Study of Drill String Safety Valves: Adam T. Bourgoyne, JR., Lsu

St udy of Drill St ring Safet y Valves
ADAM T. BOURGOYNE, JR., LSU

September 6, 2002
Craft and Hawkins Department of Petroleum Engineering
Baton Rouge, LA 70803
225 578 5215
Table of Cont ent s
EXECUTIVE SUMMARY............................... 2
ACKNOWLEDGEMENT ............................................................ 4
INTRODUCTION............................................ 5
INDUSTRY AND LITERATURE SURVEY.... 8
Common Modes of Failure...................................................... 8
Literature and Patent Search................................................. 12
DEVELOPMENT OF NOVEL DSSV DESIGN
......................................................................... 15
PHASE II DESIGN..................................................................... 15
PHASE II STATIC TEST RESULTS......................................... 18
Torque to Open under Differential Pressure ...................... 19
Torque to Close DSSV under Equalized Pressure ............. 20
Post Test Valve Inspection................................................... 21
Finite Element Analysis of Bearing Bore ............................ 21
PHASE III DESIGN.................................................................... 24
Design Changes ..................................................................... 25
Testing of Belleville Springs................................................. 27
Assembly Procedure.............................................................. 28
Phase III DSSV Test Program............................................... 29
Theoretical Operating Torque.............................................. 31
Test Results ............................................................................ 34
DEVELOPMENT OF DSSV STORAGE
STAND............................................................ 38
DESIGN ..................................................................................... 38
Hydraulic Pump ...................................................................... 39
Hydraulic Cylinders ............................................................... 40
Fluid Reservoir........................................................................ 40
Hydraulic Fluid ....................................................................... 41
Control Panel........................................................................... 41
Storage Structure.................................................................... 41
ASSEMBLY AND TESTING..................................................... 42
CONCLUSIONS AND
RECOMMENDATIONS................................ 44
CONCLUSIONS......................................................................... 45
RECOMMENDATIONS............................................................. 45
BIBLIOGRAPHY........................................... 46
PHASE III PROTOTYPE DSSV.................... 47
PHASE IV PROTOTYPE DSSV
STORAGE STAND......................................... 58
E X E C U T I V E S U M M A R Y
2
Execut ive Summary
This L SU study was funded by the Minerals Management Services U. S. Department of the
Interior, W ashington, D.C., under Contract N umber 14-35-001-30749. This report has not
been reviewed by the Minerals Management Service and approved for publication. A pproval does not
signify that the contents necessarily reflect the views and policy of the Service, nor does mention of trade
names or commercial products constitute endorsement or recommendation for use.
study of blowout preventer pressure test results by the Minerals Management Service
(MMS) identified the Drill String Safety Valve (DSSV) as one of the least reliable
components of the well control system. MMS sponsored a project at LSU to study this
problem and to make recommendations for improving the prevention of blowouts
through drill pipe. This final report presents the results that were obtained as part of this MMS
project. The study was completed in several phases by groups of senior engineering students
working on a senior design project. It was conducted as an interdisciplinary project which involved
students and supervising faculty members from the LSU departments of Petroleum and Mechanical
Engineering. The first phase of the study included a review of the common failure modes of drill
string safety valves and a review of the patent literature and existing product descriptions. Failure to
close against high flow rates and under high pressure was the most serious failure mode identified.
After considerable brainstorming, the group returned to a quarter-turn ball valve as the most feasible
concept that would allow a DSSV to be stripped in the borehole and wireline tools to pass through
an open valve. The final design concept selected was a trunnion mounted ball with spring loaded
seats. A new DSSV valve design was completed and a prototype valve was constructed. The valve
was tested under static conditions at various levels of internal pressure. The closing torque was
measured with no pressure differential across the valve seats for pressures up to 10,000 psi. The
opening torque was measured with various levels of differential pressure across the seat up to 4,000
psi. The operating torque of the new design was less affected by the internal pressure level in the
valve than the commercially available valves tested.
The second phase of the study was conducted in a summer term and consisted of a force analysis on
the bearing surfaces for an upward pressure differential across a closed valve. Modifications to the
prototype valve were recommended based on the results of the force analysis and an examination of
the valve parts after Phase I testing was completed.
In the third phase of the study, which was conducted by a different team, the valve design was
modified based on the work done in Phase I and Phase II. The design goal was a working pressure
of 5000 psi and an operating torque of less than 400 ft-lbf for all conditions. The prototype was
Chapt er
1
A
3
modified and a flow loop for testing drillstring safety valves under flowing conditions at a desired
flow rate and internal pressure was also constructed. Theoretical operating torque versus differential
pressure curves were calculated for the prototype design prior to testing. The prototype design was
tested along with several of the most commonly used commercially available drillstring safety valves.
Although the measured operating torque was higher than the theoretical values calculated, the new
design met the design goal.
It was noted during the Phase III testing that the operating torque requirements of a commercially
available DSSV that was tested with the new prototype had increased significantly over previously
measurements made during early tests. It was hypothesized that mud debris from testing and
weathering of valve components during periods of outside storage had increased friction between
the valve moving parts. Internal clearances could also have changed after conduction a number of
valve operations under high pressure and high torque conditions. Phase IV of the study was the
design and construction of a valve storage system capable of maintaining a drillstring safety valve in a
like new condition on the rig floor in an offshore environment. A storage stand was designed that
could store up to four DSSV units submerged in environmentally friendly oil and allow any of the
four valves to be retrieved with an air hoist in less than 10 seconds. Each DSSV is stored in a
cylindrical chamber and rests on a piston that can quickly elevate the DSSV by opening a valve that
allows hydraulic pressure to be applied beneath the piston. The storage stand was demonstrated at
an LSU/ MMS Well Control Workshop.
As a result of the study, the following recommendations are made regarding the use of Drill String
Safety Valves:
1. The DSSV intended for use as a stabbing valve to stop flow through the drillstring during tripping
operations should not be used in the drillstring for other operations. The stabbing valve should be
maintained in a like-new condition and used only during periodic pressure testing with fresh water.
2. Operators and/ or drilling contractors should check threads, valve wrench, and lift sub on the stabbing
valve and actuate the stabbing valve close and open each tour.
3. Operators and/ or drilling contractors should use a drillstring float whenever practical to provide
redundant protection against a high-rate flow through the drill-string during tripping operations.
4. When floats are not used, shear rams are recommended for redundant protection against blowouts
through the drillstring during tripping operations.
5. Drill String Safety Valves should not be the only means for stopping flow from the drillstring at the
surface when reverse circulating the well during completion operations. Flow should be routed through
hydraulically operated valves and a choke manifold.
6. Drill string safety valves should not be the only means for stopping flow through the drill string when
significant piping and flow restrictions are present above the valve.
4
Acknowledgement
The work for this interdisciplinary project was supervised by Dr. A. T. (Ted) Bourgoyne, Jr. and Dr.
L. S. (Scott) Stephens. The work was done by the following students:
Angelique Lawless;
Karl Eubanks;
Richard Duncan;
Daniel Reed;
Arthur J. Boudreaux;
Lance Fish;
Mark A. Casemore;
Elliot D. Coleman;
Thomas T. Core, Jr;
Brad W. Smith;
Jason Micas;
Jarield Francis;
Brian Puls; and
Hugh Kilpatrick.
Jim Cunningham of M&M International was the industry sponsor for the project and provided
valuable guidance and assistance. Mr. Cunningham also donated considerable shop time and supplies
and made his manufacturing and testing facilities available to the students on a number of occasions.
I N T R O D U C T I O N
5
Int roduct ion
In this chapter, the basicdrill string safety valve terminology will be reviewed. In addition, the current
problem areas identified in this study from a literature review and from consultation with industry
representatives will be presented.
afety valves are ball valves used to stop flow through the drill string that begin when the drill
string is being withdrawn from the well. Shown in Figure 1 is a photograph of a traditional
TIW safety valve. This photograph was taken during a visit to a valve manufacturing facility.
The valve has been disassembled
here to show the main working
components.
The portion of the safety valve shown
on the right side of Figure 1 would
accept the upper valve seat and spring
and screw down over the ball.
1
After
assembly, the ball floats between the
upper and lower seats and seals when
pressure is applied against the ball. The
spring assists in providing a low-
pressure seal. The valve is closed by
rotating the ball by means of an
operating crank or valve stem that fits
into a circular hole in the valve body.
The patent has expired on this simple
design which is now available from
several manufacturers in addition to
Texas Iron Works (TIW) from which it
took its name. The name TIW valve is
often used as the generic name for a
Drillstring Safety Valve (DSSV).

1
The portion on the right side of the safety valve is shown upside down to expose the female threads that make-up on the
male threads located just below the ball and lower seat.
Chapt er
2
S
Figure 1: Photograph of the traditional TIW Drill String
Safety Valve.
6
Shown in Figure 2 are the traditional locations
of safety valves which include upper and
lower kelly cocks as well as a stabbing valve.
Government regulations require that a safety
valve and operating wrench be maintained on
the rig floor at all times.
Displayed in Figure 3 is a photograph of a
safety valve made-up on top of a section of
drillpipe which has been cutaway so that the
ball and seats may be observed. The safety
valve shown is a one-piece valve design that
eliminates the need for threads in the valve
body area. This not only decreases the number
of possible leak paths, but also eliminates the
problem of ball locking due to excessive
torque. The valve is operated by means of an
operating-wrench that is inserted into the valve
stem and turned one-quarter turn.
Shown in Figure 4 are the blowout preventer
component pressure test results for the U. S.
Outer Continental Shelf during 1993 and
1994. Note that the pressure test failure rate
for drillstring safety valves and inside blowout
preventers was about 25%. This was especially
troublesome, since the level of redundant
protection for blowouts through the inside of
Swivel
Lower kelly cock
Kelly
Saver sub
DSSV
(Stabbing valve)
Upper kelly cock
Figure 2: Schematic showing traditional locations of
safety valves.
Figure 3: Photograph of a safety valve which has
been cutaway and made-up on top of a section of
drillpipe.
Drillstring Valves & I-BOP
25%
Choke
Manifold
Valves
29%
Connections & Other
19%
Choke
& Kill
Valves
11%
Rams
9%
Annular
8%
After Hauser, 1995
Figure 3: Results compiled from Blowout
Preventer Component Pressure Tests for the U.
S. Outer Continental Shelf during 1993 and 1994.
7
the drillstring can be much less than for flow through the annulus. Even though the choke manifold
also had a high pressure test failure rate, a failure in this component is not as serious because these
valves are not primary blowout barriers. Failure of one of these valves generally would not lead to a
blowout.
Several Objectives were defined for this project as follows: (1) to verify that safety valve failures are
a significant problem during actual well control operations, not just during routine testing, (2) to
identify common modes of safety valve failure, (3) to identify alternative methods for preventing
blowouts through drillpipe during tripping operations, (4) to design a test apparatus for evaluating
alternative safety valve designs, and (5) to make preliminary recommendations for reducing the risk
of blowouts through drillpipe. A secondary objective of the project was to provide meaningful
educational experience for LSU Engineering students working on their senior design projects.
The approach used in this study had four major aspects:
1. The students first talked with industry experts to better define the safety valve problems that
have occurred in the past. This aspect of the study dealt with problems occurring during well
control operations rather than problems occurring during routine pressure tests. The students
also visited valve manufacturers and talked with design engineers to get a better understanding
of the design problems and the various features already available.
2. The students conducted a literature and patent review to determine alternatives to the ball valve
design that had been studied in the past, and to identify auxiliary devices that could be used with
a safety valve to improve reliability.
3. The students then developed an alternative valve design and a test apparatus for comparing the
performance of the prototype valve design with several commercially available valves. As part
of the evaluation of the test apparatus, experimental data were obtained on valve performance
for several different valve designs.
4. Finally, the students developed a design of an improved storage stand for maintaining safety
valves readily accessible and in a like-new condition in an offshore rig-floor environment.
The various aspects of the research task will be presented in separate chapters.
D E V E L O P M E N T O F N O V E L D S S V D E S I G N
8
Indust ry and Lit erat ure Survey
The results of consultation with industry ex perts provided much valuable information, some of which had
been published in the literature
obil Oil Company conducted an industry survey in 1994 and identified 29 safety valve failures
during well control operations over an unspecified period. The survey was conducted after
Mobil experienced a number of problems in 1993 with stabbing valves leaking after being
stripped into a well in a threatened blowout situation. The survey also identified several
common failure modes for safety valves. The Mobil Oil study also agreed with experiences reported by
other industry experts that were consulted.
Common Modes of Failure
The common modes of failure identified in the industry survey conducted by Mobil included:
1. Failure to close due to high torque (valve lock up);
2. Failure to seal due to flow erosion;
3. Failure to seal against pressure from below;
4. Failure to seal against pressure from above;
5. Failure to seal against pressure from outside; and
6. Failures to open due to high torque (valve lock up).
The most serious of these failure modes is failure of the valve to close due to high torque. Blowouts
that occur due to this failure mode continue to occur. The author is aware of a blowout in 1996 and
another in 2001 that resulted in loss of life after a DSSV could not be closed.
2
About 400 ft-lbf is
generally regarded as an upper limit of torque that can be applied manually with an operating wrench. If
the torque required to completely close the valve is exceeded before the valve is fully closed, the failures
associated with partially closed valves occurs. High torque values are caused by the build up of pressure
in the valve as the valve begins to restrict the flow. The pressure pushes the valve stem further into and
against the valve body and the ball is forced against the upper seat. These two actions create can friction
forces that can not be overcome, especially if the ball and stem dimensions are out of tolerance due to

