Note: Descriptions are shown in the official language in which they were submitted.
~42~Z8
FIELD OF THE INVENTION
This invention generally pertains to logging while drilling
apparatus, systems and methods and more particularly pertains to systems,
apparatus, and methods utilizing mud pulsations for telemetry to transmit
signals representing one or more downhole parameters to the earth's surface.
BACKGROUND OF THE INVENTION
Many efforts have been made to develop successful logging while
drilling systems, as suggested by the following examples: Karcher, United
States Patent No. 2,096,279 proposes a system utilizing electrical conductors
inside the drill pipe. Heilhecker, United States Patent No. 3,825,078
proposes a system utilizing extendable loops of wire inside the drill pipe.
Silverman, United States Patent No. 2,354,887 proposes a system utilizing
inductive coupling of a coil or coils with the drill pipe near the drill
bit with measurement of the induced electrical potential at the earth's
~- surface. Arps, United States Patent No. 2,787,759 and Claycomb, United
States Patent No. 3,488,629 propose systems in which pulsed restrictions to
the drilling mud flow produce pressure pulse signals at the earth's surface.
Other related United States Patents are Nos. 3,186,222, 3,315,224, 3,408,561,
3,732,728, 3,737,845, 3,949,354 and 4,001,774.
,. ` .~;
~3- ~
l~.Z~228
Each of the abovementioned proposals has had some
drawback of sufficient consequence to prevent its commercial
- acceptance. For example, the inconvenience and time in-
volved for the large number of connections and disconnections
of electrical connectors is a significant drawback in
systems such as proposed by Karcher. Though an induced
electric potential system such as propo~ed by Silverman may
be considered operable for a short distance, the signal to
noise ratio of such a system prohibits its use as a practical
matter in deep wells.
When modern jet bit drilling became commonplace
and very large mud volumes and high mud pressuxes were
employed, the systems as proposed by Arps, proved to be
unreliable and subject to rapid deterioration. The intro-
duction of a controlled restriction into the very powerfulmud stream, of necessity, required large and powerful apparatus
and operation was unsatisfactory because of ~apid wear and
very high energy requirements.
The environment is ~ery hostile at the bottom of a
well during drilling. Drill bit and drill collar vibrations
may be in the order of 50 g. The temperature is frequently
as much as 400 F. The bottom hole pressure can be more
than 15,000 psi. The drilling ~luid ~lowing through the
drill collars and drill bit is highly abrasive. With present
drilling equipment including imp~oved drill bits, the con-
tinued drilling time with a particular bit can be in the
order of lO0 - 300 hours and sometimes longer before it
becomes necessary to change the drill bit. Accordingly, a
downhole formation condition sensing and signal txansmitting
--4--
~.Z~228
unit mounted near the drill bit must be capable of operating
unattended for long periods of time without adjustment and
with a continuing source of elect~ical power. Also, the
signal communication apparatus must be capable of trans-
mitting a continuing usable signal or signals to the earth'ssurface after each additional joint of drill pipe is conven~
tionally added to the drilling string as the drilled bore-
hole is increased in depth.
In general, systems using mud pulsations for
telemetry are considered the most practical since the drill-
ing operation is least disturbed. To date, however, the
reliability that has been achieved with such systems is not
satisfactory. The previous methods such as those of Arps
and Claycomb, utilize the insertion of a controlled restriction
into the mud flow circuit. ~owever, when the mud flow surpasses
600 gpm and pump pressures pass 3000 psi, controlling this
large energy by varying a restriction to produce telemetry
signals is complicated and requires powerful downhole machinery.
~ general objective of the present invention is to
pro~ide a successful logging while drilling system of the
type utilizing mud pulsations for telemetry to transmit
signals representing one or more downhole parameters to the
earth's surface.
More specifically, it is an objective of the
invention to provide such a system wherein the amount of
energy that is required to generate a strong pressure pulse
at a tool near the drill bit is significantly reduced.
Another objecti~e o~ the invention is the utili-
zation of an existing large source of energy for the pro-
duction of the mud pulsations.
4228
According to one broad aspect of the invention, thereis provided for use in a system for conducting drilling oper-
ations employing a string of drill pipe extending from the
earth's surface having a drilling means such as a drill bit,
hydraulic drill motor, or the like at the lower end, a pump by
which drilling fluid is forced downwardly through the drill
string interior and drilling means to flow back to the surface
through the well annulus, the drilling means imposing a restric-
tion to the drilling fluid flow forming a high pressure zone in
the interior of the drill string and low pressure zone in the
well annulus, a telemetering system comprising, a drilling
fluid bypass above the drilling means providing fluid communi-
cation between the interior of the drill string and the well
annulus, the bypass being defined in part by a valve seat, a
valve stem moveable to and away from said valve seat forming
a valve to close and open said bypass, a means for detecting
the magnitude of a downhole parameter and for producing an
electrical signal representing said magnitude, and electro-
magnetic solenoid means responsive to said electrical signal to
rapidly operate said valve to generate pressure pulses in the
drilling fluid, and means at the earth's surface to detect such
pressure pulses and to provide a measure of the magnitude of
said parameter.
According to another broad aspect of the invention,
there is provided for use in a system for conducting drilling
operations employing a string of drill pipe extending from the
earth's surface having a drilling means such as a drill bit,
hydraulic drill motor, or the like at the lower end, a pump by
which drilling fluid is forced downwardly through the drill
string interior and dri.lling means to flow back to the surface
through the well annulus, the drilling means imposing a restric-
tion to the drilling fluid flow forming high pressure zone in
~-6-
4;~2~3
the interior of the drill string and low pressure zone in the
well annulus, a telemetering system comprising, a drilling
fluid bypass above the drilling means providing fluid communi-
cation between the interior of the drill string and the well
annulus, the bypass having an electrically energizable valve
therein capable of rapid operation to open or close said bypass,
a means for detecting the magnitude of a downhole parameter
and for producing an electrical signal representing said magni-
tude, an electrical energy source, means responsive to said
signal for supplying a relatively large amount of electrical
power to initiate opening said valve, and substantially less
power when the valve is open or closed, to generate pressure
pulses in the drilling fluid, and means at the earth's surface
to detect such pressure pulses, and to provide a measure of the
magnitude of said parameter.
According to a further broad aspect of the invention,
there is provided for use in a system for conducting drilling
operations employing a string of drill pipe extending from the
earth's surface having a drilling means such as a drill bit,
hydraulic drill motor, or the like at the lower end, a pump by
which drilling fluid is forced downwardly through the drill
string interior and drilling means to flow back to the surface
through the well annulus, the drilling means imposing a restric-
tion to the drilling fluid flow forming a high pressure zone in
the interior of the drill string and low pressure zone in the
well annulus, a telemetering system comprising, a drilling fluid
bypass above the drilling means providing fluid communication
between the interior of the drill string and the well annulus,
the bypass being defined in part by a valve seat, a valve stem
moveabl~ to and away from said valve seat forming a valve cap-
able of rapid operation to close or open said bypass, a means
for detecting the magnitude of a downhole parameter and for
~ 6a-
~.Z~2;~3
producing an electrical signal representing said magnitude, a
cylinder in communication with the bypass having a compensating
piston therein connected to said valve stem so that fluid pres-
sure exerts a first hydraulic force on the compensating piston
in the direction corresponding to the opening of the valve and
fluid pressure exerts a second hydraulic force on the valve
stem in the direction corresponding to the closing of the valve,
the net hydraulic force on the valve stem being proportional to
the difference between said first force and said second force,
and means responsive to said electrical signal to rapidly move
said valve stem to generate pressure pulses in the drilling
fluid, and means at the earth's surface to detect such pressure
pulses and to provide a measure of the magnitude of said para- -
meter.
The invention will now be described in greater detail
with reference to the accompanying drawings.
-6b-
~L~.2~ZZ~3
BRIEF DESCRIPTION OF THE DRAWINGS
.
Fig. 1 is a schematic illustration of a conventional
rotary drilling rig showing apparatus of the present invention
incorporated therein.
Fig. 2A is a schematic illustration of a negative
mud pressure pulse generator with its valve in the open
position.
Fig. 2B is a schematic illustration of the negative
mud pressure pulse generator of Fig. 2A, with its valve in
the closed position.
Fig. 3~ is a schematic illustration of a physical
embodiment of the negative mud pressure pulse generator of
Figs. 2A and 2B, together with instrumentation and sensor
sections in place in a drill string near the drill bit.
Fig. 3B is a drawing of the negative mud pressure
pulse generator of Figs. 2A and 2B taken in proportional di-
mensions from an engineering assembly drawing used in actual
manufacture of the device.
