PO186110 DOI: 10.

2118/186110-PA Date: 12-January-18 Stage: Page: 49 Total Pages: 11

Effect of R Ratio on Performance

of Injection-Pressure-Operated
Gas Lift Valves
K. L. Decker (retired), Decker Technology, Incorporated

Summary
An injection-pressure-operated (IPO) gas lift valve’s closing force is supplied by a nitrogen charge, acting on the effective area of the
bellows (Ab), or a spring force. Traditionally, the opening force is the production pressure acting on the area of the port (Ap) plus the
injection pressure acting on the bellows effective area minus the area of the port (Ab  Ap). Traditionally, the ratio Ap/Ab is referred to
as the R ratio, and has been considered constant for every valve of the same make, model, and port size. The actual R ratio is not a con-
stant and neither is it equal to the published value. The published R ratios are embedded in all gas lift design programs, and are an inte-
gral part of the gas lift design process. This paper describes the consequences of using an assumed R ratio in a gas lift design that is not
the same as the actual R ratio of the valve that will be installed in the well.

Static-Force-Balance Equation
The force from a spring or a nitrogen-charged dome, acting on the effective area of the bellows, creates a contact area between the ball
and port. When using a square-edge seat, this contact area is defined by an outer-seal diameter and the port diameter. Production pres-
sure acts on the area defined by the port diameter (Ap), and annulus pressure acts on the area defined by the outer-seal diameter (As).
The contact area is an annular ring on the surface of the port where neither production nor annulus pressure is acting. The width of this
area changes with the valve dome pressure, strength of port material, port size, and lap band. Fig. 1 shows a typical nitrogen-charged
gas lift valve, and Fig. 2 shows the typical square-edge valve/port geometry. Fig. 3 shows the typical 45 chamfer valve/port geometry.
With this refinement of the areas where pressure is applied, the static-force-balance equation for IPO valves with square-edge
ports becomes

Pd Ab ¼ Pcf ðAb –As Þ þ Ptf Ap : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð1Þ

The R ratio, as defined by numerous publications and standards (Specification 19G2 2015) and used throughout the industry, is
Ap/Ab. This definition does not account for the sealing area nor the lap band. This paper will use the following definition of the R ratio
to account for the sealing area:

R ¼ As =Ab : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð2Þ

For chamfered ports, the ball makes contact on the chamfered surface. In this case, there is an outer sealing area and an inner sealing
area that occurs as a result of a lap band. The outer diameter of the lap band defines the annulus-pressure sealing area (As). The inner di-
ameter of the lap band defines the production-pressure sealing area (Ai).
Eq. 2 is the calculated R ratio, and can be obtained from the properties of the valve. It generally is not possible to directly measure
the seal area or the lap-band width, and, as a consequence, determining the R ratio by means of Eq. 2 should be considered only
an estimate.
The most-reliable and -accurate method of determining the R ratio is by means of pressure testing. Both the opening (PvoT) and clos-
ing pressure (PvcT) tests must be performed, and the R ratio is calculated with the following equation:

R ¼ ðPvoT – PvcT Þ=PvoT : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð3Þ

During the design phase, the following two equations can be used with the R ratio as defined by either Eq. 2 or Eq. 3:

PvcT ¼ Pd ¼ Pcf ð1  RÞ þ Ptf Ap =Ab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð4Þ

PvoT ¼ Pd =ð1  RÞ: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð5Þ

It should be noted that the traditional definition of the R ratio cannot be substituted into Eq. 4 for the quantity Ap/Ab.

Discussion
The opening pressure (PvoT) is commonly used as the basis for preparing a valve for service in a well. The reason stems from the ease
with which it can be tested and from the concept that the valve acts as a backpressure regulator. The gas lift shop technician takes great
effort to ensure that the PvoT is within þ / 34 kPa (þ/5 psi) of requested. The dome pressure is normally charged to a slightly higher
pressure than intended. The PvoT is adjusted by lightly tapping the core valve of the nitrogen dome, thus releasing a small amount of
nitrogen until the desired pressure is achieved. The practical ability of adjusting a very high pressure to an accuracy less than þ/ 34

Copyright V
C 2018 Society of Petroleum Engineers

Original SPE manuscript received for review 13 October 2016. Revised manuscript received for review 24 April 2017. Paper (SPE 186110) peer approved 28 April 2017.

February 2018 SPE Production & Operations 49

ID: jaganm Time: 19:17 I Path: S:/PO##/Vol00000/170027/Comp/APPFile/SA-PO##170027

 PO186110 DOI: 10.2118/186110-PA Date: 12-January-18 Stage: Page: 50 Total Pages: 11

kPag (þ/5 psi) by the tap of a hammer on the core valve is generally not within the skill set of gas lift shop technicians. In fact, some
“local” practices may allow greater deviations.