2
In both cases, other human errors also contributed to the loss of life.
Chapt er
3
M
9
Figure 4: Photograph of Amoco Goldsby Blowout.
wear. If the ball and stem are put under too much pressure, local stress deformations create metal to
metal contacts with the associated high friction surface. The ball of a two-piece valve often locks if too
much make-up torque is applied across the valve body. Tong placement is critical when tightening
across this type of valve.
Amoco conducted a series of safety valve tests in 1990 at their research lab after a blowout in the
Goldsby field due to multiple DSSV failures. A photograph of the Amoco Goldsby blowout is shown
in Figure 4. Note the flow of gas and saltwater exiting the drillpipe. Amoco estimated the saltwater
flow rate to be about 17 barrels per minute or 700 gallons per minute during this blowout. Three safety
valves were available on this job. After the first valve failed, the second was successfully stabbed into the
top of the first valve. The third valve was stabbed into the top of the second valve after it failed. It was
reported that the valves failed during closure for two reasons. First, the large amount of torque required
to close the valve under the existing flowing conditions prevented a quick closure. Second, the valve
balls and seats eroded before the valves could be fully closed.
1 0
The Goldsby blowout let the high-pressure fluid
move from below the valve, through the valve,
and out the top. In addition, several other failure
modes can occur. A failed safety valve installed
on top of the drillpipe with additional piping
connected above it can let high-pressure fluid
move from below the valve, through the valve
body, to the outside atmosphere. When pressure-
testing equipment installed on top of the safety
valve, a failure can occur allowing flow from
above the valve, through the valve, and into the
drillpipe below. This prevents a valid pressure
test from being performed. After stripping a
stabbing valve into the well, a failed safety valve
can let pressure move from the annular space
around the valve, through the valve body, and
into the drillpipe. This can cause surface pressure
readings to be misleading and could cause
mistakes to be made during the well control
operations. The common leak paths are illustrated
in Figures 5-7.
One of the most common failure modes
identified was the failure of the valve to open
even after pressures are equalized across the ball.
It is sometimes necessary to freeze a plug of ice-
mud below the safety valve so that it can be
replaced with pressure on the drillpipe. Ball lock-
up is caused by pressure against the valve stem, by
pressure forcing the ball into the seat, and by
alignment problems that develop under stress.
Eroded balls and seats are also a common cause
Upper seat
Stem
Stem link
Body
Ball
Seal
Lower seat
Seal
Figure 5: Fluid Escape past upper Seat
Upper seat
Stem
Stem link
Body
Ball
Seal
Lower seat
Seal
Figure 6: Fluid Escape through Valve Stem
Upper seat
Stem
Stem link
Body
Ball
Seal
Lower seat
Seal
Figure 7: Fluid Escape through lower Seat
Caused by Human Error
(Partially Closed Valve)
Figure 8: Erosion of Ball and Seat by Drilling Fluid
1 1
of failure identified in this study. The erosion
could be due by flow of mud solids through
the valve as it is being closed. However, if
the ball is not aligned perfectly in the open
position, erosion in an upper or lower Kelly
valve will also occur during normal drilling
operations (Figure 8). In addition, erosion is
caused by wireline work done through the
valve. Failed elastomers can also cause leaks
through the stem and around the valve seats.
In addition, stress cracks in the valve body
or threads also can occur.
A valve seat cut by fluid erosion due to a
slightly over-closed valve is shown in Figure
9. Wear on the valve stem stop can
sometimes allow too much rotation of the
ball. Deformation of the valve stem stop can
be caused by applying excessive torque when
using a cheater bar. Design constraints
resulting from maximum outside diameter,
minimum inside diameter, and a high
working pressure requirements make it
difficult to design robust valve stops.
Wireline cuts to the inside diameter of the
ball as shown in Figure 10 can reduce the
amount of over-rotation that can be
tolerated without leaking. Once fluid begins
leaking past the valve seat, erosion can
enlarge the leak path.
Figure 11 is an example of a failure in a
DSSV stem. The valve stem accepts the
valve wrench which then allows the valve to
be operated. The valve stem is recessed into
the valve body. Upon removal of the
operating wrench, the DSSV can be stripped
into the well without danger of the valve
stem catching on a ledge or in a casing
connection. The stem shown has wear in the
metal that serves as the valve stem stop and
is also cut by fluid erosion through the center
of the stem.
Although not illustrated in a photograph,
another common problem is valve lock-up
due to hardened solids left in the valve after
Figure 9: Erosion of Seat by Drilling Fluid in over-
closed valve.
Figure 10: DSSV Ball cut by wireline ( slickline)
work done through the valve.
Side View
Interior
View
Exterior
View
Valve Stem
Stop
Eroded
Hole
Eroded
Hole
Figure 11: DSSV Stem showing wear on Valve Stem
Stop and Erosion by Fluid
1 2
circulating mud and cement through the valve. Incomplete removal of cement from well control
equipment can lead to a non-operational safety device.
Lit erat ure and Pat ent Search
There was good coverage of devices to prevent blowouts through the drillpipe in the patent literature.
Twenty-three patents were reviewed and a number of alternatives to ball valves have been patented.
The research team concluded that ball valves are best suited to the need for full opening valves with a
small outside diameter that can be stripped into the well under pressure. Auxiliary equipment that
compliments the use of safety valves and increases the number of barriers to a blowout through the
drillstring was also found in the patent and literature review. Much of this auxiliary equipment had also
been identified through discussions with industry experts. The auxiliary equipment identified for added
blowout barriers included shear rams, floats or check valve placed in the drill collars near the bottom of
the drillstring, a drop-in check valve, a wireline retrievable check valve, a velocity triggered check valve,
and a double valve assembly. Shear rams can be used to cut through the drillpipe and close the well on
top of the drillpipe if the safety valve fails. The disadvantage of shearing the drillpipe and dropping it to
bottom is that it can make it more difficult to be able to eventually circulate kill mud to the bottom of
the well.
Floats (check valves) installed in the drill collars are widely used by some operators to make it easier to
stab and close safety valves at the surface. Both flapper and dart type check valves are available. Even if
the check valve leaks, the flow rate is
generally reduced enough so that the
safety valve can be successfully closed
without cutting out the valve. Some
operators seldom use floats for the
following reasons:
1. Extra time is needed to fill the
inside of the pipe when lowering
pipe into the well;
2. Higher surge pressures occur
when pipe is lowered into the
well;
3. Excessive loads of rock cuttings
in the annular mud can be
detected by back-flow through
the drillstring when a connection
is made; and
4. The shut-in drillpipe pressure is
more difficult to read the after
taking a kick.
The drop-in or wireline retrievable
check valve (Figure 12) overcomes
many of the objections to a float in
the drill collars. A sub that will accept
Figure 12: Schematic of a Drop-in / Wireline Retrievable
Check Valve
1 3
Threads
Seat
Ball
Passage
Venturi
Figure 13: Schematic of a Velocity Triggered
Trip Valve
Figure 14: Schematic of a Double DSSV
a check valve is always run in the drill string near
bottom. Just before it becomes necessary to pull
the drillstring from the well, the check valve
assembly is dropped into an open drillpipe
connection and pumped to bottom where it
latches into the sub. If the well tries to blowout
during tripping operations, the check valve will
stop the flow and make it easy to stab and close
the surface DSSV as part of the shut-in
procedure. In the event wireline work below the
check valve becomes necessary, the drop-in check
valve is wireline retrievable. However, the drop-in
check valve would not provide a barrier to
blowouts when making a connection during
drilling if it is not deployed until tripping
operations begin.
An example of a velocity triggered check valve is
shown in Figure 13. This valve was designed and
tested to a limited extent during the late 70s by
Hughes Tool Company for Shell. It was lost in
the shuffle of buy-outs during the 80s. Prototype
valves are again being built by a new company.
An example of a double valve assembly is shown
in Figure 14. For high flow rates, the lower ball
may cut out, but the flow rate should be reduced
enough to allow the upper valve to be
successfully closed (if it is closed quickly before
the bottom valve becomes too eroded). The
bottom valve can also be used as a mud saver
valve since a back-up valve is available. The
disadvantage of this approach is that it is not well
suited to stabbing valves, because of the extra
weight that must be handled. Even a single
stabbing valve for 4.5-in. or 5-in. drillpipe weighs
more than 100lbs. To minimize the weight of a
double valve, one manufacturer recently
developed a double ball, single body design.
New Pr ot ot ype Desi gn
An interdisciplinary team was assembled which
included seniors and supervising faculty in the
LSU Petroleum Engineering and Mechanical
Engineering Departments. The group was
charged with the development of an improved
DSSV design that would reduce the tendency for
1 4
valve lock-up when attempting to close or open against a high pressure differential or with a high
internal pressure. After reviewing the patent literature and discussing the design constraints, the group
concluded that the current approach of a quarter-turn ball valve was best. The main design constraints
resulted from the need to be able to strip into a well under pressure and to be able to perform wireline
work through the valve. It was decided that an effort would be made to design a trunnion-mounted
ball with spring loaded seats in order to reduce the pressure effects on valve operation. It was also
decided that prototypes of the new design would be constructed and tested together with commercially
available valves to evaluate the effectiveness of the new design. The development of a new valve design
and a valve test loop is presented in the next chapter.
1 5
Development of Novel DSSV Design
D rill S tring S afety V alves (D S S V s) are used to prevent blowouts through the
drillstring during tripping operations. S everal case history reviews of well control
events have shown evidence of severe problems with D S S V s. Of those problems, valve
lock up is most significant resulting in failure to open or close due to high torque.
his Chapter describes the design and testing of a prototype low torque DSSV of the ball valve type. The
prototype design relies upon a trunnion-mounted ball and pseudo-fixed seats to achieve low torque
operation, with the idealized design goal being a constant actuation torque independent of valve internal
pressure. Development of the novel DSSV took place in Phase II and Phase III of the project. Phase II
included the design, construction, and testing of an initial prototype. In Phase III, the design was improved
based on the results of Phase II and finite element modeling of stresses in the bearing surfaces. Additional testing was
then completed. Commercially available valves were tested with the new prototype to provide a basis for evaluation of
the new ball and seat design concept.
Phase II Design
Drilling operators have relied upon a manually operated DSSV over remotely actuated valves for
stabbing valves [Chilingarian, 1981]. This simplifies stabbing and operating the valve, regardless of the
height of the drillstring in the derrick. In addition, a hydraulic actuator could not be permanently affixed
to the DSSV because of the need to be able to strip the valve into the well under pressure. One
significant disadvantage of manually operated DSSVs is that the available actuating torque to open and
close the valve is limited by the physical strength of the operator. When the torque required for
operating the valve exceeds the torque applied by the operator, then the valve is said to lock up. Valve
lock up is generally considered to occur when
the torque that is required for actuating the
valve is greater than 500 ft-lbs. Because of
dimensional constraints, it is also difficult to
manufacture valve stops that can tolerate
excessive torque without permanent
deformation.
In many ball valve designs, the torque that is
required to either open or close the valve is
largely a linear function of internal pressure
and/ or differential pressure across the ball.
As such, at some critical pressure the ball valve
locks up due to increased friction between
Chapt er
4
T
0
100
200
300
400
500
600
700
800
0 500 1000 1500 2000 2500 3000 3500
Pressure (psi)
A
c
t
u
a
t
i
n
g

