controls of water cresting in horizontal well

8/10/2019 Controls of water cresting in horizontal well

1/11

Copyright 2007, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Europec/EAGE Annual Conference andExhibition held in London, United Kingdom, 1114 June 2007.

This paper was selected for presentation by an SPE Program Committee following review ofinformation contained in an abstract submitted by the author(s). Contents of the paper, aspresented, have not been reviewed by the Society of Petroleum Engineers and are subject tocorrection by the author(s). The material, as presented, does not necessarily reflect anyposition of the Society of Petroleum Engineers, its officers, or members. Papers presented atSPE meetings are subject to publication review by Editorial Committees of the Society ofPetroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paperfor commercial purposes without the written consent of the Society of Petroleum Engineers isprohibited. Permission to reproduce in print is restricted to an abstract of not more than300 words; illustrations may not be copied. The abstract must contain conspicuousacknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.Box 833836, Richardson, Texas 75083-3836 U.S.A., fax 01-972-952-9435.

AbstractIt is widely agreed that gas reservoirs with a component ofwater drive should be produced at high rates to minimize thevolume of gas which is trapped at high pressure by theadvancing water (often termed outrunning the aquifer). Yethigh production rates are also associated with coning (invertical wells) or cresting (in horizontal wells) of the

encroaching water, leading to early water breakthrough. Invertical wells, the formation of an inverse gas cone means thathigh gas rates can be maintained post-breakthrough untilalmost the whole perforated interval is flowing water.However, in horizontal wells, water breakthrough is a serious

threat to gas deliverability, because the inverse coningmechanism does not apply and the well rapidly loads withwater. Consequently, it is not clear whether producing at highrates is the best strategy to maximize recovery in gasreservoirs developed using horizontal wells.

We investigate the risk associated with producing

horizontal wells at high rates by simulating gas recovery andaquifer response over a broad range of reservoir properties andproduction scenarios. We find that high rates always result inlower gas recovery unless the ratio of vertical to horizontal

permeability is very low, in which case water cresting issuppressed. However, there are many instances whereaccelerating production recovers only slightly less gas overmuch shorter timescales, so may be economically favorable.Rate sensitivity increases in low permeability reservoirs withthin gas columns, because these conditions increase thetendency for water cresting, and decreases in reservoirs with

strong aquifer support, since water breakthrough occursregardless of the rate at which the well is produced. Ourresults can be used as a reference framework to rapidly assessgas production behavior and aquifer response within a wide

range of field development scenarios.

IntroductionHorizontal and highly deviated wells are increasingly being

used in gas field developments worldwide.1-5

Large-borehorizontal wells can deliver significantly higher gasproduction rates than conventional completions,2 reducingfield development costs by allowing reserves to be targetedwith fewer wells.4-6However, realizing the potential of high-

productivity gas wells requires an understanding of thesubsurface risks to deliverability, to ensure sustained gasproduction, maximize profitability, and establish large-bore

completions as an economically viable development option. Akey subsurface risk in gas reservoirs with a component ofwater drive is early water breakthrough.3-5,7,8 In large-bore

horizontal wells, early water breakthrough is a particularlyserious threat to deliverability, because of the significantreduction in gas flow capacity associated with flowingentrained water to the surface.4,9-11At best, substantial waterproduction will require expensive processing facilities; atworst, it will effectively kill the well.3-5,12

Based on material balance considerations, it is widelyagreed that gas reservoirs with a component of water driveshould be produced at high rates. This approach (oftendescribed as outrunning the aquifer) maximizes gas recovery

by reducing the volume of gas which is trapped at highpressure by the advancing water.13-18 In this context, high

productivity horizontal wells might be expected to make apositive contribution to gas recovery, because they cangenerally produce at much higher rates than vertical wells.However, material balance approaches assume that the gas-water contact (GWC) remains flat duringproduction.13,14,16,18,20-22 Yet bottom-water drive gas reservoirsare associated with coning (in vertical wells) or cresting (in

horizontal wells) of the GWC towards the well.7,23-28

Crestingoccurs when viscous forces associated with pressuredrawdown overcome gravity forces resulting from the densitycontrast between gas and water, causing a crest or cone of

water to be drawn upwards towards the producing well29(Fig.1). Water crest behavior has been described using analytical

approaches to predict a critical rate above which waterbreakthrough is expected,

30,31 and time to water

breakthrough.32,33

These approaches imply a sensitivity of gasrecovery to production rate that conflicts with material balancetechniques, as the severity of water cresting is increased withaccelerated production, and therefore water breakthrough is

expected earlier at higher rates. Water crest development alsobecomes more significant as the separation between the welland GWC is reduced, the horizontal reservoir permeabilitydecreases, and the vertical permeability increases.29

In gas fields developed with vertical wells, it has been

suggested that the benefits of outrunning the aquifer outweigh

SPE 107169

Controls on Water Cresting in High-Productivity Horizontal Gas WellsR.P. Sech, SPE, M.D. Jackson, SPE, and Gary Hampson, Imperial College London


