Está en la página 1de 5

A New Look at Predicting

Gas-Well Load-Up
Steve B. Coleman, SPE, consultant; Hartley B. Clay, SPE, and
David G. McCurdy, SPE, Exxon Co. U.S.A.; and H. Lee Norris III,
SPE, Exxon Production Research Co.

SPE 20280
Summary. This paper discusses Introduction Wellbore Liquid Sources
results of field tests conducted to veri- As natural gas is produced from depletion- Before examining the wellbore-liquid-Ioad-
fy minimum flow rate (critical rate) re- drive reservoirs, the energy available to ing mechanism, we must first consider the
transport the produced fluids to the surface source of the liquids. There are two obvious
quired to keep low-pressure gas declines. This transport energy eventually sources: liquids condensed from the gas ow-
wells unloaded and compares results becomes low enough that flow rates are re- ing to wellbore heat loss and free liquids
to previous work. This paper also duced and fluids produced with the gas are produced into the wellbore with the gas.
no longer carried to the surface but are held Both liquid hydrocarbons and water may be
covers liquid yield effects, liquid
up in the wellbore. These liquids accumulate present, depending on the specific reservoir.
sources, verification that wellhead in the wellbore over time, and cause addi- In examining these sources, one might
conditions control onset of load-up, tional hydrostatic backpressure on the reser- tend to minimize the impact of condensed
and effects of temperature, gas/liquid voir, which results in continued reduction water, particularly at low reservoir pres-
of the available transport energy. In most sures. Because the gas is saturated with
gravities, well bore diameter, and cases, if this condition is allowed to con- water at reservoir conditions, a plot like Fig.
packer/tubing setting depth. tinue, the wellbore will accumulate sufficient 2 can be constructed to show the impact of
fluids to balance the available reservoir ener- condensed water for a typical8,OOO-ft, low-
gy completely and cause the well to die. This pressure gas well. As shown, the amount of
phenomenon is known as gas-well load-up. water condensed increases exponentially as
As Fig. 1 shows, load-up can easily be rec- the static reservoir pressure declines. This
ognized on a typical gas-well L-lO chart by is unfortunate because, as reservoir pres-
the characteristic exponential rate decline sures decline, the amount of load fluid re-
caused by accumulating wellbore liquids. quired to balance the reservoir hydrostati-
Numerous papers have offered methods cally and to kill a well also declines, com-
for predicting and controlling the onset of pounding the problem.
load-up.1-6 Turner et al.'s 1 method for Other problems may also occur as a result
predicting when gas-well load-up will occur of gas-well load-up. The near-wellbore
is most widely used. They compared two region of the reservoir may begin to become
physical models for transporting fluids up saturated with liquids, causing the relative
vertical conduits: liquid film movement permeability to gas to decrease, further
along the pipe walls and liquid droplets en- reducing the well's potential to remain
trained in the high-velocity gas core. A com- productive. Also, condensed water can be
parison of these two models with field test damaging to formations containing swelling
data yielded the conclusion that the onset of clays because it is low in total chlorides
load-up could be predicted adequately with «500 ppm).
an equation developed from liquid droplet
theory (Stokes law), but that a 20% upward Critical· Rate Theory-
adjustment of the equation was necessary. Llquld.Droplet Model
Turner et al. also suggested that in most in- As Turner et al. showed, a free-falling par-
stances wellhead conditions controlled the ticle in a fluid medium will reach a terminal
onset of liquid load-up and that liquid/gas velocity that is a function of the particle size,
ratios in the range of 1 to 130 bbllMMscf shape, and density and of the fluid-medium
did not influence the minimum lift velocity. density and viscosity. Applying this concept
Examination of Turner et al.'s published to liquid droplets in a flowing column of gas,
data indicates that most of the wells used in we can calculate the terminal velocity, vI'
the comparison had wellhead flowing pres- of the drop using
sures (WHFP's) above 500 psi. Because v t = 1.912{[q'4 (PL -Pg) 'A ]/P g 'h}, . (1)
gas-well load-up problems generally worsen
with continued decline in reservoir energy, which assumes a fixed droplet shape, size,
this paper focuses on wells with lower reser- and drag coefficient and includes the +20 %
voir pressures that are experiencing liquid adjustment suggested by Turner et al.
load-up and have WHFP's below 500 psi. Applying this terminal velocity equation
to wellbores and correcting to standard con-
Copyright 1991 Society of Petroleum Engineers ditions, we can determine the minimum gas