T
o
r
q
u
e

(
f
t
-
l
b
s
)
LSU Ideal Goal
Lock Up
Typical Valve
Figure 15: Typical Torque versus Internal DSSV
Pressure with no Pressure above the DSSV.
1 6
internal components. Figure 15 illustrates the problem of lock up for a typical manually actuated ball
valve. An ideal DSSV ball valve design is one where the required actuation torque is constant and
independent of the valve pressure, up to the maximum design sealing pressure. For such a valve, lock
up will never occur as long as the constant required actuation torque is below the lock up torque value.
Figure 15 also illustrates this idealized concept.
The objective of Phase I was to design, construct and test a DSSV with an actuation torque vs.
differential pressure curve that approaches the idealized, constant torque design. This work focuses
upon the low torque ability of the DSSV to open under large differential pressures and to close under
high flow conditions. It considers low torque ability of the DSSV to close under equalized pressure as a
secondary design criterion, as much less torque is required when the pressure above the DSSV is equal
to the internal DSSV pressure.
Ball valve designs commonly employ either a trunnion-mounted ball with floating seats or a floating ball
with fixed seats [Piper, 1985]. In both cases, the pressure of the fluid being sealed generates the sealing
force between the ball and the seats. The seats are then said to be energized. Of course, due to the
friction between the ball and the seats, and under ideal conditions, the required actuation torque for
these designs increases linearly with this pressure. In order to achieve the design goal of an actuation
torque independent of differential pressure, the fluid being sealed cannot be used to energize the seats.
One way of achieving this
performance is to mount both the
seats and the ball in the valve body.
In this manner all forces which act
on the ball and the seats are directly
transferred to the valve body. Such
a design then requires a separate
means to energize the seats with a
force large enough to provide
adequate sealing up to the
maximum rated pressure. This is
the basis for the design resulting
from this project.
Figure 16 shows an assembly
drawing of the LSU Phase II
prototype DSSV. The valve size is
6 3/ 8" OD x 2 1/ 4" ID x 25 3/ 4"
tall with 17-4 stainless steel upper
and lower seats and ball. No
surface coatings were used on the
valve to reduce friction. O-rings
between the seats and the ball were
used to create the major dynamic
seal. O-rings are also used
throughout the valve for the
secondary static seals. The ball is
mounted in a set of sleeve bearings
on two sides, through the actuation Figure 16: Phase II Prototype DSSV, 6 3/ 8 OD x 2 1/ 4 ID
1 7
stem and stem link. The tolerances between the ball, stem link, stem and bearings are kept small such
that the ball can "float" only a few thousandths of an inch before engaging the sleeve bearings. In
essence the ball is trunnion mounted. Using this mounting, the pressure acting against the ball is
transferred to the valve body through the bearings. The bearing load capacity is thus a limiting factor
for the valve design. Stresses are concentrated in the bearing bore rather than being distributed across
the ball and seat. Commercially available bearings were reviewed for this service and those with the
greatest load capacity for the given space constraint were selected. Even so, the bearing load capacity
limits the differential pressure capability of the valve to 4000 psi. This is only 80% of the target design
specification of a differential pressure of 5000 psi. It should also be noted that the torque required to
actuate the valve must increase with differential pressure to the extent that the bearing frictional force
increases. However, the pressure range was sufficient to allow testing of the new design concept.
The valve design uses a ball housed inside a canister for sealing. The canister assembly was provided by
M&M International and is one of their patented design concepts. The bottom seat is part of the
canister itself, while the top seat is a separate component that is loaded and held in place by a helical coil
compression spring. This spring provides the force which energizes the top seat against the ball and
provides a constant sealing force independent of differential pressure when the valve is sealing from
below. For a target design specification of a maximum differential pressure of 4000 psi, the sealing
pressure between the ball and seats is taken as 1.0-1.5 times this value. Based upon simple assumptions
regarding the contact area between the seat O-rings and the ball in the loaded condition, the
compression spring was designed to provide a maximum force of 4000 lbf. Shims were designed to
adjust this sealing force as needed during testing. Figure 17 shows the resulting force-deflection curve
for the spring over three cycles as tested using an Instron machine.
The canister is held into place by a locking cap which is threaded into the top of the valve body.
Therefore, any pressure acting against the bottom of the canister is directly transferred to the valve body
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 500 1000 1500 2000 2500 3000 3500 4000 4500
Applied Load (LB)
D
e
f
l
e
c
t
i
o
n

(
i
n
c
h
e
s
)
Figure 17: Load versus Deflection for Phase II Loading Spring for Top DSSV Seat
1 8
through the locking cap. Together, the bearings, canister, spring and locking cap provide a design
where the differential pressure contribution to the actuation torque depends primarily upon the sleeve
bearing internal friction. Optimal selection of frictionless bearings then results in a low torque DSSV
design when sealing a differential pressure from the below. Another limitation of the prototype design
shown in Figure 16 is that when pressure is applied from above, the low torque operation is lost. This
is not considered to be a significant limitation as the pressure when sealing from above can be
controlled by the operator. Finally, thrust bearings are mounted on the actuation stems to reduce the
friction between internal components due to the pressure difference between the interior and exterior of
the valve. This design can be termed a trunnion-mounted ball, pseudo-fixed seat design. Table 1
below summarizes the specification of the Phase II DSSV.
Table 1: Phase II Prototype Design - Trunnion Mounted Ball, Pseudo-Fixed Seats
Internal Component Material 17-4 Stainless Steel
Sealing Surface Viton O-rings
Max. Body Pressure 10,000 psi
Max. Differential Pressure 4,000 psi (Sealing from Below)
Low Torque Operation Sealing from Below Only
Primary Sealing Surface Top Seat, sealing from above and below
Phase II Static Test Results
Three types of static pressure tests were performed on the prototype safety valve. These were a
hydrostatic test on the valve body, a measurement of torque required to open under a differential
pressure, and a measurement of torque required to close under 100% equalized pressure. No tests
under flow conditions were conducted in this first series of tests. All tests were performed using water,
pressurized by a hydraulic test stand at a local valve manufacturers facility. These tests were similar to
those performed by the commercial DSSV manufacturer. Torque readings were made by a calibrated
digital torque wrench. The hydrostatic pressure test required the valve body and stem seals to effectively
seal twice the maximum allowable working pressure of the valve for 5 minutes. The valve exceeded
this performance by sealing a pressure of 10,000 psi for a 15-minute period. Torque measurements
under differential and 100% equalized pressure were obtained as part of the following testing
procedure:
1. Begin with valve in closed position;
2. Pressurize valve from below to the desired level;
3. Hold at this pressure for 5 minutes and check for evidence of leakage across the ball;
4. Open the valve using the torque wrench to obtain the torque required to open under differential pressure (
The valve is now at 100% equalized pressure on both sides of the ball);
5. Close the valve using the torque wrench to obtain the torque required to close under 100% equalized
pressure.
1 9
Torque t o Open under Different ial Pressure
The torque to open the Phase II prototype valve was measured for various internal pressures with no
pressure above the DSSV. Similar measurements were made on two additional DSSV valves provided
by two different commercial valve manufacturers. All of the valves are of 2 1/ 4 ID bores. Both
commercial DSSVs tested were one-piece designs rather than the very low cost two-piece design
patented by Texas Iron Works (TIW). One of the commercial DSSVs tested was a premium valve
with several special features for reducing torque. As was discussed earlier, due to limitations in the
bearing load capacity, the maximum rated differential pressure for the prototype LSU DSSV is 4000
psi. The results of the tests, which are shown in Figure 18, reflect this pressure rating as differential
pressure tests were performed up to this maximum limit.
During these tests, the valve stem for the LSU DSSV yielded when excessive torque was accidentally
applied while adjusting the pneumatic operator pressure in the test stand. The stem was re-machined,
but this difficulty resulted in only four data points for these tests. The results indicate that the actuation
torque to open the two commercial DSSVs was largely linear, while that for the LSU valve varied non-
linearly. This suggests that certain components within the valve shifted under pressure due to incorrect
tolerance during assembly, and that surfaces other than the sleeve bearings and O-ring seats were loaded
and in contact. Finally, the data does show a cross-over between the endpoints, where at low pressures
the LSU DSSV requires more actuation torque than the commercial DSSVs but at high pressures it
requires less actuation torque.
Data for the commercial DSSV1 shown in Figure 18 was obtained after the valve had been used in a
number of preliminary experiments and had been allowed to weather outdoors for several months. The
data for DSSV2 was obtained with a new valve. Data published for DSSV2 [Tarr, 1998] after many
0
50
100
150
200
250
300
350
400
450
500
0 500 1000 1500 2000 2500 3000 3500 4000 4500
Differential Pressure (psi)
O
p
e
n
i
n
g

