FR2475111A1 - SYSTEMS AND METHODS FOR MEASURING SURFACE HOLES DURING DRILLING - Google Patents

Abstract

The invention relates to systems and methods for measuring sounding holes during drilling. IN AN OIL WELL DRILLING TOWER 20, BETWEEN TREPAN 33 AND DRILL COLLAR 26, BETWEEN TREPAN 33 AND DRILL COLLAR 26, A SPECIAL TELEMEASURE TRANSMITTER 50 IS PROVIDED FOR CONTAINING A VALVE 40, A TREATMENT ASSEMBLY ELECTRONICS 96 AND SENSORS 101 COMMANDING THIS VALVE 40. ASSEMBLY 96 EMITS CODED ELECTRIC PULSES REPRESENTING A PARAMETER TO BE MEASURED BY SENSOR 101, WHILE OPENINGS AND CLOSURES OF VALVE 40 ARE GENERATED BY CREATING PRESSURE PULSES THROUGHOUT THE RISING COLUMN 24. THE INVENTION IS USED TO PROVIDE THE GROUND SURFACE WITH INDICATIONS RELATING TO THE BOTTOM OF AN OIL WELL BOREHOLE.

Description

24 7511 1

  The present invention relates, in general, to measurements made in the course of

  drilling a borehole in the soil and

  in particular, systems, apparatus,

  and methods in which hydraulic shock waves are used in the drilling mud column to transmit signals on the ground surface representing one or more parameters of the bottom of the hole. The invention also relates to

  systems and methods for detecting such

  in the presence of noise.

  The present invention relates to systems

  to transmit on the ground surface data from the bottom of a well in the course of

drilling of the latter.

  For a long time, it has been accepted in the oil industry that obtaining data about the bottom of a hole during the drilling of a

  well could provide relevant information

  to the operator in charge of drilling. Information

  such as the actual weight on the bit, the inclination and the range of the borehole, the face of the tool, the pressure of the fluid, as well as

  the temperature at the bottom of the hole and the radio

  activity of the substances that surround or

  against the trephine, would all be expressed by quantities of interest to the drilling operator. A number of proposals have been

  issued in the prior art with a view to

  these quantities during drilling, for the trans-

  then put on the surface of the ground. In the technical

  than earlier, different transmission systems have been proposed for this purpose. In this regard reference may be made, for example, to U.S. Patent Nos. 2,787,795 granted to JJ Arps, 2,887,298 to HD Hampton, 4,078,620 to JH Westlake. et al., 4,001,773 granted to the names of AE Lamel et al., 3,964. 556 granted to the names of Marvin Gearhart et al., 3,983,948 granted in the name of J.D. Jeter and 3,791,043 granted in the name of M.K. Russell. All the patents listed below

  above are mentioned here for reference.

  From the practical point of view, the most

  of these earlier proposals is probably

  that of transmitting signals

  pressure pulses in the drilling fluid.

  Various methods have been suggested in the

  than prior to producing these sludge pulsations either by a controlled restriction of the sludge circuit by means of a flow restrictor valve suitably installed in the main stream of sludge, or by means of a bypass valve interposed between the In the prior art, it has been suggested to produce these sludge pressure pulses by means of valves having an internal diameter of the drill string (high pressure side) and the annular space surrounding the drill string (low pressure side). for the purpose of reducing the flow of mud inside the drill string or deriving a certain part of this flow towards the

  low pressure zone in the annulus

  surrounding the drill string. These valves

  necessarily work slowly because, when

  that they are used inside the train of

  stems, they have to control very large volumes

  slurry carriers and, when used to control a bypass, because of the very high pressure differences, these valves are also slowly actuated by a motor. For example, in response to a measurement below the

  soil surface, a motor damper of this type

  tilted between the inside of the drill string and the annular space has the effect of producing slow reductions and elevations of pressure

  mud. These reductions and increases in

  are then detected on the soil surface.

  In order to better understand the operation of a slow acting motor valve of the type suggested by the prior art, reference is made to FIG. 1A which shows the opening and closing of a valve of this type as a function of time. . Referring more specifically to this FIG. 1A, the time, t, is plotted on the abscissa, while the degree of opening of the flap, R, is

carried on the ordinate.

  R = S (t) (1) SO where SO indicates the total area of the opening and S (t) indicates the area that is open at time t during the opening or closing process of the valve. Therefore, when R = O, the valve is closed and when R = 1, the valve is completely

  open. The moments intervening in the function-

of the flap are as follows:

  a (v) = OA1 is the moment when the key

  start to open; t (V) = OB1 is the moment at which the flap is

completely open; -

  t (v) = OC1 is the moment at which the valve c begins to close; t (v) = OD1 is the moment at which the flap is

completely closed.

  Time lapse: T (v) = t (v) v) - tv) = t tv) (2) a b a d c

  Ta (V) will be called "opening or closing time"

  "Time lapse: T (v) = t (v) - t (v) (3) bcb TbV) will be called" open flow time. "Therefore, the total operating period of the flap is : T (V) = 2T (V) + T (v) (4) ta * b In the tests described above, we had T (v) = 1 second, Tb (V) = 2 seconds and, by a therefore, the total operating time of the valve was Tt (V) = 4 seconds, which were relatively slow opening and closing of the valve.

  to reductions and elevations of a

  corresponding pressure in the mud pressure at

the soil surface (see Figure lB).

  It can be seen that the sludge pressure decreased from its normal value which was, for example, 70 kg / cm 2 (when the valve was closed) to its minimum value of 52.5 kg / cm 2 (when the valve was open ). The moments occurring in the pressure variations thus observed were as follows: t (s) = OE1 is the moment at which the sludge pressure begins to decrease from its normal value of 70 kg / cm 2; t () = OF1. is the moment at which the sludge pressure reaches its minimum value at 52.5 kg / cm2, at which point it is maintained until the moment t (S) = OG = osest e moent- ticS (s) = OG1 is the moment at which the pressure of the mud begins to rise; t (s) = OH1 is the moment at which the mud pressure reaches its normal value of

70 kg / cm2.

  Therefore, the pressure decreased during the time period T (S) = t (s) t (S), then time T1 = tlb la, then it remained constant over the lapse of time Ts) = tSc) - t ) and then, it rose from its value reduced to its normal value during the lasdt sT = (s) - t (S) Therefore, the time period 3 - - Therefore, the 3 Id the

  total time during which the mud flows

  The milk through the bypass valve as a result of a single operation of the main valve was: Tt (S) = T (S) + T2 (S) + TS). In FIG. 1A, quantities (e.g. , t (v) tb (V) t (v) t (v) T (V) T (v) T (y) a 'c, d' aa TbV) and TtV)) with the exponent "lv" to indicate that these quantities relate to the functioning of the

  flapper located below the surface of the ground.

  On the other hand, in FIG. 1B, the quantities (s) (s) (s) (s) (s) (s) (s) (s) the t1b, t let1 T1 T T3 and Tt are lb 'lc' 1 '2' 3 t indicated by the exponent "s" to indicate that these quantities relate to the measurements made in

  ground surface. This distinction between

  exhibiting "v" and exhibiting "s" is essential in order to understand

  perfectly some of the new features

  of the present invention. In this respect, it is essential to distinguish between

  and the effect or, in other words, between the phenomena

  butts occurring at the bottom of the hole near the valve and those occurring at the detector installed on the surface. An essential feature of the previously proposed system is based on the following relationships:

T (S) = T (V) (6)

i

- T2) = T (V) (7)

2 b T3S) = T (v) (8) 3 a

  These relationships demonstrate that the

  period of reduction or increase of pressure at the soil surface was the same as the corresponding period of opening and closing of the damper, while the period during which

247511 1

  the pressure was almost constant (at a

  value) was the same as the period

  the valve was fully open.

  In other words, reduction and subsequent elevation

  the pressure of the mud on the surface of the

  ground were in exact correspondence with the opening

  ture and closing the valve. In the present specification, it will be stipulated that this condition, expressed by the relations (6), (7) and (8), is related to a "regime of slow variations of pressure". As suggested in the art

  earlier, the regime of slow variations in

  is not suitable for practicing tele-

  measurement during drilling operations, in particular

  bind, when measuring multiple parameters

  at the bottom of a hole. At the moment when a first para-

  meter has been measured, coded, transmitted to the surface, then decoded, the borehole of the well may have deepened and the second parameter may no longer be available for measurement. Time lapse

  relatively long times are necessary to

  the data measured in an appropriate form

  required for detection and registration. The entire logging process is tedious and

  lengthy. In addition, additional difficulties

  These effects result from various spurious effects such as the pulsations due to the sludge pump and the noise associated with the various

  drilling. It is believed that a motor valve function-

  slowly as the one suggested in the

  previous speaker is not able to answer

  the usual conditions imposed in the indus-

trie.

  Certain objects of the present invention

  are made using shock waves

  hydraulic systems for the purpose of telemetering

  logging trainings during drilling. These

  shock waves are produced by a clapper

  vation acting very rapidly (almost instantaneously).

  for all practical applications), this valve being interposed between the inside of the train of

  stems and the annular space surrounding it.

  When this bypass valve opens suddenly,

  the pressure prevailing in the immediate vicinity of this

  pet falls and returns almost instantly to his

  normal value by emitting a clear negative impulse

  while, on the other hand, when the bypass valve closes abruptly, a clear impulse

  positive is issued. The elasticity is used

  tee of the mud column to facilitate training

  and the transmission of these shock waves. This phenomenon

  is similar to the well-known water hammer that has been encountered previously in hydraulic transmission systems. (See, for example, John Parmakian, Water Hammer Analysis, Prentice Hall, Inc., New York, N.Y., U.S.A. 1955 or V. L. Streeter and E. B. Wylie, "Hydraulic Transients".

  McGraw-Hill Book Co., New York, NY, USA).

  Important features of the present invention, for example, the formation and

  detection of hydraulic shock waves, are re-

  schematically shown in Figures 2A and 2B. The graph in Fig. 2A shows the openings and closings of a quick acting valve producing shock waves, while the graph of Fig. 2B shows the variations

  pressure detected on the soil surface and results in

  both of the operation of the valve as shown in Figure 2A. Symbols such as A1, B1, C1, D1, + V) (v) V) (v) _ (v) _V l (V) tttt 1 tV) tv), tv), t Tv) Tav), Tv ) and T) of Figure 2A have a meaning similar to that

  corresponding symbols of FIG. - Any-

  times, the time scales of FIGS. 1A, 1B, 2A and 2B have been considerably distorted in order to

  facilitate the description and make the presentation more

  'clear. In examining FIG. 2A, it should firstly be noted that the opening and closing times of the valve according to the present invention are several orders of magnitude shorter than the corresponding times obtained by means of the motor valve described in FIG. referring to Figure lA. In the previously suggested system (Figure lA),

  had T (v) = i second, whereas, according to the pre-

  invention (Figure 2A), we have T (v) = 5 milli-

  * 15 seconds. It is the same with regard to the

  the length of time during which a flap remains open.

  In the previously suggested system (Figure-1A), we had T (v) = 2 seconds while in Figure 2A we have T (v) = 100 milliseconds. Therefore, for all practical applications, it can be considered that the openings and closures of the valve shown

  in Figure 2A are instantaneous or almost instantaneous

born.

  Rapid openings and closures

  or almost instantaneous flaps have a

  - important and long-lasting influence on the efficiency

  of a telemetry system used to

  take measurements during a drilling operation.

  Pressure variations detected on the surface

  in accordance with the present invention (FIG.

  re 2B) have absolutely no similarity to the

  pressure variations obtained by means of a

  slow acting fart (Figure lB). Previously, it was pointed out that equations (6), (7) and (8) show the relationships between cases

  illustrated in FIG. 1A and those illustrated in FIG.

  Similar relationships do not exist between cases

  illustrated in Figures 2A and 2B.

  As shown in FIGS. 1A and 1B, the opening of the valve caused a corresponding reduction in mud pressure at the soil surface and, on the other hand, the closure of this valve gave rise to a corresponding rise in pressure. In order to put the situation well in

  relief, it will be repeated that according to the prior art

  the opening of the valve produced a single phenomenon, in particular a reduction of the pressure, while the subsequent closing of the valve gave

  another unique phenomenon, namely an elevation

  pressure. On the other hand, according to

  invention, the rapid opening of the flap (as

  represented in FIG. 2A) gives rise to two phenomena

  namely, rapid reduction and

  pressure (negative impulse)

  1WI as shown in FIG. 2B) and this, contrarily

  the case shown in FIGS. 1A and 1B in which subsequent opening and closing of the valve is necessary to produce

  a reduction and a subsequent increase in the pressure

  if we. In addition, the rapid closing of the valve (as shown in FIG.

  2A) gives rise to a subsequent elevation and reduction of slurry pressure (positive pulse "N" as shown in FIG. 2B). This rise and subsequent pressure reduction do not occur in the systems suggested by the prior art. Following the

  In the present invention, two shock waves are pro-

  driven by a single operation of the valve. A waveform such as that represented in FIG. 2B, which comprises both a negative pulse and a positive pulse, will be called hereinafter

  "valve wave". The pressure pulses associated with

  to a valve wave oht a starting speed of several hundred kg per cm2 per second and

they are short-lived.

  It is interesting to underline the rapidity of the phenomena associated with the observed flap waves. The moments occurring in Figure 2B are the following:

  t (s) = OK is the moment of appearance of the impulse

negative sion tWl;

  t (s) = OL is the moment at which the negative impulse

Madame M weakens;

  t (s) = OM is the moment of appearance of the impulse

positive "N" t;

  t4 (s) = ON is the moment at which the impulse posi-

tive "N" weakens.

  The lapse of time T (s) representing the n

  "length" of the negative pulse "P" (or lim-

  positive pulse "N") is 100 milliseconds, while the time T (S) flowing between m

  the appearance of the negative impulse "M" and the

  positive impulse "N" is 110 milli-

  seconds. Therefore, in Figure 2B, that is to say in which we have: T (s) = T (s) + T (s) (9) u n m

  the total period of flow is 210 millis-

  while in Figure lB (see equation 5), the total flow period was T (S) = 4 t seconds. The graphs of Figures 1A, 1B, 2A and 2B have been simplified and idealized by eliminating wave series and other extraneous effects. It should also be noted (see FIG. 2B) that the bypass valve is at least partially open during the lapse of time between the times t 1 and t (s). During this time, it becomes 75 1 t 1

11 2451

  produces a slow decrease in pressure which is removed at the point of detection by means of a suitable filter. This pressure decrease is not

  shown in the graph of Figure 2B.

  It should also be emphasized that the numerical values shown in FIGS. 2A

  and 2B are given merely by way of example.

  It is understood that these values do not limit

  the invention to a particular example.

  It will be stipulated that the method explained with reference to FIGS. 2A and 2B is in relation

  with a "hydraulic shockwave regime".

  Therefore, a distinction is made between the hydraulic shock wave regime shown in Figures 2A and 2B, and the regime of slow variations

  pressure shown in Figures 1A and MB.

  By providing a regime of hydraulic shock waves, a telemetry system was obtained

  through which a large amount of information can be

  can be transmitted per unit of time. This system is much better designed to meet the usual conditions imposed in the industry than the one based on

  on the regime of slow variations of pressure.

  The valve according to the present invention is actuated by the output of one or more sensors detecting one or more parameters of the bottom of a hole made in the ground, near the bit. Only one measure of each parameter

  is represented by a succession of valve waves.

  Each valve wave corresponds to a single opening

  ture and at a single closing of the valve.

  When detected on the surface,

  these different successive valve waves (repre-

  sensing the useful signal) are usually associated

  various parasitic signals such as those produced by the operation of the pump and by

  other drilling operations. In a special system

  drilling, a large pump is used.

  surface-mounted to lower the sludge into the drill pipe and through the bit, and then return it to the surface via the annular space formed between the drill string and the well bore. According to the present invention, the effects

  nuisances due to the pump are eliminated by a

  in which account is taken of the periodicity

  of these effects. Other effects associated with

  Drilling patterns usually appear as a noise signal with a relatively wide frequency spectrum. In some cases, this noise signal is a white noise while in other cases it differs considerably from a white noise. In order to eliminate this noise signal, a digital filtering system is used

  which can be a suitable filter, a conformal filter

  pulse generator or a tip-forming filter.

  The adapted filter carries, at its maximum, the ratio

  signal / noise at the point of reception; a filter

  impulse trainer reduces, at a minimum, the difference

  effective difference between a desired output-and the output

  actual tie, while a pointer-shaped filter.

  It transforms the useful signal by contracting it into a signal sufficiently acute that it can be distinguished from a background noise. We adopt a special technique to adapt-these filters to

  objectives of the present invention. A technique-

  of this type requires memorization and reproduc-

  two reference signals. The first reference signal is a valve wave produced by the opening and closing of the valve, while the second reference signal represents noise due to drilling operations. Detection and storage of the first reference signal is carried out by eliminating the weight exerted on the

  biting and stopping the actual drilling (but still

  holding the normal operation of slurry pumps).

  As a result, a signal is obtained that is free of ambient noise. The detection and storage of the second reference signal is performed as drilling progresses over a period of time during which the valve is closed. A suitable digital computing system is provided for receiving the data representing one or both reference signals, this system extracting from these data a storage function for the matched filter, the

  pulse shaping filter or forcing filter

dot matrix.

  One aspect of the present invention relates to improvements in which the bistable operation of the valve assembly 40 of the special telemetry instrument is involved.

  50. Another aspect of the present invention

  identify the installation of a special mechanical system

  working hydraulically and ensuring periodic-

  the positive displacement of the valve 40 towards its

  closing position. In addition, a system is

  electrical system which prevents the operation of the valve 40 in the event of an electrical failure occurring in

  the apparatus installed at the bottom of the borehole.

  Other aspects of the present invention

  tion is concerned with improvements in which the power source 95 and the

  source 104 of the special instrument for tele-

  50. These improvements serve to increase

  to be considerably the number of satisfactory valve operations that can be achieved without having to recharge or replace the installed battery

at the bottom of the borehole.

