentially adjusted to match the

data. Well histories were also consulted todetermine if any skin due to drilling or opera-tional considerations could be expected. In thisexample, none was expected.

• Flowing crossflow—The process was repeatedfor flowing conditions and several rates wereanalyzed. Shutting in all but one net-producinglayer at a time can speed up processing. Theproduction index and non-Darcy skin factors ofeach layer were then adjusted to match thedata. The final calibrated model provided agood match to all the data.

The calibrated NODAL analysis model wasthen used to determine the estimated incremen-tal production for two different shutoff options.The first option would completely shut off all pro-duction from the lowest layer, Layer 5 (below).This option leaves Layers 1 to 4 open, and the netresult is an increase in oil production from 2966 to4294 BOPD [471 to 682 m3/d]. Water productionwould decrease from 26,510 to 12,742 BWPD[4212 to 2025 m3/d]. The second option wouldinvolve sealing off the nonhydrocarbon-producingLayers 1, 2 and 5, and producing from Layers 3 and 4. This option results in oil productionincreasing to 4613 BOPD [733 m3/d], which is onlyabout 300 BPD [47 m3/d] more than option 1. The

Laye

rs

Zonal flow rates, STB/D-5000 0 5000 10,000 15,000 20,000

Calculated oilCalculated water

Measured oilMeasured water

L1

L2

L3

L4

L5

>Matching NODAL analysis with production measurements. The bluebars represent water flow while the green bars are oil flow measured byproduction logging tools. The circles represent the results of the NODALanalysis. Layers 2 and 5 are fully watered out. Layer 1 is taking on waterand some oil, as indicated by the negative flow rates, because it haslower in-situ reservoir pressure than the flowing wellbore pressure.

difference between current performance and thatpredicted from shutting in one or more layers wasused as the basis for justifying the treatments.

The production log data showed that waterwas being produced from all but one of the upperlayers. Most of the unwanted water came fromthe lowest layer. Because of reduced formationpressures, the uppermost layer was stealing asmall quantity of the oil and water being pro-duced below. As expected, the liquid volumesentering this thief zone decreased as productionincreased. At the expected high production ratessuch losses were considered tolerable. The oper-ator decided on option 1, setting a plug justbelow Layer 4, completely isolating Layer 5.

Production logs—Accurate production logs,such as those from the PS PLATFORM ProductionServices measurements can show water entryinto the wellbore.12 This tool can determine flowand holdup for each fluid phase in vertical, devi-ated and horizontal wellbores.13 The addition ofnew optical and electrical sensors incorporatinglocal probe measurements and phase-velocitymeasurements have resulted in major improve-ments in the diagnosis in both complex and sim-ple wells with three-phase flow. Such advancesin reliable and accurate production logging, par-ticularly in deviated wells with high water cuts,represent a major step forward in identifying andunderstanding water-problem types.

For example, an operator drilled a horizontalwell in the Gulf of Mexico through a small gassand that was producing excessive water after ashort time on production. In this well, the mostlikely source of the unacceptable water wasthought to be edge water from the lower aquifer.If the edge water was entering at the heel of thewell, then a cost-effective solution would be torun coiled tubing into the well and cement theportion around the heel, leaving the coiled tubingin place to allow production from the toe of thewell. This would delay further water productionuntil the water advanced past the cement plug.However, if water was coming from the toe of thewell, then it was possible to cement the lowerportion of the well using coiled tubing and apacker in the screen. A final scenario, waterentering from the middle of the well, would makeit difficult to isolate the water entry and continueproduction from the toe and heel. The operatorneeded to know the exact entry point of thewater production to take proper remedial action.

The logging program included the basic PSPLATFORM tool string along with the GHOST GasHoldup Optical Sensor Tool and the RSTProReservoir Saturation Tool run on coiled tubing.The GHOST, FloView holdups and spinner-derivedfluid velocity represent fluids inside the com-pletion screen, while the TPHL log and WFLmeasurements respond to flow both inside andoutside the screen (left).

The WFL water velocity measurements arecombined with the GHOST and TPHL holdup mea-surements to calculate the water flow-rate pro-file. In this example, more than 50% of the waterproduction is coming from the toe of the well,flowing behind the screen and in the openholegravel-pack annulus. The GHOST measurementalso identified additional water entering midwayalong the horizontal wellbore at X450 ft [X137 m].Since most of the gas is coming from the toe ofthe well, the operator decided to continue pro-duction without further intervention.