T
o
r
q
u
e

(
f
t
-
l
b
s
)
Phase II Prototype
Commercial DSSV1
Commercial DSSV2 (Premium)
Figure 18: Torque to Open Pressurized DSSV with no Pressure above valve.
2 0
cycles and after exposure to a water-base mud containing sand fell slightly above the data for DSSV1.
Experience during the development of the testing program indicated that the operating torque for
DSSVs under a differential pressure depended to a great extent upon the previous history of the valve.
Torque t o Close DSSV under Equalized Pressure
Figure 19 shows the torque required to close the Phase II prototype design as a function of pressure
when the pressure above the valve is equal to the pressure within the valve. Similar data for another
commercially available valve is provided for a qualitative comparison and is labeled Commercial
DSSV1. The data shown for the commercially available valve was taken from catalog curves
published by the manufacturer. Both the Phase II Prototype valve and the commercially available valve
are of 2 1/ 4 ID bore and have a one piece body design. This data indicates that the Phase II Prototype
required a largely constant torque to actuate against equalized pressure. The required actuation torque
varied between 60-80 ft-lbs over an equalized pressure range of 0-10,000 psi. This compared favorably
to the data taken from product datasheets of the commercially available valve. These data shows that
for 100% equalized pressure, the Phase II Prototype achieved the goal of an actuation torque which is
largely independent of the internal valve pressure.
0
50
100
150
200
250
300
350
400
450
500
0 2000 4000 6000 8000 10000 12000
Equalized Pressure (psi)
C
l
o
s
i
n
g

T
o
r
q
u
e

(
f
t
-
l
b
s
)
Phase II Prototype
Commercial DSSV1
Maximum Goal
Figure 19: Torque to Close DSSV as a function of Equalized Pressure, 2-1/ 4 ID Valves
(Pressure above Valve Equal to Pressure within Valve)
2 1
Post Test Valve Inspect ion
Upon completion of the testing, the valve was disassembled and inspected. Inspection revealed
significant galling
3
between the lower seat and the ball, between the ball and stem link, and between the
stem link and stem. This wear indicates incorrect dimensional tolerance after assembly of the parts.
Indeed, an assembly review revealed that the O-ring groove for the bottom seat was machined to the
wrong size. The wear pattern between the ball and the stem link indicated that the ball was slightly off
center. Finally, wear between the stem and the canister windows indicated that the bearing tolerances
were too loose and the canister acted as a bearing surface. Each of these deficiencies would be
expected to increase actuation torque for the ball valve as a function of differential pressure.
The results achieved with the Phase II prototype indicated that the basic design concept was valid and
the research team recommended that this design concept be kept for the Phase III design team. It was
also recommended that additional analytical work be done to model stress and strain in the bearing
assembly before doing the Phase III design work.
Finit e Element Analysis of Bearing Bore
During the Phase II design of the valve, the bearing bore was identified as a key element to the success
of the new design concept. The bearing bore was designed to fit into the outer shell or body of the
valve. The double stem bearing bore assembly must handle loads as high as 20,000 psi to achieve a
DSSV with a 10,000 psi working pressure and a 2.25-in internal diameter. In terms of load handling,
the bearing surface must be able to handle high loads with very little deflection. A Finite Element
Analysis (FEA) was used to better understand how displacements and stresses on the bearing surface act
under certain load conditions. ANSYS 5.0, a commercial software package, was used to perform the
finite element analysis.
The finite element model should approximately model the actual bearing assembly in two dimensions.
The actual bearing assemble is shown in Figure 20. A review of how the stem and bearing bore fit
together and the how the bearing bore and the casing fit together shows that they can be modeled in 2-

3
Galling was indicated by adhesive wear where incomplete cold welding occurred, leaving large streaks in the surface.
1.750"
CASING
1.250"
STEM
BEARING BORE
2.250"
THRUST BEARING
Figure 20: Bearing Assembly
2 2
D as two concentric circles. The bearing was modeled as a plane 82 2-D 8-node structural-solid with
the key-option set to three in the ANSYS software package. What this means is that there are eight
nodes with each having two degrees of freedom to move in the x and y coordinate directions. The
key-option of three allows the model to be plane stress with a thickness to be input. The two sets of o-
rings were left out of the model because they are used only as a seal surface and not a bearing surface.
The model assumes that the valve body is rigid, allowing no deflection of the outer surface. The bearing
material was assumed to be homogeneous, having a modulus of elasticity of 30 Mpsi and a Poissons
ratio of 0.292. The compressive strength of the bearing material was 135,000 psi in an annealed
condition. The effect of a Melonite coating on the bearing surface was not included in the model
analysis. Radial clearances of 0.001 to 0.003 inches were investigated.
The mesh used in the Finite Element Analysis can be seen in Figures 21 and 22. Figure 21 shows the
calculated displacement for a load of 10,000 lbf (internal valve pressure of 5,000 psi) and a radial
clearance between the bearing surfaces of 0.003 inches.
ANSYS 5.3
SEP 11 1997
09:55:16
PLOT NO. 1
DISPLACEMENT
STEP=1
SUB=1
TIME=1
RSYS=0
DMX=.404E-03
SEPC=.823181
DSCA=281.477
ZV=1
DIST=1.251
Z-BUFFER
Figure 21 Visualization of Calculated Displacements for Bearing Bore using a Radial Clearance of 0.003 inches
and an Internal DSSV Pressure of 5000 psi
2 3
Figure 22 shows the calculated stress distribution on the bearing bore for the same conditions. Based on
the results of the analysis, the bearing design should be capable of handling an internal pressure of
10,000 psi. However, it was decided that construction of a DSSV with a trunnion-mounted ball and a
working pressure of 10,000 psi was beyond the scope of the current project.
ANSYS 5.3
SEP 11 1997
10:06:54
PLOT NO. 1
NODAL SOLUTION
STEP=1
SUB=1
TIME=1
SY (AVG)
RSYS=0
DMX=.404E-03
SMN=-32789
SMNB=-32891
SMX=9350
SMXB=10530
-32789
-28107
-23425
-18742
-14060
-9378
-4696
-14
4668
9350
Figure 22 Calculated Stress Distribution for Bearing bore using a Radial Clearance of 0.003 and an Internal DSSV
Pressure of 5000 psi. (Stress shown by ANSYS in psi with gray scale)
2 4
Phase III Design
Figure 23 shows an assembled drawing of the Phase
III valve. Detailed machine drawings are provided
in Appendix A. The valve size is 6 3/ 8 OD x 2
ID x 27 tall. All internal components are 17-4
stainless steel coated using a nitride process. The
surface coatings are used on the valve to reduce
friction. The valve uses Teflon O-rings between the
seats and the ball to create the major dynamic seal.
Butyl O-rings are used throughout the valve for
secondary static seals. The ball is mounted in a set of
sleeve bearings (the stems and bearing bores). The
tolerances between the ball, stem links, stems, and
bearing bores are small so that the ball is allowed to
float only three thousandths of an inch before
engaging the bearing surfaces. This design is
considered a trunnion mounted ball valve. Using this
mounting, the pressure forces acting against a closed
ball are transferred to the valve casing through the
through the stems and bearing bores.
The Phase III Prototype valve design, like the Phase II design, used a ball housed inside a canister for
sealing. The bottom seat is part of the lower canister itself, while the upper seat is a separate
component that is loaded by five Belleville springs. The springs provide the force which energizes the
top seat against the ball and provides a constant sealing force independent of differential pressure when
the valve is being sealed from below. For a design specification of 5000 psi for the maximum
differential pressure, the sealing force between the seat and the ball was taken as 1.1-1.5 times this value.
Based upon assumptions regarding the contact area between the seat, O-rings, and the ball in the loaded
condition, the Belleville springs was designed to provide a maximum force of 5500 lbf.
The canister is held into place by a locking cap which is threaded into the top of the valve body.
Therefore, any pressure acting on the bottom of the canister is directly transferred to the valve casing
through the locking cap. Optimal selection of frictionless bearings then results in a low torque DSSV
design when sealing a differential pressure from below. Finally, thrust bearings are mounted on the
stems to reduce the friction between the internal components due to the pressure difference between
the interior and exterior of the valve. The Phase III design was also a trunnion mounted ball, pseudo-
fixed seat design. Table 1 below summarizes the performance specifications of the valve as it is
presently designed.
Table 2: Phase III Prototype Trunnion Mounted DSSV with Pseudo Fixed Seats
Internal Component Material 17-4 Stainless Steel Nitride Process Coating
Sealing Surfaces Teflon O-rings
Max. Body Pressure 7,500 psi
Max. Differential Pressure 5,000 psi (Sealing from Below)
Low Torque Operation Sealing from Below Only
Primary Sealing Surface Top and Bottom Seats (Sealing from Above and Below)
UPPER
SEAT
BEARING
SURFACE
BEARING BORE
STEM LINK
BALL
LOCKING
CAP
CANISTER CAP
THRUST
CASING
LOWER
CANISTER
WASHER
SPRINGS
UPPER
CANISTER
STEM BEARING
Figure 23: Phase III Prototype DSSV Design
2 5
Design Changes
The bearing design was changed significantly
from the Phase II design which had utilized a
commercially available sleeve type bearing with
two separate press-fitted bearing surfaces. One
bearing surface was to be press-fit onto the
stem (inner sleeve), while the other was to be
press-fit into the casing (outer sleeve). A
schematic of this can be seen in Figure 24.
The design was changed for several reasons.
First, press-fitting a bearing surface into the
valve body with perfect concentricity is difficult
to accomplish. Second, replacing the bearing
surfaces would be difficult. For example,
during testing procedures if the bearing surface
is galled or damaged, it would have to be
replaced. Since the parts are press-fit and in a
recess, the only way to remove them is to
machine them off. It is very difficult to
machine the bearing surfaces without damaging
the stems. Also, it would be difficult to remove
the bearing surfaces from the valve body. These
would also have to be removed by machining.
For the required forces and size constraints no
commercially available bearing could be found.
Therefore, a new bearing was designed from
scratch. Also, results from last years testing
showed localized yielding on the bearing
surfaces because they were not hard enough.
Instead of using a press-fit bearing surface on
the stem, the new bearing surface was designed
to be an integral part of the stem. The original
bearing surface ring was designed to be 17-4
stainless. This was not changed because the
stems were also 17-4 stainless. However, we
had to increase the size of the thrust bearing in
order to compensate for the increased diameter
of the stem. An isometric view of the new
stem with its integral bearing surface is shown in
Figure 25.
The second design change was with the outer
bearing surface. Instead of using a press-fit ring
with an internal bearing surface into the casing,
we used a bearing bore with interference fit at a
OUTER SLEEVE
INNER
SLEEVE
STEM
Figure 24: Phase II Bearing Design
BEARING
SURFACE
(MELONITE)

