You are on page 1of 44

Enhanced Gas Lift Valve Performance for Sharp Edged Seat Using Larger Ball Sizes

A Thesis

By

Fathi Elldakli, BSc, MSc

In

Petroleum Engineering

Submitted to the Graduate Faculty


of Texas Tech University in
Partial Fulfillment of
the Requirements for
the Degree of

Master of Science

Approved

Dr. Mohamed Soliman


Chair of Committee

Dr. Mehdi Shahri

Dr. Talal Gamadi

Dr. Seyedhossein Emadibaladehi

Mark Sheridan
Dean of the Graduate School

December, 2015
Copyright 2015, Fathi Elldakli
Texas Tech University, Fathi Elldakli, December 2015

ACKNOWLEDGMENTS

First, I must acknowledge my unlimited thanks to Allah. I am totally sure that this work
would have never become truth, without His guidance.
I would like to express my sincere gratitude to my advisor, Professor Mohamed Soliman,
for his guidance, support, and advice he provided throughout this work. I found myself
extremely lucky to have a supervisor who cared so much about my work, and who
responded to my inquiries so promptly.
I would like to thank Dr. Mehdi Shahri for his help in running the equipment and the
viable information he offered.
I would like to thank Dr. Talal Gamadi for his encouragement and recommendations.
I would also like to acknowledge Dr. Seyedhossein Emadibaladehi for his suggestions
and support.
I am appreciative of Science and Technology - Landmark for their unlimited help to build
the model.
Last but not least, I would like to thank my family for their support and patience
throughout my entire life and particularly through the process of pursuing the Master’s
degree.

ii
Texas Tech University, Fathi Elldakli, December 2015

TABLE OF CONTENTS

ACKNOWLEDGMENTS ........................................................................................ii
ABSTRACT ..........................................................................................................iv
LIST OF TABLES ................................................................................................. v
LIST OF FIGURES ...............................................................................................vi
NOMENCLATURE .............................................................................................. vii
CHAPTER 1 : INTRODUCTION ........................................................................... 1
CHAPTER 2 : LITERATURE REVIEW ................................................................. 2
CHAPTER 3 : GLV TESTING ............................................................................... 8
CHAPTER 4 : RESULTS .................................................................................... 22
CHAPTER 5 : MODELING ................................................................................. 25
CHAPTER 6 : CONCLUSIONS .......................................................................... 31
CHAPTER 7 : RECOMMENDATIONS ............................................................... 32
REFERENCES ................................................................................................... 33
APPENDIX ......................................................................................................... 34

iii
Texas Tech University, Fathi Elldakli, December 2015

ABSTRACT

The required stem travel to achieve minimum fully open position is dependent on: stem
size, geometry and ball size. Six different port diameters (3/16”, 1/4”, 5/16”, 3/8”, 7/16”
and 1/2") can be configured inside 1.5” gas lift valves (GLV) in which the ball size is
1/16” in diameter larger than the port size.
Laboratory testing for sharp edged seats showed that the actual flow area is smaller than
the theoretically calculated area. This is due to Bellows stacking before the stem reaches
its fully open position. Consequently, the valve stem and the ball partially restrict the
flow causing the flow rate through the valve to decline.
The purpose of this work is to examine the possibility of improving the efficiency of the
GLV by using a larger ball size than conventionally used. For each port, different ball
sizes were tested at different stem positions under the same injection pressure &
temperature.
Results obtained from benchmark valve testing displayed an increasing trend in the flow
rate as the ball size increases under the same stem travel.

iv
Texas Tech University, Fathi Elldakli, December 2015

LIST OF TABLES

3-1 GLV port and ball sizes specifications ....................................................................... 13


3-2 Minimum stem travel for sharp edged seat with various ball sizes ............................ 17
3-3 Port tests ..................................................................................................................... 20

v
Texas Tech University, Fathi Elldakli, December 2015

LIST OF FIGURES

2-1 Brear oil ejector ............................................................................................................ 2


