+ All Categories
Home > Documents > SPE-14507-PA

SPE-14507-PA

Date post: 13-Apr-2018
Category:
Upload: jessica-king
View: 214 times
Download: 0 times
Share this document with a friend

of 7

Transcript
  • 7/26/2019 SPE-14507-PA

    1/7

    A Simplified Method for

    Determining

    Gas Well Deliverability

    Robert

    W Chase SPE Marietta C

    Ted

    M

    Anthony

    SPE, Marathon Oil Co

    Summary This paper presents a simplified method for predicting the performance of a gas well. A method for determining the

    deliverability

    of

    an u nfra ctu red gas well by use

    of

    a single-point flow test and a dimensionless Vogel-type inflow performance

    curve was proposed by Mishra and Caudle. Their procedure necessitates the calculation of real-gas pseudopressures for shut-in and

    flowing bottomhole pressures BHP) obtained from pressure-buildup and stabilized-flow tests, respectively. This paper offers a

    simplification

    of

    this technique in which a range

    of

    pressure values is defined over which pressure-squared terms can be substituted

    for pseudopressures. A comparison is made between results obtained from analysis of well-test data on several gas wells made with

    conventional multipoint test methods, with the Mishra-Caudle technique, and with the simplified method presented in this paper.

    The simplified method offers the engineer who might not have access to a pseudopressure computer program

    or

    pseudopressure

    tables a method for estimating gas-well deliverabilities. The method of Mishra and Caudle and the simplified method were both

    observed to yield slightly conservative estimates of gas-well deliverabilities compared with the deliverabilities calculated from

    multipoint flow-test analyses. The simplified technique was found to be useful for predicting the performance

    of

    fractured gas wells

    as well as unfractured wells.

    Introduction

    Predicting the performance of gas wells is a process that has relied

    almost exclusively on some form

    of

    multipoint well-testing proce

    dure. The conventional backpressure

    or

    flow-after-flow, 1 the

    isochronal,

    2

    and the modified isochronal tests

    3

    have been used to

    predict the short- and long-term stabilized deliverability

    of gas

    wells.

    In a typical multipoint deliverability test, a well is produced at

    a minimum

    offour

    different flow rates with shut-in periods

    ofv r-

    ious lengths separating flow periods. Pressure is monitored during

    both the flow and shut-in periods throughout the test. Analysis of

    the BHP vs. flow rate yields results that, when plotted on log-log

    paper as shown in Fig. 1, produce a straight line that reflects the

    stabilized deliverability behavior

    of

    a gas well.

    The empirically derived relationship given by Eq.

    1

    represents

    the equation of a stabilized deliverability curve such as the one

    shown in Fig.

    1.

    q=C p1-plt)n

    1

    The constant C reflects the position

    of

    the stabilized deliverabil

    ity curveon the log-log plot. The value of the exponent,

    n,

    is equal

    to the reciprocal

    of

    the slope

    of

    the stabilized deliverability curve

    a nd normally has a va lue bet wee n 0 .5 and 1.0 .

    The stabilized deliverability curve

    or

    its equation may be used

    to predict the ability

    of

    a well to produce against a given sandface

    backpressure. The absolute open flow AOF)

    of

    the well is also

    frequently calculated. The AOF is the theoretical maximum flow

    rate a well can maintain against a zero surface backpressure. The

    AOF is used mainly in comparing wells and by regulatory bodies

    in establishing production allowables.

    Multipoint backpressure tests yield very reliable deliverability

    projections when correctly conducted in the field. Frequently, how

    ever, these tests require a commitment of manpower, equipment,

    and time that may render the tests cost-prohibitive. This is particu

    larly true in the case oflow-permeability reservoirs, where testing

    times may be very long. The problem is further compounded in

    terms of lost revenues

    if

    gas must be flared throughout the test.

    Alternative methods for forecasting gas-well deliverability have

    been proposed by several authors.

    4

    -

    6

    A replot

    of

    the stabilized

    deliverability curve shown in Fig. o n Cartesian coordinate graph

    paper Fig. 2 produces an inflow performance,

    or

    IPR, curve simi

    lar to those observed in the testing

    of

    oil and gas producing wells.

    Russell et

    al.

    4 showed that IPR curves constructed with Eq.

    1

    gave

    predicted gas-production rates lower than those observed in the field.

    Copyright 1988 Society of Petroleum Engineers

    1090

    Russell

    et al.

    proposed an equation that depicted gas-inflow per

    formance more accurately:

    Tsckh p1-p;y.)

    q= 2

    re 3 )

    50.34PscTp. p z p n

    - -

    s

    r

    w

    4

    Greene

    5

    documented that Neely6 rewrote

    Eq.

    2 by collecting the

    parameters that were constant for a given well in a constant, C1

    yielding the gas-well inflow performance equation:

    c

    p1-p;y.)

    q= 3)

    p. p z p

    The constant C1in Eq. 3 may be determined from a single flow

    test

    if

    the shut-in BHP

    is

    known. The constant C l will not vary

    with flow rate; however, it may change over the life

    of

    the well

    because

    of

    changes in the producing condition

    of

    the wellbore

    or

    formation.