2/11

2 SPE 107169

the penalties associated with coning.24 McMullen andBassiouni (2000)28 explained this in the following way.Producing at high rates does indeed induce the formation of a

water cone, leading to earlier breakthrough. However, thehigh mobility of the gas means that high gas rates and low

water-gas ratios can be maintained post-breakthrough, owingto an inverted coning effect, until almost the whole perforated

interval is flowing water (Fig. 2). Their results suggest thathigh production rates are never detrimental to overall gas

production. McMullen and Bassiouni (2000)28

also suggestthat higher gas recovery is obtained in formations with

intermediate values of permeability (c. 10-100 mD). This isbecause well productivity is still high due to the high mobilityof the gas, but aquifer influx is restricted. As the permeabilityincreases above this range, the aquifer can respond more

rapidly to production, leading to higher abandonment pressure;as the permeability decreases below this range, gas well

productivity begins to decline. Finally, they suggest thatrestricting well completions to the top of the formation doesnot improve recovery. Completing vertical wells far away

from the GWC has conventionally been used as a strategy tomitigate against coning, because the tendency for coning is

increased as the stand-off decreases between the completioninterval and the GWC.

29However, they argue that decreasing

the completion interval reduces well productivity, and has nobenefit because coning is not a significant threat.

It is not clear whether these same conclusions will applyto horizontal wells, particularly those with high potential

productivity. Production capacity has been severely impairedby water influx in horizontal and highly deviated gas wellsproducing at significantly higher rates than the maximum of30 MMscf/d considered by McMullen and Bassiouni(2000).1,3,28 Generally, water influx results in the need to

significantly choke back production, because very high waterrates are difficult to handle at the surface;

3,4 expensive well

intervention is often required to locate the source of the waterand shut it off.

3At least one horizontal gas well in the North

Sea has experienced very early abandonment caused byexcessive water production.12 Despite this risk, several gasreservoirs have been developed recently using a small number

of high-rate horizontal wells to optimize recovery and netpresent value.2-6,8In developments such as these, it is desirable

to design a reservoir management strategy that avoids earlywater production.

The aim of this study is to determine quantitatively thecontrols on gas production and water breakthrough in dry gas

reservoirs producing from horizontal wells at high rates, in thepresence of an active bottom-water aquifer. We are

particularly interested in the risks associated with (1) veryhigh production rates (up to 500 MMscf/day 5); (2) uncertaintyin aquifer size and support; (3) reservoir permeability andpermeability anisotropy; (4) gas column thickness; and (5)

well location with respect to the GWC. We use a simplehomogenous simulation model to investigate these key

uncertainties. The results of this study provide a benchmark inrelation to which the effects of geologic heterogeneity on gasrecovery and aquifer behavior can be considered; we arecurrently investigating such effects using simulation models

that include a detailed description of facies architecture inshallow-marine reservoirs. Early water breakthrough in gas

reservoirs has been attributed to geologic heterogeneity inseveral cases.1,3,7,12,34

Figure 1: Simulated water saturation at breakthrough to ahorizontal well. The gas-saturated zone has been filtered out;only water saturation is shown. The gas-water interface respondsto viscous forces induced by gas production by forming anaxisymmetric crest geometry below the well.

MethodologyWe simulate gas production in the presence of an activebottom water aquifer using a simple three-dimensional model

(Fig. 3). The reservoir zone measures 900 m in the x and y(horizontal) directions and has a maximum thickness of 50 m.A bottom-water aquifer is attached to the base of the reservoir.The aquifer measures 1800 m in the x and y directions,extending beyond the reservoir. Aquifer thickness is 100 m.We change the effective size of the aquifer by modifying theaquifer porosity away from the reservoir connection.28 This

basic model geometry is used for all simulations.Reservoir and aquifer porosity (adjacent to the reservoir

connection) is constant in all simulations (= 0.25). We use therelative permeability and capillary pressure data measured forgas-water systems in the experimental work of Chierici et al.(1963)35and simulated in the coning study of Kabir (1983)25

(Table 1). The relative permeability to water (krw) at residualgas saturation (Sgr) is high and yields a higher water mobility

than considered in other coning studies.28,34

Production is viaa single 330 m (1000 ft) horizontal well with 0.18 m (7 inch)internal-tubing diameter, which is consistent with thedimensions of high-rate gas wells utilized in the Columbus

Basin, offshore Trinidad.4The well is located in the centre of

the model area. Initial reservoir pressure is 4094 psia, which

represents the average pressure of 12 reported water drive gasreservoirs from the Cassia and Mahogany Fields of theColumbus Basin.

8 Reservoir permeability (kh), permeability

anisotropy (kv/kh), gas column thickness, and stand-off

between the well and the GWC, are varied in a sensitivityanalysis which is described in the following section.