42 Oilfield Review

Measureddepth, ft

X200

X300

X400

X500

X600

Devi > 90° Gas

Water

TPHL TPHL

True vertical depthftX070 X055

Gas

GHOST

Holdup1 0

WaterWater profile

WFLwater flow rate

1 0

Deviation

Gamma rayAPI20 70

85 95deg WFLwater velocity

ft/min0 500

Gas profile

Gas flow rateB/D B/D0 1200 0 25,000

Gas

Holdup

Water

Waterentry

Waterentry

> Downhole flow profile. Track 1 contains gamma ray (green) and wellbore deviation (solid black) fromopenhole logs. The measured depth is shown in track 2. In track 3, gas (red) and water holdup (blue)measured by the GHOST Gas Holdup Optical Sensor Tool clearly identify water entering the horizontalsection of the wellbore at X450 ft and X640 ft. Track 4 shows gas (red) and water (blue) contributionsacross the entire wellbore and annulus, which is plotted against the wellbore trajectory profile. Theseindependent phase holdups are derived from the TPHL three-phase holdup log. Increasing water in theprofile can be seen as the wellbore turns more vertical above X350. Track 5 shows TPHL gas (red) andwater (blue) holdup logs. The WFL Water Flow Log water-velocity measurements (blue circles) areshown in Track 6. Track 7 contains a water flow-rate profile computed from the TPHL holdup and WFLvelocity. Track 8 contains the gas flow-rate profile computed using GHOST holdup data.

Spring 2000 43

Through-casing imaging tools, such as the USIUltraSonic Imager tool can help evaluate thequality of the cement job in a well and identifyflow channels behind casing. For example, in awell in New Mexico that was producing onlywater, the existence of a channel above the per-forations was confirmed (above). The well beganproducing oil after a cement squeeze and is cur-rently flowing 50 BOPD [8 m3/d] and no water.

Special Diagnostics for Vertical CommunicationWater crossflow has two clearly defined forms. Inaddition to crossflow in the reservoir, which hasalready been discussed, crossflow also occursinside the wellbore. Both kinds of crossflow areinterdependent and deserve careful consideration.

A potential for wellbore crossflow existswhenever the wellbore penetrates multiple lay-ers at different pressures. The pressure differ-ence is maintained only when and where there iscontinuous isolation between each layer. Thisimplies that reservoir crossflow and wellborecrossflow are mutually exclusive for any pair oflayers. Some reservoirs, for example those withstacked sand channels, have local shale barriersextending hundreds of meters. However, suchreservoirs may contain globally distant verticalconnections that lead to crossflow and pressurecommunication even though they exhibit localisolation with transient pressure variationsbetween layers during a choke-back test. Thisgives a mixture of the watered-out layer prob-lems with and without crossflow.

Identifying the presence of crossflow in theformation is critical. Watered-out layers withoutcrossflow can be easily treated at the wellbore,

while there are no simple solutions if the layersare not isolated by impermeable barriers.Additionally, watered-out layers without cross-flow will be subject to crossflow within the well-bore during shut-in. Several diagnostic methodsare useful in determining vertical communication.

Multirate tests—With little additional effort, aproduction log can be turned into a multirate pro-duction log, or ‘multilayer test,’ by measuring theproduction rate of each layer at several differentproducing pressures with station measurementspositioned between each layer. This helps deter-mine the productivity index and average reservoirpressure for each layer.14 In this way, crossflowpotential can be assessed using NODAL analysis.

Wireline-conveyed formation testers—Wireline formation pressure measurements,such as those from the MDT tool or the RFTRepeat Formation Tester tool can show if the lay-ers are in pressure communication.15 If layershave different pressures and are not in wellborecommunication, then they are isolated (below). Ifthey show the same pressure, they may be incommunication or they may have simply beenproduced (and injected) at similar rates, givingthe same pressure.

12. Lenn C, Kuchuk F, Rounce J and Hook P: “Horizontal WellPerformance Evaluation and Fluid Entry Mechanisms,”paper SPE 49089, presented at the SPE Annual TechnicalConference and Exhibition, New Orleans, Louisiana, USA,September 28-30, 1998.

13. Akhnoukh R, Leighton J, Bigno Y, Bouroumeau-Fuseau P,Quin E, Catala G, Silipigno L, Hemmingway J, HorkowitzJ, Hervé X, Whittaker C, Kusaka K, Markel D and Martin A: “Keeping Producing Wells Healthy,” Oilfield Review 11, no. 1 (Spring 1999): 30-47.

14. Hegeman P and Pelissier-Combescure J: “ProductionLogging for Reservoir Testing,” Oilfield Review 9, no. 2(Spring 1997): 16-20.

15. AL Shahri AM, AL Ubaidan AA, Kibsgaard P and Kuchuk F:“Monitoring Areal and Vertical Sweep and ReservoirPressure in the Ghawar Field using Multiprobe WirelineFormation Tester,” paper SPE 48956, presented at theSPE Annual Technical Conference and Exhibition, New Orleans, Louisiana, USA, September 27-30, 1998.

X100

X200

X300

Depth, ft

Channel Channel

Perforations

> A channel that produces water. The image ofthe cement in the annulus behind casing helpedto identify a water channel. The USI UltraSonicImager tool images—amplitude (track 1) andtransit time (track 2)—confirm that a large openchannel exists in the cement annulus behind thecasing just above the perforations.

X100

X000

5200 5400 5600 5800 6000 6200 6400

X300

Dept

h, ft

Pressure, psi

X500

X700

X400

X600

X200Upper Jurassic

Tarbut

Ness

Etive

Rannoch

Formations

Currentreservoirpressures

Initialreservoirpressures

> Pressure measurements showing layer isolation. Pressure measure-ments, such as those from the MDT tool, can be used in in-fill wells toestablish the pressure in each layer after a period of production in thefield. When pressure differences exist between layers due to differen-tial depletion, they show that the layers are isolated from each otherby vertical permeability barriers.