Fig. 3C is a schematic diagram of a radioactivity
-20 type sensor and associated instrumentation.
Fig. 3D is a schematic diagram of a temperature
type sensor and associated instrumentation.
Fig. 3E is a schematic diagram of typical instru-
mentation for controlling actuation of the valve of the negative
mud pressure pulse generator.
Fig. 3F is a schematic illustration of one type of
self-contained downhole power source that may be utilized.
Fig. 3G is a schematic illustration of another type
of self-contained downhole power source that may be utilized.
~.24228
Fig. 4 is a schematic illustration showing typical
aboveground equipment in accordance with a preferred embodi-
ment of the invention, wherein the downhole parameter being
sensed is radioactivity.
Fig. 5 is a graphic illustration, in idealized form,
showing certain wave forms and pulses and time relationships
to aid in explanation of the signal extractor portion 102 of
Fig. 4.
Fig. 6 is a schematic block diagram showing component
105 of the signal extractor 102 of Fig. 4 in further detail.
Fig. 7 is a schematic block diagram showing component
107 of the signal extractor 102 of Fig. 4 in further detail.
Fig. 8 is a schematic block diagram showing another
form of aboveground equipment that may be utilized.
Fig. ~ is a schematic block diagram showing still
another form of aboveground equipment that may be utilized.
~ig. 10 is a schematic block diagram showing an alter-
nate timing pulse generator that may be utilized.
Fig. ll is a schematic block diagram showing still
another form of aboveground equipment that may be utilized.
iL~.24228
DETAILED DESCRIPTION OF INVENTION
Before proceeding with description of preEerred
embodiments of the invention, it is believed that understand-
ing will be enhanced by discussion of some basic factors.
In a 10,000' length of 4 1/2" drill pipe, the mud
volume inside the pipe is of the order of 5,000 gallons.
Assuming that the bulk elastic modulus for compressed drill-
ing mud is 400,000, then discharging .5 gallons of fluid
will cause a pressure drop of 40 psi, (if we consider the
5,000 gallons as being in a simple tank). It can be assumed,
therefore, that discharging mud near the bottom of such
drill pipe at the rate of 0.125 gallons/sec. will cause a
signal of 10 psi/sec. at the surface. We shall refer to
the rate of change of pressure as the ~ index and in this
case the ~ index is equal to 10.
Three important experiments were performed;
1. Measurements were made in a test well at
1,800' and moderate differential pressures
of 1,000 psi across a valve at the bottom.
2. Measurements were made in an oil field
drilling well at 8,000' and low differential
pressures of 400 psi.
3. Measurements were made in a second oil
field drilling well at 5,000' and high dif-
ferential pressures (1,600 psi~.
All three series of experiments indicated that
the ~ index of the pressure pulse received at the surface
when the valve is suddenly opened was substantially higher
than calculated. The reasons for this are: (a) highly com-
pressed drilling mud may have an
~.2 ~2 Z ~
elastic modulus somewhat higher than 400,00Q; (b) there is
some wave guide acti~n by the d~ill pipe that causes the
signal to travel much more ~avorably tha~ it would in a
large tank of the same volu~le; and (c~ the sudden opening
of a valve at the bott~m o~ the well causes a h~ghe~
index than in the case o~ the large tank because of the
elasticity of the mud column above it.
In a typical 15,000 foot drill string (with the
bottom end closed off), if a marker were placed at the top
of the mud column, this marker would drop some 110 feet when
3,000 psi mud pump pressure is applied (3,0QO ~si is a
rather typical mud pump pressure in deep wells~. One can,
therefore, consider the mud column as being continually
compressed by some 100 feet and acting ~s a long spring in
which a large amount of potential energy is sto~ed. When
a valve at the bottom of the drill pipe is suddenly opened,
this potential energy is released, causing a large negative
mud pressure pulse; such mud pressure pulse being substan-
tially larger than would be the case if the mud were incom-
pressible.
In the experiments conducted at 5,00Q' in a drillingwell, a small passageway (.056 in.2 area2 between the inside
of the drill collar and the annulus, was opened and shut
in accordance with a controlled sequence. The pressure
across the valve was 1,600 psi and the discharge was
calculated to be approximately .25 gallons/sec. The
volume of mud inside the drill pipe was approximately 2,500
gallons and assuming an elastic modulus for the mud of
400,000', the pressure drop was calculated to be 40 psi/sec.
.. . . .
1~.24ZZB
(again using the as~umption that mud column was a simple
tank). In the tests the pressure drop ~t the surface was
measured to be over 100 psi/sec. or considerably more than
would be expected from the simple tank calculation. The
following conclusion was ~eached: With high pressures exist-
ing across the drill bit (1,000 psi or more2, large sharp
signals can be developed at the surface by opening and
closing a very small b~pass valve at the sub-surface near
the drill bit. Valves having an opening of .05 in.2 can
produce large signal~ from a 5,QQQ' depth and the reduction
in signal magnitude from depths between 2,50a' and 5,0Q0'
have been found to be very small; thus, indicating that the
signal attenuation is small.
The system of the present inventi~n has a number
of important advantages: The rapid discha~ge ~t a rate of
, as little as 0.125 gallons/sec will generate a "sharp"
pulse, that is a pulse containing a high rate of change of
pressure, i.e., a high ~ index (e.g. 40). ~urthermore, the
rapid opening of the bypass valve will also minimi~e wear
for the following reasons: When the bypass valve is closed,
there obviously is no wear on the valve seat. ~hen the valve
is open (and the valve area is large co~pared to a restriction
or restrictions followin~ it~, the valve will be exposed
to low velocity fluid and, consequently, the we~r will be
mostly in the following restrictiDn or restrictions which
can be made expendable and of very non-errodable material
such as boron carbide. Wear occurs in the bypass valve
only when it is in the process of opening or closing, i.e.,
is "cracked" and the velocity through the valve seat is
then very high. The valve operation should, therefore, be
~.2~228
as fast as possible for openin~ and closing and there is
no limit to the desirable speed. The rate of discharge
through the valve should ~lso be ~ast but there is an upper
limit beyond which f~ster discharge does not benefit.
The reason for this is the limit to high frequency trans-
mission through the mud. ~requencies higher than about 100
Hz are strongly attenuated and are of little value in build-
ing up a fast pulse at the surface. To determine the maxi-
mum useful rate of discharge, it was necessary to set up
experiments on a full scale using real drilling oil wells
and long lengths of conventional drill pipe. The experimental
arrangements comprised a special large ~al~e followed by
an adjustable orifice.
Changing the orifice size can deter~ine the flow
rate in gallons per second. It was determined that flows
larger than about .3 gallons per second produced little
improvement in the signal. In comparing the signals from
a depth of 5,012 feet, three different orifice sizes were
tested, .509" diameter, .427" diameter and .268" diameter.
It was determined that the .2~8" diameter orifice generated
a signal at the surface nearly as intense as the one gene-
rated by the .509" diameter orifice.
-12-
~L~.2~228
Referring now to Fig. 1, there is schematically illus-
trated a typical drilling rig 10 including a mud circulating pump
12 connected to a discharge pipe 14, a standpipe 16, a high
pressure flexible rotary hose 18, a swivel 20 and a drilling
string 22, comprising the usual drill pipe and d~ill collars,
and a jet type bit 26. A short distance above the bit 26,
and mounted within drill collar 24, is a negative mud pressure
pulse generator 28 and a sensing and instrumentation unit
30.
The negative mud pressure pulse generator 28 is of
a special design. It generates a series of programmed pulses
and, each pulse consists of a short momentary reduction in
mud pressure. In one embodiment, this is accomplished by
means including a valve that momentarily opens a passageway
between the inside and the outside o~ the drill collar 24,
i.e., the valve controls a passageway between the inside of
the drill collar 24 and the annulus 29 formed by the ~utside
of the drill collar and the well bore.
Aboveground equipment, generally designated as 32,
is connected to a pressure transducer 100, which in turn is
connected to standpipe 16. Alternatively, the transducer
lO0 could be connected into the stationary portion of swivel
20, if desired.
Figs. 2A and 2B show the negative mud pressure
pulse generator 28 in diagramatic form to facilitate ex-
planation of its function and manner of operation. The
negative mud pressure pulse generator comprises a valve
-13-
~ .2~22~3
inlet chamber 42, a valve outlet chamber 44, and a compen-
sator chamber 72. The valve inlet chamber 42 is hydraulically
connected via an inlet passageway 38 to the inside of the
drill collar 24. The valve inlet chamber 42 is also hy-
draulically connected via a passageway 48 to the valve outletchamber 44. Hydraulic flow through passageway 48 is con-
trolled by the cooperation of a valve 36 with its seat 37.
