Gas Lift System
Gas Lift System
Gas Lift System
net/publication/319112819
CITATIONS READS
0 8,059
1 author:
Fathi Elldakli
University of Wyoming
13 PUBLICATIONS 21 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Testing, Modeling and Optimizing a Modified Design of GLV Seat View project
All content following this page was uploaded by Fathi Elldakli on 14 August 2017.
Abstract
Generally, a gas lift is a flexible, and a reliable artificial lift system with the ability to cover a wide range of production
rates. Gas lift systems are a closed system empowered by high- pressured gas. The entire process is used to reduce the
wellbore fluid pressure gradient by supplementing gas through an external source to withdraw more liquid from the
reservoir under higher drawdown. Many parameters affect a gas lift system design, such as changes in the wellhead and
bottomhole pressures (BHPs), produced fluid type, and productivity index of the wellbore. As these parameters change,
the gas injection pressure changes.
Gas lift system demands a surface compressing unit and in the well gas lifts valves (GLVs). Overall, a gas lift system is a
forgiving method of enhanced production, in other words, even a poor gas lift design may increase production. To achieve
a higher ramp in fluid production rate using gas lift, however, a more sophisticated design of each compartment of the
system is required.
Keywords: Gas lift valves (GLVs); Bottomhole Pressures(BHPs); Gas liquid Ratio (GLR)
Fathi Elldakli. Gas Lift System. Pet Petro Chem Eng J 2017, 1(4): 000121.
Copyright© Fathi Elldakli.
3
Petroleum & Petrochemical Engineering Journal
ensure that the suitable amount of gas is injected at all that does not need downhole pump. However, this system
times, gas entry must be controlled by utilizing a is unable to significantly reduce BHP. The minimum
downhole device. In modern practice, gas lift valves are required pressure gradient is approximately 0.22 psi/ft
used for downhole gas injection control. Because of its and rarely goes below 0.15 psi/ft as suggested by
importance, various gas lift valves have been developed Pablano, & Fairuzov [5]. Therefore, gas lift system is a
over the years. good candidate for projects where the BHP is constant.
Gas lift system may be used to resume production from The higher production rate may be attained when the
a well since the primary production ceases until its bottom GLV installed at the deepest point. Before starting
abandonment. Also, it may be used to increase the gas injection, the required injection pressure has to be
production from low producing wells. Initially, continuous calculated based on the hydrostatic pressure inside the
gas lift may be used to produce at high liquid rates up to tubing at the depth where the valve will be installed.
50,000 STB/D [4]. Later, as both formation pressure and
liquid rates gradually declined; replacing it with an The location of first valve must be determined for kick-
intermittent gas lift system ensures that production goals off. The high pressure gas in the annulus forces the liquid
are met. out of the tubing through U-tubing. Figure 2 demonstrates
the point of injection.
Gas lift system is the only form of artificial lift system
Fathi Elldakli. Gas Lift System. Pet Petro Chem Eng J 2017, 1(4): 000121.
Copyright© Fathi Elldakli.
4
Petroleum & Petrochemical Engineering Journal
Installing GLVs at deep depths is vitally important, and closes and lower valve starts to unload the well and so
depends on the valve design, and its opening and closing forth. Usually the lowest GLV is an open orifice plate,
pressures. The order of GLVs installation is of great which is open all the time. In the Figure 4, the gas is
importance in the lifting process. If the installation is not injected from the casing and the reservoir fluid is
in the correct order, the gas lift system will fail. In a gas produced through the tubing. Figure 5 demonstrates the
lift system design, the valves’ operating pressures drop as valve depth determination with respect to the flowing
it gets deeper. Therefore, as the injection pressure drops, tubing pressure.
the upper valve operating at a higher injection pressure
Figure 4: String design for GL system. Figure 5: the depth of the valves.
