-------
Figures (Continued)
Number
B-3. CBL liquid flow test - Phase II ................ 55
B-4. Neutron activation tool liquid flow test - Phase I. . . 59
B-5. Neutron activation tool liquid flow test - Phase II. . . 60
B-6. Neutron activation tool liquid flow test - Phase III. . 62
B-7. Neutron activation tool liquid flow test - Phase IV. . 63
B-8. Testing for a hole in the long string ............ 66
B-9. Leak Test Well .......................... 70
B-10. Neutron activation tool liquid flow test .......... 74
B-11. Radial differential temperature survey ....... ... 76
B-12. RDT scan no-flow condition ......... '. ...... 77
B-13. RDT scan flow condition ................... 78
B-14, Leak Test Well. . . . ................... 83
B-15. Monitoring wells. . ....... . ............... 84
B-16. Standpipe ................. ............ 87
B-17. Pressure decline over time ................. 90
B-18. Changes in water level - 700-ft zone. ......... 92
B-19. Changes in water level - 900-ft zone. ... ...... 94
B-20. Leak Test Well ......................... 97
B-21. Atlas Wireline Services SONAN Leak Test Well. .. 99
B-22. Atlas Wireline Services SONAN Leak Test Well. . 100
B-23. Leak Test Well .................. ....... 102
B-24. Leak Test Well ..... ............... ...... 104
B-25. Gearhart Industries, Inc., continuous flow survey. 105
B-26. Gearhart Industries, lnc;, continuous flow survey. 106-
B-27. Gearhart Industries, Inc.,, continuous flow survey. 1 07
B-28. Gearhart Industries, Inc., continuous flow survey. 108
B-29. Leak Test Well ............... : ...... 1 10
B-30. Tracer runs showing fluid movement
during injection at 1/2 bpm 300 psi ..... ... 111
B-31. Slug #6 ejected at 1,1 25-ft channel down check. 112
B-32. Slug #7 ejected at 1,1 25-ft channel down check. 113
B-33. Leak Test Well ......... ... ........... ' 114
B-34: Leak Test Well ....... . : ......... ... 116
B-35. Leak Test Well. .....:.:... ..... . . ---- 118
B-36. Leak Test Well; ---- ............. ..... 122
VIII
-------
Tables
Number Page
B-1. Procedure for Drilling the Monitoring Wells 52
B-2. Effects of Adding Pressure to Depress Water .... 69
B-3. Depth to Water in Monitoring Wells , . 84
B-4. Water-Level Decline in Standpipe 87
B-5. Pressure Decline Over Time 89
B-6. Flow through Profile Nipple 90
B-7. Changes in Water Level - Monitoring Wei). No. 1 . . 91
B-8. Changes in Water Level - Monitoring Well No. 2 . . 93
B-9. Flow of Water (gpm) through Holes
at Different Pressures 94
B-10. Leak Test Well - SONAN Data 98
B-11. Oxygen Activation Log Data, Leak Test Well -
November 3, 1987 119
B-12. Oxygen Activation Log Data, Leak Test Well -
September 14, 1988 123
-------
Acknowledgments
This report reflects the work done to date on unique research
wells designed for testing methods of determining the mechanical
integrity of injection wells. The successful, design of the wells is due
to the time and effort which an unusually able advisory panel was
willing to devote to the project.
Grateful acknowledgment is made to the advisory group for their
Contributions:
Terry Anderson
Halliburton Cementing Services
Dick Angel
Phillips Petroleum
Al Bryant
Schlumberger Well Service
Mike Cantrell
Oklahoma Basic Economy Corp.
Cecil Hill
Baker Packers
Gene Littell
Litteil and Randolph Engineering
Tal Oden
Oklahoma Corporation Commission
R. C. Peckham
U.S. EPA, Region VI
Gary Batcheller, Schlumberger Well Service, and Alerdo Maffi,
Tom Hansen Company, both made incalculable Contributions to the
project through their advice, encouragement and participation in a
training course for EPA and State employees on October 16, 17 and
18, 1985.
-------
Introduction
Underground injection control regulations of the United States
Environmental Protection Agency (U.S. EPA) require that new
injection wells demonstrate mechanical integrity prior to operation and
all injection wells demonstrate such integrity at least every 5 years.
The regulations state that an injection well has mechanical
integrity if: .
(1) There is no significant leak in the casing, tubing or packer.
(2) There is no significant fluid movement into an underground
source of drinking water through vertical channels adjacent to
the injection well bore.
The initial research project to examine the question of mechanical
integrity was funded July 1, 1981. The three-phased project was to
determine the state-of-the-art for mechanical integrity testing of
injection wells and to test specific field methods to determine their
adequacy.
The first phase of the project resulted in a report, "Methods for
Determining the Mechanical Integrity of Class II Injection Wells."
Although this report represented the state-of-the-art for determining
mechanical integrity for Class II wells, the technology described may
be applied to other classes of injection wells.
-------
Mechanical integrity Test wens
The second and third phases of the project involved construction
and testing of wells designed to evaluate various tools and techniques
used to determine mechanical integrity of injection wells. The test
wells - two "logging wells," a "leak test well," a "calibration Well,"
and three "monitoring wells" - were designed for developing methods
testing the integrity of the tubing, casing and packer; locating fluid
movement in channels behind the casing; and testing for channels in
the cement behind the casing. The wells are located on a 110-acre
site approximately 11 miles west of Ada, Oklahoma.
Logging Well No. 1
The purpose of this well is to determine the present capability in
the industry for evaluating the cement bond between the
cement/casing and cement/formation coupling in injection wells, and
to provide a test facility for evaluating new tools developed for cement
evaluation.
After much discussion among members of the advisory group, it
was determined that the best method to simulate poor cement
bonding, or channels in the cement, would be to attach water-filled
polyvinyl chloride (PVC) pipes to the outside of the casing. Thus,
PVC pipe was attached to the outside of the casing to cover either
90, 60, 30 or 6 degrees of the 360 degree radial surface of the pipe
(Figure 1).
Having installed the "channels" on the casing, attention was
turned to a second vital factor in the completion of this well, the
quality of the cement job. The planned cementing program was
designed to provide the most favorable conditions for obtaining
excellent bonding of the cement to the casing and coupling of the
cement to the formations, so that the "channels" identified by the
logging tools would be those purposely created for the project.
A thorough review of the logs run to evaluate the cement bonding
indicates that about 60 percent of the well has good cement bonding
and provides an excellent facility for determining the sensitivity of
various down-hole cement evaluation techniques. The other 40
percent of the well provides an opportunity for testing techniques for
repairing channels in cement, and for evaluating the success of the
repair efforts.
The well specifications, along with a detailed description of the
installation process, are provided in Appendix A.
Cement Evaluation
With the completion of the well, the actual testing portion of the
project, determining the present capability for evaluating the cement,
-------
12-1/4" dia borehole
9 5/8-in dia
easing
8 5/8-in dia.
casing
7-in dia
casing
840-tt depth
5 1/2-in dia.
casing
4 1/2-in dia
casing
tt depth
^
\m m m
[ * * *
{
mmmmmm
mmmm.mm
m mmmmm
mmmmmm
*
mmmmmu
ซ
"H"
mm mmmm
t m mmmm
mmmm**
mmmmmm
mmmmmm
mmmmmm
mmmmmm
*
*
f-^tmmm
mmmm
mmmm
mmmm
mmmm
mmmm
mmmm
mmmm
mm m m
mmmm
mmmm
mmmm
mmmm
mmmm
mmmm
mmmm
mmmm
mmmm
mmmm
mmmm
mmmm
mm m
m m m
mmm
mmm
mmm
mmm
mmm
mmm
mmm
mmm
cement-filled annulus
X
formation wall
well casing
8 3/4-in dia borehole
Figure 1. Logging Well No. 1.
-------
was ready to proceed, uontact was mauซ wim as many
companies as possible to determine the types of tools that are being
used for evaluating cement in a well and to run as many different
tools as possible in the well. At the initial contact with each company
contracted to log the well, the construction and purpose of the well
were fully explained, except for the location of the manmade
channels. Each company representative was also asked to provide a
complete interpretation of the condition of the cement in the well,
based on the information from the -company's log, prior to their
leaving the site.
Nine companies have produced 16 logs on the well, two
companies have refused to run a log on the well.
Logging Tools
Basically, two generations of logging tools have been run in the
well: the "cement bond" tool, consisting of single transmitter/single
receiver or single transmitter/dual receivers; and the "cement
evaluation" tool which has eight ultrasonic transducers.
The typical cement bond tool presents a log with the following
data: gamma-ray and casing collar locator (CCL), which are included
for depth control; transit time (TT), which measures the time it takes
for a certain level sound wave to travel from the transmitter to the
receiver; amplitude, which measures the strength of the first
compressional cycle of the returning sound wave; and a graphic
representation of the wave form, which displays the manner in which
the received sound wave varies with time. This representation is
called variable density log (VDL), seismic spectrum, or micro-
seismogram, and is a function of the property of the material through
which the signal is transmitted.
There are various transmitter/receiver spacings available, the
most common being a single transmitter with a single receiver located
3 feet away (Figure 2). Other tools include the single transmitter/
single receiver with 4-foot or 5-foot spacing, or a single transmitter/
dual receiver with 3-foot/5-fobt spacing (Figure 3).
The "second generation" tools for determining the adequacy of
cement bonding include a tool having eight ultrasonic transducers
spiraled around it to survey the circumference of the casing (Figure
4). The information presented on the log from these tools includes
casing ovality; average casing I.D.; casing collars; hole deviation; fluid
velocity; eccentering of the tool; rotation of the tool; gamma-ray,
maximum and minimum cement compressive strength; average of the
energy returned to all eight transducers; and cement distribution
around the casing. '
-------
Electronic
Section
Trans.
Acoustical
Section
Receiver
Borehole
Liquid
Casing
Bonded Cement
Sheath
Sonic Pulse
Path
_ _
Formation
Figure 2. Cement bond tool - single transmitter/receiver.
Log Interpretation
The bonding of cement to casing can be measured quantitatively,
but the bonding, or rather the coupling, of cement to the formation is
only a qualitative estimate. Therefore, when attempting to evaluate
cement in a well, it is extremely important to obtain as much
information as possible.
The components of the sound wave that are of primary interest
when analyzing a "bond log" are the casing, formation and fluid
(mud) signals. Each medium has different characteristics, thus the
-------
Transmitter
3 ft Receiver
5 ft Receiver
Figure 3. Cement bond tool - single transmitter/dual receiver.
sound waves will have different amplitudes and velocities. Figure 5
indicates these wave forms and a composite signal.
A recommended approach to evaluating the cement bond log is to
first determine the information available from the graphic
representation of the wave form (VDL), then examine the amplitude
curve to see if the two are in agreement. For example, if the casing
diameter and transmitter/receiver spacing are known, the transit time
for the casing arrivals can be predicted. Figure 6 is a chart that, for
practical field or reference purposes, gives an idea of the approximate
transit time for the casing signal for various tool spacing and casing
I.D. By examining the VDL, the time, in microseconds of the first
arrival, can be determined. This time can then be checked against the
chart to determine if it is a casing signal.
The fluid, or mud, wave has a velocity of about 189
microseconds/foot, and its arrival can be predicted, if the tool spacing
is known, by multiplying the tool spacing by 189. The fluid wave has a
-------
Eight
Ultrasonic
Transducers
Fluid
Velocity
Transducer
.3X^,3 Casing
ฃ:H?~ Bonded -.--
3-I-I-I-I-I Cement Sheath
-- Formation
Figure 4. Second generation tooL
destructive interference; thus when it enters the receiver, distortion of
the wave occurs. Because of this, the only part of the VDL that is
useful for interpretive purposes is that part that reaches the receiver
prior to the arrival of the fluid wave.
The "second generation" tools generate a pulse of ultrasonic
energy from each of the eight focused transducers that are arranged
around the circumference of the tool. The strength and duration of the
echoes reflected from the casing and cement are used to form an
image of the cement distribution and quality around the casing. This
information and the cement compressive strengths are two very
useful pieces of data for evaluating the casing/cement bonding in a
well.
As stated earlier, 16 different logs have been produced from the
well. Appendix A contains a detailed comparison of specific sections
of the well that have been logged by "first generation" tools with
single transmitter/receiver, 3-, 4- or 5-foot spacing; and the second
generation ultrasonic logging tool.
-------
1r
Fluid
Casing
Ir
Formation
Composite
Figure 5. Composite wave form.
Logging Well No. 2
Analysis of the tests completed on Logging Well No. 1 identified
two areas that needed further investigation. As has been stated,
certain logging tools located all of the 90 and 60 degree channels, all
but one of the 30 degree channels, but none of the 6 degree
channels. Thus, the detection limit for the tools is somewhere
between 30 and 6 degrees. The second area of concern relates to
the inability of presently available equipment to evaluate cement
behind fiberglass pipe.
To address these concerns, Logging Well No. 2 was designed
and constructed with both steel and fiberglass pipe and with channels
-------
Aft
Din
4
5
6
7 .
3
235
250
265
280
4
292
307
322
337
5
349
364
379
394
A ft - Receiver spacing
D in - Casing diameter
Figure 6. Field reference transit time.
on the steel pipe covering 30. 25, 20, 15 and 10 degrees of the 360
degree radial surface of the pipe (Figure 7).
The well specifications, along with a detailed description of the
drilling and completion process, are provided in Appendix A.
Cement Evaluation
With the completion of the well, the process of determining the
capability for locating channels on steel pipe and evaluating cement
behind fiberglass pipe was ready to begin. Only those logging tools
that produced satisfactory logs in Well No. 1 were run on Well No. 2,
with the exception of some experimental tools that have been run in
the well but are not yet available for general use.
Ten logs have been produced on the well. Of these, four were
produced from experimental tools.
Logging Tools
Three generations of logging tools have been run in the well. The
"cement bond" tool consists of a single transmitter/dual receiver with
the receivers spaced 3 feet and 5 feet from the transmitter. The
"cement evaluation" tool has eight ultrasonic transducers spiraled
around a 2-foot section of the tool. Prototype tools are not yet in
commercial use.
Well Logging Conclusions and Recommendations
Well Completion
Greater care must be exercised in planning the cement job and In
carrying out that plan when cementing injection wells, especially
Class I wells where cement is to be circulated to the surface around
the long string. The plan should include equipment and activities that
will enhance the possibility for obtaining the best cement bonding
possible. This should include the use of a caliper log to determine
exact hole size to better estimate the volume of cement necessary to
-------
5 1/2-in dia Casing
l57S-ft Depth
Fiberglass Casing
8 3/4-in dia. Borehole
Figure 7. Logging Well No. 2.
complete the well; properly conditioned drilling mud prior to beginning
the cementing operation; centralizers, to ensure that the casing is
centered in the hole; pre-flush, to help clean out the hole prior to
pumping cement; rotating and/or reciprocating the pipe during the
cementing operation to further aid in cleaning out the hole; and at
least 100 percent excess cement. The experience in cementing
Logging Well No. 1 indicates that those areas where the greatest
volume of cement flowed past had cleaner holes and better cement
bonds, thus the use of 100 percent excess cement will enhance the
probability of a good cement job throughout the casing length.
10
-------
Yn Logging Well No. 1, although cement was circulated to the
surface, after the cement set for 72 hours, the top of the cement
behind the casing was 132 feet below land surface. This "fall back"
of the cement behind the casing must be monitored and corrected so
that there is cement fill-up behind the casing to the surface of the
ground.
Because of the problems encountered in cementing Logging Well
No. 1, extreme care was taken to ensure that a good cement job was
obtained in Logging Well No. 2. The top of cement behind the casing
was 42 feet below land surface in this well, as opposed to 132 feet in
No. 1. Approximately 1,075 sacks of cement were used to cement
Well No. 2 with about 500 sacks circulated to the surface during the
cementing operation.
Logging Equipment
None of the 'egging tools presently available located any of the 6
degree channel? in Logging Well No. 1. The second generation tools
located all of the 30, 60 and 90 degree channels that were designed
into and couM be identified in the well. A calibrated single
transmitter/dual .eceiver cement bond tool with 3-foot/5-foot spacing
located the 60 and 90 degree channels and all but one of the 30
degree channels. The other cement bond tools with single
transmitter/single receiver 3-foot, 4-foot, or 5-foot spacing presented
very inconsistent results.
The 3-foot spacing is the best currently available for measuring
and evaluating the amplitude of the first compressional arrival and the
attenuation of this signal is a measure of the bonding of the cement
to the casing. However, this spacing is not satisfactory for
determining data on or evaluating the relationship of the cement to
the formation. Five-foot spacing between the transmitter and receiver
is the best currently available for evaluating the relationship of the
cement to the formation, but it is not accuratetfor determining bonding
to the casing. Four-foot spacing is being used; however, it does mot
have satisfactory resolution for evaluating the relationship of the
cement to either the casing or formation.