    Greene noted that a valid IPR curve could be constructed for a

    well from a single C1-factor determination and a known shut-in

    BHP. This could be done by assuming values

    of Pwf

    calculating

    corresponding p g and z values, and substituting in Eq. 3 to find

    corresponding values

    of

    BHP could then be plotted vs. flow rate

    to obtain the IPR curve.

    VogeJ7

    extensively studied the inflow performance

    of

    solution

    gas-drive reservoirs. He suggested that the dimensionless IPR curve

    shown in Fig . 3 could b e used to generate actual IPR c urves for

    wells in which oil and gas were flowing. With Vogel s method,

    only a value for shut-in BHP and a single flow-test point are nec

    essary to generate an IPR curve for a well completed in a solution

    gas-drive reservoir. Brown

    8

    reported that field experience has

    shown that Vogel s dimensionless IPR curve also yields good ap

    proximations of flow behavior when the method is applied to wells

    producing oil, gas, and water. Vogel suggested that dimensionless

    IPR curves could be constructed for wells producing only liquids

    or

    only gas, as shown in Fig. 3. Ho we ve r, he did not propo se an

    actual dimensionless IPR curve that could be used to predict gas

    well performance.

    Mishra and Caudle

    9

    presented a method for pre dic ting th e

    deliverability

    of

    a gas well in an unfractured reservoir. The ana

    lytic solution to the gas diffusivity equation for flow

    of

    a real gas

    SPE Reservoir Engineering, August 1988

  • 7/26/2019 SPE-14507-PA

    2/7

    / / t.... fron i nt J I

    I

    rrrli

    t ~ o ~ ; t l ~ P i C f I D

    ~ - - - - - - . . J - - - - - - - - - - - J - - - - 7 1 0 : - - - - J . - - - L . . J . - - - - J - - . . . J . . . . ~ 1 0 0

    q MMscflD

    10000

    i=-;;-;

    -379

    lOJ

    p o

    i

    -

    j

    - I--

    I

    j

    N

    I

    1-

    - t

    Stabili d

    T

    o

    ,;,a

    1000

    O.li

    ,ability

    I

    i

    I

    2

    lC

    N

    Go

    I

    I

    I

    I

    /

    I I '

    r- i - --i - , -I-H- H

    ~ ; ; : : :

    ? - . : - -

    Two-phase Gas

    Oil

    Flow

    =

    r

    ....

    oL . . .

    .......

    L.

    J

    0 .0

    p p 1.0

    wE

    Fig. 1-Stabllized dellverabllity curve for modified Isochronal

    test

    (after Ref. 11). Fig.

    3-Dimenslonless

    IPR curves (after VogeI

    7

    .

    2.00

    1.90

    1.80

    1.70

    1.60

    1.50

    1.40

    i

    1..30

    1.20

    1.10

    -

    1.00

    I

    0.90

    0.80

    0.70

    0.60

    0.50

    0.40

    0..30

    0.20

    0.10

    0.00

    0.00 2.00

    4 6

    FLOWRATE (MMet/D)

    8,00

    10,00

    12.00

    --------

    0.

    0 .

    .7

    0

    .

    ,

    0.5

    .4

    -

    O.J -

    .2

    \

    .1

    \

    0.2

    0.4 0 .

    0.

    qlq ,..

    Fig. 2-IPR curve generated from stabilized deliverabll lty

    curve.

    in a porous medium under stabilized or pseudo-steady-state flow

    conditions and a broad range

    of

    rock and fluid properties were used

    to generate dimensionless flow rate and pseudopressure data and

    to derive the following empirical relationship:

    q

    5 _

    = [1 5 p

    p

    >< P

    p

    R - I ] 4

    qrnax 4

    where

    pp r -dp 5

    o p z

    Eq. 4 is the empirical relationship representing the dimension

    less IPR curve shown in Fig. 4. This curve can be used to deter

    mine an actual IPR curve for calculating the current deliverability

    of an unfractured gas well.

    Mishra and Caudle proposed a second empirical equation and

    dimensionless IPR curve for calculating the future deliverability

    of

    an unfractured gas well:

    qrnaxf

    5 _

    = [1 0.4 PpRjPpRi] . (6)

    qrnaxi 3

    SPE Reservoir Engineering, August 1988

    Fig. 4-Dlmenslonless IPR curve for current conditions (af

    ter Mishra and Caudle

    9

    .

    Eq. 6 is the empirical relationship representing the dimensionless

    IPR curve shown in Fig. 5.

    Note that the average reservoir pressure, fiR was used in the

    development

    of

    the dimensionless IPR curves.

    In

    most testing proce

    dures, a well is shut in and BHP is allowed to build up until it ap

    proaches some level

    of

    stabilization. The value for the static BHP

    is normally used in lieu of the true average reservoir pressure, which

    can be determined only by transient-pressure analysis techniques.

    The approximation

    of

    true average reservoir pressure with static

    BHP will introduce some inaccuracy in the prediction of flow rates,

    probably on the conservative side.