Advantages of Gas Lift gas lift system does not work well for old casing wells, or
long flow lines with small inside diameter (ID). As
In terms of production rate ranges and depth of lift, the
Fleshmanand Lekic and Osuji pointed out, gas lift process
gas lift system is flexible and can rarely be matched by
is limited to the BHP and fluid properties such as scale
other artificial lift methods if required injection gas
formation, existence of paraffin and corrosion because
pressure and volume are existed. Gas lift is one of the
these properties may increase the friction in the tubular
most flexible artificial lift techniques which even
[7-8].
unappropriated design will still lift some fluid. Highly
deviated wells with high formation GLR and sand
production are good candidates for gas lift when artificial Gas Lift Valve (GLV)
lift is needed.
GLV is the heart of a gas lift system because of
controlling production rates. The GLV is a backpressure
Limitations of Gas Lift regulator which functions based on the differential
Large well spacing for onshore wells and limited pressure between the injection gas pressure and the
offshore platforms space for compressors may restrict the production fluid pressure [9]. The GLV regulates the
gas lift application. Gas lift is rarely applicable to single pressure on its upstream side to its downstream. The
well installation and not appropriate for widely-spaced architectural design of each GLV is as important as the
wells due to the difficulty of locating the power system depth of installation and number of GLVs used in each
centrally. Gas lift is not the option for viscous crude oil, installation. Poor designs may result in installing dozens
super-saturated brine or emulsion fluid. In addition, the of GLVs to unload well that may not to be necessary.
Fathi Elldakli. Gas Lift System. Pet Petro Chem Eng J 2017, 1(4): 000121.
Copyright© Fathi Elldakli.
5
Petroleum & Petrochemical Engineering Journal
Therefore, calibrating each GLV to achieve the best the basic principles used in the design were quickly
performance at the wellbore is vital to the artificial lift adopted by almost all GLV manufacturers and are still
cycle of each well. Because a GLV consists of many used with little modification. Brown describes four types
movable mechanical compartments, achieving synergy of GLVs: casing pressure operated valve, throttling valve,
between all those compartments should result in the best fluid operated valve, and combination valve [3]. The
performance. pressure operated valve is used in most gas lift
installations. The gas lift valve has a loading element,
The performance of a GLV affects both the technical which is a spring, a nitrogen charged bellows, or a
and economic aspects of fluid lifting process. The main combination of the two. The loading element provides the
duty of a GLV is to allow the required amount of injected downward closing force in a gas lift valve.
gas enter the wellbore to unload the well. If the designed
parameters such as pressure column, water cut, GLR and Pressure Operated GLV
well deliverability change, the GLVs may be installed at
In the pressure operated valves, the valve’s behavior is
different depths to adjust the gas lift system accordingly.
controlled by injection pressure, production pressure, or
Figure 6 demonstrates the GLV at different depths.
both. GLVs are easily controlled by changing the surface
injection pressure. Designing the optimum gas lift system
is most important part of gas lift design for any
application, off-shore or on-shore. GLVs performance
must be tested to ensure sufficient gas flow enters the
wellbore to lift the predicted volume of formation fluid.
Any failure in choosing the right GLV size will result in an
ineffective gas lift system. GLVs are divided into two main
categories: injection pressure operated, IPO and
production pressure operated, PPO. The operation
mechanism of both types is the same.
Fathi Elldakli. Gas Lift System. Pet Petro Chem Eng J 2017, 1(4): 000121.
Copyright© Fathi Elldakli.
6
Petroleum & Petrochemical Engineering Journal
Each GLV is designed to stay closed until assured GLV Behavior Based on Flow Patterns
conditions of pressure in the annulus and tubing happen.
