Gas Lift Manual TOC Sample
Gas Lift Manual TOC Sample
Gas Lift Manual TOC Sample
GAS
LIFT
M A N U A L
TakacsFM (i-xvi).qxd 6/23/05 9:54 AM Page vii
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xv
1. Introduction to Gas Lifting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
1.1 Artificial Lifting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
1.2 Short History of Gas Lifting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
1.3 Basic Features of Gas Lifting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8
2. Production Engineering Fundamentals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9
2.2 Properties of Oilfield Fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9
2.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9
2.2.2 Basic thermodynamic properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10
2.2.3 Liquid property correlations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12
2.2.3.1 Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12
2.2.3.2 Crude oil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
2.2.4 Properties of natural gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15
2.2.4.1 Behavior of gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15
2.2.4.2 Gas property correlations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18
2.3 Inflow Performance of Oil Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21
2.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21
2.3.2 Basic concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21
2.3.3 The productivity index concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23
2.3.4 Inflow performance relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
2.3.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
2.3.4.2 Vogel's IPR correlation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
2.3.4.3 Fetkovich's method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26
2.4 Single-phase Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26
2.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26
2.4.2 Basic principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27
2.4.3 Pressure drop calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
2.4.3.1 Single-phase liquid flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
2.4.3.2 Single-phase gas flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32
TakacsFM (i-xvi).qxd 6/23/05 9:54 AM Page viii
CONTENTS | ix
CONTENTS | xi
CONTENTS| xiii
Preface
The first use of air to lift water to the surface was reported in the mines of Chemnitz, Hungary around the middle
of the 18th century. First in the wells of Pennsylvania, at around 1864, oil was lifted by compressed air that was later
displaced by natural gas as the lifting medium. Throughout its almost 150-year long history, gas lifting proved to be
one of the most popular methods of lifting liquids from wells. It can be applied in oil wells with high gas production
rates where other artificial methods are plagued with frequent failures or cannot be used at all. It is especially suited to
the offshore environment where extremely high liquid volumes have to be lifted.
Gas lifting can be used throughout the whole lifespan of an oil well: from the time it dies until its abandonment.
At early production times, higher liquid rates are achieved by continuous flow gas lift. As the field depletes and
formation pressure and liquid rates decrease, easy conversion to intermittent gas lift ensures that production goals are
met. Close to well abandonment, another version of gas lifting—chamber lift—can be applied. Because of these
features, gas lifting is probably the most flexible means of artificial lifting available today.
I wrote Gas Lift Handbook to become a handbook of up-to-date gas lift theories and practices and to cover the latest
developments in this important field of artificial lifting technology. Since I present a complete review of gas lift
technology and include references to all of the important literature sources, the practicing engineer can use the book
as a reference on the subject. I tried to distill in this book all the experience I gathered during my 30-year teaching and
consulting career in order to present a text that systematically introduces the reader to the subject matter. It would be
a personal gratification if, like its predecessor Modern Sucker-Rod Pumping published by PennWell in 1993, this book, too,
would be chosen for graduate level courses at different petroleum engineering schools.
It is fully understood by anyone in the industry that describing multiphase flow in oil wells is the basis of solving
most of the problems in gas lifting. Since this is an area where almost revolutionary achievements happened in the last
20 years, I fully describe the pressure drop calculation procedures for vertical, inclined, and horizontal wells including
the latest mechanistic models. The chapter on production engineering fundamentals includes, in addition to a very
detailed treatment of multiphase flow, a review of fluid properties, well inflow performance, basic hydraulics, well
temperature, and systems analysis.
Further chapters systematically introduce the reader to the hardware of gas lifting, reflecting the latest
developments in gas lift valve and other equipment designs. The great variety of gas lift valves, their constructional and
operational details are fully discussed along with the latest achievements on describing their dynamic performance. The
description of gas lift installation types helps the engineer select the right combination of well equipment. The chapter
on continuous flow gas lift fully describes the different ways to optimize the wells’ operation, including the latest
optimization theories (lift gas allocation to wells, systems analysis, etc.). Unloading and surface control of continuous
flow gas lift wells round up this chapter.
TakacsFM (i-xvi).qxd 6/23/05 9:54 AM Page xvi
The discussion of intermittent gas lift includes conventional, chamber, and plunger-assisted installations and
describes the performance, design, and optimization of such wells. A detailed treatment of the surface gas lift system
follows, including the operation and design of the complete system consisting of the compressor station, and the
distribution and gathering facilities. The last chapter includes practical advice on the analysis and troubleshooting of
gas lift installations. All necessary calculations are fully discussed, and the many charts in the appendices are intended
to help field engineers.
Nowadays, personal computers belong to every petroleum engineer's desk and this book was designed with this
fact in mind. Since most design and analysis problems in gas lifting are too complex to be solved by the conventional
tools of the engineer, I heavily relied on computerized solutions in the text.
While writing this book I burned a lot of midnight oil and many times had regretfully neglected those I love
most—my family. Their patience and understanding is appreciated.
This is already the third project I have worked on with PennWell. Ms. Marla Patterson (PennWell) and Ms. Sue
Rhodes Dodd (Amethyst Enterprises) were always most helpful and forgiving. A special “thank you” to both of them.
