EP0026706B1 - Process and device for determining the directional parameters of a continuously explored well - Google Patents

Description

The present invention relates to a method and apparatus for determining direction parameters of a well as a function of depth, and more particularly to a method and apparatus which use the measurement signals of an accelerometer and a magnetometer to three sensitive axes housed in a probe exploring the well. The probe is continuously moved into the well during the measurement. In the most common case, the accelerometer signal is prefiltered, then combined with the magnetometer signal to rid it of the alteration it undergoes due to the displacement of the probe in the well, then subjected to pass filtering. - very selective bottom, and finally again combined with the signal from the magnetometer for the determination of well direction parameters.

The earth's crust is made up of layers of natures, of thicknesses of various inclinations, and it has long been apparent that all information concerning the successive layers, and in particular their inclination, was of definite interest in fields such as that of oil research. However, since such information on the inclination of the layers is not directly accessible in the current state of the art, it is conventionally used to use probes, which are moved in a well passing through these layers, and which provide information on their orientation in relation to the layers crossed by the well.

It is easy to understand that this information representative of a relative orientation is insufficient and that it is therefore necessary, in order to know the orientation of the layers, to determine the three-dimensional topographic orientation of the well and the position taken by the probe in the well at the depth of investigation.

It is also well known in aviation techniques that the three-dimensional topographic orientation of an airplane or a rocket can be determined by using the measurement signals of an accelerometer and a magnetometer with three sensitive axes. . These signals are immediately usable when the aircraft is flying at a constant speed and has a regular trajectory. When there are sudden disturbances or accelerations, the signals from the accelerometer and magnetometer generally lose their interest in this orientation determination.

It is within this very general framework that the process described in United States patent no. 3,862,499 issued to CE Isham et al, as well as that of the present invention according to the preamble of claim 1, which relates determining parameters representative of the topographic orientation of the well and, although optionally, determining the position taken by the probe in the well.

According to the known method, the probe is lowered into the well and stabilized at a certain depth. The signals from an accelerometer and a magnetometer mounted in the probe are noted while the probe is fixed in the well in the absence of any disturbance. These stationary component signals are combined to obtain two well direction parameters, namely the deflection angle, defined as the angle formed between the longitudinal axis of the well and the vertical, and the azimuth, defined as the angle formed between two vertical planes, one of which contains the longitudinal axis of the well and the other the north direction. Then the probe is moved into the well and stabilized at another depth. New signals are produced when the probe is fixed, and are combined to obtain new values for the deflection angle and azimuth.

This process, despite the fact that it made it possible to obtain satisfactory precision on the parameters measured at each station of the probe, has several limitations, and in particular the two significant drawbacks below.

The first, immediate, is that the need to stabilize the probe for each measurement causes a detrimental increase in the duration of exploration of the well.

The second is that the orientation of the well between two measurements being physically unobservable, any variation in orientation, which appears and disappears between two successive measurements, critically increases the error on the location of the points of the well located below of this variation in orientation.

In this context, the object of the present invention is to propose a method and an apparatus for determining parameters of a well which is faster than the known method previously described.

Another object of the present invention is to propose a method and an apparatus making it possible to physically determine variations in the orientation of the well at any point of an explored longitudinal portion of this well.

The method of the invention according to the preamble of claim 1, is characterized in that the phases consisting in producing said acceleration and location signals and in moving the probe are simultaneous and substantially continuous, and in that said phase determination of the direction parameters includes a virtual stabilization step by which the effects of the displacement of the probe are eliminated, in the components of one of the signals, constituting a signal to be stabilized, by means of the components of the other signal, constituting a stabilizing signal, and an intermediate low-pass filtering operation involving at least stabilized components of said signal to be stabilized and by which frequency variations greater than the maximum frequency of variations attributable to the acceleration of the signal are eliminated gravity.

Preferably, said phase of determining the steering parameters further comprises a preliminary step to said virtual stabilization step, comprising an operation of prefiltering the components of the acceleration signal, by which the signal variations having a frequency greater than the greatest possible frequency of the rotational movement are substantially attenuated of the probe around its longitudinal axis.

In a simple manner, one uses respectively, to produce said acceleration and locating signals, an accelerometer and a direction indicator, this accelerometer and this direction indicator each having a first and a second sensitive transverse axis, perpendicular to each other and to the longitudinal axis of the probe, and a third sensitive axis, of longitudinal direction and coincident with the axis of the probe, said signals each comprising two transverse axial components and a longitudinal axial component and said direction indicator being for example a magnetometer giving in the coordinate system of its three sensitive axes the direction of the earth's magnetic field vector.

The preliminary step of the direction parameter determination phase comprises determining a transverse diagonal component of the stabilizing signal from the transverse axial components of this signal and eliminating the effects of rotation, using the axial and diagonal components. transverse of this same signal, in the transverse axial components of the signal to be stabilized to obtain stabilized components in rotation, corresponding to a reference position of the probe around its longitudinal axis.