VITON
O'RING
SEAL
SURFACE

Figure 25: Phase III Stem Design
INTERFERENCE FIT
SURFACE
SEAL
SURFACE
BEARING
SURFACE
(MELONITE)
VITON
O'RING
GROOVE
Figure 26: Phase III Bearing Bore Design
2 6
non-critical area. An isometric cross section of
the bearing bore can be seen in Figure 26. This
bearing bore design eliminated several
problems. First, the radial thickness of the
bearing bore eliminated any chance of
deformation during the press-fit. Since the
smallest outer diameter of the bearing bore is
the only surface of the part that will be
interference fit, a slight deformation at this point
will not effect the bearing surface. Second, the
new design of the bearing bore allows for easy
removal. If the bearing bore needs to be
removed from the valve, it can be pressed out
of the valve casing with little difficulty. The bearing bore was constructed of 17-4 stainless steel. The
new bearing bore and stem design
4
is illustrated in Figure 27.
With the new design changes, the possible stresses were re-calculated using bearing design equations and
the factors of safety were checked. The Maximum Allowable Working Pressure for the valve is 5000
psi. With the valve in the closed position, this differential pressure causes a force of 10,000 lbs to act
upon the two stems and bearing bores. The diametrical clearance between the bearing surfaces on the
stem and the bearing bore is 0.003 inches. Hertzian contact stresses were used to calculate the
maximum pressure on the bearing. This calculation provided an independent analysis of the Finite
Element Analysis model done previously by the first design team. The contact area was calculated using
the following equation:
Where:
F = Force between stem and bore;

4
The valve stem used in the Phase III Prototype was modified to fit a pneumatic valve operator in the flow loop used for
testing the valve against flow. The stem is shown protruding out of the valve body. The valve stem used on a working DSSV
valve would be recessed and require the use of a valve wrench for manual operation.
STEM
BEARING BORE
THRUST BEARING
Figure 27: Mating Stem and Bearing Bore
( ) ( )
2
1
2 1
2
2
2
1
2
1
1 1
1 1

2
1
1
1
1
1
]
1
,
_
,
_
d d
E E
L
F
a

aL
F
P

2
max
2 7
L = Length of bearing surface;
= Poissons Ratio;
E = Modulus of Elasticity;
d1 = Diameter of bearing surface on stem;
d2 = Diameter of bearing surface on bearing bore;
1 = Denotes stem material properties; and
2 = Denotes bearing bore material properties.
The static factor of safety was calculated by dividing the yield strength by the maximum pressure
(Pmax). The factor of safety for a 5000 psi working pressure is 4.4. This is the ideal situation in which
only the intended bearing surface is being used. For a 10,000 psi working pressure, the static factor
of safety would be about 2.2.
There is a second scenario in which the factor of safety will change. The seal surfaces between the
bearing bore and stem could come into contact when pressure is applied to the valve. This is because
the required diametrical clearance between the sealing surfaces is only slightly above that of the bearing
clearance (approximately 0.002). In the case that the sealing surfaces and the bearing surface do come
into contact, the contact area would increase thus causing the maximum pressure to decrease. The
worst case would be if the sealing surface absorbed all loads and the intended bearing surface absorbed
none. This could happen if the tolerances between the stem bearing surface and the bearing bore were
not met. This case was also modeled using the equations given above. A factor of safety of 2.37 was
calculated for a working pressure of 5000 psi. For a working pressure of 10,000 psi, the factor of
safety would be about 1.18.
Test ing of Belleville Springs
The Belleville springs were used in the Phase III design to force the seat against the ball. The Belleville
washer springs were designed to have an inside height of 0.080, and a thickness of 0.150, giving a
height over thickness ratio (the h/ t ratio) of 0.533. This h/ t ratio gives an approximately linear
response when deflection is plotted against force. The springs as received from the manufacturer
differed from the original design. The received
inside height was 0.070 and the thickness is
0.105. The h/ t ratio is 0.667 and the
calculated spring response is less linear. The
force of the received springs was computed to
be 1115 lbf at 0.065 of deflection. Five springs
should produce a force of 5575 lbf at the
0.065 of deflection specified in the design.
To check the computed spring values, the as
built springs were tested in an Instron
compression tester. First, each spring was tested
individually to determine the individual spring
constants. The springs were loaded between
flat steel plates. Second, a stack of five springs
Spring Constant
= 61098.46
lbs/in
y = 118673x
2
+ 74983x - 343.23
0
1000
2000
3000
4000
5000
6000
0 0.02 0.04 0.06 0.08
Deflection (in)
L
o
a
d

(
l
b
s
)
Figure 28: Load versus Deflection for the stack of Five
Springs used in Phase III Prototype DSSV
2 8
was tested. This test was done to verify that (1) the actual spring load was near the computed spring
load and (2) that the total load could be computed using the sum of the individual spring constants and
the same deflection for all five springs. The results of the tests showed that a force of 5032 lbf would
be generated for the stack of five springs used in the Phase III Prototype. Shown in Figure 28 are the
test results.
Assembly Procedure
The Drill String Safety Valve consists of many different integrating
parts. Therefore, assembling the parts in the correct order is a must.
The following is the procedure that was used to assemble the valve.
1. The bearing bore o-rings were greased and then installed on the
bearing bore. With the casing in a vertical position, the bearing
bores were inserted as shown in Figure 29.
2. The thrust bearings were greased with a light coat of grease and
placed onto the stems. The stem O-rings were placed into their
respective grooves. The O-rings and bearing surfaces were also
greased. The stems were inserted into the bearing bores. Finally, the
slots on the stem that accept the stem links were positioned
vertically. Figure 29 also illustrates this assembly.
3. The Teflon O-ring was inserted into the O-ring groove in the lower
canister. It was set by placing the ball on the O-ring and striking
the ball firmly with a rubber mallet. The ball was then removed.
The next step was to take the lower canister and insert the stem
links such that the stem links were also vertically positioned. Once
the stem links were in place, the ball was then inserted into the
lower canister with the slot in the ball aligned with the stem links.
A schematic of this can be seen in Figure 30.
4. The Teflon O-ring in the upper seat was inserted in the upper seat
O-ring groove and was set using the same method used to set
the lower Teflon o-ring. The upper canister was placed on top of
the lower canister and then fastened together by using a snap ring.
The upper seat was then placed into the upper canister with the
Teflon o-ring in contact with the ball.
5. The next step was to insert the five springs with the concave sides
down. Once the springs were in place, the upper canister cap was
placed on the upper canister. For the upper canister cap to be
attached to the upper canister, the springs were compressed using a
suitable shop press.
6. The whole inner canister was placed into the press and then
compressed until a snap ring could be inserted into a snap ring
groove to connect the upper canister and the upper canister cap.
7. Once step 6 was completed, the inner canister was one integrated
part. The inner canister was removed from the press and all of the
necessary outer o-rings were greased and installed. A schematic of
the inner canister can be seen in Figure 31.
BEARING BORE
STEM
THRUST
BEARING
CASING
Figure 29: Step 1 of Assembly
STEM LINK
BALL
LOWER
CANISTER
Figure 30: Step 2 of Assembly
BELLEVILLE
SPRINGS
STEM LINK
BALL
CANISTER CAP
UPPER
SEAT
LOWER
CANISTER
UPPER
CANISTER
Figure 31: Assembled Inner
Canister
2 9
8. The final step was to insert the inner canister into the valve body. Once the inner canister was situated
properly in the casing, the upper casing cap with its attached and greased o-ring was screwed on and the
assembly was thus completed.
Phase III DSSV Test Program
The initial test conducted on the Phase III Prototype were hydrostatic tests conducted at a commercial
valve manufacturing facility. The test was similar to the test performed after manufacturer of a
commercial valve. This test began with the water-filled valve in the closed position and one end of the
valve fitted with a threaded cap. The threaded cap was ported to allow hydrostatic pressure to be
applied. A high-pressure pump was used to increase the internal pressure of half of the valve. To pass
the test, the pressure must be held for ten minutes with no significant leaks or drops in pressure. The
opposite end of the valve is then tested in a similar manner. The Phase III Prototype was tested
successfully in this manner to 7500 psi, which was 1.50 times working pressure of the valve. The DSSV
was then removed from the test apparatus and actuated to ensure no deformation occurred with the
moving parts of the valve. Finally, the valve was disassembled and inspected for internal damage or
deformation of internal components.
The second type of tests conducted on the Phase III Prototype was a measurement of the torque to
operate the valve with internal pressure. For these tests, the pressure above the ball and upper seat was
equal to the internal pressure so that there was no pressure differential across the ball and seat. These
tests were conducted at the LSU Research Well Facility operated by the Department of Petroleum
Engineering. Similar tests were also performed on two commercially available DSSVs using the LSU
facility. Additional test results of this type had been reported in the literature for other valves. This
allowed the performance of the Phase III prototype to be compared to several other valve designs.
A schematic of the test stand used for the equalized pressure tests is shown in Figure 32. The bottom of
the valve was threaded into the test stand.
The bottom of the test stand is fitted
with adapters that allow for the
connection of a hydrostatic pressure
source. Pressurized was provided by a
pneumatic pump capable of providing
pressure up to 15,000 psi. The test stand
cap is threaded into the top of the valve,
and pressure sensors were threaded into
the top of the cap. Torque and position
sensors were attached to the valve stem.
The signals from these sensors are sent to
a computer with a sample rate of three
samples per second. The high-pressure
pump was used to charge the fluid to the
desired pressure. As the valve was closed
and opened, the torque and position was
recorded. Tests were repeated three
times for pressures of 1000, 2000, 3000,
4000 and 5000 psi. When the tests were
completed, the DSSV was disassembled,
High
Pressure
Pump
Instrumentation
Pressure
Gauge
Test
stand
Drill
String
Safety
Valve
Figure 32: Equalized Pressure Test Apparatus
3 0
and the internal components were again inspected for damage and deformation. The test data is
presented and compared graphically to other DSSVs in the next section of this report
The third type of test conducted on the Phase III DSSV measured the closing torque when the valve
was closed with water flowing through the valve. A flow loop was constructed at the LSU Research
Well Facility for these tests. A schematic of the flow loop is shown in Figure 33. A mud tank filled with
water was connected to a Halliburton HT400 positive displacement pump. The pump was connected
to a SWACO automatic choke in parallel with the design valve. Torque and position sensors were
attached to the valve stem. The signals from these sensors were sent to the computer configured with
Choke
Instrumentation
Choke
Drill
String
Saf et y
Val ve
Posi ti ve Di spl acement Pump
Mud Tank
Figure 33: Schematic of Test Flow Loop for Closing DSSV against Flow
Figure 34: Photograph of Pump and Test Stand
3 1
instrumentation software. The pump was used to circulate the fluid at the desired flow rate. A
downstream pressure was held against the DSSV by another choke places in series with the DSSV. The
choke in parallel with the DSSV was used as a pressure release device. This choke was set so that it is
normally closed at pressures below a set pressure psi. When the choke was closed, the entire flow
passed through the DSSV until the DSSV closes and the pressure in the system builds to the set
pressure. The automatic choke then opens and holds the pressure constant at the set pressure. This
arrangement approximately simulates the pressure build-up under a DSSV after it closed a flowing well.
Torque and position during closure were recorded three times for each of six flow rates ranging from
100 to 350 gallons per minute.
DSSV failure to close has also occurred during completion operations involving reverse circulation
through the drillstring. When a DSSV is used as a safety valve with downstream piping attached to the
top of the valve, significant pressure can develop at the DSSV due to flow through the downstream
piping. The purpose of the choke in series with the DSSV (Figure 33) was to simulate the effect of
downstream piping.
Theoret ical Operat ing Torque
Calculations were made in order to predict the torque required for operation of the Phase III DSSV.
These calculations were based on the design of the valve assuming that all design dimensions were
achieved during construction.
Before calculating torque to operate the DSSV in a test condition, calculations were made for operating
the valve with no pressure, as would be the case in a routine actuation test done on the rig floor. The
only force applied to the valve in this situation is the 5,000-lbf force from the springs. Assuming that the
ball is exactly centered, the contacting surfaces are the interfaces between the ball and Teflon O-rings,
and the interface between the bearing O-ring seals and the bearing bores (Figure 35).The friction
between the ball and the Teflon seals was
calculated using the normal force and the
coefficient of friction between the two surfaces.
The frictional force caused by one rubber
bearing O-ring seal was calculated with the
following equation [Warring, 1981]:
Where:
fc = Friction per unit length due to O-ring
compression (lbf/ in)
L = Length of seal rubbing surface (in)
BALL BALL BALL BALL
STEM STEM STEM STEM
CASING CASING CASING CASING
TEFLON
SEALS
TEFLON
SEALS
TEFLON
SEALS
TEFLON
SEALS
PRESSURE / FLOW
LOWER
CANISTER
LOWER
CANISTER
LOWER
CANISTER
LOWER
CANISTER

