Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02406306 2002-10-16
APPARATUS TO IMPROVE THE QUALITY OF WELL CEMENT,
PREVENT GAS MIGRATION AND MEASURE WAITING-ON-CEMENT TIME
BACKGROUND OF THE INVENTION
The Field of the Invention
The present invention describes an apparatus that is useful to improve several
aspects of well cementing. Its purpose is to provide pressure pulsations to
the
cement slurry in oil or gas wells, beginning as soon as practical after the
cement
shiny has been pumped into place, and continuing until the cement makes the
transition from liquid slurry to a solid. These pressure pulsations impart a
reciprocating motion to the fluids in the well as illustrated in Figure 1. In
this
figure, a large spring is used to illustrate the compressibility of the
fluids. When
the cement solidifies to the extent that its compressive strength exceeds the
pressufe of the cement pulsations, the pulsations automatically uncouple. This
solidification process automatically begins at the bottom of the well and
proceeds
towards the surface because of the temperature gradient in wells.
The use of this apparatus improves the seal provided by the cement,
specifically to reduce or eliminate gas migration, to measure the waiting-on-
cement time and to improve the accuracy of cement evaluation logs, as is
described and illustrated in this document. This apparatus provides all of
these
benefits, at the same time, during a single application. The apparatus is
relatively simple and inexpensive to construct and it is relatively easy to
use,
without requiring extensive training. Its design has been optimized for safe
and
effective use at well sites.
CA 02406306 2002-10-16
The Prior Art
Well Cement Seal
After a well has been drilled, casing is typically lowered into the well bore
and is
cemented in place by pumping liquid cement slurry into the annular space
S between the casing and the well bore. The objective is for the cement to
solidify
to form a seal, to isolate formation fluids to their respective formations
thereby
preventing leakage to other formations. The properties of the set cement
approximate the properties of natural formations with respect to the ability
to
isolate the formations from each other, but under some conditions, it may
allow
the leakage of formation fluids. This leakage usually occurs between
formations
with higher fluid pressures at greater depths to formations with lower #luid
pressures at lesser depths. Leakage may also occur all the way to the surface.
Gas leakage is most common because of the high mobility of gases such as
methane, carbon dioxide, hydrogen sulfide, etc., that are encountered in
formations. The migration of liquids such as liquid hydrocarbons, water,
brine,
etc., also may occur, but it is either less common or less commonly detected.
Because it is a common oilfield term, the term "gas migrationp will be used in
this
discussion to denote the leakage of all fluids, including liquids.
Type I Gas Migration
Two types of gas migration are known to the industry, but the term is often
used to denote one or the other without distinguishing between them. This is
because one type of gas migration is often more common in a particular
geographical area and it is usually simply called "gas migration" in that
area.
One type is more specifically called "gas flow after cementing'. When it
occurs,
the migration of gas from gas formations into the cement slurry may occur soon
CA 02406306 2002-10-16
after the cement has been pumped, before it has become a solid. It is a common
problem in some fields and may occur on an unpredictable basis in others. The
consequences range from "gas cut cementp to ublow outs to the surface. The
basic cause is the loss of hydrostatic pressure within the cement column as it
makes the transformation from liquid slung to a solid. This allows gas to
enter
the slurry and cut channels capable of conducting large flows of gas. When the
cement solidifies, the channels become permanent.
Good evidence from field and laboratory measurements has shown that
development of gel strength in the static column of cement is primarily
responsible for the loss of hydrostatic pressure. The combination of gelation
and
the very slight shrinkage of the slurry, due to the chemical reactions
involved in
the solidification process, transfers the weight of the slurry to the walls of
the
casing and well bore. This removes weight from the fluid column, eventually
completely eliminating the hydrostatic pressure.
This wilt be called "type I gas migration" for the purposes of this
discussion.
The apparatus described here prevents gel strength development, because the
pressure pulses from the surface result in movement of the liquid slung in the
well. This movement prevents the formation of the electrostatic intermolecular
bonds that result in gel strength development. Movement continues until such
time, during the solidification process, when the compressive strength of the
semisolid cement is more than the magnitude of the pressure pulses.
Type II Gas Migration
Another type of gas migration occurs some time after the cement has set, often
after weeks or months. This type is more specifically called "annular casing
2S pressure" or "microannuiar gas migration". The volume of gas is generally
much
z
CA 02406306 2002-10-16
less, but pressures can be quite high. It may be necessary to collect the gas
at
the surface and inject it into production lines, to prevent overpressurization
of the
surfiace casing. It is often necessary to apply expensive remedial procedures
to
prevent the build up of high annular pressures when wells are abandoned.