    The dimensionless IPR curves developed by Mishra and Caudle

    eliminate the need for multipoint testing and provide a new and

    simpler technique for estimating gas-well deliverability. The method

    simplifies the process

    of

    deliverability testing by reducing the

    amount

    of

    well-test data required for analysis to a single shut-in

    BHP value and a single stabilized flow-test point. The analysis

    of

    the data acquired from a field test requires the calculation

    of

    real

    gas pseudopressures, and unfortunately, not all gas producers have

    the capability of making pseudopressure calculations.

    The purpose

    of

    this research is two-fold. First, a simplification

    of the Mishra-Caudle method is sought in which a range of pres

    sures

    is

    defined over which pseudopressure ratios can be approxi

    mated by pressure or pressure-squared terms. This simplification

    1091

  • 7/26/2019 SPE-14507-PA

    3/7

    0.9

    + - - I - - + - + - - I - - + - - - f - - ~ - + - - + L . . . . j /

    Fig. 5-Dlmenslonless PR curve for future conditions after

    Mishra and Caudle ).

    O.B t - - t - - t -- j - --- , f - - j - - - - -- f --- .L.J----1

    0.7 +--I--+--+---I--f--+--I---,..j::V:...-...-I----l

    O.B

    +--+--+--+--+-+--1--JL/=---f..--l---l

    /

    I 0.5 t - - - t - - t - - t - - l - - - - - j f - - - j .L-- - - - - - - - - l

    0.4 +--IC--+--f--I--::;j.-.:::...../+-f--+--J.-----1

    O.J

    +---+--+---1/----;;>- L/-+--+-+---+---Il--...l

    0.2 t - - - - - - ; 7 f - - f - - - - 1 - - - - - - l - - - I - - ~

    /

    0.'

    - t - - - - f :: ; , .L-+-- t - - l- - f - - - l --+-- - -+-- -+--

    /

    10

    PRESSURE SQUARED

    Ix

    10' p. . . . )

    1

    0.9

    0.

    0.7

    III

    = t

    0.6

    co_

    co

    i l i

    a:

    0 .

    9 ~

    0

    lll. l

    0.3

    O.Z

    0.1

    0

    0 Z

    6

    O.B

    .6.4

    q-rlci

    0.2

    1 .

    1

    1.3

    1.Z

    1.1

    1

    III

    0.9

    = t

    o_

    o.

    o

    l l i

    0. 7

    ~ ~

    0.6

    III

    ..

    0 .

    o-

    0

    0.3

    O.Z

    0.1

    0

    0

    Z

    PRESSURE

    x

    10'

    P .)

    Fig.

    -Curve of

    pseudopressure

    VS.

    pressure for a O.6-gravlty

    gas.

    wouldpermit an engineerwho may not have access to pseudopres

    sure tables or a pseudopressure computer program to estimate gas

    well deliverabilities.

    Secondly, the Mishra-Caudle dimensionless IPR curves were de

    veloped for real-gas flow in an unfractured reservoir.

    It

    was desired

    to determine whether the method would yield accurate results when

    applied to the testing

    of

    wells that had beenhydraulically

    fractured.

    udopr

    u

    pproximation

    It

    has

    been shown that gas flow behavior can

    be

    most accurately

    described using the pseudopressure function,

    Pp

    which takes into

    account the deviation of gas viscosity and the gas deviation factor

    with pressure. The pseudopressure function is givenby the integral

    expression shown in Eq. 5. .

    At

    low pressures, the product of viscosity and compressibility

    remains essentially constant, and the pseudopressure function can

    be

    approximated by

    Fig. 7-Curve of pseudopressure VS. pressure squared for a

    O.6-gravlty gas.

    where the combination

    of

    terms in parentheses also remains essen

    tially constant.

    As a general guideline, Wattenbarger and RameylO suggested

    that for pressures 3,000 psia [> 20.7 MPa], pres

    surewill accurately approximate pseudopressure. These guidelines

    were developed for application with the diffusivity equation in

    transient-pressure analysis.

    In

    the context

    of

    the modification to the Mishra-Caudle method

    sought in this work, the relationship between pseudopressure and

    pressure

    or

    pressure squared was also examined. A computer pro

    gram was used to generate a table

    of

    data comprising correspond

    ing values of pressure, pressure squared, and pseudopressure.

    Graphs of pressure squared vs. pseudopressure and pressure vs.

    pseudopressure were plotted for 0.60-, 0.65-, 0.70-, and

    0.75-gravity gases. Differences between the curves for different

    gravity gases were noted, but all basically followed similar trends.

    For

    this reason, only the results obtained for a O.60-gravity gas

    are reproduced in Figs. 6 and 7.

    The portions of the curves shown in Figs. 6 and 7 that

    are

    linear

    are

    of

    particular interest in this research. The curve

    of

    pseudopres

    sure vs. pressure in Fig. 6 was viewed to

    be

    nearly linear above

    about 2,900 psi [20 MPa]. This was confirmed by an incremental

    computation of slopes along the line at 50-psi [345-kPa] intervals.

    The curve of pseudopressure vs. pressure squared in Fig. 7 was

    also viewedto be nearly linearfrom 0 to about 2,000 psi [0 to 13.8

    MPa]. This trend was confirmed by calculating slope values at points

    50

    psi [345 kPa] apart along the curve. Over this pressure range,

    the change in slope between any two pressures was less than 1

    .