Nitrogen dome Bellows Valve stem Port Lower packing

Upper packing Gas inlet Back-check
Gas outlet

Fig. 1—Typical injection-pressure-operated gas lift valve.

T/C ball on square-edge

MONEL seat

Outside-seal area (As )

Inside-seal area (Ai )

Port-diameter (Ap )
Deformation or
lap-band width

Fig. 2—Ball on square-edge MONELTM port.

T/C on chamfer port

Outside-seal area (As )

Contact point
h
dt
wi

Inside-seal area (Ai )

nd
ba
p-
La

Port-area (Ap )

Fig. 3—Ball on 458 chamfer tungsten carbide port.

The technician is adjusting the dome pressure (Pd) to achieve the requested PvoT. The test pressure is applied to the effective area of
the bellows (Ab) minus the area the valve seals (As) on the port. The technician does not know what area the valve seals and does not
care. If the seal area is larger than the published amount caused by deformation or a lap band, the amount of dome pressure required to
balance the requested PvoT will be less than calculated. Because the dome pressure (PvcT) is not tested, neither the gas lift well designer
nor the gas lift shop technician knows what the actual R ratio is, even though PvoT may be within specifications.
PvoT is a useful test when the valve is closed and the engineer wants to know the annulus or production pressure conditions that will
cause the valve to open. When the valve is open and flowing, PvoT has no value. When the valve is open and flowing, flow performance
is dictated by the dome pressure (Pd or PvcT). The consequences of a valve installed in a well with an R ratio larger than published will
be a PvcT lower than calculated and a higher flow rate at the design annulus pressure, and the valve will remain open at the expected

50 February 2018 SPE Production & Operations

ID: jaganm Time: 19:18 I Path: S:/PO##/Vol00000/170027/Comp/APPFile/SA-PO##170027

 PO186110 DOI: 10.2118/186110-PA Date: 12-January-18 Stage: Page: 51 Total Pages: 11

production-closing pressure. Normally, this is an advantage because most gas lift valve flow rates fall well short of the flow rate pre-
dicted with the Thornhill-Craver Method (Cook and Dotterweich 1946). It is expected that the annulus pressure will drop approximately
138–207 kPa (20–30 psi) when the next-lower valve uncovers, and this should cause the upper valve to close. Depending on the actual
R ratio of the valve, even with a 138–207 kPa (20–30 psi) annulus pressure drop, the valve could remain open.
In well-troubleshooting cases where the gas lift valves are pulled for inspection, it may be found that the PvoT of the valves is within
specifications. Because PvoT is the only criterion commonly used for judging the performance of a valve and because the PvoT is within
specifications, the engineer normally invokes a problem with the well to explain why the valve did not close as expected—temperature,
flowing tubing-pressure correlation, a bad valve, or a host of other downhole unknowns. The actual reason could be that the R ratio was
larger than expected but because closing pressure is never tested, the true R ratio of the valve is never known, and the engineer will
never know the true cause of the problem.
It might be reasoned that if all the valves in the well have similar R ratios, the system will operate at a lower annulus pressure, and
everything will be fine. But this is not the case. As the tubing pressure at the valve increases, the difference between calculated PvcT and
the true PvcT increases linearly. As a result, the valves at shallow depths may operate as expected, but as injection works deeper, the
valves will become more and more prone to remain open. It is possible that the lower valves could remain open and flowing even after
the annulus pressure drops. It may not be possible to inject a sufficient amount of gas at the surface, possibly creating issues related to
production-system stability.
A string of valves each with ports of the same size but each valve having a much different R ratio is also possible. If the valves were
designed with a common pressure drop at each depth, transfers will not occur as expected. The valves with large R ratios will remain
open, and those with normal R ratios will close when annulus pressure drops. This will lead to valves closing and reopening multiple
times. It may take much longer for the well to unload to the operating valve, and during that time, there will be a great deal of slugging
at the surface.

