Vous êtes sur la page 1sur 13

Distingtiished ,Author Series ,y~l~llq

n w%%?+%+
~:,
-
X..?*:? -,..
.7 ~+:
...
Nodal Systems Analysis of
,: Oil and Gas Wells
By KermitE. Brown,SPE, and

Kermit E. Brown is F. M. Stevenson Professor of Petroleum En9ineerin9 at the U. of


Tulsa. Since 7966 @wh has served es head of the Petroleum Egineerhg Dept., vice
president of research, and chairman of the Resources Engineering Div. He has conducted
James F, Lea, SPE

many courses m gas lift, nwltiphase flow, and inflow peiformamx ad served as a
Distinguished Lectwsr dudg 1969-70. Brow holds a ES deg,ee in niech?icaf ad
petroleum engineering from, Texas A&MU. and MS and PhD deg!ees from the U. Of
Texas, both in petroleum engineering. Brown sewed as the SPE faculty advisor for the U.
of Tulsa student chapter during 1982-83. He also sewed on the SPE board during
1969-72, the Education and Pm fessioalism Committee during 1966-67, and the
Education and Accreditation Committee dudg 1964-66 ad was Balcoes Sectim
chairman during 1964-65. He is currently o the Public Service Award Committee,
James F. Lea is a research associate in the Production Mechanics Group of Amoco
Production Co,, in Tulsa, He works on computer hnplen?entation of existing design and
analysis methods for a,fiificial lilt md improved application techniques. Previous~, he
worked with Pratt & Whitney Aircrati and .% 0;/ Co. and taught engimserig science at
the tmiwrsity level. Lea holds BS and MS degrees in mechanical egieering and a PhD
degree in thermal{ fluid science from Southern Methodist U., Dallas.

Summary
Nodal 1 analysis, defined as a systems approach to the 5. To check each component in the well system to
optimization of oil and gas wells, is used to ev61uate determine whether it is restricting the flow.mte
tboruughly a complek producing system. Every unnecessatiy.
component in a producing well or all wells in a 6. To permit quick recognition by the operwors
producing system can be optimized to achieve the management and engineering staff of ways to increase
objective flow rate most economically. Ml present production rates.
componentsbeginning with the static resemoir Theie are numerous oil and. gas wefls aruund the
pressure, endkg with the separator, and including world that have not been optimized to achieve an
inflow performance, as weU as. flow across the objective rate e~lciently. In fact, many may have been
completion, up the tublig string (inChIdlng 811Y completed in scb a m~er tit their maximum
downhole restrictions and safety valves), across the potenti81 kite cannot be achieved. Also, many wells
surface choke (if applicable), thrbugh horizontal flow placed on anificial lift do n6t achkve the efficiency
lines, and into the separation factilties-are tiulyzed. key shtiuld.
The pruductioi optimization of oil and gas wells by
Introduction nodal systems analvsis has contributed to improved
The objectives of nodBI aualysis are as follows. completion techniques, pfiduction, and efficiency for
1. To determine tlie flow me at which an existing many wells. @thou h this type of analysis was
oil or gas yell wifl produce considering wellbore proposed by Gilbert i. m 1954, it has been used
geometry and completion limitations (first by natural extensively in the U.S. only in the last few yeari. One
flow). principul ieason for tbk was the changing of allowable
2. To determine under what flow condhions (which producing. rates, and another has been the development
may be related to time) a well will load or die. of computer technology that allows rapid calculation of
3. To select the most economical time for the complex algorithms and provides ea.sify understood
installation of afiticial lift and to assist in the selection data.
of the optimum lift method, Past conservation practices in the U.S. more or less
4. To optimize the system to produce the objective restricted operaors t6 2- and 2 IA-in. [5.08- and
flow rate most economically. 6.35-cm] tubing and 4 shots/ft [13.1 shots/m] for
WYW 19s5societyof PetroleumEwi.eefs pmfomting. The use of larger tubing (41Aand 51Ain.

OCTOBER 1985 175I


&P, = P, - Pwf, = LOSS IN POROUSMEOIUM
AP2 = Pw,-Pwf = LOSS AcROSS COMPLETION
A% = k -%. = REsTRlcTl ON
dP4 = PKv-Po~v = SAFETY VALVE
AP5 = Pw~- PD~c = t SURFACECHOKE
AP6 = PO$C-P5,P= IN FLOWLINE
AP7 = Pwf -Pwh = TOTAL LOSS Ilq TUBING
nPs = Pw~-P,ep = r FLOWLINE

Fig. lPossible pressure losses in complete system.