    The linear portion of the curve of pseudopressure vs. pressure

    squared is of great significance in terms of simplifyingdeliverabil

    ity testing procedures for wells with shut-in

    BHP s

    less than about

    2,000 psi [13.8 MPa]. Because the curve shown in Fig. 7 passes

    through the origin, the equation describing the linear portion of the

    curve can be written as the equation of a straight line having a

    y

    intercept of zero:

    Pp

    =

    _2_ p

    2

    . 7)

    LiZi

    pp=mp2 (9)

    Substituting this equation for

    Pp

    values as they are used in Fig.

    4 and Eq. 4 yields the following relationship:

    Likewise, at high pressures, the pseudopressure function can

    be

    approximated

    by

    pp= 2

    P

    i

    p

    2

    8

    LiZi

    Ppwf mpJi.

    = = 2

    10)

    PpR

    mpR

    A similar relat ionship can be obtained for Fig . 5 and Eq. 6 .

    1092

    SPE Reservoir Engineering, August 1988

  • 7/26/2019 SPE-14507-PA

    4/7

    TABLE

    1 MODIFIED

    ISOCHRONAL TEST DATA from Ref. 9

    Sandface

    Pressure Pseudo- Gas Flow

    Duration

    Pressure Squared

    pressure Rate

    hours) psia)

    psia

    2

    ) psi

    2

    /cp) McflD)

    Initial Shut-in

    20 1,948

    3 80x

    10

    6

    3 15x10

    8

    0

    Flow 1

    12

    1,784

    3 18x 10

    6

    2 67x10

    8

    4,500

    Shut-in

    12

    1,927 3.71 x

    10

    6

    3 08x 10

    8

    0

    Flow 2

    12

    1,680

    2 82x10

    6

    2.39 x 10

    8

    5,600

    Shut-in

    12

    1 911

    3 65x10

    6

    3 04x10

    8

    0

    Flow 3 12 1,546

    2 39x10

    6

    2.04 x 10

    8

    6,850

    Shut-in

    12 1,887

    3 56x 10

    6

    2.97 x 10

    8

    0

    Flow 4

    12 1,355

    1 84x

    10

    6

    1 58x

    10

    8

    8,250

    Extended flow

    81

    1,233

    1 52x

    10

    6

    1.32 x 10

    8

    8,000

    Final Shut-in

    120 1,948

    3.80 x 10

    6

    3 15x 10

    8

    0

    TABLE 2-COMPARISON OF AOF S

    FROM

    MULTIPOINT

    AND SINGLEPOINT TESTS

    Because the slope along the curve from 0 to 2,000 psi [0 to 13.8

    MPa] is essentially constant, it can readily be seen that the slope

    terms in Eq.

    1

    cancel, leaving a ratio

    of

    pressure-squared terms

    equal to a ratio

    of

    pseudopressure terms. This key simplification

    permits the substitution

    of

    pressure-squared values for pseudopres

    sures in the dimensionless IPR curves

    of

    Figs. 4 and 5 and in Eqs.

    4 and 6, for wells with average reservoir pressures

    or

    static BHP s

    less than about 2,000 psia [13.8 MPa].

    Turning to the curve

    of

    pseudopressure vs. pressure in Fig. 7,

    the linear portion

    ofthis

    curve begins at about 2,900 psi [20 MPa].

    This linear trend, however, does not have the origin as an inter

    cept, thus prohibiting the substitution of pressure values for pseu

    dopressures in the dimensionless IPR expressions. Therefore, it can

    be concluded that above about 2,000 psi [13.8 MPa], pseudopres

    sure values should be used in working with the dimensionless IPR

    expressions proposed by Mishra and Caudle.

    12.00.00 10.00

    .00

    Outflow

    Perfonaane. . Cur ve

    2.00 4.00

    P

    19ol8

    2.00

    -y- - - - - - - - - - - - - - - - - - - - - - - - - - - ,

    1.90

    1.80

    1.70

    1.60

    1.50

    1.40

    1.30

    1.20

    1.10

    _

    1.00

    cr

    0.90

    I 0.80

    0.70

    0.60

    0.50

    0.40

    0.30

    0.20

    0.10

    0.00

    - - - . - - - - - r - - - r - - - - - - - - r - - - - - - - - - - - r . . - ~ _ _ _ _ _ _ 1

    0.00

    Fig. 8 Production performance curves constructed for a gas

    weil at current conditions with the dimensionless IPR curve.

    FLOWRATE MMclID)

    dimensionless IPR curve used in the last two techniques being em

    pirically derived on the basis

    of

    curve-fitting a data trend. Mul

    tipoint well-test analysis depends on the use

    of

    actual test data to

    determine a unique stabilized deliverability curve and should there

    fore still be viewed as the most accurate method for calculating

    deliverability. Clearly, a large-scale evaluation

    of

    test data is needed

    to substantiate the use

    of

    single-point testing methods as a replace

    ment for multipoint test methods.

    The first three AOF s reported in Table 2 pertain to hydraulical

    ly stimulated wells completed in low-permeability reservoirs.