The significant fact remains that none of the tools located
channels smaller than 30 degrees in the well. Such channels
represent a significant avenue for movement of fluid and methods
must be developed to locate these and even smaller channels. It is
recommended that the logging industry continue research efforts
toward increasing the sensitivity of the logging tools.
The research conducted on Logging Well No. 1 indicates that with
the presently available tools, the ideal approach for evaluating the
cement in an injection well is to run both the second generation tool
.and a calibrated cement bond tool with single transmitter/dual receiver
-------
3-foot/5-foot spacing. This combination gives the most information for
interpretive purposes.
An alternative to this approach is the use of either the second
generation tool or a calibrated "bond tool" with single transmitter/dual
receiver 3-foot/5-foot spacing. The second generation tool gives no
information on the cement/formation coupling, but gives excellent
information on the cement/casing bonding and its presentation allows
for easy interpretation. The "cement bond" tool provides information
on both casing/cement bonding and coupling to the formation, but is
somewhat harder to interpret and may be less sensitive in some
specific situations.
Calibration of both tools is imperative for reliable data to be
produced. The size and weight of the casing must be available for
use with the second generation tool. A standard shop calibration of
the cement bond tool is essential to its use and must be included for
there to be any chance of obtaining reliable information. Quality
control on the "cement bond" tools can be included, to some degree,
on site, in that certain checks can be made to determine whether or
not the tool is working properly.
Some of the checks that can be made include:
1 If the well contains free pipe, the chevron effect must be
obvious. The chevron effect is the "W" seen opposite casing
collars in free pipe. Figure 8 indicates a bond log with free
pipe. Note the well-developed chevron effect opposite the
casing collars.
2. In free pipe, certain casing diameters call for certain
amplitude readings. For example, for 5-inch (I.D.) casing the
amplitude should read about 74 millivolts (mv); 7 inch - 60
mv; 8 inch - 55 mv; 9 inch - 30-35 mv. Such information can
be used to determine if the tool has been calibrated. Figure 8
indicates an amplitude reading of over 60 mv in 9-inch free
pipe. This indicates that the tool was not calibrated.
3. In free pipe, the transit time from a properly centered tool
should be constant, except for the influence of the casing
collars. In Figure 8, the decrease in transit time (immediately
below the arrow) and the corresponding decrease in
amplitude, indicates a slightly eccentered tool. Figure 9 also
indicates free pipe. Note, however, the wavy free pipe signal
on the VDL and the significant drift of the transit time (over
one-half of a chart division). This indicates an improperly
centered tool.
4 The fluid wave should be visible on each wave form
presentation (VDL). If the fluid wave is not visible, this
indicates a low response tool and its use should be
questioned.
12
-------
8
Figure 8. Amplitude, chevron and free pipe.
An API committee is presently working on standardizing a
calibration system. This work should be encouraged and continued
until an adequate system is developed that can be used throughout
the industry.
-------
Figure 9. Tool not centered.
The cement bond tools and second generation logging tools run
in Logging Well No. 2 located the 30, 25, 20 and 15 degree channels.
However, none 'consistently located the 10 degree channels on the
casing. A 10 degree channel on the 5 1/2-inch casing represents
about 1/2 inch.
14
-------
One of the prototype tools consistently located the 10 degree
channels. This is very encouraging; however, much more research
needs to be completed on these new tools to be assured of their
capability for locating channels in cement.
Only one tool presented data that could be used to evaluate the
cement behind fiberglass casing. This second generation tool had
been modified to respond to the resonance of fiberglass and seemed
to work well in the test well. However, the tool is not available
commercially and will not be available in the foreseeable future.
Additional research will be done in these two areas during the
next 3 years of the mechanical integrity research project.
Log Interpretation
A significant problem that became apparent during the research
efforts, especially on Logging Well No. 1, was the inability of some
logging company personnel to interpret the logs they produced.
Several companies had good equipment, but onsite personnel
apparently did not have sufficient knowledge of the equipment to
properly operate th' system or interpret the log. This is a critical
concern if these too , are to be used to evaluate the cement in an
injection well. The egging company, injection well owner, or
regulatory agency must have trained personnel that are capable of
evaluating a log to determine as much information as possible on the
quality of the cement behind each casing string.
Leak Test Well
The purpose of the Leak Test Well is to provide a facility to
develop methods for testing the integrity of the tubing, casing and
packer and for testing the capability of various down-hole tools to
detect fluid movement behind the casing.
The design of the well generally corresponds to a typical salt
water disposal well used in the oil and gas industry. That is, it
includes the use of surface casing, long string, tubing and packer.
The deviation from the norm in this well includes two packers and a
sliding sleeve on the injection tubing and a 2 3/8-inch tubing attached
to the outside of the long string and running to the surface (Figure
10). Detailed discussion, on well design and installation is provided in
Appendix B.
Flow into the well can be controlled so that the injected fluids are
directed into the 2 3/8-inch injection tubing, or to the 2 3/8-inch leak
string. Returned flows can be controlled from the 2 3/8-inch leak
string and also from the annulus of the 5 1/2-inch casing (Figure 11).
Monitoring wells have been constructed to each of the zones!
open to the Leak Test Well (Figures 12, 13, and 14). These are
multipurpose wells that are being used to monitor pressure changes
-------
iiH
4-
5-
ป
.V."
;&
,*****
**** *
gat
A A *
ooo
ฃฃ8SS3&&8S&i
^H_I
t 1
* 1
\
P
_^
III
V""
ZZ
M
I
/
&
^
aSul M
10
I
<
%"
I
f
r
*~
:::
>/:"'"
Xvlv
A * * * A
A A A A A
A A A A A
fjQO ft
rrr-8 ' , ' '
'.. 710 ft
tins ft
7
,,, ,,,, ., ...'. 935 ft
6
1057 ft Depth of
Upper Packer
Cement
1. Baker Model "C-1
2. 2 3/S-in Tubing'
3. Baker Model "L"
4. Baker Model "R"
5. Baker Model "Ad
, 6. 2 3/8-in Tubing
7. Baker Model "R"
8. Baker Model "F"
9. 5 1/2-in Long Strir
10. Perforations
1084 ft Depth of
Lower Packer
t ion ft
HMIIM i .1 1 1 ii >.;1 1 e\l n
1130ft
Injection Zones '
C-1 " Tandem Tension Packer
g'
"L" Sliding Sleeve
"R" Profile Nipple
"Ad-1 " Tension Packer
Profile Nipple
Profile Nipple
Figure 10. Leak Test Well.
16
-------
5 1/2-in Casing
,_Head
2 3/8-in Long
String Tubing
High Pressure
Steel Pipe
/
2 3/8-in Leak
String
Flow Return to
Water Supply Tanks
5 1/2-in Annulus Connection
Figure 11. Injection well head and flow lines.
in the respective zones during injection and to determine the
capability to evaluate certain types of cement. Well No. 1 was
cemented using a latex cement, No. 2, a spherelite cement, and No.
3, a foam cement. Detailed discussions on monitoring well design and
installation are provided in Appendix B.
A number of tests have been conducted or are planned for the
Leak Test Well. In addition to a test to determine the hydraulic
conductivity of the injection zone, such tests include:
Test
1. Acoustic Cement Bond Tool
2. Nuclear Activation Technique
3. Testing for Hole in Casing
4. Ada Pressure Test
5. Nuclear Activation (PDK-100)
6. Radial Differential Temperature
7. Nuclear Activation Technique
( Oxygen Activation)
8. Mechanical Integrity Research
(Mud in Annulus)
(Pressure Monitoring)
(Standard Pressure Test)
(Volume vs Pressure/Monitoring Wells)
(Volume vs Pressure/Hole Size)
9. Noise Survey
10. Temperature Survey
11. Continuous Flow Survey
Date Completed
1/13/87
1/24/87
2/02/87
12/04/85
4/08/87
4/27/87
8/31/87
2/22/88
2/15/88
9/10/85
10/13/87
-------
9 5/8-in Conductor Pipe-
80 ft
680ft
Latex Cement
2-in Steel Linepipe
20 ft Long
I
Bottom of 41/2-in.
Steel Casing
,685 ft
30 ft
Shale Cup
20 ft Injection
I Zone
' 705 ft
4 1/2-in to 2-in Swedge
Surface
-12 1/4-in Borehole
Neat Cement
4 1/2-in Casing
Centralizer on Each Collar
8-in Borehole
Johnson "K" Packer,
4 1/2-in
9ft
3 7/8-in Borehole
- 2-in Stainless Steel Screen
.010-in Shot, 30 ft Long
4710ftTD
Figure 12. Monitoring Well #1.
12. Radioactive Tracer Survey 5/19/88
13. Differential Temperature Survey 11/04/87
14. Annulus Pressure Changes Due to Temperature Planned
15. Helium Leak Test Planned
16. "Mule Tail" Test Planned
17. Tracers Involving Monitoring Wells Planned
18. Oxygen Activation (Water Quality and Flow) Planned
18
-------
9 5/8-in Conductor Pipe
80ft
905ft
Spherelite Cement
Bottom
Steel
20ft
of 4 1/2-in I
Casing Nv
30ft
1910 ft
Injection
20 ft Zone
?930ft
4 1/2-in to 2-in Swedge
Surface
12 i/4-in Borehole
Neat Cement
4 1/2-in Casing
Centralizer on Each Collar
8-in Borehole
Johnson "K" Packer,
4 1/2-in
2-in Steel Linepipe
20 ft Long
3 7/8-in Borehole
-.2-in'Stainless Steel Screen
.010-in Shot, 30 ft Long
4 935 ft TD
Figure 13. Monitoring Well #2.
-------
4 1/2-in to 2-in Swedge
miiiiisiiii
*
9 5/8-in Condi
11
Foam Cei
2-in Steel Line
20ft Lo
jctor P pe -V
80ft i
30ft
A
ig
2
i
4
2
r '
Oft
k1110tt
)ft
Inject
Zone
1130 ft
i
K,
<
P: :
sir.
V
o
'ik
=
"ii1"*1*!
"^Z>
,Z^->
"4^r><
,|ll|i|lllllllll|i Surface
L 12 1/4-in Borehole
J| Neat Cement
, /\ 1/2 in Casing
Centralizer on Each Collar
4 7 7/8-in Borehole
,,,.r Johnson "K" Packer,
4 1/2-in
Bottom of 4 1/2-in
Steel Casing
- 2->n Stainless Steel Screen
.01 0-in Shot, 20 ft Long
Perforations
*^
4 1130 ft TD
Figure 14. Monitoring well #3.
20
-------
Conclusions
The viability of deep well injection as a waste management
alternative rests on the ability to emplace wastes in geologic
formations beneath and isolated from underground sources of
drinking water. Wastewater injected into wells may escape through
leaks in the well casing caused by mechanical failure within the well
or by migration of wastewater between the well's outer casing and
well bore because of faulty cementing. Although technology is
available to construct and operate injection wells that meet
mechanical integrity criteria, technology for determining injection well
mechanical integrity must be tested and proven to ensure that sound
decisions are made and that underground sources of drinking water
are protected.
No one test provides sufficient information to make a
determination of the mechanical integrity of an injection well. This
determination is made from a combination of tests which individually
provide pieges of information that must be evaluated together to
provide a basis for making an informed judgment regarding the
mechanical integrity of an injection well.
The research facilities described in this document offer industry,
state regulatory agencies and EPA a unique capability to test and
evaluate a variety of methods for determining mechanical integrity.
Such testing can" improve the confidence of industry and permitting
agencies in'approved methods and can speed the acceptance of new
methods that may be developed.
-------
References
Batcheller, Gary W., Schlumberger Well Services, Cement Evaluation
Seminar.
Gearhart, undated material, Pulse Echo Loq, Cement Evaluation,
Casing Inspection.
Maffi, A., Tom Hansen Co., Cement Evaluation Seminar.
Tom Hansen Co., undated material, Lazer Logging Systems Cement
Bond Log.
Schiumberger, undated publication, Cement Bond, Variable Density
Log.
Schlumberger, undated publication, Cement Evaluation Tool.
Thornhill .Jerry T. and Benefield, B.G., Mechanical Integrity
Research, Proceedings of the International Symposium on
Subsurface Injection of Liquid Wastes, New Orleans, Louisiana,
March 3-5, 1986.
22
-------
APPENDIX A
Logging Well Design Specifications and Installation
Procedures
Logging Well No. 1
Logging Test Well Material Specifications
Casing Weight Grade
3 joints of 9 5/8-inch 53.5#/ft N-80
3 joints of 9 5/8-inch 36.0#/tt K-55
8 joints of 8 5/8-inch 24.0#/ft J-55
4 joints of 7-inch 23.0#/ft K-55
2 joints of 5 1/2-inch 23.0#/ft N-80
3 joints of 5 1/2-inch 17.0#/ft J-55
3 joints of 5 1/2-inch 15.5#/ft J-55
2 joints of 4 1/2-inch 13.5#/ft N-80
4 joints of 41/2-inch 11.6#/ft J-55
5 joints of 41/2-inch 9.5#/ft J-55
Eguipment Size Grade
Swage nipple 5 1/2-inch x 4-1/2-inch K-55
Swage nipple 7-inch x 5 1/2-inch J-55
Swage nipple 8 5/8-inch x 7-inch J-55
Swage nipple 9 5/8-inch x 8-5/8-inch J-55
Swage nipple 9 5/8-inch x 5-1/2-inch J-55
10 centralizers 4 1/2-inch
7 centralizers 5 1/2-inch
3 centralizers 7-inch
7 centralizers 8 5/8-inch
5 centralizers 9 5/8-inch
Detailed Description of Well Construction
On August 14, 1984, the process of preparing the "channels"
was begun. PVC pipe, either 3/4 or 1/2 inches in diameter, was
sealed on one end, filled with water saturated with boric acid and
capped. The next step was to attach the PVC pipe to the outside of
the casing so that the channel would cover either 90, 60, 30 or 6
degrees of the 360 degree radial surface of the pipe (Figure A-1).
This was accomplished by attaching the PVC pipe to the surface of
-------
the casing, which had been sandblasted to remove mill varnish, using
fiberglass cloth and epoxy resin. Three layers of fiberglass were used
to ensure that the PVC pipe was securely sealed and attached to the
casing (Figures A-1 to A-4). This phase of the project was completed
on September 17, 1984.
The casing was then wrapped with heat tape and insulation to
prevent freezing of the channels while awaiting the availability of the
drilling rig.
The driller began moving the rig to the site on December 13,
1984. Rigging up continued on the 14th and 15th and drilling began at
2:30 p.m. on December 15, 1984. The procedure followed and the
date each step was accomplished are indicated below:
Activity
Date Completed
1) Prepare site 12/13/84
(Baulch Drilling Company)
2) Move in rig and rig up 12/15/84
3) Drill 15-inch hole to 40 feet, set 13 3/8-inch
conductor pipe (OK Cement, 35 sx) 12/15/84
4) Drill 8 3/4-inch hole to 1,530 feet. Collect
drill cuttings every 10 feet starting
at 100 feet 12/18/84
5) Condition hole for logging 12/18/84
6) Run Dual Induction Laterolog, Gamma 12/18/84
Ray, Compensated Neutron, Compensated
Density, and B.H.C. Sonic Log (Gearhart)
7) Ream 12 1/2-inch hole to 758 feet 12/20/84
8) Clean out hole to TD, condition for 12/20/84
setting casing
9) Set casing 12/20/84
10) Run 2 3/8-inch tubing and string into Baker 12/20/84
Duplex Cement shoe. Condition hole
for cementing.
11) Cement casing 12/20/84
(Halliburton, 700 sx. 50/50 Posmix)
12) Remove tubing, flush out hole 12/20/84
13) Weld steel plate between 9 5/8-inch and 12/27/84
13 3/8-inch casing. Install 9 5/8-inch
x 5 1/2-inch swage nipple on
9 5/8-inch casing. Screw locking
cap on swage.
14) Install rock pad and cement slab 9/10/85
24
-------
Figure A-1. Preparing fiberglass with epoxy resin.
-------
Figure A-2. Applying initial fiberglass layer.
26
-------
Figure A-3. Completed channel - prior to using wire brush to remove
excess.
-------
Figure A-4. Removing excess fiberglass with wire brush.
28
-------
The design of the Logging Well presumed that the channels
would remain in place during the process of setting and cementing
the casing. The driller took special precautions when moving the pipe
from the pipe racks to the rig, and during the pipe setting process to
ensure that this was the case. No problems were encountered during
the entire casing setting and well cementing operation, thus there was
a high degree of confidence that the channels were not damaged and
were in place as designed. Later review of logs run on the well
reinforced this confidence.