  Another aspect of the present invention relates to improvements to pulse codes

24,751 1 1

  over time that only short impulse

  are of virtually constant duration are trans-

  while the lapse of time between successive pulses constitutes the measurements of the amplitude of the parameter in question. In addition, a system is described for improving the accuracy and precision in the transmission and detection of

  mud pressure pulses emitted by the team-

at the bottom of the borehole, this system involving the formation (by the equipment provided at the bottom of the hole) and the transmission of a group of at least three mud pressure pulses unequally spaced for each pulse individual

  transmitting information, as well as 1 T installa-

  appropriate equipment at the surface to detect and translate groups of transmitted pulses. The new features of the

  invention will be described more particularly

  below. Implementation and implementation

  of the invention, as well as other objects and advantages

  will be better understood when reading

  of the following description of certain examples

  given by way of illustration with reference to the accompanying drawings, in which Figures 1A, 1B, 2A and 2B are

  partly referring to the description

  of the prior art and in the field of

  this invention; these figures, as well as the others

  also refer to the summary of the invention,

  as well as the description of the embodiments

favorite;

  Figure IA shows schematic

  the operation of a slow-acting valve of the type suggested in the prior art; FIG. 1B schematically represents the variations

  pressure detected on the soil surface and results in

  both of the operation of the valve represented by

  gure 1A; FIGS. 1A and 1B describe a state hereinafter referred to as the "regime of slow pressure variations"; Figure 2A shows schematically

  the operation of a quick-action valve

  the present invention; FIG. 2B diagrammatically represents the pressure variations detected on the surface of the

  ground and resulting from the operation of the valve

  shown in Figure 2A; Figures 2A and 2B describe a state hereinafter referred to as "hydraulic shock wave regime"; Fig. 3 schematically and generally shows a well drilling system equipped for simultaneous drilling and measurements in accordance with certain aspects of the present invention; FIG. 4A schematically represents a part of the equipment in the basement comprising a special telemetry instrument according to the present invention; Figure 4B schematically shows part of the equipment illustrated in Figure 4A; Figure 5A shows schematically

  and in more detail the assembly of

  electronic element in the dashed rectangle of FIG. 4A; FIG. 5B schematically shows a power source comprising a capacitor charging and discharging system for providing the power and energy required for

  actuate the flap of the special tele-

measured;

2 447511

  FIG. 5C schematically represents the electronic circuit that can be used to ensure the automatic shutdown of the drive source of the valve of the special telemetry instrument, while FIGS. 5D and 5E are graphs intended to facilitate the explanation of the cutoff.

  automatic control of the driving source of the signal

  lisation; Figures 6A, -6B and 6C show

  schematically the operation of the hydrau-

  "automatic closing" of the signal flap

  lisation; Figure 6D is a technical drawing of the system shown in Figures 6A, 6B and 6C; Figure 6E shows schematically

  an "integrated security" electronic system

  cable to the check valve; Fig. 7A schematically shows the structure of the housing and the "protector" for the special telemetry instrument, Fig. 7B schematically shows the shape of the cross section of the centering elements that can be used with the structure shown in Fig. 7A; FIG. 7C shows schematically

  special assembly elements that can be used

  adapted to connect the subsections of the housing portion 250b of Fig. 7A; FIGS. 8A to 8E are graphs representing the pressure variations measured at the ground surface and corresponding to different values of Tav) (opening or closing times of a valve) and of Tbv) (open flow time) ), the graphs of these figures show the results of some tests that have been performed by the Applicant to obtain the optimum conditions for a hydraulic shock wave regime; more specifically, FIGS. 8A to 8E can be described as follows:

  FIG. 8A corresponds to T (V) = 1 se-

  conde and TbV) = 2 seconds; the millisecond figure and T b (v) the millisecond figure and Tb (v) the millisecond figure and Tb (v) the millisecond figure and Tb (v) Figure 8B corresponds to Ta vJ = 20 (= 1 second; 8C corresponds to Ta (V) - 60 = 0.5 seconds, 8D corresponds to Ta (V) = 20 = 0.25 seconds, 8E corresponds to T (v) = 5 - a = 10 seconds, 8F is an exact reproduction of the signal of pressure indicating a valve wave received at the surface from a depth of 2,980 m during the actual drilling of an oil well

  in the East Texas region of the United States of America

America;

  Figure 9 is a sketchy illustration of

  representing specific equipment in front of

  be used on the ground in conjunction with a

  signaling system operating by pulses of pressure emitted from the bottom of a borehole in accordance with the present invention;

  device comprising a filter adapted for the purpose of

  erase erratic noises when they are white noises; FIGS. 10A to IOG illustrate, by means of graphs, certain waveforms and certain time-varying pulses,

  these figures being given in order to facilitate the ex-

  the operation of the equipment

  felt in Figure 9; in FIGS. 10A-1OC, as well as in FIGS. 1OD-IOG, the time-related axes are placed one below the other so that these 24751l 'signals and these waveforms can be compared in their relation Mutual as a function of time; more specifically, FIGS. 10A to 1G can be described as follows

  Figure IOA consists of three graphs

  S ques showing three components of a signal detected at the top of a borehole; these components respectively represent a signal carrying a

  information, a pump noise or, when

  reads several pumps in tandem, noise from the pump group and erratic noise; Figure lOB has three graphs

  representative.respectively the delayed signal

  information, delayed pump noise and delayed erratic noise; the delay is Tp representing the operating period of the pump (when using several pumps, although not sinusoidal, the pressure variations are always periodic, since the tandem pumps are kept relatively close to the "in" state. phase"); Figure 1OC shows two graphs

  showing respectively the differences in

  corresponding numbers of FIGS. 10A and 10B; a die

  these graphs represent the erratic noise, tan-

  say that the other graph represents a signal carrying information; Figure lOD illustrates a function representing the output of a digital filter or cross-correlator used in the embodiments of the present invention; this function is virtually analogous to that

  representing the signal carrying information

  glossy in FIG. the digital filter used in this case can be a matched filter, a pulse shaping filter or a tip forming filter; FIG. 10E shows a function similar to that of FIG. 10D, but delayed by a suitable period of time; FIG. 1OF represents a function similar to that of FIG. 10E, but inverted in time; FIG. 1OG represents the resultant of a comparison of the graphs of FIGS. 1B and F and shows the moments corresponding to the pulses appearing in coincidence in these FIGS.

graphics; -

  Figure It shows schematically

  certain operations carried out by a digital filter

  America; Figure 12 schematically shows a system for storing an information carrier signal or a noise signal; Figure 13 schematically shows a portion of the surface equipment including a correlator for noise suppression; FIG. 14 schematically represents a portion of the surface equipment comprising a

  filter adapted for the suppression of noises

  that these are not white noises; Figure 15 schematically shows a portion of the surface equipment having a pulse shaping filter; Figure 16 schematically illustrates

  certain operations performed by a filter con-

  pulse trainer; Figure 17 schematically shows a portion of the surface equipment comprising a tip forming filter, the latter being used to transform a double valve wave into a corresponding pair of tips; Figs. 18A-18F show six possible choices for a peak offset in a pair of peaks produced by means of the system shown in Fig. 17; Fig. 19 schematically shows a portion of the surface equipment comprising a tip forming filter which is used to transform an individual valve wave into a single corresponding tip; Figs. 20A to 20F show six possible choices for a single tip offset produced by means of the system shown in Fig. 19; FIGS. 21A to 21C schematically represent certain operations associated with a

  spike formation filter for different

  tards; more specifically, Figure 21A corresponds to

  lays at a desired peak at the time index '; Fig. 21B corresponds to a desired peak at time index 1 and Fig. 21C corresponds to a desired peak at time index 2; Fig. 22 schematically shows a system for determining the performance parameter P of a peak-forming filter; Fig. 23 is a graph showing how the filter efficiency P may vary depending on the offset of the peaks for a fixed duration filter; Fig. 24 is a graph showing how the performance parameter P of a peak-forming filter can vary in

  function of the filtering length (or duration of

  storing) for a fixed offset of the tips, FIG. 25 includes graphs showing how the output parameter P of a tip-forming filter can vary depending on the length and the filtering delay;

  Figure 26A shows schematic

  a pulse coding system over time according to the prior art;

  Figure 26B shows schematic

  a time pulse encoding system according to the present invention, a system in which

  which the amplitude of the parameter to be trans-

  mis is represented by the amount of time flowing

  between short individual pulses successively

  are of virtually constant duration; Fig. 26C also schematically illustrates the pulse pulse coding system illustrated in Fig. 26B;

  Figure 26D shows schematic

  a time pulse encoding system of the type illustrated in FIGS. 26B and 26C, but in which "triple group" pulses are used; Figure 27 is a block diagram

  showing a "code translator" used to

  causing a system to receive coded signals as "triple group" pulses over time;

  Fig. 28A is a block diagram

  illustrating in more detail the circuit of selection

  316 of FIG. 27; FIGS. 28B, 28C, 28D and 28E are

  graphics that facilitate explanation and understanding

  grasping the functioning of the circuit of the

gure 28A;

  Figure 29 is a block diagram

  nel of the circuit provided at the bottom of a borehole for emitting the "triple group" pulses of the type illustrated in Figure 26D; FIG. 30 is a diagram showing the principles of the circuit making it possible to realize the

  coding of the pulses in time according to the in-

  vention. It should be noted that in some of the

  figures above, reference figures are

  identical references have been applied to similar elements. In this case, we will not resume the

  description and functions of these elements in

  the extent that this is not necessary to explain

  the operation of these embodiments.

  I. GENERAL DESCRIPTION OF THE TRANSMISSION APPARATUS

DATA SESSION DURING DRILLING.

  Figure 3 is a general view of a

  specific system in which the principles are adopted

  principles of the present invention. The turnover figure

  FIG. 20 denotes a conventional tower for the

  oil tower, this tower comprising a rotary table 21, a driving rod 22, a pipe 23, a riser 24, a rod of

  25 and a drill collar 26. One or more

  mud pumps 27 and a mud pit 28 are connected in the usual way to feed the

  riser in mud - drilling under pressure.

  The high pressure sludge is pumped down into the drill string via the drill pipe 25 and the conventional drill collars 26, then to

  through the special telemetry instrument 50,

  that drill bit 31. This drill bit 31 includes the

  habitual planes throwing drilling jets and

  schematically represented by the reference numeral

  33. As represented, the diameters

  very necklaces 26 and the tele-

  measure 50 are exaggerated compared to that of the

  drill pipe 25 to illustrate more clearly

  mechanisms. The drilling mud goes down through the drill string as represented by

  the arrows illustrated, then it goes back

  to the annular space formed between the drill pipe and the wall of the well hole. When it reaches the surface, the mud is spilled again

  in the pit (by means of unrepresented pipes

  ssentées) o rock fragments and other well debris can decant, then be filtered before mud is taken and returned

  in circulation by means of the pump or pumps.

  Between the trephine 33 and the collar of

  drilling 26, the special assembly of the trans-

  telemetry transmitter or telemetry instrument

  referred to by reference numeral 50. This assem-

  50 telemetry transmitter

  takes a housing 250 enclosing a valve assembly or more simply a valve 40, an assembly

  electronic processing 96 and sensors 101.

  The valve 40 is designed to temporarily derive a certain portion of the sludge from within the

  drill collar in the annular space 60. Nor-

  (when dapplet 40 is closed), the total

  The drilling mud must be driven off by the jets 33 and, consequently, the sludge is under a very high pressure (of the order of 140 to 210 kg / cm 2) in the riser 24. When the valve 40 is opened under the control of a sensor 101 and the electronic processing assembly 96,

  some of the sludge is drifted

  the total resistance to flow is

  mentally reduced and a pressure change

  can be detected in the riser 24.

  The electronic processing assembly 96 transmits

  a series of coded electrical pulses repre-

  sensing the parameter to be measured by a selected sensor 101, while corresponding openings and closures of the valve 40 take place under the effect of the corresponding pulses

- 75111

  resulting pressure in the column

aunt 24.

  Reference numeral 51 denotes a pressure transducer producing a voltage representing the pressure changes occurring in riser 24. This signal represents these pressure changes.

  is processed by an electronic assembly 53 emits

  both suitable signals to be recorded in a recorder 54 or any other

  display apparatus. The graph of the record

  54 is driven by a command representing the depth of the bit by well known means

(not shown)

  II. GENERAL DESCRIPTION OF THE SPECIAL TRANSMITTER

TELEMETRY.

Figure 4A shows some

  details of the special telemetry transmitter 50.

  Some of these details, as well as others, have also been described in the aforementioned U.S. Patent Application No. 857,677

  submitted by S.A. Scherbatskoy to whom

  This request follows. Figure 4A is a schematic view. In a real instrument, the housing 250 which encloses the valve 40, the electronic processing assembly 96 and the sensors 101, is subdivided into two sections 250a and 250b. The upper section 250a (above the dotted line 249) contiguous the valve assembly-40 and the mechanical elements associated therewith and,

  As will be highlighted below, this section has

  a diameter substantially greater than that of section 250b. The lower section 250b - (in des-,

  under the dashed line 249) contains the

  the appearance of electronic processing 96,

  101 and the mechanical elements associated therewith and, as will be explained below, it has a diameter substantially smaller than that of the

  upper section 250a. As shown in

  4A, the drilling mud goes down into the

  special telemetry system 250a, 250b (as indicated in

  the arrows 65), then it passes through the

  33 and returns (as indicated by arrows 66) to the surface

  in the annular space 60, to finally terminate

  in the pit 28 via pipes (not shown). The valve assembly 40 includes a rod 68 and a seat 69. This rod and seat

  of valve are made in such a way that the sec-

  cross section of the closure A is slightly larger than the cross section B of the piston

  70. Accordingly, when the

  In the chamber 77 is greater than that prevailing in the chamber 78, the valve stem 68 is driven downwards, the valve 40 tending to close more and more hermetically.

  as different pressure is applied

stronger.

  The fluid pressure (sludge) prevailing in the chamber 77 is, at any moment, substantially equal to the fluid pressure (sludge) prevailing in

  the drill collar designated by the reference number

  number 26 in Figure 3 and the reference numeral

  50, in Figure 4A, because of the

  77a in the wall of the assembly.

  250. A fluid filter 77b is interposed in

  passage 77a to prevent debris and

  solids to enter room 77.

  When the valve 40 is closed, the fluid pressure (sludge) prevailing in the chamber 78 is equal to the fluid pressure (sludge) prevailing in the annular space 60. When the valve 40 is open and the pumps are operating, the mud flows from the chamber 77 to the chamber 78 and, via the orifice 81, it reaches the annular space 60 through

  undergoing corresponding pressure drops.

  An electromagnetic solenoid double effect 79 is provided to open or close

  the valve 40 in response to the distributing electrical current

fired by drivers 90.

  It will be assumed that P6o indicates the pres-

  mud flow prevailing in the annular space 60, P77, the pressure prevailing in the chamber 77 and P78, the pressure prevailing in the chamber 78. Therefore,

  when the valve 40 is closed, we have P78 = P60.

  When the pumps 27 operate and the valve 40 is "closed" or almost closed and that P77> '78,

  the valve stem 68 is pushed towards the seat 69.

  When the valve 40 is in the "open" position (i.e. when it is moved upwards

  in the drawing), the mud then circulates

  77 to the annular space 60 and, because of the resistance offered to the flow by the orifice C (FIG. 4B), the following relation is: P77 P78> P60- The chambers 83 and 94 are filled with an oil of very low viscosity (for example, "Dow Corning 200 Fluid" oil, preferably with a viscosity of 5 centistokes or less) and

  are connected to each other by the passage 86.

  Thanks to a floating piston 82, the pressure P83 prevailing in the oil-filled chamber 83 is, at any moment, equal to the pressure P78.

  at any time, we have P 78 = 83 P84 Encountered-

  Therefore, when the valve 40 is "open", since P84 and P77> P84, the valve 40 is pushed to the open position "under a force F = (zone B) (P77-P84). we can say that the valve 40 is "bistablell,

  that is, when it is "open", it tends to

  to remain "open" while, when "closed", it tends to remain "closed". In addition, when it is almost open, it tends to move to the open position and when it is almost closed, it tends to move to the closed position. Consequently, this valve 40 can "switch" from one state to another with an energy

  relatively weak. It can be considered that the function

  The valve is the mechanical equivalent of the well known electric bistable multivibrator

the technique of electronics.

  FIG. 4B shows the valve 40 in position

  while in Figure 4A it is

closed position.

  Referring again to FIG. 4A, reference numeral 91 indicates an electrical "pressure switch" which is electrically conductive when P77> '78 (the pumps are operating), while it is non-electrically conductive when

  P77 = P 78 (pumps are stopped). As a result

  quence, the wire 92 going from the pressure switch

  91 to the power source 93 can put the cou-

  in or out of circuit. Similarly, by means of a

  electronic counter 94 and a sequential switch

  electromagnetic system 95, any of the four

  The sensors 101 can be connected effectively to the

  electronic processing appearance 96 by stopping and sequentially operating the slurry pumps 27, or stopping these pumps, and then operating them according to a predetermined code that can be interpreted by the circuit

mounted in the element 94.

  III. DESCRIPTION OF THE TREATMENT ASSEMBLY

  ELECTRONICS OF THE SPECIAL INSTRUMENT

TELEMETRY.

  The operation of the bistable flap 40 and the sequential switch 95 providing the selective electrical connection of the different

  sensors 101 with the electrical treatment assembly

96.

  For more details on the electronic processing assembly 96, we

  refer to Figure 5A in which the same figures

  references refer to the same elements as

Figure 4A.

  Different types of sensors emitting electrical signals representing an existing parameter at the bottom of a borehole are well known. For example, gamma-ray sensors, temperature sensors, pressure sensors,

  gasses, magnetic compasses, inclinometers in-

  stressors, magnetometers,

  gyrostatic compasses and many other sensors.

  For the example given for illustrative purposes

  In FIG. 5A, a gamma ray sensor such as a ionization chamber or a Geiger counter or a scintillation counter (provided with a suitable electronic circuit) is chosen. All of these elements can be designed to produce a DC voltage proportional to the ray flux

  gamma that is intercepted by the sensor.

  It is understood that the transition from one type of sensor to another under the control of the switching mechanism 95 illustrated in FIG. 4A is well known according to the state of the art (in most cases an electronic commutation

  is preferred to the illustrated mechanical switch).

247511 1

  As a result, in Figure SA, to make the description

* lighter, only one sensor is represented

  1011. Similarly, the power source 93 and the

  mute pressure actuator 91 (shown in FIG. 4A) are not illustrated.

very in figure SA.

  In FIG. SA, the sensor 101 is connected

  cascaded to the analog / digital converter

  102, the processor 103 and the drive source 104. This driving source 104 is connected to the windings 105 and 106 of the double-acting solenorde indicated at 79 in Figure 4A. This driving source 104 may be similar to that represented by the

  Figure 3E of the related patent application. The function

  The sensor 101 emits an analog output electrical signal represented by the curve 10a1 of the graph taken immediately above the box 101 indicating the sensor. This curve represents the output of the sensor based

  the depth at which the radio transmitter

  measure 50 is located in the borehole. The analog / digital converter transforms the

  analog signal from the sensor 101 in a naked form.

  by measuring successively the amplitude of a large number of ordinates of the curve lOla and by translating each individual ordinate into a binary number represented by a binary word. This process is well known in the art and it is not

  necessary to give an explanation here. Toute-

  However, it is important to know that if the graph 10a1 can represent, in hours, the variations of the signal emitted by the transducer, the graph 102a represents a single ordinate (for example, AB of the curve 101b). Therefore, the time scale of

  the x-axis of Figure 102a would be expen-

  in seconds, while the entire graph 102a represents a 12-bit binary word; in fact, it represents the decimal 2649. Therefore, each 12-bit word in chart 102a represents a single

  ordinate such as the ordinate AB of the graph 110a.

  The usual binary coding involves rest times between each binary word. After this time of

  rest, a start impulse or impulse

  is transmitted to indicate the beginning of the

  time period attributed to the binary word. This impulse

  Preliminary is not part of the binary word, but is used to indicate which binary word is about to begin. This binary word is then transmitted and it indicates the value of an ordinate of the graph 10a1; then comes a rest period followed by the next binary word representing the amplitude of the next ordinate and so on,

  in a quick succession. Therefore, the curve

  Continued from Chart 10a is represented by a series.

  numbers or binary words each representing a single point of the graph 1l0a. In this respect, it is important to know that between each binary word there is always a rest period. This idle time (during which no signal is transmitted) frequently has a length corresponding to several binary words and is used for a significant purpose which will be explained below. In order to-allow the

  decoding on the surface, the clock n0 1 must be

  steadily constant (and in synchronism with hor-

  corresponding box 212 or 309 located on the surface) and emits a series of pulses equidistant in time in a manner well known in the art

of electronics.

  The graph 103a represents a single bit of the binary word 102a and, in this case also, the x-axis is very different from that of the previous graphs. The time taken in the

  graph 103a is expressed in milliseconds, since

  this graph is only one bit.

  Each individual bit is converted into two electrical pulses each of a duration tX and separated by a time interval tya. The graph 104a is a replica of the graph 103a which has been considerably amplified by the driving source 104. The electric pulse 104b is applied. at the winding 105 of the solemn

  (which is the winding of the valve in the "open" position

  while the electrical pulse 104c is applied to the solenoid winding 106 (which is the winding of the valve in the "closed" position 1). 104b and it is closed by the pulse 104c, so that it remains in the "open" position approximately during the time t * The tx times are set so that the solenoid windings are controlled properly, while the time ty is calculated to open the valve 40 during the correct lapse of time.These two times are determined and

Controlled by the clock n0 2.