2-2 Well with gas lift system .............................................................................................. 4
2-3 Early gas (air) sift without valves ................................................................................. 4
2-4 Jet collar........................................................................................................................ 5
2-5 Kick-off valve ............................................................................................................... 5
3-1 Gas-Lift Valve - main parts .......................................................................................... 9
3-2 Gas-Lift Valve, IPO and PPO respectively ................................................................ 10
3-3 GLV operations .......................................................................................................... 11
3-4 Flow area .................................................................................................................... 11
3-5 Flow regimes .............................................................................................................. 12
3-6 GLV seat and stem ..................................................................................................... 13
3-7 Flow area .................................................................................................................... 14
3-8 Effect of the dome charged-pressure on the bellows staking ..................................... 15
3-9 Benchmark valve ........................................................................................................ 16
3-10 Effect of the ball size on the effective area (frustum area) ....................................... 18
3-11 Schematic of blow-down dynamic test facility ........................................................ 20
3-12 Recorded data during the test ................................................................................... 21
3-13 Curve fitting techniques............................................................................................ 21
4-1 Effect of the ball size on the flow area for 1/4" Sharp Edged Seat ............................ 22
4-2 Effect of the ball size on the flow rate for 1/4" Sharp Edged Seat ............................. 23
4-3 effect of the ball size on the stem travel ..................................................................... 23
4-4 Coanda Effect ............................................................................................................. 24
5-1 Computational Fluid Dynamic procedure .................................................................. 26
5-2 Model domain ............................................................................................................. 27
5-3 Model meshing (grids) for 1/4" port with 5/16" ball size ........................................... 28
5-4 Model validation for 1/4" port with 5/16" ball size .................................................... 28
5-5 Model validation for 1/4" port with 9/16" ball size .................................................... 29
5-6 Pressure distribution inside the GLV.......................................................................... 29
5-7 Effect of the ball size on stem travel and flow rate .................................................... 30
A-1 Effect of the ball size on the flow area for 5/16" port ............................................... 34
A-2 Effect of the ball size on flow rate for 5/16" port ...................................................... 35
A-3 Effect of the ball size on flow area for 3/8" port ....................................................... 35
A-4 Effect of the ball size on the flow rate for 3/8" port .................................................. 36

vi
Texas Tech University, Fathi Elldakli, December 2015

NOMENCLATURE

BHP Bottom hole pressure, psig

CCPT Constant casing pressure test, psig

CPPT Constant production pressure test, psig

CFD Computational Fluid Dynamics

DAQ Data acquisition system

GLV Gas lift valve

GLR Gas liquid ration, SCF/STB

IPO Injection pressure operated

Pinj Injection pressure, psig

Pp Production pressure, psig

Ppc Closing production pressure, psig

Ptran Transition pressure, psig

rb Ball radius, inch

rp Port radius, inch

Y Theoretical stem travel, inch

vii
Texas Tech University, Fathi Elldakli, December 2015

CHAPTER 1 : INTRODUCTION

The objective of this work is to examine effects of the ball size on dynamic flow
performance of each gas lift valve. This study was performed both experimentally and
numerically.
If the reservoir pressure is not high enough to provide acceptable flow rate in the well,
artificial lift methods can be used to increase the flow rate. Artificial lift is a method used
to reduce the wellbore Bottom Hole Pressure (BHP) to produce at a higher production
rate. Artificial lift methods may be used to produce from a well that is no longer
producing or to increase the flow rate from a well that is producing at low rate.
The artificial lift methods are classified in two categories; pump- lift and gas-lift. The
gas-lift technology is based on injecting gas into the lower part of the production tubing,
to decrease the hydrostatic pressure in the well. Gas-lift system seeks to reduce the
backpressure in the wellbore caused by flowing fluids in the production tubing which
leads to increase the inflow from the reservoir, consequently the production rate. Gas lift
is a flexible method which can be easily modified to the changing conditions of a well.
In order to inject gas from the casing-tubing annulus into the tubing or vice versa, a gas
lift valve is used. The gas lift valve is a critical element of the gas lift system, because it
controls the amount of the gas entering the production string. Flow area, area generated
by stem movement, is the most important factor on gas lift valve flow performance.
Theoretical based calculations showed that the minimum stem travel using larger ball size
improves the stem travel when compared to the conventional ball size. This improvement
should have rather a higher impact on the GLV performance. Therefore, in this study,
three sharp-edged seats (orifice ports) are tested with various ball sizes to investigate
effects of the ball size.
To study effects of the ball size on the pressure distribution through the gas lift valve,
numerical model is built. The model developed is based on experimental results.
The model helped to justify the experimental results and proved that the flow rate
increases using larger ball size.

1
Texas Tech University, Fathi Elldakli, December 2015

CHAPTER 2 : LITERATURE REVIEW

The artificial lift method of gas-lift is required for producing many millions of barrels of
oil per day, worldwide. Gas lift system was applied to lift water in Chemnitz, Hungary in
the mid-18th century (Shaw, 1939). This system used single point injection of air into the
liquid stream. The first practical application of air lift was in 1846 when an American
named Cockford lifted oil from some wells in Pennsylvania. The first US patent for gas
lift called an “oil ejector” was issued to A. Brear in 1865 as shown in figure 2-1.
Air lift continued until the mid-1920’s when gas lift became more widely available. In
1900 gas lift system with air was first used in large-scale oil field applications in Texas.
In 1920 natural gas was used instead of air for safety purpose because natural gas has less
explosion risk.
Early installations of gas lift system were used mainly for continuous flow, with a major
pitfall of one gas injection point around the tubing string. For deep wells, this required
excessively high kick-of pressures.

