You must log in to edit PetroWiki. Help with editing

Content of PetroWiki is intended for personal use only and to supplement, not replace, engineering judgment. SPE disclaims any and all liability for your use of such content. More information


Gas lift installation design

PetroWiki
Jump to navigation Jump to search

Continuous-flow gas lift is analogous to natural flow, but there are generally two distinct flowing-pressure traverses. The traverse below the point of gas injection includes only formation gas; whereas, the traverse above the point of gas injection includes both the formation and injection gases. These two distinct flowing-pressure traverses and their corresponding gas/liquid ratios (GLR) are illustrated in Fig. 1.

Design methods

There are numerous gas lift installation design methods offered in the literature. Several installation designs require unique valve construction or gas lift-valve injection-gas throughput performance. Only two design techniques are illustrated in this page:

  • A design based on a constant decrease in the operating injection-gas pressure for each succeeding lower valve (this design is essentially the same as the API gas lift design technique in RP 11V6[1])
  • An alternative design for wells requiring high injection-gas rates.

The API design can be used on the majority of wells in the US. However, when high-volume lift and high injection-gas rates are required, gas lift valve performance should be considered in the design. Both of these techniques use the simple single-element-type, unbalanced, gas lift valve with a nitrogen-charged bellows. This type of valve is the most widely used in the industry and is available from all major gas lift equipment manufacturers.

Gas lift installation design calculations are divided into two parts:

  • Determination of the gas lift valve depths
  • Calculation of the test-rack opening pressures of the gas lift valves

The opening pressures are calculated after the valve depths because the operating injection-gas and flowing-production pressures and temperatures during unloading are based on these valve depths.

The primary objective of this page is to outline in detail installation design methods for calculating the valve depths and the test-rack opening pressures of the gas lift valves that will unload a well to a maximum depth of lift for the available injection-gas volume and pressure. The unloading operations, as illustrated by the two-pen pressure recorder chart in Fig. 2, should be automatic. The static-load-fluid level was near the surface in the casing and tubing before initial unloading began. The wellhead pressure remains relatively constant during U-tubing operations before injection gas enters the tubing for the first time through the top gas lift valve. A surge in wellhead tubing pressure and a decrease in the injection-gas casing pressure occur as the depth of gas injection transfers to each lower gas lift valve. As each lower gas lift valve is uncovered, the valve immediately above closes, and the point of gas injection transfers from the upper to the lower valve. All gas lift valves above an operating valve should be closed and the valves below should be open in a properly designed gas lift installation.

Description of unloading operations

The depths of the unloading gas lift valves are calculated to unload the kill (load) fluid to the design depth of the operating valve with the injection-gas pressure and gas volume available at the wellsite. As the injection gas is initially injected into the casing annulus, the injection-gas pressure downstream of the control device on the injection-gas line increases as the load-fluid level in the casing annulus is lowered during U-tubing of the load fluid. The load fluid is transferred into the tubing through the open gas lift valves in a well with a packer, or through the open gas lift valves and lower end of the tubing in a well without a packer. Initial gas lift operations begin after the first gas lift valve is uncovered and injection gas enters the tubing at this top-valve depth.

The pressures in the casing and tubing are essentially equal at the instant a gas lift valve is uncovered. Immediately after injection gas begins to enter the tubing through the next lower gas lift valve, the injection-gas pressure in the casing begins to decrease because the newly uncovered gas lift valve is set to remain open at a lower injection-gas pressure than the unloading valve above. Less and less injection gas enters the tubing through the upper unloading valve. The injection-gas rate through the newly uncovered valve increases until the injection-gas pressure in the casing decreases to the closing pressure of the upper unloading valve. The depth of gas-injection transfer is complete when all injection gas is entering the tubing through the lower valve and all upper gas lift valves are closed. The principles of continuous-flow operation are illustrated by a pressure/depth diagram shown in Fig. 6.

As injection gas enters the tubing through a newly uncovered valve, the flowing-production pressure decreases. The injection-gas pressure in the casing begins to increase from a decreasing opening force from a lower flowing-production pressure at the valve depth and the need for stroking the valve stem to increase the injection-gas rate into the tubing for uncovering the next lower valve. The increase in the injection-gas pressure above the initial valve opening pressure at valve depth for passing the injection-gas rate to establish the flowing-production transfer pressure must be determined. This maximum injection-gas pressure required to stroke the valve stem sufficiently to pass the injection-gas rate necessary to transfer the depth of gas injection to the next lower valve depends on the gas lift valve performance. The valve port inner diameter (ID), bellow-assembly load rate, and linear stem travel control the gas lift valve performance. The design maximum injection-gas pressure for establishing the flowing-production transfer pressure from a lower valve during unloading should not result in reopening any of the upper gas lift valves. In Fig. 6, the injection gas is entering the production conduit through the fourth gas lift valve and the three upper unloading gas lift valves are closed. Although the bottom gas lift valve is open, no injection gas can enter this valve at depth D5 because the flowing-production pressure exceeds the injection-gas pressure at this depth. The flowing-pressure-at-depth traverse gradient, gpfa, above the operating gas lift valve depth, Dov , includes the injection- plus the formation-gas production, and the flowing-pressure-at-depth traverse gradient, gpfb, below Dov contains only formation-gas production.