  When telemetry information between a sensor and the ground surface, appropriate rest times are provided between the transmission of the successive binary words. Thanks to these times of

  rest, one can memorize, in an electronic memory,

  appropriate equipment as part of the equipment of

  surface, noise caused solely by the opera-

  drilling (without the damper wave). The systems and processes required for this purpose will be described below.

  IV. DESCRIPTION OF THE POWER SOURCE FOR THE

SPECIAL TELEMETRY TRANSMITTER.

  As has been pointed out previously,

  the valve 40 illustrated in FIG. 4A must be in ac-

  very fast, and for this purpose it requires

  considerable power. (Following appropriate tests

  requested, it was determined that a valve of this t required about 1/2 to 3/4 horsepower to

at the required speed).

  Although this power is very important, it is only applied over a very short period of time and, therefore, it is not necessary to

  only a small amount of energy per operation.

  During actual operation performed during the tests, it was found that a power of

  1/2 horsepower applied for about 40 milligrams

  seconds provided the energy required to ensure a satisfactory single operation of the valve. This energy can be calculated at about 15 Joules. A group of accumulators of sufficiently small size to be housed in the housing 250b of FIG. 7A can provide about 4 million

  of Joules without requiring a refill or replacement

  cement. As a result, the system is able to ensure 130.000 complete functioning of the flap

  (opening plus closing). In reality, the consumer

  energy consumption is less than 15 Joules per

  tioning. Thanks to inductance Q, and to the

  the kinetic dance of solenoid winding, the

  Cumulative current is relatively slow and takes place along a curved slope as represented by curve 272A of Figure 5C and by curves 300, 301 of Figure 6E. Therefore, the total energy per pulse is substantially less than 15 Joules and it was measured at 9 Joules, thus ensuring

  the capacity of 216,000 complete operations of the

  fart (an even greater capacity is obtained by using the circuit described hereinafter with reference to

  in Figure 5C). From the description above,

  we can see that batteries can perfectly

  provide the required energy at the bottom of a

  drilling for a practical telemetry instrument.

  However, the need for the very high power required (1/2 horsepower) poses difficult problems. It is clear that the solution to this problem would involve the accumulation of energy in a mechanism that could be brought to release it abruptly (in a short time), thus providing the necessary short impulses of high

  power. A mechanism of this type is that

  "water hammer" which has been used in the instrument described in the patent application

  related, but which was sometimes insufficient.

  In other mechanisms previously envisaged,

  we used compressed air, springs

  awarded and others. Capacitor energy storage systems require high capacities: the energy accumulated in a capacitor varies according to the first power of the capacitor.

  capacity, as well as the square of the accumulated

  and, since low-inductance and fast-acting solenoid drive windings

  are necessary, it is obvious that we must use

  low voltage devices, the calculations

  have indicated that capacitors

  too important shots would be necessary.

  After further evaluation, it is

  appeared that an operational system could be

  tional. By mathematical analysis, as well as by experiments and tests, it has been determined that the following series of optimum circuit parameters could be established:

  1. Inductance of solenoid winding

  noid: 0.1 henry in operating position and

  0.07 henry in non-functioning position (ie

  ie a solenoid with conical armature).

24751I1

2. Resistance of winding

solenord: 4.5 ohms.

  3. Voltage at which energy is

accumulated: 50 volts.

  4. Capacity of the accumulator capacitor

lation: 10,000 microfarads.

5. Current capacity of the circuit

Attack: 10 amps.

  It has been determined that to ensure rapid operation of the solenoid, low inductance windings are desirable. It was also determined that the current capacities of the

  electronic circuits of attack could be

  well above 10 amps. However,

  low voltage requires far too high

  Vees. Recent progress made in

  so-called molten salt accumulators have provided

  these energy of a very good compactness. The same

  recent technology has also developed

  with extraordinarily high values

  (10 farads in a space as small as 16.38 cm3).

  These capacitors have proved unacceptable because high temperature heating (SOOOC) is required; it was felt that this

  was not feasible without taking into account the cost

  bitif. As a result, efforts must always be made. As a result of extensive and extensive research, it was finally discovered that a tantalum capacitor realized in accordance with

  the latest progress could satisfy the special

  requirements imposed by optimizing the

  parameters and factors listed above to match the characteristics

of these capacitors.

  From the above discussion it can be seen that a battery charge can ensure

  at least 216,000 complete operations of the flap.

  Assuming that the telemetry system can provide continuous and adequate data by the

  transmission of five pulses per minute, the system

  The system is able to operate continuously in a borehole for a period of 440 hours. It should be emphasized, however, that very often, continuous operation is not essential. The instrument can be used only intermittently under the control of the circuit controlled by the switch 91, as well as

  by the elements 94 and 95 of Figure 4A.

Moreover, as will be explained below

  after, when taking advantage of the improved circuit illustrated in Figure 5C, it can achieve a greater number of valve operations. We

  considers that operation on the basis of an impulse

  per second is practical.

  Another parameter that needs to be determined is proper charging of the capacitor after discharge. This capacitor can

  to be loaded by a resistance connected to the

  emulator (or to another energy source), but this procedure is sometimes slow because, when the capacitor is partially charged, the current passing through the resistance decreases and, at the end of the charging cycle, the current of charge is getting close to zero. If the ohmic value of the resistor is low, the accumulators must conduct an excessive momentary current> since the initial shock of current occurring during the charge cycle would exceed the value

  provided for the maximum lifetime of accumula-

  tors. The best method is to load the

  capacitor by a constant current device.

  The capacitor then charges at an optimum current.

  mum corresponding to the optimum discharge current-

  for the particular type of accumulator allowing - a maximum accumulation of energy. By correctly determining the charge current, a significant increase (sometimes by a factor of 2 or 3) in the amount of energy available in a

  given type of accumulator. The devices with

  are well known and are integrated electronic circuits readily available for

wide range of current values.

  FIG. 5B diagrammatically shows a power source that can be incorporated in the power source 104 of FIG. 4A, this source comprising a capacitor charging and discharging system in order to provide the power and energy required for the windings of the power source. solenoid

  79. In Figure 5B, the reference figure 450

  an accumulator, a turbo-generator or a

  another source of electrical potential

  tinu, the reference numeral 451 designates the

  constant current and reference numeral 452, the capacitor. This capacitor is charged by the constant current device 451, while

  that it is discharged via conductor 453-. The con-

  driver 454 provides the regular constant power required for electronic system balance

  installed at the bottom of the borehole.

  V. DESCRIPTION OF THE "CLOSED" SIGNALING VALVE

AUTOMATIC TURE "HYDRAULIC.

  There is also a system

  in case of malfunction that could happen.

  produce when the flap is "glued," in the open position for a long time. A system

  automatic closing of the flap in case of

  This type of failure (indicated by reference numeral 269 in Figure 4A) is shown

  schematically in FIGS. 6A, 6B and 6C.

  As discussed further in this specification, the valve is designed to

  a bistable operation or a hy-

  draulic; in other words, when it is opened by a pulse emitted by the winding 105 of the

  solenoid, it tends to remain in a position of

  verture and, subsequently, when closed by

  a pulse emitted by the winding 106 of the solenoid

  noting, he tends to remain in a position of

  meture. Due to faulty electrical or mechanical operation, the valve may remain in the open position. It is to highlight that

  if this malfunction occurs, the

  drill can continue to unfold. A

  some wear may appear at port 81 illus-

  tré in Figure 4A. However, it is not desirable

  table that the hydraulic circulation system of the sludge is disturbed by the opening of the valve for long periods; Moreover, although the drilling operation can continue, it is very advantageous that the valve remains closed most of the time and that it is open only for

  produce the short pulses required to obtain

the hydraulic shock wave.

  In the schematic drawings of the

  6A, 6B, 6C, the rod 100 is used to

  push the valve in the closed position while exert-

  downward force on the illustrated rod 80

  in Figure 4B (armature shaft of the solenoid).

  Referring now to the figures

  6A, 6B, 6C and 6D, the upper end of the mech-

  is exposed to the "mud lying in the foragel pipe", that is, the mud undergoing

  the hydrostatic pressure plus the different pressure

  bit, or in other words, the difference between

  pressure between the inside of the instrument

  and the annular space 60. If the pumps do not function

  the pressure prevailing in zone 111 is only hydrostatic; Moreover, when

  the pumps work, the pressure is not only

  hydrostatic, but also differential.

  Since the differential pressure is of the order of 70 to 140 kg / cm 2, a large pressure change occurs in zone 111 during the

  start-up of the pumps (ie

  pressure range of 70 to 140 kg / cm 2). In FIG. 6A, if the pumps do not work, the zones 112,

  113 are subject to the pressure prevailing in the space

  nular, since the tube 114 is connected to the chamber 84 which contains oil under the pressure prevailing in the annular space (see Figure 4A) and also because the orifice 115 connects the

areas 112 and 113.

  It will now be assumed that the pumps are turned on. The pressure prevailing in the zone 111 then rises substantially (that is to say from 70 to 140 kg / cm 2), the piston 116 is pushed down by compressing the spring 107 (not shown

  in Figure 6B), while the oil under high pressure

  In the region 112, the piston 108 pushes the piston 108 downwardly compressing the spring 110 (not shown). As a result, when the pumps are turned on, the elements shown in FIG. 6A assume the configuration illustrated in FIG. 6B, while the two pistons 116 and 108 occupy the position at the bottom of the stroke, the rod 100 also extending towards the down as shown

in the drawing.

  24751l41 Now, because of the orifice 115 and the action of the spring 110, the piston 68 is pushed

  upward at a speed determined by the size

  orifice 115, the spring restoring constant 110 and the viscosity of the oil present in the zones 112 and 113. This speed can be easily adjusted and brought to any desired value; for example, it is possible to set a speed at which the piston 108 returns to its initial position at the top of the race in a period of about

  1 minute. As a result, after one minute, the system

  tem takes the configuration shown in Figure 6C.

  For the same reasons, when the pump is stopped, under the action of the spring 107 and the orifice 115, the piston 116 goes back to its initial position

illustrated in Figure 6A.

  As a result, it can be seen that each time the mud pump starts, the rod moves downward a distance "d" as shown in Figure 6B, and then returns to its normal retracted position. Since, during a normal drilling operation, the pump is stopped each time a drill pipe joint is added, at this time (usually 9 m lengths), the rod 100 executes a single stroke. down, to return then to his

initial top position.

  As has been pointed out previously,

  the rod 100 is designed so that, when

  it is in extension downwards, it pushes the stem 80 of the solenoid armature (FIG.

  the bottom by closing the flap. Therefore, the

  FIGS. 6A, 6B, 6C and 6D are a "security device," that is, if,

  the valve remains "stuck" in the open position

  due to an electrical or mechanical malfunction, the valve is then necessarily forced to close after a maximum length of

9 m drilling.

  Figure 6D is the technical drawing of the device schematically illustrated in Figures 6A, 6B and 6C. In the real instrument, the

  device shown in Figure 6D is placed in the

  right 269 shown in FIG. 4A. In FIG. 6D, the same reference numerals are used to designate the similar elements of FIG. 6A.

6B, 6C and 4A.

  VI. DESCRIPTION OF THE ELECTRONIC SYSTEM WITH SAFETY

  INTEGRATED "FOR THE SIGNALING VALVE.

  The hydraulic "automatic closing" system described with reference to FIGS. 6A, 6B, 6C and 6D will automatically close the valve each time the sludge pumps are stopped and restarted, thus making it possible to remedy any mechanical sticking of the valve. However, it is a case requiring even more attention:

  the "closing" electrical circuit 103, 109 illus-

  in figure S.5A suffers a failure for

  for whatever reason (for example, a roll-up

  burned solenoid), the valve then reopens electrically shortly after the device

  hydraulic "automatic closing" closed it.

  FIG. 6E shows an electrical system preventing the valve from operating in

  case of electrical failure of the device

  installed at the bottom of the borehole. The reference S designates the winding of the solenoid which "closes" the

  valve, while the reference S2 designates the coil

  the solenoid which "opens" it. The resistor R1 is connected in series with the part of the circuit

  104 of the solenoid which drives the winding

  "closure" S of the latter. The resistor R2 is connected in series with the part of the circuit

  104 of the solenord that drives the wind

  The resistors have a very low ohmic value (about 0.05 to 0.2 ohm) and it is understood that the operation of the system described in detail with reference to Figure A of FIG. This specification is as follows: the "opening" electrical current pulse is

  issued in the first place and is represented

  in Figure 6E by pulse 300; the "closing" electric current pulse is produced later (after a time t y) and is shown schematically at 301 in FIG. 6E. It should be noted that these electrical pulses 300 and 301 represent the current delivered by the

  solenoid windings and not the applied voltage

  the resistors R1 and R2 produce the voltage drops R1i, and R2i2, i1 and i2 indicate the currents flowing through the respective windings of the solenord, therefore, if one of the windings S1 and S2 is burned or is in open circuit, no current does not pass and no corresponding pulse

  produced (in the same way, in the event of any

  what other electrical failure, the

  will not go through one or both resistors

R1, R2).

  The time interval t in FIG. 6E, as well as the time period tx, have been explained.

  and defined above with reference to FIG.

  The delay of the element 302 is equal to t y. In other words, at the exit of the element

  302, an impulse identical to the impulse

  input, but delayed by the time t y. These delay systems are well known and do not require

here no description.

  Since the delay of the element 302 is equal to ty, the pulse represented by the reference numeral 303 will be in coincidence with

* time with pulse 301.

  Circuit 304 is an anti-coincidence circuit (also known as an "OR" gate) and its output 305 emits an electrical signal only when it receives one of pulses 301, 303, but does not transmit no output signal

  when the two pulses 301 and 303 are pre-

  paths. Reference numeral 306 designates a

  relay actuated by the signal conveyed by the con-

  305, this relay being designed to cut the power supply to the instrument installed at the bottom of the borehole. Therefore, if

  there is only a "closing" impulse (without the impulse

  opening pressure ") or if there is only one opening pulse" (without the "closing" pulse), the power supply to the power source installed at the bottom of the borehole is cut and then , the hydraulic "automatic closing" system described with reference to FIG.

mechanical closure.

  As an alternative to the system illustrated in FIG. 6E, the relay 306 (which may, of course, be an electronic switch having transistors) may be designed to cut off the current supplying only the circuit provided for the opening solenoid ". some

  advantages, since both the

  "One of the objectives envisaged is to ensure the" closure "of the valve, and an electronic meter 314 can be installed between the" OR "circuit and the relay 306 so that the power supply of the valve current does not

  not be cut by a single malfunction

  electricity. For example, the power will be cut off only after 2, 4 or 8 operations

successive defective.

  VII. DESCRIPTION OF THE AUTOMATIC CUTTING SYSTEM

  OF THE MOTOR SOURCE OF THE SIGNALING VALVE.

  As pointed out earlier in this specification, a very

  The quick release of the valve 40 illustrated in FIG. 4A is important.

  As. The required shock wave will not be produced if the damper operates slowly. Given

  that the valve and its control mechanism represent

  feel a considerable mass, a strong power is needed to open or close this valve at the desired time. This power is of the order of 1/2 to 3/4 horsepower and can be provided by a power source of the type described in Chapter IV of this specification. As in all systems of this type, a power margin is required to ensure at all times

  the opening or closing of the valve on order.

  The different electronic "logic circuits", as well as the different "power source circuits" shown in FIG. SA are designed to emit rectangular voltage pulses

  104B and 104C with a duration of approximately 40 to 50 milli-

  seconds to ensure the excitement of the windings

  105 and 106 solenoid for a period of time sufficient to allow the operation of the valve. FIG. 5E shows in more detail the voltage pulse 104b of FIG. At moment 0, the voltage is applied suddenly by

  the driving source 104 and it rises almost

  up to the value indicated by the

  fre 270, then it stays at that value for 50 milliseconds and then it is cut off and falls back (again almost instantly) to the

value 0.

  It is very instructive to study the movement of the valve by measuring the current flow in the solenord's drive winding and building a graph (see Figure 5D). From this graph, one can study quantitatively the behavior of the valve. FIG. 5D represents a graph of this type in the form of an oscillogram of the current as a function of time. (This current is measured, for example, by the voltage applied to the

  resistor R1 or R2 of Figure 6E).

  It is important to understand that it is the current passing through the winding of the solenoid that determines the force exerted on the stem 68 of the valve (represented in FIG. 4A), since both the

  amperes determine the electromagnetic attraction

  tick. Since the windings of the solenoid

  northe have an inductor, the current will not accumulate

  not instantly when applying a

  suddenly as shown in FIG. 5E. If the solenoid consists of a simple inductor, the current will accumulate then according to a simple exponential curve 271 shown in dashed lines in FIG. 5D. In reality, there is a very different phenomenon: when the valve works (opens or closes), there is a sudden movement of the solenorde armature 79 of Figure 4B

  and a counter-electromotive force is produced.

  This counter-electromotive force is caused by the speed of the armature which rapidly changes (elevates) the inductance of the winding in cause of the solenoid 79. In FIG. 5D, the curve 271 is the approximate curve of the current as a function of the time in the winding of the solenoid 79 when the armature of the latter and the stem 68 of the valve are

  "blocked" in the "open" or "closed" position.

  The solid line curve 272 of FIG.

  present the actual accumulation of current is pro-

  when the flap is not blocked, that means

  ie under actual operating conditions (opening or closing). Curves 272 for

  openings or closures are very similar.

  It is found that after the application of the voltage, the curve 272 rises progressively (due to the inductance of the respective windings 105 and 106 of the solenord) until it reaches, in the example illustrated, the value 4 amps at the moment

  T which is 20 milliseconds. Then, it happens

  a sudden drop in current reaching the

  their lower than 2.2 amps at the moment T1, which is

  25 milliseconds. After the moment T1 = 25 milli-

  seconds, the current rises again according to the usual "exponential" curve until it reaches, asymptotically, the value

  about 10 amps after the lapse of time

  about 60 milliseconds (this value is determined by the resistance of the solenorde winding which,

  in the example given, is 4.7 ohms).

  Examining the curve 272 of FIG. D, it will be seen that the valve 40 begins to open or close at the moment TO0 = 20 milliseconds and that it completes its movement at the moment T1 = 25 milliseconds. As has been pointed out above, an almost identical situation occurs during the "opening" or "closing" of the flap and

  curve 272 indicates that at the moment T = 20 milli-

  seconds, the flapper starts its movement, which

  is completed at time T1 = 25 milliseconds.

  It is important to note that the moment T1 = milliseconds shown in FIG. 5D is given as a specific example and depends on a number of factors. Thus, at high differential pressures, T1 will be greater than 25 milliseconds and may reach 30, 35 or 40 milliseconds. Just stipulate that the moment

  T1 in Figure 5D indicates the moment at which the function

  of the valve has been completed and which flow between T1 and 50 milliseconds is actually

  ndifferent, since both the functioning of the

  pet is already finished. This extra time is a "safety factor" that, even under adverse conditions, the valve always works

  when the voltage pulse is applied.

  According to the present invention, it is

  reads the signal transmitted at the time T1 to cut off all power supply to the solenorde 79. As a result, all the current

  between the moment T and the time of 50 milliseconds.

  (thus substantially reducing the total amount of energy required to operate the valve). It should be noted that the "safety factor"

  above question is maintained in its entirety.

  gralité; the current continues to be applied until

  that the functioning of the flap (opening

or closing) ends.