[11.43 and 13.97 cm]) and 16 shots/ft [52.5 shots/m] models of other welf components can be used to
is common today. complete the p=dicted well pe,ffocmsnce.
Although the increase in flow rates in hlgh- Fig. 1 shows ,components that make up a detailed
productivity wells has popularized nodal analysis, it is, flowirtg wefl system. Beginning with the reservoir and
nevertheless, an excellent tool for low-rate wells (both procecdin~ to tie separator, the components are (1)
oil and gas) as well as for all aititicid lift wells. Some resemoir pressure, (2) well productivity, (3) wellbore
of the greatest percentage increases in production rates completion, (4] tubing string, (5) possible downitole
have occurred in low-rate oil wells (from 10 to 30 B/D re@ctive device, (6) tubing, (7) safety valve, (8)
[1.59 to 4.77 m3/d]) and low-rate gas wells (from 50 tubing, (9) surface choke, (10) flowline, and (11)
Ilp to 100 to 200 Mscf/D [1416 Up to 2832 to 5663 std separator.
m3 /d]). Numerous gas wells have needed adjustments To optimize tie system effectively, each component
in tubitg sizes; surface pressures, etc., to prolong the must be evaluated separately and then as a group to
onset of liquid Ioadlng problems. Nodsf analysis cm evaluate tbe entire well producing system. The effect
be used to estimate the benefits of such changes before of the chang&of any one component on the entire
they am made. system is ve~ impomant and can be displayed
Oneof ,jhe most impommt aspects of nodal analysis graphically yitb well analysis. Some aspects Of the
is to recognize wells that should be producing at rates IPR component are covered in Appendix A; discussion
higher than their current rate. Therefore, it can serve of myltiphase- flow pressure-drop correlations for
as an excellent tool to verify that a problem exists and pipelines is found in Appendix B,
that additional testing is necessacy. For example, The most common positions for nodal analysis
assume that a well is producing 320 B/D [51 m3 /d] of graphlcd solutions are listed below.
oil. Applying nodal analysis .to this well shows that it 1. At the center of the producing intefial, at the
is capable of producing 510 B/D [81 m3/d], This bottom of the well. This isolates the wells inflow
difference may be attributed to several factors, but performance.
nodal analysis can determine which component ii 2. At the top of the well (wellhead). This isolates
restricting the rate or can .detetine that iricomect data the flowline or the effects of surface mressure on
are the cause of tbe higher predicted rate, A basic production.
requirement for weli analysis is the ability to detine 3., Differential pressure solutioris (Ap) across the
the current inflow performance relationship (IPR) of completion intecwi to evaluate the effect of the
the well. Accurate well test data must be obtained and number of perforations on production in gravel-packed
the proper IPR applied for successful analysis, Then or standard completion wells.

1752 JOURNAL OF PETROLEUM TECHNOLOGY


o RATE

Fig. 2-Constructed
\

+
c~
q...

IPR curve.
+
BHP
or.
AP

c
u
x
Fig. 3Constructed
RATE +
TUBING
INTAKE
CURVE

tuping intake curve,

4. Solutions at the separutor, especially with gas-lift Specific ExantpIes


wells. This isolates the effect of separator pressure on A liited number of examples are presented here;
production. numerous examples, however, appear in the
5. Other solution positions for graphical solution are literature. I-5
at surface chokes, safety valves, tapered string Two specific subjects have been selected for
connection points, and dowuhole restrictions. example solutions.
The user must understand how pressure-flow 1. The effect of the downhole completion on flow
components of the weIl are grouped to form a rate is illustrated. An example solution for both a
graphicul solution at a node point. For example, if the gravel-packed well and a stundtwd perforated well is
solution is plotted at the bottom of the well (center of presented. Procedures to optimize the completions sre
completed intmvat), then the reservoir and the cmttined.
completion effects can be isolated completely from the 2. Quick recognition of those wells with a greater
enthe piping and production system. predicted potential thatr the present production rate is
Caution should be taken in ne~lecting even 200 to covered. These situations may be caused by a
300 ft [61 to 91 m] of casing flow fmm the center of restriction in one of the components in the system.
the completed interval to the bottom of the tubing.
Because of lower velocities, the larger pipe may not be
flushed out with produced fluids. This large section of Gravel-Packed Oti and Gas WeIls
pipe still can be neady full of completion fluids (water A paper presented by Jones et al. 4 seemed to be the
turd mud), even though the well may be producing catalyst that started operators looking more closely at
100% oil. Numerous flowing-pressure surveys have their completions. This paper nlso suggests procedures
verified this occurrence. A major company recently for determining whether a wells inflow capability is
surveyed a will producing 1,600 B/D [254 m3/d] of restricted by lack of area open to flow, by skin caused
oil up 2~-in. [7.3-cm] tubing. Because of a dogleg, by mud infiltration, etc.
tubing was set 1,000 ft [305 m] off bottom in the Ledlow and Granger3 have prepared an excellent
11,000-ft [3353-m] well, Both water and mud were summary of background material on gravel packing,
found in the 7-in. [17.8-cm] casing below the tubing, including detaifs on mechanical running procedures
even though rbe well produced 100% oil. Cleaning and selection of gravel size.
Wk well resulted in an increase of the rate to more The nodaf aua.lysis procedure for a gmvel-pscked
than 2 ,0i30 B/D [318 m3/d] of oil. This points out one well, illustrated with a sequence of figures, is
type of practical limitation of nodal analysis when presented here. The appropriate details, additional
tubing-pressum-drnp calculations are used to calculate references, and equations can be found in Ref. 3.
accurately a bottomhole flowing pressure (BHFP). The foflowing procedure is vtild for either an oil or
Here, the unalysis showed that the rate should be gas well with tie solution node at bottomhole.
higher and, hence, served as a diagnostic tool that 1. Prepare the node IPR curve (Fig. 2). (This step
prompted the mnning of a prsssure traverse. Irr many assumes no Ap across the completion. )
cases, the anafysis predicts what should be expected, 2. Prepare the node outflow curve (tubing intake
and the operator is advised to look for problems if the curve in Fig. 3), which is the surface pressure plus the
well is producing below that prediction. tubing pressure drnp plotted as a function of rate.

OCTOBER 1985 1753


3 API AP2 AP3
/

4
BHP

flP

(AP=O
T \
w (
cl
RATE + RATE +

Fig. 4Transfer Ap. Fig. 5Construct Ap across gravel pack.