The valve outlet chamber 44 is hydraulically connected via
an outlet passageway 51 to the annulus 29. Interposed
in the outlet passageway 51 are first and second compensator
orifices 52, 53. The chamber 40 between the orifices 52,
53 is hydraulically connected via a conduit 74 to the com-
pensator chamber 72. The inlet chamber 42 communicates with
compensator chamber 72 via a cylinder 49, which receives a
compensating piston 50 that is connected to the valve 36 by
means o~ a shaft 46. The valve 36 is also connected, by
means of a shaft 47 (see Figs. 3A and 3B) to an actuator
device 54.
The function and operation of the negative mud
pressure pulse generator 28 will now be explained. Fig.
2B shows the valve 36 of the ne~ative mud pressure pulse
generator 28 in the "closed" condition. In this figure,
the striated part indicates "high" pressure and the blank
part indîcates "low" pressure. (Pressure magnitudes, such
as "high", "low" and "intermediate" are relative pressures,
i.e., the ~ erence between the pressure at a given location
and the annulus pressure which is here considered to be zero;
the actual or real pressure would be equal to these magni~
tudes plus the hydrostatic head, which may be 10,000 psi or
higher.)
-14-
~z~zz~
The effective area of the valve 36 is made some-
what larger than the effective area of the piston 50 on the
shaft side and, consequently, when the valve 36 is closed or
nearly closed, the force on the shaft 46 is in the direction
shown by the arrow in Fig. 2B and may be equal to about
1,000 X (a - a') where a is the effective area of the valve
36 and a' is the effective area of the compensating piston
50 on the shaft side.
Fig. 2A shows the valve 36 in the "open" condition,
i.e., permitting mud flow from valve inlet chamber 42 to
valve outlet chamber 44 and via outlet passageway 51 to the
annulus 2~. The first and second compensator orifices 52
and 53 each provide a predetermined restriction to the mud
flow and each causes a pressure drop. Consequently, the
pressure inside the cha~ber 72 can be made to have any value
between the maximum pressure inside chamber 44 and the
minimum value at the exit of outlet passageway 51 which
corresponds to the pressure inside the annulus 29.
As is pointed out above, in ~ig. 2~, as in Fig. 2B, the
striated part indicates "high" pressure and the blank part at the
exit of outlet passageway 51 is "low" pressure. During the valve
"open" flow condition, the mud encounters two restrictions
to flbw: orifice 52 and orifice 53, as a consequence of
which, the pressure in the chamber 40 is intermediate between
the "high" pressure indicated by the striated section and
the "low" pressure at the exit of outlet passageway 51.
This "intermediate" pressure is indicated by the stippled
area in Fig. 2A. This "intermediate" pressure is originated
in the chamber 40 b`etween orifices 52 and 53 and communicates
-15-
42Z8
via conduit 74 to the compensator chamber 72. The pressure
in this compensator chamber 72 can, consequently, be adjusted
to any reasonable value between the "high" pressure in valve
outlet chamber 44 and the "low" pressure at the exit of
outlet passageway 51. Ihe proportioning of the sizes of the
orifices 52 and 53, therefore, controls the pressure in compen-
sator chamber 72 and, consequently, the force on compensator
piston 50. If the orifice 53 were the same size as orifice
52, then the pressure in chamber 40 (and compensator chamber
72) would be about midway between that of valve outlet
cham~er 44 and the annulus 29. As the size of orifice 53
is made larger than that of orifice 52, the pressure in
compensator chamber 72 will be relatively decreased, and
as the size of orifice 53 is made smaller than that of
orifice 52, the pressure in compensator chamber 72 will be
relatively increased. ~or example, if orifice 53 is made
small compared to orifice 52, the pressure in compensator
chamber 72 will be high and, therefore, the force on the
head of piston 50 wil] be high and tend to close the valve
36. On the other hand, if orifice 53 is large compared to
orifice 52, the pressure in chamber 72 will be low, thus,
tending to allow the valve 36 to remain open. It is seen,
therefore, that the force on the head of piston 50 can be
adjusted between wide limits, thus, providing a means for
adjusting the action of the valve 36.
It is important to note that the force tending to
close the valve 36 in Fig. 2B, and the force tending to open
the valve 36 in Fig. 2A, are determined by first and second
independent parameters, i.e., the force tending to close
.
-16-
~24Z28
the valve is derived from the effective area differences of
the valve 36 and the rod si~e of compensator piston 50;
whereas the force tending to open the valve is derived from
the relative sizes of the orifices 52 and 53. By suitably
ad~justing these parameters, the valve 36 can be made to open
or close by the application of a small external mechanical
force.
It is also important to note that the valve 36 has
a "bi-stable" action, i.e., the valve is "flipped" or "toggled"
from the "open" to the "closed" position or vice versa. In
other words, the first said independent parameter is chosen
so that when the valve is within the region of nearly closed
to fully closed, a predominant force of predetermined ma~nitude
in the valve "close" direction is applied and maintained;
and the second said independent parameter is chosen so that
when the valve is within the ~egion of nearly open to fully
open, a predominant force of predetermined magnitude in the
valve "open" direction is applied and maintained.
Thus, it is apparent that the negative mud pres-
sure pulse generator 28 of the present invention utilizes
existing energy derived from the mud pressure in such a
manner so as to greatly reduce the amount of external
energy required to operate the valve 36 and, in addition, to
impart to the valve 36 a "bi-stable" or "tog~le" action.
Further discussion of the negative mud pressure
pulse generator 28 will be facilitated by reference to Figs.
3A and 3B, which will now be described. Fig. 3~ illustrates
in schematic form a physical embodiment of ~he negative mud
-17-
$~.Z~Z28
pressure pulse generator 28 and associated downhole equip-
ment as it would be installed in the drilling apparatus o~
Fig. 1. The reference numerals that are applied in Figs. 1,
2A and 2B refer to corresponding parts when applied to ~ig.
3A. In Fig. 3~, a sub 58, which is typically 6 3/4" O.D.
and 3' long, supports an inner housing 56 by means of arms,
or perforated or slotted support members Cnot shown~. The
inner housing 56 contains the negati~e mud pressure pulse
generator 28 and carries at its lower end portion instrumentation
sections 62, 66 and sensor section 64. The mud from inside
the drill collar 24 passes around ~he housing 56 in the
direction of the ar~ows. ~ filter 60 prevents mud solids
from entering the housing. The valve 36 is shown to be
operated by an actuating device 54. ~hen the valve 36 is
open, as shown in Fig. 2~, some mud is bypassed into the
annulus 2~. The bent arrows show the direction of this
bypassed mud. The pressure that ~orces the mud into the
annulus 29 is the pressure across the jets of bit 26. ~hen
valve 36 is closed, the bypass to the annulus 29 is closed.
2Q The floating piston 76 separates chambe~ 72 from
an oil filled chamber 78. Actuating device 54 is mounted
within an oil ~illed chamber 80. ~n e~ualizing passageway
82, connects chamber 78 with chambex 80. Thus, in cooperation
with floating piston 76 and passageway 74, the chambers 72,
78 and 80 are maintained at essentially the same p~essure as
the chamber 40. Passageway 82 is partially shown in dashed
lines in Fig. 3A and is not sho~n in Fig. 3B since it is
located in adifferent plane from the cross section shown.
-18-
.. .. _, . ~ ..... . .
~.24228
Numeral 68 ~ep~esents a standard drill collar and
numeral 69 a box-box sub. Section 66 is 2 3/8" in diameter
and fits into a standard 15' 6 3/4" ~.D. - 3 1/4" I.D. drill
collar. The unit 30 is provided ~ith special centralize~
arms 70 which fit snu~ly into box-box sub 6q. The centralizer
arms 70 are designed to centralize the unit 3Q while allow-
ing free passage o ~ud.
Fig. 3B bears the corresponding reference numerals
of Figs. 2A, 2B and 3A and shows the negative mud pressure
pulse generator 28 in sufficient proportion and detail to
illustrate to one skilled in the art its actual construc-
tion. It may be noted that in ~ig. 3B the actuating device
54 comprises a pair of electrical solenoids arranged in
opposition. The winding 55 of the upper solenoid is dis-
posed to exert a force in the upward direction on its arma-
ture 57, while the winding 59 of the lowex solenoid is dis-
posed to exert a force in the downward direction on its
armature 61. The armatures 57, 61 are loosely coupled to a
mechanical linkage 63 that is fixed to the shaft 47 so that
a "hammer" effect is achieved; i.e., when a solenoid winding
is energized, its armature moves a short distance before
picking up the load of shaft 47 ~ith a hammer like impact.