  The electronic circuit used to achieve the above purpose is illustrated in Figure 5C in which the reference numeral 104 designates the power source illustrated in Figure 4A. Between the driving source 104 and the earth, is interspersed

  resistance (R1 or R2) of low value

  relative to the resistance of the solenord), for example, 0.2 ohm. As a result, the voltage applied to this resistor is proportional to the current supplying the particular winding 105 or 106 of the solenorde. (Two circuits of the type shown in FIG. 5C are necessary: a first for the driving source of the opening solenorde and a

  second, for the driving source of the iron solen

  meture; however, to simplify the illustration, only one circuit is shown in FIG. 5C). The reference numeral 273 designates a conventional amplifier and, at the output of the latter, the voltage curve 272a of FIG. 5C will be a replica of the curve 272 of FIG. 5D. Reference numeral 274 denotes a differentiator (well known in the electronics art) producing an output voltage proportional to the first time derivative of its input voltage. Curve 275 represents this derived voltage. By observing the curve 272 or 272a, it can be seen that the derivative (slope) of the curve is always positive, except for the lapse of time between T0 and

  T1 in which the slope (derivative) is negative.

  On curve 275, only pulse 276 is negative.

  ve. Reference numeral 277 represents a

  classic trainer designed to let pass

  that the pulse 276 as represented in the graph

  The reference numeral 279 represents a delay electronic circuit (also well known in the art) emitting an output pulse 276b which is a replica of the pulse.

  input, but which is delayed by T1-T0.

  Therefore, as represented in the

  graph 280, the pulse 276b is emitted lightly

  beyond the moment Ti. This pulse 276b is applied to an electronic switch 281 which is designed to cut off power to the power source 104, thereby stopping the flow of power.

  running almost immediately after the

  valve 40 (opening or closing). This com

  Electronic mutator 281 is designed to put the power source 104 back into service after a suitable period of time. This process is repeated when

4 75 1 11

  the next pulse emission 104a (or 104b).

  It is important to note that the energy savings that can be realized by exploiting this aspect of the present invention can be very important. Since very strong powers

  important to operate the

  pet 40 at the required high speed, these savings are very sensitive and, in the illustrated example,

  they could increase up to five times

lifetime of accumulators.

  VIII. DESCRIPTION OF THE STRUCTURE OF THE PROTECTOR

  AND HOUSING FOR THE SPECIAL INSTRUMENT

TELEMETRY.

  An important feature of the

  the same measurement during drilling according to the

  invention is its practical character, that is,

  that is, the convenience and ease with which it can be adapted to drill string, as well as existing instruments and equipment for drilling oil wells. In prior art tests, large special dwellings

  steel of a length of 9 m or more and a diameter of

  203mm meter are required to enclose complicated instruments; Moreover, transporting these dwellings from one place to another requires

  special vehicles. In the apparatus of the pre-

  In this invention, since no flap mechanism is interposed in the main sludge stream, the special housing can be removed from a large weight, a very long length and a high price (about 9 m). long) and only a short section of drill collar

  we call it "protector"). In the form of real

  As a practical embodiment of the present invention, this protector has a length of 914 mm and a diameter of 171 mm (instead of the length of 9 m

  which was necessary previously).

  Consequently, one of the characteristics

  important aspects of the present invention lies in

  in the fact that one does not have to use

  special items of high weight and length. This characteristic is advantageous, in particular when it is necessary to obtain, at the bottom of a borehole, magnetic measurements such as compass indications (for example,

  drilling in a deviated hole), which necessitates

  use non-magnetic drill collars. The

  non-magnetic drill collars are not only

  significant weight (2 to 3 tonnes), but also

  an extremely high price (each 20,000

  dollars from the United States monetary system of

  However, they must be made of a strictly non-magnetic material such as "Monel K". In the construction of the apparatus of the present invention, standard "API" I (= "American Petroleum Institute") drill collars having outer diameters are used.

  from 152 to 228 mm (the most

  RANTES). All standard TAPIII collars

  have an inside diameter of 71.4 + 1.587 - 0 mm.

  Simplicity, small dimensions and consistency

  coaxial truction of the valve system of the pre-

  invention and associated elements

  special feature the entire driving source in question and associated equipment may be installed in a pressure-resistant tube of sufficiently small diameter to be able to be inserted into the inner bore ( 71.4 mm) of a standardized drill collar "API" without unduly disturbing the

  mud flow. Some sensors can be

  as close as possible to the drill bit.

  In particular, a gamma-ray sensor installed 4? 511i at the bottom of the borehole must be able to

  detect the penetration of the bit into a forma-

  lithology given as soon as this penetration takes place. In addition, some sensors such as a compass inclinometer installed at the bottom of a borehole, require precise indexing vis-à-vis the "face of

  the tool "that is used in directional drilling

  nel. In addition, a compass inclinometer must be installed at a significant distance from any magnetic or paramagnetic material. Moreover, when using a compass inclinometer, the housings 250a and 250b of FIG. 7A must be carefully indexed angularly with respect to each other.

  protector 253 which, in turn, is indexed

  to the "curved protector" t used in directional drilling. The "curved protector" has an indexing mark 253a and the angle of this mark must

  to be in a constant angular relationship and

  The known angle (representing the angle between the indexing marks 253a and 254a) is then introduced into the calculation involved the range and angle to a vertical plane

of the 'courbet protector'.

  Fig. 7A is a diagram showing the special telemetry instrument 50 and illustrating the system in which the "special long tool" is removed and in which also only a short section of drill collar is needed, as well as it has been mentioned previously. In Fig. 7A, a housing designated by reference numeral 250 is constituted of an upper portion 250a and a lower portion 250b as described above with reference to Fig. 4A. The upper portion

? 475,111

  250a is locked inside a short protection

  254 (length: approximately 914 mm only).

  This short protector is specially bored to an inside diameter (for example, 114 mm) sufficient to house the valve assembly 40 and also to allow free flow of the drilling mud into this upper portion 250a via the

  passages 61 which are also represented in

  Figure 4A. The housing 250a has a small diameter (preferably an outer diameter of 68.2 mm or less). A user-supplied drill collar 255 (oil company or drilling contractor) is usually 9 m long,

  while its weight and price are important.

  As noted above, the diameter

  For example, the standard API "frage" collar is 71.4 - 0 + 1.587 mm. Centering members 256 are provided for the lower housing 250b. These members have a diameter slightly smaller than the inside diameter of the standard drill collar

  I "API" I, for example, an outer diameter of 69.8 mm.

  A weak game is very important to prevent the "tbroutagell" when the instrument is subjected to

  vibrations during drilling. The passage of evacuation

  85 is identical to that shown in Figure 4A. The housing 250b is suspended inside the protector 254 by fasteners (not shown). The cross section of the centering members 256 (as shown in Fig. 7B) is designed to provide slots or passages 258 for free flow.

drilling mud.

  In reality, the lower section 250b

  housing consists of several sub-sectors

  which are connected to each other by a special element shown in FIG. 7C. As

24751 1

-52

  this figure 7C shows, each sub-section

  door, at its upper end, a groove 260 and, at its lower end, a projection or a tooth 261. The projection 261 of a sub-section is engaged exactly in the groove 260 of the adjacent subsection. The adjacent subsections are held by an assembly sleeve 262 which fits exactly in the ends of the subsections. Circular openings 263 made in the subsections are aligned with respective threaded openings 264 in the assembly sleeve 262 and the members are secured by screws 265. The special assembly member illustrated in FIG. precise angular indexing when the protector

  253 is a "curved protector". As we have

  line, the angle formed between the markers

  253a and 254a must be known in order to calculate the angle formed with respect to the vertical of the "curved protector". It is also necessary to know the angular displacement between the axes of a magnetometer / inclinometer and the mark 254a, this angular displacement not having to vary.

  during the drilling operation (it is preferable

  but not essential that the angle formed between

  one of the horizontal axes of the magnetometer / inclination

  meter and the indexing mark 254a is equal to zero).

  For this purpose, the instrument 250b is assembled to the angular indexing teeth 261 as shown in FIGS.

Figures 7C and 7A.

  To achieve an effective telemetry system, two conditions must be

  consideration. One of them relates to

  optimal conditions to obtain the shock wave regime. Hydraulic. The other condition

  is concerned with obtaining shock waves from an inten-

  sufficient to remove the effects of

foreign noises.

  IX. OPTIMAL CONDITIONS FOR DETERMINING THE

  HYDRAULIC SHOCK WAVE SYSTEM (DETERMIN-

  NATION OF PARAMETERS K1 (or K2) and TbV) A series of experiments were conducted to determine the optimal conditions for

  the regime of hydraulic shock waves.

  The appearance of a hydraulic shock wave

  It is analogous to that of the so-called water hammer effect. By abruptly stopping the flow in a localized area of a pipe, one

  suddenly raises the pressure in this area. This-

  the initially localized rise in pressure propagates along the pipe in the form of a "water hammer". We know perfectly well that

  abrupt and localized change (reduction or increase

  velocity) produces a corresponding localized change (rise or fall) of pressure while, on the other hand, a sudden change in

  localized pressure creates a sudden change

  that and localized speed. Because of the elasticity

  ticity and fluid inertia, the change that occurs, is also transmitted from the element

  where it emanates, to the volumetric elements

  neighboring metrics at a speed which is the velocity of compression wave propagation. The problem of propagation of shock waves is

  extremely complex. To meet the conditions

  practical requirements, it is necessary to

  complete a parameter that is most representative for obtaining a clearly defined shock wave. Two parameters will be considered which will be called "parameter K" and "parameter K2". When one or other of these

  parameters exceeds an appropriate value, we ob-

  holds a clearly defined shock wave.

(a) K1 parameter

  This parameter is the average of the

  velocity of the sludge flow through the bypass valve during the cover (or closure) period of the latter: either V (t) the velocity of the sludge flow through the bypass valve, varying according to the time (in cm / second). At the moment t = 0, when the valve starts to open, the speed is equal to O or V (O) = 0O At the moment t = T (v), when the valve is completely open, the speed of this valve reaches a certain value Vf which is the derivation speed during the total flow period. Therefore: V (Ta (V)) = Ve (10) Consequently, the parameter K1, which is the average of the change of velocity during the period Ta (V) is K = Vf ()

1 T V

  a K1 is measured in cm / sec. It will be assumed that when K1 exceeds an appropriate threshold value, that is, when

K1> M1 (12),

  a clearly defined shock wave is obtained. In the experiments carried out, it was determined that MI 2 x 105 cm / sec 2 (13) (b) Parameter K 2 This parameter represents the average of change from the opening section of the flap to

during the period T (V).

at

2 4 7 5 1 1 1

  Previously (see equation (1)), S (t) has been defined as the section of the valve which is open at time t. Therefore, at time t = O, we have S (O) = 0 and at the moment t = Ta (V), we have (y) S a S (Ta (V)) = S (14) where SO indicates the total opening section of the valve. Parameter 2 is: SO K2 = cm2 / sec. (15) Ta It will be assumed that when K2 exceeds an appropriate threshold value, that is, when

K2> M2 (16),

  a clearly defined shock wave is obtained.

  In the experiments carried out, he was determined

than M-100 cm2 / second.

  Generally speaking, K1 is proportionally

  at K2. Parameter K2 is probably more useful because it directly indicates how the valve should be designed and operated. There is also a parameter Tb

  (see B1C1 in Figure 2A) which must be taken into account

  Embodiment for the description of FIGS. 8A to 8E.

  Each of these figures corresponds to a set (V) or e T

  numerical values of K1 and Tb () or K2 and Tb (.

  FIGS. 8A to 8E show the effect exerted by the variations of K and Tb (V) or K K1 Tb oud

  and Tb (V) to switch from the prespecific variation

  slow motion at the hydraulic shockwave regime.

  More specifically, each of these figures shows how the pressure detected at the surface of the ground (ordinate) varies as a function of time, t (abscissa). In these experiments, the dimension of

  the orifice is 0.5 cm2. Experimental data

  have been obtained in a number of

  well. These wells were selected in the United States

  United States of America (Oklahoma, Western Texas and Eastern Texas), as well as the Netherlands. In addition, some of these tests were performed in an "experimental well" that was explicitly drilled

  to practice telemetry experiments.

  By performing the experiments above,

  account has been taken of the wide variety of

  existing plants for pumping sludge, as well as various disturbing effects. There are many types of mud pumps: "Single Duplex" t, "Double Duplex", "Single Triplex", "Double Triplex"; moreover, the pressure variations of the pumps for a given mean sludge pressure are widely different from one installation to another. The removal of the important disturbing signals relating to the pressure of the sludge is complex. Pump pressure signals - emitted by a single "duplex system" can be 10 or even 20 times higher than those emitted by a carefully tuned "Triplex Double" system. In order to determine the optimal values of K2 (or K1) and Tb (V), the drilling operation was stopped and

  used a very good pump "Triplex" (regular).

  Therefore, the graphs of Figs. 8A to 8E are not representative graphs of a typical condition, but they represent a condition in which the various noises (from pumps and other sources) are minimized, to establish then the average by calculation and by drawing a plot to obtain values (y)

  optimal for parameters K2 (or K1) and Tb (V).

  The corresponding values of K2 (or K1) and 3LeTv) Tb (V) for each of FIGS. 8A to 8E are given in the table below:

BOARD

  K2 (in cm2 / second) Tb (v) (in seconds) Figure 8A 0.5 2 Figure 8B 2.5 1 Figure 8C 8.5 0.5 Figure 8D 2.5 0.25 Figure 8E 100 0.1

  The graphs in Figures 8A to 8E

  present average figures obtained in a large

  number of tests. In these tests, the normal pressure

  Rising in the riser was 210 kg / cm2 and pressure variations were in the range of 7 kg / cm2. The above tests were carried out using different types of valves: the valves driven by a motor, the rotary valves, the mushroom valves, etc. Figure 8F is an exact replica of the pressure pattern recorded in the riser during tests conducted at 2.980 m with a pressure of 196 kg / cm2 in the riser during a real well drilling operation. of oil in the western Texas region of

United States of America.

  The graph of Figure 8A was obtained using a slow-action valve. The numerical values of the parameters involved in the

  Figure 8A were as follows: K2 = 0.5 cm2 / sec

* count and Tb (v) = 2 seconds, that is, values

  similar to those suggested in the prior art.

  (Figures 1A and 1B). Accordingly, Fig. 8A and Fig. 1B both show the regime of slow pressure pulses. On the other hand, Figure 8E was obtained using a quick acting valve for which the following values were: K2 = 100 cm2 / second and Tb (V) = 10 1 second. As a result, Figure 8E bO seod. Ecosqeclfiue8 represents the hydraulic shock wave regime, while the valve wave of Figure 8E is

  very similar to that of Figure 2B.

  Figures 8B. 8C and 8D, defined in the table above, represent the transition from the slow pressure

  hydraulic shock wave regime.

In the essays illustrated in

  8B, 8C and 8D, the conditions were now

  held as similar as possible. The instrument was located near the bottom of the drill holes to a depth of about 2.438 m, the viscosity of the slurry was about 40 "Funnel" and the weight was 1.5 kg / l. When it was "open", the flapper

  had an effective opening section of 0.7 cm 2.

  The normal riser pressure was 210 kg / cm 2 and the valve used in these tests was similar to valve 40, but had been modified to allow a slower action (without the bistable action); in other words, the valve was a simple pressure-balanced valve and the flow was regulated by a throttling provided at the inlet passage. It should be noted that the action of the flap for FIG. 8B was very fast, but did not produce

  the desired regime of hydraulic shock waves.

  However, from the outset, it was found that faster action was desirable. Discharge rate

  was around 19 liters / second2.

  By adjusting the intake throttle

  and the exit constriction, as well as the

  electrical power supply to the solenoids

  that we have obtained different speeds of

the valve.

  From the above, it is found that no shock wave is produced when K2 = 0.5 cm2 / second, while a wave is obtained.

  almost ideal shock when K2 = 100 cm2 / second.

  X. SHOCK WAVE FORMATION FOR DELETION

NOISE (OTHER VARIANT).

  Another parameter will be introduced that expresses a requirement for the intensity of

  the shock wave. Two different methods will be

  annuities. One of them is based on a parameter K3 which represents the amount of sludge (measured in cm3) passing through the valve during the period Ta (V) (this quantity is known as the "quantity in motion") . The other method is based on a parameter K4 which represents the

  average flow of the sludge stream during the

  Ta (V) Then: quantity of sludge passing during the period Ta () K4a () Ta (17) Ta The period of opening of the valve, that is to say the period Ta (V) a To simplify the problem, it will be assumed that the rate of increase of the flow rate during the period Ta (V) is constant and is equal to K1. Since then:

V (t) = K1 t in cm / second (18).

  It will also be assumed that the speed

  of increasing the opening of the valve is con-

  aunt and is equal to K2. Since then:

S (t) = K2 t in cm2 / second (19).

  As a result, the volume passing through the valve over time Ta (V) is Ta (y) = K1K2t dt 2 (K1t2Ta) 3cc

O (20)

  Therefore, the parameter K3 is the quantity (in cm3) of fluid passing through the valve during the period T (V). This parameter is the quantity moving during the period of one year.

  single opening and a single closing of the flap.

  Another method consists in taking, instead of the parameter K3, a parameter K4 representing the flux during the period Ta v, that is to say: K 4 a (V) (21) a

  XI, GENERAL METHOD FOR NOISE SUPPRESSION.

  The general method for decoding the signals emitted by the pressure transducer 51 will now be considered. FIG. 9 represents the apparatus system, while FIGS. 10A to 10G illustrate certain waveforms and certain pulses involved in decoding the signals using the illustrated system

in Figure 9.

  The signal emitted by the transducer

  pressure 51 comprises a signal carrying information

  useful information together with para-

  sites that tend to hide the useful signal.

  The signal bearing the useful information represents the coded message provided by the valve 40 in response

  to a sensor. There are various para-

  sites. One of these signals, which is produced by the pump 27, has an intense constant component

  the mud pressure generated by the pump.

  This component is involved in the circulation of the sludge through the drilling pipe, as well as in the return of the sludge by the annular space formed between this drill pipe and the wall of the hole. To this component, is superimposed

  an alternative component produced by the movement

  back and forth of the pump pistons. In order to improve the reception, it is

  desirable to remove, from the exit of

  51, the constant component intervening in

  the pressure produced by the pump 27. As a result

  quence, a frequency selective filter 150 is

  corded to the transducer 51 to transmit frequencies

  range from 0.1 to Hertz and to attenuate frequencies outside this range. The frequencies contained

  in the constant pressure component are less than

0.1 Hertz.

  In the terminology applied to this specification, it is appropriate to distinguish between the expression "filter" applied to the frequency selective filter 150 and the expression

  "digital filter", which will be used in the

  following description of the invention. In a "filter" such as filter 150, conventional filtering is performed by means of analog type electronic networks whose behavior is usually handled in the frequency domain. The term "filter" can be used to refer to a so-called "wave filter", "Shea filter" (see, for example, "Transmission Networks and Wave Filters" by TE Shea, D. VanNostrand Co., New York, NY, USA, 1929) and other filters such as filters "Chebyshevil" and "Butterworth" On the other hand, a digital filter such as a matched filter, a

  pulse shaping filter or a shaped filter

  The spike processor is more fruitfully processed in a time domain. The output of a filter 24751 1 l

  is obtained by convolving the program

  digital input with the coefficients of

  filter deformation. A digital filter is a computer. The signal transmitted to the output terminal 151 of the filter 150 will be expressed by a function F (t) which is: F (t) = B (t) + P (t) + U (t) (22) o B (t) ) is the signal carrying the useful information,

  r (t) is the disturbing signal emitted by the varia-

  periodical pressure of the pump (noise of

  pump), while U (t) represents an erratic noise

  than. The erratic noise is produced by various effects such as the action of the teeth of a drill bit being drilled, by the gears of the mechanism of the

  stem train and other devices

  in rotary drilling operations. In some cases, U (t) is about a white noise whereas in other cases the difference between

  U (t) and a white noise can be important.

  The coded message expressed by the information-carrying signal B (t) is a series of words

  binary each containing a succession of bits.