3/11

SPE 107169 3

B

Gas coning water breakthroughlow GWR

C

Gas coning high GWR

A

Water

Gas

Water coning

B


B


C

Gas coning high GWR

C

Gas coning high GWR

A

Water

Gas

Water coning

A

Water

Gas

Water coning

D

A CB

CumulativeGasProduction

Time

CumulativeWaterProduction

Low

High

Early Late

Low

High

DD

AA CB

CumulativeGasProduction

Time

CumulativeWaterProduction

Low

High

Early Late

Low

High

Figure 2: Series of schematic cross-sections illustrating coningbehavior in a gas-water system produced with a vertical well (after McMullan and Bassiouni, 2000)

28: (A) during early production, the

gas-water interface deforms below the well to form a prominentcone, (B) after water breakthrough, the gas-water interface invertsto create a gas-cone which protects the well from increasedwater influx, and (C) the high mobility of gas relative to watermaintains the gas-cone as the GWC rises, ensuring a low water-gas ratio until virtually all of the producing interval is flowingwater. (D)Cumulative recovery profiles for gas and water from theproduction scenario shown above. McMullan and Bassiouni(2000)

28simulated a similar response with a maximum production

rate of 20 MMscf/d, noting that nearly all gas is recovered prior to

a rapid increase in water production as the well eventually floods.

Table 1: Relative permeability and capillary pressure (in PSI) dataused in the simulation models.

Figure 3: Simulation model used in the sensitivity analysis. Thereservoir is shown in red; the aquifer in blue. The reservoirmeasures 900 m x 900 m areally, and a maximum of 50 mvertically. The aquifer measures 1800 m x 1800 m areally, and100 m vertically. The reservoir gas pore volume (GPV) is 68 Bcffor a gas column thickness of 50m.

We measure the sensitivity of reservoir performance to

various factors by considering the total volume of gasrecovered prior to well abandonment, when (1) water breaks

through to the well, or (2) reservoir pressure depletes to theminimum flowing bottom-hole pressure (BHP) of 500 psia. Inreal reservoir cases, gas production is likely to continue afterbreakthrough, but at significantly reduced rates and increased

cost because of the need to handle large volumes of water.However, this study focuses on the impact of production

strategy on water-free gas production in high-rate horizontalwells. Production rate is the primary target for eachsimulation, and is varied in the sensitivity analysis. The targetrate may decrease during production in order to comply with

the BHP constraint.

0.60000.00001.00000.80

0.53000.12500.65000.70

0.45000.22500.37000.60

0.37000.37500.20000.50

0.30000.50000.08500.40

0.23000.62500.02000.30

0.15000.75000.00000.20

0.07000.87500.00000.10

0.00001.00000.00000.00

Pc

krw

krg

Sg

0.60000.00001.00000.80

0.53000.12500.65000.70

0.45000.22500.37000.60

0.37000.37500.20000.50

0.30000.50000.08500.40

0.23000.62500.02000.30

0.15000.75000.00000.20

0.07000.87500.00000.10

0.00001.00000.00000.00

Pc

krw

krg

Sg

from Chierici et al. (1963)


4/11

4 SPE 107169

0.01, 0.1, 1kv/kh

5, 75, 250, 400, 750, 1000, 1400, 2000, 3000 (mD)kh

10, 20, 30, 40, 50 (m)Standoff

10, 20, 30, 40, 50 (m)Gas column thickness

1, 2.5, 5, 10, 25, 50, 200 (ratio of aquifer PV to GPV)Aquifer size

50, 100, 200, 300, 500 (MMscf/d)Production rate

the following parameter settings are varied for each simulation run, depending upon the sensitivity analysis

500 (psia)BHP limit

4094 (psia)Initial reservoir pressure

0.25Porosity

0.18 m (7 inches)Tubing diameter

330 m (1000 ft)Well length

the following parameter settings remain fixed for each simulation run

SettingParameter

0.01, 0.1, 1kv/kh

5, 75, 250, 400, 750, 1000, 1400, 2000, 3000 (mD)kh

10, 20, 30, 40, 50 (m)Standoff

10, 20, 30, 40, 50 (m)Gas column thickness

1, 2.5, 5, 10, 25, 50, 200 (ratio of aquifer PV to GPV)Aquifer size

50, 100, 200, 300, 500 (MMscf/d)Production rate

the following parameter settings are varied for each simulation run, depending upon the sensitivity analysis

500 (psia)BHP limit

4094 (psia)Initial reservoir pressure

0.25Porosity

0.18 m (7 inches)Tubing diameter

330 m (1000 ft)Well length

the following parameter settings remain fixed for each simulation run

SettingParameter

Flow is simulated using a gridding scheme that is locallyrefined around the well, and coarsened within the aquifer. Thehighest resolution cells are located closest to the well, and

measure 0.3 m perpendicular to the length of the horizontalwell, and 50 m parallel to the well (Fig. 3). The width of each

grid cell increases exponentially away from the well in the xdirection (perpendicular to the well) for a distance of 150 m.