This "hamme~" action has a beneficial effect upon the opening
and closing operations of the valve 36. Suitable solenoids
for this application a~e the Size 6EC, medium stroke, conical
face, type manufactured by Ledex, Inc., of Dayton, Ohio.
Reverting n~w to discussion of the negative mud
pressure pulse generator 28, there are several further
factors and features that should be considered.
-19-
~.Z 4~ Z ~
The orifices 52, 53 are made to have smaller
opening areas than that of the passa~eway 48, so that the
velocity of mud flow over the se~ling surfaces of valve 36
and its seat 37 is significantly reduced when compared to
the velocity of mud flow through the orifices 52, 53; thus,
concentrating ~ear on the orifices ~2, 53, which are made of
wear resistant material (such as boron carbide~ and which
are also made readily replaceable in the "field", as indicated
in Fig. 3B. These small non-erodable orifices 52, 53 make
the ne~ative mud pressure pulse ~enerator 28 completely
"fail safe", i.e., no matter ~hat happens to the operation
of valve 36 (such as being stuck in the open position~ the
amount of mud that is allowed to flow through the orifices
52, 53 would have no significant adverse effects on the
drilling. A further advantage of making the orifices 52, 53
readily replaceable in the "field" is that they can be
charged to best suit varying weights and viscosities of mud.
Because the negative mud pressure pulse generator
28 is exposed to severe vibration forces, the design must
provide for stability of the valve 36 in both its open or
closed position. The requisite stability is provided by the
"hydraulic detent" or "bi-stable" action of the valve 36
which was previously herein described.
The vertical acceleration encountered in drilling
is more severe in the upward than in the downward direction.
When the teeth of drill bit 26 encounter a hard rock, the
drill bit and drill collars 24 are fvrced upwards, i.e.,
accelerated in the upward direction; but once the drill bit
is raised upward and out of contact with the rock, there is
-20-
~.2~228
little force other than the acceleration due to gravity tha~
forces the drill bit and drill collars downwardly. Conse-
quently, the acceleration upward can be several hundred g's
but the acceleration downward is only of the order of 1 g.
The valve 36, therefore, must be designed so that when in
the closed position, high upwards acceleration tends to keep
it closed, i.e., makes it seat better, and high downward
acceleration (assumed to be small~ tends to open the valve.
This has been accomplished in the design, as can be seen
from Figs. 3A and 3B.
I determined, by conducting various tests and ex-
periments, that a force of appro~imately 34 pounds would be
required to actuate the valve 36 when the first and second
independent parameters hereinbefore described had been chosen
to provide appropriate "hydraulic detent" or "bi-stable"
action to achieve adequate stability for the valve 36.
With good engineering safety factors added, the required
force became 70-100 pounds. The application of force of
this magnitude over the required distance of valve travel,
20 with electromagnetic drive solenoids of reasonable size,
would require about 350 watts of electric power; i.e., nearly
1/2 horsepower. With such a large power requirement it
would appear at first glance that the energy needed for the
number of actuations of the valve 36 that would be necessary
for successful operation would be far beyond the capacity
of any available self-contained downhole power source.
This apparent energy problem is overcome, however, when it
is considered that the negative mud pressure pulse generator
28 of the present invention provides a very rapid action for
30 the valve 36; i.e., the valve 36 can be made to open (or
-21-
~.2~28
to close) with the application of the required 350 watts for
only about 20 milliseconds. The amount o~ energy required
to open (or close) the valve is, therefore,
350 - 20 _ ~ .002 watt hours
1000 60 60
There are available modern high density batteries of reason-
able size and capable of being included in the space pro-
vided withln the drill collar 24 and which can easily pro-
vide 2,000 watt hours of energy. Therefore (even without
recharging, as is described later herein~ ~ reasonable
battery can provide enough energy to operate the valve 36
about one million times. Assuming that the valve is operated
once every four seconds, a single battery charge is able to
operate the valve continuously for over one month. It is an
important requirement in logging while drilling that the
downhole apparatus be capable of operation unattended (i~e.,
without battery recharge~ for at lease the length of time
- between "round trips", i.e., the time that a single bit can
drill without replacement, the best bits last only about
100-300 hours and, therefore, the 30 day figure above is
more than adequate.
The practical design o~ the negative mud pressure
pulse generator 28 is a complex matter. In my experience,
although careful calculations were made using much of modern
hydrodynamic theory, in the final stages, many o~ the para-
meters had to be determined by empirical methods. ~n impor-
tant reason for this is because the "viscosity" of drilling
mud is thixotropic and the dynamic behavior i$ quite dif-
ferent from that of liquids having classical or so called
~.Z ~2 Z ~
Newtonian viscosity. Dxilling mud "weight" (grams per cc)
and "viscosity" vary over wide ranges and consideration
must be given to the fact that "weight" ùsually varies
over a much smaller range than "viscosity". Drilling mud
usually contains not only colloidal particles in sUspension
but also larger grains of sand and other particles.
An experimental set up was designed to determine
the minimum size of the discharge orifice (which controls the
rate of fluid discharge into the annulus~. In this set up,
a large "servo" valve ~1" diameter~ ~as followed by smaller
replaceable orifices. In 8,0QQ' and 5,0~0' well depth ex-
periments careful measurements were made of the magnitude
of the negative mud pre~sure pul$e at the surface as a
function of the size o the discharge orifice. ~s this size
was successively reduced, the magnitude of the pulse at
the surface seemed almost independent of the size o~ the
orifice until the surprisingly small .05 in.2 orifice area
was reached,at which time a slight reduction in pulse mag-
nitude was observed. This phenomenon was quite unexpected,
but was later understood after careful consi.deration of the
elastic properties of the mud column and the stored poten-
tial energy therein as was hereinabove explained. This
discovery produced the realization that a s~all negative mud
pressure pulse generatol- could produce useful si~nal$ at
the surface. Calculations were thereafter made and it was
determined that the "servo" principle for the valve actuation
was not necessary and the "servo" valve approach ~as ~ban-
doned. The direct, very fast acting, ne~ative mud pressure
pulse generator of this invention was thereupon designed and
has proved to be successful.
-23-
~.Z4~28
In a negative mud pressure pulse generator 28 of
practical design the ~ollowing dimensions may be considered
as typical: orifice 52, 0.500" in diameter; orifice 53,
0.306" in diameter; stroke of valve 36, 0.125"; diameter of
piston 50, 0.383"; diamete~ of ~alve 36 at its seating
surface, 0.430"; angle of seat 37 relative to axis of ~ralve
movement, 60; diameter of openin~; at seat 37 or passage-
way 48, 0.375"; diameter of valve shaft 46, 47, 0.187".
In Fig. 3~ there is schematically illustrated a
special type o battery that is well adapted to powering the
downhole equipment of the present invention.
Deep oil wells frequently have high bottom hole
temperatures 300-400F and many electric batteries cannot
operate at this temperature. There is, however, an ex-
ception; the modern molten salt batteries. They operate
well at high temperatures of 400-500C or even higher but
will not operate properly at lower temperatures principally
because the electrolyte solidifies and ceases to conduct
electrically. A litl ium aluminum iron sulphide molten salt
battery is manufactured by the Eagle Pitcher Co., Joplin
Missouri. Other manufacturers also manufacture high energy
molten salt batteries that are especially intended for
electric vehicle use. These batteries are very well adapted
for high temperature operation.
As illustrated in Fig. 3F, I provide an arrangement
that will "start up" the battery before it is immersed into
the hot environment of the oil well and will maintain it
charged when in use. In Fig. 3F, reference numeral 155
designates the battery proper; reference numeral 156 designates
-24-
~!.Z4~28
heating elements that are arranged to provide a small amount of heating to
the battery 155; and reference numeral 157 designates a jacket containing
thermal insulation, as for example, a material known as "Super Insulation"*
manufactured by the Union Carbide Co., New York, N.Y. or "Multifoil"*,
manufactured by The Thermo Electron Co., Waltham, ~lass. Initially an
external voltage (not shown) is applied to the terminal 158 (while the
instrument is at the surface and before immersion into the well). This
voltage activates the heating elements 156 and the battery electrolyte melts.