  A single bit of a binary word is produced by a single "operation" (ie by a single opening and a single closure) of the valve 40.

  single operation produces a shock wave

  hydraulics manifesting itself on the surface of the ground by a single valve wave such as the valve wave illustrated in Figure 2B. Accordingly, the message expressed by B (t) is in the form of a coded valve wave sequence, each of which is of the type shown in FIG. 2B. Figures IOA to IOG represent different steps to be

  performed to separate the information-carrying signal

  B (t) interference. In order to facilitate

24? 5111

  When expressed, B (t) has been expressed in FIG. 10A by a single valve wave rather than a succession

  coded valve wave. Therefore, the clafeat wave shown in FIG. 10A is of the same type S as a single wave of

valve of the type represented in

  Figure 2B. However, a slight change

  comes in the notation. In Figure IOA, we have

  awarded the exponent "s" which appears in Figure 2B.

  Therefore, in Figures IOA to IOG, different times have been designated t1, t2, .... t15, t16 without the exponent "s". Various graphs of Figs. 10A-IOG have been annotated appropriately. In order to make the presentation easier and clearer, the time scales corresponding to

  these graphics have been distorted.

  In order to eliminate noise signals (pump noise and erratic noise) and to emit a signal representing the message

  coded, three successive operational steps are planned.

  sive that can be identified as follows: Step 1: During this step, a signal having three components as shown in Figure

  IOA is transformed into a signal having two

  as shown in Figure IOC. This step

  for the purpose of suppressing pump noise P (t).

  Following this step, a "valve wave" of the type shown in FIG. 10A is transformed into a double wave. "This double wave is represented

in figure IOC.

  Step 2 This step is to remove the

erratic noise signal.

  Step 3: During this step, each wave

  double the type shown in Figure 1OD is trans-

  formed in a single pulse of the type shown in Figure IOG. As a result, a coded sequence of individual pulses representing, under 64, is obtained.

  digital form, the parameter measured by the sensor-

  101 at an appropriate depth in the borehole.

  XII. SUPPRESSION OF PUMPS NOISE (STEP N 1).

  We will now examine Figure 10A. This figure shows the three components of the signal F (t) as expressed by equation (22). These components are as follows: the shut-off wave B (t), the pump noise P (t) and the erratic noise U (t). As noted above, the

  signal F (t) was obtained by means of the filter 150.

  This filter is connected to a delay element 152 effectively delaying the input signal F (t) of a

  time Tp which is a period of the oscillation pro-

  The signal obtained on the output lead 153 of the delay element 152 can therefore be expressed by F (t-Tp). The three p components of the signal F (t-Tp) are indicated in p

  figure lOB. These components are: the wave of

  delayed pet B (t-Tp), delayed pump noise p P (t-Tp) and delayed erratic noise U (t-Tp). The lapse of time T depends on the speed of rotation of the pump and, since this speed is not

  not constant, the delay T is a delay-variable.

  Therefore, the delay element 152 must be appropriately controlled in response to the speed of rotation of the pump 27. As a result, the element

  152 is designed to receive, via the

  154, synchronization pulses emitted by a generator 155 which is mechanically driven

  by the pump to emit an appropriate number of

  impulses per revolution of it. For this purpose,

  a chain transmission is provided.

  152 may be the double "Reticon" analog delay line, model SAD-1024, sold by "Reticon Corporation", Sunnyvale,

California, U.S.A.

  It will be assumed that the pump 27 performs N1 strokes per second. Therefore, Tp = 1 / Nl. Pulse generator 155 outputs timing pulses at a relatively high rate N2 which is a multiple of N1. Therefore, N2 = KN1

  o K is a constant chosen at the value of 512.

  As a result, if the pump runs one stroke per second, the signal generator will then have

  produce 512 pulses per second. It is obvious

  that the pulsation rate of the mud pump 27 varies with time and, therefore, N2 will vary so that the delay produced by the element 152 is always equal to a period of the mud pressure oscillations produced by

the pump 27.

  The signal F (t-Tp) emitted by the delay element p1 152 is applied to an input lead 153 of a subtractor 160. This subtractor 160 also receives, at its input conductor 161, the signal F ( t) issued by filter 150 and, at its

  output conductor 162, it emits a different signal

  which is: x (t) = F (t) - F (t-Tp) = B (t) -B (t-Tp) + P (t) -P (t-Tb) + U (t) U (t-Tp) (23) Since P (t) is periodic and has a period Tp, we have: P (t) - P (t-Tp) = O (24) p Therefore, because of the periodicity of the pulsations produced by the sludge pump 27, the noise emitted by it is removed and the

  signal obtained at the output conductor 162 of the sub-

  Tractor 160 can then be expressed as follows: x (t) = b (t) + u (t) (25) o (26) b (t) = B (t) - B (t-Tp) is the carrier signal information and u (t) = U (t) - U (t-Tp) (27)

is an erratic noise signal.

  The information carrier signal b (t) and the noise signal u (t) are represented in

  figure lOC. We can then see that in fact

  killing step n 1 as described above, converting the information-carrying signal B (t) shown in FIG. 10A and having the shape of a valve wave into a different carrier signal

  information item b (t) as shown in FIG. 10C.

  The signal b (t) will be called "double wave" by

  traste with the signal B (t) which represents a "valve wave". A double wave comprises two valve waves such as valve waves "IA" and "B" in Figure 10C. These valve waves are separated from each other by a time Tp. The valve wave "A" is analogous to

  that of Figure 10A, while the wave of

  pet "B" represents the inverted form of the wave of

valve "A".

  The signal x (t) (equation (25)) is furthermore applied to the input conductor 162 of an analog / digital converter 163 which is controlled by a clock 178. In this case,

  to the output conductor 164 of the analog converter

  logic / digital, a signal expressed by: xt = bt + ut (28) o, according to the symbols used here, x.t, bt and ut are digital versions of the signals

  analogous x (t), b (t) and u (t) respectively.

  The signals xt and ut are in the form of time series such that xt = (- * * 2 x_2, x1, x0, x1, ..., x9 ,, a) (29) and ut =. u_2, ul, u0, uI1, *., u9, ...) (30) respectively, while the signal bt is a wave of finite length bt = (bo, b1, b2, .. bn) (31)

  XIII. ERRATIC NOISE SUPPRESSION WHEN

  THESE ARE WHITE NOISES (STEP N 2).

  The mixture of a double wave bt and a noise signal ut is then applied to a filter

  numerical 170 of length (n + 1) having a function

memorisation.

  (32) In this embodiment, a digital filter known as

  "adapted filter", while the function is chosen

  memorization and to optimize the functionality

  filtering. Optimum conditions are obtained when the signal-to-noise ratio at the

  Filter output 170 is at its maximum value.

  (For a description of suitable filters, see,

  for example, a publication by Sven Treitel and

  E.A. Robinson, "Optimum Digital Filters for Signal-

  to-Noise Ratio Enhancement ", Geophysical Prospec-

  ting, volume 17, No. 3, 1969, pages 248-293, or

  E.A. Robinson, "Statistical Communi-

  with Radar and Seismic Signals ",

  Hafner Publishing Company, NY, N.Y., U.S.A.

1967, pp. 250-269).

  The storage function of the matched filter 170 is performed so that it can be controlled to ensure, at all times during the measuring operations, the optimum conditions for the operation of the filter. This

  filter control is carried out by means of a

  user 172 which receives appropriate data from a storage and call element 173 and

as will be described below.

  A signal yt obtained at the output conductor 174 of the matched filter 170 may be expressed by a convolution of the input function xt and the storage function at. Since then: Yt = xt * at: aOxt + alxt-1 + ... + anxt-n (33)

  o the asterisk indicates a convolution. In sub-

  stituant, in this formula (33), xt = bt + Ut, we obtain: vt (34) Yt = ct + vt o ct = bt a (35) ttt is the response of the filter to a pure signal input and vt ut * at (36) is the noise output. Figure -11 gives a

  block diagram showing these relationships.

  In order to ensure the optimum conditions during the operation of the matched filter 170, a certain moment is chosen, for example, the moment t = to and the instantaneous power of the output of the filter containing a signal at the moment t = to be as great. as possible compared to the

  average power of filtered noise at this time.

  Therefore, in order to detect the signal ct in the filtered output ut, the signal-to-noise ratio defined by: = (value of the signal filtered at the time t0) 2 (37) power of the filtered noise If the signal is convolved of length (n + 1) (bO, b1, ..., bn) with the filter

  of length (n + 1), we obtain a series of outputs

  length (2n + 1) (c0, c1, ... cn, .. c2n1, c2) where c is the central value of this output series. From then on, when to = tn,> becomes C = (aobn + albn + * + anbo) 2 (38) nn 1 = 1n = 2 2 [Evn] E [vn] where E [v2n] is the mean value of the noise output power. We will now assume that the erratic noise ut is a white noise. It can then be demonstrated (see, for example, a publication by Sven Treitel and EA Robinson, "Optimum Digital Filters for Signal to Noise Ratio Enhancement," Geoph., Prosp., Vol. XVII, No. 3, 1969, pp. 240-293). that the maximum value of the signal-to-noise ratio can be obtained when: (a0, a1, ..., an) = (kbn, kbn1, ..., kb0) (39) where the value 1 has been chosen for k. Therefore, in the presence of a signal embedded in the noise, when the latter is a white noise, the optimal conditions can be obtained when the filter memory is given by the inverted signal, in particular by the

  sequence of coefficients (bn, bn_1 .. *, bo).

  The memory of the filter 170 is determined at all times by means of the computer 172 which is

  connected to this filter by a channel 175. The expression

  channel, used in this specification.

  tion, means appropriate conductors, connections or transmission elements in accordance with the requirements of a particular case. A storage and call element 173 is provided for storing the function bt for subsequent transmission thereof to the computer 172 via the channel 176. This computer is intended to reverse the input data expressed by a sequence (bo, b1, ..., bn) in order to obtain, at its output channel 175, a sequence (bn, bn_1., 1 bo) which, in turn, is applied via the channel 175 to the filter adapted where it is stored as memory

  filter according to equation (39).

  In addition, the filtered output yt, obtains

  to the output conductor 174 of the matched filter 170, is applied to a digital-to-analog converter 181. Since yt represents a signal in digital form, the corresponding analog function obtained at the output conductor 182

  of the digital-to-analog converter 181 will be

  awarded by y (t) according to the symbols used here. It should be noted that maximizing the signal-to-noise ratio in the filtered output Yt is equivalent to minimizing the noise signal vt (or its analog equivalent v (t)) compared to the information-carrying signal ct ( or its analogical equivalent c (t)). Therefore v (t) << "c (t) (40) and y (t) r c (t) Nb (t) (41)

  Consequently, as it is

  shown in Figure lOBD, the output function y (t) of the matched filter closely resembles the function

b (t) shown in FIG.

  An important feature of the present invention is to store (for later reproduction) the function bt at

  means of the storage and calling element 173.

  The process required for memory will now be described.

  use the function bt with reference to figure 12.

  This process comprises the following different steps: Step (a) Stopping the drilling operations, that is to say that the bit 31 is lifted a short distance above the bottom, the transmission systems are maintained and lifting at a standstill and

  the rotation of the element 21 is stopped.

  Step (b) The pump 27 continues to operate in the same manner as during normal drilling operations, i.e., at a uniform rate and at a pressure representative of that used during actual measurements taken during drilling". All other sources of disturbance such

  that the taking of an alternating electric current

  From generators, the operation of cranes, etc., are shut down. The effects of

  "swell" and other noise sources are eliminated.

  as far as possible in the case of

  drilling operations carried out at sea (eg

choosing a calm day).

  Step (c) As described above and illustrated in FIG. 5A, the underground coding is determined by

  a "clock" that is rigorously in sync with

  With the clock installed in the surface equipment, we can determine, on the surface, the moment when an individual impulse

  in the basement, for example, the impulse

  opening; moreover, if we know the speed of transmission through the mud column, we can

  also know the exact moment at which the impulse

  hydraulic pressure will be received at the surface. Therefore, we can receive a single "wave" on the surface and

  to know in advance when the wave appears

  even if it is likely to be concealed by

  a noise. (In many cases, the individual wave

  duel will emerge from the noise, so that we can

  practice a visual observation on the oscilloscope).

  The transient hydraulic phenomenon produced by the valve 40 is received by the transducer 51 at a

known moment.

  Step (d) The signal received by the transducer 51 is transmitted by the filter 150 to selectively transmit frequencies in the

  range from 0.1 to 10 Hertz. Since the operations

  Drilling rations were stopped (as described in step (a) above), the erratic noise U (t) is negligible and, therefore, the signal obtained at the output of the filter 150 has the form F (t) =

B (t) + P (t).

  Step (e) The signal F (t) emitted by the filter 150 passes through the delay element 152, the subtractor and the analog / digital converter 163 as discussed above. Usually, during drilling, the signal emitted at the output of the analog / digital converter 163 has the form

  xt = bt + ut o ut represents the noise due to ope-

  drilling rations. However, in this case too, since the drilling operations were

  stopped, the erratic noise signal ut is neglected.

  geable. Under these conditions, we have xt% .bt The xt signal represents, as far as possible, a "noise-free signal" which could correspond

  to a wave representing the signal carrying information

  mation. Step (f) The function xt bt is stored. The process steps (a), (b), (c), (d) and (e) described

  above are carried out by means of the system illus-

  FIG. 12. In this system, the output of the analog / digital converter 163 is applied to the input of the storage and calling element.

  173 to register the function xt bt.

  It should be noted that, in the frequency domain, the storage function at and of the matched filter 170 can be expressed as follows: ## EQU1 ## (b) (f) (42) of indicates the frequency and B * (f) is the complex conjugate number of signal transformation

bt following Fourier.

  The erratic noise suppression, when the noise is a white noise, can, in

  in some cases, be carried out by means of an

  correlator rather than by means of a suitable filter such as the matched filter 170 illustrated in Figure 9. For this purpose, it is necessary to modify the diagram of Figure 9 by eliminating the matched filter 174, the computer 172 and the memorisation and appeal element 173. It is therefore appropriate to use a

  autocorrelator. The input terminals of the car

  correlator must be connected to the

  164 of the analog-to-digital converter 163. At the same time, the drivers of

  out of the autocorrelator must be connected

  dice to the input drivers 174 of the converter

* digital / analog 181. The output of the converts

  Digital / Analogue can be processed at

  the delay element 190, of the element

  polarity pourer 192, AND gate 193, etc., as shown in FIG. 9. However, in some cases, the output of the digital-to-analog converter can be directly applied to a recorder.

  XIV. TRANSFORMATION OF A CODED WAVE SEQUENCE

  DOUBLES IN AN ENCODED SEQUENCE OF IMPULSES

SHORT (STEP 3).

  We will again consider the operating system illustrated in Figure 9. Now we have

  obtained, to the output driver 182 of the converters

  181, a signal representing

  by a function y (t) which is represented in FIG. 19D and which has a shape similar to that of

  the double wave b (t), that is to say y (t) .- p b (t).

  The function y (t) -.- # b (t) represents a single bit of the signal converted into a digital signal and actuating the valve 40. It is obvious that a 247511l

  such a function is not very convenient for

  a very short period of time corresponding to a single opening and closing of the

  40. As a result, as has been pointed out

  In step 3, it is necessary to trans-

  to form a double wave in a single impulse

  coinciding with the operation of the valve. This object is produced by means of a delay element controlled by a clock 191 in combination with the polarity inverter element 192 and an AND gate 193 (coincidence network) according to the system shown in FIG. 9. Via the conductor 182, the delay element 190 receives the signal y (t)

  issued by the digital-to-analog converter 181.

  This delay element is controlled by a clock 191 in order to obtain, at the output conductor 195, a

  delay equal to the period of time Tm.

  B (t-Tm) shown in FIG. 10E is applied by the conductor 195 to an inverter element of polarity 192 to produce, at the output conductor 197 of the element 192, an inverted delayed double wave, expressed by -b (t - Tm)

and illustrated in Figure 1OF.

  The signal -b (t-Tm) is applied by the conductor 197 to the gate And 193. At the same time, the signal b (t) transmitted by the digital / analog converter 181 is applied by the

  conductors 182 and 200, at the AND gate 193.

  each of the signals b (t) and -b (t-Tm) comprises pulses having a positive polarity and a negative polarity. By comparing the signal b (t) represented in FIG. 1OD with the signal -b (t-Tm) represented in FIG. 10F, it can be seen that there is, in FIG. 10D, only one pulse which is in coincidence of time with the pulse shown in Figure 10F. This pulse is transmitted in the time interval between t3 and t4 in FIG. 10D and between t and t10 in FIG. 10F. It should be noted that the moments t3 and t9 coincide, since t3 = t1 + Tm and t9 = t1 + Tm. In the same way, the moments t4 and t10 coincide) since

  t = t1 + T + TM and t 1 = t1 + Tn + Tm.

  quence, a single momentum of coincidence is

  of the double wave b (t) and is represented in FIG. As a result, the AND gate 193 which has received, at its input leads 200 and 197, signals representing the function b (t) and the function -b (t-Tm) respectively, transmits at its output conductor 210. , a single pulse of the type shown

in Figure 1OG.

  It will be recalled that, for reasons of simplicity, there is shown, in this embodiment, a single pulse which is produced and corncide substantially with a single opening and a single closure of the valve. It will be remembered that, during a real drilling during which a measurement operation is simultaneously performed, the output conductor 210 obtains a coded succession of individual pulses representing a measurement.

  by a selected sensor of a select parameter

tioned.

  The coded sequence of indi-

  viduals obtained at the output conductor 210 of the AND gate 193 is applied to a digital-to-analog converter 211 which is controlled by a

  clock 212. At the output conductor 214 of the

  digital / analog builder 211, there is obtained, in analog form, a signal representing the measurement of the selected parameter. This signal is

  recorded using the recorder 54.

  XV. USE OF A CROSS CORRELATION.

  In another embodiment of the present invention, for noise suppression, a cross-correlator may be used instead of a matched filter. There is an analogy

  between the convolution of two functions (

  represented by equation (20a)) and a correlation

  crossing. The cross-correlation of one function with another gives the same result as that which would be obtained by passing the first function through a filter (adapted filter) whose function

  memorisation is the opposite of the second function.

  (See, for example, a publication by N.A. Anstey,

  "Correlation Techniques-A Review", Geoph. Prosp.

  Volume 12, 1964, pages 355-382 or a publication of Y.W. Lee, "Statistical Theory of Communication," John Wiley & Sons, New York, N.Y., E.U.A., 1960,

page 45).

  FIG. 13 shows how a cross correlator 200 can perform the same operations as those performed by a matched filter. The cross correlator 200 has two input terminals 201 and 202 and an output terminal

  203. The xt signal emitted by the analogue converter

  Digital signal 163 is applied to the input terminal 201, while the signal bt transmitted by the storage and call element 173 is applied to the input terminal 202. A signal representing the correlation

  cross of xt and b is thus obtained at the

  It is readily apparent that the cross-correlation signal obtained at the output conductor 203 is identical to the convolution signal yt expressed by equation (33) and emitted, in FIG. 9, by the matched filter 170. Furthermore, this cross-correlation signal is processed, as shown in FIG. 13, in the same manner as the signal obtained by means of the matched filter 170 has been processed in the system illustrated in FIG. 9. The cross-correlator can be of the type known as "Model 3721A", manufactured by "Hewlett Packard Company", Palo Alto, California, USA

  XVI. REMOVAL OF ERRATIC NOISE WHEN IT IS

CI IS NOT A WHITE NOISE.