3. Transfer the differential pressure available 6. Evaluate other shot densities or perhaps other
between the node inflow and node outflow curve on hole sizes until the appropriate Ap is obtained at the
the same plot (FQ. 4) to a Ap curve. objective mte (Fig. 6). Perforation efficiency should
4. Using the appropriate equations, 3,4 calculate the be considered at thk time. A good review on
pressure drop across the completion for various rates. perfoiatiug techniques, which poiuts out such factors
Nnmerous variables have to he considered here, as the number of effective holes expected and the.
includ]ng shots per foot, gravel permeability, viscosity effect of the number of holes and hole sizes on casing
and density of the fluid, and length of the perforation strength, was presented by Bell. 6
tunnel for linear flow. Add this Ap curve on Fig. 4, as 7. The Ap across the pack can be included in the
noted in Fig. 5, IPR curve, as noted in Fig. 7.
5. Evsluate this completion (Fig. 5) to detemnine
whether the objective rnte can be achreved with an Example ProblemTypicaI Gulf Coast Well With
accepted differential across the grnvel pack. Company GraveI Pack. Below is a list of given data.
philosophies on accepted Ap values differ. A
reasonable maximum allowable Ap that has given ~, = 4,000 psi [27.6 MPa],
good results rnnges from 200 to 300 psi [1379 to 206g D = 11,000 ft [3352 m] (center of perforations),
kl%t]for single-phase gas or liquid flow. Most k = 100 md (penneabfiity to gas),
operatom will design for smaOer Aps for multiphase h = 30 ft [9.1 m] (pay interval),
flow across the pack. h, = 20 ft [6.1 m] (perforated interval),

RATES
POSSIBLE

o \
RATE + RATE +

Fig. 6Evaluation of various shot densities. Fig. 7Gravel pack solution by including Ap completion in
IPR curve.

1754 JOURNAL OF PETROLEUM TECHNOLOGY


4 8r /

3 -
q

&-
;2 -
L
8
m Pr = 4000 Psl
DEPTH = 11,000
1- K = 100 MD

I
1 I I 1 I 1 1 I Oo,
00 20 40 60 80 100 120 140 70
RATE, MMCFD RATE,MMCFD

Fig. 8IPR curve for gas wellgravel-pack analysis. Fig. 9Evaluation of tubing sizes.

40/60-mesh gravel-packed sand, around the perforation because of unconsolidated


640-acre [259-ha] spacing, forrrmtion-that is, sand flows immediately into all
8~.in. [21.9-cm] casing; 1()%-in. [27.3-cm] perforated holes until properly prepacked.
drilled hole,
Tg = 0.65,
Procedure.
screen size = 5-itr. [12.7-cm] OD, 1. The IPR curve is prepared with Darcys law, and
gas-sales-line pressure = 1,200 psi [8273 Wla], the additional turbulence pressure drop4 is included
short flowline. (Fig. 8).
2. Tubing sizes of 2%, 31A, and 41A in. [7.3, 8.89,
This well is to he gravel packed. The tubing size and 11.43 cm] are evaluated at a wellhead pressure of
and the number of shots per foot are to he evaluated 1,200 psi [8272 kpa], which is needed to flow gas into
with an undcrbalanced tubhrg-conveyed gun. It is the sales fine. From analysis of Fig. 9, 41A-in.
assumed that thereis no computable zone restriction [11.43-cm] tubing is selected. Note that, if market

RATE, MMCFD RATE, MMCFD

Fig. 1OAP available fmm sandface to tubing intake. Fig. 1lAp across gravel pack at 4, 8, 12, and 16 shotslft.

OCTOBER 1985 1755


RATE, MMCFD RATE, MMCFD

Fig. 12Completion effects included with lPRgravel- Fig. 1S-Effects of wellhead pressure-gravel-packed well.
packed well.

conditions permitted, much figher rates cotdd be surrounded by a low-permeabfity zone. They still
projected with adequate sand control. incorporate basic concepts suggested by Jones et al. 4
3. The Ap is transferred, as noted in Fig. 10. This is for gravel-packed weUs.
the Ap available across the gravel pack.
4. The Ap across tie pack for 0.75-in. [1.905-cm] ExztnpIe ProbIem and Procedure for
-diameter holes with 4, 8, 12, and 16 effective shots/ft a Perforated Weff
[13. 12,26.2, 39.4, and 52.5 effective shots/m] (Pig. Iu thk section, a sample oil well with a low GOR, a
11) should be calculated with Jones et al.s equations low bubblepoint pressure, and assumed single-phase
m with modifications of these equations adjusted to tit liquid flow across the completion will be anutyzed.
field data. The reason for thk selection is that current technology
5. Figs. 11 and 12 show the final two plots has offered solutions only for single-phase flow (gas or
indicating that 16 shots/ft [52,5 shots/m] are necessary liquid) across such completions. When two-phase flow
to obtain a Ap of about 300 psi [2068 kPa] at a rate df occurs across either a gravel-packed or a standard
58.5 MMscf/D [1.7x 106 std m3 /d]. Additional perforated well, relative permeability effects must be
perforations could bring thk AP below 200 psi considered. Additional turbulence then occurs in
i1379 kPa~. - ~ grovel-packed weUs and creates more energy losses.
6. To bring tik well on production properly, one McLeod7 noted that most of the pressure drop can
more plot (such as FQ. 13) should be made with occur across a compacted zone at the pe.tioration wall
several weffhead prcssmes so that Ap across the pack because of turbulence. He annlyzed a gas-well
can be watched through fhe observation of rate and example and showed that 90% of the totaf Ap across
wellhead pressure. Thk procedure is described by the completion, in fact, was caused by turbulence
Crouch and Packs and Brown et al. 3 across the approximately IA-in. [1.27-cm] -thick
compacted zone. (.Eefs. 3 and 7 provide more details).
Nodal Atwdysis To Evafuate a Standard To use this technique, the crashed-zone thickness,
Perforated WeU e,, the pemneabllity, kc, the perforation-tunnel
In 1983 NfcLeod7 published apaper that prompted diameter, dp, and the length, Lp, must be ~own.
operators to examine completion practices on normally Obviously, because of the many input variables
perforated wells. Although numerous prior required, the technique can only be approximate and
p~bli@tioss-10 discussed this topic and Companies iadicate trends. It is hoped that fature research in this
bad evaluated the problem, Wk paper sparked new area wiU lead to mom accumte models of pressure
interest. A modification of dds procedure is presented drup through perfomtions shot in both over- and
in Ref. 3. underbalanced condkions.
The procedure is similar to that offered for gravel-
packed weUs, except that the equations used for the Example Problem.
calculation of pressure drop acmsa the completion j, = 3,500 psi [24.1 MPa],
have been altered to model flow through a perforation D = 8,000 ft [2438 m],