Furthermore, the battery lS5 is charged by the voltage applied at 158 before
the instrument is immersed in the oil well. When the battery 155 is in its
normal operating temperature range, the circuit to the heating element 156
is opened by the thermostatic switch 159, which closes during periods when
additional heat to the battery 155 is required. When logging while drilling,
the vibration of the tool will cause the device 160 to generate a charging
current. The device 160 is described in United States Patent 3,970,877,
Russell, et al. Instead of the Russell, et al, device, a small mud flow
powered turbine and electric generator could be used to maintain the battery
charged, since only about 1 watt of continuous charging power is required.
In Figure 3G there is schematically shown another special type of
battery that may be used to power the downhole equipment of the present
invention. This battery preferably uses cells of the Lithium Sulphur type,
such as are manufactured by Power Conversion Inc., of Mt. Vernon, New
*Trade Mark -25-
~Z42Z8
York. It may also use LeClanche type cells or Lead Acid
type cells. ~11 such cells, if exposed to high temperatures
(such as those normally encountered in deep earth boreholes)
would develop high internal pressure, so that the cells
would tend to explode. In one aspect of the present inven-
tion, there is provided an arrangement (.illustrated by Fig.
3G) by which this problem is overcome. In Fig. 3~, a plurality
of individual cells-161 such as one of the above mentioned
types are connected in series between a ground terminal 162
and a positive terminal 163. Each cell preferably is provided
a conventional pressu~e release cap or.vent 164. In acco~-
dance with the invention, the cells 161 are placed in a
container or reservoir 165 which is capable of withstanding
pressures exceeding those that could be developed by the
electrolyte of the cells 161. Within the ~eservoi~ 165
there is placed a liquid 166 having the same or similar
pressure-temperature characteristics as the electrolyte,
i.e., the liquid 166 will produce vapor.pressure (when
exposed to elevated temperatures~ that is substantially
equal to the vapor pressure of the electrolyte in the cells
161. In the simple case of the LeClanche type o~ Lead Acid
type cell, the liquid 166 can be water since the container
165 is hermetic and pressure resistant, the liquid 166 cin
- this example, water) will never boil - no matter how high
25 the temperature. It will simply build up vapor pressure in
the space above the liquid 166 high enough to be in equili-
brium with the vapor pressure generated by the hot liquid
166.
The same principle can be used when the cells are
of the Lithium Sulphur type, the liquid 166 could be Sulphur
-26-
-
~l~.Z4ZZ8
Dioxide. The Sulphur Dioxide v~por generated by the cells
161 will always be in pressure equilibrium with the container
165 because the Sulphur-Dioxide liquid in this auxilliary
container 165 will always generate pressures equal to those
generated by the cells 161.
Sulphur Dioxide and water, given as examples
above, are often unsatisactory ~a~ because Sulphur Dioxide
is highly corrosive and because water ic an electric conductor
and can short out the batteries. ~n alternative substance
is dichlorodifluoromethane, popularly called F~eon and manu-
facuted by E. I. DuPont and Cs., Wilmington, Delaware. ~any
types of Freons ha~e been developed ~ith almost an unlimited
number of thermodynamic properties, i.e., pressure-tempera-
ture relations. Othe~ substances can readily be found, such
as hydrocarbon vapors, propane or butane or mixtures of
'- vapors and gases. $uffice it to say, that I enclose the
battery cells 161 in a containe~ 165 and place in this
container a small quantity of a substance having similar
pressure-temperature ~elations to that of the electrolyte in
the battery cells 161. In Figs. 3F and 3G, I show only a
small number of cells connected in series. In actuality, a
larger number is normally employed. In the manufactured
instrument of Fig 3G, I employ 17 Power Conversion Co.
~ Lithium Sulphur cells.
Another important feature of the present invention
is that the length of time the ~alve 36 is maintained "Dpen"
has no relation to the amount of energy required. The only
energy required is that expended to actuate the valve 36 to
the "open7' position. The importance of this feature is
fully appreciated from the following consideration:
~.Z42;28
- It has been determined by experiment that in order
to provide a strong signal from a depth vf 10,000 t~ 20,000,
the valve must remain "open" ~or about 1/2 to one second and
any electromechani.cal Csolenoid or other2 device operating
for this length of time would not only requi~e la~ge amounts
of energy but would o~erheat and under ~ell conditions
probably burn up from its self generated heat.
As is hereinabove pointed out, two typical sensors
are disclosed as examples of the types that can be employed
in the operation of the present invention. F~g. 3C illus-
trates a natural gamma ray sensor and its associated cir-
cuitry which in this example is o the analog type. ~ig. 3D
illustrates a temperature sensor which in this example is of
the digital type. Either one of these sensors can be connected
to the input terminal of the instrumentation illustrated by
~ig. 3E which will be hereinafter described.
With reference to Fig. 3C, a geiger counter 168 is
provided with the conventional high voltage supply ~HV. The
geiger counter 168 generates pulses and is connected through
a capacitor 169 to amplifier 171 which generates pulses at
its output that correspond to those of the geiger counter
168. A scale of 1024 circuit 172 generates one output pulse
for each 1024 geiger counter pulses and its output is shown
as pulses having a separation tl. The higher the gamma ray
intensity, the higher will be the frequency of the pulses at
the output of the scale of 1024 circuit 172 and the smaller
will be the time tl.
Fig. 3D illustrates the case of the temperature
sensor. The temperature is sensed by a thermistor 173, i.e.,
-28
ZZ8
a semiconductor that varies in resistance with temperature
(it is provided with a suitable power supply, not shown~ and
it is assumed that the output of the thermister 173 is a DC
voltage proportioned to temperature. The a~lplifier 174
amplifies this DC voltage and impresses it on an analog-to-
digital convertor 175 which in turn generates a series of
binary bytes, one after the other, each representative of a
number proportional to the sensed temperature. The outputs
of the power amplifiers 185, 186 are utilized to control
energization of the windings of the "back-to-back" coupled
solenoids (hereinabove described~ to actuate the valve 36.
When winding 55 is energized the solenoid armature 57 (see
Fig. 3B) is moved upwardly, pushing upwardly on shaft 47 to
actuate valve 36 to the "open" position. When winding 59 is
energized, the solenoid armature 61 is moved downwardly,
pulling downward on the shaft 47 to actuate the valve 36 to
the "close" position.
In the sensors utilized in the present invention,
the magnitude o~ the downhole parameter is represented by
electric pulses. The sequence o~ the pulses represents a
code (binary or other~ and this sequence represents the
magnitude of the parameter. ~ig. 3E illustrates how each
single pulse of this code is processed to operate the valve
36. In Fig. 3E, numeral l77 represents one such pulse which
is narrow in time; being only a few microseconds long. This
pulse 177 is impressed upDn the circuitry contained in
block 178. This block 178 contains a so called "one shot"
univibra~or and suitable inverting rectifying circuits well
known in the electronics art and provides cin response to
_~9_
~.;242'~8
the single input pulse) two output pulses separated in time
by tl (the first pulse is normally time coincident ~ith the
input pulse and the second appears later by an amount of
time equal to tl~ as shown by pulses 179 and 180. These
electric pulses 179, 180 are now impressed, respectively,
upon the circuitry contained in blocks 181, 182. These two
circuits are identical and are so called pulse lengthening
circuits, also well known in the electronics art. Each
input pulse is lengthened to provide output pulses 183 and
184. These pulses are re$pectively applied to the "Darlington"
power amplifiers 185 and 186 (as manufactured by Lambda Mfg.
Co. of Melville, New Y~rk, and sold under the type PMD16K100).
In the practical desIgn of the electronic logic and
power circuitry of Fig. 3E that I use in this preferred em-
bodiment, L have chosen as constants tl - 500 milliseconds
and t2 = 20 milliseconds. In operation, when a single pulse
177 is impressed on lead 167, the Darlington 185 is turned
on for 20 milliseconds and then turned off. Then 500 milli-
seconds later the Darlington 186 is turned on ~or 20 milli-
seconds and then tu~ned off. Thus, the valve 36 is openedfor 500 milliseconds without requiring any energy during
this period. Energy is required only during the short 20
millisecond periods that are required to actu~te the valve
~ 36 to the "open" or to the "close" position. The figures
given above are for illustrative purposes only. Suffice it
to say that by making the action of the valve 36: (a~ ve~y
fast and (b) bi-stable; very high pressures and volumes of
mud can be valved wîthout the necessity of employing large
amounts of energy and as hereinabove described, relatively
-30-
~.24Z28
small energy batteries can operate the valve about one
million times.
