METHOD FOR DETERMINING THE ACOUSTIC CHARACTERISTICS OF A MUD FILTER CAKE

Abstract

At least one acoustic sensor registers a pressure response to a low-frequency non-oscillating pressure signal generated in a borehole by at least one source. At least one characteristic of a transient process of pressure change is determined from the registered signal. A thickness of a mudcake is determined. Then, based on the values received, a mudcake piezoconductivity or a fluid mobility or both are determined.

Description

FIELD OF THE DISCLOSURE

The present invention relates to methods for acoustic characteristic determination of a mudcake created during well drilling, namely, fluid mobility and cake piezoconductivity.

BACKGROUND OF THE DISCLOSURE

A mudcake is created during drilling with a drilling mud, fed into a borehole through a drilling string and removed through holes in a drill bit for lubrication of the drill bit during drilling and removal of drilled rocks to the surface. A layer of mudcake is formed as the fluid mixes with cuttings and/or other solids and circulates up through an annular space between an outer side of the drill string and the borehole wall. The mixture covers the borehole wall and forms a layer of mudcake. Isolation of a formation from an internal part of the borehole is one of the cake functions. The cake layer is frequently called as a mud cake or a filter cake.

Direct determination method of filter cake characteristics by sampling while drilling is known, as described in the WO 2009/139992. In this method a low-frequency acoustic sensor in a listening mode is used to estimate a mudcake cake pressure diffusivity (piezoconductivity) κ, which is directly associated with sealing characteristics of the mudcake. A piston of a pretest chamber, or of any other device, was used as a device to create harmonic or periodic pressure oscillations. However generating of pressure oscillations is not always possible in practice.

SUMMARY OF THE DISCLOSURE

The invention provides for creation of a simple and effective method to determine mudcake characteristics in a borehole, which allows to determine mudcake parameters by using a pressure source, generating a non-oscillating signal (e.g., a single step-wise pressure pulse).

The method comprises registering by at least one acoustic sensor a pressure response to a low-frequency non-oscillating pressure signal generated in a borehole by at least one source. At least one characteristic of a transient process of pressure change is determined from the registered signal. A thickness of a mudcake is determined. Then based on the values received a mudcake piezoconductivity or a fluid mobility or both are determined.

Characteristics of the transient process are an exponent of a transient component of a solution, a moment of time when the transient component of the solution reaches its maximum, and a value of maximum pressure attained during the transient process.

Either natural or artificial sources can be used as the at least one source of the non-oscillating pressure signals.

Low-frequency acoustic sensors/sources/transducers, a low-frequency well pressure modulation, etc., can be used as the artificial sources.

Hydrophones, transducers, accelerometers, pressure transmitters, etc., can be used as the acoustic sensors to register the pressure response.

The source generating the low-frequency pressure signals at the same time may be the acoustic sensor.

The source generating the low-frequency pressure signals or the acoustic sensor of the low-frequency pressure signals or both can be installed on a packer.

The source generating the low-frequency pressure signals or the acoustic sensor of the low-frequency pressure signals or both can be installed on a sampling probe.

The source generating the low-frequency pressure signals or the acoustic sensor of the low-frequency pressure signals or both can be installed on a backup shoe.

Several sources generating the low-frequency pressure signals and mounted in several places can be used.

The thickness of the mudcake is determined by echo-impulse measurements comprising a supply of short high-frequency (HF) signals to the formation, and registration of arrival time of reflected echo-signals.

During mudcake thickness determination it is preferable to supply the HF signals from at least two positions at different distances from the mudcake.

BRIEF DESCRIPTION OF THE FIGURE

The invention is explained by a drawing where FIG. 1 shows a ratio of pressure at a side of a mudcake where a sensor is installed, to a pressure amplitude at its other side for different permeability values.

DETAILED DESCRIPTION

To obtain parameters of a formation and parameters of a mudcake, propagation of a pressure pulse through the formation and the mudcake can be decoupled. Given that a dispersion pressure wavelength in the mudcake, λ_mc is much less than a wavelength in the formation, α_for and that a thickness of the mudcake, h_mc is much less than a radius, R_b of the borehole, description of pressure signal propagation through the mudcake may be reduced to a simple one-dimensional problem.