Figure 2-1 Brear oil ejector


Book6

Historically, gas was injected essentially uncontrolled into the bottom of the well and gas
lift application was limited to shallow wells because of low achievable injection pressures
(Takacs, 2005). In the mid-1930s the introducing of spring-operated differential valve
and the development of unloading process consisting of multiple well injection points
allowed gas lift to be used for deep wells. The spring-loaded deferential valve opened if

2
Texas Tech University, Fathi Elldakli, December 2015

there was enough pressure deference between casing and tubing. These valves were fixed
in place on the tubing and some valves were improved that could be mechanically opened
from the surface (Brown, 1980), but these all had problems during the operations, and if
they failed the entire tubing had to be pulled out to replace the valve.
In 1944 the first pressure-operated gas lift valve was patented by W.R. King (King,
1944). This valve uses bellows that replaced the mechanical spring to control gas
injection. Wireline retrievable valves were later discovered to change the valve without
replacing the entire tubing.
Over 25,000 Patents in gas lift valves have been issued since 1900 in the US only, but the
fundamental idea of the King valve is still the most widely-used today (Takacs, 2005).
Gas lift valves being used nowadays are one-way valves that allow the injected gas to
pass through to the tubing and block oil from passing through back to the annulus.
Figure 2-2 shows an oil well with gas-lift system. Gas flows in the annulus then through a
gas-lift valve into the tubing. When the injection pressure is higher than the pressure in
the tubing, the injected gas flows through the production tubing together with the
produced fluids.
The following development of gas lift system was done by Brown, Canalizo and
Robertson in 1961. Prior to 1864: some laboratory experiments performed with possible
one or two practical applications. 1864- 1900: This period consisted of lifting by
compressed air injected through the annulus or tubing. Numerous patents were issued for
foot-pieces, etc.

3
Texas Tech University, Fathi Elldakli, December 2015

Figure 2-2 Well with gas lift system

Gas Lift Equipment improvement

The technical development of gas lift system can be grouped into stages which are
described as follow:

1- Straight gas injection which employed no valve and consisted mainly of U-tubing
the gas around the bottom of the tubing. Several types of early gas and air lift are
shown in figure 3-2

Figure 2-3 Early gas (air) sift without valves


(API Gas Lift Manual)

4
Texas Tech University, Fathi Elldakli, December 2015

2- Jet collars (Figure 2-4) were placed up the string to allow gas to enter the tubing
at higher position. This will help to reduce the required kicking pressure around
the bottom.

Figure 2-4 Jet collar

3- Kick-off valves(Figures 2-5) were next employed to provide a means for closing
off gas after a lower valve was uncovered. The early kick-off valves were
intended to operate on a 1-20 psi differential pressure . The differential pressure
was improved by intrducing the spring-loaded differential valve which operated at
about 100 psi differential pressure. The kick-off valve was a crude forerunner of
the modern gas lift flow valve.

Figure 2-5 Kick-off valve

5
Texas Tech University, Fathi Elldakli, December 2015

In 1940, W. R. King introduced his bellows charged gas lift valve. The bellows in the
King valve is protected from excessive well pressure by sealing the bellows compartment
from the well fluids. The success of the King valve is evidenced by the fact that the
fundamental theories used in the design were quickly implemented by almost all valve
manufacturers. King Gas Lift Valve still uses with slight adjustment made.
Most valves, like the king valve, are called injection pressure operated valves because the
pressure of the injected gas generates force to open or close the bellows.
A lot of work was conducted on gas lift valve performance and gas flow behavior
through the gas lift valve at Tulsa University Artificial Lift Projects (TUALP). Acuana’s
gas lift valve performance data shows three distinct flow regions. In 1985, Biglargbigi
(Biglargbigi, 1985) performed simple valve testing. He presented a method to determine
valve throughput. His method is similar to the method used by Teledyne Merla (Merla,
1980).
Neiberding (Neiberding, 1988) performed dynamic tests on the Camco R-20 valve. He
developed semi empirical correlation for throttling and orifice flow. Acuna (Acuna,
1989) used Neiberding’s work as a basis for the normalization for one-inch nitrogen
charged gas-lift valves. He presented two correlations for throttling flow. Neiberding’s
and Acuna’s orifice flow equations are similar to the equation for orifice meter.
Decker (Decker, K., 1968) modelled a spring charged pressure operated valve. Decker
recognized the complex nature of the throttling flow phenomena. He used a force balance
equation to calculate a pressure profile on the ball. Decker, L. (Decker, et al., 1976)
developed an analytical method to determine the pressure response of a bellows operated
nitrogen charged gas lift valve.
Hepguler (Hepguler, 1988) determined the effective pressure acting on the ball of the
valve by replacing the bellows with a load cell. He was able to use his extensive data to
model the gas lift valve. It also predicted the valve flow performance in the throttling
flow region.
Sagar (Sagar, 1991) used convergent –divergent nozzle theory to determine pressure
profile acting on the surface of the ball.