In a typical embodiment of this apparatus, the
weight of the entire valve mechanism 36 o ~gs. 2~ or 3A,
including the solenoid a~matu~e 54, shaft 46 and piston 50
is approximately 9 ounces. The valve 36 has been designed
to operate at a differential pressure o~ 1,600 psi and pro-
portioned to operate at optimum perormance, including the
consequence that the ~orce required to open and shut the
valve 36 must exceed the force due to vertical acceleration
of all the apparatus near the bit 26.
Assuming a vibration figure of 60 g and the weight
of 9 ounces, maximum vertical force on valve 36 due to the
vibration of the tool 56 will be about 34 pounds. To be
certain that the valve 36 will not open accidentally, the
force keeping the v~lve closed in Fig. 2B and the orce
keeping the valve open in Fig. 2A must both exceed about 34
pounds. By suitable choice of the first and second indepen-
dent parameters hereinabove described, a "balanced'' conditîon
is achieved. By ''balanced'' is meant that the force required
to open the valve 36 is equal to the force required to close
it.
Above ~round equipment utilized with the present
invention, particularly as to methods and apparatus for
eliminating interferring efects that are present in the
output of pressure transducer 100, can take various forms,
as will now ~e described.
l~,.Z4ZZ8
Fig. 4 shows typical above ground equipment in
accordance with a preferred embodinlent of the invention,
wherein the downhole parameter being sensed is the radio-
activity of formations traversed by the bore while drilling
is in progress. The corresponding portion of the loggin~
equipment which is below the earth's surface has been pre-
viously described and shown in Figs. 2A, 2B, and 3A-G.
Referring now to Fig. 4, pressure transducer 100
connected to the standpipe 16 converts the variation of mud
pressure within the standpipe into a varying electrical
voltage. This voltage represents a mixture of two component
signals: the useful, information carrying signal and the
interferring signal. The information carrying signal is a
succession of short, negative mud pressure pulses produced
by the sudden opening and closing of the valve 36. The
interferring signal is in the form of relatively slow and
'- periodic pressure ~ariations which are generated by the
strokes of the mud pump 12. These mud pump signals tend to
mask or obscure the information one desires to obtain by
utilizing the short negative mud pressure pulses.
One of the objectives of this invention is to
recover, from the "contaminated" signal produced by the
transducer, a "clean" signal which gives the desired infor-
mation. This is accomplished by means of a signal extractor
102 which is applied to the output terminal 101 of the
pressure transducer 100. The signal extractor eliminates
the interferring effects and produces across its output
terminal 108 a succession of pulses from which the infor-
mation regarding the downhole parameter can be readily
obtained.
32
~.2~'28
~ he signal extractor 102 is controlled in a pre-
determined manner by a succession of timing pulses obtained
from a pulse generator 111 and applied to the control terminals
113, 114. The pulse generator 111 is mechanically driven by
S the mud pump 12 to produce an appropriate number of timing
pulses per revoluti~n of the pump. A chain drive trans-
mission assembl~ 112 is provided for this purpose.
The "clean'' information carrying signal obtained
from the extractor 102 is in the form of pulses derived from
the actuation of valve 36 of generator 28. The relevant
information is provided by the time intervals separating the
pulses. A time-to-amplitude convertor 115 connected to the
signal extractor output terminal 108 converts these pulses
derived from the actuation of the valve 36 of generator 28
into signals having magnitudes representing the intervals
therebetween. The convertor 115 is a well known electronic
device and can be made up of components manufactured by the
Burr-Brown company of luscon, Arizona, U. S. A. For further
detailed description of time-to-amplitude converters see:
M. Bertolaccini and S. Cova,'''I,og'ic Design of High Precision
Time 'to'Pul'se' Hei'ght' C'o'n'ver't'e'~s", Nuclear Instruments and
Methods 121 (1974), pp. 547-566, North Holland Publishing
Co .,
The signals derived fr~m the convertor 115 are in
turn applied to the input terminal 109 of a reciprocation
circuit 118. The reciprocation circuit 118 (as, for example,
-manufactured by Analog Devices, Inc. of Norwood, Mass.)
produces output voltages which are the reciprocals of the
' input voltages. Thus, if a voltage of magnitude M is applied
to reciprocation circuit 118, an output voltage having
-33-
~.Z 42'~ ~
magnitude l/M is obtained. These signals having magnitudes
l/M are in turn recorded on the chart of a recorder lZ0.
The record chart of recorder 120 is moved in correlation
with changing depth of the sensor unit 30 by a depth sensing
device 121. The depth sensing device may be, for example, a
modification or adaptation of equipment such as marketed by
The Geolograph Medeavis Company gf ~klahoma City, Oklahoma,
U. S. A.
In order to show more clearly the operating fea-
tures of the signal extractor 102, we will analyze the
behavior of the various signals which are involved. They
are shown schematically in a simplified and idealized form
as they vary with time in ~ig. 5. Let
F(t) = S(t) ~ N(t) (12
where S(t) is the useful information carryin~ signal formed
by the negative mud pressure pulses Pl, P2, and P3 aligned
along the time axis t. [See Fig. 5 (axis A)~ The times of
arrival oE these pulses, which correspond to the ti~es of
actuation of the valve 36 of generator 28, are tl, t2 and
~0 t3, respectively. The time intervals which separate these
pulses are ~1 = t2 ~ tl~ A2 = t3 - t2. ~3 t4 - t3, etc-
are indicative of the intensity of the radiation measured.
If these time intervals are large, the intensity is relatively
weak and conversely, iE they are small, the intensity is
relatively strong. The interfering signal produced by the
mud pump 12 is represented in Fig. 5 (axis A) by a periodic
but not necessarily sinusodal function N(t) having a period
T. The length of the period is related to the speed of
rotation of the pump.
-34-
~.Z4228
To facilitate explanation, ~he relative scales in
Fig. 5 have been distorted. In actual practice, there may
be 50 to 80 oscillations o~ N(t) between the time of ar~ival
of Pl and P2. Thus, ~i and ~2 may vary from 50T t~ 80T.
~owever, in Fig. 5 (axis A~ only a few oscillations oF N(t)
b^tw^en Pl and P~ ha~rre been shown. Furthermore, in actual
practice the negative mud pressure pulses Pl, P2, P3 do not
have clean rectangular forms as in Fig. 5 (axis A). More-
over, thè actual pulses are much smaller than those which
have been shown in Fig. 5 Caxis A). In actual experience,
the magnitude of Pl, P2 or P3 is about 0.1 t~ O.ûl of the
ma~imum a.mpl~tude of the pulsations N(t).
Axes A-E in Fi~. 5 are positinned one below the
othPr so that one can compare the signals in thèir time
relationship~s on~ to another. Usinsr these figures, we can
now enumerate the instrumental steps which are involved in
the operation of the signal extractor 102. These are as
follows:
Step 1 We displace the input F(t~ by an amount T, to obtain
2Q F(t - T) = S(t - T) ~ N(t - T) (2)
where S(t - T) and N(t - T) are, respectively, ~he displaced
u.seful signal and displaced interfering signal. Both
signals are shown in Fig. 5 (axis B). The signal S(t - T)
is represented by pulses Pl(a), P2(a) and P3(a~ which have
been obtained by displacing by an amount T the corresponding
pulses Pl, P2 and P3 in Fig. 5 (axis ~). The signal N(t -
T) in Fig. 5 (axis B) is shown to be in exact synchronism
with N(t) in Fig. 5 ~axis A). This is due to the periodicity
of the signal. Thus>
N(t - T) - N(t) (3)
-35-
~.Z4Z28
Step 2 We subtract the displaced input function F(t - T)
from the original input function F(t) to obtain
M(t) = F(t~ - F(t - T) (4)
Taking into account (1~, (2) and (3), we obtain
M(t~ = S(t) - S(t-T) (52
Thus, the inter:Eering signal has been eliminAted and does
not appear in M(t~. This can also be seen from inspection
of Fig. 5 (axes ~ and B).
As shown in Fig. 5 (axis C), M(t~ consists of
impulses which occur in pairs. Each pair contains a nega-
tive and a positive pulse separated one from another by a
time interval T. Thus, we observe a pair consisting of
Pl(b) and Pl(b) which is followed by a succeeding pair
consisting of P2(b) and P2(b) , then by another pair con-
sisting of p~c) and P3~ and so on.Step 3 We displace M(t) by a time T so as to obtain M(t -
T). Thus, the entire sequence of pulses in Fig. 5 (axis C)
is shi~ted along the time axis by T so as to appear as shown
in Fig. 5 (axis D). The arrangement of pulses as in pairs
has been preserved in Fig. 5 (axis D). However, each pair
such as Pl(C) and Pll~ is displaced with respect to the
pair Pl(b) and Pl~b~ ~shown in Fig. 5 (axis C)] by T.