Vertical interference test—A vertical inter-ference test performed with the MDT tool willshow effective vertical permeability near thewellbore. Vertical permeability can be deter-mined from the change in formation pressuremeasured by a pressure probe, as formation fluidis pumped from the formation by a second (sam-pling) probe located about 2.3 ft [0.7 m] fartheralong the wellbore face.16

Shale correlations—Log correlations candemonstrate whether extensive shale barriersexist across a field. Excellent shale correlationsfrom well to well suggest that reservoir layersare isolated by impermeable rock and that cross-flow is unlikely.

Spinner survey during shut-in—A productionlog (spinner) may detect wellbore crossflow dur-ing well shut-in, a clear sign of a pressure differ-ence between isolated layers.

Choke-back test—Choke-back tests or produc-tion data can provide a useful diagnosis of verti-cal communication through the detection ofpressure differences.

Water-Control SolutionsEach problem type has solution options thatrange from the simple and relatively inexpen-sive mechanical and chemical solutions, to themore complex and expensive reworked comple-tion solutions. Multiple water-control problemsare common, and often a combination of solu-tions may be required. Today, in addition to thetraditional solutions described above, there arenew, innovative and cost-effective solutions forwater-control problems.

Mechanical solutions—In many near-wellbore problems, such as casing leaks, flowbehind casing, rising bottom water and watered-out layers without crossflow, mechanical orinflatable plugs are often the solution of choice.The PosiSet mechanical plugback tool can bedeployed on coiled tubing or wireline, and is afield-proven technology that ensures reliablewellbore shutoff in cased- and openhole environ-ments (right).

When the wellbore must be kept open tolevels deeper than the point of water entry, athrough-tubing patch may be the answer. Forexample, a new coiled tubing- or wireline-deployed, inside-casing patch called the PatchFlexsleeve has been used successfully in many appli-cations worldwide (far right). It is particularly wellsuited to through-tubing water or gas shutoff,injection-profile modifications and zonal isolation.The inflatable sleeves are custom-built to matchthe length of the perforated intervals and can

withstand wellbore crossflow pressures. Once set,the sleeve becomes a composite liner inside thecasing that is millable using through-tubing tech-niques if a subsequent squeeze operation isdesired, or it can be reperforated later to allowreentry to the zones. The only disadvantage of thecomposite liner is a reduction of less than 1 in. [2.5 cm] in the wellbore diameter. However, othermechanical patch remedies take up even more ofthe available casing inner diameter.

Shell UK Exploration and Production reducedwater cut in a North Sea well from 85% to 10%by using a PatchFlex sleeve to isolate the water-producing intervals. The PS PLATFORM loggingtool quantified fluid contributions from each pro-ducing zone. Two 4-ft [1.2-m] perforated intervalswere identified as producing most of theunwanted water. The RST readings confirmed the

44 Oilfield Review

> PosiSET mechanical plugback tool application.The PosiSET through-tubing plug is used fornear-wellbore water shutoff. The wireline- orcoiled tubing-deployed plug uses a positiveanchoring system with upper and lower slip-anchors (top) that isolate water-producing layers in both open and cased holes (bottom).

> The PatchFlex sleeve. A flexible compositecylinder made of carbon fiber, thermosettingresins and a rubber skin, the PatchFlex sleeve isbuilt around an inflatable setting element that isattached to a running tool and run into a well on a wireline. When the sleeve is positioned oppositethe area to be treated, a pump within the runningtool inflates the sleeve using well fluid. The resinsare then heated until fully polymerized. The inflat-able setting element is then deflated and extractedto leave a hard, pressure-resistant sleeve that fitssnugly, even in damaged or corroded casing.

Spring 2000 45

high water saturation in the water-producingintervals. In addition, the RST saturation analysisidentified two more unperforated oil zones belowthe other producing zones. A traditional bridgeplug could shut off the water-producing zone, butwould also block the new oil zones beneath.Using PatchFlex technology, Shell shut off thewater-producing zones and produced the new oilzones below them.

Chemical solutions—Chemical treatmentsrequire accurate fluid placement. Coiled tubingwith inflatable packers can help place most treat-ment fluids in the target zone without risk to oilzones. Coiled tubing dual injection is a process ofpumping protective fluid down the coiled tubingto the casing annulus and delivering the treat-ment fluid through the coiled tubing (right).

SqueezeCRETE cement is another keyweapon in the armory of water-control solu-tions.17 Its low fluid loss and capability to pene-trate microfractures narrower than 160 micronsmake it ideal for remedial treatment of tubingleaks caused by flow behind pipe. Once set, thiscement shows high compressive strength, lowpermeability and high resistance to chemicalattack. SqueezeCRETE treatment is often usedwith common cement for shutting off perfora-tions when the problem is watered-out layers, orrising bottom water or OWCs. Other applicationsinclude sealing gravel packs, casing leaks orchannels behind casing.