6
Texas Tech University, Fathi Elldakli, December 2015

Gas lift system is not the most efficient artificial lift technique, but it is often more
technically and economically feasible. Some advantages of gas lift system are:
- Gas-lift requires few moving parts, and therefore is suitable also when solids
(such as sand) are produced.
- Gas-lift wells have downhole equipment with low cost and long service life. The
major equipment is the gas compressor, which is located on the surface which
allows for easy maintenance, while the downhole equipment mainly consists of
valves.
- Gas-lift is very flexible to changes in well conditions and production rates.
- Gas lift system has also some disadvantages. First, gas is circulated through the
wells, which requires a high gas compressor capacity, as more gas is compressed
than what is produced. There is also a problem related to the stability. This
problem where wells with gas-lift show an oscillatory behavior, where production
levels vary significantly with time.

7
Texas Tech University, Fathi Elldakli, December 2015

CHAPTER 3 : GLV TESTING

The gas lift system requires a source of gas, compressor and special tubing completion.
However, the most important part of this system is the GLV because it controls the
amount of the injected gas thus the amount of the produced fluid; it allows the injected
gas to pass the valve at the optimum depth and rate.
The GLV is the heart of the system and any failure in the designing the GLV will lead to
low production or no production.
Gas flow testing of gas lift valves is required to validate valve operation and measure
flow capacity. The gas lift valve tests guidelines given in API Recommended Practice
11V2 and ISO 17078-2. Shahri (2011) proposed a method to test the gals-lift valves based on
the concept of blow-down; the valve can be tested in a few seconds. This method is applied to
calculate accurate flow rate through the GLV in the critical flow region.
The gas lift valve has a gas dome section filled with high pressure gas, usually nitrogen,
connected to bellows which provides the movement to the valve stem and a metal seat
housed in a valve body. The seat provides the flow area for the injected gas to pass
through the valve. Also, the valve equipped with check valve to avoid back flow. Figure
3-1 shows the main parts of the gas lift valve.

The closing force for a gas lift valve can be either a gas pressure charge in the dome
section act on the bellows area or a spring force or a both a dome section charge pressure
and a spring. The valve will remain closed until the set closing force is exceeded. The
major initial opening force for most gas lift valves is the pressure exerted over the
effective bellows area less the stem-seat contact area. Other forces that help to open the
valve come from production pressure acting on the port area.
When classifying GLVs, industry uses the terms Injection pressure operated GLV and
tubing or production pressure operated GLV to represent valves that are more sensitive to
casing or tubing pressure as shown in figure 3-2. Pressure operation means that the
valve’s behavior is controlled by injection pressure, production pressure, or both. Gas lift
valves are can be controlled by changing the surface injection pressure and the operation
mechanism of either type of GLVs is the same.

8
Texas Tech University, Fathi Elldakli, December 2015

Figure 3-1 Gas-Lift Valve - main parts

Injection Pressure or Casing Pressure Operated Valves (IPO)


In this valve, the casing pressure (injection pressure) contacts the larger area of the
bellows and plays the main role of valve’s operation. During the unloading process, the
drop in the casing pressure is used to close the valves in order. The advantages of this
type of valves is that when the desired injection point is reached then an extra casing
pressure drop can be designed in ensuring the upper valves are closed. In addition, the
variation in tubing pressure is very unlikely to result in the unloading valves re-opening.

Production Pressure or Tubing Operated Valves (PPO)


In the PPO valves the flow path is reversed and thus the tubing pressure is acting on the
larger area of the bellows making the valve primarily sensitive to the tubing pressure. The
drop in the tubing pressure as gas is injected is used to close the valve. Figure 3-2 shows
IPO and PPO valves.

9
Texas Tech University, Fathi Elldakli, December 2015

Figure 3-2 Gas-Lift Valve, IPO and PPO respectively

When we inject the gas from the surface, the injected gas enters the valve through the
inlet ports. Depending on pressure conditions, the valve stem tip opens or closes the flow
area. In the closed position dome charge pressure (Pd) acting on the area of the bellows
(Ab) provides enough closing force to keep the valve stem on the port. The other forces
acting on the valve stem work in the other direction and try to open the valve. The greater
opening force comes from the injection pressure acting on the difference area of the
bellows and port area. Another force help to open the valve comes from production
pressure acting on the port area as shown in figure 3-3. When we inject the gas and the
upward force is greater than downward force the stim tip starts to move and the valve
starts to open. At any valve stem position, the area open to flow equals the lateral area of
the frustum of a right circular cone and when this area equals the port area the valve is
fully open (figure 3-4).