Similarily, the pair P2~C) and P2(C) is displaced with
respect to the pair p2(b)and P2(b) by T, and so on.
Step 4 We compare the displaced pulses in Fig. 5 (axis D)
with those in Fig. 5 (axis C). We note that some of these
in Fig. 5 (axis D) are in time coincidence with some of the
pulses in Fig. 5 ~axis C). The instances at which coincidence
occurs are recorded in Fig. 5 ~axis E) as pulses Pl(d),
36 ~~~~
~.Z~Z28
pz(d) and p3(d~. Thus,
Pl(d) coincides with Pl(b) and Pl(C)
P2(d) coincides with P2(b) and P2(C)
p3(d~ coincides with P3~ and p3(c)
The times at which the pulses pl(d~, P2(d) and p3(d) occur
1 ' 2 and t3 ~ T-
The pulses Pl~d~, P2(d2 and p3(d) correspond to
the pulses Pl, P2 and P3 shown in Fig. S Caxis A). Con-
sequently, the pulses in Fig. 5 (axis E) also represent this
useful f~mction which now is S(t-T) since it has only been
displaced by T. It is evident that the pulses in Fig. 5
(axis E) provide the information which we are seeking to
obtain. The time interval between Pl(d~ and P2(d~ is Al,
and the time interval between P2(d) and p3(d) is ~2' etc..
The quantities ~1~ A2, etc. are indicative of the radiation
measured by the gamma ray detector.
The above steps will now be considered as they
relate to the performance of the signal extractor 102 and
more particularly to that of its two component parts de-
signated in Fig. 4 as 105 and 107 and shown schematically in
Figs. 6 and 7, respectively.
The component 105 receives at its input terminal
101 (which is the same as that of the signal extractor 102 of
Fig. 4) the signal F(t~. As shown in Fig. 6, this signal is
transmitted through an amplifier 130 to the input terminal
131 of a delay network 132. The delay network delays F(t)
by T, thus, pr~ducing at its output terminal 134 the signal
F(t-T). This signal is a sum of two component slgnals S(t-
T) and N(t-T) which are shown in Fig. 5 (axis B).
-37-
~L~.Z4228
The signal F(t-T) is applied to one input terminal
134 of a subtractor 135. The other input terminal 136 of
the subtractor receives directly the signal F(t~, which is
transmitted from termin~l 10l. by means oE conductor 137.
Thus, at the output terminal 106 of the subtractor 135 we
obtain the difference signal M(t~ ~ F~t~ - F(t-T). This
is shown in Fig. 5 (axis C).
The delay network 132 is provided with control
terminal 113 which receives a signal controlling the delay
T. It is important that the length of the delay T be the
same as the period of mud pressure oscillati~ns produced by
the mud pump 12.
The amount of the delay T is controlled by the
timing impulses derived rom pulse generator 111 shown also
in Fig. 4 and applied via conductor 110 to the control
terminal 113. It is noted that the delay T is the same as
the period of oscillation of mud pressure produced in the
successive strokes of the mud pump 12. Consequently, the
frequency of these timing pulses must be controlled by the
rotation o~ the pump.
Assume that the pump produces Nl strokes per
second. Thus, T = l/Nl. The pulse generator 111 produces
timing pulses at a relatively high rate N2, which is a
multiple of Nl. Thus, N2 ~ KNl, where K is a constant which
has been chosen to be 512. Thus, if the strokes of the pump
are one per second this would require the signal generator
to produce 512 pulses per second. It is apparent that ~he
rate of pulsation of the mud pump 12 varies with time and,
accordingly, N2 will vary so as to insure that the delay
~.24ZZ8
produced by delay network 132 will always be equal to one
period of the mud pressure oscillations produced by the mud
pump 12.
The delay network 132 which is controlled, as
described above, may be a ~eticon ~odel SAD-1024 Dual Analog
Delay Line as marketed by Reticon Corporation, Sunnyvale,
California, U. S. ~..
The instrumental steps hexebefore described are
the steps 1 and 2 performed by the component 105 of the
signal extractor 102. We ha~e transformed the input signal
F(t) [represented by its components in Fig. 5 (axis A)] into
an output signal M(t) which appears as a succession of pairs
of pulses and is shown in Fig. 5 (axis C). We ~ill now
proceed to describe further instrumental steps ~hich are
required in order to accomplish the desired objectives.
These are performed by the component 107 of the signal
extrac~or 102.
We refer now to Fig. 7. The signal ~(t2 is now
applied through conductor 140 to a delay network 141. This
delay network is identical to that designated as 132 in Fig.
6. It receives, at its control terminal 114, the same
control signal which was applied to the control terminal 113
of the delay network 105. Consequently, the amount of delay
produced by delay network 141 is T and the signal appearing
at the output of 141 is M(t-T) as shown in ~ig. 5 ~axis D).
This output signal is transmitted through an amplifier 143
to one input terminal 145 of an AND gate 146. At the same
time, the undelayed signal M~t) is applied through the
conductor 147 and amplifier 148 to the other input terminal
3~ --
~L~.Z4228
149 of the ~ND gate 146. These two input signals M(t) and
M(t-T) which are applied to the ~ND gate 146 are shown in
Fig. 5 (axes A and D), respectively. ~e have previously
observed that some impulses shown in Fig. 5 (axis C) occur
in coincidence with imp~llses in Fig. 5 (a~is D). Those
impulses that occur in coincidence appear in the output of
the ~ND gate 146. They are designated in Fig. 5 (axis E) as
Pl(d), P2(d) and p3(d). These coincident pulses are the
output of pulses of the component 107, and consequently of
the signal extractor 102.
It is thus apparent that by means of the component
107, we have per~ormed the instrumental steps 3 and 4. We
have transformed the signal M(t) shown in ~ig. 5 (a~is C)
into the signal S(t-T) shown in Fig. 5 (axis E). The latter
provides the quantities Al, A2, ~3, etc., which represent
the information it was desired to obtain. It should be
recalled that the sîgnal S(t-T) is represented by a succes-
sion of pulses as shown in Fig. 5 (axis E). These pulses
are transmitted to the time-to-amplitude convertor 115 to
produce at the output of the time-to-amplitude convertor 115
signals of various magnitude such as Al, ~2~ ~3, etc., that
represent time intervals between the arrival of pulses.
These signals are in turn fed to and transformed by the
reciprocation circuit 118 of Fig. 4 into other reciprocal signals
having magnitudes 1~ 2, l!A3, respectively. These
reciprocal signals are recorded by recorder 120 of Fig. 4.
It is apparent that the quantities l!Al, l!A2 and l/A3
represent the intensity of radioactivity of formations
sensed by the sensor unit 30 at various depths in the borehole.
-~0-
~L~.Z4Z28
We have described above an instrumental means for
performing lo~ical steps leading from the function F(t) to a
function S(t-T). These steps ha~e been performed by re-
presenting these f~mctions in an analog (non-digital) form.
Alternati~ely, i~ de~ired, the entire process can be digi-
talized, as shown diagramatically by Fig. 8. In ~ig. 8, the
output of the pressure transducer 100 is fed to an analog-
to-digital con~ertor 103, the output of which is fed to a
digital computer 104. The operations indicated in Fig. 8
are performed by the elements designated 122, 123, 124, 125
and 126 in the dig~tal computer 104. Timing signals from a
pulse generator 111 or 140 are introduced to the digital
computer 104 in order to control the delays in accordance
with the pump speed. The operations indicated in the dotted
rectangle of ~ig. 8 are performed mathematically in a sequence
which may be flow charted. The output of the computer 104
is fed to a digital-to-analog convertor 127, the output of
which is fed to the recorder 120.
In Fig. 9. there is shown an arrangement similar
in some respects to that of Fig. 4, but wherein the data to
be obtained and recorded are the temperature at the location
of sensor unit 30 o~ Fig. 1. In Fig. 9, these data, as
presented to the signal extractor 102 are in digital form
(see Fig. 3D). The signal extractor 102 of Fig. ~ is identi-
cal to that of Fig. ~, but the time-to-amplitude convertor
115 and the reciprocation circuit 118 of ~ig. 4 are replaced
by a digital-to-analog convertor 141. The output signals o~
an appropriate pulse generator will be applied to the control
terminal 110 of the signal extractor 102.