The preliminary step comprises the operations consisting in: determining a transverse diagonal component of the locating signal from the transverse axial components of this signal; determine from this transverse diagonal component and the longitudinal axial component of this same locating signal the sign of the difference between a first angle, formed between said vector of fixed direction and the longitudinal axis of the probe, and a limit angle of predetermined value; defining the stabilizing and stabilizing signals, respectively by the locating and acceleration signals when the sign of said difference is positive, and by the acceleration and locating signals when this sign is negative; and determining a transverse diagonal component of the stabilizing signal from its transverse axial components when said stabilizing signal is defined by said acceleration signal.

When the sign of the difference determined during said preliminary step is positive, said final step comprises an operation of reintroducing the effects of rotation of the probe, supplying from two stabilized transverse axial components of the acceleration signal and transverse components diagonal and axial of the locating signal, two transverse axial components of the acceleration signal which are no longer stabilized again with respect to said reference position of the probe around its longitudinal axis.

Under these current conditions of exploration of the well, it is advantageous that the low-pass filtering eliminates, by a rapidly increasing attenuation from 3dB, the signal variations having a frequency higher than 8.10- ^Z Hz and that the pre-filtration consists an attenuation, increasing from 3dB, of the signal variations having a frequency greater than 2.5 Hz.

An apparatus according to the invention is described in claim 6.

• -La figure 1 est une vue schématique représentant, en coupe un puits exploré par une sonde;
• - La figure 2 est un diagramme fonctionnel (flow-chart) représentant les principales opérations de la phase de détermination de valeurs de paramètres dans le procédé de l'invention;
• -Les figures 3a et 3b sont des représentations schématiques de circuits multifilaires pour la circulation et le traitement des composantes des signaux d'accélération et de repérage dans la phase de détermination des paramètres, et sur lesquelles chaque conducteur est assigné à une seule composante d'un des deux signaux; ces figures sont raccordées par les faisceaux de conducteurs 0₁ et Y'2%
• -La figure 4 est un diagrame représentant les caractéristiques du filtre de préfiltrage;
• -La figure 5 est un diagramme représentant les caractéristiques du filtre passe-bas.

A particular embodiment of the invention will be described below, by way of indication and in no way limitative, with reference or annexed drawing, in which,

• FIG. 1 is a schematic view showing, in section, a well explored by a probe;
• FIG. 2 is a functional diagram (flow-chart) representing the main operations of the phase of determining parameter values in the method of the invention;
• FIGS. 3a and 3b are schematic representations of multi-wire circuits for the circulation and processing of the components of the acceleration and location signals in the phase of determining the parameters, and on which each conductor is assigned to a single component of one of the two signals; these figures are connected by the bundles of conductors 0 ₁ and Y'2%
• FIG. 4 is a diagram representing the characteristics of the prefiltering filter;
• FIG. 5 is a diagram representing the characteristics of the low-pass filter.

As previously indicated, the method of the invention aims to determine different parameters, related to the topographic orientation taken by a well 1 at a given depth.

To this end, there is an elongated probe 2 which is initially lowered into the well 1 by means of a cable 3 secured to the probe and wound on a winch 4.

Between the winch and the upper edge of the well, the cable passes over a measuring wheel 5 connected to a counter 6 recording the rotations of the wheel 5. The depth at which the probe is in the well, which obviously depends on the length of cable unwound from the winch, can in a known manner be deduced from the indication of the counter 6.

The probe 2 comprises centering poles 7 enabling it to always adopt a position in the well in which its longitudinal axis 2 a is, over the length of this probe at least, substantially coincident with the longitudinal axis 1 a of the well, l orientation of the probe axis thus assimilating to the orientation of the well at the exploration depth.

Inside the probe housed an accelerometer 8 and a magnetometer 9 firmly attached to the probe.

The accelerometer provides a signal with three axial components whose amplitudes represent the lengths of the projections, on three respective sensitive axes, of the vector associated with the set of accelerations undergone by the probe, and the magnetometer provides a signal with three axial components of which the amplitudes represent the lengths of the projections, on three respective sensitive axes, of the vector associated with the magnetic field passing through the probe, that is to say in practice with the terrestrial magnetic field.

However, the magnetometer 9 could be replaced by a gyroscope delivering a three-component signal constituting location information of the probe with respect to the characteristic direction of the gyroscope, or by any other direction indicator, provided on the one hand that the direction of the vector represented by the signal that this indicator would provide is fixed and known and on the other hand that it is different from the vertical.

The sensitive axes of the accelerometer and the magnetometer form a fixed rectangular trihedron with respect to the probe, the accelerometer and the magnetometer having a first sensitive axis in the longitudinal direction of the probe and two transverse sensitive axes.

The probe having been lowered into the well to a known depth is raised using the winch and the cable at a substantially constant speed while the accelerometer and the magnetometer produce their respective signals, which are transmitted to the surface by the cable. 3, and recovered at the surface in correlation with the signal from counter 6.

Due in particular to the irregularities in the wall of the well and the elasticity of the cable 3, the probe 2 is subjected to accelerations which, in addition to the acceleration of gravity, include the acceleration due to the movement of the probe 2 in the well. Indeed, on the one hand the probe undergoes transverse movements and shocks against the wall and on the other hand, despite the fact that the cable is rewound at substantially constant speed, the probe advances in the longitudinal direction by jerky progressions a movement called "yo-yo". In addition, the probe generally undergoes an additional movement of rotation about its longitudinal axis.