Example Case
Assume we are designing a 11=2-in. valve with a 4.76-mm (0.187-in.) MONELTM alloy 400 port (Special Metals Corporation1) for a
well where the annulus pressure is 10 170 kPag (1,475 psig) and the production pressure is 4480 kPag (650 psig). For these conditions,
estimate the dome pressure to be 9650 kPag (1,400 psig). From a previous study,2 the diameter that the valve seals on the port was cor-
related to the dome pressure, bellows size, and port material. By use of a graph from that study, it is found that the diameter that the
valve seals on the port is approximately 5.48 mm (0.216 in.). The R ratio for this seal diameter is As/Ab ¼ 0.0476. This R ratio is larger
than the published R ratio (0.038), and is representative of how the valve will actually perform. With the valve’s true R ratio in Eq. 4,
the dome pressure would be Pd ¼ 10,170  (1  0.0476) þ 4,480  0.178/4.97. Pd ¼ 9846 kPag (1,428 psig). By use of Eq. 5 with the
true R ratio, the opening pressure would be PvoT ¼ 9846/(1  0.0476). PvoT 5 10 338 kPag (1,499 psig).
Consider how this valve would be prepared if the engineer used the published R ratio and port area with the same design pressures.
The dome pressure is calculated as Pd ¼ 10,170  (1  0.038) þ 4,480  0.187/4.97. Pd ¼ 9952 kPag (1,443 psig). The PvoT is calculated
as 9,952/(1  0.038). PvoT ¼ 10 345 kPag (1,500 psig).
The PvoT using published R ratios is the same as when using the true R ratio, and because the gas lift shop prepares the valve solely
on the basis of PvoT, it is assumed the R ratio has no influence on performance, but that is false. When the gas lift shop technician pre-
pares this valve, the dome pressure (Pd) will be adjusted until 10 335 kPag (1,499 psig) applied on the area Ab  As opens the valve. The
technician will be preparing the valve with the valve’s true R ratio. If the closing/dome pressure had been tested, it would have been
PvcT ¼ 9846 kPag (1,428 psig), not 9955 kPag (1,443 psig). The R ratio will be 0.047, not 0.038. The calculated PvoT has very little sen-
sitivity to the R ratio, but the PvcT is very sensitive to the R ratio.
At the design annulus pressure, this valve will not throttle to a closed position when the production pressure drops to 4480 kPag
(650 psig). A computer program (Decker 2016a) was used to graph the performance of the example valve with a variety of R ratios. As
shown in Fig. 4, with an actual R ratio of 0.047, the valve will flow approximately 16 990 std m3/d (600 Mscf/D) when production pres-
sure is at 4480 kPag (650 psig). Had the R ratio actually been 0.038 as published, the valve would have throttled closed at a production
pressure of 4480 kPag (650 psig) at the design annulus pressure.
When the next-lower valve begins injecting, the annulus pressure is expected to fall. Fig. 5 shows the flow-rate performance when
the annulus pressure decreases by 110 kPa (16 psi). If the annulus pressure drops by this amount, the valve will throttle closed at the
design production-closing pressure of 4480 kPag (650 psig). This performance mimics the description of transfer to the next-lower
valve—the upper valve flows at near-maximum capacity until the next-lower valve uncovers at which time the annulus pressure will
drop by approximately 138 kPag (20 psig), and the upper valve will close. In this case, the higher R ratio actually enhanced the perform-
ance of the upper unloading valve.
The previous exercise demonstrates three principles:
• First, whether using Eq. 4 with tested R ratios or the standard force-balance equation with published R ratios, the PvoT is the same.
This will be the case for the vast majority of conditions. Because the test criterion for service in the field is PvoT, the conclusion would
be that it does not make any difference which equation or R ratio is used. This is a false conclusion. Calculation of PvoT has very little
sensitivity to the R ratio, but PvcT is very sensitive to the R ratio, and performance of the gas lift valve in the well is based on PvcT.
• Second, in the vast majority of cases, the actual R ratio of a valve will be larger than published, and as a consequence, the PvcT
will be lower than anticipated. The published R ratio assumes the valve seals at a diameter 0.152 mm (0.006 in.) larger than the
port diameter. When using MONEL alloy 400 port material, the port surface will be deformed by the force of the dome pressure
to create a contact area larger than the published amounts. When closing pressures are tested on valves being prepared for service
in the field, the tested R ratio using Eq. 3 is consistently larger than published. The reason is that both the deformation of the port
and the lap band are greater than published.
• Third, if a valve actually had the published R ratio, the performance in the well would be rather disappointing. When the R ratio is
slightly larger, the performance is actually enhanced, and the valve will perform exactly as expected. However, when the R ratio
becomes too large, the valve will remain open even with annulus-pressure drops of 138–207 kPa (20–30 psi).
During valve preparation, the gas lift technician will adjust the dome pressure to achieve the PvoT computed with the standard force-
balance equation and published R ratios. The resulting PvcT will be a function of the valve’s actual R ratio. This PvcT is compared with

1
Special Metals Corporation, headquartered in New Hartford, New York, USA, the MONEL trademark owner. The company acquired the trademark from Inco Alloys International in 1998.
In 2006, Special Metals was acquired by Precision Castparts Corporation of Portland, Oregon.
2
Decker, K. 2017. Gas Lift Valve R Ratios. Paper submitted to SPE peer review 27 June 2017.