10
Texas Tech University, Fathi Elldakli, December 2015

Figure 3-3 GLV operations

Figure 3-4 Flow area

Depending of the position of the ball with respect to the port, the gas flow regime may
change. Theoretically when the flowing area is equal to the port area, the valve stem is
100% fully open. When the valve is fully open, there is no restriction and we expect the
valve passes the maximum gas volume. This position is representing the orifice flow
where the minimum area is the port area. Orifice flow performance can be divided into
two distinct regions: critical and subcritical (Figure 3-5).

When we inject the gas at a constant injection pressure as the down pressure (production
pressure) decreases the flow rate increases until the critical pressure is reached. In case of
critical flow, dropping downstream pressure does not affect the upstream flow rate.

11
Texas Tech University, Fathi Elldakli, December 2015

In the orifice flow, the injection pressure (opening force) can overcome the dome charged
pressure (closing force) even if the down pressure (production pressure) ceases. So, the
valve remains open.
When the injection pressure is not sufficient to keep the valve open and overcome the
bellows-charged pressure, the flow-rate reaches its maximum value at critical pressure
and extra drop in downstream (tubing) pressure causes low flow rate. This flow regime is
known as throttling flow. In throttling flow regime, the open area to flow is smaller than
the port area. At this case, the downstream pressure affects the production flowrate.

Figure 3-5 Flow regimes

There are two parts that can be easily changed based on the gas rate required for the GL
system design. These parts are: - stem tip (ball size) and seat (port) as shown in figure 3-
6. Sharp edged seat has 0.987” OD and 0.42” height. For the 1.5 GLV, we have 6
different ports, 3/16, ¼”, 5/16”, 3/8”, 7/16” and ½” as shown in the table 3-1. For each
port the ball size 1/16” larger than the port diameter as recommended by API. The flow
area can be calculated based on the ball stem position.

12
Texas Tech University, Fathi Elldakli, December 2015

Figure 3-6 GLV seat and stem


Implementing the geometrical calculations based on the frustum area (figure 3-7), the
relative theoretical min distance for the GLV to become initially fully open can be
calculated.

Table 3-1 GLV port and ball sizes specifications


Port Size(in) Ap (in2) Ball size, (in)
(1/16” + Port size)

3/16 0.03 1/4

1/4 0.05 5/16

5/16 0.08 3/8

3/8 0.11 7/16

7/16 0.15 8/16

1/2 0.2 9/16

13
Texas Tech University, Fathi Elldakli, December 2015

Figure 3-7 Flow area


Problem:

Bellows have helical mechanical structures that act like spring with more stabilized
performance. The bellows provide the most important function of the GLV by allowing
the valve stem to move on and off the seat to generate the flow area which is very
important parameter in each GLV. During the operations and with respect to the pressure,
the bellows are exposed to the opening force (injection pressure) and closing force
(dome-charged pressure). Therefore, the bellows compressed and stretched many times
due to the axial load. The axial loading on the bellows is limited. If the load exceeds the
bellows linear margin, the bellows stiffness changes and outer and inner convolution
come to lie on each other so the bellows may go under non-linear stacking behavior. As a
result, the GLV stem does not travel appropriate distance to create a flow area equal to
the port area. Therefore, the flow path will be restricted by the GLV's ball and the actual
throughput flow in the GLV becomes less than the theoretically calculated value.

Experimental results indicated that the GLV does not often fully open under actual
operations due to the bellows stacking phenomenon. As the dome charged pressure
increases, the test racks opening pressure, which is the GLV opening pressure, increases.
This causes the bellows to stake sooner due to the higher load rate as shown in figure 3-8.
This immature stacking result in flow restrictions and the required upstream area to flow
is no longer achieved.

The stem travel in the actual GLV system is based on the difference between the opening
and closing forces, and the bellows-assembly load rate. The equivalent port area for a
partially open valve is defined by the lateral surface area of the frustum of a right circular

14
Texas Tech University, Fathi Elldakli, December 2015

cone. The aim of this work is to improve the flow area with less stem travel. So, in case
the bellows stacking, we still have enough area to pass the injected gas. How can we
make the actual linear travel equal or close to the theoretical calculated one.