Log Interpretation
Figures A-5, A-6 and A-7 are portions of logs from the Logging
Well that compare single transmitter/single receiver with either 3-, 4-,
or 5- foot spacing; single transmitter/dual receiver with both 3-foot
and 5-foot spacing; and the second generation logs. Each
interpretation is based solely on information from the log.
Figure A-5a indicates a log section from a tool with single
transmitter/single receiver, 3-foot spacing. The fluid wave should
enter the receiver at about 567 microseconds; however, it is
indistinguishable on this log. Casing signals are present in the upper
part of the VDL, but no other interpretation can be made. The
amplitude curve indicates poor bonding in the upper part of the
section, which seems to agree with the VDL, and casing/cement
bonding throughout the remainder of the section.
Interpretation: Excellent casing/cement bonding throughout
the section with a possible channel or micro-
annulus in the upper part of the section. Coup-
ling to the formation cannot be determined.
Figure A-5b is the same section from a single transmitter/single
receiver tool, with 4-foot spacing. The fluid wave should enter the
receiver at about 756 microseconds, as indicated on the log. Casing
signals may be present in the upper part of the section, although the
signals are weak. The amplitude curve indicates a problem in the
upper part and casing/cement bonding for the remainder of the
section.
Interpretation: Excellent casing/cement bonding throughout
the section with a possible channel or micro-
annulus in the upper part of the section. No
coupling to the formation.
Figure A-5c is the same section from a tool with single
transmitter/single receiver, 5-foot spacing. The fluid wave should
enter the receiver at about 945 microseconds. In this case, the fluid,
wave is pushed far enough to the right to allow the formation signals
to be seen. The VDL indicates formation signals throughout the
section, and some casing signals in the upper part of the section. The
-------
Figure A-5a. Single receiver 3-foot spacing.
30
-------
Figure A-5b. Single receiver 4-foot spacing.
-------
Figure A-5c. Single receiver 5-foot spacing.
32
-------
amplitude curve indicates a problem in the upper part with
casing/cement bond throughout the remainder of the section.
Interpretation: Excellent casing/cement bonding throughout
the section with a possible channel or
microannulus in the upper part of the section.
Possible coupling of cement to the formation.
Figure A-5d is the same section from a tool with single
transmitter/dual receivers and a 3-foot/5-foot spacing. The VDL shows
formation signals throughout the section and casing signals in the
upper and lower parts of the section. The amplitude curve indicates a
problem in the upper part with casing/cement bonding indicated
throughout the remainder of the section.
Interpretation: Excellent casing/cement bonding throughout
the section with either channels or microannuli
in the upper and lower parts of the section.
Possible coupling to formation.
Figure A-5e is the same section from one of the second
generation logs. The white areas on the bond image portion of the log
indicate the presence of poor casing/cement bonding. The minimum
compressive strength curve also indicates some problems in three
areas that coincide with the white areas on the bond image portion of
the log.
Interpretation: Excellent casing/cement bonding with three
channels or microannuli in the upper, middle
and lower parts of the section. No information
is available on formation coupling.
A second comparison of the different logging tools run in the
Logging Well is shown in the Figure A-6 series. Figure A-6a is a log
section from a tool with single transmitter/single receiver, 3-foot
spacing. The fluid wave should enter the receiver at about 867
microseconds; however, it is not distinguishable on this log. There
may be formation signals in the presentation; however, they are not
distinctive, making it difficult to determine the character of the log.
The amplitude curve only registers in two places on the log, and the
transit time curve is not definitive.
Interpretation: Excellent casing/cement bonding throughout
the section.
Figure A-6b is the same section from a tool with single
transmitter/single receiver, 4-foot spacing. The fluid signal is very
strong, with the VDL indicating casing/cement bonding but no
formation coupling. The amplitude curve indicates casing/cement
bonding..
Interpretation: Excellent casing/cement bonding, no formation
coupling.
-------
Figure A-5d. Dual receiver 3-foot/5-foot spacing.
34
-------
Figure A-5e. Second generation log - Company A.
-------
Figure A-6a. Single receiver 3-foot spacing.
36
-------
r
03
.-a .
3
JL
1
Sv-a
0)
2 '
13
^
<
-, .. , ,, ,,n^',.
/',
Figure A-6b. Single receiver 4-foot spacing.
-------
Figure A-6c represents a log rrom a 1001 wun smyio
transmitter/single receiver, 5-foot spacing. It shows very clear
formation signals on the VDL, with one area near the center of the
section showing casing signal. The amplitude curve shows
casing/cement bonding, with a possible problem in the center of the
section.
Interpretation: Good casing/cement bonding with the
possibility of a channel or microannulus in the
middle of the section. Formation signals
indicate coupling to the formation.
Figure A-6d represents a log from a tool with single
transmitter/dual receivers with 3-foot/5-foot spacing. The VDL shows
formation signals throughout the section and definite casing signals
near the middle and in the lower part of the section. The amplitude
curve indicates cement/casing bonding, with a very slight response in
the middle of the section.
Interpretation: Excellent casing/cement bonding with the
possibility of channels or microannuli in the
middle and lower parts of the section.
Formation signals indicate coupling of the
formation to the casing.
Figure A-6e is a log of the same section of the hole from a
second generation tool. The bonding display indicates two channels
or microannuli, one in the middle and one in the lower section. The
average cement strength curve correlates with the bonding display.
Interpretation: Excellent casing/cement bonding with the
exception of two channels or microannuli in .the
middle and lower part of the section. No
information on formation coupling.
The third comparison of results from various logging tools
involves the occurrence of "fast formations," those formations that
exhibit a travel time that is equal to or faster than the travel time for
casing.
Figure A-7a is a section from a tool with a single transmitter/single
receiver with 3-foot spacing. The VDL indicates what appears to be
casing signals in the upper part of the section and formation signals in
the lower part of the section. The VDL signal in the lower part is
called formation signal because no chevrons are visible opposite the
casing collars. The amplitude curve indicates some casing signals in
three parts of the section.
Interpretation: Poor casing/cement bonding in the upper part.
The amplitude curve and VDL are contradictory
in the lower part of the section. The VDL
indicates casing/cement bonding and the
amplitude indicates poor bonding in one area.
38
-------
Figure A-6c. Single receiver 5-foot spacing.
-------
*ซfc^^
Figure A-6d. Dual receiver 3-foot/5-foot spacing.
40
-------
Figure A-6e. Second generation log - Company B.
-------
1*8
Figure A-7a. Single receiver 3-foot spacing.
42
-------
Figure /V-7b is a log of the same section with a tool with single
transmitter/single receiver, 4-foot spacing. The VDL indicates
casing/cement bonding in most of the section with one area of
formation signal in the lower part. The amplitude curve indicates
casing signal in the lower part.
Interpretation: Good casing/cement bonding in the upper part
of the section. No formation coupling in the
upper part. The amplitude curve and VDL are
contradictory in the lower part of the section.
Figure A-7c is a log of the same section from a tool with single
transmitter/single receiver, 5-foot spacing. The VDL indicates
formation signals throughout the section with possible casing signals
in the upper part. The amplitude curve indicates poor casing/cement
bonding throughout most of the section except for about 10 feet in
the upper part and the lower 20 feet.
Interpretation: Poor cement/casing bonding throughout the
section except for about 10 feet in the upper
part and the lowermost 20 feet. The amplitude
curve and VDL are contradictory in the lower
part. The amplitude curve reads over 70 mv,
which indicates free pipe.
Figure A-7d indicates a log of the same section from a tool with
single transmitter/dual receiver with 3-foot/5-foot spacing. The VDL
indicates formation signals throughout the section with casing signals
in the upper part. The amplitude curve indicates cement/casing
bonding with one possible problem in the upper part of the log.
Interpretation: Excellent cement/casing bonding throughout
most of the section. Possible channels or
microannuli in the upper part of the section.
Formation coupling throughout most of the log.
Figure A-7e is the same section from one of the second
generation logs. The bond image part of the log indicates three
channels or microannuli in the upper two-thirds of the log. The
minimum compressive strength curve supports that some problem
exists in these areas.
Interpretation: Excellent cement/casing bonding with three
channels or microannuli. No information on
formation coupling.
As can be seen from these examples, casing signals and
formation signals are very difficult to differentiate when a fast
formation is involved.
-------
Figure A-7b. Single receiver 4-foot spacing.
44
-------
Figure A-7c. Single receiver 5-foot spacing.
-------
I
1
Figure A-7d. Dual receiver 3-foot/5-foot spacing.
46
-------
Figure A-7e. Second generation log -.Company A.
-------
Logging Well No. 2
Logging Test Well Material Specifications
Casing Weight Grade
2 joints of 8 5/8-inch conductor pipe
39 joints of 5 1/2-inch 17.0#/ft J-55
5 joints fiberglass with
steel in collars
Eguipment
43 centralizers
Detailed Description of Well Construction
On May 26, 1987, the process of preparing the channels was
begun. Essentially the same process that was used on Logging Well
No. 1 was repeated, with the exception of the size of channel
material. PVC pipe, 1/4-inch polyethylene tubing, and fiberglass were
used to create the 30, 25, 20, 15 and 10 degree channels. The
excess resin and fiberglass were sandblasted to ensure that the
channels were the correct size, and the completed work was moved
to the drilling rig on June 25, 1987.
The driller began moving to the site on June 22, 1987, and began
drilling the rat hole on June 23, 1987. The procedure followed for this
well included:
Activity Date Completed
1) Prepare site (Baulch Drilling Company) 6/22/87
2) Move in rig and rig up 6/22/87
3) Drill 12 1/4-inch hole to 80 feet: 6/24/87
set conductor pipe
4) Drill 8 3/4-inch hole to 1,575 feet 6/26/87
5) Condition hole for logging 6/26/87
6) Run Dual Induction - SFL and Microlog 6/26/87
(Schlumberger also ran some
experimental open hole logs. They logged
the well from 1:00 a.m. until 9:15 p.m.)
7) Set casing 6/27/87
8) Cement casing (Halliburton, 1,075 sacks 6/27/87
50/50 Posmix)
9) Install rock pad and cement slab 7/08/87
No significant problems were encountered in completing the well,
and the first log was run on July 14, 1987.
48
-------
APPENDIX B
Leak Test Well Design and Testing Criteria and Test
Summaries
Leak Test Well
The purpose of the "Leak Test Well" is to provide a facility to
develop methods for testing the integrity of the tubing, casing and
packer and for testing the capability of various down-hole tools to
detect fluid movement behind the casing.
The design of the well generally corresponds to a typical salt
water disposal well used in the oil and gas industry. That is, it
includes the use of surface casing, long string, tubing and packer.
The deviation from the norm in this well includes two packers and a
sliding sleeve on the injection tubing and a 2 3/8-inch tubing attached
to the outside of the long string and running to the surface (Figure
10). ,
The depth to which surface casing was set was based on
Oklahoma Corporation Commission regulatory requirements to extend
below the occurrence of ground water having 10,000 mg/L or less
total dissolved solids. The selection of the depth for locating .the
profile nipples and the injection zone was based on porosity data from
Compensated Density and Compensated Neutron logs from the
Logging Well.
The 2 3/8-inch tubing, outside the 5 1/2-inch long string, extends
from a 1/4-inch hole in the long string at 1,070 feet below land
surface to the surface of the ground. The hole was drilled using a 1/4-
inch bit, and a 2 3/8-inch elbow was welded to the casing so that the
tubing could be attached. Profile nipples were placed in the tubing
opposite the 680- to 710-foot sand at 700 feet and the 905- to 935-
foot sand at 920 feet. These nipples will control the leakage of fluid
from the tubing, in that fluid can exit the tubing at either of the nipples
or be brought to the surface.
Casing and equipment for the Leak Test Well include:
571 feet of 13 3/8-inch casing
1,215 feet of 5 1/2-inch long string
1,070 feet of 2 3/8-inch tubing outside the long string
1,120 feet of 2 3/8-inch tubing inside the long string
Baker Model "L" Sliding Sleeve
Baker Model "AD-1" Tension Packer
-------
* DelKfcJr IVIUUBI O-U lezllUOIII I Cl loiui I I aorvci
Baker Model "R" Profile Nipple 1.78
Baker Model "RW" Profile Nipple 1.81
Baker Model "F" Profile Nipple 1.87
Baker 5 1/2-inch Float Shoe
Hinderliter 10FSF Wellhead for dual completions (5 1/2-inch and
2 3/8-inch) .
3 centralizers 5 1/2-inch '
The surface equipment for the Leak Test Well consists of two
100-barrel fiberglass tanks, a 10-horsepower electric powered
injection pump, high pressure injection flow lines and schedule 40
plastic return flow lines (Figure 11). The water supply is from the City
of Ada, Oklahoma. The control accessories are an air chamber, which
smooths out the pumping actions of the pump pistons; a pressure
control valve, which can be set to any predetermined pressure from
10 to 600 psi; a check valve which prevents back flow in the injection
line; a strainer to catch foreign material that may be pumped into the
line; a flow meter to record the number of barrels of liquid pumped; a
flow outlet pipe used to calibrate the flow meter; a control valve to
regulate the flow to the injection well; a thermometer to determine the
temperature of the injected fluids; and pressure gauges to indicate
the injection pressure (Figure B-1).
Flow into the well can be controlled so that the injected fluids are
directed into the 2 3/8-inch injection tubing, the tubing/casing anriulus
or to the 2 3/8-inch outside tubing. Returned flows can be controlled
from the 2 3/8-inch outside tubing and also from the annulus of the 5
1/2 inch casing.
Three monitoring wells were constructed around the Leak Test
Well to depths of 710, 935 and 1,130 feet. The casing and equipment
for the monitoring wells included:
A/o. 1
15 Baker Centralizers, 4 1/2-inch
Halliburton Super Seal Float Shoe
671 feet of 4 1/2-inch steel casing, 10.6 #/ft
2-inch Johnson stainless steel well screen, .010-inch slot with bottom
plate
20 feet of 2-inch line pipe on top of screen
2- to 4-inch shale cup
Johnson "K" type packer, 4 1/2-inch
80 feet of 9 5/8-inch conductor pipe, 36 .#/ft
4 1/2- to 2-inch steel swage/cap
A/o. 2
21 Baker Centralizers, 4 1/2-inch
Halliburton Super Seal Float Shoe
905 feet of 4 1/2-inch steel casing, 10.6 #/ft
50
-------
Flow Return Line
*
Flow Thermometer
Meter
From Water
Supply Tank
o
To
Injection
Well
Pressure
Gauges
Figure B-1. Injection pump and control accessories.
2-inch Johnson stainless steel well screen, .011-inch slot with bottom
plate
21 feet of 2-inch line pipe on top of screen
Johnson "K" type packer, 4 1/2-inch
80 feet of 9-5/8" conductor pipe, 37 #/ft
4 1/2- to 2-inch steel swage/cap
A/o. 3
27 Baker Centralizers, 4 1/2-inch
Halliburton Super Seal Float Shoe
1,130 feet of 4 1/2-inch steel casing, 10.6 #/ft
2-inch Johnson stainless steel well screen, .010-inch slot with bottom
plate
20 feet of 2-inch line pipe on top of screen
80 feet of 9 5/8-inch conductor pipe, 36 #/ft
4 1/2- to 2-inch steel swage/cap
Wells No. 1 and 2 were drilled using air rotary to prevent
contamination of the zones to be monitored by drilling fluids. Well No.
3 could not be completed using air so it was drilled with mud rotary.
The procedure followed for drilling the wells and the dates
involved are indicated in Table B-1.
-------
Table B-1. Procedure for Drilling the Monitoring Wells
Activity
Date Completed
Monitoring Well No. 1
1. Rig up, air drill 9-inch diameter hole to 82
feet, ream 12 1/2-inch hole to 82 feet, set
and cement 9 5/8-inch conductor pipe
2. Drill 8-inch hole to 680 feet, run 4 1/2-inch
steel casing with centralizer on each collar,
cement with latex cement (Halliburton)
3. Drill remaining 30 feet with 3 7/8-inch bit
4. Set 2-inch by 30-foot Johnson stainless steel
screen
Monitoring Well No. 2
i. Rig up, air drill 12 i/4-inch diameter hole
2. Cement conductor pipe
3. Drill 8-inch hole to 905 feet, run 70 sacks of
gel to stabilize hole
4. Run 4 1/2-inch steel casing with centralizer
on each collar
5. Cement with spherelite cement (Halliburton)
6. Drill to 935 feet with 3 7/8-inch tricone bit, set
2-inch by 30-foot stainless steel screen with
bottom plate
Monitoring Well No. 3
1. Rig up. drill 12 1/4-inch diameter hole to 80
feet, set and cement 9 5/3-inch conductor
pipe |
Dig mud handling pit
Drill 7 7/8-inch hole to 1,130 feet, run 4 1/2-
inch steel casing with float shoe
Cement with foam cement (Halliburton)
2.
3.
4.
5.
6.