February 2018 SPE Production & Operations 51

ID: jaganm Time: 19:18 I Path: S:/PO##/Vol00000/170027/Comp/APPFile/SA-PO##170027

 PO186110 DOI: 10.2118/186110-PA Date: 12-January-18 Stage: Page: 52 Total Pages: 11

the PvcT computed with the standard force-balance equation and published R ratio. The difference will be equivalent to the amount of
annulus pressure drop required to close the valve at production-closing pressure.

28 320
(1,000)

22 650
(800) R = 0.047 (actual)

Flow Rate [std m3/d (Mscf/D)]

16 990
(600)

11 330
(400)

5660
(200) R = 0.038 (published)

R = 0.036 (calculated)

0
0 3445 6895 10 340
(500) (1,000) (1,500)
Downstream Pressure [kPa (psi)]

Fig. 4—Example valve at design Pcf.

28 320
(1,000)

22 650
(800) R = 0.047
Flow Rate [std m3/d (Mscf/D)]

Pcf = 10 170 kPa (1,475 psi)

16 990
(600)

11 330
(400)

5660
(200)
R = 0.047
Pcf = 10 060 kPa (1,460 psi)
0
0 3445 6895 10 340
(500) (1,000) (1,500)
Production Pressure [kPa (psi)]

Fig. 5—Example valve with annulus-pressure drop.

A spreadsheet program (Decker 2016b) was created to explore the range of PvcT for a wide variety of port sizes, dome pressures, and
production pressures. A positive DPvcT indicates the PvcT calculated with the standard force-balance equation, and published R ratio
will be higher than when using the actual R ratio and Eq. 4. In other words, the valve used in the well will have a lower PvcT by the
amount of DPvcT. Fig. 6 shows the results for a 1-in. valve with a MONEL 400 alloy port with square edge. Fig. 7 shows the results for
a 11=2-in. valve with a MONEL 400 alloy port with square edge.
When the DPvcT is 103 kPa (15 psi) or less, the flow performance will be normal. When the DPvcT is 103–172 kPa (15–25 psi), the
flow performance will be greater than normal. When the lower valve uncovers and the annulus pressure drops 103–172 kPa (15–25
psi), the valve will throttle closed at the production-closing pressure. If the DPvcT is greater than 172 kPa (25 psi), the valve will not
close when annulus-pressure drops. This could lead to valve interference and multipointing.

Square-Edge Port Summary (No Lap Band). Newly manufactured valves with MONEL ports will receive a very light lap. The ma-
jority of the contact area is the result of the deformation caused by the force of the dome pressure. Valves prepared in the field may be
subject to a more aggressive lapping regimen.

52 February 2018 SPE Production & Operations

ID: jaganm Time: 19:18 I Path: S:/PO##/Vol00000/170027/Comp/APPFile/SA-PO##170027

 PO186110 DOI: 10.2118/186110-PA Date: 12-January-18 Stage: Page: 53 Total Pages: 11

1600
(232)
1-in. valve with 4.76-mm square-edge MONEL port
deformation only - No lap band

1400
(203)
2,100 kgs/sq-cm (30 ksi)
4,220 kgs/sq-cm (60 ksi)
6,330 kgs/sq-cm (90 ksi)
1200
(174)

1000
ΔPvcT [kPa (psi)] (145)

High-production pressure
800
(116)
Low-production pressure

600
(87)

400
(58)

200 Possible multipoint

(29)
Enhanced performance

Normal performance
0
4000 6000 8000 10 000 12 000 14 000 16 000 18 000 20 000 22 000
(580) (870) (1,160) (1,450) (1,740) (2,030) (2,320) (2,610) (2,900) (3,190)
Annulus Pressure [kPag (psig)]

Fig. 6—DPvcT vs. port material for 1-in. valve.

Any port material with compressive yield strengths greater than approximately 2800 kg/cm2 (40 ksi) will serve admirably in the nor-
mal pressure ranges commonly used today. At low annulus pressures and very high compressive-yield-strength port materials, the Win-
kler-Eads (Winkler and Eads 1993) or VPCTM (Valve-Performance Clearinghouse3) performance models will accurately describe
performance, but the performance may be disappointing and insufficient to unload the well. A larger R ratio than published is actually
advantageous to a certain extent.
At high pressures [greater than 13 790 kPag (2,000 psig)], deformation of the port surface will cause the R ratio to be very high. This
will cause the PvcT to be much lower than anticipated. In these cases, it is essential that Eqs. 4 and 5 be used during design and that the
R ratio be tested before delivery to the field.