JPT • March 1991 329


"Wellbores with
concentric tubing
strings that terminate
some distance above
the completion Interval
will generally not
follow a qc calculation
based on flowing
wellhead conditions."
Fig. 1-Gas-well L-10 chart.

flow rate, qc' for the continuous removal wellhead, historical well test records were Next, as Table 1 shows, the liquid/gas ra-
of liquids from a wellbore: used to obtain WHFP's. tio of the wells examined ranged from 1 to
Table 1 presents a summary of the data 22.5 bbl/MMscf. This had no influence in
qc =3.06pvt AITz. . .............. (2) collected from both the critical-rate tests and determining the onset of load-up. In other
This rate, also known as the critical rate, production-chart data base. Fig. 4 plots these words, regardless of the amount of liquid
can be used to predict the onset of gas-well same data, comparing the observed and cal- present, the onset of load-up will occur at
load-up. Fig. 3 plots this critical rate for culated critical rates. A first examination of the same gas flow rate, all other factors
typical 2Ys-in. tubing with water as the the plot shows that the unadjusted liquid- being equal. This is consistent with Turner
droplet fluid. droplet model tends to offer a better match et at. 's findings.
of the field data for these low-pressure wells. Next, a comparison of the actual water
Comparison to Field Data Thus, the terminal-velocity equation (Eq. production rates with calculations for
1) can be rewritten without the 20% ad- condensed-water volumes indicates that the
To verify this theory and Turner et at. 's con- primary source of load fluid was condensed
justment:
clusions for wells experiencing liquid load- water in most cases.
up with WHFP's below 500 psi, two sources v t = 1.593{[u \4 (PL -Pg) \4]/ Pg y, }. . (3) Finally, for several of the critical-rate
of field data were collected and analyzed. A closer examination of the plot and raw tests, the wells were produced to a three-
First, 17 tests were conducted to deter- data yields several other valuable observa- phase separator to measure gas and liquid
mine the rate and WHFP at which load-up tions and explains some of the anomalies. production. In these tests, liquid production
would begin. These tests examined wells First, the data examined support the as- stopped completely once the well began to
that would consistently flow at a stable rate sumption that in most cases wellhead condi- exhibit load-up behavior. This verified that
above their calculated qc' For the test, tions control the onset of load-up. This is droplet terminal velocity had been reached
gradual, stepwise increases in WHFP were consistent with Turner et at. 's observations. and produced liquids were being held up in
introduced until the wells exhibited the typi- Second, in several of the critical-rate tests, the wellbore. Additional information regard-
cal exponential rate decline denoting the on- a relatively large step was inadvertently im- ing wellbore liquid-holdup behavior is
set of load-up. WHFP and gas-rate chart posed when the WHFP was raised to force provided in Part 2 of this series. 8
recorders were used to document the test and the well into load-up. In doing so, the exact
to provide a time history of the variables. point of load-up was overshot. This helps Influence of Important Variables
Fig. 1 shows a typical test chart. explain why some of these data points (solid This examination of critical-rate theory
The second source of data consisted of circles) plot below the theoretical curve. would not be complete without a discussion
numerous gas-well 8-day L-lO production Third, it was observed that a number of the of the influence of different variables on the
charts examined for similar rate-decline ten- data points (open points), which lie signifi- accuracy of the calculation.
dencies. In each case, an interpretation of cantly below the theoretical line, exhibited liq- Up to this point, the discussion of qc has
the rate and WHFP was made for the point uid-slugging behavior denoted by a ragged been directed toward using water as the load
that indicated the onset ofload-up. Because trace on the L-IO chart. These wells should fluid. This is a valid assumption in most
the production charts for these wells were be not expected to fit the liquid-droplet mod- cases. However, some situations do exist for
generally located some distance from the el theory and are considered anomalies. rich-gas reservoirs where the wellbore con-