Otherwise, surface casing may be ruptured after wells are abandoned. The
basic cause is due to the very slight shrinkage of the cement slurry when it
makes the transition from liquid slurry to a solid and additional shrinkage
that
occurs during the initial and later stages of the development of compressive
strength. Most of the shrinkage occurs during the initial stages of
compressive
strength development.
This will be called "type II gas migration° for the purposes of this
discussion.
The apparatus described here reduces or prevents this type of gas migration by
maintaining intimate contact between the cement slurry and the casing until
the
cement develops initial compressive strength.
Gas Migration Control
The control of both types of gas migration is one of the most costly and
challenging technical problems in well cementing. Most of the following
discussion refers to efforts to prevent type I gas migration. Efforts to
prevent
type II gas migration have heretofore been relatively unsuccessful. Various
chemical additives have been used to control gas migration in cement slurries.
Some of the additives appear to be completely ineffective, while others appear
to
have different degrees of effectiveness. But al! are expensive and most are of
limited applicability. Such additives typically increase the cost of cementing
casing by a factor of two to five times.
A
CA 02406306 2002-10-16
Relatively few mechanical methods have been used to control gas migration,
but two are in common use. One involves cementing a short column of cement
across a gas zone known to be a problem and leaving a column of drilling fluid
over the cement slurry to maintain hydrostatic pressure as the cement sets up.
This technique may be used with a cement "staging tool" to complete the
cementing operation. Another involves the use of a "liner packer" to prevent
gas
from leaking through the seal between a casing liner into other strings of
casing.
Both methods enjoy a degree of success, but they significantly increase the
cost
and are limited in applicability.
As stated above, type I gas migration can be prevented if gelation of the
cement slurry can be prevented or delayed until the chemical reactions
involved
in solidification develop enough viscosity to prevent the movement of gas
within
the slurry. This can be accomplished by mechanics! agitation. It has been
reported that slowly rotating the casing, after the cement has been placed but
before the slurry sets up, can prevent gas migration. Clearly this method is
limited to applications where casing rotation is practical. Special equipment
is
required to accurately control the torque. Rotation must be stopped when the
drag on the casing at the bottom of the well becomes too high and before
torque
builds to the point that the casing might be twisted off. This may occur
before the
cement is viscous enough to prevent gas migration at shallower depths. This is
because cement slurries begin to thicken and set up at the bottom of wells
first,
because the temperature is higher.
Another method that has been proposed is to vibrate the casing. An apparatus
was constructed to do this, but it was never applied to cementing operations
due
c
CA 02406306 2002-10-16
to the high cost and concerns about parting the casing, due to the very high
stresses that could develop in the walls of the casing during resonant
vibrations.
Waiting-On-Cement Time
After the cement slurry has been pumped into a well, a condition known in the
field as "waiting~n-cement' time or WOC occurs, before operations on the well
can be resumed. This time is required for the cement to develop sufFcient
compressive strength that any stresses applied to the casing, as operations
are
resumed, will not damage the cement sheath. The time required may be
substantial. During offshore applications, the rig time during WOC may be more
costly than slurry and pumping costs. Extensive laboratory testing together
with
detailed knowledge of the temperature gradients, circulation temperatures and
complex computer models have been used to minimize WOC. Downhole
measurements of circulating temperatures and static temperature build up are
necessary for accurate predictions and, during the rare instances where they
have been made, they have shown that many of the models and current tables
are incorrect. Due to the cost and time required to collect the data to refine
the
models, the current state of this technology is inadequate. The current
practice
is to allow conservatively long times, to prevent problems. The cost of this
practice to the industry is very high.
The direct measurement of the WOC time is a by-product of the operation of
the apparatus described in this document.
Evaluating Cement Quality
Cement evaluation logs are used after the cement has been allowed to solidify,
to determine if the placement of cement between the casing and well bore has
been successful. There are several types of logs used to evaluate cement
CA 02406306 2002-10-16
placement but, for the purpose of this discussion, the term "bond log"
includes all
devices that rely on acoustic logging principles to evaluate the placement of
cement. It also refers to the test results of such devices.
Bond logs heretofore have been subject to a number of shortcomings. A good
bond log generally means that the casing is sealed by an adequate amount and
quality of cement, but a poor bond log does not prove poor cement. There are
many reasons why a poor bond Jog may be obtained that does not pertain to the
actual condition of the cement around the casing.