    It

    can

    be seen that the AOF s computed with the Mishra-Caudle method

    and the simplified method compare favorably with the AOF s found

    with conventionalmethods for the three fractured wells. Addition

    al comparisons are needed, however,

    to

    confirm the validity

    of

    using

    unfractured well dimensionless IPR curves for fractured wells. Fur

    ther research has been conducted by Chase and Williams

    12

    and by

    Chase

    13

    to develop dimensionless IPR curves specifically for pre

    dicting the performance

    of

    fractured gas wells.

    Predicting GasWe ellverabUlty

    With

    imensionless IPR urves

    The well-test data appearing in Table 1 were used in conjunction

    with the dimensionless IPR curve shown in Fig. 4 to construct the

    current deliverability IPR curve shown in Fig. 8. The procedure

    for constructing the IPR curve by the simplified pressure-squared

    method is detailed in the Appendix.

    The IPR curve is used to estimate the ability

    of

    a porous medium

    to flow gas into the wellbore when a given backpressure is main

    tained at the sandface. Because deliverability at the wellhead

    is

    nor

    mally desired, either an outflow performance curve or a tubing

    performance curve must be constructed. This performance

    prediction process is commonly called nodal analysis.

    Greene

    5

    gives several examples

    of

    the use

    of

    tubing-pressure

    curves in conjunction with IPR curves to analyze gas-well perform

    ance. The tubing-head pressure or outflow performance curve is

    148

    491

    692

    8,400

    8,940

    10 341

    Pressure-Squared

    AOF

    MscflD)

    149

    492

    692

    8,700

    9,080

    10,526

    Mishra-Caudle

    AOF

    MscflD)

    167

    547

    696

    9,000

    9,500

    10,988

    Multipoint Test

    AOF

    MscflD)

    iscussion of Results

    Field-test data were obtained on several gas wells with shut-in aver

    age reservoir pressures

  • 7/26/2019 SPE-14507-PA

    5/7

    a plotof tubing-headpressure vs. flow rate,

    n

    represents the ability

    of

    the gas and l iquid,

    if

    present, to flow up the tubing.

    multiple phases

    are

    flowing up the tubing, multiphase flow

    correlations are necessary to generate outflow performancecurves.

    These outflow performance curves normally exhibit a humpbacked

    shape.

    The

    apex of such a curve is designated as the flow point

    and represents themaximum flowing tubing-head pressure and the

    minimum sustainable flow rate possible for the well.

    Any flowing

    BHP

    algorithm may

    be

    used to generate the out

    flow performance curve fo r a dry gas well. A dry gas well will

    have no flow point

    or

    apex

    on

    its outflow performance curve.

    Outflow performance curves are valid only for the current static

    BHP

    and

    IPR

    curve

    of

    a well.

    To

    predict future deliverability, fu

    ture

    IPR

    curves must be constructed using Fig. 5 or Eq. 6 Corre

    sponding outflow performance curves must then be calculated for

    each future IPR curve.

    A second type of curve, called the tubing performance curve

    or

    constant tubinghead pressure curve, can also be constructed with

    eithermultiphase flow correlations when multiple phases are flowing

    or single-phase flow correlations for dry gas flow. The tubing per

    formance curve is a plot of the flowing BHP required to produce

    a well at various gas flow rates through a given-size tubing string

    at a constant flowing tubinghead pressure. This type of curve is

    also shown in Fig. 8.

    The

    intersection

    of

    the constant tubinghead

    pressure curve with the

    IPR

    curve yields the stabilized flow rate

    that the well can maintain against the specified constant tubinghead

    pressure.

    As static BHP falls, it is necessary to constructfuture IPR curves

    corresponding to the new static

    BHP s.

    The constant tubinghead

    pressure curve, however, applies throughout the flowing life of the

    well and need not be recalculated unless tubing size is changed.

    onclusions

    A simplification

    of

    a method developed by Mishra and Caudle for

    predicting gas-well performancewas proposed in this research. The

    procedure requires a value for the shut-in BHP and theflowing BHP

    and associated flow rate from a single stabilized flow test for a gas

    well. These data

    are

    used to construct IPR curves for the gas well

    by use

    of

    the dimensionless

    IPR

    curves developed by Mishra and

    Caudle.

    1

    Pressure-squaredvalues can be substituted for pseudopressutes

    in the dimensionless IPR graphs and equations developed by Mis

    hra and Caudle

    if

    the average reservoir pressure

    or

    static BHP for

    a gas well is less than about 2 ,000 psi [13.8 MPa].

    2. Above average reservoir pressures of about 2,000 psi [13.8

    MPa], pseudopressures must be used in the process of construct

    ing

    IPR

    curves from the dimensionless plots

    of

    Mishra and Caudle.

    3.

    The

    simplifiedmethod and the Mishra-Caudle technique both

    yield reasonable, conservative estimates

    of

    gas-well deliverabili

    ties compared with values obtained with conventional multipoint

    test methods.

    4.

    The

    simplified method and theMishra-Caudle technique were

    used successfully to predict the deliverabilities

    of

    several fractured

    gas wells completed in low-permeability reservoirs, even though

    the dimensionless IPR curves were developed for unfractured wells.