iT56 JOURNALOFPETROLEUM
TECHNOLOGY
3.0 - 3.0
DEPTH = 8000
R = 3500 Psl
2.5 - 2.5
[\. TUBING I.D. = 2.992

~ 2.0 -
L
&-
; 1.5-
n.
I
m 1.0-
DEPTH= 800L7
,5 - Pr = 3500
Pwh = 140 PSI

! 1 I 1 (
5000 6000
I +! \, I ,\l
o 1000 2000 3000 4000 o 1000 2000 3000 4000 5000 6000
RATE, BID RATE, BID

Fig. 14-IPR and tubing curves for peqorated oil well. Fig. 15Transfer for Ap curve-perforated oil well.

36API [0.84-g/cm3] oil, D = 8,000 ft [2438 m],


Solution GOR = 180 scf/bbl [32 std m3 /m3], 2~-in. [7.3-cm] tubing,
80-acre [32. 3-h2] spacing, p, = 2,100 psi [[4.5 MPa],
5Win. [13.97-cm] casing, 35API [0.85-g/cm3] oil,
8 k-in. [21.59-cm] hole, 50% water [yW= 1.07],
Lp =4-in. [10.16-cm] perforation tunnel (see solution GOR=300 scf/bbl [54 std m3/m3],
Table 6 of Ref. 7 for tabulated values), sepamtor pressure =60 psig [413 kla],
e. around perforated tunnel = O.5 in. [1.27 cm], flowline len=@=4,000 ft [1219 m],
ph = 800 psi [5515 kPa], well test 500 B/D [79.5 m3/d] at 1,740 psi [12 MPa],
h = 30 ft [9.1 m], pb = 2,400 psi [16.6 MPa],
hp = 20 ft [6.1 m], -yX = 0.7, and
78 = 0.7, tubing size = 2 Win. [6.35-cm] ID.
T = 180F [82C], and
p WA = 140 psig [965 kpa]. Sufficient gas pressure is available (2,000 psi [13.8
MPa]) to inject gas near the bottom, and a total
gas/liquid ratio of 800 scf/bbl [143 std m3/m3] is
Procedure.
maintained for gas lift. The flowline might be
1. Prepare the IPR ctt~e with Darcys law,
restricting the rate. With nodal analysis, a graphical
assuming no Ap across the completion.
solution can be generated auicklv at the wellhead
2. Plot the node outflow curve (tubhg intake) for
2x- 2%,, ~d 31h-in. [6.03-, 7.3-, and S.w-mnl location. - -
Examination of the results in F]g. 17 indicates that
tubing, This dekmmines the pressure requited at the
the flowline is a restriction because the Dmssure loss in
bottom of tubing for flow through the tubing. Steps 1
the flowline (21%-in. [6.35-cm] ID) sho~s a significant
(IPR) and 2 (tubing intake) a~ shown in Fig. 14.
increase in pressure loss with rate and is angled
Assume 3 k-in. [8.89-cm] tubing is selected.
sharply upward at the intersection point between the
3., T~sfer the Ap curve, as shown in Fig. 15.
two cuwes shown. The intersection point of the
4. Using the appmpriite equations fmm McLeod7
pressure required at the flowline intake and the IPR
(and as discussed by Brown et al. 3), determine the
pmsure minus the pressure drop in the well from
Aps across the completions listed. in Table 1.
sandface to the wellhead is the point of predicted flow
An an~ysis of Fig. 16 shows the importance of
from the well.
perforating undeinlanced. Of course, the best fluids
A 3- and 4-in. [7.62- and 10. 16-cm] flowline is then
and techniques should be used.
evaluated on the same plot. As soon as the slope of
the flowline intake pressure VS. rate becomes small
Recognition of Components Causing Restricted (showing very little increase of Ap with rate), thin the
Flow Rates in a WeIl flowline diameter is sufficiently large. The diameter
Example ProbIemAnafysk of Flowline Capacity. should not be oversized because additional slugging
The following well is on gas lift. and head@ may occur. Some operators just add a
OCTOBER 1985 i757
TABLE I-SAMPLE COMPL5T10Ns FOR PERFORATED OIL WELLS

Feet Perforation kc as%of


Number
ShotslFt
?erforated Condition k, Formation
1 4 20 Overbalanced with 10
filtered salt water
2 a 20 Overbalanced with 10
salt water
3 4 20 Underbalanced
with 30
filtered salt water
4 8 20 Underbalanced
With 30
filtered salt water