Advances are continually being made in the design and application of acoustic
logs, but they all share one major shortcoming, namely they require very good
physical contact between the cement and the casing in order to conduct sound
into the cement and formation. A gap of only a few thousandths of an inch
between the cement and casing may be detected by the log as no cement at all
between the casing and well bore. Any one of a number of processes, such as
temperature cycling, cement shrinkage, casing contraction, etc., can cause
gaps
to form as the cement sets. Such a gap or "microannulus° may not result
in any
operational problems. They are often of no practical consequence.
If a poor bond log is obtained, remedial cement squeeze operations are
generally performed until a satisfactory borrc! log is obtained aaoss
intervals
where a good cement seal is considered to be critical. Several squeeze
operati~s may be required. A substantial body of field evidence indicates that
these squeeze operations are often unnecessary and that the problem is the
bond log rather than the quality of the cement. That is to say, the cement
bond
log is taken to represcnt the extent of displacement of the drilling fluid by
the
CA 02406306 2002-10-16
cement slung during cement placement and it is overly sensitive to factors
that
do not relate to the quality of the cement.
Strategies have been developed to improve the ability of bond logs to
accurately represent the completeness of the drilling fluid displacement by
the
cement. For example, it is now common practice to run bond logs with the
casing pressurized. The expansion of the casing when pressurized at the
surface to pressures of the order of 1,000 to 2,000 pounds per square inch
(psi)
may close any microannulus that formed during cementing thereby providing an
improved bond log.
However, there are also instances where pressurizing ,the casing does not
close this gap. Excess pressure can enlarge a microannulus andlor form cracks
in the cement sheath that are detrimental to the seal provided thereby.
Expansive cement additives have been developed with the goal to improve bond
logs. They have had a more limited application than pressurizing the casing
and
they may actually reduce the strength of the cement. All of these methods are
time consuming and of uncertain reliability. The apparatus described here
improves the contact between the solid~ed cement and the casing, thereby
improving the accuracy of bond log evaluations.
SUMMARY OF THE INVENTION
It is the objective of this document to describe an apparatus to improve
cementing operations by applying pressure pulses to the fluid phases that
exist in
a well during a well cementing operation. The apparatus has been optimized to
be relatively simple to construct and operate and to be safe and reliable for
use
in the designated hazardous areas near a wellbore, during drilling and
cementing
operations. An important aspect of the present invention is the finding that,
when
Q
CA 02406306 2002-10-16
this apparatus applies pressure pulses at the surface of the well, the
resulting
oscillatory motion is efficiently transmitted long distances down the well,
where it
prevents the development of gel strength, or reverses the process of gelation,
if it
has already occurred earlier in the operation. A remarkable aspect of this
apparatus is that it accomplishes four objectives, at the same time.
1. It reduces or eliminates gas flow after cementing (type I gas migration).
2. It reduces or eliminates microannular gas (type II gas migration).
3. It measures the waiting-on-cement time.
4. It improves the accuracy and quality of bond logs.
The preferred embodiment of the apparatus is described in detail below.
BRIEF DESCRIPTION OF FIGURES AND APPENDECES
Figure 1 illustrates the effect of pressure pulses in wells.
Figure 2 is a photograph of a "commercial prototype air pulse generator"
(CPAPG). Externally, it consists of three parts, (1 ) a water flow control
section
attached to the left side of an enclosure, (2) an enclosure that houses
various
controls and (3) an air flow control section attached to the right side of the
enclosure. The enGosure is a stainless steel box about 10 inches high, 15
inches wide and 8.5 inches deep, inGuding the recessed lid. A carbon steel
mounting flange, welded to the back of the enclosure, increases the height and
width by one inch on each side. The total width as shown, including the
valves,
is about 24 inches and its total weight is about 50 pounds, depending on the
type
of connections used and if flow measurement devices are installed.
Figure 3 illustrates one type of well hookup.
Figure 4 is a schematic diagram of the major components in the enclosure.
The enclosure has a triple hinge that allows the lid to turn down as shown in
0
CA 02406306 2002-10-16
Figure 2. It also allows the front panel of the enclosure to pivot down on the
same hinge. This provides convenient access to the inside of the enGosure for
servicing. The aspect of the diagram in Figure 4 shows the location of the
components when this panel is so opened. That is, the triple line through the
middle of the drawing represents the location of the hinge. The items shown
below this line on the illustration represent the items on the front panel of
Figure
2. They are upside down, compared to Figure 2, but they show the correct
orientation when this panel is opened.
Figure 5 is a calibration curve for the rate of air delivery of an air
compressor.
Figure 6 shows the results of some field tests using this apparatus.
Attachment 1 is a parts list and shop drawings to construct this apparatus.