  When the erratic noise is a white noise, the autocorrelation qt of the noise function is zero for t 0. We shall now consider the case in which the parasitic noise τ has a known autocorrelation function q in which the coefficients qt are not necessarily zero for t 4 O. This is the case of a "noise set in autocorrelation" unlike a pure white noise

  only the coefficient of autocorrelation

  not to zero is qo. Fig. 14 shows a suitable embodiment of a matched filter and associated elements therein. In this case, it is necessary to memorize not only the information carrier signal bt (for example, by means of the element 173), but also the noise signal ut. Accordingly, the signal illustrated in FIG. 14 comprises two storage and call elements 173 and 224. The storage element

  and call 173 performs a function which is identical

  to the element designated by the same reference numeral in FIGS. 9 and 12. It is used to memorize and then produce the function bt. On the other hand, the memory and call element

  224 has the function of memorizing, then reproducing

  to obtain the noise function ut. The data represented

  sensing the functions bt and ut provided by the storage elements 173 and 224 respectively

  are applied via channels 225 and 226 respectively.

  This computer 228 has the function of transforming the input signal from the channels 225 and 226 into the required data.

  to determine the memory function of the thread

  220. The latter data are applied

  The notation will then be the same as that described above, with the exception that, in this case, it must be taken into account that the noise ut is no longer a white noise. The adapted filters discussed here are

  indeterminate in the sense of an amplification factor

  arbitrary decision which, for reasons of commu-

dity, is equal to unity.

  We will use the same definition of the signal / noise ratio f. Therefore, we have: cn n f E [v 2] (43) It is desirable to bring to a maximum by assuming that the noise ut is of the type set to autocorrelation. A it is appropriate to introduce the notation a = (aO, a -, ... an) denotes the vector of (n + 1) rows which ó the memory of the matched filter 220. In addition b = (bn, bnlV A .., bo) is the vector of (n + 1) rows that defies time of the signal bt and qo ... qe * the input value that moment,

matrix.

(44) characterizes e. - (45)

iit the in-

0 0

qn *. * qo

247 75111

  is the autocorrelation matrix of (n + 1) by (n + 1) of the noise. We can then write: = (ab ') (ba') (46) aqal 'o the prime sign (') designates the transposition of

the matrix.

  In order to bring the value "to its maximum

  mum, the quantity (46) will be differentiated vis-à-vis the filter vector a and the result will be equal to

zero.

  We obtain a relation qa '= b' (47) which can be transcribed in the form 0 n 15.a1 bn_ 15. =. * n | ab (48) g L o q0 a nb0 This is the matrix formulation

  of a group of (n + 1) linear simultaneous equations

  areas in the (n + 1) filter coefficients in-

  known (ao, al, ..., an). The solution gives the optimum optimum matched filter in the presence of a noise set in autocorrelation. Equation (48) can be solved by the Wiener-Levinson recursion technique (see N. Levinson, "The Wiener

  RMS error test in Filter Design and Prediction

  "Day of Mat and Phys., 1947, Volume 25, pages 261-278 and S. Treitel and E. A. Robinson,

  "Seismic Wave Propagation in Terms of Communica-

  Theory ", Geophysics, 1966, volume 31, pages

  17-32). This recurrent method is very effective

  therefore, it is possible to calculate adapted filters of great length by means of

  228. The known quantities involved

  in this calculation are the matrix of autocorrelation

  noise enumeration q and the time inverse of the signal wave bn t, while the unknown quantities are the filter coefficients at. These coefficients

  The filter functions represent the memo function.

adapted filter 220.

  The calculations necessary to determine

  the memory function of the adapted filter

  220 are carried out by the computer 228.

  memoranda and appeal 173 and 224, the

  The receiver receives data on the functions b and ut respectively. When receiving t

  qt, the noise autocorrelation matrix is cal-

  abutment and, when receiving the function bt v

  the inverse time of this signal is determined.

  Then the unknown filter coefficients at are calculated and then transmitted to the matched filter

220 via the output channel 230.

  The output of the matched filter 220 is applied to a digital-to-analog converter

  181 and is subject to a complementary treatment

  in the same manner as the output of the matched filter 170 has been processed in the illustrated system.

in Figure 9.

  In the frequency domain, the function

  The storage pattern of the matched filter 220 may be expressed as follows: A (f) = e-2rfn (49) where B * (f) indicates the Fourier transformation of the time inverse of the signal b = bl, ..., bn), while Q (f) indicates the power spectrum of the noise in the interval

  (f + df). The physical meaning of ltexpres-

  sion (49) is simple. More spectrum of amplification

  of IB (f) I the signal is wide, the more the spectrum of

  power density Q (f) of the noise in the interfering

  valle (f, f + df) is reduced and the filter adapts

  transmit frequencies in this range.

  Therefore, if the spectral power density Q (f) of the noise is low in a certain interval of the frequency band occupied by the signal, the adapted filter is essentially transparent.

  (very low attenuation) in this interval.

  Consideration will now be given to the storage and signal calling elements 173 and 224. The process necessary for storing the bt signal by means of the element 173 has been previously described in steps (a) to (f) performed

  using the system illustrated in Figure 12.

  A different method is needed to memorize the noise signal ut by means of the element 224. As previously pointed out

  about figure 12, we can receive and memorize

  Raise a "noise-free signal". In the same way

  because of the synchronism between

  "clocks" of the basement and the surface, one can receive

  see and memorize a "signal-free noise", that is, the signal received by the transducer 51

  during a normal drilling operation (

  all the noises that are part of such

  operation, but no information-bearing signal).

  In this case, one can also use the system

  shown in Figure 12 to illustrate the processes

  sus required. The steps to obtain a

  The registration of the function u (t) may be the following: Step (a) Apply a weight to the bit and

  normal drilling operations are carried out.

  Step (B) A time is chosen at which there is no information-carrying signal, i.e.

  rest time between binary words.

  24751! I Step) A signal is obtained representing the

  pressure variation of the drilling fluid at the trans-

  driver 51; This signal is transmitted by the filter 150. Because of the moment chosen in step (p) above, the signal b (t) is non-existent and, consequently, the signal obtained at the output of the

  Filter 150 has the form F (t) = P (t) + U (t).

  Step (4) The signal P (t) is removed from the pump noises. This elimination is carried out by means of

  the delay element 152 and the subtractor 160.

  Then we apply the signal obtained to the converts

  Analog / digital source 163. Since there is no information-carrying signal, bt = 0 and, therefore, the signal transmitted at the output of the analog-to-digital converter 163 has the form

Xt = Ut.

  Step (ú) We obtain a record of the

  function xt = ut using the memory element

  22 to the output of the analog / digital converter 163 as shown in FIG. 12. To summarize what has been said above, it can be seen that, if the noise is white, the matched filter 170 and the associated elements (as shown in Figure 9) provide the optimum value of the signal-to-noise ratio i. If the noise is not

  a white noise, but has a function of autocorrela-

  known filter, the matched filter 220 and associated elements (as shown in FIG.

the optimum value of È.

  XVII. CONFORMER FILTER OF TYPE IIMPULSIONS

WIENER.

  Figure 15 shows a portion of

  surface equipment including a functional filter

  according to a principle different from that

  adapted filter of Figure 9 or Figure 14.

  The matched filter of FIG. 9 or FIG. 14 is optimum in that it is a linear filter optimizing the signal / noise ratio. On the other hand, the filter 240 of FIG. 15, which is called "pulse shaping filter" or "Wiener filter", is an optimum filter in the sense that it is

  a linear filter that minimizes the difference

  tive between a desired output and a real output.

  (For a description of a filter of this type, see,

  for example, the publication of E. A. Robinson and

  Sven Treitel, "Principles of Digital Wiener Filte-

  ring ", Geophysical Prospecting 15, 1967, pages

  312-333 or a publication of Sven Treitel and E.A.

  Robinson, "The Design of High-Resolution Digital

  Filters ", IEEE Transactions on Geoscience Electro-

  nics, volume GE-4, No. 1, 1966, pages 25-38).

  The pulse shaping filter 240 of FIG. 15 receives, via its input channel, data relating to the function xt = bt + ut and supplied by the analog / digital converter 163. This pulse shaping filter is a filter of length (m + 1) having memory ft = (fox fl, * - 'fa) (50)

  which, with a low weight error, transforms the

  trength of length (n + 1) xt = (xO, xl, **> ..., xn) in an output of length (m + n + 1) zt = (Zo, zl, .., zm + n) . A model of a filter of this type is shown in Figure 16. In this model, there are three signals, including: (1) the input signal xt, (2) the actual output signal zt and (3) the desired output signal bt. The signal bt is a wave

  double the type shown in Figure 10C.

  The output signal zt is a convo

  lution of the filter storage function ft with the input function xt, that is: zt = ft * xt (51) The problem is to determine the storage function ft so that the actual output zt be as close as possible (with low weight error) to the desired output bt. To select the storage function, the following quantity is minimized: I = (sum of squared errors between the desired output and the filtered signal wave)

+ \ (filtered noise power).

  o v is a pre-assigned weighting parameter.

  The calculations necessary to minimize I are performed by the computer 245 having three input channels 246, 247 and 248 respectively. A storage and call element 173 transmits to the computer 245 and via the

  channel 246, the data relating to the function bt.

  In the same way, the storage and call element 224 transmits to the computer 245 and via the

  channel 247, the data relating to the function ut.

  Channel 248 is used to transmit, at the

  245, the data relating to the functions Xt which are also applied to the input lead 241 of the pulse shaping filter 240.

  When receiving the signals from

  bt, ut and xt applied by the channels 246, 247 and 248 respectively, the computer 245 is designed to perform certain calculations which will be described hereinafter, as well as to transmit, to the shaping filter 240 and via the

  251, the data required for the

  the storage function of the filter 240.

  Therefore, the actual output zt of the filter is, with

  an error of low weight, as close as pos-

  sible of the desired output bt. In other words: zt. bt (52)

24751 * 1

as shown in FIG.

  The operations carried out by means of the computer 245 will now be considered in greater detail. In the symbols, the quantity I to be reduced to the minimum is: m + n I = (bt - t) 2 + VE [v] (53 t = oo the notation E [] indicates the mean of set and om vt => Ut = L fs Ut-s s = O

  represents the filtered noise. By simplifying the expression

  for I, we obtain: m + nm 2 mm I = t 7- fs xt-s) + Efsqt-sft (54) t = O s = 0 s = O tO In this case: qt-s = E [uts utt] ot is a fictitious time index and o qt-s is the autocorrelation of the received noise. Differentiating

  the expression for I vis-à-vis each of the coefficients

  filter ficients and by putting the derivatives equal to zero, we obtain a group of simultaneous equations, namely: m Tfs rt-s + qt-s = gt (56) s = 0 for t = 1, 2, *. ., m. In the equations above, the quantities rt-s, qt-s and gt are known, while

  that the quantities fs are unknown.

2475 1I

  The calculations made by the computer 245 are used to determine the parameters rts, qt-s and gt according to the input functions applied to the

  channels 248, 247 and 246 respectively, to resolve

  then equations (56) for the unknown quantities f s. The parameters rt st5, -s' and gt are defined as follows: The parameter rt-s is the autocorrelation

  of the input signal xt which is applied to the computer

  The parameter qt is the autocorrelation of the noise signal ut which is applied to the computer 245 by the channel 247. The parameters gt are the vector product coefficients between the desired output bt and the input xt. Therefore, we have: m> gt1 bs XS-t (5-7) s = O for t = 0, 1, 2, ..., m. In the expression for

  gt, the desired output bs is applied to the computer

  245 through channel 246, while input xt is applied by channel 248. The parameter J is a weighting parameter to which a value

  appropriate is affected as described below.

  Therefore, the parameters rts, qts and gt are determined by the computer 245, which then solves the equations to obtain, at the output channel 251, the quantities f s. These quantities S are applied to the memory of the pulse shaping filter 240. The actual output zt of the filter 240 is, with a low weight error, also

  as close as possible to the desired output bt.

  Since the matrix of the equa-

  the matrix [rt_s + Vqts] is under

  form of an autocorrelation matrix, these equa-

  can be solved effectively by the

247511I

  recurring method. This recurrent method has been described in the following two publications:

  N. Levinson, "The Wiener R4S (root mean square

  terion = effective criterion, in Filter Design and Prediction ", appendix B by N. Wiener," Extra-

  Polation, Interpolation, and Smoothing of Stationa-

  Time Series, "John Wiley, New York, NY, 1949, and Enders A. Robinson," Statistical Communication and Detection with Special Reference to Digital Data Processing of Radar and Seismic Signals, "pp. 274-279, Hafner Publishing Company, New York. York

N.Y., U.A., 1967.

  It should be noted that the machine time

  required to solve the simultaneous equations

  above for a filter with m coefficients is proportional

  m for the recurrent method, compared with

  to the standard method of solving

  simultaneous equations. Another advantage of adopting this recurrent method is that it does not require, for the

  memorizing in the computer, that a pro-

  portional to m rather than m2, as is the case

in classical methods.

  When designing a filter con-

  impulse trainer, two conditions may be taken into consideration, including: (a) conform as closely as

  possible function zt in function desired bt.

  (b) Produce an output power

  as low as possible if the noise

  portun is the only entry for this filter.

  In many practical cases, a filter is needed to simultaneously fulfill both of the above conditions and so we are faced with the problem of finding a

  appropriate compromise between these two conditions.

247S111

  Therefore, we choose, for the parameter \, an appropriate value ensuring the relative weighting between

these two conditions.

  There are cases in which the value zero is assigned to V. In this case, the expression (53) takes a simpler form, notably: m + n I '= (bt zt) (58) t = O and the 245 computer does not require the data

  representing ut. In this case, the memory element

The call and call 224 shown in FIG. 15 is deleted and, consequently, the computer 245 has only two input channels, in particular the channel 246.

and the channel 248.

  It will now be observed that the return

  Both a pulse shaping filter and an adaptation filter are not exactly alike; in other words, for a given input, the outputs of these filters are not

  identical. The expression yt rbt, which is

  cable to a suitable filter, has been used above

  to indicate that the signal expressed by

  feeling the output of a suitable filter) is getting closer

closely of the double bt wave.

  As a result, it was pointed out that the same graph in Figure 1OD represented the

  functions y (t), as well as the functions b (t).

  It should also be noted that the expression

  zt bt, which is applicable to a conformal filter

  pulse generator, has been used above for

  indicate that the signal expressed by Zt (represent

  the output of a pulse shaping filter

  sions) is closely related to the double bt wave. As a result, it was pointed out that the same graph of Figure 10D represented the function z (t), as did the function b (t). Strictly speaking, the same graph in Figure 10D does not

  should not be used to represent the functions

  b (t), y (t) and z (t). However, since the functions y (t) and z (t) closely approximate the function b (t), it is thought that,

  for the purpose of the presentation, it is appropriate to use

  be the same graph as Figure 1OD to de-

  write the performance of a suitable filter and that of

  a pulse shaping filter.

XVIII. FORMATION OF WAVE POINTS

  The operating system shown in FIG. 17 will now be considered. At the output conductor 162 of the subtractor 160, both the information carrier signal b (t) and the noise signal u (t) have been obtained. The signal b (t) is the double wave shown in FIG. The combination

  signals b (t) and u (t) is applied to the conver-

  analog / digital weaver 163, thus emitting, at the output conductor 164 of the converter,

  signals converted into digital bt and ut, these

  signals corresponding to the analog signals b (t) and u (t) respectively. The combination of these two signals bt and ut is, in turn, applied to the input lead 300 of a tip forming filter 351. This tip forming filter is

  a particular case of a filter shaper of im-

  Wiener-type impulses in which the form de-

  sirée is simply a tip (for a descrip-

  tion of a tip-forming filter, see, for example,

  for example, the publication of S. Treitel and E.A.

  Robinson, "The Design of High-Resolution Digital

  Filter "IEBE Transactions on Geoscience Electro-

  , volume GE-4, No. 1, 1966, pages 25-38).

  It will be recalled now that a double wave b (t) such as that represented in FIG.

  1OC consists of two waves of valve, that is,

  ie a valve wave "A" I and a valve wave "B". Valve wave "BI" follows the valve wave "A" 'after a time T p The tip forming filter to be used in the embodiment

  illustrated in Figure 17 has the function of transforming

  sea the valve wave ^^ A ", as well as the valve wave" Bu at a respective point clearly defined. Therefore, a double bt valve wave is transformed, by means of the formant filter of

  tips 351, in a pair of points.

  FIGS. 18A to 18F respectively represent six possible positions of a pair of points (for example M1 and N1) with respect to the double wave applied to the input terminal 300 of the tip-forming filter 351. It will be assumed that that Tk is the lapse of time flowing between the points M1 and Nj, this lapse of time being the same in all of Figures 18A to 18F. We will assume that H1 is the point

  intersection of the M1 point with the abscis axis

  its (expressed in milliseconds). Therefore, the

  OH1 (in milliseconds) represents the offset

  of time spikes vis-à-vis the double wave.

  Consequently, in FIG. 18A, the time shift OH1 = O, that is to say that the initial peak M is placed at the very beginning of the double wave. The five cases represented in FIGS. 18B to 18F correspond to increasing values of the offset

of time OH1.

  One of these figures represents the

  their optimal time offset for which the resolution of the peak-forming filter is the highest. For this optimum time shift, the output signal from the tip-forming filter is much sharper than any other time offset. A method for obtaining the optimal value of the time offset, the optimal length of the memory of the filter and the optimal value of the lapse of time will be described below.

time Tk.

  A double peak obtained at the output terminal of the filter 351 represents a single bit of the digitally converted signal and actuates the valve 40. As has been pointed out with respect to FIG. 9, it is desirable to transform a

  double point in one point or single pulse.

  In this case, a treatment system similar to that shown in FIG.

  -chapter XIV of the specification, entitled "Trans-

  forming a coded sequence of double waves into a coded sequence of short pulses - step 311) * Accordingly, an element is provided for

  delay 303 controlled by a clock 304 in combination with

  In this case, a polarity inverting element 306 and an AND gate 307 (coincidence lattice) are provided. These elements are mounted in the same manner as that shown in FIG. 9. However, in FIG.

  the delay is different from that of Figure 9.

  In particular, in FIG. 17, the delay element 303 must emit an output signal which is delayed by the input signal by a quantity Tk while,

  in FIG. 9, the delay produced by the element cor-

  respondent 193 is T_. A coded succession of individual pulses obtained at the output conductor of the AND gate 307 is applied to a digital-to-analog converter 308 controlled by a clock 309. At the output lead of the digital-to-analog converter 308, an analog signal is obtained which represents the measurement of a selected parameter in the borehole. This

  signal is recorded using the recorder 54.

  It should be noted that, in certain cases, depending on the specific electronic circuit chosen for the different elements shown in FIG. 17, the inverter-polarity element 306 may not be necessary, since its function can be fulfilled by an appropriate design of FIG. the door AND.

  Figure 19 shows another system

  noise suppression method by means of a tip-forming filter. In Fig. 17, there is provided a special member for eliminating pump noise (i.e. delay element 152 in combination with subtractor 160). On the other hand, in

  19, the noise suppression method has been simplified

  fied. Thus, the signal processing in which a delay element 152 and a subtractor 160 are used is eliminated. Consequently, in FIG. 19, the signal F (t) obtained at the output terminal 151 of the filter 150 is converted to digital by means of an analog / digital converter

  350, to then be applied to a formal filter

  spike driver 351a. This filter 351a is designed differently from the tip forming filter 351 illustrated in FIG. 17. In FIG. 17, the tip forming filter 351 was designed to convert a double wave of the type shown in FIGS. 18A-18F into a pair of separate tips. from one another by a time Tk. On the other hand, the tip forming filter 351a of FIG.

  was designed to convert a single wave of

  fart in one point. Figures 20A-20F show different positions of a tip

  unique vis-à-vis a single valve wave.

  It will be recalled now that each

  individual wave of valve applied to the terminal-

  filter 351a and that each tip indicates

  247i5ii vidual obtained at the output terminal of this filter 351a represent a single bit in the signals

  converted into digital and actuating the valve 40.