Figure 3-8 Effect of the dome charged-pressure on the bellows staking


(Shahri 2011)
A GLV is a variable orifice until maximum stem travel or a fully open port area is
attained. As the ball (stem) moves away from the ball/seat contact, the area opens to flow
is generated until the flow area equals or exceeds the fully open port area. The GLV
opening mechanism is gradual.
To study the GLV flow behavior at different stem travel position, we need to use
Benchmark valve (figure 3-9). The benchmark valve has the body of an actual GLV with
a controlled flowing area, which is used for these sets of experiments. The only physical
difference between a true GLV and the benchmark valve is the bellows assembly. The
benchmark valve is not equipped with bellows or dome section, while the stem position
in relation to the valve seat is manually adjustable.

15
Texas Tech University, Fathi Elldakli, December 2015

Figure 3-9 Benchmark valve


The theoretical minimum stem travel (Y) may be calculated using Eq. 1 developed by
Kulkarni (2005). The stem travel in Eq.1 is a function of the port radius(𝑟𝑝 ), and ball
radius(𝑟𝑏 ).

−𝑟𝑝2 +√𝑟𝑝4 +4.𝑟𝑏2 .𝑟𝑝2


−1
𝑌 = 𝑟𝑝 . 𝑡𝑎𝑛 [𝑐𝑜𝑠 ( )] − (√𝑟𝑏2 − 𝑟𝑝2 ) ………………………… (1)
2.𝑟𝑏2

In this research, the minimum stem travel for each port with different balls sizes was
calculated using the same equation applying different ball sizes. The results are compared
with minimum stem travel when the ball is only 1/16” larger than the port. Table 1 bears
the results. The table illustrates the improvement of the theoretical calculation of
minimum stem travel with larger ball size. As the ball size increases for the same port the
minimum stem travel required for theoretical fully open calculation decreases down to
25%.

16
Texas Tech University, Fathi Elldakli, December 2015

Table 3-2 Minimum stem travel for sharp edged seat with various ball sizes

Based on the theoretical minimum stem travel, ball size and port size, the open area to
flow is calculated using equation 2. At the same stem travel, the larger ball size provides
larger flow area which will result in improve the gas lift valve performance as shown in
figure 3-11.

2
𝐴 = 𝜋(𝑟𝑏 + 𝑟𝑝 )√(𝑟𝑏 − 𝑟𝑝 ) + 𝑦 2 …………………………….. (2)

17
Texas Tech University, Fathi Elldakli, December 2015

Flow Area 1/4" port


Ball
0.3
9/16"
Ball
0.25
Flow Area [inch]
8/16"
Ball
0.2 7/16"

0.15

0.1

0.05

0
0 0.05 0.1 0.15 0.2
Stem Travel [inch]

Figure 3-10 Effect of the ball size on the effective area (frustum area)

GLV test
GLV Tests can be done in two ways; one test method is to hold the upstream pressure
(casing pressure) constant while dropping the downstream pressure (tubing pressure).
As the tubing pressure is dropped, the flow rate through the valve starts increasing till
reaches its maximum value and remains constant. These types of tests are called constant
casing pressure test (CCPT). Another test method is to hold the downstream pressure
constant while varying the upstream pressure. As the upstream pressure increases, the
flow through the valve also increases. This method is called the constant production
pressure test (CPPT). The testing procedure was based on a constant production pressure
test (CPPT), in which the upstream pressure varies while the downstream pressure is
constant. In this experimental work, the downstream production pressure is atmospheric
pressure. Different sizes of seats were tested to determine gas passage through GLV 1-
1/2-in. IPO was used in all experiments. The benchmark valves with identical seats were

18
Texas Tech University, Fathi Elldakli, December 2015

installed inside the encapsulating tester, and then the stem travel was adjusted at six
different positions relative to the ball/seat contact area as shown in figure 3-11. The
methodology behind this technique is simply discharging a certain known volume of gas
at a certain upstream pressure. Each test will become complete when the upstream
pressure reaches the final downstream pressure which is ambient pressure.
Three different port sizes (1/4”, 5/16” and 3/8”) were tested at 25%, 50%, 75%, 100%,
125% and 150% of fully open with different ball sizes for each port which is shown in
Table 3. At each ball position the gas was injected through the valve and the data was
gathered digitally using a Data Acquisition System (DAS).
The pressure-time data were recorded using a very fast DAS. This system is capable of
recording up to 12,000 samples per second per channel. In our experiments, the sample
rate was set at 100 samples per second per channel (figure 3-13). The recorded data is
pressure versus time. Due to the high resolution and the amount of recorded data, the data
is scattered and need to be curve fitted in order to be used for the calculation. Regression
curve-fit method was implemented on the data using different techniques and the best fit
found (figure 13-4).