The behavior of each GLV under operating condition
When the valve becomes open, (in IPO GLVs) it allows the
strongly depends on the injection gas pressure, the port
gas or fluid to enter the tubing pass from the annulus.
size and the produced fluid pressure. The optimum
GLVs may also be arranged to permit flow from the tubing
performance of a GLV is when the valve operates under
to the annulus. Figure 8, illustrates the basic operating
critical flow conditions in which the injection gas pressure
principles involved in a GLV operation. Apparatus used to
is high enough to compensate the downstream fluid
apply energy to keep the valve closed are: (1) a metal
pressure and therefore maximum lifting production is
bellows charged with a pressurized gas, usually nitrogen;
achieved. Following are some flow behaviors through the
and/or (2) an evacuated metal bellows and a spring in
GLV:
compression. The valve operating pressure, in both cases,
is adjusted at the surface before being run into the well.
Critical flow is defined as the state when the
All GLVs have check valves to prevent back flow.
magnitude of the fluid stream velocity reaches sound
velocity at the minimum flow area where, is supposedly
the port area in a GLV. The pressure at the throat is the
critical pressure (i.e. for air flowing under isentropic
conditions this ideally 0.528 x injection pressure (𝑃𝑖 )).
Fathi Elldakli. Gas Lift System. Pet Petro Chem Eng J 2017, 1(4): 000121.
Copyright© Fathi Elldakli.
7
Petroleum & Petrochemical Engineering Journal
under two different flow patterns of throttling and orifice 3. Pressure-operated valves get opened and closed by
flow [11]. Throttling flow resembles flow through a injection and /or production pressure.
variable area venturi; whereas orifice flow is very similar
to gas flow through a fixed choke. Figure 9 depicts Mechanically controlled valves open and close from the
schematic performance curves in the gas rate against surface by some mechanical means, most commonly by
production pressure for these two flow patterns. The wire line. This type of the valves is used in intermitted lift
orifice flow model can be divided into two regions: wells; its biggest disadvantage is the unloading operation
subcritical and critical. At critical flow, flowing gas had to be accomplished manually.
velocity across the valve port equals to the sound velocity,
and injection rate remains at its maximum value The specific gravity differential valve, on the other
regardless any further decrease in tubing pressure. This hand, was used for unloading continuous flow
flow pattern occurs when the stem in the GLV at its installations. Its main element is a flexible diaphragm that
maximum travel so the port area is the smallest passing opens and closes the valve based on the pressures acting
flow area. on its two sides. A control force is provided by a liquid
charge of a set specific gravity, and the valve opens based
on the actual fluid gradient at the valve depth. This valve
proved to be an excellent option, however because of its
bulkiness was later replaced by modern types.
Fathi Elldakli. Gas Lift System. Pet Petro Chem Eng J 2017, 1(4): 000121.
Copyright© Fathi Elldakli.
8
Petroleum & Petrochemical Engineering Journal
𝑃2 = down stream pressure, psi Bigilarbigi developed the facilities and procedures for
𝑔 = acceleration of gravity, 32.2 ft/sq sec. valve testing [15]. He adjusted constant injection pressure
𝐾 = Cp/Cv spcific heat ratio test (CIPT). Based on the experimental result he built
𝑟 = 𝑃2 /𝑃1 > ro statistical model to simulate the flow performance of
ro = (2/K+1)^ k/(K-1) nitrogen and spring loaded valves and predict valve
𝑍 = comprisipility factor performance. An error in the static test facility made all
𝐺 = specific gravity of gas ( air = 1) his test data invalid.
𝑇 = inlet temperature of choke, R
Nieberding corrected the static test facility error and
Winkler and Camp studied the valve performance developed statistic models for throttling and orifice flow
considering the effect of the operating forces on bellows based on test data [11]. He used the test rack opening
load rate and linear stem travel for many valves and they pressure (𝑃𝑡𝑟𝑜 ) as a reference to distinguish between
found that a simple mathematical model to predict gas lift throttling and orifice flow. He pointed out that the valve
valves is questionable due to the effect of these factors behaves like an orifice if the injection pressure is above
[9]. Also, they introduced the Benchmark valve to obtain the test rack opening pressure. However, if the injection
the valve discharge coefficients (𝐶𝑑 )which may be used to pressure becomes lower than the test rack opening
calculate the flow rate for actual gas lift valves. pressure but greater than the valve closing pressure (𝑃𝑣𝑐 ),
the valve will behave like a throttling valve. He also used
Decker pointed out that as the flow area changes, the methane as injected gas instead of air to examine his test
𝐶𝑑 also changes and cannot be used as a constant in procedure. In addition, he developed a correlation (2-2)
Thornhill-Craver equation [14]. He also described the best to predict gas passage.