Lap-Band Effects
The previous analysis and summary assumed the seal area was the result only of the deformation of the MONEL port because of the
dome pressure acting on the effective area of the bellows. When the ball and seat are lapped together, a larger seal area than for defor-
mation alone results. The larger seal area will yield a larger R ratio that will cause the PvcT of the valve to be much lower than antici-
pated. Figs. 8 and 9 show the results of an analysis to determine the effect of the lap band on the PvcT.
Fig. 10 shows the example-valve performance with a 0.51-mm (0.020-in.) lap band. When the annulus pressure is at design pressure,
the valve is flowing approximately 23 000 std m3/d (850 Mscf/D) at the design production-closing pressure. If the annulus pressure
drops 110 kPa (16 psi), the flow rate drops to approximately 15 570 std m3/d (550 Mscf/D) at the design production-closing pressure,
and when the annulus pressure drops 207 kPa (30 psi), the valve closes.
It is possible that the annulus pressure could drop 207 kPa (30 psig) when the next lower valve begins flowing, but most designs do
not plan for this much pressure drop between valves. And even if it does drop 207 kPa (30 psi), this valve will reopen with modest
increases of either the production or annulus pressure. When this valve was evaluated with no lap band but with an R ratio larger than
the published R ratio, the performance was enhanced. With a 0.51-mm (0.020-in.) lap band, there is a high probability that this valve
will remain open at the design-production closing pressure even with a drop in annulus pressure.

Square-Edge MONEL Alloy 400 Port Summary (With Lap Band)

For 1-in. valves, a lap band of 0.25 mm (0.010 in.) or less will have enhanced performance at annulus pressures up to 10 340 kPag
(1,500 psig). Beyond 10 340 kPag (1,500 psig), the amount of annulus pressure drop between valves increases significantly or there is a

3
Valve-Performance Clearinghouse (VPCTM) administered by Louisiana State University. 2014 to present. Baton Rouge, Louisiana. e-mail: [email protected].

February 2018 SPE Production & Operations 53

ID: jaganm Time: 19:18 I Path: S:/PO##/Vol00000/170027/Comp/APPFile/SA-PO##170027

 PO186110 DOI: 10.2118/186110-PA Date: 12-January-18 Stage: Page: 54 Total Pages: 11

risk of multipointing. Any lap band greater than 0.38 mm (0.015 in.) will require larger-than-normal pressure drops. When annulus pres-
sure increases beyond 11 720 kPag (1,700 psig), very large pressure drops will be required.

1600
(232)
1.5-in. valve with 4.75-mm square-edge MONEL port
deformation only - No lap band

1400
(203) 2,100 kgs/sq-cm (30 ksi)
4,220 kgs/sq-cm (60 ksi)
6,330 kgs/sq-cm (90 ksi)

1200
(174)

High-production pressure
1000
(145)
ΔPvcT [kPa (psi)]

800
(116)
Low-production pressure

600
(87)

400
(58)

200
(29) Possible multipoint
Enhanced performance

Normal performance
0
4000 6000 8000 10 000 12 000 14 000 16 000 18 000 20 000 22 000
(580) (870) (1,160) (1,450) (1,740) (2,030) (2,320) (2,610) (2,900) (3,190)
Annulus Pressure [kPag (psig)]

Fig. 7—DPvcT vs. port material for 11=2-in. valve.

The 11=2-in. valves fair better than 1-in. valves. For port sizes 4.75 mm (0.187 in.) or less with lap bands up to 0.38 mm (0.015 in.),
performance will be enhanced. A lap band of 0.51 mm (0.020 in.) will perform well at annulus pressures up to 8620 kPag (1,250 psig).
A lap band of 0.76 mm (0.030 in.) will require very large pressure drops, and will work only at annulus pressures less than 8275 kPag
(1,200 psig).

Tungsten Carbide Ports

Both STELLITETM and tungsten carbide (TC) have compressive yield strengths high enough to prevent deformation of the port surface.
However, both STELLITE (Kennamental STELLITE4) and TC require lapping to achieve a seal at the ball/port interface. As a consequence,
the seal area could actually be larger than MONEL alloy 400 ports, depending on the skill of technicians lapping the valve to the seat.
When a TC seat is used, a 45 chamfer is ground onto the port surface, and the diameter of the ball is increased. The contact diame-
ter of the ball on the port is larger than the port diameter. When the ball is lapped to the seat, the lap band extends equally to both sides
of the contact diameter. Fig. 3 illustrates typical ball-on-chamfer port geometry.
It is possible to place a TC ball on a square-edge TC seat. In those cases, a small chamfer is ground on the seat similar to a ball on
MONEL alloy 400 seat. The TC ball is then lapped to the seat surface. The resulting valve/seat geometry is the same as for a ball on a
MONEL alloy 400 square-edge seat but with a slightly larger lap band. Typically, the lap-band width is 0.25–0.51 mm (0.010–0.020 in.).
Manufacturers have chosen to use the diameter at which the ball makes contact with the chamfered seat surface as the area of the
port. Of course, this is not the actual area of the port, but it was necessary to use this diameter to continue to use the traditional defini-
tion of the R ratio (Ap/Ab) and the standard force-balance equation. Published data for TC ports are shown in Table 1. The published
data for R ratios assume that the lap-band width is nonexistent. Of course, this is not possible—a contact area of some width must exist
to maintain a seal. A computer-aided drawing of the ball on the seat shows that contact is made at slightly different diameters than pub-
lished (see Table 2).
Trying to predict the sealing area for lapped seats is difficult because of the craftsmanship of the technicians lapping the ball and
seat together. The resulting lap band can be anywhere from 0.25 mm (0.010 in.) to 1.02 mm (0.040 in.) wide. As a consequence, the
actual R ratio of TC ports will always be greater than published and will be unpredictable.