Gábor Takács
July 2005
Chapter 01 (1-8).qxd 6/20/05 11:25 AM Page 1
Wells may die for two main reasons: either their flowing bottomhole pressure drops below
the total pressure losses in the well, or the opposite happens and pressure losses in the well
become greater than the bottomhole pressure needed for moving the wellstream to the surface.
The first case occurs when a gradual decrease in reservoir pressure takes place because of the
removal of fluids from the underground reservoir. The second case involves an increasing flow
resistance in the well, generally caused by (a) an increase in the density of the flowing fluid as a
result of decreased gas production; or (b) various mechanical problems like a small tubing size,
downhole restrictions, etc. Surface conditions, such as separator pressure or flowline size, also
directly impact total pressure losses and can prevent a well from flowing.
Artificial lifting methods are used to produce fluids from wells already dead or to increase the
production rate from flowing wells; and several lifting mechanisms are available to choose from.
One widely used type of artificial lift method uses a pump set below the liquid level in the well to
increase the pressure so as to overcome flowing pressure losses that occur along the flow path to
the surface. Other lifting methods use compressed gas, injected periodically below the liquid
present in the well tubing and use the expansion energy of the gas to displace a liquid slug to the
surface. The third mechanism works on a completely different principle: instead of increasing
the pressure in the well, flowing pressure losses are decreased by a continuous injection of high-
pressure gas into the wellstream. This enables the actual bottomhole pressure to move well fluids
to the surface.
Although all artificial lift methods can be distinguished based on the previous three basic
mechanisms, the usual classification is somewhat different and is discussed here.
Chapter 02 (9-168).qxd 6/20/05 11:26 AM Page 17
Example 2–4. Find the deviation factor for a natural gas at p = 1,200 psia (8.28 MPa) and T = 200 ˚F
(366 K), if the pseudocritical parameters are ppc = 630 psia (4.35 MPa) and Tpc = 420 ˚R (233 K).
Solution
The pseudoreduced parameters are
ppr = 1,200 / 630 = 1.9
Tpr = ( 200 + 460 ) / 420 = 1.57
Using these values, from Figure 2–2:
Z = 0.85
The Engineering Equation of State enables the direct calculation of volume factors for gases. Equation 2.15 can be
written for a given number of moles in the following form:
pV
ZTa ( )
pV
—–– = —––
ZTa sc
2.19
This equation can be solved for Bg, which is the ratio of actual volume to the volume at standard conditions:
V pscZTa
Bg = —— = ——— 2.20
Vsc pZscTsc
Substituting into this the values psc = 14.7 psia, Tsc = 520 ˚R, and Zsc = 1, one arrives at
ZT
Bg = 0.0283—––a 2.21
p
Example 2–5. What is the actual volume of the gas, if its volume measured at standard conditions is
1.2 Mscf (33.9 m3)? Other data are identical to those of Example 2–4.
Solution
The volume factor of the gas, from Equation 2.21:
Bg = 0.0283 0.85 ( 200 + 460 ) / 1200 = 0.013
Actual volume is found from Equation 2.6:
V(p,T) = Bg Vsc = 0.013 1200 = 15.6 cu ft (0.44 m3).
Gas Density
The fact that gas volume factor is an explicit function of state parameters allows a direct calculation of gas density
at any conditions. Based on the definition of volume factor, gas density can be expressed from Equation 2.7, and after
substituting the formula for gas volume factor, we get
ρsc 0.0764γgp p
ρ(p,T) = —— = ——–—— = 2.7γg —— 2.22
Bg 0.0283ZTa ZTa
The previous formula is used to find the actual density of natural gases at any pressure and temperature based on
the knowledge of their specific gravities and deviation factors.
Chapter 02 (9-168).qxd 6/20/05 11:26 AM Page 43
8 9 10 11 12 13 14 16 18 20
C h o k e S i z e, 6 4 t h o f a n I n c h
Downstream P r e s s u r e, p s i g
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
200
300
400
500
600
700
800
900
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
Upstream Pressure, psig
Fig. 2–12 Gas throughput capacity chart for different choke sizes.
Appendix C contains two gas capacity charts for different ranges of choke sizes, as well as a correction chart that
can be used to correct gas volumes at temperatures different from chart base.
Example 2–18. Find the gas volumes for the cases given in the previous example with the use of the gas
capacity chart in Figure 2–12.
Solution
For case one, start at an upstream pressure of 1,014 psia (7 MPa) and go vertically until crossing the curve
valid for a downstream pressure of 814 psia (5.6 MPa). From the intersection, draw a horizontal to the left to
the proper choke size (16/64 in.). Drop a vertical from here to the upper scale to find the gas flow rate of
1,140 Mscf/d (32,281 m3/d).
Since actual flow conditions differ from chart base values, a correction must be applied. According to
Equation 2.65, the actual gas flow rate is
21.25
qact = 1,140 0.9 –—–—– = 1,140 0.96 = 1,094 Mscf/d (30,978 m3/d).