Rigid gels are highly effective for near-wellbore shutoff of excess water (right). Unlikecement, gels can be squeezed into the target for-mation to give complete shutoff of that zone or toreach shale barriers. They have an operationaladvantage over cement treatments because theycan be jetted rather than drilled out of the well-bore. Typically based on cross-linked polymers,products like MaraSEAL and OrganoSEAL-Rsystems can be easily mixed and have a longworking life. They can be bullheaded into the for-mation to treat specific water problems such asflow behind casing and watered-out layers with-out crossflow, or selectively placed in the waterzone using coiled tubing and a packer.18

Another solution is a flowing gel that can beinjected into small faults or fractures, but onlypenetrates formations with permeabilities greaterthan 5 darcies. Large volumes (1000 to 10,000 bbl)[159 to 1589 m3] of these inexpensive fluids oftensuccessfully shut off extensive fracture systemssurrounding waterflood injector or producing

wells.19 Like rigid gels, products such as Marcitand OrganoSEAL-F systems are cross-linked poly-mers that are simple to mix, have a long (up tothree days) working time before becoming rigid,and can be pumped through completion screens.

Smart or selective fluids in the form of poly-mers and surfactants are being developed for for-mation matrix treatments near the wellbore.These treatments, called relative permeabilitymodifiers, produce a permanent gel-like material

16. Crombie A, Halford F, Hashem M, McNeal R, Thomas EC,Melbourne G and Mullins OC: “Innovations in WirelineFluid Sampling,” Oilfield Review 10, no. 3 (Autumn 1998):26-41.

17. Boisnault JM, Guillot D, Bourahla A, Tirlia T, Dahl T,Holmes C, Raiturkar AM, Maroy P, Moffett C, Mejía GP,Martínez IR, Revil P and Roemer R: “Concrete Develop-ments in Cementing Technology,” Oilfield Review 11, no. 1 (Spring 1999): 16-29.

18. These gels will not penetrate formations with perme-ability less than 25 mD.

19. O’Brien W, Stratton JJ and Lane RH: “MechanisticReservoir Modeling Improves Fissure Treatment GelDesign in Horizontal Injectors, Idd El Shargi North DomeField, Qatar,” paper SPE 56743, presented at the SPEAnnual Technical Conference and Exhibition, Houston,Texas, USA, October 3-6, 1999.

> Coiled tubing dual injection. In water-control problems where the treatment fluid placement is critical, a coiled tubing-conveyedinflatable packer (A) can be used to provide wellbore isolationbetween the oil (B) and watered-out (C) zones. In this gravel-pack example, a treatment fluid (D) to stop unwanted water entry ispumped through the coiled tubing into the lower watered-out zoneand a protective fluid (E) is simultaneously pumped through theannulus into the oil-producing zone.

> Rigid-gel application using coiled tubing. Pumping a rigid gel (A)into the watered-out zone can shut off water entry from a layerwithout crossflow. A coiled tubing inflatable packer (B) isolates the oil-producing zone (C) from the watered-out zone (D).

Oil zone B

Watered-out zone C

A

D

D E

E

Treatment fluid

Protective fluid

Tubing

Coiled tubing

PackerGravelpack

Packer

Casing

Oil zone C

Watered-out zone D

Barrier

Tubing

Coiled tubing

Packer

Casing

Packer B

A Rigid gel

to stop flow in water layers, but retain fluid behav-ior in oil layers to allow production to continue. Insome applications, they offer the potential of per-forming a selective treatment simply by using alow-cost bullheading method of placement.

Treatments for water problems in horizontalwells are most effective when the treatmentzone is isolated from the remainder of the well-bore. In cased holes, and to some extent in open-holes, this is achieved mechanically withinflatable packers. However, when a screen orliner has been run but left uncemented, suchmechanical devices are not effective in isolatingthe open annular space behind the pipe.Developed for such situations, the AnnularChemical Packer (ACP) achieves zonal isolationusing coiled tubing-deployed packers or bridgeplugs (right).20 The objective of the ACP is toachieve full circumferential coverage over a rela-tively small length while leaving the liner free ofmaterial that might obstruct fluid flow or toolpassage through the section. A low-viscosity,cement-based fluid is pumped through coiled tub-ing and a straddle-packer assembly and placedthrough the small slots in the pipe. Once placed,the fluid immediately develops high gel strengthto prevent slumping and ensures complete annu-lar filling and isolation.

Completion solutions—Alternative comple-tions, such as multilateral wells, sidetracks,coiled-tubing isolation and dual completions, cansolve difficult water problems such as risingOWCs, coning, incomplete areal sweep andgravity segregation.21 For example, coproducingwater is a preferred strategy for coning in high-value wells. It involves perforating the water legand using dual completions (below).

Injector ProblemsInjectors can induce problems if the injectionwater is not properly filtered, because it maycontain particles large enough to cause matrixplugging. Or, if it is not treated properly withproduction chemicals such as bactericide andoxygen scavengers, damage can build up. Both ofthese can increase injection pressure until a frac-ture is initiated. Initially short, these fractureswill grow in length and height to maintain injec-tivity as the fracture faces become plugged.22

When induced fractures extend vertically overseveral layers, the operator no longer has controlover the vertical sweep. It is difficult to regaincontrol of the injection profile.