4
Kennametal STELLITE, headquartered in Pittsburgh, Pennsylvania, USA, the trademark owner.

54 February 2018 SPE Production & Operations

ID: jaganm Time: 19:18 I Path: S:/PO##/Vol00000/170027/Comp/APPFile/SA-PO##170027

 PO186110 DOI: 10.2118/186110-PA Date: 12-January-18 Stage: Page: 55 Total Pages: 11

1600
(232)
1-in. valve with 4.75-mm square-edge MONEL port
lap band

1400
(203) 0.25 mm (0.010 in.) lap band
0.51 mm (0.020 in.) lap band
0.76 mm (0.030 in.) lap band

1200
(174)

1000
ΔPvcT [kPa (psi)] (145)

800
(116)

600
(87)

400
(58)

200 Possible multipoint

(29)
Enhanced performance

Normal performance
0
4000 6000 8000 10 000 12 000 14 000 16 000 18 000 20 000 22 000
(580) (870) (1,180) (1,450) (1,740) (2,030) (2,320) (2,610) (2,900) (3,190)
Annulus Pressure [kPag (psig)]

Fig. 8—DPvcT vs. lap band for 1-in. valve.

As with the MONEL alloy 400 ports, the lap-band seal area is subject to neither upstream nor downstream pressure. But because the
ball seals on the 45 chamfered surface and not the edge of the port, the force-balance equation must be altered as follows:

Pd Ab ¼ Pcf ðAb  As Þ þ Ptf Ai : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð6Þ

Both As and Ai will be larger than the equivalent square-edge port configuration. As a consequence, a valve with a 45 chamfered
TC port will have less annulus-pressure sensitivity and more production-pressure sensitivity. Eq. 7 can be used to compute PvcT. Eq. 5
can be used to compute PvoT:

PvcT ¼ Pd ¼ Pcf ð1  RÞ þ Ptf Ai =Ab : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð7Þ

A similar analysis concerning DPvcT was performed with 45 chamfered TC ports. This analysis is correlated to the width of the lap
band. Eqs. 7 and 5 were used to compute the true PvoT, PvcT and was compared with the PvoT, PvcT by use of the standard force-balance
equation and published data. Fig. 11 shows the case for a 1-in. valve with a 4.76-mm (0.187 in.) chamfered port and a 7.94-mm (0.312-
in.) ball. Fig. 12 shows the case for a 11=2-in. valve with a 4.76-mm (0.187-in.) chamfered port and a 7.94-mm (0.312-in.) ball.

45º Chamfered TC Port Summary

The published R ratios for TC ports do not allow for lap bands. But the method of preparing TC ball-and-seat sets results in creating a
lap band to enable the valve to seal. The actual R ratio of TC ports will be larger than published. If the published R ratio is used in the
design, when the valve is prepared with the calculated PvoT, the dome pressure will be lower than calculated.
For 1-in. valves, any lap band greater than 0.25 mm (0.010 in.) and used at annulus pressures greater than 13 790 kPag (2,000 psig)
will require larger-than-normal pressure drops to close at the design production-closing pressure. Lap bands of 0.51 mm (0.020 in.) and
larger require pressure drops of 207–276 kPa (30–40 psi) to work at annulus pressures of 12 066 kPag (1,750 psig). If a port has a lap
band of 0.76 mm (0.030 in.), it is imperative that the design use Eqs. 5 and 7 and that the R ratio be tested before service in the field. In
summary, a 1-in. valve with a 45 chamfered TC port will perform best when the lap band is restricted to less than 0.38 mm (0.015 in.).
For 11=2-in. valves, lap bands up to 0.76 mm (0.030 in.) will have enhanced performance at pressures up to 12 066 kPag (1,750 psig).
If the lap band is less, annulus pressures up to 17 237 kPag (2,500 psig) are possible. If the lap band is less than 0.25 mm (0.010 in.) and
annulus pressures are less than 13 790 kPag (2,000 psig), the performance is normal—that is to say disappointing. In general, a 11=2-in.
valve with a 45 chamfered TC port has the widest range of application in the enhanced-performance bracket.