Perforate zone from 1,120 to 1,130 feet, 21
shots
Swab casing, set 2-inch by 20-foot stainless
steel screen
11/17/87
11/24/87
11/28/87
11/28/87
11/30/87
12/01/87
12/03/87
12/04/87
12/08/87
12/21/87
12/01/87
12/08/87
1/09/88
1/10/88
1/14/88
1/14/88
Test Summaries
A number of specific tests are planned for the Leak Test Well. As
the tests are completed, brief summaries are prepared and forwarded
to the Underground Injection Control Program Offices in EPA
Headquarters and the regions. Summaries of those tests completed
to date are included as follows:
52
-------
Test Mo. 1: Acoustic Cement Bond Tool Test for Flow Behind
Casing
Introduction
In November 1986, personnel from Regions IV and V witnessed a
demonstration of the use of an acoustic cement bond tool for
detecting fluid movement behind casing. The demonstration was
conducted by Dresser Atlas personnel at their field office in Olney,
Illinois, and involved pumping water through the annular space
between two concentric pieces of pipe while holding the bond tool
stationary in the inner pipe.
EPA personnel suggested that the tool(s) be tested in the Leak
Test Well at the Robert S. Kerr Environmental Research Laboratory
(RSKERL) to get a better definition of the sensitivity of the tool and
the conditions under which it will or will not work.
C.D. "Mac" McGregor, a log analyst for Dresser Atlas, contacted
RSKERL personnel, and plans were made to run the tests on January
23, 24, and 25, 1987.
Test Well Conditions
The purpose of the test was to determine if flow of water at
various rates could be detected behind pipe using the data presented
by the fluid wave from a cement bond tool. The test was developed in
two phases: Phase I looked at flow immediately behind casing under
free-pipe conditions, and Phase II looked at flow in tubing behind
casing under conditions which would possibly simulate flow in a
channel in cement.
Figure B-2 indicates the configuration of the Leak Test Well for
the initial test. In this configuration, water was pumped down the
tubing/casing annulus into the injection zone. This represents flow in
the free-pipe condition, i.e., no cement behind the pipe (2 3/8-inch
tubing in this case).
Figure B-3 indicates the well configuration for the second test,
which was designed to simulate flow in a channel in cement. The
section of the well between 1,070 and 950 feet has cement behind
the 5 1/2-inch casing and thus around the 2 3/8-inch tubing. Thus the
tubing in that area represents, to some degree, a channel in the
cement. In this phase, water was pumped down the 2 3/8-inch tubing,
into the 5 1/2-inch casing and out the perforations.
Test Phase I
The test was conducted with a 1 11/16-inch OD Acoustic Cement
Bond Tool with a single transmitter/single receiver with 4-foot spacing.
The tool was placed in the injection tubing at 57 feet and the
oscilloscope was viewed in the no-flow and flow conditions.
-------
.680ft
Flow=
710ft
905ft
935ft
Injection Zones
CBL Liquid Flow Test - Phase I
1. Unseat packers #1 & #5.
2 ' Plug profile nipple #4
1057 ft Depth of 3. Fill tubing (#2) with water
Upper Packer 4. Set CBL tort in 2 3/S-.n
HH tubjnฃ at yajiable depths
5 Pump water down 5 1/2-in
casing at 3 different rates
Cement
1070ft
1. Baker Model "C-1"
Tandem Tension Packer
2. 2 3/8-in Tubing
3. Baker Model "L" Sliding
Sleeve
4. Baker Model "R"
Profile Nipple
5. Baker Model "Ad-1"
Tension Packer
1084 ft Depth of * B^Mbde^V
Lower Packer p^f fla Nipple
8. Baker Mpdel "F'1'
1100ft .Profile Nipple
9. 5 1/2-in Long String
.....,...................-,.v 1120ft
Leak Test Well
Figure B-2. CBL liquid flow test - Phase I.
54
-------
Cement Bond Log Tool
Flow=
. 905 ft
935ft
Injection Zones
CBL Liquid Flow Test - Phase II
1. Pull tubing and packers
2. Set plug in 5 1/2-in casing at
101.0 ft
3. Pull tubing
4. Fill 5 1/2-in casing with water
5. Set CBL tool in 5 1/2-in casing
at variable depths
6. Pump water down 2 3/8-in leak
tube at 3 different rates
Cement
1. 2 3/8-in Tubing
2. Baker Model "R" Profile Nipple
3. Baker Model "F" Profile Nipple
4. 5 1/2-in Long String
1120ft
Leak Test Well
Figure B-3. CBL liquid flow test - Phase II.
-------
1 I OOl "" I I 1C1OO I
Oscilloscope
Time Flow Rate Response
1:50 p.m. No flow None
2:00 p.m. 8 gpm Yes
2:03 p.m. 4 gpm Yes
2:17 p.m. 0.78 gpm Yes
2:20 p.m. 0.78 gpm + air Yes
2:24 p.m. 6 gpm + air Yes
2:30 p.m. Stopped injection Yes
The test was repeated with the tool at various depths in the well,
with the same results; that is, the oscilloscope indicated no distortion
of the fluid wave in the no-flow condition and distortion at all three
flow rates. The distortion was greater when air was added to simulate
gas movement behind the pipe.
Test - Phase II
The second test was conducted with the tubing and packers
removed and a bridge plug set, as indicated by Figure B-3. The
Acoustic Cement Bond Tool used for this test was a 3 5/8-inch OD
tool with single transmitter/single receiver with 5-foot spacing.
ป Test - Phase II
Oscilloscope
Tool Depth (feet) Flow Rate (qpm) Response
600 8 None
700 8 None
800 8 None
900 8 None
1,000 8 None
The tool was initially set immediately above the bridge plug and
the oscilloscope viewed in the no-flow condition. Water was then
pumped down the outside tubing at a rate of 8 gpm. No distortion of
the fluid wave was evident on the oscilloscope. The tool was then
moved up the hole in 100-foot increments. No distortion of the fluid
wave was evident at any depth, thus no flow was detected in the
tubing.
56
-------
Conclusions
The Phase I test results indicate that the fluid wave of the
Acoustic Cement Bond Tool responded to fluid movement behind the
tubing in a free-pipe condition, that is, with no cement behind the
pipe. A response was evident with flow as low as 0.78 gpm.
The tool used in the Phase II test did not pick up flow in the
2 3/8-inch tubing either within or above the cemented section of the
well. Thus, flow in the manmade channel behind the 5 1/2-inch casing
could not be detected under the test conditions.
One explanation for the responses observed under the test
conditions previously outlined is that under free-pipe conditions, the
paths for movement of the sound wave are through the casing arid
fluid. Thus, under static conditions, where the tool is not moving and
there is no movement of fluid in or behind the pipe, the fluid wave, as
presented on the oscilloscope, is also static. On the other hand, flow
of fluid behind the pipe while the tool is stationary affects the sound
wave as it moves through the fluid, causing a distortion of the wave.
This distortion shows up as rapid changes in amplitude in the display
of the fluid wave on the oscilloscope and indicates movement of the
fluid. Thus, under free-pipe conditions, the fluid wave has the capacity
to reflect fluid movement behind pipe.
The presence of cement behind pipe presents a much more
difficult set of conditions for identifying fluid movement with the
cement bond tool. The paths for the sound wave under these
conditions are movement along the casing and cement (small signal
because of the attenuation effect of the cement behind the casing),
movement thro'ugh the formation and movement through the fluid.
The heterogeneity of the formation, the size of the channel, and type
and amount of fluid movement will all affect the ability of the tool to
identify fluid flow in channels in cement. Thus, the capability of the
Acoustic Cement Bond Tool to identify fluid flow in channels is
unproven, though certainly not impossible.
Recommendations
Field data should be accumulated to determine the capability of
this type of tool for detecting flow behind casing in varying well
conditions, i.e., free pipe and channels in cement.
When running other tools, such as temperature or noise surveys
for detecting flow behind pipe, service companies should run the
bond tool for comparison purposes to determine if flow in channels
can be detected.
-------
Behind Casing
]..
Introduction
On January 23 and 24, 1987, personnel from the Robert S. Kerr
Environmental Research Laboratory (RSKERL) and Dresser Atlas
conducted a series of tests for determining flow behind pipe using
two neutron activation tools.
The purpose of the tests was to determine if flow of water at
various rates could be detected behind pipe using the data presented
by & pulsed neutron lifetime logging system (PDK-100) and a Cyclic
Activation Tool.
Tools Tested
Two tools were tested during the 2-day period:
A 1 11/16-inch diameter PDK-100 Tool
* A 3 5/8-inch diameter Cyclic Activation Tool
The operation of both tools is based on a nuclear activation
technique in which flowing water is irradiated with neutrons emitted by
a logging sonde. These neutrons interact with oxygen nuclei in the
water to produce nitrogen-16 (16N), which decays with a half-life of
7.13 seconds, emitting gamma radiation. The flow is then computed
from the energy and intensity response of two gamma ray detectors
mounted in the logging sonde.
Test Well Conditions
The tests were developed in four phases, the first three using the
PDK-100 Tool and the last using the Cyclic Activation Tool.
i i .
Figure B-4 indicates the configuration of the Leak Test Well for
the initial test. In this configuration, water was pumped down the
tubing/casing annulus into the injection zone with the 1 11/16-inch
diameter PDK-100 Tool'held stationary in the 2 3/8-inch injection
tubing. This condition represented flow in the free-pipe condition, i.e.,
with no cement behind the pipe (2 3/8-inch tubing in this case). A
valve at the surface on the outside 2 3/8-inch tubing was closed so
that circulation was not possible up that tubing.
Figure B-5 indicates the well configuration for the second test,
which was designed to simulate upward flow in a channel in cement.
Water, pumped down the tubing/casing annulus, moves through a
1/4-inch hole in the 5 1/2-inqh casing at 1,070 feet and up the 2 3/8-
inch outside tubing. The section of the well between 1,070 and 950
feet has cement behind the 5 1/2-inch casing and thus around the
2 3/8-inch tubing. The tubing in that area represents, to some degree,
a channel in the cement.
58
-------
1% vX
1
5-
-
_M
looo
V
s~
*.li
* II
\_
1
_^
1
~\
1,
"'
<
V
1
t
-7>
^
^ V<
(
1
1
,
>
4-
i
/,'jlM
_ Lo
" .if".!
- *J
(1
i :
.\x- 8:::::::::v:vv.'v::::::.'v:: Flow- X
:::::;x:S:::::;::::x::: ::fe: :::x x: :: 7 1 Q ft
905 ft
Injection Zones
nryK ft '
6
1057 ft Depth of
Upper Packer
NAT Liquid Flow Test - Phase I
1 . Unseat packers #1 & #5
2. Set NAT tool in 2 3/8-in tubing
at variable depths
3. Pump water down 5 1/2-in
casing at 3 different rates
Cement
1 070 ft
1 . Baker Model "C-1 " Tandem
Tension Packer
2. 2 3/8-in Tubing
14 ft Depth of 3. Baker Model "L" Sliding Sleeve
wer Packer 4. Baker Model "R" Profile Nipple
5. Baker Model "Ad-1"Tension
. Packer
6. 2 3/8-in Tubing
7. Baker Model "R" Profile Nipple
8. Baker Model "F" Profile Nipple
9. 5 1/2-in Long String
1 1 00 ft
..v.:.:.v:..y-:-:-: ::... ' l<;u u
Leak Test Well
Figure B-4. Neutron activation tool liquid flow test - Phase I.
-------
Ni
.
i"
4"
5-
1,
'&**&%&.
i
m
*>**.&?.
(ffjs'j
:i.{
8
\1
-v.;..... -\
*"v.::'*::'. :' '*
1
I
(>
4 "
4
V
ป
ooo
^
/
k 1
* 1
\
*
1
_>
4
\
1,
't
'8
# ฃฃ vl^^iii-^1-
V
>
71
<
II
1
i
1
1
j*l
yi.
!'!
S'
IP
X
'> -V
u>
;<-9ii
VJ
j5t,i\
^ir
*' -C
f-
r
^^_
L
_ซ.*ป.
iS
f
ft'MK
|ซ*f,
Hj\5>;:
'li^i
ปซ
;>.}>>'
'>.,&# K*PI
.R.;ซ<;i.iA\ft.
l|
\2i"
.&.
^ป
" 's
?I*
r "
V
S^l
!ปl!!iMi
?c3rrf
Vs
^
ฃ
680 ft
^ o fiSSS'-:''-^:!'!^:'-? Flow = 1
|vj: o:::::i:::v:i:>.::v;v>.:::
Wx^xWivXiAvX:::^::! 7m ft
on=; ft ..;:..:.:.:.:.:::.:;..
*-J 7 Injection Zones
CMR ft
~6
1057ft Depth of
Upper Packer
NAT Liquid Flow Test - Phase II
1. Unseat packer #1
2. Plug profile nipple #4
3. Set NAT tool in 2 3/8-in tubing
at variable depths
4. Pump water down 5 1/2-in
casing and up 2 3/8-in tubing
Cement
1. Baker Model "C-1" Tandem
in?nft Tension Packer
2. 2 3/8-in Tubing
3. Baker Model "I." Sliding Sleeve
4*. Baker Model "R" Profile Nipple
5. Baker Model "Ad- 1 "Tension
Packer
6. 2 3/8-in Tubing
7. Baker Model "R" Profile Nipple
8. Baker Model "F" Profile Nipple
9. 5 1/2-iii Long String
1084 ft Depth of
Lower Packer
1100 ft
11 on ft
1130ft
Leak Test Well
Figure B-5. Neutron activation tool liquid flow test - Phase II.
60
-------
Figure 8-6 indicates the well configuration for the third test, which
was designed to simulate downward flow in a channel in cement.
Water, pumped down the 2 3/8-inch outside tubing, moves through
the 1/4-inch hole in the 5 1/2-inch casing at 1,070 feet and up the 5
1/2-inch casing to the surface.
Figure B-7 indicates the well configuration for the final test, which
was designed to simulate downward flow in a channel in cement
using the larger Cyclic Activation Tool. Water, pumped down the
2 3/8-inch outside tubing, flows into the 5 1/2-inch casing through the
1/4-inch hole and out through perforations into the injection interval
from 1,120 to 1,130 feet.
Test - Phase I
This test was conducted with the PDK-100 Tool with the two
detectors located below the neutron generator so that downward flow
could be detected. With the tool located at 300 feet inside the
2 3/8-inch injection tubing, data was obtained under conditions of no
flow and flow of 8, 4 and 1 gallon per minute (gpm). Two replications
of these flow rates were conducted and flow was detected by the tool
in all instances.
Test - Phase II
This test was conducted with the PDK-100 Tool with the two
detectors located above the neutron generator to determine if flow Lip
the outside 2 3/8-inch could be detected. With the tool located at 600
feet, data was obtained under no flow, and 8 gpm flow conditions.
Flow up the outside 2 3/8-inch tubing could not be detected.
Test - Phase III
This test was conducted with the PDK-100 Tool at 600 feet with
the generator-detector configuration identical to the Phase II test.
Water was pumped down the 2 3/8-inch outside tubing and up the
5- 1/2" casing at three different rates (8,4 and 1 gpm). Upward flow
was detected in the 5 1/2-inch casing at all three flow rates.
The tool configuration was then changed with the.detectors below
the generator to determine if downward flow in the outside tubing
could be detected. Flow down the outside 2 3/8-inch tubing could not
be detected.
Test - Phase IV
This test was conducted using a 3 5/8-inch diameter Cyclic
Activation Tool. The detectors were located below the generator for
detecting flow in the 2 3/8-inch tubing as water moved down the
tubing, through the 1/4-inch hole into the 5 1/2-inch casing and out
the perforations into the injection interval. Flow rates for this test were
7.8, 6.1 and 0.79 gpm. All three flow rates were detected by the tool,
-------
9I
1 -
3.
4.
5-
&
jj^-a-j.
i
?ฃJ*2$ฃ
aiifejy
~^E$
IS
l|i
3
1
t
t
\
t
t
X
ooo
^_
t
1
* 1
r
ff
I
,ป "-1
1
I
J
t
\
I
t
4
"R
'SKSSirtMcSii'in;
k >
^ <
ซ
H
-^
~
ts
T
t
1
"i
I
i
-
4
1
J,
^
k
P*
-*-
~
'+
n
': . / :':-.-;'. -'::::: :::::
..v'.v HVV.-!.. :'.'-: 710 ft
905 ft
: -..'. "' ;' ''' ' en* ft tn-
3W= 1
ection Zones
1057 ft Depth of
Upper Packer
NAT Liquid Flow Test - Phase III
1. Unseat packer #1
2. Plug profile nipple #4
3. Set NAT tool in 2 3/8-in tubing
at variable depths
4. Pump water down 2 3/8-in
tubing and up 51/2-in casing
Cement
1. Baker Model "C-1" Tandem
1070ft Tension Packer
2. 2 3/8-in Tubing
3. Baker Model "I." Sliding Sleeve
4. Baker Model "R" Profile Nipple
5. Baker Model "Ad-1 "Tension
Packer
6. 2 3/8-in Tubing
7. Baker Model "R" Profile Nipple
8. Baker Model "F" Profile Nipple
9. 5 1/2-in Long String
1 084 ft Depth of
Lower Packer
1 1 00 ft
11TI ft
Leak Test Well
Figure B-6. Neutron activation tool liquid flow test - Phase III.