Thermal fracturing, often encountered off-shore, is caused by the stress reduction in theinjection zone from cool-down. The zone with the highest injectivity cools down first and then fractures—taking even more injection fluid andcausing poor vertical sweep (below). In thesecases, it is difficult to avoid thermal fracturing.The best strategy may be to ensure that all zonesare fractured, either thermally or hydraulically, toensure a more even injection profile. Sometimesif a high-permeability layer is adjacent to a low-permeability layer, the thermal fracture can breakinto the high-permeability zone, taking all theinjection water and leaving the low-permeabilityzone unswept.

46 Oilfield Review

> Annular Chemical Packer. ACP technology involves placement of a cement-based fluid into the annular space between an uncemented slotted liner andthe formation. The fluid is conveyed to the treatment zone using coiled tubingand injected between an inflatable packer assembly to fill the annulus over aselected interval. It is designed to set in this position forming a permanent,impermeable high-strength plug, fully isolating the volume of the annulus.

> Fighting water with dual drains. One solution to water-coning problems (left) is to perforate the water legof the formation and coproduce (middle) the water to eliminate the water cone. This low-cost approachmay increase the water cut, but improves the sweep efficiency and long-term reserve potential. Alternatively, the water and oil can be produced separately through the tubing and annulus (right).

> Thermal fracturing in an injector well.Fractures can be initiated in injector wellsthrough pressure and thermal stressinduced by cold-water entry. As a result,the vertical sweep profile is compromised.

Descaling

not

successful

Resetplug

Plug set

OK

Plug not

set

Plug set

OK

Decision tree for a well with scale. The decision tree presents different possiblescale treatment outcomes represented bybranches with the economic losses or profitsand the probabilities of reaching the end ofeach branch. Circular nodes (yellow) representchance nodes where two or more possibleoutcomes exist. The outcome of each branchis independent of any other node, and theprobability of each branch is described by a unimodal probability distribution (green)computed from Monte Carlo simulations.Square nodes (blue) represent decisions in which the branch selected is a matter of choice, with no element of chance. Thebranch endings represent revenues—calledvalue maximization. These help compare different scenarios in an optimal allocation of scarce resources.

Spring 2000 47

Evaluating RiskJustification of a treatmentin any well is based on the valueof the increased hydrocarbon produc-tion expected. The key word here is‘expected,’ which indicates a degree of uncer-tainty in the analysis. Some water-control treat-ments can guarantee substantial productionincrease. In such circumstances, the primary ele-ment of uncertainty is the job success itself. Whenthe incremental production is relatively small (orwas based on several assumptions) not only doesjob-risk come into play, but also the prediction itselfbecomes a key risk. Therefore, the value of a water-control treatment to the operator needs to be quan-tified. An analysis incorporating the multifacetedcomponents of risk can be undertaken using themethods of quantitative risk analysis (QRA).

D e c i s i o ntrees are valuable

tools to visualize andquantify all the options avail-

able to a decision-maker and theprobability of their outcomes. As an illus-

tration, PrecisionTree, provided by PalisadeCorporation, is a decision-analysis program usedwith the Excel spreadsheet program. The softwarecan be coupled to Monte Carlo methods, furnishinga ‘risked decision tree’ to analyze water-controloptions for specific wells (above).

Field-Wide Water ControlWater-control problems, diagnostic techniquesand solutions have been discussed in the contextof their application to individual wells within afield. However, if diagnostic techniques are modi-fied and extended to large number of wells in a

field, then there is greater reduction in total fieldwater handling and, in many cases, signifi-

cant enhancement in total field hydro-carbon production. By combining

the correct diagnosis withthe application of proven

solutions, water control can be aneffective reservoir management tool.

It is possible to apply individual wellwater-control strategies to a number of wells

within a field; however, in large fields, this canbecome time-consuming and inefficient. The firstobjective in a field-wide water-control program isto screen wells with the following characteristics:• The well is accessible for intervention.• The completion is robust enough to tolerate

intervention.• There is economic value to reducing water pro-

duction from that well.• The well has a water-control problem that can

be treated economically with acceptable risk.Field-wide water-control strategies often are

different from those applied on a well-by-wellbasis. For example, completion designs that haveworked effectively on single wells may need to bemodified for field-wide improvements. In one case,a South American operator was producing from alayered reservoir with distinct flow units separated

20. Elphick J, Fletcher P and Crabtree M: “Techniques forZonal Isolation in Horizontal Wells,” presented at theProduction Engineering Association Meeting, Reading,England, November 4-5, 1998.

21. Hill et al, reference 5. 22. Injectivity is a measure of how much liquid can be pumped

in a well (or zone) with a given difference between theinjection fluid pressure and formation pressure.

by shales. The operator perforated all layers andignored the variable pressures across the differentlayers. Eventually, water appeared at several lay-ers in different wells, and the subsequent pressuredepletion caused decreased oil production in theremaining layers. Originally, the operator simplyshut off water in the offending layers where thelocal geology was favorable, but field productioncontinued to decline because of increased occur-rence of water breakthrough and possible cross-flow through discontinuities in the shale barriers.Using a field-wide water-control strategy, the oper-ator moved away from commingled to single-layerproduction in each well, so that the crossflow couldnot occur and full effective drawdown on the lowoil-pressure layers was achieved. This meansfewer wells were draining each layer, but the fieldwas being swept more efficiently.