330 March 1991 • JPT


ditions cause condensation of the heavy ,.
hydrocarbons, but not water. This was ob-
Source: Mck_tta-W.h.
served in several wells examined as a part Baal I; 0.' Sp. Gr. Ga.
of this work. Turner et at. noted that in these 25
200·r R• .." rlmp.
SO Pwlg W.llhead P,.. ..
situations the density and surface tension of 90-r W.llhead Tlmp.
the liquids should be adjusted accordingly ...
CJ
VI

in calculations of qc' In most cases, ifboth ...


Z
CJ 2.
Z
water and liquid hydrocarbon are present, a
u
the denser water controls the onset ofload- ...'"
up. Examination of other variables, such as
temperature, gas gravity, and interfacial ten-
~
~
...a ..
sion (1FT), indicates that in most cases they ...u
have only a minor influence on the accura-
cy of the critical-rate calculation.
,.,.
VI

;;;.
,.
-'
The wellbore cross-sectional area obvi- <D
<D

ously is one of the most important variables


in the critical-rate calculation. Fig. 5 shows
a comparison of various wellbore diameters.
Smaller-diameter wellbores have better
droplet lift efficiencies because of increased
transport-gas velocities. This direct relation-
I •• 2 •• ,
.+---------~--------~--------~--------~--------_r--------~
• I •• ..
STATIC RESERVOIR PRESSURE, PSIA
••• •••
ship leads to a very important observation.
Wellbores with concentric tubing strings that Fig. 2-Condensed·water volume: typical 8,000·ft well. 7
terminate some distance above the comple-
tion interval will generally not follow a qc
calculation based on flowing wellhead con- .... ,T-------~-------- __ --~-- __ ~------------ __------__ --~ __,
ditions. Because of the significant impact of .. -.-- .. -..... -- .... ...........•.
~

diameter, the qc for these situations should


be calculated with the flowing conditions of
the largest-diameter segment of the well- :---
bore, rather than the wellhead conditions. NON LOAD-UP
Setting the tubing/packer of the concentric REGION