  The coded succession of peaks of the digitally converted format obtained at the output terminal of the peak-forming filter 351a is then applied to a digital-to-analog converter 352 where it is transformed into a coded succession of peaks each representing a single bit. coded information in the instrument in the basement. The succession of these bits represents, in a digital format, the measurement of the chosen parameter. However, for recording and / or display purposes, it is necessary to represent this measurement in an analog form. As a result, the signal transmitted by the digital-to-analog converter 352 is applied to a digital-to-analog converter 362 for transmitting, at the output of this converter 362, a signal having an amplitude representing the measurement of the selected parameter. This signal is

  recorded using the recorder 54.

  It should be noted that we can not obtain

  only an approximation of the conversion of a wave

  double in two points by means of the filter 351 re-

  shown in Figure 17, or the conversion of a single wave of valve into an individual tip

  by means of the filter 351a shown in FIG.

  You can not get a pure tip, that is to say

  a delta function. However, the present

  The purpose of the invention is to increase the resolution or definition, that is, to obtain a signal

  much clearer output than the signal from

  Tree. We will now consider how to design a tip-forming filter. In theory, we can achieve this object exactly as if we could use a filter

  which would be allowed to become the memory function

  infinitely long. As a general rule, if one wants to obtain an exact yield of the filter, it is also necessary to delay the desired peaks of infinite time with respect to the input wave. (See publication by J. C. Claerbout and E. A. Robinson, "The Error in Squares Inverse Filtering", Geophysics, Volume 29, 1964, pages 118-120). In practice, it is necessary to design a digital filter whose storage function has a finite duration and, hence, in the

  better conditions, we can not achieve the objective

  It is assumed that for practical reasons it is desired to consider a filter having a storage function of a duration of the order of that of an input wave. We will also assume that we have the freedom to place

  the desired peak at any chosen location.

  For example, FIGS. 18A to 18F represent six possible positions or locations for the tips formed by a filter 301 of the type shown in FIG. 17. Similarly, FIGS. 20A to 20F represent six possible positions for the points formed by the filter. 351 shown in Figure 19. The optimal position of the tips has been determined for each of these cases. Note

  that the position of a point is an important factor

  both determining the fidelity with which the output

  actual looks like the desired tip.

  A tip-forming filter is a

  particular case of a pulse shaping filter

  Wiener type which has been described above.

  As a result, the processes required to design a filter of this type are similar to those

  have been described above. We will focus on

  a low weight error for a filter

whose output is a tip.

  In order to determine the optimum value of the time shift, as well as the optimal length of the storage function for the tip forming filter 301 of Fig. 17, it is necessary to obtain a wave record.

  double bt (which is a digital version of b (t)).

  The steps necessary to obtain a registration

  of this type, that is, steps (a), (b), (c), (d), (e) and (f) have been described above.

  with reference to Figure 12. Therefore, the

  bt is recorded in element 173

  forming part of the system illustrated in Figure 12.

  In the same way, in order to determine the value

  optimal time shift and optimum length

  the function of storing the filter for-

  351a, it is necessary to obtain a recording of an individual flap wave

  Bt (which is a digital version of B (t)).

  We will consider the formative filter of

tips 351a of Figure 19.

  Different positions of a peak corresponding to different delays (FIGS. 20A to 2OF) can be expressed as follows: (1.0.0, ...... 0.0): peak at the time index O or filter tip trainer

zero delay.

  (0,1,0, ...... 1,0): peak at the time index m + n - 1 or trainer filter

of delay peaks (m + m - 1).

  (0.0 ........ 0.1): points to the time index m + n; trainer filter

delay points (m + m).

  The efficiency of a peak forming filter corresponding to different delays is schematically illustrated in Figs. 21A, 21B and 21C. In all these figures, the input wave is the same, that is to say the valve wave Bt which

  recorded and stored as described above.

  above. In Fig. 21A, the desired output is a peak (1,0,0), i.e. a peak having a

  no delay. The memory function corresponds to

  for a zero delay spike forming filter is F- (FO, F, F q Fn) and the actual output is WO = (Wo, Wn).

  Figures 21B and 21C are illustrated in these figures.

  At each position of the point corresponds

  energy error E. The minimum standard energy error

  E is a very convenient way of measuring the performance of a Wiener-type pulse shaping filter and, more particularly, a tip-forming filter. When the filter behaves perfectly, E = 0, which means that the desired and actual filter outputs match for all time values. On the other hand, the case E = 1 corresponds to the worst case, that is to say that there is no concordance between the desired outputs and the actual outputs. In place of the quantity E, it is desirable to consider the

  state complement "a" of E which will be called "para-

  PIt efficiency meter filters.

P = 1 -E (46)

  We then have a perfect efficiency of the filters when P = 1, while we have the case the

less good when P = 0.

  Figure 22 schematically illustrates the process of measuring the performance parameter

  P. A computer 400 has three channels of

  401, 402 and 404. From the storage and call element 403, the input channel 401 receives data representing a valve wave Bt; of the time shift control element 405, the input channel 402 receives peak data for different delays, while the input channel 404 receives, from the storage time control element 406, related data

  at the tips for different storage times.

  The output 410 of the computer 400 provides, by means of a measuring apparatus 411, a measurement of the output parameter P. For a constant filtration duration, it could be supposed that there must be at least one time value of offset at which the value P is as large as possible. FIG. 23 gives a diagram of the value P with respect to the time of offset of the output peaks for a family of filters of a fixed duration. It will be observed that the highest point of the curve (point M1) corresponds to a time offset ON1, the choice of this offset time leading to the optimum time shift filter. It will be recalled that the curve of FIG. 23 relates to a duration filter

fixed.

  One can also observe what happens when you increase the filter storage time to a constant time offset. The

  Figure 24 gives a diagram of the value P vis-à-

  screw the length of the filter for a shift of

  desired peak time and fixed. We can consider

  that this curve is monotonic and that it approaches asymptotically the highest value of P when the length of the filter

  increases more and more. The graphs represent

  23 and 24 are obtained by means of the system shown diagrammatically in

figure 22.

  The two design criteria

  which have been described here, are the

  time and the storage time of the filter.

  We can always improve the yield by increasing

  the duration of the storage function, however,

  since, for physical reasons, this duration can not be of indefinite length. On the other hand, one can look for the desired output time shift leading to the most

  high for a given chosen duration of the filter.

  This time lag in the output of the filter has

  no detrimental effect and can significantly improve-

this output of the filter.

  The efficiency parameter P of the filter as a function of the time offset and of a constant duration (FIG. 23) or the parameter P depending on

  the duration of the filter memorisation for a deca-

  The constant time (Figure 24) is useful, but does not provide all the indications. Ideally, we would like to study the extent to which the P value depends on the time lag and duration

  memorization for all physical values-

  these variables. One method that could be adopted for this purpose would be to draw a diagram of the value P by taking the time offset of the filter in ordinate and the storage duration of the filter, in abscissa. It would then be possible to draw the contour of the area of the values P so that one can see, at first glance, the combination of time offset and storage duration which ensures the optimum efficiency of the filter. Figure 25 gives a projection only for the contours of P, P2 and P3 * Of course, these are the largest

2 4 75 1 11

  P values which are the most interesting because it's

  precisely to these values that we obtain the best

  their filter performance. By examining these plots, one can choose the best combination between the time shift and the filter storage time.

  XIX. CODE OF PULSES IN TIME.

  Although, in the examples given,

  telemetry systems have been described in

  which one uses a system with binary code, other codes are sometimes better appropriate. For example,

  it may be preferable to adopt a code impulse

  time for a gamma ray sensor or an electronic compass inclinometer. In

  certain cases and, in particular, where the trans-

  sequential mission of several numbers is

  quise, a pulse code over time offers advantages. For some compass systems

  electronic systems, it is necessary to ensure the trans-

  sequential mission of five numbers to measure the magnetic range. By using a time pulse code telemetry system, considerable savings can be achieved both in the energy required for the battery and the time required for the battery.

transmission of information.

  A conventional pulse code transmission system is illustrated in FIG. 26A (for example, for the transmission of

  values of three parameters). A series of impulses

  voltage is transmitted and the duration (t1, t2, t3) of each pulse a, b, c represents (for example, proportionally or inversely

  proportional) the amplitude of the parameter to be trans-

  to put. It should be noted that between each pulse, a rest period is necessary to separate the end of one pulse from the beginning of the next. Thus, in FIG. 26A, the pulses a, b, c are somewhat analogous to three separate "words" -binary

  each of the following a lapse of time Tw. Well understood

  These idle times are detrimental to fast data transmission, since no information is transmitted during these rest periods. In addition, long-lasting impulses are detrimental to the telemetry system

of the present invention.

  According to the invention, a type of pulse code in time of the type represented in FIG. 26B is proposed. In this system, it is not the duration of the pulse that is a measure of the parameter,

  but the time elapsing between very short

  successive sessions. Instead of transmitting lon- :.

  impulses of variable duration, only short pulses of practically

  constant (in the telemetry system of the pre-

  invention, pulses of a duration of

  milliseconds), while time is running

  between impulses is a measure of the amplitude

  study of the parameter. Therefore, no duration is

  necessary to separate a lapse of time

  one parameter) of the next. In FIG. 26B, the parameter n0 1 is represented by the time (1) flowing between the pulse P0 and the pulse P1. The parameter n 2 is represented by the time (t2) between the pulse P1 and the pulse P2, while the parameter n 3 is represented by the time- (t3)

  between pulse P2 and pulse P3 '

  tate that, in the example above, the pulse P1 represents the end of the time period t1 and also

  the beginning of the period of time t2, whereas the impulse

  sion P2 represents the end of the time period t2 and also the beginning of the time period t3, etc. from

2 4 75 1 1 1

  there is no loss of time between each significant period of time (ie, the

time Tw of Figure 26A is zero).

  Therefore, we can see that

  reading the pulses P1, P2 and P3 both to indicate the end of a lapse of time and also the

  beginning of the next, the loss of time (

  read) is zero, while all the time used for data transmission (i.e., identifying timeslots t1, t2, t3) is a useful time. In terms of binary coding, each

  "word" (identifying a number) is followed immediately

  the next "word" and so on. It is only at the end of a transmission sequence that

  has a rest time Tp, then the sequence is repeated.

  Of course, during the next sequence,

  the time between P0, P1, P2, P3 will be

  somewhat different since the data represented by times t1, t2, t3 usually vary with time; in addition, each new information load transmitted represents, for example, a new

  existing condition in the borehole.

  Figure 30 shows the principles of the circuit for realizing the pulse code in time according to the present invention. Of course, in the practical instruments

  part of a borehole, it should be

  use modern electronic integrated circuits (for example, a bilateral switch of type "ICD40661") (see also Figure 29). In order to

  to facilitate the presentation, we will describe a simple

  a mechanical relay step by step, as well as a simple mechanical relay so as to be able to illustrate simply

  the principles of the logic of the system.

  In Figure 30, sensors are connected to

  corded to terminals 1, 2 and 3 of the switch not to

  285 having an electrical pick-up winding

  tromagnetic 286. It will be assumed that the sequence is started while this step-by-step switch

  occupies the position 110 "shown in FIG.

  The accumulator 288 produces a reference DC voltage. This DC voltage appears at the resistor 289 and it charges the capacitor 290 at a predetermined rate by the

  the value of resistance 289, the dimension of

  290 and the voltage of the accumulator 288.

  Reference numeral 291 designates a circuit of

  triggering a clear electrical impulse

  when the voltage applied to its input exceeds a predetermined value: .iine (trip voltage). The Fa circuit exit

  trigger 291 energizes the winding 286.

  while the 287 arm of the switch not to. step 285 moves to the next contact (in this case, the

  contact n 1). At the same time, the trigger circuit

  291 temporarily activates relay 292

  which discharges the capacitor 290 to the ground.

* When the arm 287 moves from the 110O position to the "1" position, the process

  repeats itself, with the exception that instead of -

  reference voltage of the accumulator 288, the output voltage of the sensor 1 is connected to the circuit, while the pulse P1 is emitted at the moment when the capacitor reaches again the tripping voltage of the circuit 291. This period of time is proportional to the value (RC) o VR indicates the resistance value of resistor 289,

  C, the capacity of the capacitor 290 and Vs, the voltage

  output of the sensor. Therefore, the time t1 is inversely proportional to the output voltage of the sensor. After excitation of the tripping circuit 291 by the voltage coming from the sensor

  n0 1, the process is repeated and, when the conden-

  The controller 290 reaches the trip voltage, the circuit 291 emits a net pulse which drives the relay 292, discharges the capacitor 290, energizes the step switch 285 and moves the arm 287.

  from the latter to the next contact.

  Therefore, the switch 285 advances stepwise and connects the sensors 1, 2, 3 one after the other to the resistor 289. The pulse emitted by the trip circuit 291 when the arm 287 is in position "011 , corresponds to the pulse P0 (of FIG. 26B), whereas the pulses emitted by the trip circuit

  when the arm 287 occupies the respective positions

  "lli 1'211 l3t 'correspond to the impulses re-

  Pl. P2 and P3 (of Figure 26B). The corresponding time periods t1, t2, t3 represent (inversely proportional) the

  voltage outputs of sensors nos. 1, 2 and 3.

  The above paragraphs describe the time pulse encoder principle that can be used in downhole equipment instead of the analog-to-digital converter 102 illustrated in FIG. 4A. Surface decoding can be performed by the conventional pulse-time decoding circuit and it does not require here

no detailed description.

  In Fig. 26C, TP 0 TP1, TP2, TP3, etc. represent the successive pulses received

  at the point of detection on the surface. These pulses appear

  at times T0, T1, T2, T3, etc. respected

  tively. In the pulse code in time described with reference to FIG. 26B, the time elapsing between successive pulses is

  used to indicate the amplitude of a parameter.

  Thus, if it is necessary to carry out tele-

  measurement of three parameters, the code can be-that represented in FIG. 26C o

  T1-T0 is the period of time representing the

meter n0 1

  T-T1 is the period of time representing the

meter n0 2

  T3-T2 is the period of time representing the

meter # 3.

In measurement operations by

  mud pulsation while drilling, it is ne-

  in some cases, the measurements should be performed with great precision. Since the speed of sound in the mud column is not always constant and the noise and attenuation conditions vary, the time lapse between pulses received at the surface is not in agreement. precise with the lapse of

  time between the corre-

  produced in equipment installed at the bottom of the borehole. In other words, there is frequently a surface uncertainty about

  exact moment at which a specific pulse arrives.

  It will be assumed that the absolute uncertainty of each moment of arrival of a pulse is plus or minus 0.2 seconds, for a total of 0.4 seconds. In order to obtain an accuracy of + 1% for T-T0 with a total absolute error of 0.4 seconds, the time between pulses must be at least 0.4 x 100 or 40 seconds. Moreover, since the device could sometimes not give a clean and clear impulse, at least

  two detection scans are needed.

  If both scans give the same answer, then the data should be "checked". In

  consequently, in order to obtain the precision and

  desired study (for a practical case, a precision

  + 1%), it takes about 80 to 120 seconds

  measured meter (that is, about 2 minutes per parameter). In the improved time pulse code according to the invention, an addition is proposed which, in many cases, can provide precision.

  much bigger. For each trans-

  P0, P1, P2, P3, it is proposed to use not a single mud pressure pulse, but a group of at least three mud pressure pulses unequally spaced as shown in FIG.

  Figure 26D (hereinafter referred to as "triple group").

  1S It is assumed that in each triple group the time differences are as follows: Time between first and second

pulse = t1.

  Time between the second and the third

pulse = t2.

  Time between first and third

pulse = t.

  Under these conditions, T0 again represents the arrival time of the triple TP0 v T1 represents the arrival time of the TP1 triple group; T2 represents the arrival time of the triple group TP2 and T3, the arrival time of the TP3 triple group and

  T - T0 is a lapse of time representing the para-

meter n 1.

  T2 - T1 is a period of time representing the

meter no 2.

  T3 - T2 is a period of time representing the

meter n0 3.

  The advantage of this system lies in

  the fact that, in the event of a momentary failure,

  not allowing the reception of an impulse

  group, this failure can be detected

  immediately, since both a group

  triple contains two pulses instead of three.

  Moreover, since the times t1, t2, t3 are unequal and known, the missing pulse of the group can be determined; in addition, given again

  that tI, t2, t3 are known, it is possible to apply

  appropriate timing and determine the T1-To, T1-T2> T2-T3 times with the same precision as in the case where all pulses are present in the triple group. The triple group system offers another advantage: since it is difficult

  to determine the exact moment of arrival of a: n

  particular drive, the triple group allows a

  much more precise determination of the moment

  riveted. For example, we could take the mo-enne

  arithmetic of the arrival times of each ipu-

  of the triple group or, by adopting

  modern computers, one could get one -

  even greater accuracy of the time of arrival.

  Figure 29 shows a block diagram

  of an installed electronic logic system

  at the bottom of a borehole and emitting the impulse

  triple groupings illustrated in Figure 26D.

  Reference numeral 101 designates a sensor (see FIG. 4A) producing an electrical voltage indicating the amplitude of a parameter at

  bottom of a borehole. The figures for

  601, 602 and 603 respectively denote a voltage-controlled oscillator, a framing apparatus and a triggering circuit emitting,

  well known, a series of electrical impulses

  separated by time periods indicating the magnitude of the voltage output of the sensor 101 and hence the parameter to be measured at the bottom of a borehole. Consequently, the time lapse between the pulses P0 and P represented in FIG. 26B is a measurement of a parameter measured by one of the sensors 101 of the

Figure 4A.

  The part of Figure 29 which is surrounded by a rectangle in broken lines,

  shows the details of the circuit emitting the impulse

  triple groupings that have been mentioned above. The reference numerals 607, 608 and 609 represent "single-wire" electronic univibrators which, in response to the pulse emitted by the trigger circuit 603, each emit an individual output pulse having the durations D1, D2 and D3 respectively represented.

  in Figure 29. Boxes 610 represent deviations

  electronic transmitters each emitting, at their output, a signal which is proportional to the first derivative

  time of the input signal (these electronic devices

  are well known in the art). In

  Consequently, their outputs will be those

  G, H and I in FIG. 29, i.e., two pulses of opposite polarity separated by respective time periods D1, D2 and D3. Cases 611 represent rectifiers transmitting only the positive pulses appearing at the outputs of the differentiators 610. The outputs of the rectifiers 611 are connected in parallel at 612 and they emit the signal J which is the desired signal

  (Also shown in Figure 26D). As a result

  quence, each individual impulse emitted by the

  trigger circuit 603 will produce three pulses

  separated by known and unequal time periods (triple group) as represented by J.

24 75 111

  In practice, the time interval D1 is very short compared to the time periods D2 and D3 and it will only be a few microseconds, while the time periods D2 and D3 are time periods of a few milliseconds to several seconds.

  hundreds of milliseconds. Therefore, in the

  lysis of the operation, we can assume that D1 0.

  Accordingly, in the output indicated at J in FIG. 29, the pulse P1 indicates the end of the output pulse coming from the element 607 (which, as pointed out above for all the applications practical, is also the beginning of the output impulse, since

  its length is assumed to be zero); lVim-

  pulse P2 indicates the end of the output pulse from the element 608, while the pulse P indicates the end of the output pulse from the element 609. Therefore (since D is assumed to be equal to zero), the time t = D2; the time t3 = D3 and the time t2 = D3 - D2. Therefore, the part of Figure 29 that is in the broken line rectangle, creates the triple group at its output at 612 (as shown in J in Figure 29) in response to

  a single pulse applied to its input.