If we can consider the components of the locating signal, coming from the magnetometer, as appreciably independent of the sudden movements of the probe, on the other hand the components of the acceleration signal, coming from the accelerometer, are, to a very large extent , representative of these parasitic movements of limited amplitude, which critically disturb it.

The phase of determining the direction parameters of the well from the signals from the accelerometer and the magnetometer therefore requires different steps and operations aimed in particular at recovering from these signals the information that they would have provided directly if they had been produced while the probe was at rest and had not undergone any rotation around its longitudinal axis.

• -S désigne un signal quelconque de nature vectorielle, de composantes axiales S_x, S_v et S_z;
• ―S_xy désigne la norme partielle ou composante diagonale de ce signal;
  
  ;
• ―S_xyz désigne la norme:
  
  du signal S;
• ―S_ξo et S_ξ désignent une même composante axiale du signal S, respectivement avant et après une opération modifiant cette composante; ξo et ξ peuvent respectivement adopter les significations suivantes: x_o et x; y_o et y; z_o et z; x_oy_o et xy;
• -Sc désigne une composante normée si
  
• ―^γS et ^µS désignent respectivement les signaux d'accélération et de repérage, de nature vectorielle, respectivement issus de l'accéléromètre et du magnétomètre et ayant les composantes axiales respectives ^γS_x, ^γS_y, ^γS_z et ^µS_x, ^µS_y et ^µS_z;
• ―^aS et ^PS (a=actif; p=passif) désignent respectivement un signal stabilisateur et un signal à stabiliser, la nature de la stabilisation étant expliquée en détail ultérieurement.

In the description given below of these different steps and operations, the following definitions will be used:

• -S denotes any signal of a vector nature, of axial components S _x , S _v and S _z ;
• ―S _xy designates the partial norm or diagonal component of this signal;
  
  ;
• ―S _xyz denotes the standard:
  
  signal S;
• ―S _ξo and S _ξ denote the same axial component of the signal S, respectively before and after an operation modifying this component; ξo and ξ can respectively take the following meanings: x _o and x; y _o and y; z _o and z; x _o y _o and xy;
• -Sc denotes a standardized component if
  
• - ^γ S and ^µ S respectively designate the acceleration and location signals, of vectorial nature, respectively from the accelerometer and the magnetometer and having the respective axial components ^γ S _x , ^γ S _y , ^γ S _z and ^µ S _x , ^µ S _y and ^µ S _z ;
• - ^a S and ^P S (a = active; p = passive) respectively designate a stabilizing signal and a signal to be stabilized, the nature of the stabilization being explained in detail later.

Referring to FIG. 2, which represents the phase for determining the values of the parameters, it can be seen that this phase comprises a preliminary stage ETO, a virtual stabilization stage ET1 itself comprising an operation D ₁ or D ₂ of elimination rotation effects, and a final step ET2 of combining the processed components of the signals ^γ S and ^µ S, the step ET1 and the final step ET2 being separated by an intermediate operation OIF of low-pass filtering F ₂ 13 or F ₂ 47.

The preliminary step ETO comprises, in addition to possible operations 1 13 and 146 of inverting the sign of the components of the signals ^γ S and ^µ S, prefiltering operations F ₁ of the signal ^γ S, of delay R ₁ of the signal ^µ S, of normalization N ₁ of the signal ^µ S, of choice with test "T = 0?" and, optionally of normalization N ₃ of the signal ^γ S.

Operations 13 and 46 consist in changing the sign of the components of the signals ^γ S and ^µ S and are only necessary when the ETO step relates to the signals directly supplied by the accelerometer and the magnetometer as representative of vectors of directions opposite to those of the vectors of acceleration on the one hand and of terrestrial magnetic field on the other hand.

The prefiltering operations F, and delay operations R, will be explained in detail later.

In addition to obtaining pre-filtered components of the acceleration signal, the preliminary ETO stage has two essential purposes. In fact, as previously mentioned, the components of the acceleration and location signals generally carry information coming from a parasitic phenomenon, namely the rotation of the probe around its axis. To eliminate the effects of this rotation on the values of the transverse axial components of one of the signals hereinafter called "signal to be stabilized", use is made, in the subsequent step of virtual stabilization ET1, of the use of the components transverse axial and a transverse component, called diagonal, of the other signal, hereinafter called "stabilizing signal". However, depending on the topographic orientation of the longitudinal axis of the probe, it may be preferable either to use the components of the signal from the magnetometer to correct the components of the signal from the accelerometer, or, conversely, to use the components of the accelerometer signal to correct the components of the magnetometer signal. The preliminary step ETO therefore appears to have the function, on the one hand, of making it possible to determine which of the two signals ^γ S and ^µ S must play the role of signal to stabilize ^p S, the other signal obviously having to play the role of signal stabilizer ^a S, and on the other hand to supply, for the needs of the virtual stabilization step ET1, the diagonal transverse component of the stabilizer signal, that is to say ^a S _xy according to the notation previously introduced.