Field-wide considerations also include thecollective influence of inflow performance ofmany wells. Local and regional geology—interms of structure and heterogeneity—influencefluid movement. For example, the hydraulic rela-tionships between producers and aquifers orinjector wells should be considered (left). Currentand future completion strategies also are impor-tant factors in the analysis. Clearly, a lengthyscoping, or screening, study is not required everytime a field-wide water-control project is under-taken. Nor should a scoping study simply be asifting mechanism for finding treatable wells.The study must fit the problem, and the operator’sextensive knowledge can often help augmentand expedite the study.

Every water-control scoping study uses engi-neering diagnostic tools to identify which wellshave high value and can be effectively treated atlow risk. The scoping study consists of twophases, the diagnostic phase and the solutionsphase. The diagnostic phase uses the operator’sregional expert knowledge and experience cou-pled with Schlumberger engineering and soft-ware to profile the nature and cause of theproblem. Wells are initially screened to select afocus area within the field, then again to identifywells that might benefit from some type of inter-vention, and finally to choose wells that are ofsufficient value to justify treatment.

WaterCASE software-based methodologyscreens candidate wells on the basis of existingdata such as production histories, existing pro-duction logs, reservoir characterization from bothnumerical and analytical models and offset treat-ment data and experience (next page). One recentstudy provided by Schlumberger in the North Seaillustrates the results of the screening process.Here a field contained nearly 100 wells with

water cuts ranging from 20% to 90%, and a fieldaverage of 60%. The scoping study made the fol-lowing determinations:• 15 wells are subsea, requiring a rig for inter-

vention, and 6 have production tree or ‘fish-in-hole’ problems, making intervention difficult.

• Of the remaining 85 wells, 20 have corrodedtubulars, increasing intervention risk.

• Of the remaining wells, 25 have significantpotential for additional productivity if thewater cut is reduced.

• Of these 25 wells, 15 have solvable problemsconsisting of casing leaks, flow behind pipe,bottom water, high-permeability layers withoutcrossflow, or fractures from injector to producer.

The results identify primary candidate wellsto take through to the second phase of the inter-vention process—developing a solutions plan.

In this phase, a spectrum of solutions includ-ing mechanical, fluid and completion options isdeveloped. The solutions spectrum is ranked byrisk, cost and benefit using Schlumberger quanti-tative risk analysis (QRA). Solutions range from‘quick hit and rapid pay’ to longer duration,‘higher cost with higher pay’ solutions.Schlumberger works jointly with the operator’sasset team to identify the most cost-effective,lowest risk and highest value treatment optionfor each well. The chosen solution for each can-didate well is fully engineered for final submis-sion and peer review prior to execution.

To maximize field-wide cost reductions, sur-face-related water-control services (page 50)should be included in the overall screening pro-cess. An integrated solution is often a combina-tion of borehole, reservoir-scale and surfacesystems. Surface facilities may contribute up to25% reduction in overall water-handling costs.

Field-Wide ProblemsEventually most oil fields are under a waterdriveeither from waterflood or a natural aquifer. Anyattempt to significantly increase the recovery fac-tor must increase at least one of the componentsof the recovery factor: displacement efficiency,areal-sweep efficiency or vertical-sweep effi-ciency. The first, displacement efficiency, can beimproved only by reducing the residual oil satu-ration with a surfactant, miscible flood or water-alternating-gas scheme. Water control improvesareal- or vertical-sweep efficiency.

Any analysis of water sweep at a field scalerequires an understanding of the geology andproper reservoir characterization. Reservoir char-acterization, particularly heterogeneity, is poorlyunderstood early in the life of the field, but grad-ually improves as dynamic production databecome available.

48 Oilfield Review

1 year

2 years

5 years

10 years

> Streamline simulation. History-matchedFrontSim water-flow streamline simulations canbe used to show well interactions and detail theexact fraction of water that flows between theinjector and producer wells. In this example with10 producers (red circles) and 5 injectors (bluecircles), the model helps visualize where injectionwater is going at 1, 2, 5 and 10 years. Unsweptregions (blue) are clearly visible near the centerof the reservoir.

>WaterCASE screen. Here a typical user interface asks spe-cific questions (left) about symptoms and diagnostic testresults that help process analysis of the water-control prob-lem. Once a sufficient set of answers is completed, problemtypes are identified and ranked by score (right) according totheir likelihood of incidence. The WaterCASE logical structureis shown superimposed above the screen display.

Spring 2000 49

In calm depositional environments such asshallow marine, continuous shales are often pre-sent, providing good vertical isolation betweenlayers, and making vertical sweep improvementpractical. Any problem with watered-out layerswithout crossflow is easily corrected at the well-bore, and in this environment, this problem dom-inates the more difficult problem of watered-outlayers with crossflow.

Eolian sands, often thick with good verticalpermeability, pose different problems for watercontrol. They can exhibit gravity fluid segre-gation, causing unwanted water entry into pro-ducing wells.