-41-
Z4228
- It is not al~ays convenient to provide a mechanical
connection to the mud pump 12, as shown by the chain drive
transmission assembly 112 in Fig. 4, and an alternate means
for generating the pulses required for the signal e~tractor
may be desirable. Fig. 10 illustrates such an alternate
means. In a typical example, the signal. extractor 102 of
~ig. 4 is provided ~t its terminal 110 with pulses at a rate
of 512 pulses per full pump stroke. It must be clearly
understood that this rate must be rigorously sychronized
with the pump strokes. All the "times" sho~n as T, tl, t2,
etc. in Fig. 5 are not so-called "real time", but are direct-
ly related to the speed of the mud pump 12 and rigorously,
T, tl, t2, etc. should be expressed, not in seconds or
minutes of "time" but in "~allons of mud". When it said
that at terminal 110 o~ Fig. 4, there are 512 pulses per
mud pump stroke, it is meant that at ter~inal 110 there are
present voltage pulses having a frequency equal to the 512th
harmonic of the pump stroke frequency. Fig. 10 shows how
this can be accomplished without mechanical connection to
the pump shaft.
In ~ig. 10, component 145 is a VC0 or "voltage
controlled oscillator" which at its output 110 produces
electric pulses the frequency o~ which is controlled by the
DC voltage applied at its input terminal 108. Component 150
is a binary divider or scaler that divides the frequency of
the pulses impressed on its input terminal 116 and generates
output pulses at its output terminal 117 having a frequency
equal to l/512th of f~equency of the input pulses~ Compon-
ent ll9 is a phase comparator that compares two inputs (one
-42-
~.242Z8
from scaler output terminal 117 and one from the output
terminal 130 of pressure transducer 100~, and provides at
its output terminal 128 a voltage which is zero volts DC
when the two inputs 117 and 130 are exactly equal in phase;
and provides a positive voltage when the input at 117 leads
the input at 130 in phase; and a negative DC voltage when
the input at 117 lags the inptlt at 130 in phase. A battery
129 is pro~ided to properly bias the VCO 145. The circuit
151, just described, is known as "phase locked loop". The
operation is best explained by an example: Assume that the
pump pulse frequency (pump stroke frequency~ is 1 Hz and the
VCO is generating 512 Hz. The output of the scaler 150 will
then generate exactly 1 Hz. The 1 Hz f~om the scaler 150
and the 1 Hz from the p~essure transducer 100 will then be
exactly matclled in frequency and phase and the output of
the comparator at terminal 128 will be zero volts, and the
VCO 145, when properly biased by battery 129, will generate
exactly 512 pulses per stroke.
~ssume now that the mud pump 12 speeds up. The
frequency at ter~lnal 130 will than be somewhat greater
than 1 ~z - i.e., 1 ~ alHz. The comparator 119 will then
provide an output at terminal 128 ~hich will no longer be
zero volts DC, but for example, ~ a2V, this small voltage
increment will be applied to the VC0 145 at terminal 10-8
and increase its frequency until the nominal 512 pulses per
second is increased to a value f such that f/512 - 1 + ~l
Thus, the frequency at terminal 110 will always
accurately follow the frequenc~ of the mud pump 12 and will
always be its 512th multlple.
-43-
~.242Zi 3
- Two arrangements for obtaining ti~ing pulses for
the signal extractor 102 have been hereinabove described
(pulse generato~ 111 of Fig. 4 ~nd the "phase locked loop"
circuit 151 of Fi~. 10). ~ third arran~ement that may be
used for ~btaining such timing pulses is illustrated by
Fig. 11 and is based on ''auto-correlation". In Fi~. 11, the
input terminal 154 of a correlator 152 is supplied by the
output of the pxessure transducer 100, and receives the
function F(t) which contains the periodic signal N(t) and
the function S(t2 which may be considered a ~andom function.
The output of the pressure transducer 100 is also applied to
the input terminal 101 o~ the signal extractor 102. The
correlator 152 is adapted to produce ac~oss its output
terminals the autocorrelation function o F(t) which is
~ff CT) = [S(t2 ~ N(t2 ] rS(t+T~ -~ N(tr~T) ] (6)
Where the bar in the above expression indicates averaging
over an appropriate period of time. The function ~ff (T)
can be expressed as
- ~ ,bSS (,T) ~ tl3nn(T~
where
~s5 (T) - S(t2S(t ~ ~ (8)
and
~ nn(T) = N(t~-N(t--~ r)
The function ~5s (T~ reaches zero at some value of T = TO
and beyond rO, we have
~ ff (T) a ~nn(') (10)
Since ~nn(T) is periodic, the function ~ff (T~ iS also
periodic and it has the period T. This function, which is
obtained in the output of the correlator 152 is in turn
applied to a pulse multiplier 153 which produce~s a succession
~L~.Z~28
of timing pulses similar to those produced by the pulse
generator 111 in Fig. 4 and which are applied to input
terminal 110 of the signal extractor 102. The pulse mult-
iplier 153 multiplies the f~equency of the input pulses by a
phase locked loop system simila~ to that of Fig. 10 or by
any other conventional means. The remaining elements in
Fig. 11 are the same as those in ~ig. 4, except, of course,
that the pul'se generator 1'11 and its chain d~ive trans-
mission assembly 112 are elim~nated.
There are commercially available instrumental
means based on auto-correlation ~or recovering a periodic
signal from a mixture o~ a perisdic and a random signal
~see, for example,' S'ta't'i's'tic`a'l'Th'e'ory' of Communications,
by Y. W. Lee, John Wiley, New York, N. Y., 1960, pp. 288-290).
The correllator 152 of Fig. 11 may be Model 3721A manufactured
by Hewlett Packard Company of Palo Alto, California. The
correllator 152 could also be one of the types described
in the following references: A. E. Hastings and J. E.
Meade "A Device for Computing Correlation ~unctions",
Revie _of''S'c _n'ti'~'ic 'Ins'trumen'ts, Vol. 23, 1952, pp. 347-349;
and F. E. Brooks, Jr. and H. W. Smith, "A Compute~ for
Correlation Functions",' ~evie of Scientific Instruments,
Vol. 23, 1952, pp. 121-126.
45-
~L~.;Z 42Z8
While I ha~e sho~n my invention in several forms,
it will be obvious to those skilled in the art that it is
not so limited, but is susceptible of various changes and
modifications without departing ~rom the spirit thereof.
I have disclosed herein, as examples, sensors for
only two downhole parameters, it is, however, to be under-
stood that sensors for various other downhole parameters
could be used as well. It is also to be understood that
sensors for a plurality of downhole parameters may be used
at the same time, in which case, conventional techniques
would be employed (sucIi as time ~h~ring, multiplexing, or
the like) to handle the data representing the plur~lity of
parameterS-
When deviated or inclined wells are drilled, aturbine or "mud motor" such as a Dynadrill, manufactured by
Smith Industries, Inc., Houston, Texas, is frequently
employed. In such case, the drill string 31 of Fig. 1, is
not rotated by the rotary table at the surface. The rotating
~46-
'~ Z ~2 Z 8
action to turn the bit 26 is der~ved from such a mud motor,
which usually is located immedi~tely above the bit 26 in
the drill string compri$ing elements 22, 24, 28, 30, of Fig.
1. When such a mud m~tor ls employed, a large pressure
drop occurs across it (since the mud motor derives its
power from the ~ud flo~. This large pressure drop can be
utilized to supply the pressure difference between the inside
of the drill strin~ and the annulus and, in such case, a
"jet" type bit need not be employed.
The p~esence of the pressure drop across the ~nud
motor merely enhances the operation of my invention so long
as the negative mud pressure pulse generator is located
above the mud motor.
The term ".~low restriction means", for purposes
herein, applies to either a jet type bit, or a mud motor,
or both. The term "high pressure zone" applies to the
drilling fluid pressure on the upstream side of the "flow
restriction means" and the term "low pressure zone" applies
to the dr-illing fluid pressure on the downstream side of
the "flow restriction means".
It is recognized that, in some instances, a
plurality of mud pumps flre employed on a single drilling
rig and these pumps are not necessarily operated in syn-
chronism.
In an example of three pumps, the periodic pres-
sure curve of Fig. 5A would, in the practical case, not be
a simple periodic function as shown b~ N(t~ but would be
the sum of three co~ponents, each component being periodic
and having its own distinct period.
47
~.Z~2Z8
By the employement of three delay systems, as
shown in Fig. 6, each synchronized with its own pump, each
periodic component of the interfering mud pulse pressure
signal can be separately nullified. Suitable interconnec-
5. tion will then produce a signal ~rom which the interferingmud pump pressure signals are eliminated.
The foregoing disclosure and the showings made in
the drawings are merely illustrative of the principles of
this invention and are not ts be interpreted in a limiting
sense.
-48-