62
-------
5 1/2-in I
Cement
1070ft
........,,,..:.:.;.:.. :.:.;.;.:.v!; 680 ft
I 3W$i$i%iiff ; Flow= j
mmmmyfm 710ft
935 ft
Injection Zones
NAT Liquid Flow Test - Phase III
1. Pull tubing and packers
2. Set plug ins 1/2-in casing at 1010 ft
3. Pull tubing
4. Set NAT tool in 5 1/2-in casing at
variable depths
5. Pump water down 2 3/8-in leak
tube at 3 different rates
1 . 2 3/8-in Tubing
2. Baker Model "R" Profile Nipple
3. Baker Model "F" Profile Nipple
4. 5 1/2-in Long String
1120ft
1130ft
Leak Test Well
Figure B-7. Neutron activation tool liquid flow test - Phase IV.
-------
arid IIUW VtJlUUlUWS wwiw
tool.
num mo vjaia oun-av^i.^0 w
Conclusions
The PDK-100 Tool was able to detect all three flow rates when
flow was up or down the 5 1/2-inch casing. The tool did not detect
any flow up or down the outside 2 3/8-inch tubing.
The Cyclic Activation Tool was able to detect all three flow rates
in the outside 2 3/8-inch tubing. In addition, the computer associated
with the tool has the capability to compute a velocity of flow for each
flow rate.
Recommendations
Additional work should be done to increase the sensitivity of the
PDK-100 Tool. It should be noted here that since the tests were
conducted, Dresser Atlas personnel have made some modifications to
the PDK-100 Tool and have been able to detect flow in outside tubing
in a well constructed very similarly to the Leak Test Well. The
adjusted tool will be retested'at the RSKERL Test Facility as soon as
it can be arranged. In the meantime, Dresser Atlas personnel will run
the tool in several wells owned by Mobil, and will make those results
available to RSKERL personnel.
The Cyclic Activation Tool should be tested under "real well"
conditions to verify the results seen during the tests on the Leak Test
Well.
The capability of this equipment to locate flow behind pipe could
be a significant breakthrough for mechanical integrity testing.
Especially the PDK-100 Tool, which can be run in tubing filled with
water or with only air present. Thus, no workover costs would be
involved in testing a well, i.e., setting plugs, pulling tubing, etc.
i ' i
64
-------
Test No. 31. Testing fop a Hole in the Long String
Introduction
On January 23, 24 and 25, 1987, test results from testing tools
for detecting flow behind casing indicated a possible hole in the 5 1/2-
inch long string of the research well. A series of tests was conducted
on the well on January 25 and 27, and February 2, 3, 10, 11 and 12
to determine whether there was a hole in the pipe.
Test Well Conditions
Figure B-8 indicates the well configuration during most of the
tests to be discussed. Any changes in the well will be noted as the
various tests are discussed.
During testing of an Acoustic Cement Bond Tool (ACBT), the
5 1/2-inch casing was full of fluid above a bridge plug and water was
being pumped down the outside 2 3/8-inch tubing at about 8 gpm.
Pumping had been in progress only about 5 minutes when water
began flowing out of the 5 1/2-inch casing at about 2 1/2 gpm.
The immediate thought was the bridge plug was leaking;
however, the Baker Packer representative was confident that the
bridge plug could not leak. In checking the setting depth for the plug it
appeared possible that it was located opposite a casing collar. The
plug was reset to ensure that it was properly set between collars.
Acoustic Cement Bond Tool
A plan was developed to systematically check the well to
determine where the leak was in the system. The first approach was
to use the ACBT to determine if flow in the 5 1/2-inch casing was
occurring. The tool was set immediately above the bridge plug, which
was set at 1,010 feet, and readings were taken to determine if flow
would be reflected by the fluid wave. The tool was then moved up the
well at 100 foot increments and readings taken, with the following
results:
Flow Indicated
1,000 No
900 No
800 No
700 No
600 No
500 No
400 No
300 Yes
250 Yes
200 Yes
-------
Cement
1070ft
680ft
710ft
-905ft
935 ft
Injection Zones
1. 2 3/8-m Tubing
2. Baker Model "R" Profile Nipple
3. Baker Model "F" Profile Nipple
4. 5 1/2-m Long String
1120ft
1130ft
Leak Test Well
Figure B-8. Testing for a hole in the long string.
66
-------
7r\e tests were rerun with the same results. Thus, the ACBT data
indicated that flow might be coming into the casing at around 300
feet.
Down-Hole TV
On January 27, 1987 Layne-Western Company brought their
down-shot camera to survey the well, to locate any hole that might be
present. The regular lens would not give a good image, so a lens that
must be used in air rather than water was tried. First the well was
swabbed so that the water level was about 870 feet below land
surface.
The camera, which had never been run in 5 1/2-inch casing,
provided an excellent picture of the casing as it was lowered down
the casing. An anomaly was seen at about 240 feet that could
possibly be a damaged area of the casing.
Pressure Test - Gas
The next test was to pressure the well with nitrogen and shut it in
to determine if any loss of pressure would occur over time. The well,
with the bridge plug still intact at 1,010 feet, was pressured to 185 psi
with nitrogen and shut in. There was no loss of pressure evident after
1 hour.
Pressure Test - Water
The gas pressure was relieved and the injection pump hooked up
to fill the 5 1/2-inch casing with water. After filling, 150 psi pressure
was added to the weight of the water column and the well was shut in
for 1 hour. No pressure drop was noted.
Pressure was relieved and water was pumped down the outside
2 3/8-inch tubing at about 8 gpm. After pumping for only 5 minutes,
water began flowing out of the 5 1/2-inch casing at a rate of about
2 1/2 gpm.
Pressure Test - Packer
Next, Baker Packer, using a full-bore packer on tubing performed
a series of pressure tests with the packer set at various depths in the
casing, as follows:
Packer
Depth Pressure Drop in Pressure
feet psi after 5 Minutes
11r 92 None
295 115 None
595 100 None
-------
The final test was !a modified "Ada Pressure Test." After
removing the bridge plug, nitrogen was used to move the state water
Ivel toward the hole in the 5 1/2-inch casing. A pressure of 380 psi
was placed on the fluid in the casing and held overnight with no loss
of pressure.
Conclusions
The series of pressure tests performed on the well clearly
indicated no hole is present in the 5 1/2-inch casing. The differential
presste bridge'plug apparently did not have sufficient pressure
differential to set securely, thus allowing the plug to leak when
injection was taking place down the outside tubing.
The ACBT and down-hole TV were inconclusive in that a doubt
still existed after reviewing data from the tests.
68
-------
Test No. 4: Ada Pressure Test
Introduction
Early in the mechanical integrity test program, it was discovered
that some wells could not be tested using the standard pressure test.
For example, a number of wells in Osage County, Oklahoma, could
not use the standard pressure test because of perforations in the long
string above the packer. EPA regional personnel suggested that a
procedure similar to the air line method for determining the water
level in a producing water well should be explored as an alternative
for testing those wells whose "special" construction would not permit
the use of standard pressure tests.
Development of Test
A possible alternative to the standard pressure test is to use
compressed air or nitrogen to depress the static water level below the
point to be tested and hold the pressure for a specific period of time.
If the pressure holds, the tubular goods above the fluid level have no
leaks.
Table B-2 indicates theoretically what would take place in the well
as pressure is added to the tubing to depress the water. With a static
fluid level of 360 feet below the land surface, the hydrostatic head at
the perforations would be 760 feet. The tubing gauge pressure and
the pressure at the fluid level would both be zero. This hydrostatic
head would exert 307 psi at a depth of 1,070 feet, and 329 psi at a
depth of 1,120 feet. As pressure is added from cylinders of
compressed gas, the gauge pressure increases and depresses the
fluid level, thus reducing the hydrostatic head by a corresponding
amount. The pressure at the gauge and pressure at the fluid level
remain equal throughout the procedure. Thus, the pressure at the
point of consideration remains constant in that the hydrostatic
pressure is replaced by gas pressure during the test.
Table B-2 Effects of Adding Pressure to Depress Water
Hydrostatic
Tubing
Gauge
Reading
(psi)
0
100
200
300
307
329
Depth to
Fluid Level
(feet)
360
591
822
1,053
1,070
1,120
Head
Above
Perforations
(feet)
760
529
298
67
50
0
psi @ Fluid
Level
0
100
200
300
307
329
psi ฉ Hole
(1,070 feet)
307
307
307
307
307
329
psi @ Pert.
(1,1 20 feet)
329
329
329
329
329
329
-------
impiemenuriy ii
On December 4, 1985, two tests were implemented to determine
if pressure testing with compressed air was a viable option for testing
special wells. The configuration of the Leak Test Well was altered
slightly for each of the tests to represent real world situations as
closely as possible (Figure B-9).
i . . i
Injection Zones
1. Surface Casing (571 ft)
2. 2 3/8-in Tubing
3. Baker Model "L" Sliding
Sleeve
4. Baker Model "R" Profile
Nipple
5. Baker Model "Ad-1"
Tension Packer
6. 2 3/8-in Tubing
7. Baker Model "R" Profile
Nipple
8. Baker Model "F" Profile
Nipple
9. 5 1/2-in Long String
10. Baker Model "C-1" Tandem
Tension Packer
1084 ft Depth of
Lower Packer
1100ft
1120ft
1130ft
Figure B-9. Leak Test Well.
The first test was conducted with the sliding sleeve closed, to
represent no leak in the system, and the second test was conducted
with the sliding sleeve open, to represent a leak in the tubing at a
depth of 1,070 feet. The fluid level in the injection tubing as measured
with an Echo Meter was 360 feet below the land surface. That
70
-------
a hydrostatic head of 710 feet at the hole at 1,070 feet arid
760 feet of head at the perforations at 1,120 feet. Assuming 2.31 feet
per psi, it would require 307 psi to depress the water level to a depth
of 1,070 feet and 329 psi to depress the water level to a depth of
1,120 feet.
Test 1
In the first test, with the sliding sleeve closed, the pressure
should have reached 329 psig before the pressure could not be
increased. This would indicate that the fluid level was at the top of the
perforations at 1,120 feet, and with the gas source closed the
pressure gauge should continue to read 329 psig.
A pressure of 380 psig was added without reaching the point
anticipated. When the cylinder was closed the pressure dropped
below 329 psig, and repeated attempts to increase the pressure gave
the same results. The source of compressed gas was exhausted and
the test was stopped without reaching the point of stabilization at 329
psi.
This behavior seems to indicate that the permeability of the
injection zone was low and the formation would not accept the water
fast enough to depress the water level as quickly as anticipated.
Thus, during the time of the test, the fluid level in the tubing was still
being depressed but had not had time to reach the desired level.
Tesf 2
In the second test, with the sliding sleeve open, 307 psig should
have depressed the water level to a depth of 1,070 feet, and it should
have been impossible to add additional pressure because pressure
was lost through the hole in the long string at that depth. However, an
inability to increase the pressure was reached at a pressure of 300
psig. The pressure remained at 300 psig after shutting off the
compressed gas source.
On May 1, 1986, the well was acidized and injectivity tests
indicated a permeability of 125 millidarcies (md). The second series of
tests was run with the results comparable to those predicted in the
table. Nitrogen was used to eliminate any explosion hazard that might
have been presented by using compressed air.
The test was rerun on February 11, 1987, using nitrogen, and the
results confirmed the response seen in May 1986.
-------
Conclusions
The test results indicate that an annulus pressure test can be run
on certain wells that have special conditions that prevent the running
of standard pressure tests. The following conditions are
recommended to assure that the test has validity for a specific well:
1. The fluid level in the zone being tested must have reached static
conditions before the test is run.
!
2. The specific gravity of the injected water must be known to
calculate the pressure required to depress the fluid level to a
specific depth.
72
-------
Test No. 5: Nuclear Activation Technique for Detecting Flow
Behind Casing
Introduction
On April 8, 1987, personnel from the Robert S. Kerr
Environmental Research Laboratory (RSKERL) and Dresser Atlas
conducted a series of tests to determine flow behind pipe using the
PDK-100 Flow Tool.
The purpose of the tests was to determine if flow of water at
various rates could be detected behind pipe from data presented by a
pulsed neutron lifetime logging system (PDK-100). The 1 11/16-inch
diameter tool had been tested on January 23 and 24, 1987, and could
detect flow immediately behind the injection tubing but could not
detect flow in the outside 2 3/8-inch tubing. The tool had been
modified for the new series of tests.
Test Well Configuration
Figure B-10 indicates the configuration of the Leak Test Well for
the test. Both packers were set, the sliding sleeve was open and
injection was maintained down the outside tubing at varying injection
rates.
Tool Testing
For each flow rate the PDK-100 was held stationary at a depth of
300' in the injection tubing. After taking two background checks, flow
was initiated down the outside tubing at a rate Of 8 gallons per minute
(gpm), 6, 4, 2, and 0.105 gpm. The results of the tests for detecting
flow are as follows:
Flow Rate (gpm) Flow Detected
8 Yes
6 Yes
4 Yes
2 Yes
0.105 No
Readings were taken three times at each flow rate.
Conclusion
The PDK-100 was able to detect four of the five flow rates with
no problem. Movement was detected for the 0.105 gpm flow but it
was probably the column of water in the tubing moving toward static
conditions, since at this extremely low flow the fluid level in the tubing
could not be maintained.
The capacity of this tool to locate flow behind casing looks very
promising. The next phase should be field testing under "real well"
conditions.
-------
680ft
710ft
905ft
935 ft
Flow=
lr|Ject'on Zones
NAT Liquid Flow Test
Pump water down 2 3/8-in tubing (6),
through sliding sleeve (3),
through injection tubing (2),
into injection zone
Baker Model "C-1 " Tandem
'Tension Packer
2 3/8-in Tubing
Baker Model "I." Sliding Sleeve
Baker Model "R" Profile Nipple
"Ad"1 "Tension
i Tubing
Baker Model "R" Profile Nipple
Baker Model "F" Profile Nipple
5 1/2-in Long String
- 1120 ft
1130ft
Leak Test Well
Figure B-10. Neutron activation tool liquid flow test.
74
-------
Test No. 6: Radial Differential Temperature Survey
Introduction
On April 27, 1987, Gearhart Industries, Inc., was contracted to
run a Radial Temperature Survey on the Leak Test Well to determine
the capability of this tool to detect vertical flow behind casing.
The Radial Differential Temperature (RDT) tool employs two
highly sensitive temperature probes which extend from the centralized
1 11/16-inch O.D. housing to contact the casing. As these probes are
rotated, they measure any difference in temperature at two points on
the casing 180 degrees apart.
Well Configuration
Figure B-11 indicates the Leak Test Well configuration for the
RDT survey. The fluid level in both the 5 1/2-inch casing., and the
2 3/8-inch tubing was 190 feet below the surface of the ground.
Radial Differential Temperature Survey
The procedure for running the RDT in the Leak Test Well was to
run a conventional temperature profile, take RDT readings at five
depths under no-flow conditions, then repeat the RDT readings at the
five depths after beginning injection down the outside 2 3/8-inch
tubing.
The temperature profile indicated that the fluid level in the 5 1/2-
inch casing was at 190 feet below land surface. RDT scans at 1,050,
1,025, 1,000, 975, and 850 feet, under no-flow conditions, showed
the same temperature for both probes as they were rotated at each
depth. (See Figure B-12.)
injection down the outside 2 3/8-inch tubing was started at 10:57
a.m., and the RDT surveys at 11:05 a.m. The scan at 975 feet began
to show a temperature difference as a result of the injection down the
tubing. The temperature differential is easily seen at 1,025 feet
(Figure B-13).
A final scan at 850 feet indicated the temperature variation one
would expect with cooler fluid flowing in a channel outside the long.
string.
Conclusion
The Radial Differential Temperature tool easily identified cooler
water flowing in the outside tubing in the Leak Test Well.
-------
I
Flow =
710ft
905ft
1^ 935 tt
Injection Zones
Cement
1070 ft
1. 2 3/8-in Tubing
2. Baker Model "R" Profile Nipple
3. Baker Model "f" Profile Nipple
4. 5 1/2-in Long String
1100ft
1120 ft
1130ft
Leak Test Well
Figure B-11. Radial differential temperature survey.