5500 lbf 5500 lbf 5500 lbf 5500 lbf
UPPER
SEAT
UPPER
SEAT
UPPER
SEAT
UPPER
SEAT
RUBBER
O'RINGS
RUBBER
O'RINGS
RUBBER
O'RINGS
RUBBER
O'RINGS
Figure 35: Phase III Prototype Schematic
showing Force acting between Ball and Seat
and in Trunnion Bearing
) ( ) ( A f L f F
n c bs
+
3 2
fn = Friction per unit area due to fluid
pressure (lbf/ in2)
A = Projected area of seal (in2)
The torque needed to overcome the
friction involved with the Teflon O-rings
and the bearing O-rings at atmospheric
pressure is given by:
When an equalized pressure is applied inside the DSSV, the friction between the ball and Teflon O-rings
is the same as the previous case. However, the stem will be forced axially into the needle thrust
bearings. Since the needle thrust bearings allow for low torque operation at high forces, the operating
torque was expected to increase only slightly with increasing equalized pressure applied to the valve.
When a pressure differential exists across the ball and upper seat such that the pressure inside the valve
becomes greater that the pressure above the valve, additional load will be transferred to the trunnion
bearings in the stem (Figure 36). When there is no pressure in the valve, the only forces on the ball are
the squeezing forces caused by the spring. When pressure applied to the valve, the pressure causes the
ball to push upward. With the tolerances set correctly in the valve, the ball will compress the spring
slightly and crush the Teflon o-ring a little more causing the stem bearing surfaces to contact. The
friction of the bearing surfaces was calculated assuming that all force due to an applied pressure
differential is transferred to the stem bearing surfaces.
The actuation torque for an applied pressure differential was estimated using:
In older valves or when excessive pressure is applied to the ball, the ball and seat come into contact the
coefficient of friction in the first term of this equation changes to that of non-lubricated Melonite. The
only term that is a function of pressure is the friction in the bearings. The coefficient of friction given by
the manufacturer for Melonite is 0.35 for dry surfaces and 0.04 for lubricated surfaces. The bearing
surfaces for the manufactured valve were greased prior to assembly. The coefficient of friction was
taken as the average of the dry and lubricated values because of the inability to lubricate the bearings
after assembly. A plot of the actuation torque as function of differential pressure is shown in Figure 37.
The plot shows that ideally the valve is expected to operate under 400 ft-lbf of actuation torque up to a
5000-psi differential pressure. Once closed, the valve would hold against higher pressure, but could not
be opened without applying an equalizing pressure above the DSSV.
lbs - ft 66 )
12 * 2
488 . 1
* 498 . 1 * * 8 . 0 ( 2 )
12
8 . 1
* 5500 * 04 . 0 ( 2
) ( 2 ) ( 2
+
+
seal bs ball spring t np

R F R F T
SPRING
REACTION 1
SPRING
FORCE
REACTION 2
R2
SPRING REACTION 1
+ PRESSURE FORCE
SPRING
FORCE
Figure 36: Force Balance with no Pressure Differential
and with 5000 psi Pressure Differential
bearing bearing Melonite ring O ring O n compressio ball spring Teflon NP
R F R L f R F T ) ( 2 + ) ( 2 ) ( 2

+
3 3
The required torque to close the valve during flow is a maximum during the last interval of rotation of
the ball. This is due to the water hammer effect when flow is suddenly stopped. The maximum
pressure experienced by the ball during a closing operation is given by:
Where:
= Density of drilling mud.
c = Speed of Sound through mud.
V = Mud Velocity through valve
before closure (assume to be zero
after closure).
g = Acceleration due to Gravity
The differential pressures caused by closing
the valve against a flow were calculated for
an unweighted drilling fluid for the range of
the actual test flow rates. The water
hammer pressures calculated are valid for a
stabbing valve as long as the closing time is
less than the time required for a pressure wave to travel to the bottom of the drillstring and then be
reflected back to the surface. This would be a valid assumption for most actual cases but not for a test
loop. The required torque was calculated using the equations for P and for TNP given above. The
results are shown in Figure 38. For flow rates up to 350 gal/ min (8.3 bbl/ min) the Phase III Prototype
DSSV should close easily. The water hammer pressure at 350 gal/ min is about 1750 psi. The
calculated results vary linearly with mud density. For 16.7 pound per gallon mud, the torque required
would be twice those shown in Figure 38.
The ball and seats, like the bearing surfaces, are both coated with Melonite. If the ball is not centered
correctly because dimensional tolerances were not achieved or if the Teflon o-rings are not seated
properly during assembly, the ball and the seats could come into contact at higher pressure differentials
and flow rates. The result would be higher actuation torque because the reaction force from the
pressure would be at a non-lubricated surface. The main reason it is desired that the force be
transferred to the bearing surfaces is because they can be lubricated.
0 1000 2000 3000 4000 5000
0
100
200
300
400
Pressure (psi)
Torque (ft-lbf)
Figure 37: Theoretical Actuation Torque as a Function of
Upward Differential Pressure

P
c V
g

3 4
Test Result s
The test results obtained in this study for the
Phase III Prototype were compared to test
results obtained for commercially available
valves. Some of the test results for
commercially available valves were obtained in
this study and some were obtained from
published results by Tarr et al [1998]. Some of
the data published by Tarr et al were obtained
after flowing a 16-ppg water base mud
containing sand through the valve. All of the
data obtained in this LSU study was done
with water or unweighted gel mud.
Operating torque as a function of pressure are
shown in Figure 39 for the Phase III
Prototype and for five commercially available DSSV designs. The results of Figure 39 were obtained by
closing the DSSV with pressure above the valve equal to the internal pressure of the valve (equalized
pressure condition). DSSV0 was a valve with the traditional two-piece TIW design with a floating ball
and fixed seats. DSSV1 was a more advanced design with a one-piece body, a floating ball, and fixed
seats. DSSV2 was a premium design similar to DSSV1 but with additional needle bearings in the stem
and special coatings for torque reduction. DSSV3 was a relatively new design with a trunnion-mounted
ball and floating seats. DSSV4 was a single body, double ball design intended for use with a top drive.
0 100 200 300
0
100
200
300
IDEAL TORQUE VS FLOW RATE
Flow Rate (gal/min)
T
o
r
q
u
e

(
f
t
*
l
b
f
)
Figure 38: Calculated Torque required for closing Phase III
Prototype against flow of unweighted Mud
0
100
200
300
400
500
600
0 2000 4000 6000 8000 10000 12000
Equalized Pressure (psi)
T
o
r
q
u
e

(
f
t
-
l
b
s
)
Phase III Theoretical
Phase III Prototype
DSSV0 (2-PieceTIW Design)
DSSV1
DSSV2 (Premium)
DSSV2 (Premium)Tarr et el, 16 ppg mud)
DSSV3 (Trunnion Ball;Floating Seat)Tarr et al
DSSV3(Trunion Ball,Floating Seats)Tarr et el - 16 ppg mud
DSSV4 (Double Ball Valve Design)Tarr el al
DSSV4(Double Ball)Tarr et el-16 ppg mud
Maximum Goal
Figure 39: Maximum Closing Torque as a function of Internal Pressure with Pressure above DSSV equal
to internal Valve pressure
3 5
Note that the operating torque for the Phase III Prototype was much less sensitive to pressure than the
commercially available valves tested. At high pressures, the Phase III Prototype could be operated at
much lower values of torque than the commercially available valves. At low internal pressures, the Phase
III Prototype required a higher torque to close the valve than the commercially available valves. This was
an expected effect of the high spring force squeezing the upper seat against the ball. There was good
agreement between the calculated theoretical torque to close the valve and the measured torque to close
the valve for internal pressures below 1000 psi. However, at high values of internal pressure the
measured torque value were about 30% higher than the calculated theoretical values. It is believed that
the reason for this discrepancy was because not all of the pressure loading was being transferred to the
stem bearings as was assumed in the theoretical analysis. Compression of the spring with increased
pressure likely increased the friction between the ball and the seat, which was not considered in the
theoretical analysis. A more complex finite-element analysis, that was beyond the scope of this project,
would be needed to consider these effects.
The maximum torque to open the Phase III DSSV with pressure inside the valve and atmospheric
pressure above the closed ball and seat are shown in Figure 40. Note that in this test, the Phase III
Prototype achieved the best results at the highest differential pressures used in the tests. DSSV2
(Premium one-piece design) was a close second at 4000 psi and gave slightly better results than the
Phase III Prototype at the lower differential pressures used in the tests. However, data taken in sandy
mud and reported by Tarr et al [1998] showed a much poorer performance for DSSV2. Although not
subjected to the sandy mud test protocol, the Phase III design would be expected to perform better
because the ball and seat are held in firm contact. Thus, solids cannot pass between the ball and seat and
enter the valve body. Note that most DSSVs should not be expected to open with less than 400 ft-lbf
of torque at differential pressures above about 1000 psi.
0
100
200
300
400
500
600
0 1000 2000 3000 4000 5000
Pressure Differential across Valve (psi)
T
o
r
q
u
e