$\begin{array}{cc}\frac{\partial P}{\partial t}-\frac{\partial }{\partial x}\left(\kappa \left(x\right)\frac{\partial P}{\partial x}\right)=0& \left(1\right)\end{array}$

where κ is piezoconductivity (pressure diffusivity), P is pressure, x is a linear co-ordinate, which is perpendicular to a mudcake surface, k is permeability, with boundary conditions of

$\begin{array}{cc}P{}_{x=0}=\varphi \left(t\right)\frac{\partial P}{\partial x}{}_{x=h_{\mathrm{mc}}}=0& \left(2\right)\end{array}$

The solution of the (1)-(2) problem is the following:

$\begin{array}{cc}P\left(x,t\right)=-\frac{2}{h_{\mathrm{mc}}}\sum _{n=0}^{\infty }\exp \left(-\kappa A_n^2t\right)\sin \left(A_nx\right)\kappa A_n{\int }_0^t\exp \left(\kappa A_n^2\lambda \right)\varphi \left(\lambda \right)\lambda & \left(3\right)\end{array}$

It means that in case of a source producing non-oscillating pressure signal, a sensor response will contain only a transient process. For example, let us consider step-wise source function:

φ(t)=η(t)−η(t−τ₀)  (4)

By simple transformations of the (3), (4) solution we can get the following expression:

$\begin{array}{cc}\mathrm{for}t\le {\tau }_0P\left(x,t\right)=\frac{-2}{h_{\mathrm{mc}}}\sum _{n=0}^{\infty }A_n^{-1}\left[1-\exp \left(-\kappa A_n^2t\right)\right]\sin \left(A_nx\right)\mathrm{and}& \left(5\right)\\ \mathrm{for}t>{\tau }_0P\left(x,t\right)=-\frac{2}{h_{\mathrm{mc}}}\sum _{n=0}^{\infty }A_n^{-1}\exp \left[\kappa A_n^2\left({\tau }_0-t\right)\right]\left\{1-\exp \left(-\kappa A_n^2{\tau }_0\right)\right\}\sin \left(A_nx\right)& \left(6\right)\end{array}$

One can see that the given solution contains only transient process. We can expect such a result since there are no sources to drive induced oscillations. The situation is shown as an example on FIG. 1, where P/P₀(t) curves for models with different permeability values are plotted. In this case a step-wise initial pressure pulse of 10 seconds was used as a source. The transient process, its maximum and further decay are sufficiently pronounced and can be used for evaluation of the mudcake permeability. This possibility results from analysis of the initial pressure growth as well as from the long-term pressure reduction. Both processes can be analyzed using the (5), (6) formulae.

In general, for any non-oscillating pressure pulse a pressure response will contain only a transient process. This process has several characteristic features which can be used to evaluate the mudcake piezoconductivity κ:

1) An exponent of a transient component of a solution;

2) A moment of time τ_max , when the transient component of the solution reaches its maximum;

3) A value of maximal pressure attained during the transient process.

Extraction of the characteristics from the sensor response (a transient process) is not too difficult. Estimation of τ_max and the maximal pressure is a simple task. To extract the above mentioned values from the signal registered by the sensor we propose to use ideas of signal treatment and filtering and phase locked loops to separate the transient and oscillating processes. This can be explained by the fact that forced vibration frequency is known (a source frequency), and a spectral content of the solution transient component is concentrated on a much lower frequency. Therefore, one can use a low frequency filter for extraction of the solution transient component. Then, it can be very easy to evaluate τ_max . Selecting a decaying part of the transient component (at t>τ_max) and taking its logarithm can help to find the time interval when the curve slope becomes constant. This indicates that the phase characterized by the presence of only one exponent remained, is already reached. The initial stage of this process is registered by the sensor and can be analyzed, using the (5), (6) solution formulae. Knowing these values and using formulae for their relationship with κ, one can easily evaluate its meaning (e.g., by simple iterations or using a conventional solver for finding roots of functions).

Using accelerometers as the sensors allows to cover wide and especially a high frequency part of low frequency spectrum (1 Hz to tens kHz); standalone pressure sensors ensure measurements of a pressure signal and can be used even when the direct contact with mudcake/formation due to some reason is undesirable or impossible, or in such places as a probe inlet, etc.