76
-------
~ ;->_-. i- -i|
1 ' " ' t~^ ^ ~ "\^ .__''.~".f"~*^ ~~ ** *"~ ' ""1'
Figure B-12. RDT scan no-flow condition.
-------
~:~_." Scan #9 . '. ;_' [i...:,1^. 1'-1l:-,-r~
"' 1025ft - -i - j :: r;~zq.r4=^T-t-n-
:~".:;1140 Hours \r^-.^.r^^=^~^~^^:
Figure B-13. RDT scan flow condition.
78
-------
Test Mo. 7: Nuclear Activation Technique for Detecting Flow
Behind Casing
Introduction
On August 28 and 29, 1987, personnel from the Rdbert S. Kerr
Environmental Research Laboratory (RSKERL); EPA Regibn IV, Atlas
Wireline and Shell Western E & P conducted a series of tests to
determine flow behind pipe using the PDK-100 pulsed neutron logging
system.
The purpose of the test was to determine if flow of water at two
different rates could be detected behind pipe in a "real world" well.
Shell personnel had agreed to the use of an abandoned 10,600-foot
gas well in which a 100 + foot channel had been identified using a
radioactive tracer survey.
Test Well Conditions
The well, Little Creek 2-6A, has 5 1/2-inch long string which hiid
been cleaned out to perforations at 4,163 feet. The test was then
conducted in two stages: with a packer set at 4,000 feet and the
PDK-100 located below the packer in the long string, and with the
packer set at 4,125 feet and the PDK-100 located within the tubing.
Test Procedure
The first objective was to determine if the previously identified
channel was still present behind the casing. This was done with a
radioactive tracer survey as follows:
A. Tracer FloLog
1. Rig up Atlas Wireline Services and go into the hole with
1 11/16-inch O.D. dual detector tracer instrument. Place
instrument 5 feet above perforations.
2. With the instrument stationary, start water injection into
the perforations at 4,162 feet with the pump truck
operating at a rate of 1/2 barrels per minute (BPM),
3. When the injection rate stabilizes, eject a slug of
radioactive iodine-131 into the flow and verify its mode of
travel. The material should travel downward past the two
radiation detectors and into the perforations. If upward
channeling exists, the material should travel up behind the
casing within the channel, passing the detectors again,
but in reverse order.
4. After channeling has been detected and the radioactive
material has moved past the instrument, move the
instrument upward rapidly, catching and recording the
travel path of the radioactive material. (The instrument is
-------
moved up and down past the slug repeatedly to accom-
plish this.)
5. Reposition the FloLog instrument 5 to 10 feet above the
perforations and repeat steps 2 through 4 to verify all
previous measurements.
6. Stop water injection and remove the Tracer FloLog
instrument from the well.
. . " . .! ; , .j '.
This procedure established that a channel existed behind the
casing from 4,162 feet to about 4,020 feet. Having established this
fact, the following procedure was used to test the PDK-100:
1 Configure the PDK-100 with the pulsed neutron source
beneath the radiation detectors so that upward flow may
be identified.
2. Go into the hoje and locate too! 5 to 10 feet above the
perforations but: below the tubing and packer.
3. Turn the PDK-lOO instrument on and record the no-flow
response.
4. Start the water injection at the rate of 1/2 BPM.
5. Turn the PDK-100 on and record the results. Adjust the
flow to 1/4 BPM and record the results.
I . , , i , . i
6. Move the PDK-100 to the mid-range of the channel.
7. Turn on and record the results at both 1/2 and 1/4 BPM.
8. Move to the top of the channel and record the results at
both flow rates.
... ,.j .. , i ... i .
9. Move out of the channel area and record the results. If no
movement is present, stop the water injection and remove
tool from the well.
10. Reset packer at 4,125 feet and rerun surveys with the
PDK-100 within the tubing.
I i
11. Rig the wireline unit down and review results of both
surveys.
80
-------
Conclusions
The first series of tests, with the tool below the tubing and
packer, included stations at 4,180, 4,150, 4,100, and 4,050 feet. The
second series, with the tool located within the tubing, included tests
at 4,100, 4,050, 4,000, 3,990, and 3,950.
The PDK-100 detected both flow rates with the tool either in the
casing or within the tubing. The top of the channel was determined to
be between 4,000 and 4,050 feet.
The PDK-100 has the potential for providing an excellent method
for detecting flow behind pipe. However, additional work needs to be
done to determine specific applications for the tool.
-------
Introduction
During the week of February 22, 1988, a series of tests was
conducted on the Leak test Well at the Mechanical Integrity Test
Facility at Ada, Oklahoma. The series included pressure tests,
monitoring well response, and volume versus pressure tests. Each
test was designed to provide information on mechanical integrity of
injection wells and to determine whether or not flow from a leak in an
injection well could be identified in an adjacent monitoring well.
1 ! '
Well Configuration
The Leak Test Well was configured as shown in Figure B-14:
surface casing set at 5^1 feet and cemented to the surface; long
string set at 1,215 feet and cemented to 925 feet; injection tubing set
on a packer at 1,084 feet; sliding sleeve in injection tubing closed;
profile nipple at 700 feet in outside tubing open; profile nipple at 920
feet in outside tubing closed. The profile nipple has three 3/16-inch
openings to allow flow from the outside tubing.
Two pressure gauges and two flow meters were installed in the
flow line between the pump and the injection well so that an accurate
determination of the injection pressure and flow to the well could be
obtained.
Three monitoring wells are located around the Leak Test Well in a
radial pattern 20 feet from the Leak Test Well (Figure B-15). Table
B-3 indicates the depths of the monitoring wells and the depth to
water in each well. The water table was measured using a weighted
steel tape and was corrected to depth below land surface.
On Monday, February 22, 1988, pressure transducers were
installed in the 710-foot and 935-foot wells to monitor any water-level
changes that might occur prior to and during the tests. The pressure
transducers were used in conjunction with an SE200 Hydrologic
Analysis System which is marketed by In-Situ, Inc. The transducers
are 0.85-inch diameter stainless steel and the SE200 was
programmed to record data (in this case water levels) every 10
minutes for 10,000 minutes.
Mud in Annulus
The Leak Test Well was completed in January 1985 with the long
strong/surface casing annulus full of native drilling mud above the top
of the cement around the long string The mud weight was recorded
as 9.7 Ib/gal upon completion of reaming the well, prior to setting and
cementing the long string. The gel strength was 3 lb/100 ft2 at 10
seconds and 4 lb/100 ft2 at 10 minutes.
82
-------
Injection Zones
1. Surface Casing (571 ft)
2. 2 3/8-in Tubing
3. Baker Model "L" Sliding
Sleeve
4. Baker Model "R" Profile
Nipple
Cement 5 Baker Mode, ซM_^
Tension Packer
6. 2 3/8-in Tubing
7. Baker Model "R" Profile
Nipple
8. Baker Model "F" Profile
Nipple
084 ft Depth Of
Lower Packer a
1100ft
1120ft
1130ft
Figure B-14. Leak Test Well.
The purpose of this test was to determine if water, injected into
the well and out the profile nipple at 700 feet, would move into a zone
open to the well bore or move up the well bore through the drilling
mud to the surface. Water was to be injected down the outside tubing
while a surface valve on the tubing/long string annulus was open so
that the flow would discharge at the surface, thus removing any air in
the system as the outside tubing and tubing/casing annulus filled with
water. In addition, a bull plug was removed from the surface casing
-------
Monitoring Well
No.1
O
Monitoring Well
No.2
O
Leak Test Well
North
Q Monitoring Well
| No.3
Monitoring Monitoring
Removable Well ' Well
Well -ซซ. 1 2
Monitoring
Well
680 ft
710 f
-
I
!
(
1
t
f
=
>!
! i
1*1
\
7 ft
3ement
905 ft
935 f
t
f
^=:
.
t
a
i
5c
j\
152 ft
/ Borehole
/
reen
1120 ft
(f
;
',
.'
i
f
JU
'
1
'!
#
i
195 ft
/-Casing
."'v:',;/"'
1130 ft
Figure B-15. Monitoring wells.
Table B-3. Depth to Water in Monitoring Welis
Monitoring Well
Depth (feet)
Water Level Below
Land Surface (feet)
710
935
1,130
157
152
195
84
-------
so that the surface casing/long string annulus was open to the
surface.
At 1:05 p.m. the pump was turned on to begin pumping water into
the outside tubing. In 8 minutes water began flowing out of the
tubing/long string annuius at the surface. At that time the valve on the
tubing/long string annulus was closed so that the flow into the outside
tubing would exit only through the profile nipple at 700 feet.
At 1:46 p.m. a significant pressure drop was noted on the
pressure gauge (from 300 psig to 70 psig), and water began flowing
at the surface through the bull plug opening in the surface casing.
The water flowing from the surface casing was clear, indicating
channeling through the mud rather than displacement of the mud.
On Friday, February 26, after all the other tests were completed,
an attempt was made to determine how much pressure would be
necessary to reopen the apparent channel in the mud and establish
flow at the surface. The bull plug in the surface casing was removed
and injection begun down the tubing/long string annulus at a pressure
of 30 psig and a flow of 3.3 gpm. Within 1 minute flow of clear water
appeared at the surface from the bull plug opening in the surface
casing. The pressure was gradually reduced and the flow decreased.
Even when the pressure gauge snowed zero pressure, there was a
trickle of water flowing from the surface casing. The low pressure
necessary to induce flow indicates that the channel created on
February 22 remained open.
The fact that a channel was created in the mud initially and the
channel did not "heal" over the 5-day test period is of concern when
considering abandoned wells and wells with inadequate casing
through underground sources of drinking water. In a report titled,
"Determining the Area of Review for Industrial Waste Disposal Wells,"
Stephen E. Barker investigated mud strengths and the pressure
required to initiate flow in abandoned walls. He stated that in addition
to the pressure required to overcome the hydrostatic head of the
borehole mud, the pressure necessary to displace the mud varies
directly with the gel strength and well depth and inversely with
borehole diameter.
0.00333 (GS) (h)
D
GS = gel strength, pounds/100 ft2
h = height of mud column or depth of well, feet
D = hole diameter, inches
P = displacement pressure, psi
The constant 0.00333 has the units ft/inch
-------
The resistance to flow in the long string/borehole annulus ot me
Leak Test Well includes the mud column in the long string/borehole
annulus (mud gradient .499 psi/ft), and the gel strength of the mud
(assume 25 lb/100 ft2)
j Mud Column
.499 psi/ft X 700 ft = 349 psi
Gel Strength
.00333 ft/in X 251W100 ft2 X 700 ft
6 in
Thus, the resistance to flow is the sum of the mud column and
the gel strength: 349 psi + 9.7 psi = 358.7 psi.
The inducement to flow includes the water column in the long
string above the 700-inch zone and the pump pressure:
Hydrostatic pressure = .434 psi/ft x 700 ft = 303.8 psi
Pump pressure - 300 psig
. '
Inducement to flow = 603.8 psi
Thus, the differential pressure available to cause flow in the long
string annulus is 245.1 psi (603.8 psi - 358.7 psi).
If the gel strength of the mud in the annulus had reached 120
lb/100 ft2, the displacement pressure due to the gel strength woud
have only been 46.6 psi and the differential pressure to cause flow in
the annulus, 208.2 psi.
It appears that at shallow depths, the probability of fluids from
leaks in the system being able to move through the mud, either to the
surface or into permeable zones, is very high when injection
pressures are additive to the hydrostatic pressure.
Pressure Monitoring
This test was designed to establish if monitoring a positive
oressure, without an initial pressure test, could detect the same leak
in the system that could be detected with a pressure test. To perform
the test a 5 1/2-foot standpipe was attached to the tubing/long string
annulus to provide the positive pressure for the test (Figure B-16).
The plastic standpipe was graduated into 1 foot increments and each
foot of the standpipe held 1 gallon of water. The base of the stand-
pipe was located 3.2 feet above the land surface.
The standpipe was attached to the well on Wednesday morning
and the system was filled with water to the 5-foot level. The water-
level decline was then measured every hour for 7 hours, until the
86
-------
Injection
Tubing
Long
Bull Plug
5ft
Standpipe
Long
String
Surface Casing
Bull Plug
Figure B-16. Standpipe.
Valve
Outside
Tubing
Land Surface
Table B-4 Water-Level Decline in Standpipe
Time
(24 hours)
0810
0910
1010
1110
1210
1310
1410
1510
1610
Elapsed Time
(hours)
0
1
2
3
4
5
6
7
8
Water Level Decline/Hour
(feet) (feet)
5.0
4.3
3.7
3.1
2.4
1.8
1.2
0.5
Below
0.0
0.7
0.6
0.6
0.7
0.6
0.6
0.7
base of standpioe
-------
water level fell below the base or me' stanop.pe. iaiwป DV*
the water-level decline in the standpipe over the period of the test.
At 1:45 p.m. on February 25, 1988 (about 29 hours after start of
the test) the bull plug in the long string was removed and a water-
L j measu'emenf was taken. The water level inthetubmg/lor*
string annulus was 16.1 feet below the base of the standp.pe.
The average decline in the standpipe was .64 ft/hour, slightly
slower thin thl tubing/casing annulus decline of .76 ft/hour This is
due S he fact that the standpipe contained 1 gallon perJtoot and fte
tubing/casing annulus contained only 7938 gaUona.per.foot The
decline noted calculates to a .01 gpm, .6 gpd or .34 bpd leak.
The data indicate that the monitoring system can detect a leak of
this size. The amount of positive pressure is not signrfcantrt is just a
means of putting the fluid level above the surface where the water
level decline can be readily observed. The pressure deferential
created by the hydrostatic column in the casing/tubing annulusand
the formation pressure is what controls the rate of decl.ne andI rateoi
volume lost. Had the test been allowed to run long enough, the rate
of decline and volume of fluid lost would have stead.ly decreased until
equilibrium was reached.
Standard Pressure Test
The standpipe was removed and the well was subjected to the
standard mechanical integrity pressure test to determine effects, of
different pressures and length of tests on the capability for' detecng
leaks. Table B-5 indicates the pressures applied and pressure decline
over time.
In some UIC programs, if a well does not lose more than 10
percent of the pressure in 30 minutes, it passes the pressure test.
The Leak Test Well failed the test in less than 2 minutesi ataฎ
pressures (Figure B-17). In each of the tests (50, 100, 200 and 400
psig), more than 75 percent of the initial pressure was lost within 15
minutes.
The water-in-annulus test, in which a well passes the mechanical
integrity test (MIT) if the water level does not decline more than 5 feet
per hour, is of concern. A water-in-annulus test would have indicated
that the Leak Test Well had mechanical integrity. As themonrtonng
showed, the water-level decline was only 0.76 ft/hour. Thus, under
the pressure differential (70 psi) created during this test by the
hydrostatic column in the annulus, the fluid loss was 01 gpm,,0.6
qph 14.4 gpd or .34 bpd. However, a tubing or packer leak, sufficient
to create only a 120 psi differential (equivalent to the 50 psig test on
Table B 6) between the annulus and the receiving formafon wouki
lose 0.8 gpm, 48 gph, 1,152 gpd or 27.4 bpd through the same holes.
88
-------
Tafafe S-5 Pressure Decline Over Time
Pressure Applied 50 psig 100psig 200 psig 400 psig
Pressure Differential 120 psi 170 psi 270 psi 470 psi
(700 ft sand)
Time (min) Pressure Decline (psig)
0
5
10
15
20
25
30
50
.25
15
10
5
3
0
100
43
22
13
7
5
3
200
77
35
18
13
7
6
400
225
117
60
34
23
15
Injected Volume Versus Pressure/Leak Detection through Monitoring
Wells
The next test run on the well involved evaluating injected volume
versus pressure, to determine whether or not a monitoring well can
detect an increase in pressure as a result of flow through a leak in
the system.
While injecting at pressures of 50, 100, 200 and 400 psig for 30
minutes, the flow into the profile nipple at 700 feet was measured
(Table B-6). At the same time the injection-volume relationship was
being determined, the In-Situ Hydrologic Unit was recording the
changes in water level in the two monitoring wells every 10 minutes
(Tables B-7 and B-8).
Injection began at 3,892 minutes elapsed time at 50 psig; 100
psig at 3,922; 200 psig at 3,952; and 400 psig at 3,982. Injection was
stopped at 4,012 minutes.
The data from the transducers clearly demonstrate that the water
level in the 700-foot sand is responding to the different rates of
injection (Table B-7, Figure B-18) while that in the 900-foot sand
(Table B-8, Figure B-19) is not. The water-level drop during the 400
psi injection probably indicates that some other zone began taking
water, since in this test the pressure was maintained as a constant,
rather than the flow rate. Additional "injection volume versus
pressure" tests need to be conducted over a longer period of time to
establish more data for determining the reason for the water-level
drop.
-------
Pressure (psig)
400 Gh
300 j-
200
100
Pressure-50
Pressure-100
Pressure-200
I Pressure-400
0 10
Time (min)
Figure B-17. Pressure decline over time.