tubing string a significant distance above the


completion interval can lead to premature
liquid-loading problems.
This may seem somewhat inconsistent
with the theory that condensing liquids are
a primary source of load fluids, because
these fluids are generally present only in the
upper portion of the wellbore. We have
found that system upsets, well shut-ins, hu-
man intervention, etc., will cause liquids to
be deposited periodically in the lower por-
tion of the wellbore. If there is a significant
difference in diameter between the upper
,..,.
+-------~----~~ __ ..
,
~~~~~--------~--_+--~~--~~~

FLOWING WELLHEAD PRESSURE, PSIA


1000

and lower portions of the wellbore, the larg-


est of the two will generally control the on-
Fig. 3-Wellbore critical rate: 27fa·in. tubing.
set of load-up.

Conclusions
1. The minimum flow rate, known as the
critical rate, qc' required to keep low-
pressure gas wells unloaded can be predicted
1 2 0 0 Q - r - - - - - - - - - - - - - - -__- -____


o
Critical Rat.T •• fa
Critical Rat. T.... - Slugglnll
----r------__--------__...__--__
10000+-----;----'-----+---;--+--,-t-----'---+---c--....__-+-~"""''_I
....•..
. "•. ··ii:~
.......... ~
::·.·.·.~·.~~·.,)il.
.
...
~'7I

... Production Chart Data


adequately with the Turner et at. liquid- {j. PrOduction Chart Data - 5111"ln,
droplet model without the 20% upward ad- • TII,no, Data - Unload.d
o Turn.r Data - loaded
justment. ~
()
2. A primary source ofload fluid for low- U)
:2
pressure gas wells can be condensed water. cr
3. Liquid/gas ratios below 22.5 bbl/ 0
UJ
1000
>
MMscf have no influence in determining the a:
UJ
U)
onset of load-up. aJ
a
4. Wells that exhibit slugging behavior
may not follow the liquid-droplet model be-
cause of a different transport mechanism.
5. Such variables as temperature, gas and
liquid gravity, and 1FT have little effect on
the critical rate, whereas wellbore diameter
and pressure have a direct and significant
impact. CALCULATED q,. MSCFD

6. In most cases, wellhead conditions can


be used to predict the onset of load-up. How- Fig. 4-Critical·rate data base.

JPT • March 1991 331


TABLE 1-CRITICAL-RATE DATA BASE

Gas
Specific Observed Calculated
Gravity Depth Tubing ID Condensate Water WHFP qc qo BHSP*
Test (air ... 1) ~ (in.) (bbVMMscf) (bbVMMscf) (psi a} (MscflD) (Mscf/D) (psig) Comment
- 1 0.582 7,812 2.441 0.1 1.9 275 726 874 548 qc test well
2 0.595 8,021 2.441 2.9 0.0 205 660 744 350 qc test well
3 0.628 8,437 2.441 5.7 3.9 212 585 737 650 qc test well
4 0.628 8,437 2.441 5.7 3.9 150 468 618 650 qc test well
5 0.620 8,042 2.441 3.5 7.1 185 573 691 355 qc test well
6 0.602 5,538 2.441 1.5 5.6 145 593 619 315 qc test well
7 0.602 5,538 2.441 1.5 5.6 145 617 619 315 qc test well
8 0.654 6,446 2.441 2.0 9.2 70 250 412 175 qc test well
9 0.668 6,026 2.441 2.9 7.0 140 607 580 223 qc test well
10 0.668 6,026 2.441 2.9 7.0 138 600 575 223 qc test well
11 0.602 6,499 2.441 1.3 0.0 130 635 586 306 qo test well
12 0.628 6,764 2.441 0.0 7.0 125 583 563 200 qc test well
13 0.672 5,678 2.441 5.9 2.3 165 649 628 329 qc test well
14 0.610 8,507 2.441 0.0 0.0 395 647 1,031 400 q c test well-ragged trace