    5. A large-scale study of well-test data is needed to determine

    whether single-point testing methods can be used as an effective

    replacement for conventional deliverability test methods.

    6. Further research

    on

    the development

    of

    dimensionless IPR

    curves for fractured wells is warranted and ongoing.

    Nomenclature

    C

    constant

    of

    stabilized deliverability equation

    C

    constant

    of

    gas inflow performance equation

    h

    reservoir thickness, ft [m]

    k

    permeability, md

    m

    slope

    n

    exponent

    of

    stabilized gas deliverability equation

    P

    average pressure,

    [ PR+PMf /2],

    psia [kPa]

    p real gas pseudopressure evaluated at pressure

    psia

    2

    /cp [kPa

    2

    /rnPa s]

    PR

    average reservoir pressure, psia [kPa]

    1094

    Psc pressure at standard conditions, psia [kPa]

    Pth tubinghead pressure, psi [kPa]

    f

    flowing BHP, psia [kPa]

    Pws static BHP, psia [kPa]

    q stabilized gas flow rate, Mcf/D [m

    3

    /d]

    r

    e

    radius of external boundary, ft [m]

    r

    w

    radius of wellbore, ft [m]

    s skin factor

    T reservoir temperature, OR

    ]

    T

    sc

    temperature at standard conditions,

    OR

    [K]

    z gas deviation factor

    J g gas viscosity, cp [mPa s]

    Subscripts

    f future conditions

    i

    initial conditions

    P

    pseudo

    cknowledgements

    We

    th nk

    Paul Hyde (Columbia Natural Resources Inc.), Will Tank

    (Marathon Oil Co.), and Jim Murtha (Marietta

    C.)

    for their input

    during the course of this research.

    References

    1

    Rawlins, E.K. and Schellhardt, M.A.: Back-Pressure Dataon Natural

    GasWells and theirApplication to Production Practices, Monograph

    7, U.S . Bur. Mines (1936).

    2. Cullender, M.H.: The Isochronal PerfonnanceMethod ofDetennining

    Flow Characteristics of Gas Wells, Trans AIME (1955) 204, 137-42.

    3.

    Katz, D.L. et al : Handbook ofNatural

    as

    Engineering McGraw Hill

    Book Co. Inc ., New York City (1959) 448.

    4.

    Russell, D.G. et aI.: ~ t h o d s for Predicti1JB Gas Well Perfonnance,

    lP T (Jan. 1966) 99-108, Trans AIME,

    237.

    5.

    Greene, W.R.: Analyzing the Performance of Gas

    Wells, lPT uly

    1983) 1378-84.

    6 Neely, A.B.: The Effect of Compressor InstaIlation On GasWell Per

    fonnance,

    HAP

    Report 65-1, Shell OilCo., Houston (Jan. 1965) 1-13.

    7.

    Vogel, J.L.: Inflow Perfonnanc Relationships for Solution-Gas Drive

    Wells,

    lP T (Jan. 1968) 83-92;

    Trans

    AIME, 243.

    8.

    Brown, K.E.: The Technology ofArtificialL ift Methods PPC Books,

    Tulsa (1977) 13-14.

    9 Mishra, S. and Caudle, B.H.:

    A

    Simplified Procedure forGas Deliver

    ability Calculations Using Dimensionless

    IPR

    Curves, paper SPE

    13231 presented at the 1984 SPE Annual Technical Conference and

    Exhibition, Houston, Sept. 16-19.

    10. Wattenbarger, R.A. and Ramey, H.J. Jr.: Gas Well TestingWith Tur-

    bulence, Damage and Wellbore

    Storage,

    lP T (Aug. 1968)

    877-87;

    Trans AIME, 243.

    i 1

    Theory and Practice of the Testing of asWells Energy Resources

    Conservation Board, Calgary (1975) 3-25.

    12. Chase, R.W. and Williams,

    M.A.T.:

    Dimensionless IPR Curves for

    Predicting the Perfonnance of Fractured Gas Wells, paper SPE 15936

    presented at the 1986 SPE Eastern Regional Meeting, Columbus, Nov.

    12-14.

    13. Chase, R.W.: DimensionlessIPR Curves for PredictingGasWell Per

    formance, paper SPE 17062 presented

    at

    the 1987SPE

    Eastern

    Region

    al Meeting, Pittsburgh, Oct.

    21-23.

    ppendix Comparison of

    Performance

    Prediction Methods

    The modified isochronal well-test data given in Table 1 were used

    in conjunction with the dimensionless IPR curve shown in Fig. 4

    to construct the current deliverability IPR curve shown in Fig. 8.

    Specifically, the static BHP

    of

    1,948 psia [13.4MPa] and the flow

    ing BHP

    of

    1,233 psia [8.5 MPa] associated with the extended

    or

    stabilized flow rate of 8,000 Mcf/D [227 x 10

    3

    m

    3

    /d] were used.

    Because the static BHP in this well was less than 2,000 psia [13.8

    MPa], pressure-squared values were substituted for pseudopressures

    in the analysis.