Beyond this region, each reservoir grid cell measures 50 m inthe x and y directions (Fig. 3). The refinement in the x

direction is necessary to resolve the water crest geometry. Gridlayering in the z (vertical) direction is set constant at 1 m

within the reservoir zone (Fig. 3). The areal local gridrefinement is maintained within the upper 15 m of the aquiferto ensure flow convergence at the reservoir-aquiferconnection. Vertical cell thickness within this aquifer section

is increased to 5 m. The remaining grid cells below thisaquifer section measure 50 m in the x and y directions, and

15 m in the zdirection (Fig. 3). In total, 41,040 grid cells areactive in the reservoir section and 12,960 cells are active in theaquifer. All flow simulations are undertaken using the

Eclipse100 black oil simulator.

Sensitivity AnalysisThe following study assesses the sensitivity of gas recovery tofour key parameters: (1) gas production rate; (2) aquiferstrength; (3) gas column thickness, and (4) well-GWC stand-off. Table 2 lists the ranges over which these parameters arevaried, and also the range of kh and kv/kh, against which the

impact of each parameter is tested. Parameter ranges arechosen to represent good reservoir quality sands typical of gasfields in the Columbus Basin. Management strategy within anumber of these fields relies upon using few high ratehorizontal wells to target thick, sand-rich reservoirs.4,8 It has

been forecasted that production rates of up to 500 MMscf/dmay be achieved from these reservoirs,

5 yet the loss of

deliverability from a single well can potentially disableproduction from an entire reservoir. Therefore, we considerthe risk of producing at high rate in a reservoir with a bottomwater-drive aquifer, against the critical parameters which may_

be difficult to characterize from subsurface datasets, and areoften associated with significant uncertainty.

Impact of Production Rate on Gas Recovery and Aquifer

Behavior. We begin by considering the sensitivity of gas

recovery to production rate for different values of reservoirpermeability and permeability anisotropy (kv/kh ratio). Each

simulation run assumes the well is located at the top of a gascolumn 50 m (164 ft) thick. Aquifer strength is fixed at

25xGPV (gas pore volume); all other reservoir and productionparameters are varied according to the values in Table 2.

We find that recovery always decreases with increasingproduction rate, regardless of reservoir permeability (Fig. 4).However, the sensitivity of recovery to rate decreases as thekv/kh ratio decreases. Well abandonment is caused by water

breakthrough in all cases, except for the highest permeabilityconsidered (3000 mD), the lowest production rate (50

MMscf/d), and the lowest kv/khratio (0.01). Recovery ceasesowing to the BHP constraint in these cases. The sensitivity ofrecovery to rate increases with decreasing permeability, except

for the lowest permeability considered (5 mD).Decreasing the kv/khratio yields higher recovery for any

given production rate. Furthermore, as kv/kh decreases, thesensitivity of recovery to reservoir permeability decreases, asdoes the penalty of producing at high rate. When the kv/khratio is low (


5/11

SPE 107169 5

In this base case scenario, maximum recovery is alwaysachieved with the lowest production rate; producing at highrates to outrun the aquifer leads only to an increased risk ofearly water breakthrough, rather than increased recovery. The

risk of early breakthrough is reduced as vertical permeabilitydecreases, but the sensitivity of recovery to the kv/kh ratio is

high, particularly when kh is low. Consequently, acceleratingproduction in the presence of a large, bottom water aquiferdoes not increase overall recovery, and is associated with

significant risk of early water breakthrough if reservoirpermeability is poorly understood. This may change if the

aquifer size is different. We consider this issue in the nextsection.

Impact of Aquifer Strength on Gas Recovery and Aquifer

Behavior. The impact of production rate and reservoirpermeability on cresting behavior may change with aquifers of

varying strength, through the dynamics of reservoir pressuredecline interacting with different levels of water influx andpressure support. Variations in aquifer strength are accountedfor within material balance expressions of gas recovery (e.g.

[8]), but are not included in analytical descriptions of watercresting because these are based on the local balance of

viscous and gravity forces. We therefore use our simulationmodel to investigate the impact on recovery of varying aquiferstrength over the range listed in Table 2. All other simulation

parameters remain the same as in the previous section. Wepresent results for low (kh= 75 mD) and high (kh= 1400 mD)

values of reservoir permeability, and low (50 MMscf/d) andhigh (300 MMscf/d) production rates.

For small aquifers (1xGPV to 2.5xGPV) we find that thewell is abandoned at the BHP limit, except for the lowestreservoir permeability, highest kv/kh ratio, and highestproduction rate (Fig. 5). Water breakthrough is observed for

these cases. As aquifer size increases, recovery decreasesowing to earlier water breakthrough for cases with highproduction rate, low reservoir permeability and high kv/khratio. However, in cases with low production rate and highreservoir permeability, and in all cases with low kv/kh ratio,

recovery increases with increasing aquifer size, and waterbreakthrough does not occur, up to an aquifer size of 25xGPV.As aquifer size increases above this, recovery decreases withincreasing aquifer size, and the well is abandoned owing towater breakthrough. As kv/khdecreases, recovery becomes lesssensitive to rate and reservoir permeability. Increasingproduction rate generally decreases recovery, except when the

kv/khis low (0.01) and the aquifer is large (>50xGPV).These results suggest that, in cases where water cresting

is expected (high rate, low permeability, and high kv/khratio),increasing aquifer size leads to earlier water breakthrough andhence lower recovery. However, in cases where cresting isinhibited (low rate, high permeability, and most significantly,

low kv/khratio), recovery increases as aquifer size increases upto a maximum value. For aquifer sizes less than this, water

breakthrough does not occur; for aquifer sizes greater thanthis, recovery decreases with increasing aquifer size owing toearlier water breakthrough.