The operation of determining ^a S _xy , is included in block N ₃ or in block N, depending, respectively, on whether the role of ^a S is held by the signal ^γ S or by the signal ^µ S. However, as the choice with test "T ₁ = 0?" suppose, as will appear below, the use of the diagonal component of one of the two signals, and very preferably of ^µ S, we first determine ^µ S _xy during the operation N ₁ , we then use ^µ S _xy to conduct the test "T ₁ = 0?" which makes it possible to decide which of the two signals should play the role of stabilizing signal ^a S and determines ^a S _xy = ^γ S _xy during the operation N ₃ if the test "T ₁ = 0?" has led to assign to ^γ S the role of stabilizing signal ^a S.

The detailed description of the various operations of the whole phase of determining the parameters generally refers below to FIGS. 3a and 3b on which are represented lines of material or virtual information circulation, each assigned, unlike the case of FIG. 2 , single component or signal standard.

At blocks I 13, I 46; F ₁ ; R ₁ , R ₂ .14, R ₂ .59; F ₂ .13 and F ₂ .47 in FIG. 2 correspond respectively to the inverters I ₁ to I ₁₃ and I ₄ to I ₆ , the pre-filtration filters F ₁ .1 to F ₁ .3, the buffer cells R ₁ . 1 to R ₁ .5, R ₂ .1 to R ₂ .4 and R ₂ .5 to R ₂ .9 and filters F ₂ .1 to F ₂ .3 and F ₂ .4 to F ₂ .7 of the figures 3a and 3b.

The blocks N ₁ to N ₄ , D ₁ and D ₂ , E ₁ , DEV 1, DEV 2, RB 1 and RB 3 AZI1.1 and AZI1.2, AZIM1 and AZIM3 are to be considered as operations in FIG. 2, and as function generators, suitable for carrying out these operations, in FIGS. 3a and 3b.

In FIGS. 3a and 3b, the notations of the components of the signals do not take account, for reasons of simplification, of the fact that these components obviously keep, at all levels, the memory of the processing operations which have been applied to them in the blocks which 'they have crossed previously.

The axial components ^γ S _xo , ^γ S _yo , ^γ S _zo and ^µ S _xo , ^µ S _yo , and ^µ S _zo of the accelerometer and magnetometer output, available at the start of the parameter value determination phase, can be considered as each having on each of the elementary time intervals Δt, a constant amplitude.

The axial components of the magnetometer, of sign possibly corrected by the inverters I ₄ , I ₅ and I ₆ , are applied to the function generator N ₁ , which delivers at its output the standard ^µ S _xyz , the normalized axial components ^µ S _x = ^µ S _xo / ^µ S _xtz , ^µ S _y = ^µ S _yo / ^µ S _xyz , ^µ Sz = ^µ Szo / ^µ S _xyz and the normalized transverse diagonal component

The axial components of the accelerometer, of sign possibly corrected by the inverters I ₁ , I ₂ and I ₃ , are applied to the identical pre-filter filters F ₁ .1 to F ₁ .3.

avec a_k=0,54-0,46 cos

If _ξo represents x _o , y _o or z _o for a component before filtering, if ξ represents x, y, z for a component after filtering, if k and I represent whole numbers and if ^γ S _ξ.iΔt represents the amplitude of the component of ξ ^γ S signal in the j ^th time interval .DELTA.t, the characteristic of the filters F ₁ to F ₁ .1 .3 is to deliver to I, an output signal such that:

with a _k = 0.54-0.46 cos

The characteristic of these filters F, is represented on FIG. 4 which carries, on the abscissa, the frequency and, on the ordinate the attenuation, in the case where the value of each component of the signal ^γ S of the accelerometer is sampled every 8.3 milliseconds (At = 8.3 ms). New filtered components therefore appear every 15.5 Δt, or approximately every 1 / 7.5 seconds. The role of filters F, is to very significantly attenuate, in the filtered components, the signal variations having a frequency greater than the maximum possible frequency of the rotational movement of the probe around its axis. We see in Figure 4 that frequencies above 2.5 Hz undergo an attenuation greater than 3 dB.

As the appearance of the filtered component ^γ S _{ξ, 15.51Δt} supposes the previous appearance of the nonfiltered component ^γ S _{ξo, (15.5) (1 + 1) Δt} the output signal of the filter F ₁ has a certain delay compared to the input signal. As the components of the accelerometer and magnetometer signals relating to the same instantaneous depth of the probe in the well must obviously be used together, the components ^µ S _x , ^µ S _Y , ^µ S _z , ^µ S _xy and the standard ^µ S _xyz of the locating signal, coming from the magnetometer, undergoes in cells R ₁ .1 to R ₁ .5 a delay equivalent to that caused by filtering F, on the components of the acceleration signal.