(
f
t
-
l
b
s
)
Phase III Prototype
DSSV1
DSSV2 (Premium)
DSSV2 (Premium) Tarr et al - 16 ppg mud
DSSV3 (Trunnion Ball, Floating Seats) Tarr et al - Water
DSSV3 (Trunnion Ball, Floating Seats) Tarr et al - 16 ppg mud
Maximum Goal
Figure 40: Maximum Torque to Open DSSV as a function of Internal Pressure with Atmospheric
Pressure above DSSV
3 6
As in the previous tests, the observed maximum torque to operate the Phase III Prototype was close to
the theoretical values calculated at low differential pressures, but about 30% higher than the theoretical
values calculated at high pressure differentials. As discussed above, it is believed that the reason for this
discrepancy was because not all of the pressure loading was being transferred to the stem bearings as
was assumed in the theoretical analysis.
The maximum torque to open the Phase III Prototype DSSV with a higher pressure inside the valve
than above, but with pressures above the valve higher than atmospheric is shown in Figure 41. This test
was more severe than the test to produce Figure 40 because of the much higher internal pressures
reached. The data points are grouped at differential pressure increments of 500 psi as described in the
test procedure. The bias pressures shown are the pressures above the ball and upper seat. For example,
at a pressure differential of 500 psi in Figure 40, the lowest point shown corresponds to the maximum
torque to operate the DSSV with 500 psi above the valve and 1000 psi in the valve. The highest point
shown corresponds to the maximum torque to operate the valve when the pressure above the ball and
upper seat is 3500 psi and the pressure in the valve is 4000 psi. The theoretical design curve for
differential pressure is for a zero pressure bias. Up to 4000-psi differential, the only data point that
exceeded 400 ft-lbf was for a 2500-psi pressure bias.
3500
4500
1000
3000
1000
2500
0
2500
0
100
200
300
400
500
600
0 1000 2000 3000 4000
DIFFERENTIAL PRESSURE (PSI)
TORQUE (FT-LBS)
Phase III Prototype
Maximum Goal
4000
3500
5500
3500
6000
500
500
500
Figure 41: Maximum Torque required to Operate the Phase III Prototype as a function of Pressure Differential
acting on the Ball and Upper Seat. (Various points shown at same differential pressure are for series of
Bias Pressures held above Ball and Seat at 500-psi increments. The Numbers shown identify the
Minimum and Maximum Pressures held above Ball and Seat.)
3 7
Figure 42 shows the maximum torque required for closing the Phase III Prototype when that valve has
unweighted drilling mud flowing through it. Flow rates of 100 to 350 gallons per minute were used in
the tests. The equalized pressure in the DSSV when it is initially open is 500 psi as set by the choke
downstream of the valve (See Figure 33). When the DSSV is closed, the line pressure increases up to
2000 psi maximum before the automatic choke in parallel with the valve began operating to hold the
pressure constant at 2000 psi. In addition, due to the quick closing of the DSSV under flow, a water
hammer pressure may occur. This pressure depends upon the flow rate and is also given in Figure 42.
It should be noted that this value is an upper bound as it is based upon instantaneous closing of the
valve in a relatively short test loop. The valve was closed quickly (~ 1 sec) as recommended based on a
previous study by BP-Amoco. This may have reduced the pressure build-up in the valve due to
throttling. A commercial DSSV was also used in the flow tests to allow a comparison to be made to
the performance of the Phase III Prototype.
The Design Goal for the Phase III Prototype was met over the full range of flow rates studied.
However, the measured values of torque were much higher than those expected based on a theoretical
analysis of the valve design. In addition, the commercial DSSV could be closed in this test at lower
values of torque then the Phase III Prototype. It is believed that the unexpected results could have been
due to a change in the internal condition of the Phase III Prototype. The prototype had been
disassembled and reassembled several times prior to running this last series of tests. Erratic test results
for commercially available DSSVs were also seen during other tests performed during the study.
1250 1000 750
WATER HAMMER
PRESSURE (PSI)
0
100
200
300
400
500
600
0 50 100 150 200 250 300 350
FLOW RATE (gal/min)
T
O
R
Q
U
E