parallel line instead of replacing the current line with a 320-acm [129-ha] spacing,
larger size, T = 200F [93C],
k = 0.12 md,
Restriction Caused by Incorrect Tubing Size. The p~h = 100 psig [689 kfa],
tubing may be either too large (causing unstable flow) hp = 15 ft [4.57 m],
or too small (reducing flow rate). This can be -yg = Q.7,
recognized immediately on a nodal plot and is as
hole size = 8% in. [21.6 cm], and
important in high-rote gas lift wells as in low-rate gas
no skin effects.
wells.
A weak gas well is chosen to show how to
deterroipe when thetubingistoolar eand to predict
Evaluate 3 Y-, 234-, 2X-, and 1%-in. [8.89-, 7.3-,
when loading will occur. The Gmy ,? cyrelationis
=commended for use in the calculation of tubing 6.35-, and 3.81-cm] tubing (1.66-in. [4.21-cm] ID)
pressure drops in gas wells that produce some liquids. and l-in. [2.5-cm] tubig (1 .@19-in. [2.66-cm] ID) for
MS well.
Example ProbIemWeak Gas Well with Note in Fig. 18thataU sizes oftubing are too large
Liquid Production. fortbis particular caseexcept thel.049-in. [2.66-cm]
-ID tubing. Unstable flow isindicated bythetubig
P, = 3,200 psi [22 MP?], curves crossing the IPR at a point to the left of the
30 bbl/MMcf [168 x 106 m3 /m3] condensate, minimnm forthelarger tubimg. The ?.O-in. [2.54-cm]
5 bbl/MMcf [28.1 X10-6 m3/m3] water, tubing shows stable flow,
D= 10,000 ft [304$ m], The same type of analysis can he made for oil wells
,4 = 15 ft [4.57 m], for various tubing sizes.

3.0 -

DEPTH = 8000
r
500
79

2,5 - TUBING I.D. = 2.992


R = 3500 PSI
5
. 2,0 -
&
.
; 1,5

s
: 1.0 -

.5 -

1500 2500 3500 4s00


02~
500
RATE, B/D RATE, B/D

Fig. 16Production vs. various perforated completions. Fig. 17Wel!head nodal plotflowline size effects.

1758 JOURNAL OF PETROLEUM TECHNOLOGY


Well Inflow und Completion Restrictions. It is very TABLE 2AOFP,S FOR HIGHER VALUES OF n
important for operutora, engineers, and managers to AOFP
recognize inflow restrictions immediately. Some [m3,,jx, o-5]
n (MMscf/D)
companies have arranged their computerized well
0.7 7 2
records to do such things as call up a group of wells in 0.s 38 11
one field in descending-kh-vahe order. In addkion, all 0.s5 90 92
other available pertinent information, includlng the 0.9 211 60
latest test data, cm also be printed out. 1.0 1,157 32S

Exsmple Problem. Compare predicted perfonnanm to


actual oilwell performance.
A closer estimate can be made from
k = 50 md (cores), kh (50)(30) BID
h = 30 ft [9.14 m] (logs), =1.56 ,
KOBO = (1,000)(0.8)(1.2) psi
35API [0.85-g/cm3] oil,
casing = 7 in. [17.78 cm], but it requires that PO and 30 are known. One can
tubing = 2X in. [6.1 cm], recognize that a 35API [0.85-g/cm3] cmde at 170F
D = 7,000 ft [2134 m], ~77C] with 400 scf/bbl [71 std m3/m3] i solution
yg = 0.65, will have a viscosity less than 1 and that the product
T = 170F [77C], poBO will be close to 1. Heavy cmdes, of course, will
p, = estimated 2,400 psi [16.5 MPa], and have high viscosities, and a larger value must be used
pwh = 250 psi [1723 k%]. in estimating the productivi~ index.
Also, a reasonable estimate at lower pressures ia that
about 500 psi [3447 kPa] is required to place 100
The latest well test shows thk well producing 600
B/D [95 m3/d] oil (no water) with a GOR of 400 scf/bbl [17.8 atd m3/m3] in solution giving a
bubblepoint pressure, pb, of 2,000 psi [13.8 MPa].
scf/bbl [71.2 std m3/m3 ] (natural flow).
Standkgs 14 correlation shows the pb to be very close
Determine whether this well is producing near ita
to 2,000 psi [13.8 MPa] for these conditions. This
capacity. It is the engineers responsibility to recognize
thk wells potential quickly and to recommend permits a quick calculation of the maximum flow rate.
addkional testing, a workover, a change in tubing, or Jp~
other action. %ax=qb+=
..-
A very quick estimate of the productivity index can
1.5 (2,000)
be estimated from the product kh in darcy-feet. = 1.5 (2,400-2,000)+
1.8
=600+ 1,667

=2,267 BID.

r
2.5 - 30
DEPTH = i 0,00W
Fwh = 100 PSI
R = 3200 PSI 25 -
2,0 -
30 B/MMCFD COND.
~ 5 B/MMCFD WATER
20 -
~- 1.5 - $ .995
x ~- 15 -
L
x
%1.0 - &
~ 10 -

7
,5 -
5 -

I c , I 1 I [t [ ! ! I I
0 50 100 150 200 250 00 500 1000 i 500 2000 2500
RATE, MCFD
RATE, MCFD
Fig. 18Tubing.diameter effects-weak gas well. Fig. 19Predicted vs. observed oilwell performance.

OCTOBER 1985 1759


3.0 7UU

2.5 [~ a

,s~%:,fl::.9g5 , } , ;mjMy!y,: ,>,

0 500 T000 1500 2000 2500 200 400 600 800 1000 1200 I
RATE, WD RATE, B)D

Fig. 20Wellhead pressure effects on ratenodal plot. Fig. 21 Pmduction vs. wellhead pressure.