est positive ou nulle, la sortie du comparateur COMP 1 se met dans l'état T,=0 (cas général), et, si u est négatif, dans l'état T₁=1 (cas particulier, le moins fréquent), T₁ étant par exemple défini par la fonction explicite T₁=1-INT 2 ^u-|u| où "INT" désigne la fonction "partie entière de". Ainsi, pour la valeur, généralement appropriée, de 5.10^-2 pour L1, la sortie T₁ du comparateur COMP 1 sera désactivée si l'angle

est supérieur ou égal à 3° (cas général).The divider DV, to which the components ^µ S _z and ^µ S _xy are then applied, performs the ratio ^µ S _xy / ^µ S _z , which represents the tangent of the angle a formed between the direction of the earth's magnetic field vector and that of of the probe axis. The information ^µ S _xy / ^µ S _z is then applied to the comparator COMP 1 which compares it to a predetermined value limit L ₁ . If the quantity

is positive or zero, the output of comparator COMP 1 is put in the state T, = 0 (general case), and, if u is negative, in the state T ₁ = 1 (particular case, the least frequent), T ₁ being for example defined by the explicit function T ₁ = 1-INT 2 ^{u- | u |} where "INT" denotes the "whole part of" function. Thus, for the generally appropriate value of 5.10 ^-2 for L1, the output T ₁ of the comparator COMP 1 will be deactivated if the angle

is greater than or equal to 3 ° (general case).

The state T, of the output of the comparator COMP 1 makes it possible to operate a switch, symbolically produced by two relays M T ₁ and MT ₁ . The MT relay closes its contacts when T ₁ = 1 ― T ₁ is equal to 1 and the MT ₁ relay closes its contacts when T ₁ is equal to 1. When T ₁ is zero, (general case), ie when T ₁ is equal to 1 (fig. 3a) the signal ^µ S of the magnetometer is used as the stabilizing signal ^a S and the signal ^γ S of the accelerometer as the signal to stabilize ^p S, which means that the signal from the magnetometer is used to correct the accelerometer signal from the probe rotation effects. Conversely when T ₁ is equal to 1 (special case), that is to say when T ₁ is zero, the stabilizing signal ^a S is the signal ^γ S of the accelerometer, which is used to correct the signal ^µ S of the magnetometer, constituting the signal to be stabilized ^p S. _

pour les deux valeurs de T₁.More concisely, the MV and MT relays define:

for the two values of T ₁ .

In the case T ₁ = 1 (special case), the components ^γ S _xo and ^γ S _yo coming from F ₁ .1 and F ₁ .2 are combined in N ₃ to obtain the diagonal transverse component

^PS_xo et ^PS_γo proviennent de F₁.1 et F₁.2 si T₁=0 (cas général) et de R₁.1 et R₁.2 si T₁=1 (cas particulier); aS_x et ^aS_y proviennent de R₁.1 et R₁.2 si T₁=0 (cas général) et de F₁.1 et F₁.2 si T₁=1 (cas particulier); et ^aS_xy provient de N₁ par R₁.4 lorsque T₁=0 (cas général), et de N₃ lorsque T₁=₁ (cas particulier). Les composantes ^PS_x et ^pS_y stabilisées sont sensiblement celles qui auraient été obtenues en absence de rotation de la sonde autour de son axe longitudinal. Les composantes ^pS_x et ^PS_y provenant des blocks D₁ ou D₂, la composante axiale longitudinale ^γS_z du signal de l'accéléromètre (définissant ^PS_z si T₁=0 et ^aS_z si T₁=1 et, si T₁=1 (cas particulier), la composante diagonale ^γS_xy=^aS_xy du signal stabilisateur, subissent ensuite, dans les blocs F₂.1 à F₂.7 un filtrage passe-bas dont la caractéristique est donnée par:

avec b_k=0,54-0,46 cos

The virtual stabilization step ET1 essentially consists in correcting the transverse axial components of the signal to be stabilized, by eliminating in these components the effects of the rotation of the probe, by means of the diagonal and axial transverse components of the stabilizing signal, in blocks D ₁ or D ₂ ; for components ^p S _xo , ^p S _yo , ^a S _x , ^a S _y , ^a S _xy input, D ₁ and D ₂ output the new components ^p S _x and ^P S _y such that:

^P S _xo and ^P S _γo come from F ₁ .1 and F ₁ .2 if T ₁ = 0 (general case) and from R ₁ .1 and R ₁ .2 if T ₁ = 1 (particular case); aS _x and ^a S _y come from R ₁ .1 and R ₁ .2 if T ₁ = 0 (general case) and from F ₁ .1 and F ₁ .2 if T ₁ = 1 (particular case); and ^a S _xy comes from N ₁ by R ₁ .4 when T ₁ = 0 (general case), and from N ₃ when T ₁ = ₁ (special case). The stabilized components ^P S _x and ^p S _y are substantially those which would have been obtained in the absence of rotation of the probe around its longitudinal axis. The components ^p S _x and ^P S _y coming from blocks D ₁ or D ₂ , the longitudinal axial component ^γ S _z of the accelerometer signal (defining ^P S _z if T ₁ = 0 and ^a S _z if T ₁ = 1 and, if T ₁ = 1 (special case), the diagonal component ^γ S _xy = ^a S _xy of the stabilizing signal, then undergo, in blocks F ₂ .1 to F ₂ .7 a low-pass filtering whose characteristic is given by:

with b _k = 0.54-0.46 cos

The characteristic of these filters F ₂ is represented in FIG. 5 which carries, on the abscissa, the frequency and on the ordinate, the transmitted amplitude, in the case where the value of each component to be filtered is sampled every 1/7, 5 seconds (At = 1 / 7.5 s). New filtered components therefore appear every 31.5At, or approximately every 4.2 seconds.