It is possible to use one or several sources and one or several acoustic sensors. It should be mentioned that often one device can operate both as a source and as a sensor, ant these states can be either combined or switched. Further, there is a flexibility as to where these sources and/or sensors are placed. Examples include but not limited to:

• • a tool packer;
  • a pad of a sampling probe;
  • a backup shoe;
  • source(s)/receiver(s) mounted standalone;
  • etc.

Wide variety of options is important and provides numerous advantages. For example, placing source(s)/sensor(s) on the packer will ensure a good contact with the mudcake; placing them on a pad of the sampling probe will ensure the response measurement in vicinity of the probe inlet thus avoiding significant pressure signal attenuation, etc.; placing them on a backup shoe can compensate noise and ensures accurate measurement of the signal component associated with pressure diffusion through the mudcake; standalone mounting ensures flexibility during measurements and designing; etc.

Low frequency measurements can be significantly improved by use of several sensors. They can be installed in different places: a pad of a sampling probe, a backup shoe, etc. This can ensure the noise reduction or removal, as well as possibility of the differential pressure measurements. This can increase ratio signal-interference, reduce requirements for dynamic range and sensitivity, facilitate reduction of the possible effect of the measurement geometry, etc.

The mudcake thickness, h_mc, is preliminary determined based on the echo-pulse measurements including short high frequency signals supply to the formation and registration of the echo-signal time of arrival (see WO 2009/139992). During the mudcake thickness determination it is favourable to supply HF signals from at least two positions at different distances from the cake.

A fluid mobility, η, in the mudcake is determined as

η=κφ/K

The cake porosity, φ, is estimated as 10-30%, K is a volume Young's modulus of the porous medium.

Claims

1. A method for determining mudcake acoustic characteristics in a borehole, the method comprising: registering by at least one acoustic sensor a response to a low-frequency non-oscillating pressure signal generated in a borehole by at least one source;determining from the registered signal at least one characteristic of a pressure change transient process;determining a thickness of a mudcake; anddetermining based on values obtained a piezoconductivity of the mudcake, a fluid mobility, or both.

2. The method of claim 1, wherein the at least one characteristic of the pressure change transient process is an exponent of a transient component of a solution, a moment of time when the transient component of the solution reaches its maximum, and a value of maximal pressure attained during the transient process.

3. The method of claim 1, wherein a natural source is used as the source of the low-frequency non-oscillating pressure signal.

4. The method of claim 2, wherein the at least one acoustic sensor is installed on a packer.

5. The method of claim 2, wherein the at least one acoustic sensor is installed on a sampling probe.

6. The method of claim 2, wherein the at least one acoustic sensor is installed on a backup shoe.

7. The method of claim 1, wherein at least one artificial source is used as the source of the low-frequency non-oscillating pressure signal.

8. The method of claim 7, wherein the source of the low-frequency non-oscillating signal is at the same time the acoustic sensor.

9. The method of claim 7, wherein the low-frequency non-oscillating pressure signal is excited by a low-frequency modulation of the well pressure.

10. The method of claim 1, wherein vibrometers are used as the acoustic sensors to register the pressure response.

11. The method of claim 1, wherein accelerometers are used as the acoustic sensors to register the pressure response.

12. The method of claim 1, wherein transducers are used as the acoustic sensors to register the pressure response.

13. The method of claim 1, wherein pressure transmitters are used as the acoustic sensors to register the pressure response.

14. The method of claim 7, wherein the source of the low-frequency non-oscillating signal or the acoustic sensor or both are installed on a packer.

15. The method of claim 7, wherein the source of the low-frequency non-oscillating signal or the acoustic sensor or both are installed on a sampling probe.

16. The method of claim 7, wherein the source of the low-frequency non-oscillating signal or the acoustic sensor or both are installed on a backup shoe.

17. The method of claim 7, wherein several sources of the low-frequency non-oscillating signal are installed at different places.

18. The method of claim 1, wherein a thickness of a mudcake is determined based on echo-impulse measurements comprising a supply of short high-frequency signals to a formation and registration of the reflected echo-signals arrival time.

19. The method of 18, wherein the short high-frequency signals are supplied at least from two positions, located at different distances from the mudcake.