February 2018 SPE Production & Operations 55

ID: jaganm Time: 19:18 I Path: S:/PO##/Vol00000/170027/Comp/APPFile/SA-PO##170027

 PO186110 DOI: 10.2118/186110-PA Date: 12-January-18 Stage: Page: 56 Total Pages: 11

600
(87)
1.5-in. valve with 4.75-mm square-edge MONEL port
lap band

0.25 mm (0.010 in.) lap band

500 0.51 mm (0.020 in.) lap band
(72) 0.76 mm (0.030 in.) lap band

400
(58)
ΔPvcT [kPa (psi)]

300
(44)

200
(29)

Possible multipoint

100
(14) Enhanced performance

Normal performance

0
4000 6000 8000 10 000 12 000 14 000 16 000 18 000 20 000 22 000
(580) (870) (1,180) (1,450) (1,740) (2,030) (2,320) (2,610) (2,900) (3,190)
Annulus Pressure [kPag (psig)]

Fig. 9—DPvcT vs. lap band for 11=2-in. valve.

42 480
(1,500)
Flow Rate [std m3/d (Mscf/D)]

Pcf - 10 170
28 320 (1,475)
(1,000)

Pcf - 10 060
14 160
(1,460)
(500)

Pcf - 9965
(1,445)
0
0 3445 6895 10 340
(500) (1,000) (1,500)
Production Pressure [kPa (psi)]

Fig. 10—Performance of example valve with 0.51-mm (0.020-in.) lap band.

56 February 2018 SPE Production & Operations

ID: jaganm Time: 19:18 I Path: S:/PO##/Vol00000/170027/Comp/APPFile/SA-PO##170027

 PO186110 DOI: 10.2118/186110-PA Date: 12-January-18 Stage: Page: 57 Total Pages: 11

Table 1—Published data for tungsten carbide ports with 45º chamfer.

Table 2—Actual properties of tungsten carbide ports with 45º chamfer.

700
(101)
1-in. valve with 4.76-mm 45° chamfer T/C port
lap band

600
(87) 0.25 mm (0.010 in.) lap band
0.51 mm (0.020 in.) lap band
0.76 mm (0.030 in.) lap band

500
(72)

400
ΔPvcT [kPa (psi)]

(58)

300
(44)

200
(29) Possible multipoint

Enhanced performance

100
(14) Normal performance

0
4000 6000 8000 10 000 12 000 14 000 16 000 18 000 20 000 22 000
(580) (870) (1,180) (1,450) (1,740) (2,030) (2,320) (2,610) (2,900) (3,190)
Annulus Pressure [kPag (psig)]

Fig. 11—DPvcT for chamfer port 1-in. valve.

February 2018 SPE Production & Operations 57

ID: jaganm Time: 19:18 I Path: S:/PO##/Vol00000/170027/Comp/APPFile/SA-PO##170027

 PO186110 DOI: 10.2118/186110-PA Date: 12-January-18 Stage: Page: 58 Total Pages: 11

300
(44)
1.5-in. valve with 4.76-mm 45° chamfer T/C port
lap band

0.25 mm (0.010 in.) lap band

250 0.51 mm (0.020 in.) lap band
(36) 0.76 mm (0.030 in.) lap band

200
(29)

ΔPvcT [kPa (psi)]

150
(22) Possible multipoint

Enhanced performance
100
(14)

Normal performance
50
(7)

0
4000 6000 8000 10 000 12 000 14 000 16 000 18 000 20 000 22 000
(580) (870) (1,180) (1,450) (1,740) (2,030) (2,320) (2,610) (2,900) (3,190)
Annulus Pressure [kPag (psig)]

Fig. 12—DPvcT for chamfer port 11=2-in. valve.