The IPR curve can be drawn quickfy and the tobing area to flow than to stimniation. Refs. 3 aad 4 provide
curve imposed on the sample plot (Fig. 19). The more details on this procedure.
intersection shows a rate of 760 BID [121 m3Id] of
oil. Effects of Wellhead And Separator Pressure
The question of whether this well is worth spending Specific cases of gas wells and gas-lift oil wells may
sufficient money to determine why the rate is less than be influenced signiftaotly by changes in separator
the prdcted rate now arises. The source of error pressure andior welfhead pressure.
could be with two. bits of information. Is the A good plot for both oil and gas wells is a
permeability of 50 md (obtained fmm cores) correct? deliverability plot of wellhead pressure vs. rnte and, in
1s there a completion problem? For this well, the turn, separator pressure vs. rate. This plot a.tso can
possibility of additional production justifies the show the loading or critical rate and offers immediate
expenditure to ron a buildup test to verify M/yOBo selection of rates based on wellhead pressures. The
and to check for skin. A high skin may indicate that sample data used to construct Fig. 19 arc used to
farther testing is needed to determine whether a rnte- construct Flg. 20 at various wellhead pcessures. From
sensitive skin exists to decide whether stimulation or tMs graph, data are used to consmtct Fig. 21, which
teperforating is required. demonstrates dte well response as a function of surface
pressu~.
Restricted Gas Well
Many operatora fail to tecognize the significance of Summm_y and Conclusions
the exponent n for gas-well IpR equations obtained
NodaJ analysis is an excellent tool for optimizing the
from four-point tests. It is common to see expcments
objective ffow rnte on boti oil aod gas wells. A
of 0.7 to 0,8 .or less in gas wells. For exnmple, the
common misconception is that often there are
following equation was obtained from a U.S. @f
insui%cient data to use thk analysis. Thk is tme in
coast well after data were plotted on log-log paper.
some cases, but mzmy amazing improvements have
been made with very few data. The use of nodal
c?gm@@W(5>oo02PW2107
Mcf/d. analysis has &so prompted the obtaining of additional
data by proper testing of numerous wells.
The operntor of tiIs well had a market of 15 Aaother common statement is that there is too much
MMscf/D [424x 103 std m3 Id]. Note that tbk well error involved in the vmious multiphase-flow tubing or
has an abso[ute o en-flow potential (AOFP) of 6,984 flowfiie correlations, completion formulas, etc., to
Mcf/D [198x 103Pm 3/d]. See Table 2 for AOFPs for obtain meaningful resufts. Because of these possible
higher values of n. errors, it is sometimes dficult to get a pmdlctive
Obviously, this well has a serious completion nodal plot to intersect at exacdy the same production
restriction. Sufficient data are nlready available to plot rate of the actual well. Even if current conditions
in the form suggested by Jones et al. 4 They suggested cannot be matched exactfy, however, the analysis can
plotting (p, 2 p ~f 2)/qg,, on the onihwe vs. qg.c on show a percentage. increase in production with a
the abscissa to evaluate the need for opening more change, for instance, in wellhead pressure. These

1760 JOORNALOF PETROLEUM TECHNOLOGY


predicted possible increases often are fairly accufate T = temperature, F [C]
even without an exact match to existing flow rates. yg = gas gravity (air= 1.0)
Two detailed illustrations are given in this paper to y,, = water gfavity
show .jhe effect of perforation shot density in both y. = oil viscosity, cp pa.s]
gravel-packed and standard perforated wells on
production.
Nodsl analysis has completely altered perforation
philosophy in the U.S. snd has encouraged higher- 1, Mach, J,, Pmano, E., and Brown, K.E.: A Nodal Approach for
density perforating and use of open-hole completions Applying SYSteInSAndysiS to the Flowing and Artificial Lift Oil
when practical. One of the most important aspects Of or Gas Well,,, paper SPE 8025 available at SPE, .Richardson,TX.
nodal tiysis is that it offers engineers and managers 2 Giiben, W. E.: , Flowing and Gas-Lift Well Performance, Drill.
and Prod Proc. , API (1954) 126-43.
a tool to recognize quickly those components that an 3, Brown, K,E. et cl,: &Prod.cdm Optimization of Oil and Gas
restricting production rates. Wells hy No&f Systems Analysis,>XTechml.g3 of Artificial Lf$
Although not discussed in this paper, nodal analysis Methods, PennWell Publishing Co., Tulsa (1984) 4.
is used to optimize all artificial lift methods. 3 Rate 4 Joes, L.G. Bloum, E.M., and Glaze, C.E.: Use of Shofi Term
MultipleRateFlowTeststo PredictPerfmmmeofWellsHaving
predictions, along with horsepower requirements for T.I+JUleme,>7paperSPE6133 presented at the 1976 SPE Amma3
sll lift methods, cm be predicted, thereby permitting Technical Conference and Exhibiticm, New Orleans, Oct. 3-6.
easier selection of lift methods. 5 Crouch, E.C. and Pack, K.J.: &SystemsAnalysis Use for the
Finally, some ve~ complex network systems, such DesiS and Evahafim of Higi-Ram Gas Wells,,, paper SPE 9424
presented.at the 1980 SPE Annual Technical Confe=?ce and Ex-
as ocean-floor gas-lift fields (including gas allocation hibition, Dalfas, Sept. 21-24.
t.omsximize rates) and most economical gas rates, can 6, Bell, W.T.: Perforating UnderbalancedEvolving Tech-
be pfedlcted with this procedure. niques,, >J, Per, Tech. (Oct. 1984) 1653-62.
Nodd analysis, however, should not be used 7 McLead, 33.0. Jr.: l%. Effect of Perforating Ccmditicmson Well
indkcriminately without the recognition of the Performance,,> 3. Per. Tech. (lam 1983) 31-39.
significance of all plots and the meaning of each 8, Locke, S,: LAnAdvanced Method for Predicting fhe Prod. ctivitj
Ratio of a Pmfmated Well,,> 3. Per. Tech. (D,,. 19S1) 2481-S8.
rslationsbip. Engineers should be tmined to understand 9 Hong, K.C.: Productivity of Petiotated Completions in Forma-
the assumptions that were used iR developing the tions With or Without Damage, J. Per. Tech. (Aug. 1975)
various mathematical models to describe well 1027-3% Tram; , AIME, 259.
10 IGotz, I. A.,. Xmeger, R.F., and Pye, D.S.: EffectofPerfomtion
components. Also, recognizing obvious ermrsnd Damage cm WdJ Productivity,,, J. Per. Tech. (Nov. 1974)
using practical judgment are necessary. Experience in 1303-1% Trans., AIME, 257.
diffenmt opemting areas can indicate the accumcy to 11 Gxay, H.E.: Verticd Flow Comelation in Gas WeUs, User
be expected from various correlations used in nodal Maumi for API 14B, .Mbswface Conmiled Safety Valve SizinS
mzalysis well models. Cbmpurer Program, App. B. API. Dallas (June 1974).
12 Vogel, J.V.: ..Inflow Performance Relationships for Soknion-Gas
Drive Well s, J. Pet. Jech. (Jan. 1968)S3-92 Trans.,AfME,
Nomenclature 243.
13 Fetkovich, M,J.; .Thc Isochronal Testing of Oil W.Us,, paper
B. = FVF, bbl/stb [m3/stock-tank m3] SPE 4529 presented at tie 1973 SPE Annuaf Meeting, Las Vegas,
Cl = numerical cnefticient seDL 30-oct. 3,