  The circuit shown in FIG. 29 can be interposed, in FIG. 4A, between a selected sensor 101 and the driving source 104.

  words, when using the impulse code system

  time in Figure 29, the converts

  Analog / Digital 102 and the processor 103 are eliminated (since they are suitable

  for a binary coding), while the source

  trice 104 is ordered directly from the exit

  of the amplifier 613 of FIG. 29.

  When using pulse code in triple group time in place of the

  binary code, it is necessary to decode these

  triple group gals on the surface. In FIGS. 9, 12, 13, 17 and 19, the signals representing the parameter at the bottom of the borehole are assumed to be in binary code. In order to modify the system so that it receives signals in the pulse code in the triple group time described with reference to FIGS. 29 and 26D, it is necessary to interpose between the filter 150 and the apparatus later on the surface, a "translator of

  special code such as that shown in Figure 27.

  For this purpose, the conductor 151 illustrated in FIGS. 9, 12, 13, 17 and 19 will be interrupted and the "code translator" will be inserted. In some cases, it may be more desirable to interpose the "code translator" between the subtractor 160 and the analog / digital converter 163 on the conductor 162 in FIGS. 9, 12, 13 and 17;

  Moreover, a person skilled in the art will clearly recognize

  the favorite place this "code translator"

must be inserted.

  Referring to Figure 27, the figure

  Reference fre 316 refers to a "selector" designed

  to emit a single output pulse

  in the triple group described with reference to FIG. 29. The reference numeral 317 denotes a time / amplitude converter, that is to say

  an electronic circuit providing, on the con-

  319, a current output voltage of

  tinu which is a predetermined function of time

  flowing between two input pulses applied

  318. These devices are well known in the electronics art and they do not require any

  detailed description. The reference number 320

  means an analog / digital converter

well known in the art.

  Figure 28A shows the selector 316 in more detail. Reference numerals 321, 322 and 323 refer to monostable univibrators designed to emit a single output pulse of a predetermined and predetermined duration in response to an input pulse. The monostable univibrator 321 emits a long pulse 13: the univibrator 322 emits a pulse of more

  short-term 12 and the 323 transceiver

  output pulse of still shorter duration ll, as indicated above each of the boxes 321, 322 and 323. The boxes 324 designate differentiators,

  that is, electronic circuits that produce

  an output proportional to the first time derivative of a signal applied to the input. These units

  are also well known and they emit.

  output signals as indicated on the curve

  produced above each box designating a deri-

  vateur. The boxes 325 designate "inverters", that is to say devices emitting an output signal which is a replica of the input signal but whose sign is inverted as shown on the curve reproduced immediately above. of each inverter. Each of the elements represented by boxes 326 includes a rectifier and these elements

  emit, at their output, a single electrical impulse

  positive pole 326a, 326b and 326c, as shown

  in the drawing. Boxes 327 designate "cir-

  coincidentally "or" AND gates "which are all well known in the art.

  327 emits an output pulse at its terminal "c" uni-

  when there is, at the entrance "a" and at the entrance

247511 1

  "bl", impulses that are in coincidence

  time. The outputs of the three circuits to be

  327 are connected in parallel to the conductor 329 and are applied to the input of the converter.

  time / amplitude 317. The time / amplitude converter 317 produces a DC output voltage which is a function

  predetermined time between two impulses

  successive entry The output of the

  The time / amplitude weaver 317 is connected to the analog / digital converter 320 which then translates the DC input voltage into pulses-coded according to the binary mode so that

well known in the art.

  The circuit shown in the figures

  27 and 28A consists of integrated circuit elements

  conventional electronic grits that are well known in the art. The general functioning of the

  "selector" requires a more detailed description.

  The pulses P1, P2 and P3 produced

  amplifier 613 of FIG. 29 are

  at source 104 of Figure 4A and

  they are transmitted on the surface as impulse

  mud pressure by valve 40. In addition,

  face, these mud pressure pulses are captured

  eg by elements of Figure 9,

  including the transducer 51, the filter 150, the elem-

  152 and the 160 subtractor.

  impulses appearing on the driver 162 of the

  Figure 9 (or Figures 12, 13 and 17) will be

  TP1, TP2 and TP3 (these impulses correspond to

  P1, P2 and P3 pulses emitted by the electronic system installed at the bottom of the hole of

drilling and illustrated in Figure 29).

  Figures 28B, 28C, 28D and 28E show

  trent the response of the circuit of Figure 28 to the pulses TP1, TP2 and TP3. When the pulse TP1 arrives and is applied on the conductor 151 (or the conductor 162) of FIG. 28A, the three monostable univibrators 321, 322 and 323 are triggered and they each emit a respective output pulse having its own characteristic, thus that a fixed and predetermined length 13, 12 and 11 respectively. Consequently, when the pulse TP1 triggers the univibrators, they produce the output voltages (pulses) A1,

  B1 and C1, as shown in FIG. 28B.

  When the pulse TP2 arrives, it can not trigger the univibrator 321 because it is already in the state "tEN CIRCUIT". However, the impulse

  TP2 triggers the univibrators 322 and 323,

  that both have returned to the "OFF" state by output pulses B2 and C2, as shown in FIG. 28B. When the pulse P arrives, it can not trigger a monostable univibrator 321 or 322, since also

  although these are already in the "ON" state.

  However, the pulse TP3 triggers the throttle 323, since the throttle has returned to the "OFF" state by producing the output pulse

C as shown in Figure 28B.

  The time frames 13, 12 and 11 of the

P P 37 21

  vibrators 321, 322 and 323 of Fig. 28A are calculated to correspond to delays due to the action of the univibrators 609, 608 and 607 of the

  Figure 29; therefore, the end of the

  pulses of the univibrators of Figure 28B is in "coincidence" and these pulses operate the

AND gates of Figure 28A.

74;? 111

  Figure 28B shows the conditions

  operating conditions when all the

  pulses TP1, TP2 and TP3 are present.

  Figure 28C shows the same con-

  than in Figure 28B, but with impetus

  missing (for example, TP1 pulse).

  Figure 28D shows the same

  conditions, the TP2 pulse being, however,

  while Figure 28E represents the same conditions, but the TP3 pulse is missing. It should be noted that it does not matter which pulse is missing (TP1, TP2 or TP3), two output pulses of univibrators

  always ending at the moment T. This characteristic

  The circuit circuit of FIG. 28A is used to produce at least two time coincidence pulses at time T regardless of the missing pulse among pulses TP1.

  TP2 and TP ô. Provided that at least two of the

  Group pulses are detected, the moment of appearance of the output pulse at 329 (Figure 28A) will be the same. The single pulse 328a of Fig. 28A is produced each time a "group" of pulses is received by the apparatus at the top of the hole, while the pulse 328a will be present whenever any two pulses occur.

  "group" are detected on the surface.

  Referring again to FIG. 28A, box 317 denotes a conventional time / boost converter producing a DC output voltage having a predetermined functional relationship with time flowing between

  successive pulses 328a. The reference figure

  320 refers to an analog / digital converter

  Conventional converter converting the amplitude of this DC output voltage into a binary word. Binary words follow each other

  in a rapid succession determined by the characteristics

  characteristics of the converter 320 and the associated clock. As a result, it can be seen that

  the apparatus of Figure 28A translates the impulse code

  time, in which the group is used.

  triple, in a binary code, while the

  similar to those following the driver 151 or 162

  Figures 9, 12, 13 and 17 will function exactly as

  in the same way as if the data had originally been transmitted in a binary

from the basement.

XX. ADDITIONAL REMARKS

  (1) In order to obtain the shock waves described in the speci-fication above, some

  limits are imposed on the value K2 (average speed-

  do not change the opening of the clapezj, so (y) than the value T () [open flow time). Experiments have shown that the Kn value should be at least 5 cm2 / secorde and that it should preferably be in the range of 20 cm2 / second to 150 cm2 / second. The value T (V) must b

  No more than 500 milliseconds and,

  it should be in the meantime

  ranging from 50 milliseconds to 150 milliseconds.

(2) Although the impulses of syn-

  Chronization (clock 155) of the described examples are emitted by the generator connected to the pump shaft or by the rigid phase loop described in the related patent. it should be noted that it is possible to provide other elements ensuring the clock frequency in synchronism with the operation of the pump. For example, we can use the

  well-known and popular "pump stroke meter"

  It is adapted to the pump connecting rod to produce an electrical impulse per stroke of the pump. The period between each of these successive pulses can be divided into an appropriate number (for example, 512 or 1.024) of equal time by a microprocessor or by a rigid phase loop or by another

  well known method in the electro-technical

  and computers. In a system of this type, it is not necessary to have access to the transmission shaft of the pump and the clock frequency equal to that of the generator 155 can be produced by the microprocessor and the switch.

pump stroke meter.

  (3) In the first part of the preface

  In this specification, the conditions of formation of hydraulic shock waves and associated "valve flanges" have been described in detail. At certain depths, for example, at shallow depths, conditions may occur in which the above-described valve wave is not suitably formed. For a valve wave of this type, it is necessary that a sufficient volume of sludge flows in the drill pipe and that there is sufficient hydrostatic pressure at the end of the transmitter. It is clearly understood that the present invention is not limited to the particular valve wave shown and can be applied to other forms of pressure pulses detectable on the surface of the valve.

  ground following the operation of the valve 40.

  (4) Different digital filters,

  in particular adapted filters, filters

  impulse meters and tip-forming filters, have been described above with many

  details. In particular, it was clearly explained

  the performance of each digital filter in

  providing for a detailed series of operations to be carried out

116-

  kill. These operations have been explained and

  by appropriate mathematical formulas.

  It is clearly understood that by adopting

  In the modern world of calculation, the skilled person can establish the necessary programs on the basis of

  descriptions given in this specification,

* while the operations described with reference to Figures 9, 12, 13, 14, 16, 17, 19, 21 may

  be performed by appropriate software.

  (5) Different digital filters that.

  have been described, can also be applied to other forms of measurement transmissions by sludge pulsations by means other than the bypass valve of the type described in this specification. These other forms may include the process based on the controlled throttling of the sludge flow circuit by means of a throttling valve-judiciously installed in the main sludge stream as described in

  U.S. Patent No. 2,787,795 ac-

  strung in the name of J.J. Arps. In general, the digital filtering systems that have been described can be applied to any type of telemetry system with logging while drilling, as well as to other forms of logging in which the equipment of

  drilling is withdrawn in order to let down the equi-

  measuring instrument in the borehole. These systems

  Digital filtering systems can be applied to telemetry systems in which pulses representing any form of energy are used, for example, electrical, electromagnetic, acoustic, and other impulses. (6) The pulse code over time

  in which the triple pulse group is used

  described above, may also be

  a well acoustic logging system to facilitate the processing of acoustic well logging signals and to obtain a highly efficient method of automatically correcting errors due to pulse jumps during the measurement of a well. transit time of acoustic waves. Processes and devices

  Acoustic logging of a well is usually

  designed to measure the transit time of an acoustic wave between a first and a second pulse. In U.S. Patent No. 3,900,824 issued to J.C. Trouiller et al. On August 19, 1975, it has been proposed to prevent pulse jumps by practicing measurement during

  of an N-1 sequence stored in a memory

  xilary and comparing this measurement with the next measurement (sequence N). U.S. Patent No. 3,900,824 is incorporated herein by reference.

  reference. The other proposed method according to the

  and based on the pulse code over time makes it possible to more effectively and reliably correct measurement errors due to

cyclic jumps.

  (7) The tri-group code of impulse

  in time offers a very wide range of applications apart from

  drilling course. It can be adopted in any

  What communication system for the transmission

  message transmission between a transmitting station and a receiving station, as well as in different types of well logging systems (and not necessarily for logging during drilling), for example, acoustic logging

(see note 6).

2475 1 11

  (8) It is understood that in order to capture and

  memorize a wave for later use

  In the digital filters described in this specification, some measurements must be taken in situ as described. he

  It is sometimes desirable to pick up an indi-

  vidual (instead of a double wave), as is the case in the embodiment of Figure 19

  the tip forming filter 351A is provided.

  In order to capture an individual wave, it is necessary to synchronize the transmission of the signal produced by

  equipment in the basement with the detection equipment

  surface. This synchronization can be performed by replacing one of the sensors 1, 2, 3 and

  4 of the surface equipment (Figure 4A) by a

  positive such as a "clock" or a generator of constant signals controlled in time ensuring operations of the valve 40 of Figure 4A, spaced uniformly over time. The operation should then proceed as follows: (a) By stopping and starting the surface slurry pumps in the proper order, the switch 91 illustrated in Figure 4A can be made to connect the modified sensor (ie to say

  the pulse generator uniformly spaced).

  Since then. a series of pulses will be emitted by the valve at known times. (Of course, corrections must be made for the travel time of the pulse between the subsoil and the

  soil surface, this time having been determined

  later by well known methods).

  (b) The surface equipment is

  controlled by its own clock which is in sync with

  in time and in phase with the transmitter

signals in the basement.

24? 51 1 41

  (c) By appropriate switching

  surface, the capture and storage can be interrupted

  of the double wave so that the memory circuit is only connected during the time of a damper wave, and then disconnected automatically during the occurrence

of the second wave.

  Of course, the same operation can be performed manually (by the operator), this operation is easily performed when this

  wave is distinct and clearly overcomes noise.

  When this wave is embedded in the noise, it is

  reads the automatic system as described

in this specification.

  (9) There are two noise signals

  disrupters who tend to conceal the reception of

  useful signal B (t) (see equation 22). One of these signals represents the pump noise P (t) and the other, the noise U (t) associated with the various drilling operations other than the operation.

  pump. In order to eliminate these disturbing

  According to the invention, three

  filter systems # 1, # 2 and # 3.

  The filtering system No. 1 is the analog filter 150. This filter is intended to eliminate the constant component of the output of the transducer representing the pressure produced by the pump 27, as well as other frequencies.

  outside the relevant range.

  The filtering system n0 2 includes

  a delay element 152 and a subtractor 160.

  This system is intended to remove or eliminate

the pump noise P (t).

  The filtering system No. 3 includes a correlator or a digital filter which can

  be a suitable filter, a filter shaper of im-

247511 1

  pulses or a tip-forming filter and also includes various associated elements such as storage and call elements, as well as computers for determining the optimum values of the storage elements for the corresponding digital filters (see FIGS. , 12, 13, 14 and 15). System No. 3

  for the purpose of eliminating or suppressing the noise U (t).

  The filtering systems - Nos. 1, 2 and 3 are connected in cascade. In the embodiments of the invention which have been described above, the filtering system No. 1 is connected to the pressure transducer 51, the system No. 2 is connected to the output conductor 151, while the system No. 3 is connected to the output conductor

164 of System No. 2.

  Each of the filtering systems

  above is a linear system. Consequently,

  the functions of these systems may be

  changed or reversed. We can start with the filtering system no 1, then interchange the order of the filtering systems n0 2 and n0 3. Similarly, in some cases, it may be superfluous to use the three filtering systems. Any two filtering systems may be sufficient and in some cases only one may suffice. Similarly,

  the system mounted between the driver 182 and the con-

  driver 210 can sometimes be removed, while digital to analog converter 211 can be

  designed to accept double waves.

  (10) When the signal formed by the method described in chapter XIII (steps a to f) is picked up and stored, it can be correlated

  crossed with the raw signal emitted by the transduc-

  51 or with the preconditioned signal on the conductor 162 of FIGS. 9-19. In the case of a cross-correlation with the raw signal transmitted to the transducer 51, the second double wave will have to be removed by appropriate means well known in the art so that it can be cross-correlated with the signal. wave

  single emitted at the output of the transducer 51.

122.

Claims (10)

  1 - Sludge circulation system used during the recording of information during drilling, system in which mud pressure pulses are produced by the opening and closing of a valve, characterized in that it comprises a sensor (101) for sensing the value of a parameter in a borehole, a first member (79) for holding the valve (60) for relatively long periods in an open or closed position, and a second element (96) responsive to this sensor to ensure, sequentially, the opening of this valve in   a relatively short period of time, and then closing this cla-   also in a relatively short period of time, thus creating   clear pressure pulses in a sequence both the value of this parameter.   2 - System according to claim 1, characterized in that the speed K of each openLra of the valve (40)   as well as the speed K of each closure of the valve, de- 2 (y)   pass the value of 1CO cm / secon-de, K = SoT, So rewrite   sensing, in cm, the opening section of the clape: when the latter is completely open, while T indicates, in   seconds, the opening or closing period of the flap.   3 - System according to claim 1, characterized in that the first element (79 :) is designed to maintain the stability of the valve in the open or closed position., 4 - System according to claim 2, characterized in that the first element ( 79) is designed to maintain the   valve stability in open or closed position.   System according to Claim 3, characterized   in that the flap (40) is an electrically controllable flap   while the second element (96) is a generic element   electric current generator for producing short current pulses for opening and closing this valve. 6 - System according to claim 1, characterized in that the opening time of the valve (40) is short enough to create, at the top of the borehole, a reduction, 123. and a pressure rise following this opening of the   valve, while the closing time of the flap is sufficient   fisely short to produce, at the top of the hole, a rise   tion, then a reduction of pressure following this closing of the valve. 7 - System according to claim 1, characterized in that the opening time or the closing time of the   valve (40) is short enough to produce, within the   mud flow system (22, 23, 24) a hydraulic shock wave after an opening of the valve and a wave   hydraulic shock after closing this valve.   8 - System according to claim 4, characterized   in that the opening time of the valve (40) is sufficiently   short way to produce, at the top of the borehole, a reduction, then a rise in pressure following the opening of the valve, while the closing time of the valve is   short enough to produce, at the top of the hole, a   then a pressure reduction following this closure. shutter valve.   9 - System according to claim 4, characterized   in that the opening time or closing time of the   pet (40) is short enough to produce, indoors   of this mud circulation system, a hy-   hydraulics after valve opening and shockwave   hydraulic after closing this valve.   System according to Claim 1, characterized   in that the speed of opening of the valve (40), as well as the closing speed of the latter, are calculated so as to produce, in this sludge circulation system (22, 23, 24), a velocity of increase or reduction rate of the sludge flow which, in absolute amplitude, exceeds 2 x 105 cm / second2.   11 - System according to claim 1, characterized   in that the relatively short period T (y) <20 milli- at - seconds.   12 - System according to claim 1, characterized   the relatively short period T (V) <5 millis- acces. 2 4 751 124.   13 - System according to claim 1, characterized   rised in that the relatively long period Tb (v) <0.25 se- count.   14 - System according to claim 1, characterized   (y) in that the relatively long period Tb () <0.1 seconds.   System according to claim 1, characterized   in that the total operating period of the flap (40) Tt (V) <270 milliseconds.   16 - System according to claim 1, characterized   rised in that the total operating period of the flap (40) Tt (V) <110 milliseconds.   17 - Sludge circulation system used during   logging information while drilling, -ystem-   in which mud pressure pulses are produced   by the recurrent operation of a flap, this system   comprising a sensor for capturing a value of a para-   meter in a borehole, an element reacting to this cap-   for supplying an electric current for actuating said valve, characterized in that it comprises   also an element (281) responsive to the movement of this key   pet (40), when it is actuated to control this current. 18 Sludge circulation system used when logging information while drilling, system   in which the mud pressure pulses are produced   by the opening and closing of a valve, characterized   in that it comprises a sensor (10) for measuring the   of a parameter in a borehole, a first element   solenoid (79) for actuating this valve (40), a second controllable element (96) comprising a capacitor (452), a power source (104, 450) for generating an electric current for charging this capacitor, as well as a resistor (451) interposed between this source and this capacitor, this resistance being provided to maintain this current at a substantially constant value during the charging of this capacitor, a third element (102, 103) 125 reacting to this sensor in order to control this second   element to make successive discharges of this   denser and apply current pulses resulting   from these discharges to this first element in order to effectively   the valve, thereby producing impulses   pressure in a sequence representing the value of this para- metre.