When cresting is suppressed, it might be expected that

reservoir behavior would be predicted by material balanceapproaches. However, an increase in recovery with increasing

aquifer strength is not predicted by material balance. It occursbecause aquifer support slows the rate of reservoir pressuredecline, allowing further reserves to be produced prior toabandonment at the BHP limit (e.g. Fig. 6). This effect has

been observed in Columbus Basin gas fields produced throughvertical wells, which are protected from water influx by the

A

B

Figure 4 A-C: Variation of gas recovery factor with productionrate. Different curve colors denote different values of reservoirpermeability (kh). Each plot corresponds to a different kv/khratio.

C


6/11

6 SPE 107169

inverse coning effect.8,28 The benefit of aquifer support hasalso been predicted in a simulation study for the Mahogany 20___

Sand in this basin, which considered a horizontal well of 7-inch diameter and similar ranges of reservoir parameters.36

Our results suggest that it is particularly important to

characterize aquifer size in reservoirs with low kh and highkv/khratio, because these conditions provide the greatest risk of

cresting leading to early water breakthrough at highproduction rates. If the kv/khratio is low (


7/11

SPE 107169 7

For high permeability cases, the sensitivity of recovery toproduction rate increases with decreasing gas columnthickness. For low permeability cases and high kv/khratio, rate

sensitivity decreases with decreasing gas column thickness,because water breakthrough occurs almost immediately for

gas column thickness of 20m (66 ft) or less. This suggests thatthere is a critical minimum gas column thickness, below which

gas recovery is severely compromised by very early waterproduction. Its value varies with production rate, reservoir kh

and kv/khratio. Our results suggest it is less than 10 m (33 ft)in reservoirs with high permeability (>75 mD) or low kv/kh

ratio (


8/11

8 SPE 107169

DiscussionWe have found that water cresting dominates aquifer responseand hence recovery in gas reservoirs with an active bottom-

water aquifer, when produced at high rates through horizontalwells, unless the kv/kh ratio is low. Cresting is controlled by

reservoir pressure drawdown, which in turn is a function ofproduction rate, permeability and aquifer strength. Anunderstanding of the interaction between these parameters isrequired to develop a reservoir management strategy which

avoids water production.In general, producing at high rates increases pressure

drawdown, leading to the formation of a water crest, earlierwater breakthrough and lower gas recovery. Consequently,producing at high rates to outrun the aquifer does not lead tohigher recovery in bottom-water drive gas reservoirs.

However, the risk of early water breakthrough associated withproducing at high rates decreases with increasing horizontal

permeability, and decreasing vertical permeability, becausereservoir drawdown is decreased for a given production rate,and water cresting is suppressed. The sensitivity of recovery

to kv/kh is high, and there is a significant production penaltyassociated with producing at high rates when vertical

permeability is underestimated, particularly if aquifer supportis strong.

We observe three recovery responses to varying aquiferstrength: (1) recovery can be enhanced as aquifer strength

increases, in cases where crest development is inhibited eitherby low production rate, high horizontal permeability, or low

vertical permeability; (2) when aquifer response is dominatedby cresting, increasing aquifer strength reduces recovery and alow production rate is recommended; and (3) acceleratingproduction rate can enhance recovery in a small number of

cases in which the aquifer is very strong and verticalpermeability is very low. This is the only situation where we

find that gas recovery is increased by producing horizontalwells at high rates to outrun a basal aquifer.

The significance of these observations must be

considered with respect to the level of uncertainty inherent toeach. Producing at a low rate appears to be favorable in terms

of ultimate gas recovered, since recovery can be enhanced ifthe aquifer strength is weaker than assumed, and the well isafforded greater protection from early water breakthrough ifthe aquifer is stronger and/or the kv/kh ratio is higher thaninitially thought. Accelerating production to target theconditions necessary to enhance recovery at higher rates

requires a good understanding of aquifer strength andpermeability anisotropy. However, these parameters oftenremain uncertain.

Despite this, there might be an economic benefit toproducing at high rates, which offsets the reduction in ultimate

gas recovery. We have integrated the results of the sensitivityanalysis in a series of contour plots, which represent thedifference in absolute recovery prior to well abandonmentbetween producing at 50 MMscf/d and 300 MMscf/d (Fig. 9).The responses are interpolated between simulation results toprovide a rapid assessment of the change in recovery thatresults from producing at high rate. Each plot provides an

assessment of when producing at high rate is not likely to bedetrimental to ultimate recovery. It can be seen that few cases

directly indicate any benefit in terms of ultimate recoveryfrom accelerating production. However, a number of casesshow less than 10% difference in recovery between 50MMscf/d and 300 MMscf/d. Producing at the high rate case in

these examples would return a similar recovery six timesfaster, which might be economically beneficial.