    Eq. 4 was used to construct the

    IPR

    curve shown in Fig. 8. Sub

    stituting 8,000 Mcf/D [227x 10

    3

    m

    3

    /d] for

    q

    and the squares

    of

    PMfequal to 1,233 psia [8.5 MPa]

    andPR

    equal to 1,948psia [13.4

    MPa] for their respective pseudopressures, a value

    of

    qmax

    =

    10,341 Mcf/D [293 x

    1

    3

    mJ/d] was obtained. This compares with

    SPE Reservoir Engineering, August 1988

  • 7/26/2019 SPE-14507-PA

    6/7

    TABLE A 1 DATA USED FOR CONSTRUCTION OF

    IPR, OUTFLOW PERFORMANCE, AND CONSTANT

    TUBINGHEAD PRESSURE CURVES

    Fig. A1 Production performance curves constructed

    for

    a

    gas well at future condit ions with

    the

    dimensionless IPR

    curve.

    12 00

    J utu re TPI. CUl Yl

    4 00 6 00 8 00 10 00

    FLOW RATE

    (MMcI'fD)

    2 00

    2 00

    .

    1 90

    :: : J : : : = . . . . . r - - - + - - - - - - O : C o n I w > I ~ 1 h . 5 0 0 J * g

    1 60

    1 50

    1 40

    1 30

    1 20

    1 10

    1 00

    0 90

    0 80

    0 70

    0 60

    0 50

    0 40

    0 30

    0 20

    0 10

    0 00

    -- , - - - , - - - , - - , - - , - - - - . - - , - - - - . - - - - - r - - -- , - - - - - - - r - - i

    0 00

    1,725

    1,739

    1 781

    1,823

    1,858

    1,884

    Constant

    tt

    =1,500 psig

    Curve Data

    (psia)

    1,679

    1 581

    1,364

    1,128

    869

    556

    334

    136

    27

    Outflow

    Performance

    Curve Data

    (psia)

    Calculated

    Current Gas

    Flow Rate

    Mcf/D

    o

    973

    2,709

    5,269

    6,989

    8,164

    8,975

    9 281

    9,414

    9,439

    9,535

    9,914

    10,159

    10,297

    1 341

    Flowing

    BHP

    (psia)

    1,948

    1,900

    1,800

    1,600

    1,400

    1,200

    1,000

    900

    850

    840

    800

    600

    400

    200

    o

    a value of 10,988 McflD [311 x 10

    3

    m

    3

    d

    obtained for the AOF

    of

    the well from the modified isochronal test analysis and an AOF

    of

    10,526 McflD [298 x

    10

    3

    m

    d found with pseudopressures.

    Additional data calculated for constructing the IPR curve are shown

    in Table A I

    The second step in predicting the performance

    of

    a gas well re

    quires a knowledge

    of

    flow behavior in the tubing. From Fig. 8,

    corresponding values

    of

    flow rate and flowing BHP were read from

    the IPR curve. Tubingheadpressure values were then computed

    for each flow rate and BHP for a 0.605-gravity gas producing

    through 5,000 ft [1525 m

    of

    2 -in. [6-cm tubing. These data ap

    pear in Table A-I and are plotted in Fig. 8 as the outflow perform

    ance

    or

    tubinghead pressure curve. Note that this curve, like the

    IPR curve, is valid only at the current static BHP conditions.

    It

    is now desired to estimate the current deliverability

    of

    this well

    when producing against a backpressure

    of

    1,500 psig [10.3 MPa

    at the wellhead. Entering Fig. 8 at a tubinghead pressure

    of

    1,514.7

    psia [10.4 MPa , proceeding horizontally to intersect the outflow

    performance curve and then vertically downward to the x axis, a

    stabilized deliverability of about 3,600 McflD [102 x 10

    3

    m

    d

    is

    determined. The expected flowing BHP for this well at these produc

    ing conditions can be determined by proceeding vertically upward

    to intersect the IPR curve, and horizontally across to the y axis to

    read a flowing BHP

    of

    1,750 psia [12.1 MPa .

    An alternative method for determining deliverability

    of

    this well

    at a constant tubinghead pressure involves selecting flow rates ar-

    bitrarily and calculating BHP for a constant tubinghead pressure

    of

    1,500 psig [10.3 MPa . This procedure was used, and the resul

    tant data also appear in Table A I The corresponding constant

    tubinghead pressure curve is plotted in Fig. 8. The intersection

    of

    this curve with the IPR curve yields a stabilized deliverability

    of

    3,600

    Mcf

    [102 x 10

    3

    m

    3

    /d at a flowing BHP

    of

    approximate

    ly 1,750 psia [12.1 MPa .

    Unlike the outflow performance curve, the constant tubinghead

    pressure curve

    is

    valid for future IPR curves constructed for declin

    ing static average reservoir pressures as long as the flow-string con

    figuration in the well remains the same.