dp = perforation-tunnel diameter, in. [cm] 14 Suindig, M.B.: Inflow Performance Relationships for Damased
Wells Pmd.cing by Solution-Gas Drive,, > J. Pet. Tech. (Nov.
D = depth, ft [m] 1970) 1399-1400.
e. = .cmshed-cone tlickness, in. [cm] 15 Eickmeier, J. R,: *How to Accurately Predict Future Well Pm-
.h=height of pay interval, h [m] ductilities,,, World Oil (May 1
1968) 99.
16 Dia.-Couto, L.E, and Gobm, N H.: <GeneralInflow Performance
hP =height of interval perforated, ft [m] Relationship for Solntion-Gas Reservoir Wells, J. Per. Tech.
J= productivity index, B/D/psi [m3/d/kPa] (Feb. 1982,-., .-. .-.
?~~-~~
k= permeability 17 Uhri, D.C. and Blount, E. M,: ,&Pivot Poim Methcd Quickly
Predicts Well Pdmmance, ,S Wmid Oil (Mw 1982) 153-64
kc = penneabiity ofcmshed zone around 1s Aga@ R. G., A1-H.ssainY, ?., and Ramey, H.J. Jr.: .A. In.
perforation, md vemgmon of Wellbore Storage md Ski Effect in Unsteady Liq-
kf = formation penneabllity, md uid Flow: L Amdvtical Treatment. Sot. Pet. Em. J. (Sect.
1970) 279-9Ll T,&?., AIME, 249.
LP =length of perfora.tion tunnel, in. [cm] 19 Agarwaf, R, G., Carter, R. D., and PoRock, C.B.: Evaluation
p = pressure, psi [kPa] and Performance Predictim of Low-Permeability Gas Wells
P_b bubblepoint pressure, psi [kPa] S&mdamd by Massive Hydraulic Fracture,,3 J. Per. Tech. (March
1979) 362-72 Trans. , AIME, 267.
p, = Kservoirpressure, psi [kPa] 20 Lea, J. F.: C-AvoidPremamm Liquid fmadig in Tight Gas Wells
pwf = BHFP, psi [~a] by Using Pmfrac and Pomfrac Test Da% Oil ??d Gas J. (Sept.
20, 19S2) 123.
P ~fi = wellhead pressure, psi [kpa]
21 Meg, H. M .1.: Production Systems Analysis of VerticaUy
Ap = pressure difference, psi ma] Fmctured Wells,, paper SPHDOE 10S42 presented 81the 1982
qb = flOWrate at the bubblepoint, Mscf/D [103 SPFJDOE Unconventional Gas Recove~ Symposium, Pittsburgh,
std m3 /d] May 16-18,
22 Greene, W.R.: ,Analyzing the Performance of Gas Wells,
q~~ = msxirnum flow rate, B/D [m3/d] Pm,, , 1978 .%utiwestem Petmlem Shmt Cows., Lubbock, TK
cIe = liquid flow rate, B/D [mS/d] (APril 20-21) 129-35.