Gas fields in the Columbus Basin have exhibitedenhanced recovery owing to a weak component of waterdrive.

8However, most of the reported field examples utilized

vertical wells which, as discussed previously, benefit from an

inverse gas cone which protects the flowing interval andmaintains gas deliverability.28 Our results indicate that

pressure support can also increase recovery in fields developedwith horizontal wells, so long as cresting is not significant.

Documented experimental and real field examples whichindicate increased gas recovery through accelerated production

invariably employ vertical well completions.16,17,25,28

This is aresult of the inverse coning phenomenon described by

A

B

Figure 8 A-B: Variation of gas recovery factor with stand-offbetween the well and the GWC. Different curve colors denotedifferent values of reservoir permeability (kh); different curvesymbols denote different production rates. Each plotcorresponds to a different kv/khratio. Key as in Figure 5C.


9/11

SPE 107169 9

_______

Figure 9: Contour plots showing the

difference in absolute recovery betweenlow (50 MMscf/d) and high (300 MMscf/d)production rates.


10/11

10 SPE 107169

McMullan and Bassiouni (2000)28 (Fig. 2). The protectionafforded to vertical wells by this mechanism is not available tohorizontal completions. Consequently, although horizontal

wells can produce at higher rates than vertical completions,consideration should be given to post-water breakthrough

performance when water influx is inevitable. Of course, thefundamental assumption of the recovery forecasts we present

is that production ceases as soon as the water contact reachesthe well. We use this production constraint because water

breakthrough has such a negative impact on production thatavoiding water breakthrough is highly desirable. However, we

recognize that continued gas production in association withwater could enhance the benefit of accelerated production.

We have considered production via high-rate horizontalwells in a simple homogenous reservoir with a bottom-water

aquifer. Reservoirs with edge-water drive should not beassociated with significant cresting towards horizontal wells,

because the distance between the well and the GWC is muchlarger. Consequently, producing at high rates in reservoirswith edge-water drive may lead to improved recovery.

Moreover, a homogenous model cannot account for the impactof permeability heterogeneity on water cresting behavior. It is

uncertain to what extent reservoir architecture will modify ourfindings. Depositional heterogeneity can provide preferentialfluid flow pathways which might exacerbate water cresting,and baffles to flow which might benefit high-rate gasproduction by suppressing cresting or protecting the wellfrom the aquifer. Our results suggest that enhanced recovery

from a high rate horizontal gas well is most likely to occurbecause heterogeneity is suppressing cresting, rather thanbecause production is outrunning the aquifer. However, inorder to examine the influence of heterogeneity on gasrecovery and dynamic aquifer behavior, a detailed description

of heterogeneity is required.

ConclusionsOur results suggest that accelerated production via large-borehorizontal wells in gas reservoirs with a component of bottomwater drive does not increase the ultimate recovery of gas,because aquifer response is dominated by water cresting.

Consequently, there is no benefit to ultimate recovery inattempting to outrun a basal aquifer. Cresting initiated by

producing at high rates results in poor sweep efficiency andearly water breakthrough, unless the ratio of vertical tohorizontal permeability is very low, in which case watercresting is suppressed.

Rate sensitivity increases in low permeability reservoirswith thin gas columns, because these conditions increase the

tendency for water cresting, and decreases in reservoirs withstrong aquifer support, because water breakthrough occursregardless of the rate at which the well is produced. In a smallnumber of cases, with very strong aquifer support and very

low vertical permeability, gas recovery is increased byproducing at high rates. Water influx is inevitable, and gas

remains trapped at high pressure if the reservoir is produced atlow rate. However, the increase in recovery is small, and theassociated risk is high, as recovery will decrease at higherrates if the vertical permeability is higher than predicted.

Despite this, there are many instances where acceleratingproduction recovers only slightly less gas over much shorter

timescales, so may be economically favorable.The dynamics of water cresting are neglected in material

balance approaches, which may over-estimate the benefit of

accelerated production; moreover, gas production fromhorizontal wells is much more significantly affected by early

water breakthrough than in vertical wells, because verticalwells are protected from water production by an inverse gas

cone which forms in response to the high mobility of gas.

AcknowledgementsFunding for this work is gratefully acknowledged from the UK

Engineering and Physical Sciences Research Council(EPSRC) and BP Trinidad and Tobago (bpTT). The authorswould like to thank Schlumberger Geoquest for providing theEclipse 100 simulation package and Roxar for providing the

RMS

software.

NomenclatureBHP = bottom-hole pressure, psiaGPV = gaspore volume, cubic feet

GWC = gas-water contactkh = horizontal permeability, mD

kv/kh = ratio of vertical to horizontal permeabilitykrg = gas relative permeabilitykrw = water relative permeabilityPV = pore volume, cubic feetqg = gas flow rate, MMscf/dSg = gas saturation

Sgr = residual gas saturationx = distance inx-directiony = distance iny-directionz = distance inz-direction

References1. Baron, R. P. and Pearce, A. J., 1996, Understanding the

Performance of a Low-Permeability Gas Reservoir: Hyde Field,

Southern North Sea SPE Reservoir Engineering, August, p.210-214.