Initial installation design considerations

Continuous-flow installation designs vary depending on whether complete and precise well data are known. Reliable inflow well performance and an accurate multiphase-flow correlation are required to establish the approximate point of gas injection in deep wells. When the well data are limited or questionable, the exact point of gas injection cannot be calculated accurately in many wells. If there is insufficient injection-gas pressure to reach the bottom of the well, a desired depth of gas injection may not be possible. If there is no change in injection-gas pressure or well conditions, the point of gas injection should remain at the maximum depth for the life of the gas lift installation.

Retrievable gas lift valve mandrels are installed (usually with dummy valves in place) in many wells before little, if any, well-production information is available. The engineer must locate these mandrels in wells before gas lift is required. The design considerations are similar for wells with a changing point of gas injection. In general, many gas lift installations are in this category, in which accurate well data are unknown or limited and the point of gas injection is unknown and/or changing as the reservoir is depleted.

Assumptions and safety factors

Safety factors are used for continuous-flow gas lift installation design with unbalanced, single-element, gas lift valves when the load rate and the gas throughput performance of the valve are not considered in the calculations. The initial gas lift valve opening pressures are based on the static force-balance equations. Safety factors allow the injection-gas and/or the flowing-production pressure to increase at valve depth, which is needed to properly stroke the valve stem and provide the equivalent port area required to pass the injection-gas rate necessary for unloading and gas lifting most wells. The following safety factors compensate for the fact that most operators set the gas lift valves to the nearest tubing joint. The actual depth of the gas lift valve is usually within 15 ft of the calculated depth.

  1. The operating injection-gas pressure used for the installation design calculations should be the average and not the maximum injection-gas pressure available at the wellsite for most wells. In special cases, a kick-off pressure can be used.
  2. The unloading daily production rate is assumed equal to the design daily production rate. Generally, the actual unloading daily production rate may be less than the design production rate and can be controlled at the surface by the injection-gas rate.
  3. No formation gas is produced during the unloading operations. The total gas/liquid ratio is based on the daily injection-gas rate available for unloading the well.
  4. The flowing-pressure-at-depth traverses above the unloading gas lift valves are assumed to be straight lines for the design calculations.
  5. The unloading flowing-temperature-at-depth traverse is assumed to be a straight rather than a curved line between an assigned unloading flowing wellhead temperature, Twhu, and the bottomhole temperature, Twsd.

The design surface unloading flowing temperature generally is assumed to be lower than the final, operating temperature. A final flowing temperature that is slightly higher than the design temperature increases the initial opening pressure of a bellows-charged gas lift valve and aids in keeping the upper valves closed while lifting from a lower gas lift valve.

  1. An assigned valve-spacing pressure differential, ΔPsD, of 20 to 60 psi across a valve for unloading is used by many gas lift design engineers. As a result, the actual minimum flowing-production pressure required to uncover the next lower unloading gas lift valve is greater by the assigned ΔPsD.
  2. The flowing-pressure traverse below the point of gas injection for locating the valve depths is normally assumed to be the static-load-fluid gradient. Once formation production occurs, the actual flowing pressure gradient decreases in most wells.

Orifice-check valve

An orifice being used for gas lifting a well should include a reverse-flow check valve. The check disk, or dart, should be closed by gravity or spring loaded. In a well with a packer, the check portion should remain closed to prevent debris from accumulating on top of the packer when this valve is below the working fluid level and is not the operating valve. An inlet screen is recommended for orifice-check valves with a small choke to prevent possible plugging. The individual openings in the inlet screen should be smaller than the choke in the orifice-check valve.

A properly designed continuous-flow gas lift installation with an orifice-check valve does not have a higher injection-gas requirement than the same well with an injection-pressure-operated gas lift valve. The injection-gas rate for lifting a well is controlled by the metering device on the injection-gas line at the surface. An orifice-check valve rather than a more expensive and complicated pressure-operated gas lift valve should be considered for the bottom valve in most continuous-flow installations.