Fluvial and deltaic depositional environmentstypically create sand channels. These may varyfrom well-stacked sands with good horizontaland vertical continuity to isolated channels withpoor communication. Since various problemtypes can occur in this setting, good sand char-acterization is important.

Carbonate reservoirs have their own chal-lenges, including frequent natural fractures lead-ing to water entry from a water layer, or throughfractures connecting injectors and producingwells. Additionally, large dissolution channelsfrom underground water flow, sometimes severalmeters across, can create superhighways to flow,often with premature water breakthrough. These

may be considered subsets of fracture-inducedwater problems. Shutting off this type of channelis extremely difficult.

Many operators are reluctant to proactivelycontrol water prior to breakthrough, so mostaction is remedial. Proactive water control wouldinclude choking back zones with higher permeabil-ity to create a more uniform sweep, but this wouldmean sacrificing early cash flow for an uncertainreturn due to incomplete knowledge of hetero-geneity. However, the production (and injection)profile can be improved through selective stimula-tion of zones with lower permeability. This is a

Typical surface water facilities and relativecosts. The surface water-management facilitiesinclude primary oil, water and gas separators,water-polishing systems to remove residual oil from the water, solids-filter systems as well as chemical treatments. These ensure that thereinjected water is compatible with the receivingformation and does not cause other problemssuch as scale deposits and corrosion in the well-bore system, and reservoir damage. Also shownare typical relative water-cycle costs from theproducing well (lifting costs of 17%), chemicals13%, removal and processing costs (includingseparation 9%, de-oiling 14%, and filtering 15%),pumping 27% and finally reinjection well costs5%. Estimates of average water-handling costs of 50 cents per barrel were based on the assump-tion that the fields were onshore and the wellswere 6000 to 8000 feet [1828 to 2438 m] deep, andproducing 1000 BOPD [159 m3/d] and injecting5000 BWPD [795 m3/d].

particularly attractive option because of the capa-bility of using coiled tubing to precisely placesmall hydraulic fractures. The improvement in hor-izontal drilling techniques, including multilateralsand coiled tubing, also is allowing a greater rangeof viable solutions for complex reservoir problems.However, the predominantly reactive mode forwater control, and hence sweep improvement, islikely to continue until more precise early reservoircharacterization is achieved.

Based on knowledge—or even a rough esti-mate—of the reservoir volume and the frac-tional-flow curve, the expected recovery can beestimated assuming production continues to agiven water cut. By comparing the expectedrecovery with the ultimate recovery indicated bythe WOR semilog plots, one can use field-widediagnostics to estimate how well the reservoir is

being swept. If the WOR is less than the frac-tional-flow curve indicates, then there isbypassed oil (above).23 If the oil production isaccelerated, then it must account for its time-delayed value when calculating net presentvalue—the value of the oil as it is producedminus its value when it would have been pro-duced. If the oil is incremental, then the water-control operation can assume all the value tohelp justify the costs of operation. Incrementaloil is often more valuable than accelerated oil.

Surface Facilities Surface facilities separate water from oil and pro-cess it to an acceptable specification suitable fordisposal to the environment or for reinjection(below). Gas is sent to a gas-processing plant or

simply flared, while the oil is processed in a‘water-polishing’ stage in which water is removedfrom the oil down to the 0.5 to 1.0% level, depend-ing on delivery requirements. Water is reinjectedfor both disposal and pressure maintenance. In atypical water-treatment facility for injection pur-poses, all water streams from each stage of sepa-ration are further de-oiled to a level compatiblewith discharge to the environment or receiving for-mation, typically between 10 and 40 ppm. Thisincludes filtering through a 10- to 50-micron filterto remove solids, making the water more compat-ible with the formation prior to reinjection.

Chemical treatments including emulsionbreakers, biocides, polyelectrolytes and oxygenscavengers are added to the water to condition itfor reinjection, and corrosion inhibitors and anti-scale chemicals are added to protect tubularsand downhole equipment. When water is pro-duced at high rates, chemical additives consti-tute up to 20% of the surface water-handlingcosts. Surface equipment and facilities accountfor the remaining 80%.

In practice, surface solutions start downhole.Partial downhole oil-water separation in the well-bore can eliminate some of the costs of liftingwater. An alternative to simultaneous downholeseparation and reinjection is downhole segre-gated production whereby water and hydrocar-bons are produced separately—avoiding the needfor surface separation capability. Finally, chemicaltreatments, such as emulsion breakers, antiscaleand corrosion inhibitors injected downhole canprepare fluids for efficient surface treatment.24

50 Oilfield Review

1.0

0.75

0.5

0.25

0.00 10 20 30 40 50 60 70

Frac

tiona

l flo

w, w

ater

-cut

frac

tion

Water saturation, %

A B

Water cut 95%

Final formationwater saturation,38%

Final formationwater saturation,58%

Fractional-flow prediction.The two fractional-flow plotsshow how a multilayer reservoirmight perform under differentassumptions. The two curvesshow a large difference in thefinal formation water-saturationvalue at the same water-cutflow rate. Assuming that reser-voir layers water out accordingto their flow capacity, Curve Ashows a substantial amount ofoil still remaining in the forma-tion. Assuming layers water out from bottom to top, Curve Bshows nearly all the oil isrecovered.