20
30
Table B-6.
Tim'e
Start Test
0800
0830
0900
0930
Flow through Profil
Elapsed Time
(min)
3892
3892-3922
3922-3952
3952-3982
3982-4012
3 Nipple
Injection
Pressure
(psig)
0
50
100
200
400
Differential
Pressure
(psi)
70
120
170
270
470
Injection
Volume
(spm)
0.01
0.8
1.0
1.2
2.7
Volume Versus Pressure-Hole Size
The next test performed was to determine the volume of water
which could be pumped through a 1/32-, 1/16-, 1/8-, and 3/16-inch
orifice at varying pressures. Orifices of these sizes were drilled into
caps fitted onto the injection line. With all valves closed to the well,
90
-------
different pfegsuteS were applied and the flow rate was measured
a stopwatch and a graduated bucket.
Table B-7. Changes in Water Level - Monitoring Well No. 1
Elapsed Time
(min)
3810
3820
3830
3840
3850
3860
3870
3880
3890
3900
3910
3920
3930
3940
3950
3960
3970
3980
3990
4000
4010
4020
' 4030
4040
4046
Water Level
(ft)
167.06
157.06
157.06
157.06
157.06
157.06
157.05
157.05
157.04
157.05
157.07
157.07
157.08
157.07
157.11
157.12
157.12
157.12
157.17
157.05
157.02
157.06
157.10
157.09
157.11
Change in Water level
(ft)
0.06
0.06
0.06
0.06
0.06
C.06
0.05
0.05
C.04
0.05,
0.07
0.07
0.08
0.07
0.11
0.12
0.12
0.12
0.17
0.05
0.02
0.06
0.10
0.09
0.11 .
-------
157.2 ,
Water
Level
(ft)
157.1
157.0
Water Level
3,800 3,900 4,000
Elapsed Time (min)
Figure B-18. Changes in water level - 700-ft zone.
4,
100
Table B-9 indicates the results of this test. !t should be noted that
the differential pressure between the casing/tubing annulus (gauge
pressure + hydrostatic pressure) and the formation pressure
opposite the hole determines the volume of a flow through a certain
size hole in a well casing.
92
-------
table B-8. Changes in Water Level - Monitoring Well No. 2
Elapsed Time
(mm)
3810
3820
3830
3840
3850
3860
3870
3880
3890
3900
391 Q
3920
3930
3940
3950
3960
3970
3980
3990
4000
4010
4020
4030
4040
4046
Water Level
(ft)
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.04
152.03
152.03
152.04
152.03
152.04
Change in Water Level
(ft)
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.04
0.03
0.03
0.04
-------
152.5 r-
Water
Level
(ft)
Water Level
152 &^~
3,800
3,900
4,000
4,100
| Elapsed Time (min)
Figure 4-19- Changes in water level - 900-ft zone.
Tatjle B-9. Flow of Water (gpm) through Holes at Different Pressures
Hole Size (inch)
Pressure (psig)
50
100
200
400
1/32
0.675
0.088
0.115
0.160
1716
(3.65
0.85
1.13
1.60
1/8
2.45
3.31
4.50'
**
3/16
5.5
ill
Barely reached 50 psi
Could not get above 220 psi
94
-------
Goncfusions
The tests run during this week provide a number of points that
need to be considered in determining mechanical integrity of injection
wells:
1. A channel was created in the mud in the long string/well bore
annulus on Monday and the channel was still open on Friday.
2. Data from the monitoring system (standpipe) were sufficient to
detect a leak in the Leak Test Well.
3. The Leak Test Well failed the standard pressure test in less than
2 minutes.
4. The Leak Test Well would have passed a water-in-annulus test
even though it failed the standard pressure test. The water-in-
annulus test needs more study to determine if it is indeed a valid
test.
5. The water level in the 700-foot zone clearly responded to the
injection rates in the Leak Test Well, while the water level in the
900-foot zone showed no response to the injection.
-------
Introduction
On February 15, 1988, personnel from Atlas Wireline Services
conducted a noise survey on the Leak Test Well at the Mechanical
Integrity test Facility. This survey, termed SONAN, was conducted to
determine if flow through the outside tubing in the well could be
identified.
!
f ' : ' i
Well Configuration
The well was configured as shown in Figure B-20, with injection
maintained down the 2 3/8-inch outside tubing. The noise tool was
located inside the 2 3/8-inch injection tubing.
Noise Survey
The tool used by Atlas Wireline has an outside diameter of
1 11/16 inches, a length of 3 feet, a bottom hole temperature rating of
350 degrees and a bottom hole pressure rating of 15,000 psi. The
tool has the capability to take data at Steven frequencies as shown on
Table B-10. On the table, HPO1 relates to 200 HTZ, HPO2 to 600
HTZ, HPO3 to 1,000 HTZ, HPO4 to 2,000 HTZ, HPO5 to 4,000 HTZ,
HP06 TO 6,000 HTZ and HPO7 to 8,000 HTZ. On this particular test,
readings were taken only through 2,000 HTZ, thus the numbers
recorded for HPO5, 6 and 7 are meaningless.
Data was taken at. 25 stations in the well (Table B-10). At least
four readings are taken at each station to identify the frequency
character of the sound sources.
The curves presented on Figures 8-21 and B-22 and the data on
Table B-10 are developed by recording the peak millivolt reading for
each frequency at each station. The readings for HPO1 represent the
noise levels for 200 HTZ and up; HPO2 for 600 HTZ and up; HP03
for 1,000 HTZ and up; and HPO4 for 2,000 and up.
I ' ' . ;
!
Conclusions
The attached copies of the Sonan Log (Figures B-21 and B-22),
present data indicating increased noise at the hole in the long string
at 1,070 feet. However, two noise anomalies (980 and 860 feet) have
not been fully explained.
R.M. McKinley, Exxon Production Research, Houston, Texas, is
one of the foremost authorities on the noise tool. He has stated that
extraneous sources of sound are the greatest impediment to noise log
quality control. These extraneous sources may be surface equipment
noise or inadvertent flow past the sonde or continued movement of
the logging tool during measurement.
96
-------
Injection Zones
1. Surface Casing (571 ft)
2. 2 3/8-in Tubing
3. Baker Model "L" Sliding
. Sleeve
4. Baker Model "R" Profile
Nipple
Cement 5_ Baker Model "Ad-1"
Tension Packer
6. 2 3/8-in Tubing
7. Baker Model "R" Profile
Nipple'
8. Baker Model "F" Profile
Nipple
1084 ft Depth of g 5 1/2.in Long String
Figure B-20. Leak Test Well.
Another interesting aspect of the log interpretation relates to
single versus multiphase flow. Multiphase flow is characterized by
lower frequency sound than single phase flow, and is indicated by the
^separation of the 200 HTZ curve from the other curves. This
'phenomenon seems to be quite vividly portrayed on the log. Thus, air
was apparently being pumped down the. well along with the water.
Much additional testing is necessary to fully understand the
capabilities of the noise tool in mechanical integrity testing and to fully
-------
\
Table B-10. Leak Test Well - SONAN Data
Depth Corrected Data for Line
HPO1 HP02 HP03 HPO4
: , 650.00 58 13
700.00 48 20
750.00 51 10
800.00 74 16
860.00 132 11
. 900.00 83 22
940.00 106 11
980.00 144 12
1000.00 101 17
1020.00 90 11
1040.00 59 10
1050.00 135 29
1055.00 119 16
1060.00 137 13
1065.00 14 16
1070.00 136 9
1075.00 86 14
1080.00 89 11
1090.00 ,64 11
1100.00 73 10
1120.00 62 14
1140.00 35 20
1160.00 26 8
1180.00 20 15
1200.00 17 7
interpret data presented on
that the following articles be
when dealing with noise surv
10
13
9
12
9
12
10
11
11
8
9
7
11
8
8
8
10
8
9
8
13
17
8
7
6
:he noise
9 made a
reys:
98
9 ;
7
6
4
7
7
7
7
7
6
7
4
6
7
7
6
9
.7
7
I
7
8
,_,
7
4
4
Length, Size
HPO5 HPO6 HPO7
,"'3 '"' 2" ' 1 " '
3 2 1
32 1
3 2 1
1 ii
321
4 2 1
3 2 1
321
3 2 1
3 2 1
3 2 1
3 2 1
3 2 1
321
3 2 1
3 2 1
32 1
3 2 1 ' "\
3 2 1
32 1
3 2 1
32 1
3 2 1
3 2 1
3 2 1
log. It is strongly recommended
part of a library for assistance
-------
Figure B-21. Atlas Wireline Services SONAN Leak Test Well.
MoKinley, R.M., Bower, P.M., Rumble, R.C., "The Structure and
Interpretation of Noise from Flow Behind Cemented Casing," Journal
of Petroleum Technology (March 1973), pages 329-338.
McKinley, R.M.; Bower, P.M., "Specialized Applications of Noise
Logging," Journal of Petroleum Technology (November 1979), pages
1387-1395.
McKinley, R.M., "Production Logging," SPE Paper 10035, undated.
-------
a*
=1
1
i
N
H
y
=i
-t
H
o
a
to
o
0
o
o
o
0
E!"
:ซ
31
i'l
!
3i
I
1
i
3
-1
-1
J
-1
i
3!
fl
3,'
3!
}
i
>
i
i
f
i
V
Figure B-22. Atlas Wireline Services SONAN Leak Test Well:
100
-------
Test No. 70: Temperature Survey
Introduction
On September 10, 1985, personnel from the Tom Hansen
Company ran a temperature survey in the Leak Test Well. The
purpose of the survey was to determine the geothermal gradient, the
.fluid level, and, by including a casing collar locator, the location of the
casing collars, packers and sliding sleeve in the well.
Well Configuration
The well had surface casing set at 571 feet and cemented to the
surface, and long string set to 1,215 feet with the top of the cement
at about 920 feet. When the injection tubing was set, calculations
indicated the upper packer should be at 1,057 feet, and the lower
packer at 1,084 feet, with the sliding sleeve at about 1,070 feet
(Figure B-23).
Temperature Survey
The sequence of events leading to the temperature survey
included:
9-5-85 The 5 1/2-inch casing was filled with water and the zone
1,120-1,130 perforated with 2 shots per'foot {SIEDP - deep
penetrating shot). The average hole size should be .411
inches and the average depth of penetration of the shot, 19
inches. The water level was measured and was 20 feet below
the land surface.
9-9-85 the water level was 46 feet below the land surface. The well
was swabbed before setting tubing and packers. Operators
set the tubing and packers and prepared to run the
temperature survey.
Two runs were made in the well, each run taking about 35
minutes. A geothermal gradient could not be obtained since the work
done in setting the tubing (swabbing, etc.) resulted in moving fluid
from the casing into the injection zone. As a result, the log reflected
the results one would expect after injection has taken place.
The temperature went from 76 ฐF at the surface to 75.5 ฐF at total
depth in the first run. The second run indicated a surface temperature
of 80ฐF and a bottom hole temperature of 73.5ฐF. The fluid level in
the second run was about 474 feet below the land surface.
A very interesting part of the log presentation was the casing
collar locator. It indicated the upper packer at a depth of 1,054 feet,
sliding sleeve at 1,066 feet and the lower packer at 1,084 feet. This
compares favorably with the calculated locations made prior to setting
the tubing and packers.
-------
Injection Zones
1. Surface Casing (571 ft)
2. 2 3/8-in Tubing
3. Baker Model "L" Sliding
Sleeve
4. Baker Model "R" Profile
Nipple
5. Baker Model "Ad-i"
Tension Packer
6. 2 3/8-in Tubing
7. Baker Model "R" Profile
Nipple
8. Baker Model "F" Profile
Nipple
9. 5 1/2-in Long String
10. Baker Model "C-1" Tandem
Tension Packer
1120ft
1130ft
Figure B-23. Leak Test Well.
](i ,. ; . , | '
Conclusions
The CCL portion of the log was good for comparing the
calculated location of bofh packers and the sliding sleeve. The
temperature log gave no significant information since a geothermal
gradient was not obtained. A base log .to establish the geothermal
gradient is extremely important for developing a meaningful log.
Temperature, differential temperature and radial differential
temperature logs will be run on the well at a later date.
102
-------
Test No. 11: Continuous Flow Survey
Introduction
On October 13, 1987, personnel from Gearhart Industries, Inc.,
ran a continuous flow survey on the Leak Test Well. The purpose was
to determine whether or not the survey could identify leaks in the
well, and, in turn, whether or not flow was exiting through the
perforations.
Well Configuration
The well was configured as shown in Figure B-24. The sliding
sleeve was closed so that all flow down the injection tubing went to
the perforations at 1,120 to 1,130 feet.
Flow Meter Survey
The survey was run by pumping water down the injection tubing
at different rates and checking those rates with calculations of flow
from data collected by the flow meter. Also, readings were taken at
specific intervals to determine if flow was leaving the injection tubing.
Flow rates from the pump were about 7, 6, 1 and 1/2 gallons per
minute. However, accurate determinations of flow rates were not
possible owing to a malfunction of the flow meter in the flow line to
the well.
Twenty-one different "runs" were made in the well under flow and
no-flow conditions. Each run was indicated on a log as a file with the
following information presented: speed of tool movement, either up
or down in feet per minute; depth interval; average revolutions/second
of the spinner in a clockwise and a counterclockwise direction; and
the flow rate, in revolutions per second, in the clockwise and
counterclockwise direction.
In the initial test (File 1), the tool was held stationary at 193 feet
and water was injected down the injection tubing at a rate of about 7
gpm. The calculated flow rate past the tool was 265 barrels per day
(7.7 gpm).
The flow was changed to about 6 gpm, and the tool (File 2)
indicated a flow of 221 bpd (6.4 gpm) (Figure B-25).
For File 3, the flow rate was changed to about 1 gpm and the tool
indicated a flow of 44 bpd (1.3 gpm).
The flow rate for File 4 was about .5 gpm and the calculated flow
from the tool data was 15 bpd (.43 gpm).
The remaining runs (files) involved moving the tool up or down in
the tubing under no-flow conditions to check the flow meter, stop
checks above and below the sliding sleeve to determine whether
-------
Injection Zones
1. Surface Casing (571 ft)
2. 2 3/8-in Tubing
3. Baker Model "L" Sliding
Sleeve
4. Baker Model "R" Profile
Nipple
5. Baker Model "Ad-1"
Tension Packer
6. 2 3/8-in Tubing
7. Baker Model "R" Profile
Nipple
8. Baker Model "F" Profile
Nipple
9. 5 1/2-in Long String
10. Baker Model "C-1" Tandem
Tension Packer
Figure 6-24. Leak Test Well.
leakage was occurring out the sleeve and a check of the flow out the
perforations.
Figures B-25 to B-28 indicate specific runs and comments of the
operator based on data provided by the survey.
104
-------
ฃ DATE FLOUM SERIAL * PROGRAM MODE JOB * FILE
AS 13 OCT 87 O3S 3-ZOOI-04 STAT 0 2
.....:
F
1
,=
1
.P.?
T/
.!?.S
MI
i
1
N
j,
ay
t
T
-
N.
bo
r=
=H
. , :
193
DEPTH
STOP CHECK (FLOUM)
SO CCU NHZ S3
AVERAGE CU
'-ง' REV /S6C 5'
-- AVERAGE CCU --
-5
-CCU
5
T~?
cr
_._i
t
4
REV/SEC 5
FLOU RATE *CU
R
j
EV
1 1
1 i
1
L
1 j
1
. n_
.... .. i
/SEC
!
j
'
, i
--
-
' '
I
-
5
1 t
Figure B-25. Qearhart Industries, Inc., continuous flow survey.
Conclusions
The continuous flow survey indicated that there was a possible
slight loss of fluid through the sliding sleeve and the flow was going
out the perforations about equally.
Continuous flow surveys are a useful tool for determining leaks in
tubing, casing or packers. Additional tests will be done to correlate
more accurately the flow rate from the pump with the estimate from
the tool.
-------
TIME DATE FLOUM SERIAL * PROGRAM MODE JOB * FILE
!:Sl:43 13 OCT 17 03! 3-2O01-04 UP 0 15
. ......
FT/'MIM -100
1024
DEPTH
STOP CHECK (FLOUH)
BO CCU NHZ 93
-s
.--.,
'^S
.--. AVERAGE , CU. .--..,
' REV /SEC
..-
REV/SEC
-CCU FLOW RATE +CU
i I
'P. SPEED. .-DOWN., 1080
FTXMIM -100 DEPTH "^REV/SEC
CCU FLOU RATE *Cll
Figure B-26. Gearhart Industries, Inc., continuous flow survey.