Advantages of an orifice-check valve

The orifice-check valve is the simplest of all types of operating valves and has a very low possibility of malfunction. It can be used as a "flag" because of the change in the surface injection-gas pressure downstream of the control valve when the orifice-check valve is uncovered and becomes the point of gas injection. Fig. 3 illustrates an unloading operation using an orifice-check valve on bottom. The heading flowing wellhead tubing pressure is the result of the opening and closing of the unloading gas lift valves because of a 24/64-in. choke in the flowline and a frictional drag mechanism in the valve to prevent stem shatter. After the orifice-check valve is uncovered at approximately 3:00 a.m., there is no heading. The operating injection-gas pressure decrease is the result of low reservoir deliverability and not the gas lift system. A properly sized orifice-check valve can prevent severe heading or surging in a continuous-flow gas lift installation by ensuring a constant orifice size. No injection-gas pressure increase is required to stroke an orifice-check valve, and the orifice size is always known because it is equal to the choke size in the valve. The orifice-check valve is always open and passes gas as long as injection-gas pressure at valve depth exceeds the flowing-production pressure at the same depth. A properly sized orifice is required to control the injection-gas volume for gas lifting some wells. One application is gas lifting one zone of a dual gas lift installation with a common injection-gas source in the casing annulus. A design pressure differential of at least 100 to 200 psi across the orifice is necessary to ensure a reasonably accurate gas-passage prediction.

Disadvantage of the orifice-check valve

If the injection-gas-line pressure is high, relative to the flowing-production pressure at the orifice-check valve depth, freezing can occur at the surface if wet gas is used. The weak wells with an orifice-check operating valve will continue to consume injection gas at lower injection-gas-line pressure than stronger wells with higher flowing-production pressures at the depth of the operating orifice-check valve.

A hole in the tubing or a leaking packer is indistinguishable from an orifice-check valve during a normal, uninterrupted, continuous-flow gas lift operation. An orifice-check valve generally is not recommended for a small closed rotative gas lift system when costly makeup gas is required to charge the system after a shutdown. A properly set injection-pressure-operated gas lift valve closes after a slight decrease in the injection-gas pressure and prevents the unnecessary loss of injection gas from the casing annulus and the small high-pressure system.

Depth of the top gas lift valve

The top gas lift valve should be located at the maximum depth that permits U-tubing the load fluid from this depth with the available injection-gas pressure. If the well is loaded to the surface with a kill fluid, the depth of the top valve can be calculated with one of the following equations.

RTENOTITLE....................(1)

RTENOTITLE....................(2)

or

RTENOTITLE....................(3)

where

Dv1 = depth of top valve, ft,
Pko = surface kick-off or average field injection-gas pressure (optional), psig,
Pwhu = surface wellhead U-tubing (unloading) pressure, psig,
ΔPsD = assigned spacing pressure differential at valve depth, psi,
gls = static load (kill)-fluid pressure gradient, psi/ft,
and
ggio = injection-gas pressure-at-depth gradient, psi/ft.

Eq. 1 does not include the increase in the injection-gas pressure to the valve depth, Dv1. This equation is widely used because of a safety factor from neglecting this increase in gas pressure with depth. Eq. 2 yields the same depth as a graphical solution without any pressure drop across the top gas lift valve at the instant this valve is uncovered. In other words, the top valve is not uncovered if the actual kick-off injection-gas pressure is less than the design value or if the U-tubing wellhead pressure is higher than assumed. Eq. 3 includes injection-gas column weight and an assigned pressure differential at the instant the top valve is uncovered.

The surface U-tubing wellhead pressure is less than the flowing wellhead pressure for most installations. The difference between these two pressures increases for longer flowlines and higher production rates. The wellhead U-tubing pressure is approximately equal to the separator or production-header pressure because the rate of load fluid transfer is very low during the U-tubing operation and no injection gas can enter the flowline until the top gas lift valve is uncovered. Gas lift operations do not begin until injection gas enters the production conduit through the top valve. Flowing wellhead pressure should be used to locate the depths of the remaining gas lift valves.

A load-fluid traverse based on gls can be drawn from the wellhead U-tubing pressure to the intersection of the kick-off injection-gas pressure-at-depth curve (PkoD traverse) on a pressure/depth plot. The top valve may be located at this intersection, which is the same depth as calculated with Eq. 2. An arbitrary pressure drop across the top gas lift valve can be assumed in conjunction with the graphical method, and this technique is the same as Eq. 3. If no gas pressure increase with depth is assumed, this method becomes similar to the calculation of Dv1 with Eq. 1. For simplicity, Eq. 4 is often used for top-valve spacing calculations.

RTENOTITLE....................(4)

Flowing pressure at depth

Accurate flowing-pressure-at-depth predictions are essential for good continuous-flow gas lift installation design and analysis. When computer programs for gas lift installation design and analysis are unavailable for daily routine calculations, the gas lift designers must rely on published gradient curves to determine flowing pressures at depth. Many oil-producing companies have their own multiphase-flow correlations and publish in-house gradient curves. Gradient curves are available from the gas lift manufacturers and are published in books that can be purchased. Where possible, use field data to verify the accuracy of the computer program calculations and gradient curves. It is not the purpose of this chapter to compare the various multiphase-flow correlations or published gradient curves.