>

>

Spring 2000 51

Well pad factory concept—Existing separa-tion technologies and multiphase pumping arereadily available for commercial use as a “wellpad factory.” Oil, water and gas are separatedclose to the wellhead area and the unwantedwater and gas are reinjected for pressure main-tenance or disposal with multiphase pumps.

Conventional surface facilities—Conven-tional gravity-separation facilities can bedesigned for specific production profiles. Withbest practices and technologies, surface facili-ties can provide substantial savings in the water-removal chain (right). For example, centrifugalseparation performed by Framo Engineering—technology derived from multiphase pumpingpractices—could soon provide important opera-tional and capital savings by reducing theamount and size of equipment, and chemical-injection costs. Centrifugal separation could beextended to the well pad factory. Other specificwater-conditioning technologies used to reducethe concentration of water in oil to extremely lowlevels include water polishing, which can reducethe water content down to the 40 ppm level;ultrapolishing systems that reduce the waterdown to the 5 ppm level; and fine solids removalto filter debris such as sand down to 2-micronparticle size.

As worldwide daily water productionincreases, surface facilities, which were not orig-inally designed to handle large volumes of water,are being retrofitted with equipment that canhandle higher water fractions economically.Today, some reservoirs are being produced cost-effectively with over 95% water cut. In well-known reservoirs, such improvements inwater-handling services at surface facilities areunlocking additional recoverable reserves.

The LASMO Plc Apertura project in the Daciónfield in Venezuela is an example of a water-controlstrategy used to improve the economics of field-wide oil production by reducing the bottlenecks inthe water-handling capabilities of surface facili-ties. Managed by the LASMO-Schlumbergeralliance, the project, which began in April 1998,consists of three phases:• Complete an intensive upgrading and debottle-

necking of surface facilities to increaseprocessing capacity 50%, from 20,000 B/D[3178 m3/d] at 50% water cut to 80,000 B/D[12,712 m3/d] at 60% water cut, increasing oil production from 10,000 to 30,000 BOPD[1589 to 4767 m3/d].

• Install new production facilities with process-ing capacity of 360,000 B/D [57,204 m3/d] at75% water cut, reaching a 90,000 BOPD[14,300 m3/d] oil-processing capacity.

• Retrofit the water-handling module in thefuture to boost the mature-field water-handlingcapacity to cope with up to 90% water cut,allowing an economic final production phase of up to 600,000 B/D [95,340 m3/d] and 30,000 BOPD.

In this particular field-wide redevelopmentproject, water-control services and managementhave unlocked reserves by doubling the crude-oilrecovery factor from 14% to nearly 35%.

A Look at the FutureThe goals of reducing the costs of excess pro-duced water and unlocking additional recover-able reserves from mature fields appear difficult,but some quick victories are within reach.Understanding water-flow problems and theirsolutions is now a key component of today’sreservoir engineering.

Making the best of what we have is the firststep in water control, requiring a detailed under-standing of the assets, resources, activities andcosts associated with handling produced water.Opportunities may then become apparent to

reduce the costs of traditional practices and mate-rials (chemicals) and identify where future poten-tial cost increases can be controlled. Technicalinnovation will enable larger gross volumes to behandled with existing facilities. The total produc-tion system, from reservoir to custody transferpoint for oil and final resting place for water, mustbe considered. In many operator and service com-panies, research and development programs arecurrently targeted at developing appropriate toolsto manage this wave of produced water.

Finally, an integrated approach to water con-trol in every well from reservoir to disposal (orback to reservoir for pressure maintenance) willbring immediate and long-term cost-savings.Integrated water management services is envi-sioned as the key to reservoir production opti-mization by providing the means for producingadditional recoverable reserves. While water-control services will provide the bulk of progress,a downhole factory—built on the well pad fac-tory concept—will minimize produced water-handling costs, and optimized facilitiesprocesses could turn waste into a commodity,which will further enhance the recovery factor.Nevertheless, the real money comes from thepotential increase in oil production. —RH

23. Dake LP: “The Practice of Reservoir Engineering,” inDevelopments of Petroleum Science 36. Oxford, England:Elsevier, 1994: 445-450.

24. Crabtree M, Eslinger D, Fletcher P, Miller M, Johnson Aand King G: “Fighting Scale—Removal and Prevention,”Oilfield Review 11, no. 3 (Autumn 1999): 30-45.

Oilpump

Water pump

Interface levelcontrol valve

Water meter

Degasser

Hydrocyclone

First-stageseparator Second-stage separator

Flow path for removal of oil-contaminated water

> Surface water polishing. Oil is removed from producedwater prior to disposal into ariver or sea, or injection backinto the reservoir (top). Thehydrocyclone unit (bottom) ispositioned downstream of thewater outlets on the separatorand upstream of the degasser.Its function is to remove anyentrained oil from the waterand return it to the separationprocess before water is sent to the degasser.

Hatch

Oil compartment

Dirty-watercompartment

Individualhydrocyclones

Clean water

Clean-watercompartment

Dirty water

Oil reject

Hydrocyclone cross section