technique to perform gas lift valve performance test. He
discovered that the stem travel and opening and closing P inj ×(P inj −P pd )
Qgas = 1240.3 × Ap × Cd × Y (2)
pressures cannot be calculated from static force balance. 1−β 4 ×T inj ×Z inj ×γ g
Based on control valve capacity test procedures he used
flow coefficient (𝐶𝑣 ) concept instead of 𝐶𝑑 and presented Where,
flow coefficients as a function of stem travel. Mscf
Qgs = Gasthroughput, ;
d
Tulsa University Artificial Lift Program (TUALP) Ap = Port area, in2 ;
started to study the gas lift system in 1983. An orifice and Cd = Discharge coefficient;
throttling flow models were built to define the gas Y = Gas expansion factor;
passage through gas lift valve. The researchers created Pinj = Upstream gas injection pressure, psig ;
the flow performance family in orifice, transition and Ppd = Downstream gas injection pressure, psig ;
throttling flow as shown in Figure 10. β = Square root of ratio of port to inlet area ;
Tinj = Temperature gas injection , R ;
Zinj = Comprisibility factor at injection pressure ;
γg = Gas specific gravity .
Fathi Elldakli. Gas Lift System. Pet Petro Chem Eng J 2017, 1(4): 000121.
Copyright© Fathi Elldakli.
9
Petroleum & Petrochemical Engineering Journal
(Merla make N-15R valve) and developed empirical can be used to calculate the dome temperature by
equations quantifying bellows performance. He also following equation,
developed a model using a Quasi-one dimensional flow 𝑇𝑑 = 0.3 × 𝑇𝑝𝑑 + 0.7 × 𝑇𝑖𝑛𝑗 (4)
theory for a convergent-divergent nozzle to explain the Where,
complex high velocity gas flow through the ball seat Td = Valve dome temperature, F;
configuration. During his study he pointed out that Tpd = Production fluid temperature, F;
pressure operated valve is sensitive to performance Tinj = Injection gas temperature, F;
parameters. A small variation of 1% in injection pressure
and temperature can change the maximum flow rate by as Both, Bertovic’s and Faustinelli’s models assumed one
much as 30%. Rodriguez tested Nieberding and Acuña dimensional flow and the ball-seat area is only a function
models at various pressures operated gas-lift valves to of the dimensionless stem position, which may not be
examine these models performance [18]. He developed accurate to use in transition flow. In their models, they
empirical model and verified that constant production used only three different valve closing pressures. So the
pressure test (CPPT) and the constant injection pressure calculated coefficient is not valid for other valve closing
test (CIPT) methods are equivalent for valve performance pressures.
testing. Escalante developed a single equation model for
orifice and throttling flow performance [19]. She Shahri developed simple, rapid, and very inexpensive
described the ratio of the effective flow area (𝐴𝑒𝑓𝑓 ) to the method to test the GLV in the critical flow region [4]. Also,
port area (𝐴𝑃 ) as effective area factor ( 𝐴𝑓 ). She figured he examined the effectiveness of Thornhill-Craver (TC)
out that the relationship between 𝐶𝑑 × 𝑌 × 𝐴𝑓 and the equation and corrected the equation for the discharge
valve’s throttling behavior is linear. For throttling flow, coefficient value. Another way to model the gas flow
she introduced a new term, dimensionless stem position behavior is using numerical technique such as
(𝑁𝑆 ) which is sensitive to changes in valves closing Computational Fluid Dynamics (CFD). CFD can be defined
pressure. On the other hand, she used 𝐴𝑓 = one for orifice as the field that uses computer resources to simulate flow
flow. The final form of gas passage equation was: related problems. This field has been developing for the
P iod ×(P iod −P pd ) past 30 years. Due to the introduction of new computer
Qgs = 1240.3 × Ap × Cd × Y × Af (3) specifications, the CFD has been used extensively. Before
T v ×Z v ×S g
computational fluid dynamics was introduced, some
Where, scales were difficult to simulate.