The role of the filters F ₂ is to eliminate, from the filtered components, the variations in the amplitude having a frequency greater than the maximum frequency of the amplitude variations which are attributable to the acceleration of gravity and which essentially derive from the variations the angle formed between the vertical and the longitudinal axis of the probe. We see in Figure 5 that frequencies above 8.10- ² Hz undergo attenuation above 3 dB and very rapidly increasing.

As the appearance of a filtered component S _{ξ, (31.5) lΔt} supposes the previous appearance of the unfiltered component S _{ξo, (31.5) (I + 1) Δt} the components at the output of the filters F2.1 to F2.7 undergo a delay of 31.5 Δt. To eliminate the effects of this delay, the unfiltered components undergo equivalent delays in the buffer cells R ₂ .1 to R2.9.

After the low-pass filtering, the components of the accelerometer signal are normalized. When T ₁ = 0 (general case), the components of ^γ S = ^p S are normalized in N ₂ , which provides the norm ^γ S _xyz = ^p S _xyz and the diagonal normalized components ^γ S _xy = ^p S _xy and axial ^γ S _x = ^p S _x , ^γ S _y = ^p S _y and ^γ S _z = ^p S ₂ . When T ₁ = 1 (special case), the components of ^γ S = ^a S are normalized in N ₄ , which provides the norm ^γ S _xyz = ^a S _xyz and the normalized longitudinal components ^γ S _z = ^a S _z and diagonal ^γ S _xy = ^a S _xy .

In addition, when T ₁ = 0 (general case), new transverse components ^γ S _x = ^p S _x and ^γ S _y = ^p S _y of the accelerometer signal are obtained in E ₁ , at the output of N ₂ , using the transverse components ^a S _x = ^µ S _x , ^a S _y = ^µ S _y and ^a S _xy = µ ^S _xy , of the locating signal from the magnetometer. This operation E ₁ constitutes the reverse operation of the operation D ₁ previously mentioned and has the effect of reintroducing, into the components of the accelerometer signal, the information relating to the rotation of the probe around its longitudinal axis.

If ^γ S _xo and ^γ S _yo are the components of ^γ S at the exit of N ₂ and ^µ S _x , ^µ S _y , ^µ S _xy the transverse components of ^µ S at the exit of R ₂ .1, R ₂ . 2 and R ₂ .4, the new components of ^γ S at the output of E ₁ are:

It should be noted here that these components ^γ S _x and ^γ S _y are not at all identical or proportional to the components of the accelerometer output signal. If, in fact, these new components ^γ S _x and ^γ S _y again contain the information relating to the rotation of the probe around its longitudinal axis relative to a reference position, on the other hand they are rid of disturbing information originating impacts of the probe against the wall of the well.

The final step ET2 of combining the components of the acceleration and locating signals results, by different operations described below, in the determination of different parameters representative of the topographic orientation of the well and of the position of the probe in the well relative to a reference position corresponding to a setting of the probe for the rotational movements around its longitudinal axis.

The diagonal transverse components ^γ S _xy and longitudinal ^γ S _z of the accelerometer signal, normalized in N ₂ or in N ₄ , are combined to obtain the value of a first parameter, DEV, representing the angle β formed between the vertical and the longitudinal axis of the probe.

If T ₁ = 0 (general case), the DEV parameter is obtained in DEV 1 which provides the information with the same name DEV 1, and if T ₁ = 1, DEV is obtained in DEV 2, providing the information DEV 2. The DEV 1 and DEV 2 function generators are identical and provide the information defined by

et

Autrement dit J(N,D) est égal à:

si D est négatif, et à

si D est positif, 2π étant ajoutés si

est négatif.In the case T ₁ = 0 (general case), the information DEV 1 is, in the comparator COMP 2, compared with an angle L2 of predetermined value, for example equal to 0.5 °; according to the result of this comparison, the value of two other pieces of information RB 1 and AZIM 1 is multiplied by 0 or 1, which will be defined later. This is, schematically, represented by the possibility, for the comparator COMP 2, of controlling two relays MT ₂ .1 and MT ₂ .2 closed or switched to ground. The comparator COMP 2 and the relays MT ₂ .1 and MT ₂ .2 are equivalent to a test "T ₂ = 0?" in which T ₂ is a function with value 1 if the angle v defined by v = DEV 1-L2 is positive or zero and with value zero if v is negative. The function T ₂ can for example take the explicit form: T _S = INT 2 ^{v- | vl} where INT designates the function "whole part of". To define the information RB1 and AZIM 1, previously mentioned, it is advantageous to define two functions, H and J, of two variables N and D, such as:

and

In other words J (N, D) is equal to:

if D is negative, and at

if D is positive, 2π being added if

is negative.