2. Teng, D., Nettleship, G., Hicking, S. and Hindmarsh, K., 1998,High Rate Gas Well Design: Issues and Solutions - Goodwyn

Gas Condensate, NWS, Australia SPE paper 50081 presentedat the SPE Asia Pacific Oil and Gas Conference and Exhibition,

12-14 October, Perth, Australia.3. Veeken, C. A. M., Chin, H. V., Ross, R. W. and Newell, M. D.,

2000, Monitoring and Control of Water Influx in StrongAquifer Drive Gas Fields Offshore Sarawak SPE paper 64402

presented at the SPE Asia Pacific Oil and Gas Conference and

Exhibition, Brisbane, Australia, Oct.16-18.4. Lumsden, P.J., Balgobin, C. J., Bodnar, D., Brayshaw, A. C.,

Dyer, B. L., Gainski, M., Hennington, E. R., Burch, T. K.,Farmer, C. L. and Saldana, M. A., 2002, The Kapok Field A

step change for Trinidad gas developments SPE paper 75670presented at the SPE Gas Technology Symposium, Calgary,

Canada, April 30 May 2.5. Kromah, M. J., Lumsden, P. J., Hennington, E. R. and

Brayshaw, A. C., 2003, Trinidads First 500 MMscfd well:

Fact or Fiction? SPE paper 81045 presented at the SPE LatinAmerica and Caribbean Petroleum Engineering Conference,

Port-of-Spain, Trinidad, West Indies, April 27-30.6. Pucknell, J., Holder, G., and Seesahai, T., 2003, Amherstia

Field, Trinidad: Breaking Barriers in the New Millennium SPE

paper 81012 presented at the SPE Latin America and Caribbean


11/11

SPE 107169 11

Petroleum Engineering Conference, Port-of-Spain, Trinidad,West Indies, April 27-30.

7. Hower, T. L. and Jones, R. E., 1991, Predicting Recovery ofGas Reservoirs Under Waterdrive Conditions SPE paper 22937

presented at the SPE Annual Technical Conference andExhibition, Dallas, Oct. 6-9.

8. Hallam, R. and Kin, M. L., 2003, Performance of Trinidad Gas

Reservoirs (Cassia, Immortelle, Flamboyant, Mahogany,Amherstia and Teak) SPE paper 81010 presented at the SPE

Latin America and Caribbean Petroleum EngineeringConference, Port-of-Spain, Trinidad, West Indies, April 27-30.

9. Turner, R. G., Hubbard, M. G. and Dukler, A. E., 1969,

Analysis and Prediction of Minimum Flow Rate for theContinuous Removal of Liquids from Gas Wells Journal of

Petroleum Technology, Nov., p. 1475-148210. Coleman, S.B., Hartley, B.C., McCurdy, D.G., and Norris III,

H.L., 1991, A New Look at Predicting Gas-Well Load-up,Journal of Petroleum Technology, March, p.329-338.

11. Sutton, R. P., Cox, S. A., Glynn Williams J., E., Stoltz, R. P. andGilbert, J. V., 2003, Gas Well Performance at Subcritical

Rates SPE paper 80887 presented at the SPE Production andOperations Symposium, 22-25 March, Oklahoma City,

Oklahoma.12. Gringarten, A. C., von Schroeter, T., Rolfsvaag, T. and Bruner,

J., 2003, Use of Downhole Permanent Pressure Gauge Data to

Diagnose Production Problems in a North Sea Horizontal WellSPE paper 84470 presented at the SPE Annual Technical

Conference and Exhibition, 5-8 October, Denver, Colorado.13. Bruns, J. R., Fetkovitch, M. J., and Meitzen, V. C., 1965, The

Effect of Water Influx on p/z-Cumulative Gas ProductionCurvesJournal of Petroleum Technology, March, 1965. p. 287-

29114. Agarwal, R.G., Al-Hussainy, R. and Ramey Jr., H.J., 1965, The

Importance of Water Influx in Gas Reservoir, Journal of

Petroleum Technology, November, p.1336-1342.15. Matthes, G., Jackson, R. F., Schler, S. and Marudiak, O. P.,

1972, Reservoir Evaluation and Deliverability Study, BierwangField, West Germany SPE paper 3736 presented at the SPEEuropean Spring Meeting, Amsterdam, May 16-18.

16. Lutes, J. L., Chiang, C. P., Rossen, R. H. and Brady, M. M.,1977, Accelerated Blowdown of a Strong Water-Drive Gas

Reservoir Journal of Petroleum Technology, December, p.1533-1538.