Performance Summary
The actual R ratio of a valve will be higher than those published. The R ratio is not a constant but varies as a function of the dome pres-
sure, port size, port material, and lap-band width. Valves with TC ports will always have larger R ratios than those published, caused by
the lap-band width. The only way to accurately determine the R ratio is to test the closing pressure.
The current definition of the R ratio (Ap/Ab) must be amended to include the deformation of the port material and lap band. Because
of the variation of port materials and lap-band widths, the true R ratio (As/Ab) is a variable that changes with each valve. Publishing con-
stant R ratios is an attempt to simplify the design process, but this comes at the expense of sending valves to the field that will not per-
form as expected.
When we recognize that a seal area must exist between the ball and the seat (deformation or lap band), the standard force-balance
equation is no longer appropriate. We must begin with Eqs. 4 and 5 in our design calculations with tested R ratios if we expect to solve
gas lifting problems. When this is performed, the calculation of dome pressure and opening pressure will be correct and reflective of
how the valve will actually perform.
Using the published R ratio with the standard force-balance equation during design will lead to valves used in the field that have
lower dome pressures (PvcT) than expected. In those cases, valves will remain open at production-closing pressures, and may not close
with annulus-pressure drops.
One of the test criteria imposed by users on suppliers of gas lift valves is that the valves/ports have near-zero leak. It is reasoned that
a small leak will eventually become a large leak and that is the cause of multipointing in a well. The opposite is actually happening. To
satisfy the producers’ demand for near-zero leak, the manufacturer will use a soft MONEL 400 alloy port and then lap it to the ball,
with the end result being a very large lap band and R ratio. The valve will not leak, but it will cause gas lifting problems because of the
large lap band. Requesting a valve with near-zero leak is prudent, but this request must be accompanied with a request to test the closing
pressure and design with the true R ratio. Only then will the user be able to predict the performance of the valve.
The standard force-balance equation with published R ratios does not accurately reflect how the valve will perform, but in the case
of very small lap bands [0.76 mm (0.010 in.)] and slightly larger than published R ratios, the flow performance will be enhanced. In
those cases, the actual PvcT of the valve will be 69–138 kPa (10–20 psi) lower than calculated. The current force-balance equation with
published R ratios is misleading, but in certain circumstances and for low annulus pressures, it works.
The higher the production pressure is, the greater the DPvcT deviation is. This occurs near the bottom of the well. Valves with larger
R ratios may work satisfactorily near the top of the well where production pressures are relatively low, but that same valve will not per-
form as expected near the bottom of the well where production pressures are higher.
The PvoT has been chosen as the sole quality criterion for preparing valves for service in the field. The PvoT does not describe how
the valve will perform in the well—it describes only the opening and closing pressures. The PvcT will describe how the valve will

58 February 2018 SPE Production & Operations

ID: jaganm Time: 19:18 I Path: S:/PO##/Vol00000/170027/Comp/APPFile/SA-PO##170027

 PO186110 DOI: 10.2118/186110-PA Date: 12-January-18 Stage: Page: 59 Total Pages: 11

perform in the well. When PvoT is the sole test criteria for service in the field, our expectations will not be reflected by actual perform-
ance. We must begin testing closing pressure before valves are sent to the field if we expect to produce gas-lift well designs that perform
as intended.

Nomenclature
Ab ¼ effective area of bellows, cm2 (in.2)
Ai ¼ inner seal area of valve on port, cm2 (in.2)
Ap ¼ true area of port (D2p p=4), cm2 (in.2)
As ¼ outer-seal area of valve on port, cm2 (in.2)
Pcf ¼ annulus pressure at depth of valve, kPag (psig)
Pd ¼ pressure in nitrogen dome, kPag (psig)
Ptf ¼ production pressure at depth of valve, kPag (psig)
PvoT ¼ opening pressure at temperature with downstream atmospheric pressure, kPa (psi)
PvcT ¼ closing pressure at temperature when upstream and downstream pressures are equal, kPa (psi)

Acknowledgments
This paper would not have been readable or understandable without Jim Hall’s diligent editorial review and comments. Thanks also to
Sid Smith for reviewing and providing historical context.

References
Cook, H. L. and Dotterweich, F. H. 1946. Report on the Calibration of Positive Flow Beans Manufactured by Thornhill-Craver Company, Inc., Houston,
Texas, August 10, 1946. Kingsville, Texas: Department of Engineering, Texas College of Arts and Industries.
Decker, K. L. 2016a. API Simplified w/Custom Trim. Proprietary 2016.
Decker, K. L. 2016b. R Ratios—IPO and PPO valves.xlsx. Proprietary 2016.
Specification 19G2, Flow-Control Devices for Side-Pocket Mandrels. 2015. Washington, DC: API.
Winkler, H. W. and Eads, P. T. 1993. Applying the Basic Performance Concepts of Single-element Unbalanced Gas-lift Valves for Installation Design.
SPE Prod & Fac 8 (3): 211–216. SPE-21636-PA. https://doi.org/10.2118/21636-PA.

SI Conversion Factors
psig ¼ 6.895 kPag in. ¼ 2.54 cm sq. in. ¼ 6.452 cm2

Ken Decker has worked in the gas lift industry for 36 years with Teledyne Merla, Otis Engineering, and Decker Technology Incor-
porated. He is currently retired, but continues to enjoy the gas lift challenge, and can be expected to publish a few more papers
on the subject before he really retires.

February 2018 SPE Production & Operations 59

ID: jaganm Time: 19:18 I Path: S:/PO##/Vol00000/170027/Comp/APPFile/SA-PO##170027