106
-------
ป it
F
S
W
PF
ป
FO
N~
-O
~f
DU
So'
4
"FT /Kit? -ioo
JQ81
DEPTH
i
loai
DEPTH
SrOf CHECK tPLOUH) ,
60 CCU NHZ S3
_ . - - __*ฅ MM A S E C W_.- -
"S
^~~t
AVEaAGESECCU -- S
. , RJIV/S6C 5
-CCU FLOW RATE ซCU
T7
ซ
h^
^
-J
b2Ji
iiU
TT1
r7T
cc
Ll
w
R
FL
EV
0 *
/SI
S
:C
A:
i
*c
-r
=?
J
!^-
S
-S REV/SEC 5
-- AVERAGE CCU --
... ....... .-
-- AVERAGE CU --
-5" REV/SEC S
STOP CHECK (FLOUM) .
60 CCU NHZ S3'
10 = 54 13 OCT 87 O3S 3-JOOI-O4 STAT O 17
ME DATE FLOUM SERIAL t PROGRAM MODE JOB * FIUE
Figure B-27. Gearhart Industries, Inc., continuous flow survey.
-------
ME DATE
9!S2 13 OCT 87
FLOUM
SERIAL *
03S
PROGRAM MODE JOB
G-2OO1-CM DOUN
' FIUE
o 20
SPEED HBOUN
_......_._ ฃ.
STOP CHECK (FLOUM)
'SO CCU NHZ 83
1 103
DEPTH
p_
REV/SEC
-- AVERAGE CCU --
-5
R6V/S6C
-CCU FLOW RATE *CU
S
....
REV/SEC
SPEED -DOUN
"FT/nTiT "-Too"
-5
-S
STOP CHECK IFLOUMI
0 CCU NHZ 83
REV/SEC
-- AVERAGE CCU --
REV~7s"EC
-- AVERAGE CU --
RE"V>"S"E"C
13 13 OCT 37 O3S
E DATE FLOUM SERIAL ซ
3-2001-04 DOUN
PROGRAM MODE JOB
0 20
t FILE
Figure B-28. Gsarhart Industries, Inc., continuous flow survey.
108
-------
Test No. 12: Radioactive Tracer Survey
Introduction
On May 19, 1988, personnel from the Tom Hansen Company
conducted a radioactive tracer survey on the Leak Test Well. The
purpose of the survey was to determine if there were leaks in the
tubing, casing or packer, or channeling in the cement in the area of
the perforations in the long string.
Well Configuration
The well was configured as shown in Figure B-29. The sliding
sleeve was closed so that material injected down the injection tubing
would exit through the perforations from 1,120 to 1,130 feet.
Radioactive Tracer Survey
The first run was to determine a background. The tool used to
develop the base log included a casing collar locator and low-
sensitivity gamma ray.
The survey was conducted while pumping approximately 1/2 bprn
at 300 psig, as follows:
1. Ejected tracer at about 1,000 feet and followed it down the tubing
and out the perforations. Most of the material went out the lower
part of the perforations. Eight runs were taken to trace the
material (Figure B-30).
2. Ejected tracer at 1,125 feet and checked for movement with a
detector at 1,139 feet. No channeling detected (Figure B-31).
3. Ejected tracer at 1,125 feet and checked for movement with a
detector at 1,135 feet (time drive). Some of the material was
detected (Figure B-32), which indicates a channel in the cement
below the perforations.
4. Ejected tracer at 1,107 feet with a detector at 1,115 feet. No leak
indicated.
5. Ejected tracer at 1,102 feet with a detector at 1,110 feet. No leak
indicated.
6. Ejected tracer at 1,066 feet with a detector at 1,076 feet. No leak
indicated.
7. Ejected tracer at 1,052 feet with a detector at 1,060 feet. No leak
indicated.
8. Ejected tracer at 1,042 feet with a detector at 1,050 feet. No leak
indicated.
-------
si 680 ft
710ft
905 ft
a 935ft
Injection Zones
1. Surface Casing (571 ft)
2. 2 3/8-in Tubing
3. Baker Model "L" Sliding
Sleeve
4. Baker Model "R" Profile
Nipple
5. Baker Model "Ad-1"
Tension Packer
6. 2 3/8-in Tubing
7. Baker Model "R" Profile
Nipple
8. Baker Model "F" Profile
Nipple
. 9. 5 1/2-in Long Strirtg
10. Baker Model "C-1" Tandem
Tension Packer
Figure B-29. Leak Test Well.
1 i
Conclusions
'"" !' ;:" , ! "" ' ;
The radioactive tracer survey indicated a slight channel down
from 1,130 to 1,135 feet. There was no indication of channeling at
1,139 feet. There were no indications of leaks in the tubing or packer'
or channels above the perforations.
110
-------
Figure B-30. Tracer runs showing fluid movement during injection at 1/2
bprti 300 psi.
-------
-r-ii-
-+,>
"I"
L;J!-
Figure B-31. Slug #6 ejected at 1,125-ft channel down check.
112
i )
-------
- -
-_.;-;--;
_ --;
-*"-//-{
\=^=
J-5-j
'!===
^
/I ' I
^%d
tn i
II : 1
;"--
~: -' '-'-
' l ' ;
; :
i i
; ' \ !
! '
M"| , r-
! 1 1 1
' - ,
! i
\-s^tsnr-
. ' 1 ;
-=;=
Figure B-32. Slug #7 ejected at 1,125-ft channel down check.
-------
.
Cement
1070ft
_
nqoft
11:20 ft
1130ft
Figure B-33. Leak Test Well.
Injection Zones
114
-------
Test No. 13: Differential Temperature Survey
Introduction
On November 4, 1987, personnel from Schlumberger Well
Services ran a differential temperature survey in the Leak Test Well.
The survey was run after injection had been ongoing for about 5
hours and the well had been shut in for 16 hours.
Well Configuration
The well was configured as shown in Figure B-34. The injection
tubing had been pulled and injection was taking place down the long
string into the perforations at 1,120 to 1,130 feet. The surface valve
on the outside tubing was closed so that no fluid could move through
this area.
Differential Temperature Survey
The log presentation included curves for gamma ray, casing
collars, temperature gradient arid differential temperature.
The temperature gradient was from 65.5 ฐF at about 200 feet to
73ฐF at 1,215 feet. The differential temperature curve indicated a
slight (0.2ฐF) change in temperature at the base of the injection zone.
The fluid level was indicated at about 173 feet below land surface by
the temperature gradient curve.
Conclusion
The differential temperature log indicated there are no leaks in the
long string.
-------
Injection Zones
Figure B-34. Leak Test Well.
116
-------
Test No. 14: Nuclear Activation Technique for Detecting Flow
Behind Casing
Introduction
On November 3, 1987, personnel from the Robert S. Kerr
Environmental Research Laboratory (RSKERL) and Atlas Wireline
Service conducted a series of tests to determine flow behind pipe
using an oxygen activation tool.
The purpose of the tests was to determine if flow could be
detected behind pipe in the Leak Test Well and, if possible, the
detection limit of the tool.
Well Configuration
Figure B-35 indicates the configuration of the Leak Test Well. A
packer was set at 1,084 feet and a profile nipple was open at 700
feet. Injection was maintained down the injection tubing/long string
annulus, out the 1/4-inch hole in the long string and up the outside
tubing.
Tool Test
The test was conducted with the Atias Wireline 1 11/16-inch
diameter oxygen activation tool (Serial No. 24334) located in the
2 3/8-inch injection tubing. Stationary "no flow" background gamma
ray count rates were taken for both the long spaced (LS) and short
spaced (SS) detectors at depths of 300, 800 and 1,000 feet.
A background count rate was computed for each depth of
investigation by determining the inelastic gamma ray and oxygen
count rates for three no-flow measurements at each station. For each
no-flow measurement, the ratio of the oxygen count rate to the
inelastic count rate was computed, and the average of these ratios
was determined. The result of this activity gives a long-space factor
and short-space factor that are then multiplied times the measured
inelastic iong space and inelastic short space count rate, respectively,
to compute the proper background.
After determining the background factors for each depth
investigated, the too! was moved down the well at speeds of 15 feet
per minute and 30 feet per minute to check the velocity calculations.
The finai part of the test involved injecting water oown the tubing/iong
string annulus at different flow rates and determining what flow could
be detected coming up the outside tubing. Flow measurements were
taken at depths of 1,000, 800 and 660 feet.
Table 8-11 is a summary of specific data taken at a depth of
1,000 feet. The determination of interest during this investigation was
a fiow or no-flow indication. The velocity data are aiso of interest,
although not critical to this series of tests.
-------
Injection Zones
1084 ft Depth of g.
Lower Packer
Surface Casing (571 ft)
2 3/8-in Tubing
Baker Model "L" Sliding
Sleeve
Baker Model "R" Profile
Nipple
Baker Model "Ad-1"
Tension Packer
2 3/8-in Tubing
Baker Model "R" Profile
Nipple
Baker Model "F" Profile
Nipple
5 1/2-in Long String
1100ft
11213ft
1130ft
Figure B-35. Leak Test Well.
the criterion for flow indication is that the long space count rate
must be greater than 1.0 counts/second after subtracting the
background reading. Thus, from Table B-10 flows were indicated at
stations 17 through 26, 36, and 37.
118
-------
7ab)8 B"1t. Oxygen Activation Log Data, Leak Test Well -Novembers, 1987
Depth
(feet)
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
iฐpo
1000
1000
1000
1000
1000
1000
1000
1000
Station
11
12
13
17
18
19
20
21
22
23
24
25
26
28
29
30
31
32
33
34
35
36
37
38
39
40
Flow
SS
.35
-.01
.05
5.18
3.52
3.68
3.17
5.24
6.29
4.50
5.46
6.36
5.59
.26
.80
1.99
.95
1.33
.02
-.49
3.47
3.02
2.82
.60
-.11
.05
Ind."
LS
.19
.16
.11
3.02
3.35
2.60
3.36
4.32
3.73
3i91
2.88
3.16
2.60
.43
.24
.51
.32.
.17
.04
.004
.99
1.19
-1.05
.45
.12
.29
Velocity
None
None
None
14 ft/min
155.88 ft/min
21 .88 ft/min
Q
39.43 ft/min
14.69 ft/min
54.65 ft/min
11. 97 ft/min
10.94 ft/min
9.98 ft/min
0
0
5.67 ft/min
0
0
0
. 0
6.09 ft/min
8.18 ft/min
7.78 ft/min
0
0
0
Comments
Not injecting
Not injecting
Not injecting
Injecting .86 gpm
Injecting .86 gpm
Injecting .86 gpm
Injecting .86 gpm
Injecting 4 gpm
Injecting 4 gpm
Injecting 4 gpm
Injecting 1 .5 gpm
Injecting 1.5 gpm
Injecting 1 .5 gpm
Injecting .46 gpm
Injecting .46 gpm
Injecting .46 gprn
Injecting .46 gpm
Injecting .32 gprn
Injecting .32 gprn
Injecting .32 gprn
Injecting .75 gprn
Injecting .75 gpm
Injecting .75 gpm
No injection
No injection
No injection
With background subtracted
-------
AS previously SVaiWU, Ilia vtnucuy iiieaauioiiienio aio 111101001.. .y
but are not significant in the use of the tool for determining flow
behind pipe at this point in the development of the tool, with one
exception: one must determine the sensitivity of the tool, i.e., the
slowest velocity the tool can identify as flow. The criteria for a valid
velocity measurement are:
1. The flow indication signal for the SS must be at least three times
the error bar.
2. The flow indication for the LS must be at least two times the error
' bar. ^ " .' ' '. ', j " [ " \
3. The LS signal must be less than the SS signal.
4. Neither signal can be zero.
,;!'"', 1 , " |, :,".., i . ,1 .1 .
If any of these criteria is not met, the velocity should be shown as
zero in the data listing. A review of the data sheets from this test
indicates that the velocity measurements meet these criteria.
Conclusions
The 1 11/16-inch oxygen activation tool was successful in
detecting flow up the outside tubing in each of the tests while
injecting at .86, 4, 1.5 and .75 gallons per minute. The tool did not
detect flow at the .46 or the .32 gpm rates.
" | :" "; | . ' " [
The minimum velocity the tool was able to detect during the tests
was 3 ft/min. The results of this and other tests indicate that the
velocity range of the tool in its present configuration is approximately
3 to 100 ft/min.
120
-------
Test Mo. 15: Nuclear Activation Technique for Detecting Flow
Behind Casing
Introduction
On September 14, 1988, personnel from the Robert S. Kerr
Environmental Research Laboratory (RSKERL) and Atlas Wireline
Service conducted a series of tests to determine flow behind pipe
using an oxygen activation tool.
The purpose of the tests was to determine if flow could be
detected behind pipe in the Leak Test Well, both in 2 3/8-inch tubing
and in a channel in the mud system, and, if possible, the detection
limit of the tool.
Well Configuration
Figure B-36 indicates the configuration of the Leak Test Well. A
packer was set at 1,084 feet and a profile nipple was open at 700
feet. Injection was maintained down the injection tubing/long string
annulus<> out the 1/4-inch hole in the long string and up the outside
tubing, out the tubing through the profile nipple at 700 feet and
through a channel in the mud to the surface of the ground.
Tool Test ;
The test was conducted with the Atlas Wireline i 11/16-inch
diameter oxygen activation tool located in the 2 3/8-inch injection
tubing. Stationary "ho flow" background gamma ray count rates were
taken for both the long spaced (LS) and short spaced (SS) detectors
at a depth of 1,075 feet, which was below the injection activity.
Readings were taken during injection at depths of 300, 600 and 1,000
feet to determine both flow/no-flow and velocity.
A background count rate was computed for the 1,075 feet depth
by determining the inelastic gamma ray and oxygen count races for
three no-flow measurements at this station. For each no-flow
measurement, the ratio of the oxygen count rate to the inelastic count
rate was computed, and the average of these ratios was determined.
The result of this activity gives a long-space factor and short-space
factor that are then multiplied times the measured inelastic long space
and inelastic short space count rate, respectively, to compute the
proper background.
After determining the background factor, the final part of the test
involved injecting water down the tubing/long string annulus at
different flow rates and determining what flow could be detected
coming up the outside tubing, and through the channel in the mud
from 700 feet to the surface of the ground.
Table B-12 is a summary of specific data taken during the test.
The determination of interest during this investigation was a flow or
-------
,1 ln n n,
Injection Zones
1. Surface Casing (571 ft)
2. 2 3/8-in Tubing
3. Baker Model "L" Sliding
Sleeve
4. Baker Model "R" Profile
Nipple
5. Baker Model "Ad-1"
Tension Packer
6. 2 3/8-in Tubing
7. Baker Model "R" Profile
Nipple
8. Baker Model "F" Profile
Nipple
084 n Depth of g. 5 i/2-in Long String
Lower Packer
Figure 8-36. Leak Test Well.
no-flow indication within bo^h the Outside tubing and the channel in
the mud. The velocity data are of interest, although not critical to this
series of tests.
The criterion for flow indication is that the long space count rate
must be greater than 1.6 counts/second after subtracting the
background reading. Thus, from Table B-12, flows were indicated at
stations 3 (1,000 feet) and 4 through 9.
122
-------
table B-12. Oxygen Activation Log Data,, Leak Test Well -
September 14,1988
Depth
(feet)
1075
1075
1075 .
1(500
1000
600
600
300
300
300
Station
0
1
2
3
4
5
6
7
8
. 9
Flow
SS
-.61
.16
.51
1.24
.89
71.70
71.17
93.87
57,01
62.19
Ind.
LS
.03
-.01
-.02
1.72
1.68
29.10
26.19
23.76
10.05
10.07
Velocity
None
None
None
0
0
8.49 ft/min
7.66 ft/min
5.57ft/min
4.41 ft/min
4.20 ft/min
Comments
Below injection
Below injection
Below injection
Tubing flow
Tubing flow
Channel flow
Channel flow
Channel flow
Channel flow
. Channel flow
The tests began with a flow of approximately 20 gpm coining from
the pump. Stations 3, 4, 6 and 7 were taken at that flow rate with the
stations opposite the 2 3/8-inch outside tubing (stations 3 and 4) and
the channel in the mud (stations 5, 6 and 7). Although flow was
detected at each station, a much higher flow indication was seen at
Stations 5, 6 and 7. Stations 8 and 9 were taken opposite the channel"
but at a flow rate of about 10 gpm. A reduced flow indication is
evident for these stations.
Conclusions
The 1 11/16-inch Oxygen activation tool was successful in
detecting flow at all stations, although the flow indication was much
.lower at the stations opposite the 2 3/8-inch outside tubing than in
those stations opposite the channel in the mud system. This was
probably due to the larger size of the mud channel.
Additional tests should be run with this tool in real wells to provide
data for evaluating the total capability of the tool for detecting flow
behind pipe.
US GOVERNMENT PRINTING OFFICE:1990-748-159/ao487 REPRINt
-------
-------