    For

    example,

    if

    the static

    reservoir pressure in the well

    of

    Table 1 declines to 1,800 psia [12.4

    MPa , a new IPR curve must be found. Substituting values for the

    squares

    of

    1,800 and 1,948 psia [12.4 and 13.4 MPa], respective

    ly, in

    PyRflppRi ,

    and 10,341 Mcf/D [293 x

    10

    3

    m

    3

    /d for

    qm xi

    in Eq. 6 or Fig. 5, a value for the new AOF or qmaxf=9,353

    McflD [265x 10

    3

    m

    3

    /d is found. The value for

    qmaxf

    and the new

    static reservoir pressure

    of

    1,800 psia [12.4 MPa are used in con

    junctionwith Fig. 4

    or

    Eq. 4 to construct the new IPR curve shown

    in Fig.

    A I

    Data for this curve appear in Table A-2.

    It

    can be seen from Fig. A-I that if a 1 5OO-psig [1O.3-MPa

    tubinghead pressure is maintained on this well, the stabilized deliver

    ability

    of

    the well will fall from 3,600 to 1,600 McflD [102 to

    45 x 10

    3

    m

    3

    /

    d

    as static reservoir pressure falls from 1,948 to

    1,800 psia [13.4 to 12.4 MPa .

    TABLE A 2 DATA USED FOR CONSTRUCTION

    OF

    FUTURE PERFORMANCE CURVES

    TABLE

    A a DATA

    FOR CONSTRUCTION OF

    PERFORMANCE CURVES FROM MODiFIED ISOCHRONAL

    TEST DATA

    Constant

    tt = 1,500 psig

    Curve Data

    (psia)

    Initial static reservoir pressure, psia

    Future static reservoir pressure, psia

    Initial j\OF rate, McflD

    Future AOF rate,

    Mcf/D

    1,823

    1,797

    1,775

    1,756

    1,742

    1,732

    1,725

    1,724

    Flowing

    BHP

    (psia)

    1,800

    1,700

    1,600

    1,400

    1,200

    1,000

    800

    600

    400

    200

    o

    Gas

    Production

    Rate

    Mcf/D

    o

    1,866

    3 351

    5 5 1

    6,910

    7,849

    8,478

    8,895

    9,159

    9,306

    9,353

    1,948

    1,800

    1 341

    9,353

    1,723

    1,730

    1,747

    1,785

    1 821

    Gas

    Flow Rate

    (McflD)

    10,988

    10,000

    9,000

    8,000

    7,000

    6,000

    5,000

    4,000

    3,000

    2,000

    1,000

    5

    o

    P

    ws = 1,948 psia

    P

    ws = 1,800 psia Constant

    Flowing Flowing Pth = 1,500 psig

    BHP BHP Curve Data

    (psia) (psia) (psia)

    o 0

    743 0

    1,036 721

    1,249 1,002

    1,416 1,204

    1,553 1,362

    1,665 1,489

    1,758 1,592

    1,833 1,675

    1,890 1,737

    1,930 1,780

    1,942 1,794

    1,948 1,800

    SPE Reservoir Engineering, August 1988 1095

  • 7/26/2019 SPE-14507-PA

    7/7

    Fig.

    A Productlon

    performance curves constructed for a

    gas well at current and future conditions with the stabilized

    dellverablllty curve.

    5 etrl onveslon actors

    cp

    x

    1.0* E 03

    Pa s

    ft3 x 2.831 685 E 0 2 m

    3

    psi

    x

    6.894 757

    E OO

    kPa

    Eq. A-I was used to compute the data in Table

    A 3

    and con

    struct the IPR curves in Fig. A-2 for static reservoir pressures

    of

    1,948 and 1,800 psia [13.4 and 12.4 MPa]. Plotting the 1,500-psig

    [10.3-MPa] constant tubinghead pressure curve yields expected

    stabilized deliverabilities

    of

    4,000 and 2,000 Mcf/D [113 and

    57 x 10

    3

    m

    3

    /d] for the respectiveR values.

    It

    can

    be

    seen that the values for deliverabilities obtained with

    the modified isochronal test analysis were somewhat higher than

    the values obtained with the dimensionless IPR curves. One rea

    son for the differences may be that pressure-squared values were

    used instead

    of

    pseudopressures in calculating flow rates from the

    dimensionless IPR curves.

    12.000.00.00 6.00 8.00

    FLOW RATE

    MWctIOt

    2.00

    2.00

    1.10

    1.80

    1.70

    1.60

    1.50

    1.40

    1.30

    1 2 0

    1.10

    1:00

    5

    f

    0.70

    0.60

    0.50

    0.40

    0.30

    0.20

    0.10

    0.00

    - - - - - - - . .. . - - - - - . . . . - - - - - .- - r - - - - - - - - - . - - - l

    0.00

    For comparative purposes, the equation for the stabilized deliver

    ability curve resulting from the analysis

    of

    themodified isochronal

    test data in Table 1 was determined to

    be

    Converslon factor Is exact. SPERE

    q=0 00124 p

    R2

    PMf2 O.60.

    .

    A-I

    Original

    SPE

    manuscript received for review

    Nov. 6, 1985.

    Paper accepted for publication

    March

    16, 1987.

    Revised manuscript received

    Feb. 11, 1988.

    Paper

    SPE

    14507 first

    presented at the 1985 SPE Eastern Regional Meeting held

    In

    Morgantown WV, Nov. 6-8.

    1096

    SPE Reservoir Engineering, August 1988


Recommended