19
Texas Tech University, Fathi Elldakli, December 2015

Figure 3-11 Schematic of blow-down dynamic test facility

Table 3-3 Port tests


Ball Ball Position

Port Size 25% 50% 75% 100% 125% 150%

5/16"

6/16"
1/4"
7/16"

8/16"

6/16"

5/16" 7/16"

8/16"

7/16"
6/16"
8/16"

20
Texas Tech University, Fathi Elldakli, December 2015

Figure 3-12 Recorded data during the test

Figure 3-13 Curve fitting techniques

21
Texas Tech University, Fathi Elldakli, December 2015

CHAPTER 4 : RESULTS

Applying the newer concept of having a bit larger balls than conventional one leads to a
smaller ball-stem travel requirement. Figure 4-1 shows that as the ball size increases, the
flow area also increases at the same stem travel.

Figure 4-1 Effect of the ball size on the flow area for 1/4" Sharp Edged Seat

Calculating gas flow rate at new stem positions is another factor that demonstrates the
effect of the ball size and illustrated in Figure 4-2.
There are two main reasons for the effect of the ball size. The first reason is the contact
point between the larger ball and the top of the seat set is higher. So less ball movement
well generate the same flow area compared to smaller ball. Figure 4-3 shows the
difference in stem travel for different ball sizes.

22
Texas Tech University, Fathi Elldakli, December 2015

Figure 4-2 Effect of the ball size on the flow rate for 1/4" Sharp Edged Seat

Figure 4-3 effect of the ball size on the stem travel

The second reason is the Coanda Effect. Coanda effect is the ability of a fluid to be
attracted to a nearby surface. When the gas stream flows past the ball, some of the gas

23
Texas Tech University, Fathi Elldakli, December 2015

follows the contour of the ball and only leaves after flow a certain distance toward to the
port. Actually, the ball is pulling the gas around its surface. Figure 4-8 demonstrates the
Coanda effect. Since the ball is fixed to the valve body, it cannot move resulting in
direction the gas stream toward the port which increases the flow rate. In all cases it was
found that the larger the ball size, the greater was the efficiency.

Figure 4-4 Coanda Effect

24
Texas Tech University, Fathi Elldakli, December 2015

CHAPTER 5 : MODELING

Experimental method is a proper method that can be used to study the flow behavior
through the GLVs. However, the experimental methods are constrained to a particular
measurement points where we can install devices to record data. That’s why the
experimental results may not clear in invisible area and need to be justified.
To study the flow behavior and the effect of different parameters on GLV performance,
numerical method using computational fluid dynamic (CFD) can be applied.
CFD uses numerical methods to solve and analyze problem that involves fluid flow. CFD
modeling is built using governing equations; momentum, the conservation of mass, and
energy to describe and predict fluid flow behavior through a domain. These equations can
be solved using software tools.
The main goal of CFD is to obtain large amount of data at a large number of points to
understand the system behavior in depth and use this data to improve the system
performance. The CFD simulation process contains of three steps that are involved in the
analysis of the fluid flow:
1. Pre Processing: This is the first step of CFD simulation process, defining the geometry
with dimensions to define the domain of interest. This domain is then divided into
smaller segments known as mesh generation and then defining the boundary conditions.
2. Solver: After building the geometry and define the boundary conditions, computer
software will be used to solve the flow equations. The software has the capability of
solving the equations at every probe-point that has been already defined during the pre-
processing step. The numerical methods are also defined at this stage and we solve the
whole problem.
3. Post-processing: The next step after getting the results is analyzing the results with
varies techniques such as contour plots, vector plot, streamlines, and data curve Figure 5-
1 illustrates the CFD process.

25
Texas Tech University, Fathi Elldakli, December 2015

Figure 5-1 Computational Fluid Dynamic procedure

The CFD model built using commercially available software (Fluent from ANSYS) to
simulate the fluid flow behavior. The geometry is based on the experiments apparatus
dimensions. Figure 5-2 illustrates the parts of the equipment that has been modeled.
Based on the geometry where the flow behavior changes, higher number of grids were
used for accuracy purpose. The number of grids has a range from 145000 to 155000.
Figure 5-3 shows an example of grids for 1/4” port with 5/16” ball at 100% fully open.
The modeling is validated by matching laboratory experiments for eight (8) different
cases. The model results showed excellent matching of the experiments results with error
less than 3%. Figures 5-4 and 5-5 show two validated cases.