15 0.610 9,445 2.441 0.0 0.0 255 612 821 450 q c test well-ragged trace
16 0.628 6,984 2.441 5.0 17.6 355 952 962 613 qc test well
17 0.620 6,034 2.441 2.9 3.6 105 430 520 182 q c test well-ragged trace
18 0.643 5,338 2.441 0.0 0.0 99 396 494 400 L-l0 chart-ragged trace
19 0.654 7,632 2.441 0.6 2.7 70 164 410 650 L-l0 chart-ragged trace
20 0.652 5,342 2.441 0.0 0.0 43 329 323 350 L·l0 chart
21 0.646 5.147 2.441 0.0 0.0 52 267 356 284 L-l0 chart
22 0.595 7,763 2.441 0.0 0.0 352 640 983 663 L-l0 chart
23 0.595 7,763 2.441 0.0 0.0 225 615 780 663 L·l0 chart
24 0.595 6,261 2.441 0.0 0.0 495 1,072 1,174 466 L-l0 chart
25 0.622 6,900 2.441 0.0 0.0 94 748 488 N/A L-l0 chart
26 0.661 7,428 2.441 3.4 12.3 65 276 395 568 L·l0 chart
27 0.682 4,680 2.441 2.2 3.4 59 500 371 N/A L-l0 chart
28 0.650 5,011 2.441 1.7 1.9 50 366 348 168 L·l0 chart-ragged trace
29 0.634 5,745 2.441 2.9 5.4 39 324 311 284 L-l0 chart
30 0.658 ·6,491 2.441 2.9 9.8 97 90 484 1,000 L-l0 chart-ragged trace
31 0.628 6,443 2.441 4.0 17.5 60 220 389 421 L-l0 chart-ragged trace
32 0.628 6,443 2.441 4.0 17.5 90 355 478 421 L·l0 chart-ragged trace
33 0.683 6,582 2.441 8.7 9.7 50 338 341 122 L-l0 chart
34 0.599 6,898 2.441 1.8 11.0 60 401 398 N/A L·l0 chart
35 0.599 6,898 2.441 1.8 11.0 80 450 460 N/A L·l0 chart
36 0.663 6.351 2.441 14.8 5.1 107 471 508 257 L·l0 chart
37 0.709 6,722 2.441 5.9 0.0 135 372 553 306 L·l0 chart-ragged trace
38 0.600 7,600 2.441 2.5 1.6 131 518 590 781 L·l0 chart
39 0.660 6,120 2.441 4.3 5.7 130 330 51;j2 462 L·l0 chart-ragged trace
40 0.617 6,880 2.441 0.0 0.0 82 511 460 181 L·l0 chart
41 0.674 6,556 2.441 0.0 0.0 90 558 461 273 L·l0 chart
42 0.662 6,301 2.441 1.8 1.5 100 493 491 362 L·l0 chart
43 0.642 4,751 2.441 2.4 6.5 183 627 676 673 L·l0 chart
44 0.651 5,065 2.441 3.5 1.8 120 518 542 202 L·l0 chart
45 0.610 6,285 2.441 0.0 1.0 47 358 349 212 L-l0 chart
46 0.600 6,335 2.441 0.0 0.0 315 885 924 440 L·l0 chart
47 0.651 8,439 2.441 7.5 0.0 165 712 638 447 L·l0 chart
48 0.620 8,158 2.441 4.6 4.9 75 408 438 450 L-l0 chart
49 0.750 8,508 2.441 0.0 2.3 380 666 924 1,100 L·l0 chart
50 0.625 8,466 2.441 2.4 5.9 155 648 630 725 L·l0 chart-ragged trace
51 0.625 8,466 2.441 2.4 5.9 145 564 608 725 L·l0 chart-ragged trace
52 0.621 8,504 2.441 4.5 0.0 235 781 782 728 L·l0 chart
53 0.621 8,504 2.441 4.5 0.0 225 755 764 728 L·l0 chart
54 0.717 8,440 2.441 5.3 1.0 165 620 610 725 L-l0 chart
55 0.688 6,796 2.441 1.1 3.8 49 430 335 244 L·l0 chart
56 0.674 6,381 2.441 6.5 8.6 59 397 372 154 L·l0 chart
• Bottomhole static pressure.

ever, for concentric tubing strings where the v( = terminal velocity of particle, this study. Special thanks go to the Opera-
tubing/packer is a significant distance from ft/sec tions personnel and the Production/Reser-
the completion interval, flowing conditions z= gas compressibility factor voir Technology support staff of Exxon's
of the largest-diameter segment should be Pg = gas phase density, Ibm/ft3 South Texas Div., who provided invaluable
used to predict the wellbore critical rate. PL = liquid-phase density, Ibm/ft3 cooperation and assistance in gathering and