Mscf
Qgs = Gasthroughput, ; In 1930, two dimensional (2D) methods were
d
Ap = Port area, in2 ; established to solve the linearized potential equations
Aeff using conformal transformations of the flow [22].
Af = ; Richardson is the first one who introduced the earliest
Ap
Cd = Discharge coefficient; type of calculations for modern CFD [23]. Actually, his
book "Weather prediction by numerical process," is the
Y = Gas expansion factor;
Piod = Upstream gas injection pressure, psig ; foundation for modern CFD and numerical meteorology.
He used finite differences method and divided the
Ppd = Downstream gas injection pressure, psig ;
physical space in cells. T3 group, a Consulting and
Sg = Gas specific gravity .
Software Development Company developed three-
dimensional method [24]. They used computers to model
Bertovic accomplished comprehensive research [20]. fluid flow using Navier-Stokes equations. In 1967 Hess
He performed experimental and theoretical studies of the published the first paper with three-dimensional model
flow performance for one inch GLV. The model uses six [25]. In their paper, they presented programs called Panel
coefficients to predict flow regimes, orifice, throttling and Methods which discretizes the surface of the geometry
transition. with panels.
To study GLV behavior, Faustinelli built a model for all Turzoand Takacs developed a numerical solution for
three flow regimes [21]. To calculate volumetric flow the GLV behavior that generated results in agreement
rates through the valve, the model uses only four with the API specifications [26-27]. In their approach,
empirical parameters. Furthermore, Faustinelli developed they solved five sets of equations including Navier-Stokes
a numerical model to study the effect of production and equation, state of the fluid which in compressible,
fluid temperature on bellows effectiveness. This model conservation of mass, energy, and the enthalpy changes
Fathi Elldakli. Gas Lift System. Pet Petro Chem Eng J 2017, 1(4): 000121.
Copyright© Fathi Elldakli.
10
Petroleum & Petrochemical Engineering Journal
due to change in internal energy at each position. The 14. Decker KL (1986) Computer Modeling of Gas-Lift
main caveat of this approach is associated with the Valve Performance. Offshore Technology Conference,
correct assignment of pressure distribution pattern on Houston, Texas.
the ball tip.
15. Biglarbigi K (1985) Gas Passage Performance of Gas-
References Lift Valves. Tulsa, pp: 206.
1. Brantly JE (1961) History of Petroleum Engineering. 16. Hepguler G (1988) Dynamic Model of Gas-Lift
Boyed Printing Co, Dalals, Texas. Performance. The University of Tulsa, Tulsa, pp: 179.
2. Cloud WF (1937) Petroleum Production. Norman: 17. Sagar R (1991) An Improved Dynamic Model of Gas-
University of Oklahoma, pp: 613. Lift Valve Performance. The University of Tulsa,
Tulsa, pp: 217.
3. Brown KE (1984) The Technology of Artificial Lift
Methods. Vol 2a, PennWell Publishing Company, 18. Rodriguez M (1992) Normalization of Nitrogen-
Tulsa, Oklahoma. Charged Gas-Lift Valve Performance. The University
of Tulsa, Tulsa, pp: 151.
4. Shahri MA (2011) Simplified and Rapid Method for
Determining Flow Characteristics of Every Gas-Lift 19. Escalante S (1994) Flow Performance Modeling of
Valv. Texas Tech University, USA, pp: 1-128. Pressure Operated Gas-Lift Valves. The University of
Tulsa, Tulsa, pp: 104.