–—––——
√ 0.65610
The second case involves critical flow and the vertical line from the upstream pressure should be drawn
to the upper boundary line of the downstream pressures constituting critical flow. Using the same procedure
as in the first case, a gas flow rate of 1,360 Mscf/d (38,511 m3/d) is read from the chart. Since flow conditions
are similar to case one, again a correction factor of 0.96 is used to find the actual flow rate:
qact = 1,360 0.96 = 1,305 Mscf/d (36,953 m3/d).
Chapter 02 (9-168).qxd 6/20/05 11:26 AM Page 66
The flow pattern map used by the authors is given in Figure 2–19. The coordinate axes are the gas and liquid velocity
numbers defined before. As seen, four flow patterns are distinguished: bubble, slug, transition, and mist. The boundaries
of the flow patterns (shown in the figure for air-oil flow in an 8-cm pipe) are found from the following equations:
Ngv = L1 + L2 Nlv bubble-slug boundary 2.102
2.2
2.0
1.8
1.6
1.4
1.2 L2
1.0 L1
0.8
0.6
0.4
10 100 1000
Nd
Functions L1 and L2 previously are evaluated from Figure 2–24 in the function of the pipe diameter number Nd.
According to Duns-Ros’ general approach, measured liquid holdup data were correlated by using the dimensionless
groups defined previously as independent, and the dimensionless slip velocity number as the dependent parameter. The
latter is defined as follows:
—– —–
√ √
4 ρl 4 ρl
S = vs ––– = 1.938vs ––– 2.105
gσl σl
After calculating the dimensionless slip velocity number, the real slip velocity vs can be expressed from the previous
equation:
—–
√
ρl
4
vs = 0.52S ––– 2.106
σl
As derived in Two-Phase Flow Concepts (Section 2.5.2.1), knowledge of the slip velocity enables one to find liquid
holdup (see Equation 2.78) from the formula reproduced here:
—————––––
√
vs – vm + (vm – vs)2 +4vsvsl
ε l = ———————————– 2.107
2vs
Chapter 02 (9-168).qxd 6/20/05 11:26 AM Page 85
Practically identical to the Orkiszewski model. The only difference is in slug flow, where the authors eliminated the discontinuities in
liquid holdup occurring when the original model is applied.
Chierici et al. [55] adopted the Orkiszewski model for the prediction of flow patterns and the calculation of
pressure gradients in all but the slug flow pattern. Therefore, this correlation can be regarded as a slightly modified
version of the original. For this reason, only the modifications are detailed as follows.
In slug flow, Chierici et al. used the drift-flux approach for the determination of the liquid holdup. According to
this theory, liquid holdup is found from the following basic equation (see Equation 2.76):
vsg
ε l = 1 ———— 2.152
C0vm + vb
The authors assumed the distribution factor to be C0 = 1, and calculated the bubble rise velocity vb as suggested
by Griffith and Wallis [52]:
—–
vb = C1C2 √ gd 2.153
As with the Orkiszewski model, C1 is found from Figure 2–34. When finding the other parameter C2, the authors
showed that the method proposed by Orkiszewski is defective. There are no problems in the NRel ≤ 6,000 range, but
at higher Reynolds numbers the extrapolation functions given by Orkiszewski (see Equations 2.140 –2.143) result in
discontinuities.
To prevent the discontinuities caused by the Orkiszewski extrapolations of Figure 2–35, Chierici et al. proposed
the use of the following formula:
1
C2 = –————— vm 2.154
1 – 0.2 —–
vb
The authors proved that the use of the previous equation eliminates the discontinuities of the calculated C2 values
and ensures the convergence of pressure gradient calculations in slug flow.
The first correlation developed for all pipe inclination angles was based on a great number of data gathered
from a large-scale flow loop. Flow patterns are determined for a horizontal direction only and are solely used as
correlating parameters. Its main strength is that it enables a simple treatment of inclined wells and the description of
the well-flowline system.
Chapter 02 (9-168).qxd 6/20/05 11:26 AM Page 108
Gas core liquid entrainment is limited to values E ≤ 1.0 since part of the total liquid content is situated in the liquid
film covering the pipe wall. The definition of the critical vapor velocity vcrit figuring in the previous formulas is
——–
√
vsgµg ρg
vcrit = 104 × –—— –—– 2.240
σl ρl
µg = gas viscosity, cP
The elevation term can now be calculated from Equation 2.222, by substituting into the mixture density the gas
core density defined as follows:
The frictional pressure drop occurs due to the high-velocity flow of the gas core on the wavy liquid film covering
the pipe inside wall. Since the thickness of this film is not very significant (typically less than 5% of the pipe diameter),
the basic formula to be used in this special case neglects the changes in diameter:
dp ρ vsg 2
( )
—– = 1.294 × 10–3 fc —c ———–
dl f d 1 – λlc ( ) 2.242
In the previous formula, friction factor in the liquid film is found, according to Hasan and Kabir from the formula:
——––—
√
4 µg
fc = 0.024 (1 +75λlc) –––—— 2.243
ρgvsgd
µg = gas viscosity, cP
Gas core density being defined by Equation 2.240, frictional pressure gradient is easily found from Equation 2.241.