a = interfacial tension, dynes/cm manipulating data and in preparing the nec-
Nomenclature essary technical review packages.
A = flow area of conduit, ft2 Acknowledgments References
p = pressure, psia We express our appreciation to the manage- r. Turner, R. G. et al.: "Analysis and Prediction
qc = critical gas flow rate, ment of Exxon Co. U.S.A. and Exxon Pro- of Minimum Flow rate for the Continuous
MMscflD duction Research Co. for the support and Removal of Liquids From Gas Wells," JPT
T = temperature, OR encouragement given while we conducted (Nov. 1969) 1475-82; Trans., AIME, 246.

332 March 1991 • JPT


,····fSS2~~~';""';mss:':
:
.. .. . ~~ ".-.
... ~.
~
~ ••
~ ~ ~
.;•. !.~ .
.. .:... .. ~-.:. ~ .
.

...... .... .. :- .. -:-- ..


~ .. ~ .. -~. :

. . .;. . ·;··:nT.
~.
-·~--·:··t-1-(
t········~···· ~ .... ,. ;.L-,...,....,.,---,--..,.---,-...,...J,.
1·±,.--....;..-;.....;......;....+....;...~1.~.1'-1I-"-1A....;..-..;........;.....;....;....;...;...;.1.l-••--...;....-i-.;.....i-+~..;,........

FLOWING WELLHEAD PRESSURE. PSIA

Fig. 5-Wellbore critical rate VS. WHFP.


. McCurdy Norris
2. Hutlas, E.J. and Granberry, W.R.: "A Practi- 51 Metric Conversion Factors
cal Approach to Removing Gas Well Liquids," Steve .. ca ...... lsacOnsultlngengio
bbl x 1.589 873 E-Ol = m3 neer In Corpus Christi, TX.He previous.
JPT (Aug. 1972) 916-22. ft x 3.048* E-Ol = m
3. Iiobi, M.l. and Ikoku, C.U.: "Minimum Gas "worked at Exxon's SoutIt T'-Dlv. In
ft3 x 2.831 685 E-02 = m 3
Corpus Chl'lstl. hoIds~a as
Flow Rate for Continuous Liquid Removal in OF (OF-32)/1.8 = °C
Gas Wells," paper SPE 10170 presented at the in. x 2.54* E+OO = em
degree in mechnlcal from
1981 SPE Annual Technical Conference and psi x 6.894 757 E+OO = kPa Tax_Alelll.
Exhibition, San Antonio, Oct. 4-7.
engineer in the Raa_oII',.r•.-omrtOlCIGY
'Conversion factor is exact.
4. Libson, T.N. and Henry, J.R.: "Case Histo-
ries: Identification of and Remedial Action for
Liquid Loading in Gas Wells-Intermediate
Provenance
Shelf Gas Play," JPT (April 1980) 685-93. Original SPE manuscript, Optimizing Re-
5. Weeks, S.G.: "Small-Diameter Concentric covery of Natural Gas From Depletion-
Tubing Extends Economic Life of High Drive Reservoirs: Part I-A Study to
Water/Sour Gas Edwards Producers," JPT Verify the Minimum Flow Rate Required
(Sept. 1982) 1947-50. ", H.. .... 1IG11'1'1tl III ta 8118f1l1Ol"4
for Continuous Removal of Liquids From f'8 Be arcb SpecIaIIat In the Fa.ciIItIes De-
6. Orris, P.W. and DeMoss, E.E.: "Liquid
Removal From Gas Wells-Gas Lifting With
Low-Pressure Gas-WeDs, received for SIgn SectIon at Exxon PJoduction Re-
Reservoir Gas, " paper presented at the South- review Nov. 16, 1989. Paper (SPE 20280) search Co. In Houston. Be holds a as
western Petroleum Short Course, Texas Tech accepted for pUblication Feb. 9, 1990. Re- degree from the U. of· In AuStm
U., Lubbock, Apri118-19, 1%8. vised manuscript received Dec. 21, 1990. and MSand Stanford
7. Engineering Data Book, 10th edition, Gas ~~. alt~.~_~.·_fl:teerl...
Processors Assn., Tulsa (1987). JPT
8. Coleman, S.B. et al.: "Understanding Gas-
Well Load-Up Behavior," JPT(March 1991)
334-38.

JPT • March 1991 333

También podría gustarte