The two axial transverse components of the signal to be stabilized ^P S _x , ^p S _y , rid of the effects of rotation of the probe and filtered, coming from N ₂ when T ₁ = 0 (general case) and from F ₂ .6 and F ₂ .7 when T ₁ = 1, the normalized longitudinal component ^P S _z of this same signal, coming from N ₂ when T ₁ = 0 (general case) and from R ₂ .9 when T ₁ = 1, and the diagonal components and longitudinal ^a S _xy and ^a S _z of the stabilizing vector, coming from R ₂ .4 and R ₂ .3 when T ₁ = 0 (general case) and from N ₄ when T ₁ = 1, are combined to obtain the value of a second parameter, AZIM, representing the angle ξ formed between the horizontal trace of the vertical plane passing through the longitudinal axis of the probe and the horizontal projection of the earth's magnetic field vector.

• AZIM 1=J(N,D) avec
  
• Après le test "T₂=0 ?", l'information AZIM 1 devient AZIM 2 telle que AZIM 2=T₂.AZIM 1. Pour T₁=1, le bloc AZIM 3 réalise la fonction élaborant l'information AZIM 3, définie par: AZIM 3=J(N,D) avec
  
  Le paramètre AZIM est donc égal à AZIM 2 si T₁=0 (cas général) et à AZIM 3 si T₁=1.

For T ₁ = 0 (general case), the AZIM 1 block performs the function elaborating the information of the same name, AZIM 1 previously mentioned and defined by:

• AZIM 1 = J (N, D) with
  
• After the test "T ₂ = 0?", The AZIM 1 information becomes AZIM 2 such that AZIM 2 = T _2. AZIM 1. For T ₁ = 1, the AZIM 3 block performs the function producing the AZIM 3 information, defined by: AZIM 3 = J (N, D) with
  
  The AZIM parameter is therefore equal to AZIM 2 if T ₁ = 0 (general case) and to AZIM 3 if T ₁ = 1.

et

Lorsque T₁=1, la combinaison des six composantes axiales des signaux est réalisée par AZI1.2, de la même façon, c'est-à-dire avec les mêmes espressions pour N et D. Le paramètre AZI 1 est donc égal à AZI1.1 si T₁=0 et à AZI1.2 si T₁=1.The three axial components ^γ S _x , ^γ S _y and ^γ S _z of the accelerometer signal; containing the effects of rotation of the probe, that is to say coming, when T ₁ = O (general case) from F ₁ as far as ^γ S _x and ^γ S _y and from N ₂ for ^γ S _z , and, when T ₁ = 1, of R ₂ .5 and R ₂ .6 for ^γ S _x and ^γ S _y , and de N ₄ for ^γ S _z , and the three axial components ^µ S _x , ^µ S _y and ^µ S _z of the magnetometer signal, also containing the effects of rotation of the probe, i.e. coming from, when T ₁ = 0 (general case), from R ₂ .1, R ₂ .2 and R ₂ .3 and, when T ₁ = 1, of R ₂ .7, R ₂ : 8 and R ₂ .9, are combined to obtain the value of a third parameter, AZI 1, representing the angle δ formed between the horizontal projection of the earth's magnetic field vector and the horizontal projection of a vector perpendicular to the longitudinal axis of the probe and joining this axis to a fixed point P of the probe, distant from this same axis. This combination is made, when T ₁ = 0 (general case) by AZI1.1 which provides the information AZI1.1 such that AZI1.1 = J (N, D) with

and

When T ₁ = 1, the combination of the six axial components of the signals is carried out by AZI1.2, in the same way, that is to say with the same expressions for N and D. The parameter AZI 1 is therefore equal to AZI1.1 if T ₁ = 0 and AZI1.2 if T ₁ = 1.

The two transverse axial components ^γ S _x and ^γ S _y of the accelerometer signal, containing the rotation effects of the. probe, that is to say coming from E1 when T ₁ = 0 (general case) and from R ₂ .5 and R ₂ .6 when T ₁ = 1, are combined respectively in RB1 and RB3 to obtain the value d 'a fourth parameter, RB, representing the maximum angle θ, or dihedral angle, formed between a vertical plane containing the longitudinal axis of the probe and a plane containing the axis of the probe and passing through the fixed point P of that -this. The information RB1 and RB3 are expressed by the same combination of components, namely J (N, D) with N = ^γ S _y and D = - ^γ S _x . After the test "T ₂ = 0?", The information RB1 becomes RB2 such that RB2 = T ₂ .RB 1. The parameter RB is therefore equal to RB ₂ if T ₁ = 0 and to RB3 if T ₁ = 1.

In FIG. 3b, the double contact relay T ₁ T ₁ , controlled by the comparator COMP 1, schematically represents the connection of the phase for determining the value of the parameters to a display operation AFF of these parameters. Thus this relay T, T, makes it possible to obtain, at the end of the determination phase, the parameters DEV, AZIM, AZI1 and RB which, in an explicit form, are expressed by:

It is however possible and possibly advantageous, to determine during the final step ET2, the value of other parameters such as Sin i, i being the angle of inclination of the earth's magnetic field vector. This possibility is illustrated in FIG. 3b (case T ₁ = 1). The Sin i parameter is given by:

On the other hand, the display of quantities such as the norm ^µ S _xyz of the signal of the magnetometer, and the norm ^γ S _xyz of the signal of the accelerometer, after low-pass filtering, makes it possible to exercise control over the meaning actual values obtained for the different parameters.