Attachment 2 is operating instructions for using this apparatus.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
Referring to Figure 2, the large valve on the right side of the enclosure is a
"Schrader" pilot operated air valve. The port on the top of the valve, in the
photograph, is attached to an air supply such as a portable air compressor or
a
rig air compressor. The static air pressure from the air supply is typically
100 psi,
but it may vary over a wide range from 10 psi, or less, to 1,000 psi, or more.
The
flow capability is typically 185 standard cubic feet per minute (cfm), but it
can
vary over a wide range, from 10 cfim, or less, to 1,000 cfm, or more. When the
valve is in the "open" position, compressed air is conducted through the valve
to
the coupling shown on the bottom of the valve in the photograph, which is
typically connected to the well head by a 2 inch diameter air hose as shown in
Figure 3. When this valve is "closed", air flow returns from the well through
this
hose and is exhausted to the atmosphere through the exhaust port of the valve.
~n .
CA 02406306 2002-10-16
During operation, water is typically added to the annulus to keep it full, or
nearly so, to compensate for the loss of fluid volume due to cement hydration
reactions, fluid loss from the well bore, etc. This is accomplished by the
water
flow control section. Typically, clean rig water at a static pressure of
approximately 35-50 psi is attached to the top of this valve with a hose. The
pressure can vary over a wide range from 5 psi, or less, to 200 psi, or more.
The
air and water pressures and filows, that are useful for the purposes of this
invention, extend over a large range, but the values for the preferred
embodiment
are limited by the ratings of the components.
The water flow into the well should be sufficient to maintain the annulus full
of
water, or at least to prevent the level from declining too much. During each
cycle, when the air pressure in the well is less than the water pressure, a
check
valve allows water to flow into the well. The flow of the water is controlled
by
manually adjusting a valve. The flow of compressed air from the Schrader valve
primarily controls the cement pulsation process. The volume of the water flow
does not significantly contribute to cement pulsations. When the well annulus
is
completely full of water, excess water will be exhausted to the atmosphere,
from
the Schrader valve exhaust, at the end of each exhaust cycle. This condition
provides the criterion that the annulus is full of water.
The logic and operation of this apparatus are illustrated in detail in Figure
4. In
the discussion that follows, the number in parenthesis after each major
component is mentioned for the first time ident~ies the item number on
Attachment 1. "I" denotes "Internal" and "E" denotes "External". In the
figure, the
pressure of the air flow into the well is sensed at port P1 by air switches
AS1 (1,27) and AS2(1,26) at ports P3 and P4, respectively. When the pressure
at
i~
CA 02406306 2002-10-16
P1 drops to the adjustable set point of AS1, for example, 5 psi, then AS1
sends a
pulse of compressed air to air valve AV(1,24) and switches its position to
connect
its air supply through the multiport ball valve BV1 (1,8), located on the
enGosure
panel, to ball valve BV2(E,8) located on the "dome" of the Schrader valve
SV(E,9), switching the Schrader valve. This causes the Schrader valve to
connect the "air supply" to the "to well° air hose and start the
process of
supplying compressed air to the annulus of the well. When the pressure at P1
increases to the adjustable set point of AS2, for example, 100 psi, then AS2
sends a pulse of compressed air to air valve AV and switches its position to
exhaust the air through the multiport ball valve BV1 to ball valve BV2 on the
dome of the Schrader valve SV, switching back the Schrader valve. This causes
the Schrader valve to divert the "to well" hose to the "exhaust". The
compressed
air in the annulus of the well is then exhausted until the pressure drops to
the set
point of AS1, repeating the cycle. This cycle repeats continuously until it is
stopped, or modified, by the operator.
The ball valve BV2 serves the purpose of a shut off, or emergency shut off,
valve. When it is rotated 180 degrees, the pressure is exhausted from the dome
of the Schrader valve and it immediately switches to the exhaust mode,
whatever
the status of the automatic air controller cycle. When the multiport ball
valve BV1
is rotated 90 degrees, the automatic air controller is bypassed and the
Schrader
valve can be cycled manually with ball valve BV2, in the unlikely event that
the
automatic air controller fails.
Pressure gauge G1(1,5) displays the pressure at port P1, and is useful to
monitor the air pressure pulses in the well. When ball valve BV3(1,6) is
rotated
180 degrees, G1 displays the air pressure available to operate the pneumatic
CA 02406306 2002-10-16
controls, generally the same as the maximum static supply air pressure at port
P2. When the Schrader valve opens to provide air to the annulus of the well,
the
pressure at port P2 might initially fall to such a low value that the
pneumatic
controls would not function property. Check valve CV1 (1,18) maintains the
pressure in the pneumatic control system until the pressure at port P2 builds
up
later in the pulse cycle. Sufficient air storage volume, for this purpose, is
provided by the components of the control system. The pressure of G1 also
allows the operator to monitor the condition of the filter F(1,20). In the
unlikely
event it becomes plugged, the pressure in that part of the system may drop to
less than that required to operate the pneumatic controls and require the
operator to change the filter.