The widely accepted multiphase-flow correlations and mechanistic models are based on pseudo-steady state flow without serious heading through a clean production conduit with an unrestricted cross-sectional area. Accurate pressures cannot be obtained from gradient curves based on these correlations if the conduit is partially plugged with paraffin or scale. Emulsions also can prevent the application of these correlations and gradient curves. The applicability of a particular correlation or set of gradient curves for a given well can be established only by comparing a measured flowing pressure to a pressure at depth determined from the correlation or gradient curves. The measured production data must be accurate and repeatable before discounting the multiphase-flow correlations or gradient curves.

A set of typical gradient curves is given in Fig. 4. These gradient curves are used in the example installation design calculations in Example 1. GLR and not gas/oil ratio (GOR) is used for these installation design calculations.

Most gradient curves display GLR rather than GOR. For this reason, the first step in the application of gradient curves is to convert GOR to GLR, if only GOR is reported and the well produces water. The GLR can be calculated for a given GOR and water cut with Eq. 5.

RTENOTITLE....................(5)

where

Rglf = formation gas/liquid ratio, scf/STB,
fo = oil cut (l – fw), fraction,
and
Rgo = gas/oil ratio, scf/STB.

Example 1

Given:

  • Rgo = 500 scf/STB
  • Water cut fw = 0.60 (60%)

Calculate the formation GLR: Rglf = (1 – 0.6) 500 = 200 scf/STB.

When gradient curves are used, the depth is a relative depth and may be shifted, whereas pressure is never shifted. If a flowing-pressure-at-depth traverse is being traced, the pressures on the pressure/depth plot must always overlie the same pressures on the gradient curves. For deviated wells where friction is small, use true vertical depths rather than measured depths in a graphical design.

Flowing temperature at depth

The accurate prediction of the flowing-production fluid temperature at valve depth is important in the design and analysis of many gas lift installations with nitrogen-charged gas lift valves. The temperature of a wireline-retrievable valve is assumed to be the same as the temperature of the flowing fluids at the valve depth. A retrievable gas lift valve is located in a mandrel pocket inside the tubing and is in contact with the production from the well. The temperature of a conventional valve is between the flowing fluid temperature and the geothermal temperature for the well but is normally closer to the flowing fluid temperature because steel has higher thermal conductivity than gas.

Kirkpatrick[2] published one of the most widely used flowing-temperature-gradient correlations in 1959. The family of flowing-temperature-gradient curves in Fig. 5 is based on data from high-water-cut wells being produced by gas lift through 2 7/8-in.-OD tubing over a wide range of production rates. Although the correlation does not include several important parameters, such as GLR and fluid properties, the estimated surface temperature and temperatures at depth have proved to be reasonably accurate for many gas lift operations. Sagar et al.[3] published another flowing-temperature correlation. This empirical method for calculating flowing-temperature profiles is far more rigorous and is based on well data from several areas. The calculation procedure can be programmed easily for predicting surface flowing temperatures in vertical and inclined wells. However, the best approach, when possible, is to measure the temperature-at-depth traverse in the actual gas lift well.


Nomenclature

Dv1 = depth of top valve, ft
fo = oil cut, fraction
fw = water cut, fraction
Fp = production-pressure factor, dimensionless
ggio = static injection-gas pressure at depth gradient, psi/ft
glc = average pressure gradient for liquid production in chamber, psi/ft
gls = static load (kill)-fluid pressure gradient, psi/ft
PbvD = nitrogen-charged bellows pressure at valve temperature, psig
Pko = surface kick-off or average field injection-gas pressure (optional), psig
Ppfd = flowing-production pressure at Dd based on design qlt and Rglu, psig
PpfD = flowing-production pressure at valve depth, psig
Pwhu = wellhead U-tubing unloading pressure, psig
ΔPsD = assigned spacing pressure differential at valve depth, psi

References

  1. API RP 11V6, Recommended Practice for Design of Continuous Flow Gas Lift Installations Using Injection Pressure Operated Valves, second edition. 1999. Washington, DC: API.
  2. Kirkpatrick, C.V. 1959. Advances in Gas Lift Technology. Drill. & Prod. Prac. (March): 24.
  3. Sagar, R., Doty, D.R., and Schmidt, Z. 1991. Predicting Temperature Profiles in a Flowing Well. SPE Prod Eng 6 (4): 441-448. SPE-19702-PA. http://dx.doi.org/10.2118/19702-PA.

Noteworthy papers in OnePetro

Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read

External links

Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro

See also

Gas lift system design

Gas lift installation design methods

Intermittent-flow gas lift installation design

Gas lift for unusual environments

Gas lift

PEH:Gas_Lift

Category