As mentioned above, the value of L1 should be chosen to be quite low. preferably less than or equal to 5.10 ^-2 (5.10 ^-2 = tg 3 °). Indeed, the signal ^γ S of the accelerometer being very disturbed by the accelerations undergone by the probe due to its movement, it is advantageous to limit as much as possible the use of the signal S of the accelerometer as a stabilizing signal ^a S to get rid of the magnetometer signal from the effects of rotation of the probe, therefore to limit the cases T ₁ = 1 as much as possible.

Although, in the method of the invention, the phase of determining the value of the parameters can, using the preceding indications, be carried out according to various methods, and for example by means of a hardware device specially designed for To this end and corresponding to the diagram of FIGS. 3a and 3b, it appeared that the most suitable way consisted in resorting to automatic data processing by means of a computer. In such a component, the blocks in FIGS. 2, 3a and 3b represent subroutines, with the exception of the comparators in FIG. 3a which represent tests, and relays in FIGS. 3a and 3b, which represent conditional connections.

Claims (6)

1. A method of determining at least two direction parameters (DEV, AZIM) of a well as a function of depth, the method comprising the steps consisting in; producing a three-component acceleration signal (y_s) representing a set of accelerations undergone by a sonde which is moved in the well, the accelerations being detected along three reference axes referred to the sonde; producing a three-component laying signal (µ_s) representing a vector of fixed direction which is different from the vertical and which is referred to the said three reference axes; and determining the said direction parameters by combining components of the said signals; the method being characterized in that the steps consisting in moving the sonde and in producing the said acceleration and laying signals are substantially simultaneous and continuous, and in that the said step of determining the direction parameters includes a virtual stabilization stage (ET1) in which the effects of sonde displacement are eliminated from the components of one of the said signals, constituting a signal to be stabilized (P_s), by means of the components of the other signal, constituting a stabilizing signal (as), and an intermediate low-pass filtering operation (OIF) applied to at least one of the stabilized components of the said signal to be stabilized, thereby eliminating from these components variations of frequency greater than the maximum frequency of variations attributable to acceleration due to gravity.

2. A method according to claim 1, characterized in that the said step of determining the direction parameters further comprises a stage (ETO) preliminary to the said virtual stabilization stage and including a prefiltering operation on the acceleration signal components (y_s) thereby substantially attenuating in these components signal variations at a frequency higher than the highest possible frequency of sonde rotation movement about its longitudinal axis.

3. A method according to claim 1 or claim 2, charaterized in that since the acceleration (y_s) and laying (µ_s) signals comprise two transverse axial components and one longitudinal component each in a system of axes referred to the longitudinal axis of the sonde, a preliminary stage (ETO) is provided before the virtual stabilization stage (ET1) in which a transverse diagonal component of the stabilizing signal (as) is determined from the transverse axial components of this signal, and in that the said rotation effects are eliminated from the transverse axial components of the signal to be stabilized (p_s) by means of the transverse axial and diagonal components of the stabilizing signal, in order to obtain rotation-stabilized components corresponding to a reference position of the sonde about its longitudinal axis.

4. A method according to claim 3, characterized in that the preliminary stage (ETO) comprises the operations consisting in: determining a transverse diagonal component of the laying signal (µ_s) from the transverse axial components of this signal; determining from this transverse diagonal component and from the longitudinal axial component of this same laying signal the sign of the difference between a first angle formed between the said vector of fixed direction and the longitudinal axis of the sonde, and a limit angle of predetermined value; defining the stabilizing signals (as) and the signals to be stabilized (p_s) respectively as the laying (µ_s) and the acceleration (γ_s) signals when the sign of the said difference is positive, and as the acceleration and the laying signals when this sign is negative; and determining a transverse diagonal component of the stabilizing signal from its transverse axial components when the stabilizing signal is defined as the said acceleration signal.

5. A method according to claim 4, characterized in that when the sign of the difference determined during the said preliminary stage (ETO) is positive, an operation is provided, after the intermediate filtering, to re-introduce the effects of sonde displacement, thereby supplying on the basis of the two stabilized transverse axial components of the acceleration signal (y_s) and the diagonal and axial transverse components of the laying signal (µ_s), two transverse axial components of the acceleration signal which are again not stabilized relative to the said reference position of the sonde.

6. Apparatus for determining at least two direction parameters (DEV, AZIM) of a well, the apparatus comprising:

a sonde (2)

a winch (4) and a cable (3) for moving the sonde in the well (1);

bows (7) for centring the sonde in the well;

an accelerometer (8) included in the sonde to produce first signals (y_s) representative of the acceleration vectors to which the sonde is subjected when the sonde is moved in the well;

a magnetometer (9) or a gyroscope, included in the sonde, for producing second signals (µ_s) representative of a vector of fixed direction different from the vertical; and

means for processing and combining the said first and second signals to determine the said direction parameters;

characterized in that the said means for processing and combining include means (F₁, R₁, D₁, D₂, R₂) for removing the effects of sonde movement from the first signals by combining components of the first and second signals, and low-pass filter means (F₂) for eliminating from the components of the first signals as thus processed variations of frequency greater than the maximum frequency of variation attributable to acceleration due to gravity.

Images