26
Texas Tech University, Fathi Elldakli, December 2015

Figure 5-2 Model domain

27
Texas Tech University, Fathi Elldakli, December 2015

Figure 5-3 Model meshing (grids) for 1/4" port with 5/16" ball size

Figure 5-4 Model validation for 1/4" port with 5/16" ball size

28
Texas Tech University, Fathi Elldakli, December 2015

Figure 5-5 Model validation for 1/4" port with 9/16" ball size
To develop an in depth understanding of pressure distribution, fluid flow inside the valve,
and two cases were compared:
Case 1: port diameter 1/4” and ball size 5/16”
Case 2: port diameter 1/4” and ball size 9/16”
The visualization at the same plan for the two cases as shown in Figure 5-6. It
demonstrates that the case “2” has a better pressure distribution at the effective area
which justifies the improvement of the flow rate using larger ball size as appearance in
Figure 5-7.

Case 1 Case 2

Figure 5-6 Pressure distribution inside the GLV

29
Texas Tech University, Fathi Elldakli, December 2015

Effect of the Ball Size on GLV Flow Rate


1/4" port, Sharp Edged Seat
Flow Rate [MSCF/D] 1200

1000

800
Ball
600 9/16"
400

200

0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
Stem Travel [inch]

Figure 5-7 Effect of the ball size on stem travel and flow rate

30
Texas Tech University, Fathi Elldakli, December 2015

CHAPTER 6 : CONCLUSIONS

 The effect of the ball size on the gas lift valve performance has been
evaluated.
 It was realized that bellows stacking is a limiting factor in gas throughput
capacity.
 The experiment results showed that larger ball size provides higher flow rate
at same stem travel for all ball positions.
 Using CFD model, the effect of the ball size on the overall performance of the
GLV was studied.
 Pressure distribution inside the GLV is improved using larger ball size.
 Computational Fluid Dynamics is a preferable method of modeling fluid flow
and is a cost-effective method for evaluating the performance of GLV. Using
CFD model, aid to investigate the effect of the ball size on the pressure
distribution.

31
Texas Tech University, Fathi Elldakli, December 2015

CHAPTER 7 : RECOMMENDATIONS

 It is recommended to investigate effects of the ball size on the GLV performance


using actual GLV.

 It is also recommended to do more works on ports (3/16”, 7/16” and 1/2”) to find
out the best relationship between the port diameter and the ball.

 Further work is required to solve the bellows stacking problem.

32
Texas Tech University, Fathi Elldakli, December 2015

REFERENCES

- S. F. Shaw. Gas Lift Principles and Practices. Houston, TX: Gulf Publishing Co., 1939.
- API Gas Lift Manual, Book 6 of the Vocational Training Series. Third Edition, 1994.
- Gabor Takacs. Gas Lift Manual. Pennwell Corporation, 2005.
- W. R. King. Time and volume control for gas intermittent. In US Patent Number
2339487, 1944.
- Biglarbigi, K., “Gas passage performance of Gas-Lift Valves.” M. S. Thesis, University
of Tulsa, Tulsa, Oklahoma (1985).
- Teledyne Merla: “Gas Lift Manual,” Garland, Texas (1980).
- Nieberding, M., Normalization of Nitrogen Loaded Gas-Lift Valve Performance Data.
M.Sc. Thesis, the University Of Tulsa, Tulsa, OK (1988).
- Acuna, G.A., Normalization of One Inch Nitrogen Charged Pressure Operated Gas-Lift
Valves. M.Sc. Thesis, University Of Tulsa, Tulsa, OK (1989).
- Decker, K. L., “Computer Modeling of Gas-Lift Valve Performance”, Paper OTC 5246
Presented at the 18th Annual OTC, Houston, Texas, May 5-8, 1968.
- Decker, L. A., and Undell, M.: “Analytical Methods for Determining Pressure Response
of Bellows Operated Valves”, unsolicited Paper SPE 6215, SPE of AIME, 1976.
- Hepguler, G.: Dynamic Model of Gas-Lift Performance. M.Sc. Thesis, University of
Tulsa, Tulsa, OK. (1988).
- Rajiv K. Sagar: An Improved Dynamic Model of Gas-Lift Valve Performance. M.Sc.
Thesis, University of Tulsa, Tulsa, OK. (1991).
- Mehdi Shahri, Simplified and Rapid Method for Determining Flow Characteristics of
Every Gas-Lift Valve (GLV), PhD Dissertation, Texas Tech University, 2011.

33
Texas Tech University, Fathi Elldakli, December 2015

APPENDIX A

Figure A-1 Effect of the ball size on the flow area for 5/16" port

34
Texas Tech University, Fathi Elldakli, December 2015

Figure A-2 Effect of the ball size on flow rate for 5/16" port

Figure A-3 Effect of the ball size on flow area for 3/8" port

35
Texas Tech University, Fathi Elldakli, December 2015

Figure A-4 Effect of the ball size on the flow rate for 3/8" port

36