A flushing arrangement was developed to flush out particulate materials that
might enter the system through the port P1. Whenever the air pressure at P1
drops to less than the water supply pressure at P5, a small quantity of water
flows through this system, exiting at P1. When the pressure at P1 is greater
than
the water pressure at P5, the check valve CV2(1,19) prevents back filow. The
flow rate is adjusted by a metering valve MV(1,9) and the pressure drop from
this
flow has no significant effect on the pressures between P1, P3 and P4. Before
this arrangement was developed, particulates from port P1 often rapidly
contaminated lines, plugged any filters used therein, and rendered the
automatic
air controller inoperable.
Water flow into the annulus of the well to make up for fluid loss was adjusted
by
a large ball valve BV4(E,3) connected to a large check valve CV3(E,2). This
combination allowed water to flow into the well annulus when the air pressure
fell
to less than the water pressure during each pressure pulse cycle. The flow of
CA 02406306 2002-10-16
water was supplied to the well with a separate hose, usually a 1 inch water
hose.
The pressure gauge G2(1,5) was used to monitor the pressure in the well, or
the
ball valve BV5(1,7) was rotated 180 degrees to monitor the water supply
pressure. On occasion, valve BV4 was momentarily closed so that the pressure
in the well could be monitored with G2 without any water flow, thus completely
obviating any pressure drop. Then BV4 was opened to resume water flow.
Attachment 1 is a complete parts list for a CPAPG, including detailed
construction drawings for the enclosure and two small air manifolds.
Attachment
2 is instructions for performing the cement pulsation operation.
It was found that as long as the annulus was full of water, the volume of air
injected into the annulus, or exhausted, during each cycle could be measured
and this volume could be used to monitor the cement setting process. When the
plug was bumped and cement pulsation was started, the volume of air per cycle
typically increased rapidly, within a few pulsation cycles, to a maximum
value. At
this point, it was often consistent with the volume expected if all of the
fluids in
the well were reciprocated by the cement pulsation process, but the volumes
varied aver a large range, even for offset wells in the same field. With time,
as
the cement made the transition from liquid slurry to a solid, the volume of
air
required per cycle decreased until a constant value was obtained. This
represented the compressibility of the noncementatious fluids, such as
drilling
fluid, cement preflushes and spacers, in the annulus. If the annulus was
cemented to the surface, cycle times typically became very short. The cycle
times could be easily and accurately measured with a stop watch and related to
volumes in at least finro ways.
CA 02406306 2002-10-16
One means, when a portable air compressor was used, was to calibrate the
pressurization cycle time of the compressor, connected to the CPAPG, and
delivering air into a large air tank. The volume of the tank was changed by
metering in known amounts of water. A compressor calibration curve is shown
on Figure 5. Another means was to attach a calibrated gate valve, operated as
a
choke, to the exhaust of the Schrader valve as referenced in Attachment 2. It
was concluded that it was not necessary to accurately measure the volumes to
determine the WOC time. It could be easily established by measuring the time
required until the cyGe time became (1 ) constant, without significant further
changes with time, and (2) about the magnitude that would be expected for
compressing the fluid phases in the well above the cement slurries. Figure 6
shows the results of volumes measured, for research purposes, for some very
similar East Texas gas wells with a depth of about 10,000 feet. Note the large
well-to-well variations, with no changes in any of the cementing or cement
pulsation parameters.
During the development of the CPAPG, described in detail in this document as
the preferred embodiment of this invention, a large number of alternatives
were
also developed and tested on wells. They included devices controlled
electrically, instead of the pneumatic device described here, devices with an
air
storage tank at the Schrader valve inlet and devices with a water storage tank
at
the outlet. The air storage tank allowed the rapid delivery of a pressure
spike at
the initiation of each cyGe. This was considered to be beneficial to initiate
breaking gel strength under some conditions. The water storage tank allowed
direct visualization of the water going into and out of the well and more
accurate
is
CA 02406306 2002-10-16
measurement of the displacement volumes. These alternatives significantly
increased the cost, size and complexity of the apparatus.
The foregoing descriptions may make other alternative arrangements of this
apparatus apparent to those of skill in the art. The aim of the appended
claims is
to cover aN such changes and modifications that fall within the true spirit
and
scope of the invention.
15
25
m