17. Brinkman, F. P., 1981, Increased Gas Recovery from aModerate Water Drive Reservoir Journal of Petroleum

Technology, December, p. 2475-2480.18. Al-Hashim, H. S. and Bass Jr., D. M., 1988, Effect of Aquifer

Size on the Performance of Partial Waterdrive Gas ReservoirsSPE Reservoir Engineering, May.

19. van Everdingen, A. F. and Hurst, W., 1949, The Application ofthe Laplace Transformation to Flow Problems in Reservoirs

Trans., AIME, 186,p. 305-32420. Hurst, W., 1958, The Simplification of the Material Balance

Formulas by the Laplace Transformation Trans., AIME, 213p.

21. Carter, R. D. and Tracy, G. W. (1960) An Improved Method forCalculating Water Influx. Trans., AIME, 219p. 415

22. Fetkovich, M. J., 1971, A Simplified Approach to Water InfluxCalculations Finite Aquifer Systems Journal of Petroleum

Technology, July, 1971, p. 814-828.

23. Trimble, A. E. and DeRose, W. E., 1977, Field Application ofWater-Coning Theory to Todhunters Lake Gas FieldJournal of

Petroleum Technology, May, p. 552-560.24. Pepperdine, L., 1981, The Determination of Sweep Efficiency

and Ultimate Gas Recovery from Water Driven Gas PoolsSimilar to Those Found in the Middle Devonian of North East

British Columbia Paper 81-32-48 presented at the 32nd

Annual

Technical Meeting of the Petroleum Society of CIM, Calgary,Alberta, Canada, May 3-6.

25. Kabir, C.S., 1983, Predicting Gas Well Performance ConingWater in Bottom-Water Drive Reservoirs, SPE paper 12068,

presented at the SPE Annual Technical Conference andExhibition, San Francisco, Oct 5-8.

26. Peng, C. P. and Yeh, N., 1995, Reservoir Engineering Aspects

of Horizontal Wells Application to Oil Reservoirs with Gas orWater Coning Problems SPE paper 29958 presented at the SPE

International Meeting on Petroleum Engineering, Beijing, PRChina, Nov. 14-17.

27. Ehlig-Economides, C. A., Chan, K. S. and Spath, J. B., 1996,

Production Enhancement Strategies for Strong Bottom WaterDrive Reservoirs SPE paper 36613 presented at the SPE

Annual Technical Conference and Exhibition, Denver,Colorado, Oct. 6-9.

28. McMullan, J.H., Bassiouni, Z., 2000, Optimization of Gas-Well Completion and Production Practices, SPE paper 58983,

presented at the SPEInternational Petroleum Conference andExhibition, Mexico, Feb 1-3.

29. Joshi, S.D., 1991, Horizontal Well Technology, PennWellPublishing Company, Tulsa.

30. Chaperon, I., 1986, Theoretical Study of Coning TowardHorizontal and Vertical Wells in Anisotropic Formations:Subcritical and Critical Rates, SPE paper 15377 presented at

the SPE Annual Technical Conference and Exhibition, NewOrleans, Oct. 5-8.

31. Giger, F. M., 1989, Analytical Two-Dimensional Models ofWater Cresting Before Breakthrough for Horizontal Wells SPE

Reservoir Engineering, November, pp. 409-416.32. Ozkan, E. and Raghavan, R., 1990, A Breakthrough Time

Correlation for Coning Toward Horizontal Wells SPE paper20964, presented at the SPE European Offshore PetroleumConference, The Hague, Netherlands, Oct. 22-24.

33. Papatzacos, P., Herring, T. R., Martinsen, R. and Skjaeveland,S. M., 1991, Cone Breakthrough Time for Horizontal Wells

SPE Reservoir Engineering, August, pp. 311-318.34. Cohen, M. F., 1989, Recovery Optimization in a Bottom/EdgeWater-Drive Gas Reservoir, Soehlingen Schneverdinge, SPE

paper 19068 presented at the SPE Gas Technology Symposium,Dallas, June 7-9.

35. Chierici, G. L., Ciucci, G. M. and Long, G., 1963,Experimental Research on Gas Saturation Behind the Water

Front in Gas Reservoirs Subjected to Water Drive paper 17,Proceedings of the World Petroleun Congress, Frankfurt am

Main, Germany, June 25. pp. 483-498.36. Ali-Nandalal, J. and Gunter, G. 2003., 2003, Characterising

Reservoir Performance for the Mahogany 20 Gas Sand Based onPetrophysical and Rock Typing Methods SPE paper 81048

presented at the SPE Latin America and Caribbean PetroleumEngineering Conference, Port-of-Spain, Trinidad, West Indies,

Apr 27-30.37. Dake L. P., 1978, Fundamentals of Reservoir Engineering,

Elsevier Scientific Publishing Company, New York, p.30.

SI Metric Conversion Factors

bbl 1.589 873 E-03 = m3

cp 1.0* E-03 = Pa s

ft 3.048* E-01 = m

ft3 2.831 685 E-02 = m3

md 9.869 233 E-04 = mm2

psi 6.894 757 E+00 = kPaF (F-32)/1.8 = C

controls of water cresting in horizontal well

Documents