5. Pablano E, Camacho R, Fairuzo YV (2002) Stability
Analysis of Continuous-Flow Gas-Lift Wells. Society of 20. Betrovic D, Doty D, Blais R, Schmidt Z (1997)
Petroleum Engineers, SPE Annual Technical Calculating Acurate Gas-Lift Flow Rate Incorporating
Conference and Exhibition, San Antonio, Texas. Temperature Effects. Society of Petroleum Engineers,
SPE Production Operations Symposium, Oklahoma.
6. Bellarby J (2009) Well Completion Design. 1st (Edn.),
56: 726. 21. Faustinelli JG, Doty DR (2001) Dynamic Flow
Performance Modeling of a Gas-Lift Valve. Society of
7. Fleshman R, Lekic HO (1999) Artificial Lift of High- Petroleum Engineers, SPE Latin American and
Volume Production. Oilfield review, p: 48-63. Caribbean Petroleum Engineering Conference,
Buenos Aires, Argentina.
8. Osuji LC (1994) Review of Advances in Gas Lift
Operations. Society of Petroleum Engineers. 22. Thomson LM (1973) Theoretical Aerodynamics. 4th
(Edn.), Dover Publications, New York.
9. Winkler HW, Camp GF (1987) Dynamic Performance
Testing of a Single Element Unbalanced Gas-Lift 23. Richardson LF (1965) Weather Prediction by
Valves. Society of Petroleum Engineers, SPE Numerical Process. Dover Publications, pp: 1-27.
Production Engineering, 2(3).
24. Harlow FH (2004) Fluid Dynamics in Group T-3 Los
10. King W (1944) Time and Volume Control for Gas Alamos National Laboratory. Journal of
Intermitters. US Patent No. 2,339,487. Computational Physics 195(2): 414-433.
11. Nieberding M (1988) Normalization of Nitrogen 25. Hess JL, Smith AMO (1967) Calculation of Potential
Loaded Gas-Lift Performance Data. The University of Flow about Arbitrary Bodies. Progress in Aerospace
Tulsa, Tulsa, pp: 267. Sciences 8: 1-138.
12. Acuna HG (1989) Normalization of One Inch Nitrogen 26. Turzo Z, Takacs G (2009) Special Report: CFD
Charged Pressure Analysis Gas-Lift Valves. Chapter techniques determine Gas-Lift Valve Performance. Oil
VI, pp: 251. & Gas Journal.
13. Decker AL (1976) Analytical Methods for 27. API (2001) Recommendation Practice Gas-lift Valve
Determining Pressure Response of Bellows Operated Performance Testing. 2nd (Edn.), API RP 11V2.
Valves. Society of Petroleum Engineers.
Fathi Elldakli. Gas Lift System. Pet Petro Chem Eng J 2017, 1(4): 000121.
Copyright© Fathi Elldakli.
11
Petroleum & Petrochemical Engineering Journal
28. API (1994) Gas Lift Manual book 6. 3rd (Edn.), 31. Encycloedia CE (2013) Coanda Effect. 6th (Edn.).
Exploration and Production Department, API, USA.
32. Kulkarni M (2005) Gas Lift Valve Modeling with
29. Brown KE (1967) Gas Lift Theory and Practice. Orifice Effects. Texas Tech University, Lubbock, Texas.
Englewood Cliffs NJ, Prentice-Hall, pp: 924.
33. Takacs G (2005) Gas Lift Manual. Pennwell
30. Economides MJ, Hill AD, Ehlig-Economides C, Zhu D Corporation.
(2012) Petroleum Production Systems. 2nd (Edn.),
Prentice Hall, pp: 1-730.
Fathi Elldakli. Gas Lift System. Pet Petro Chem Eng J 2017, 1(4): 000121.
Copyright© Fathi Elldakli.