(
f
t
-
l
b
s
)
LSU PHASE III
DSSV0
Maximum Goal
Poly. (LSU PHASE III)
Poly. (DSSV0)
1500 1750 500
Figure 42: Maximum Torque required for Closing against Flow
D E V E L O P M E N T O F D S S V S T O R A G E S T A N D
3 8
Development of DSSV Storage Stand
I t was observed in the test program that the torque required for operating a D S S V
changed with time and use. T his is not surprising since the internal moving parts
are ex posed to the slurries periodically pumped through the valve and to
weathering. I n addition, the lack of lubrication of the mating parts between the
stem and the ball more susceptible to binding if some parts are slightly out of
tolerance. T he purpose of this phase of the research program was to design a storage
container that will preserve D S S V s in a lik e new work ing condition and allow for
easy accessibility at the rig floor, thus resulting in improved reliability.
he design goals were determined by the student design team after talking with a number of
different drilling operators. It was found that more than one size DSSV is often required on
the rig floor at all times making it necessary for the stand to accommodate more than one
DSSV. The storage container must hold the largest diameter of DSSVs listed in API
Specification 7, which is 7.875 inches. In order to maintain the DSSVs in a like new condition, it is
important that corrosion is prevented and that debris is not allowed to solidify on the moving
components. In order to account for the quick retrieval of the DSSV, in less than 10 seconds, there
must be a fast and easy way of removing it from storage. Since the valves are also used off shore,
the container and its components must be environmentally safe. Due to the corrosive environment
offshore the stand must be made to resist all conditions and maintain its integrity for a design life of
five years. In addition, it was believed that industry would accept the new technology more readily if
the cost of the unit were below $4,000.
Design
The design proposed is a storage structure, containing four storage cylinders. Each storage cylinder
contains enough oil for a DSSV stored in the cylinder to be immersed in oil. A mechanism to
quickly lift the DSSV out of the oil bath is driven by a hydraulic pump. In the prototype, the pump
is manually actuated to supply the required fluid and pressure to lift each valve from its cylinder. A
single fluid reservoir located centrally, supplies oil to the pump. Machine drawings for the prototype
storage structure are given in Appendix B.
The storage stand consists of two basic components: the storage structure and the hydraulic system.
The storage structure includes the all the components that act as containment, including a side plates,
bottom plate, lid, and maintenance panel. The hydraulic system consists of all parts used to store the
valves in the bath as well as those components which raise or lower the valves. The basic
Chapt er
5
T
3 9
components of the hydraulic system include: the storage cylinders, pistons, pump, fluid reservoir,
control valves located on a side control panel, and the tubing with the necessary fittings. A schematic
of the hydraulic system is shown in Figure 43.
Hydraulic Pump
Because most prefabricated pumps are hand actuated and can be purchased at a much lower cost
than a customized pump the design team decided to purchase a hand actuated hydraulic pump
rather than fabricate a customized pump. In order to find the appropriate pump needed for the
hydraulic system all of the forces needed to be accounted for. The first step was to determine the
resistive forces on the piston. These forces include the weight of the safety valve, the weight of the
piston, the weight of the fluid above the piston and, the resistance caused by the friction on the o-
ring. The maximum weight of a DSSV was found to be 250 1bf. The piston that was designed
weighs 29 lbs. The maximum that the fluid would weigh is when the piston is at its lowest position
was found to be 59 lbs. The frictional force between the O-ring and cylinder wall was calculated to
be about 20 lbf. Thus the total weight that needs to be overcome was found to be about 360 1bf,
which corresponds to a pressure of five psi in the lower lift chamber. To allow for head losses,
unexpected frictional forces in the system, and ease of pump operation, a pump that could handle a
pressure of 25 psi was chosen.
The pump that was selected is a Fill-Rite hand pump intended for pumping fuel from 55-gallon
drums. The Fill-Rite is easily attachable to any threaded mounting and is small enough to fit inside the
storage structure between the storage cylinders. The fluid displacement of the pump is 35 cubic
inches per stroke, which results in a 0.5-in lift of the valves per stroke. Only 10 strokes are required
OVERFLOW
RESERVOIR
CONTROL PANEL
Figure 43 Schematic: for Hydraulic System for Proposed New DSSV Storage Stand
4 0
to raise the valves the desired 5 inches. Each stroke of the pump took less than 1 second to execute
allowing the minimum lift time requirement to be met. Also, the Fill-Rite operates at maximum
pressure of 25 psi.
A check valve was placed at the inlet of the control panel to relieve any pressure that the pump
would experience while the valves were in their lifted positions. The check valve holds the pressure
of the fluid and valve weight exerted when the valves are completely lifted. The maximum pressure
the check valve can withstand is 250 psi making it more that sufficient to handle the pressure exerted.
This was a precautionary measure implemented to improve the life and reliability of the hydraulic
pump.
Hydraulic Cylinders
The hydraulic cylinders were designed to accommodate all sizes of DSSVs listed in API spec 7. The
maximum height of a safety valve was found to be 2 inches. The maximum size that the cylinders
can hold is 30 inches. The largest outer diameter of a safety valve was determined to be 7.875 inches
so, the cylinders were made to hold valves up to 9.5 inches in diameter. An overflow pipe was
welded on to the top of the hydraulic cylinders. One-inch pipe was welded below the over flow to
allow for the flow from the lower chamber to the upper chamber. Also, one-inch pipe was welded
at the bottom of the cylinder to allow the fluid to flow out of the lower chamber. To attach the
cylinders to the bottom plate of the outer storage container, three steel feet were welded to the
bottom of each of the cylinders.
For a piston cylinder devise to work properly there are many factors that must be considered during
the design. A surface finish between 8 to 16 micro-inches was selected to minimize frictional forces.
A Nitrile O-ring material was selected to be compatible with the oil used in the hydraulic system and
oil bath. A gasket width of .210-in was selected for the pump and .275-in for the lift. Since the
pressure requirements were very low for a piston, a large tolerance could be used. This helped
provide a low-cost construction process.
One of the concerns of the proposed design was the possibility of the DSSV tipping over during
removal or resting on the inside wall of the storage cylinder. To prevent lateral movement of the
DSSV, a threaded shaft was inserted into the piston. The maximum allowable tilt angle with the rod
inserted is 7 degrees or 0.75 inches laterally. The rod was designed against bending and shear failure.
The rod is 314 stainless steel and the dimensions are 10 inches long with a minor diameter of 7/ 8
inch on the threaded part located at the base of the piston. Using the lateral load of 250 pounds and
the dimensions and material properties of the rod, it was determined that the rod had a factor of
safety of 1.3 against bending failure and 240 against shear failure.
Fluid Reservoir
The reservoir was built to store extra fluid that gets displaced when safety valves are inserted and to
collect drilling fluid solids that settles out of the oil. The reservoir is made of 10-inch OD steel pipe
with four two-inch openings for the overflow pipes that lead to each cylinder. There is also a one-
inch drain at the bottom to allow for cleaning. The reservoir was attached to the pump intake using
one-inch pipe located 10 inches from the bottom of the reservoir. It was put near the center of the
reservoir so the only oil would get sucked out to the pump in the event water and solids began
4 1
collecting in the bottom of the reservoir. Three steel feet were welded to the bottom of the reservoir
to attach it to the bottom plate.
Hydraulic Fluid
The hydraulic fluid must bathe and lubricate internal valve parts and create a non-corrosive
environment ensuring the valve is in a like-new condition if needed. Next, the fluid must be
environmentally friendly, in order to be compatible with offshore use. The hydraulic fluid must also
have the versatility to remain operable under the range of temperatures experienced and maintain a
reasonable viscosity to allow for easy pumping. The viscosity of the hydraulic fluid is not a real
factor in the performance of the system due to the low pressures and velocities. Additionally, the
head loss the hydraulic fluid experiences while being pumped through the system is negligible due to
the large diameters of tubing and valves.
The fluid selected is called EAL-224H Hydraulic Fluid, and is produced by Mobil. The EAL-224H
has excellent corrosion and rust resistance creating an excellent environment for the valves. In
addition, 224H is recommended for use between 0-180 degrees Fahrenheit. This temperature was
felt to be suitable for rig floor conditions. The 224H is not only non-toxic, but is also bio-
degradable. It is recommended in marine applications.
Cont rol Panel
Since there is only one pump for the four hydraulic lifts, a control panel was needed to direct the
fluid flow that will determine which of the DSSVs is accessed or stored. To do this, hydraulic lines
were run to a control panel containing ball valves. In order to lift a DSSV, the ball valve for that
cylinder is opened and the drain valve is closed. The valve positions are reversed when lowering a
valve into the storage position.
St orage St ruct ure
Since the DSSV Phase III valve is 27 inches long, and most commercial valves are no longer than 32
inches in length, a height for the storage stand of 3 feet was selected. In order to house a maximum
of four valves with an outer diameter of 6 to 8 inches and have a small footprint on the rig floor
area, a 3-foot cubic stand was chosen.
The materials considered for the stand construction included stainless steel, aluminum, and carbon
steel. Aluminum would have been a good choice as far as weight and moderate strength, but failed
in the area of corrosion resistance. Stainless steel was strongly considered because it is non-corrosive,
has high strength, and is very durable. However, manufacturing costs would have caused the design
goal to be exceeded. After consulting with several oil-field service personnel and steel distributors, it
was discovered that the vast majority of containers used offshore are made from regular carbon
steel that has been sandblasted and painted. The design chose carbon steel for construction of the
prototype DSSV Storage Stand.
The components of the storage structure considered were the bottom plate, side plates, back wall,
the lid, and maintenance panel. Due to the heavy weight of the hydraulic system, the bottom plate
4 2
needed to be thicker than the side and back plates. After
performing deflection calculations, one-quarter inch carbon
steel was selected for the bottom plate. The remaining
panels were constructed from 3/ 16-inch plates.
In order to make the stand portable, consideration was
given to placing a set of locking wheels on the bottom of
the stand. However, after consulting with industry personnel
familiar with rig floor conditions, this idea was discarded in
favor of the use of pad eyes that would allow the unit to be
moved with an air hoist. Originally only two pad eyes were
considered with one on each side panel. After the hydraulic
system was designed, it was decided on to use four pad
eyes (two on each side panel) to ensure that the stand could
not tip while being moved with the air hoist. In addition to
pad eyes for moving, there should be some means of
moving the stand with a forklift. To accomplish this, the
stand was elevated using two half-moon rubber skids. Lid
accessories include the hinges, handle, latches and rubber
seals. These accessories have all been designed previously by
an oil-field service fabricator, and are standard for this type
of application.
Assembly and Testing
The DSSV Storage Stand was assembled and tested at the LSU Petroleum Engineering Research
Well Facility. It was tested and found to function as designed. All design goals were met.
Shown in Figure 44 is the top portion of a DSSV located on a rig floor among other equipment. It
is in a ready-to-stab configuration with a lifting sub screwed into the top of the valve. This is the
arrangement commonly used by the drilling industry today. Shown in Figure 45 is the prototype of
the proposed DSSV Stand.
Figure 44: DSSV and Inside BOP on
Rig Floor Ready to Stab
on Drillstring if Needed
4 3
Figure 45: Proposed DSSV Storage Stand with Control Valves on Side.
C O N C L U S I O N S A N D R E C O M M E N D A T I O N S
4 4
Conclusions and Recommendations
T his study was undertak en by L S U under the sponsorship of the M M S to improve
safety and reduce risk of blowouts through the drill string. M M S has been
mandated by congress to insure that the best available and safest technology is
utiliz ed by industry in developing our natural resources on the outer continental
shelf.
robust Drill String Safety Valves DSSV is difficult to design because the DSSV must fit in the
drillstring and be full-opening enough to allow wireline work to be done through the valve.
The valves have been shown to have one of the highest failure rates of the blowout control
equipment routinely used on drilling rigs. Blowouts have occurred on a number of occasions when a
DSSV could not be closed to stop flow from the well up the drillstring. Other failure modes have
also been reported that have not caused blowouts but have made well control difficult to achieve.
Most DSSVs that are commercially available are primarily quarter-turn ball valves in which the ball
floats between fixed seats and depend on well pressure to force the closed ball against a seat to
form a seal. This design minimizes the crushing force placed on the ball and distributes stresses
through the seats over a wide area. However, interference between parts can result during ball
rotation if dimensional tolerances are not met. A relatively new DSSV design has recently become
available that utilizes a trunnion mounted fixed ball and floating seats. This design reduces the risk of
interference between parts during ball rotation but results in more concentrated stresses in the valve
stem and ball.
In this project, DSSV failure modes were studied in detail and procedures for reducing the risk of
blowouts due to a DSSV failure were identified. A novel DSSV design was explored that used a
trunnion-mounted ball and spring loaded seats was constructed and tested by an interdisciplinary
research team composed of Petroleum Engineering and Mechanical Engineering seniors and faculty.
The novel design results in operating torque being less affected by internal pressure but also results in
more concentrated stresses in the valve stem. The test program indicated that the design objectives
of the new design were achieved. The test program also revealed that the torque required to operate
a DSSV type ball valve can change with use (especially after being operated in a high torque
condition), with internal exposure to drilling slurries, and with non-use and exposure to the elements.
In addition to a new valve design, a new DSSV storage stand was designed and tested that could
maintain a DSSV on the rig floor in a like-new condition.
As a result of the work done in this study, the following conclusions and recommendations were
made:
Chapt er
6
A
C O N C L U S I O N S A N D R E C O M M E N D A T I O N S
4 5
Conclusions
1. Failure of a DSSV to close against a high flow rate or under high pressure has resulted
in a number of blowouts.
2. Test results obtained and this study and results published in the literature show that the
torque required for operating a DSSV increase with internal valve pressure and can
exceed 400 ft-lbf at pressures as low as 1000 psi.
3. The new prototype design developed in this work shows promise of reducing the
operating torque at high pressures.
4. High pressure on a DSSV valve before closing can occur when there is a significant
length of piping and or restrictions downstream of the DSSV.
5. The link between the ball and stem in a DSSV is generally a two piece design to facilitate
disassembly and rebuilding the internal components of the valve. This design is
susceptible to binding during rotation of the ball under flow or pressure if dimensional
tolerances are not closely observed during manufacturing and assembly.
6. Low cost DSSVs utilize a threaded two-piece body construction to allow the valve to
be disassembled and rebuilt in the field. The torque used to remake the body threads
can affect critical dimensional tolerances of the component placement in the valve body.
Applying excessive torque across the valve body when making up the valve in the
drillstring can bind the ball between the seats.
7. Redundant barriers to blowouts are needed for the drillstring during tripping operations
similar to the use of redundant barriers provided in the blowout preventer stack for
stopping a flow up the annulus.
8. Experience has shown that maintaining multiple DSSVs available on the rig-floor as
stabbing valves does not always prevent blowouts.
Recommendations
1. The DSSV intended for use as a stabbing valve to stop flow through the drillstring during tripping
operations should not be used in the drillstring for other operations. The stabbing valve should be
maintained in a like-new condition and used only during periodic pressure testing with fresh water.
2. Operators and/ or drilling contractors should check threads, valve wrench, and lift sub on the stabbing
valve and actuate the stabbing valve close and open each tour.
3. Operators and/ or drilling contractors should use a drillstring float whenever practical to provide
redundant protection against a high-rate flow through the drill-string during tripping operations.
4. When floats are not used, shear rams are recommended for redundant protection against blowouts
through the drillstring during tripping operations.
5. Drill String Safety Valves should not be the only means for stopping flow from the drillstring at the
surface when reverse circulating the well during completion operations. Flow should be routed through
hydraulically operated valves and a choke manifold.
6. Drill string safety valves should not be the only means for stopping flow through the drill string when
significant piping and flow restrictions are present above the valve.
B I B L I O G R A P H Y
4 6
Bibliography
1. Chilingarian, G.V., Drilling and Drilling Fluids, Elsevier Scientific Publishing
Company, Amsterdam, 1981
2. Piper, C.F, Mud Equipment Manual, Gulf Publishing Company, Houston, 1985.
3. Tarr, Brian A.: Research Targets Drillstring Safety Valve Improvements, Gas Tips
(Spring, 1996).
4. Tarr, B.A. et. Al, New Generation Drill String Safety Valves, IADC/ SPE Paper
39320, IADC/ SPE Drilling Conference (March 306, 1998), Dallas, TX.
5. Hauser, William: Minerals Management Service (MMS) Review of Blowout Preventer
(BOP) Testing and Maintenance Requirements for Drilling Activities on the Outer
Continental Shelf (OCS), paper presented at the IADC Well Control Conference of
the Americas, 1995.
6. Dayton, R. W.; Sleeve Bearing Materials. American Society for Metals, 1949.
7. Gupta, Pradeep K.; Advanced Dynamics of Rolling Elements. Springer-Verlag New
York Inc., 1984.
8. Harris, Tedric A.; Rolling Bearing Analysis. John Wiley & Sons, Inc., 1966.
9. Mischke, Charles R. and Joseph Edward Shigley; Mechanical Engineering Design.
McGraw-Hill, Inc., 1989.
10. Warring, R.H.; Seals and Sealing Handout; Gulf Publishing Company, 1981.
M A C H I N E D R A W I N G S F O R P H A S E I I I P R O T O T Y P E D S S V
4 7
Phase III Prototype DSSV
This appendix contains the machine drawings for the Phase III prototype DSSV .
Appendi x
A
4 8
4 9
5 0
5 1
5 2
5 3
5 4
5 5
5 6
5 7
Phase IV Prot ot ype DSSV St orage St and
This appendix contains the machine drawings for the Phase IV prototype DSSV Storage Stand..
Appendi x
B
M A C H I N E D R A W I N G S F O R P H A S E I V P R O T O T Y P E D S S V S T O R A G E S T A N D
5 9
6 0
6 1

Study of Drill String Safety Valves: Adam T. Bourgoyne, JR., Lsu

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Study of Drill String Safety Valves: Adam T. Bourgoyne, JR., Lsu

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Study of Drill String Safety Valves: Adam T. Bourgoyne, JR., Lsu

Uploaded by

Copyright:

Available Formats

St udy of Drill St ring Safet y Valves

ADAM T. BOURGOYNE, JR., LSU

seal bs ball spring t np

You might also like