OCTOBER 19S5 1761


23. Hagedorn, A.R. and Brown, K. E.: ?3xperimentd Study of The Fetkovich procedura 13 requifes a three-or fom-
Fmssure Gradients Occurring During Continuous Two-Phase flow-rate test plotted on log-log paper to determine sn
Flow in SmaJJ-Diameter Vertical Conduits,. J. Pet. Tech. (April
1963J 475-S4 Trans. ANE, .234.
equation in the form of a gas-well backpressure
24. Dins, H. Jr. andRos,N.CJ.: VerticalFlowof GasandLiquid equation with a coeffkient and exponent determined
M,xturesin Wells,,, Pro.., gixti World Pet. Cong. (1963) 451. from plotted data. This is equivalent to armlysis of an
25, Orkiwcwsti, J.: Wedicdng Two-Phase Prass.re Drops in Ver- oil well with gas well relationships.
tical P,pes,,, J. Pet. Tech. (J.ne 1967) 829-38; Trans., A2ME, %audmgs 14 extension of Vogels wmk accom3ts for
240.
; 26, Beggs, H.D, and Brill, J. P.: A Study of Twc-Phase Flow in In- flow-efficiency values other than 1.00. Jones et al. s4
clined Pipes,, >J. Pet. Tech. (May 1973) 607- 1.% Tram., .41ME, procedure will determine whetier sufficient mea is
255, open to flow.
27, Aziz, K., Govicr, G.W., and Fogammsi, M.: %essure Dmp in
Wells Producing Oil and Gas,,, J. Cd.. Pet. Tech. (July-Sepl.
. ..-.,.. 38-d8
1972). Futnre IPR Curves
28. Dukkr, A, E. et .[.: Gas-Liqoid Flaw in Pipelines, 1. Research The prediction of future IPR curves is critical in
Results, AGA-API Pmjmt NX-28 (May 1969). determining when a well will die or will load up or
29, Du!der, A.E. and Hubbard, M. G.: A Model for Gas-Liquid Slug
Flow in Horizontal and Near Horizontal Tubes, Ind. and Eng. when it shoufd be placed on artiticid lift. The
Chen. (1975) 14. No. 4.33747. foUowing procedures can be used (1) Fetkovich 13
30. Eaton, i. A. et il.: The Predction of Flow Paftems, Liquid procedure, (2) combination of Fetkotich and Vogels
Holdup and Ptessure Losses Occurring During Continuous TwG- equation, IS (3) Coutos 16 procedure, and the (4) PivOt
Phme Flow In HorizontalPipelines,,, J. Pet. .Tt-ch. (June 1967)
point method. 17
815-28; Tram,. , A3ME, 240.
31, CuUen&r, hM,H. and Smith, R. V.: Practical Solution of Gas-
Flow Equations for Wells and Pipelines witi Large Temperature Transient IPR Curves
Gradients,, J. l+v. Tech, (D... 1956) 281-8R Tmm., A2ME,
Oil or Gas WeIIs. A time element allowing fhe
207.
32. Poeimmnn, F.H. and Cmpenter, P. G.: The Muldpbase Flow of constmction of IPR cmves for transient conditions can
Gas, Oil and Water Through Verdcal Flow String wilh Appliw.- be brought into Darcys law. This is important in
tion to the Designof Gas-Lift Installations, Drill. and Prod. some wells because of the long stabfiza.tion time. (See
Pm,., API (1952) 251-317, Ref. 3 for discussions by several authors.)

Fractured Oil and Gus Wells. The constmction of


APPENDIX A
IPR ctnves for fractured oil or gas wells has been
Inflow Performance treated in the literature by Agsfwal et al., 18,19Lea, 20
Inflow performumc is the ability of a well to give up and Meng. 21 Fractured wells can show flush
fluids to the welJbore per unit drawdown. For flowing production initially but drop off considerably in rate at
and gas-lift wells, it is plotted normally as stock-tunk future timex.
barrels of liquid per day (abscissa) vs. bottomhole
pressure (BHP) opposite the center of the completed IPR Methods For Gas Weffs. GenemIIy, a three- or
intend (ordimte). The total volumetric flow rote, four-flow-fste testis required for a gas weJl from
includlng free gas, can also be found with production which a plot is made on log-log paper and the
values and PVT data to cdcuIate, for instartce, a total appropriate equation derived.
volume into a pump.
Brown et al. has given detailed example probIems q=cl(P2PLfV
for most methods of constmcting IPR curves. Nothing,
however, replaces good test data, and many where q is the mte of flow, Cl is a numerical
procedures, in fact, do require from one to four coeftkient, characteristic of the particular well, p, is
dlffemnt test pointsthat is, a stabl@ed rate and the shut-in rcse~oir pressure, p .f is the BHFP, and n
conesponding BHFP, as well as the static BHP, arc is a numerical exponent that is characteristic of the
usually a minimum requirement for establiahlng a psrticxdw weJL (See Ref. 22 for a discussion on gas
good IPR. well pefionnance). Also, Darcys law can be used,
and the turbulence terms should always be included6
for all but the lowest rates.
IPR Methods for Oil Wells Fructured and transient wells have Alsobeen treated
For flowing pressure above the bubblepoint, test to in the literature.
find the productivity index, or cnlculate the
productivity index from Darcys law. APPENDJX B
For two- base flow in a reservoir, apply Vogels
procedure 1? or Darcys law using relative Muftiphase Flow Correlations
permeability data. The use of multiphsse-flow-pipeline pressure-drop
For reservoir pressure greater thau bubblepoint correlations is very important in applying nodal
(P, >P~) and BHFp above or below the bubblepoint, analysis.
use a combination of a straight-line productivity index The corrclationa that am most widely used at the
above pb and Vogels 12 procedure below. present time for vetical multipbase flow were

1762 JOURhAL OF PETROLEUM TECHNOLOGY


.
developed by Hagedorn and Brown, 23 Duns smd Vertical Gas Flow. The procedu~ by (ldlender and
Ros, M Ros modification (Shell Oil Co., unpublished), Smith 31 and Poettmann aud C~enter32 are
Orkizewski, ~ Beggs and BriU,26 snd Aziz. 27 These recommended for gas-flow calculation in wells.
correlations calculate pressure drop ve~ well in certain
wells snd fields. However, one may be much better Wet Gas Wefla. We recommend the Gray
than the other under certain conditions, and field correlation 11 for wet gas wells.
pressure surveys are the onfy way to find out. Without
knowledge of a particular field, we would recommend S1 Metric Conversion Factors
beginning work with the correlations listed in the bbl X 1.589873 E01 = m3
above order. cu ft X 2.831 685 E02 = m3
ft X 3.048* E01 = m
Horizontal MuIti haae-Flow Pipeline CorreIationa. in. x 2.54* E+OO = cm
Beggs and Brill, 2? Dukler et al., 28 Dukler snd psi x 6.895757 E+OO = kPa
Hubbard, 29 Eaton er al., 30 and Dukler using Eatons
holdup28,30 nre the best horizontal-flow correlations. -conversion tam, i, exact. Jrl-
Again, we recommend to begin work using them in Origlml mans+t (SPE 14714] r%,bed in the SOcieV d Peto[eum Engineef$ .f.
the order given. tic, Aug. 19, 19S5,

. .

OCTOBER 1985 1763

Vous aimerez peut-être aussi